HMG Advance Access published June 10, 2015
1 Fingolimod (FTY720) enhances hippocampal synaptic plasticity and memory in Huntington’s disease by preventing p75NTR up-regulation and astrocyte-mediated inflammation
Andrés Miguez1,2,3, Gerardo García-Díaz Barriga1,2,3, Verónica Brito1,2,3, Marco Straccia1,2,3, Albert Giralt1,2,3, Silvia Ginés1,2,3, Josep M. Canals1,2,3, Jordi Alberch1,2,3,* 1
Departament de Biologia Cel·lular, Immunologia i Neurociències, Facultat de Medicina,
2
Institut d'Investigacions Biomèdiques August Pi i Sunyer (IDIBAPS), 08036 Barcelona, Spain
3
Centro de Investigación Biomédica en Red sobre Enfermedades Neurodegenerativas
(CIBERNED), Spain
*
To whom correspondence should be addressed at: Departament de Biologia Cel·lular,
Immunologia i Neurociències, Facultat de Medicina, Universitat de Barcelona, Casanova 143, 08036 Barcelona, Spain; Tel: +34 934035285, Fax: +34 934021907; E-mail: [email protected]
© The Author 2015. Published by Oxford University Press. All rights reserved. For Permissions, please email: [email protected]
Downloaded from http://hmg.oxfordjournals.org/ at CSU Fresno on July 22, 2015
Universitat de Barcelona, 08036 Barcelona, Spain
2
ABSTRACT Huntington’s disease (HD) is a hereditary neurodegenerative disorder characterized by motor and cognitive impairments, involving striatum, cortex and hippocampus. Synaptic and memory dysfunction in HD mouse models have been related to low levels of brain-derived neurotrophic factor (BDNF) and imbalance between TrkB and p75NTR receptors. In addition, astrocyte overactivation has also been suggested to contribute to HD cognitive deficits. Fingolimod (FTY720), a modulator of sphingosine-1 phosphate (S1P) receptors, has been shown to increase
inflammatory response. In this view, we have investigated whether FTY720 improves synaptic plasticity and memory in the R6/1 mouse model of HD, through regulation of BDNF signaling and astroglial reactivity. Chronic administration of FTY720 from pre-symptomatic stages ameliorated long-term memory deficits and dendritic spine loss in CA1 hippocampal neurons from R6/1 mice. Furthermore, FTY720 delivery prevented astrogliosis and over-activation of nuclear factor kappa beta (NF-κB) signaling in the R6/1 hippocampus, reducing tumor necrosis factor alpha (TNFα) and induced nitric oxide synthase (iNOS) levels. TNFα decrease correlated with the normalization of p75NTR expression in the hippocampus of FTY720-treated R6/1 mice, thus preventing p75NTR/TrkB imbalance. In addition, FTY720 increased cAMP levels and promoted phosphorylation of CREB and RhoA in the hippocampus of R6/1 mice, further supporting its role in the enhancement of synaptic plasticity. Our findings provide new insights into the mechanism of action of FTY720 and reveal a novel therapeutic strategy to treat memory deficits in HD.
Downloaded from http://hmg.oxfordjournals.org/ at CSU Fresno on July 22, 2015
BDNF levels and to reduce astrogliosis, proving its potential to regulate trophic support and
3
INTRODUCTION Huntington’s disease (HD) is an autosomal dominant neurodegenerative disorder caused by the expansion of a CAG tract in the exon-1 of the huntingtin gene (1). Clinical diagnosis of HD relies on the manifestation of motor abnormalities. However, cognitive and behavioral impairments are evident at least 15 years before motor diagnosis (2–4), and are currently considered the greatest burden on HD families (5). The primary regions of neurodegeneration in HD have long been considered the striatum and cerebral cortex (6, 7), but other structures involved in cognition, particularly the hippocampus, are affected in early stages of the disease
hippocampal dysfunction (10–12). Neurotrophins promote the development and survival of neurons. Among them, brainderived neurotrophic factor (BDNF) stands out for its ability to regulate synaptic plasticity and cognitive function (13). BDNF is initially synthesized as a precursor protein called pro-BDNF, which is then cleaved by proteases to generate the mature form (14). Reduction in BDNF levels affects the onset and severity of HD in mouse models (15, 16). Despite its therapeutic potential, BDNF delivery is hampered by its poor pharmacokinetic properties and its inability to cross the blood-brain barrier (17). TrkB and p75NTR receptors can bind to both BDNF and pro-BDNF (18, 19). TrkB activation promotes long-term potentiation, whereas p75NTR is regarded as a negative modulator of synaptic plasticity (20, 21). Imbalance of TrkB/p75NTR expression has been reported in the striatum and hippocampus of HD patients and mice, affecting cell survival and synaptic plasticity (12, 22). Therefore, current research is aimed at developing strategies to promote the release of endogenous BDNF and to modulate TrkB and p75NTR receptors (12, 23, 24). Fingolimod (FTY720), a compound used as an immunomodulator in Multiple Sclerosis patients (Brinkmann et al., 2010), has recently been shown to improve BDNF release in the striatum and to rescue motor deficits in mouse models of Rett’s Syndrome and HD (25, 26). FTY720 is a structural analogue of sphingosine that can cross the blood-brain barrier and reach
Downloaded from http://hmg.oxfordjournals.org/ at CSU Fresno on July 22, 2015
(8, 9). Notably, deficits in spatial and recognition memory have been recently associated to
4 the brain, where it is converted to FTY720-phosphate (FTY720-P). FTY720-P initially acts as an agonist of sphingosine-1 phosphate (S1P) receptors, but chronic exposure to the drug leads to receptor internalization and functional antagonism (27). Most of FTY720 therapeutic effects within the brain are mediated by S1P1 receptor, which is highly expressed in astrocytes (28) and to a lesser extent in neurons (25), oligodendrocytes (29) and microglia (30). Conditional deletion of S1P1 receptor from astrocytes mimics the effects of FTY720 in experimental autoimmune encephalomyelitis (EAE), suggesting a neuroprotective effect through astrocytes modulation (28). Indeed, FTY720 treatment reduced astrogliosis, nuclear factor kappa beta (NFκB) activity and nitric oxide production in EAE mice, further supporting the relevance of S1P
HD (32–35) and associated with increased NF-κB activity (36). Furthermore, expression of mutant huntingtin (mHtt) in astrocytes is sufficient to induce neurological symptoms characteristic of HD mice (37, 38), strongly suggesting a critical role of astroglial dysfunction in HD pathogenesis. Besides its capacity to modulate astrocyte reactivity, FTY720 was found to reduce the severity of dendritic spine loss in striatal neurons of EAE mice (39) and to improve cognitive function in rodent models of Alzheimer’s disease (40, 41), suggesting additional effects on structural and functional synaptic plasticity. However, the mechanism through which this compound may enhance hippocampal synaptic plasticity remains unclear, and no studies to date have examined its ability to counteract memory deficits in HD. The present work was designed to investigate the potential of FTY720 to improve synaptic and memory function in R6/1 mice through regulation of hippocampal BDNF signaling and astrocyte activation.
RESULTS Chronic administration of FTY720 ameliorates long-term memory deficits in R6/1 mice Since the most prominent effects of FTY720 within the brain are related to its action on S1P1 receptor (27), we first examined by Western blot (WB) whether S1P1 expression was altered in the hippocampus of R6/1 mice. At 20 weeks of age, S1P1 protein levels were similar in wild-
Downloaded from http://hmg.oxfordjournals.org/ at CSU Fresno on July 22, 2015
antagonism in astroglia for drug efficacy (31). Interestingly, astrogliosis has been reported in
5 type (WT) and R6/1 mice (Supplementary Material, Fig. S1), suggesting a similar sensitivity to the drug in both genotypes. To analyze whether FTY720 improves long-term memory deficits in HD, either FTY720 or vehicle solution were delivered intraperitoneally in R6/1 and WT mice starting at pre-symptomatic stages (8 weeks) and continuing until 20 weeks of age, when R6/1 mice exhibit an overt motor and cognitive phenotype. Recognition memory was evaluated at 17 weeks of age by performing the novel object recognition test (NORT). We found that, whereas WT mice spent more time exploring the new object than the old one, this preference was lost in R6/1 mice. Notably, chronically treated R6/1 mice showed a similar preference for the new object than WT mice (Fig. 1A). To further confirm this cognitive improvement, we next
During the test, vehicle-injected R6/1 mice did not show any preference for the new arm respect to the old one, whereas FTY720-treated R6/1 mice explored more the new arm (Fig. 1B). No differences were detected between vehicle- and FTY720-treated WT mice. Therefore, FTY720 administration ameliorates spatial and recognition memory deficits in R6/1 mice.
FTY720 chronic treatment prevents dendritic spine loss and PSD-95 down-regulation in the hippocampus of R6/1 mice Learning and memory impairments are often associated with dendritic spine pathology and most HD animal models show lower spine density (12, 42). To analyze whether the memory improvement observed after chronic delivery of FTY720 correlated with structural synaptic changes, we quantified the number of dendritic spines in hippocampal neurons of R6/1 mice at 20 weeks of age. Following Golgi-Cox staining, the number of second order apical dendrites was counted in pyramidal neurons of the CA1 area (Fig. 2A). Spine density was decreased by 17% in R6/1 mice, compared to WT animals. FTY720 treatment prevented spine loss in the R6/1 hippocampus (Fig. 2A and B), while it had no effect in WT mice. Maintenance and stabilization of dendritic spines can be regulated by scaffolding protein post-synaptic density-95 (PSD-95), which is known to be reduced in HD patients and animal models (43–45). Therefore, we examined the expression of PSD-95 by WB in the hippocampus at 20 weeks of age. R6/1
Downloaded from http://hmg.oxfordjournals.org/ at CSU Fresno on July 22, 2015
performed the T-maze spontaneous alternation task (T-SAT), which measures spatial memory.
6 mice showed a decrease in PSD-95 protein levels that was prevented by FTY720 treatment (Fig. 2C), indicating a role for FTY720 in modulating structural synaptic plasticity.
FTY720 attenuates astrocyte-mediated inflammation in the hippocampus of R6/1 mice Astrocyte over-activation has been reported in HD patients (34) and mouse models, where it has been suggested to contribute to poor cognitive function (36). We thus examined if there was any evidence of astrogliosis in the hippocampus of R6/1 mice at 20 weeks of age. Increased astrocyte infiltration and activation was detected in the CA1 region of the R6/1 hippocampus (Fig. 3A), with no apparent changes in astrocyte proliferation. Within the CA1 area, astrocytes
(Fig. 3D). We found no significant differences in the size and density of Iba1+ cells between WT and R6/1 mice, suggesting no obvious microgliosis (data not shown). Remarkably, chronic administration of FTY720 reduced the number, size and GFAP intensity of astrocytes within the CA1 region of R6/1 mice, proving the ability of the drug to prevent astrogliosis (Fig. 3A-D). Reactive astrocytes have been directly associated with aberrant NF-κB activity in R6/2 mice (36). In order to learn whether NF-κB activity was altered in the hippocampus of R6/1 mice, we performed a WB analysis for NF-κB and inhibitor of kappa beta α (IкBα), an inhibitory protein that sequesters NF-κB in the cytoplasm (46). Vehicle-injected WT and R6/1 mice showed similar NF-κB protein levels (Supplementary Material, Fig. S2B), but IкBα expression was strongly reduced in the R6/1 hippocampus (Fig. 3E), suggesting increased NFκB translocation to the nucleus. Immunohistochemical analysis confirmed the presence of reactive astrocytes surrounding the CA1 area displaying increased nuclear localization of NFκB p65 (Supplementary Material, Fig. S2A), indicating an over-activation of NF-кB signaling in R6/1 mice. Interestingly, FTY720 chronic treatment prevented IкBα down-regulation (Fig. 3E) and subsequent NF-κB p65 translocation to the nucleus in astroglial cells (Supplementary Material, Fig. S2A), revealing a role for FTY720 in the reduction of hippocampal NF-κB activity. Since NF-κB regulates the expression of a large number of genes associated with neuroinflammation (47), we next examined whether FTY720-induced normalization of NF-κB
Downloaded from http://hmg.oxfordjournals.org/ at CSU Fresno on July 22, 2015
displayed a significant increase in number (Fig. 3B), size (Fig. 3C) and GFAP immunoreactivity
7 signaling correlated with decreased levels of pro-inflammatory mediators in the R6/1 hippocampus. As shown by quantitative real-time PCR (qRT-PCR) analysis, FTY720-treated R6/1 mice displayed decreased mRNA levels of tumor necrosis factor alpha (TNFα) (Fig. 3F) and inducible nitric oxide synthase (iNOS) (Fig. 3G), with no significant changes in IL-1β and IL-6 expression (data not shown). Altogether, these data support a role for FTY720 in diminishing astrocyte-mediated inflammation in the R6/1 hippocampus.
FTY720 increases BDNF mRNA in the hippocampus of R6/1 mice The effects of pro-inflammatory factors on synaptic plasticity and memory have been associated
been shown to increase BDNF levels in the hippocampus of WT mice (25, 51). To find out whether FTY720 chronic treatment was able to improve BDNF synthesis in the hippocampus of R6/1 mice, we performed qRT-PCR and WB analysis at 20 weeks of age to assess mRNA and protein levels, respectively. R6/1 mice displayed a decrease in BDNF mRNA, which was prevented by FTY720 delivery (Fig. 4A). However, we found no differences in BDNF protein, neither in WT nor in R6/1 mice (Fig. 4B).
FTY720 chronic treatment prevents p75NTR/TrkB imbalance in the hippocampus of R6/1 mice Imbalance of BDNF receptors, p75NTR and TrkB, has been reported in the hippocampus of HD patients and mice and associated to synaptic and memory deficits (12). Immunohistochemical and WB analysis of the R6/1 hippocampus at 20 weeks of age revealed abnormally high levels of p75NTR (Fig. 5A and B), which correlated with decreased TrkB expression in R6/1 mice (Fig. 5C). Chronic delivery of FTY720 was sufficient to prevent up-regulation of p75NTR expression in the hippocampus of R6/1 mice (Fig. 5A and B). Moreover, TrkB protein levels were not significantly different between WT and R6/1 mice after treatment (Fig. 5C). Interestingly, FTY720 promoted activation of TrkB in the R6/1 hippocampus, as showed by the increase in TrkB phosphorylation (Fig. 5D).
Downloaded from http://hmg.oxfordjournals.org/ at CSU Fresno on July 22, 2015
with altered neurotrophic support, particularly with BDNF (48–50). Moreover, FTY720 has
8 p75NTR expression can be regulated by TNFα in hippocampal cells (52). To determine the potential contribution of TNFα to the up-regulation of p75NTR in the hippocampus of R6/1 mice, we purified primary hippocampal neurons from WT and R6/1 animals. After 21 DIV, we examined p75NTR expression under basal conditions and following stimulation with TNFα (10 ng/ml), in the presence and in the absence of FTY720-P (10 nM). p75NTR protein levels were similar in WT and R6/1 primary hippocampal neurons under basal conditions (Fig. 6). Nevertheless, incubation with TNFα significantly increased p75NTR expression in R6/1 hippocampal neurons, whereas it had no effect on WT neurons. Remarkably, addition of FTY720-P to R6/1 hippocampal cultures was sufficient to counteract the effect of TNFα on
robust effect of FTY720 on the reduction of p75NTR expression in hippocampal neurons, likely acting through inhibition of TNFα.
FTY720 chronic treatment enhances cAMP signaling in the hippocampus of R6/1 mice Activation of cyclic AMP (cAMP) signaling improves synaptic plasticity and hippocampusdependent long-term memory (53). Notably, S1P inhibits cAMP pathway (54), whereas FTY720-P has been suggested to induce sustained cAMP signaling in vitro (55). In order to determine whether FTY720 treatment modulated the levels of cAMP and its downstream targets in a HD mouse model, we examined cAMP expression in the hippocampal CA1 region by immunohistochemical analysis. Both WT and R6/1 mice showed an evident increase in cAMP immunoreactivity after FTY720 chronic administration (Fig. 7A and B). To validate the activation of cAMP downstream signaling, we next examined two well-known targets whose phosphorylation is mainly mediated by protein kinase A (PKA). First, levels of the cAMPresponse element binding (CREB) and its phosphorylated form (pCREB) were assessed in the hippocampus of both mouse genotypes. FTY720 treatment increased pCREB protein in R6/1 mice, albeit not in WT mice, without significant changes in total CREB levels (Fig. 7C). Next, we determined the protein levels of RhoA, a negative modulator of spine density and complexity of hippocampal neurons (56) whose activity can be down-regulated by cAMP
Downloaded from http://hmg.oxfordjournals.org/ at CSU Fresno on July 22, 2015
p75NTR levels (Fig. 6). These results, together with the in vivo data described above, point to a
9 through PKA-mediated phosphorylation (57). No changes were detected in RhoA expression, but phospho-RhoA (pRhoA), the inactive form of the protein, augmented in the hippocampus of FTY720-treated R6/1 mice (Fig. 7D). Overall, these data suggest that enhancement of cAMP signaling by FTY720 could contribute to improve synaptic and cognitive deficits in R6/1 mice.
DISCUSSION The present work provides the first evidence that FTY720 prevents memory decline and enhances hippocampal synaptic plasticity in a mouse model of HD. Furthermore, we uncover an unexpected role for FTY720 in preventing imbalance of BDNF receptors in the hippocampus of
In HD patients, many lines of evidence demonstrate that cognitive disturbances precede the onset of motor symptoms and cell death (4, 58). This supports the view that early deficits in learning and memory are likely caused by synaptic dysfunction or cellular atrophy, rather than being a consequence of neuronal loss. In fact, human postmortem HD brain samples and most animal models display lower dendritic spine densities on cortical, striatal and hippocampal neurons (12, 42, 59, 60). As a consequence of spine loss, reduced levels of PSD-95 have also been found in several human postmortem samples and mouse models (43–45). Our results show that FTY720 chronic treatment from pre-symptomatic stages leads to a significant amelioration of long-term memory deficits in symptomatic R6/1 mice. Consistently, the effect of FTY720 on memory improvement was accompanied by prevention of PSD-95 down-regulation and preservation of dendritic spines in CA1 pyramidal neurons. Since dendritic spine pathology underlies learning and memory impairments (61) and the CA1 area is required for acquisition and retrieval of long-term memory (62), FTY720-induced structural changes correlate with the prevention of cognitive dysfunction in R6/1 mice. Chronic neuroinflammation involving glial activation is a common feature of several neurodegenerative diseases (47) and it has been associated with deficits in neural plasticity and memory (48). On the other hand, the hippocampal CA1 region has been suggested to be particularly vulnerable to neuroinflammation and aging (63, 64). We detected evidences of a
Downloaded from http://hmg.oxfordjournals.org/ at CSU Fresno on July 22, 2015
R6/1 mice, acting through the down-regulation of TNFα and p75NTR.
10 localized inflammatory response in the CA1 area of 20 week-old R6/1 mice, characterized by astrocyte over-activation and infiltration. The presence of reactive astrocytes correlated with upregulation of NF-κB signaling and increased production of pro-inflammatory mediators, namely TNFα and iNOS. Remarkably, this astrocyte-mediated inflammation was completely prevented by FTY720 chronic treatment. Consistent with these data, FTY720 administration in EAE mice showed a neuroprotective effect by reducing astrocyte activation and nitric oxide production (28, 31). In agreement with our findings, TNFα levels positively correlate with disease progression in HD patients (65), and targeting NF-κB and TNFα up-regulation has recently been shown to be an effective therapeutic approach in R6/2 mice (36, 66). Interestingly, NF-κB,
cognitive impairment (36, 67–69). In particular, TNFα has been suggested to inhibit long-term potentiation and mediate long-term depression of synaptic transmission in the CA1 hippocampal region (70, 71). Excessive release of pro-inflammatory factors can lead to reduced production of neurotrophins, which underlies detrimental effects on synaptic plasticity, learning and memory (48). BDNF, an important neurotrophin in HD pathophysiology (15–17, 72), has been particularly linked to inflammation-induced cognitive dysfunction (49, 50). We observed that FTY720 chronic administration was able to increase BDNF mRNA in the WT hippocampus, in accordance with previous reports (25, 51). Compared to WT animals, the R6/1 mice analysed in this study displayed a decrease in BDNF mRNA, which was prevented by FTY720 delivery. However, we could not detect significant changes in BDNF protein in the hippocampus of R6/1 mice treated with FTY720. Instead, FTY720 likely exerted its main therapeutic effects by preventing imbalance of BDNF receptors in the hippocampus of R6/1 mice. In line with these findings, recent research suggests that BDNF deficits may be preceded by dysfunctional signaling of TrkB and p75NTR, which modulate BDNF-dependent plasticity at both striatal and hippocampal levels (12, 73). Consistent with this hypothesis, FTY720 delivery normalized p75NTR expression and promoted phosphorylation of TrkB, despite no obvious increase in BDNF protein levels. Overall, these data suggest that FTY720 can modulate hippocampal balance of
Downloaded from http://hmg.oxfordjournals.org/ at CSU Fresno on July 22, 2015
TNFα and iNOS have been proposed as critical mediators of neuroinflammation-induced
11 p75NTR/TrkB receptors in R6/1 mice, which we previously identified as a critical mechanism for memory and synaptic function in HD (12). We detected increased levels of p75NTR in primary hippocampal neurons from R6/1 mice, but only after TNFα treatment. Accordingly, previous research has shown that proinflammatory cytokines, such as IL-1β and TNFα, can regulate expression of p75NTR in hippocampal neurons (52, 74). In particular, TNFα has been suggested to induce p75NTR expression via NF-κB (52). Therefore, the decreased NF-κB activity observed in the hippocampus of FTY720-treated R6/1 mice might also be associated with the reduction of p75NTR levels. Taken together, these data suggest a correlation between TNFα secreted by
results provide a possible mechanistic explanation for FTY720 cognitive benefits reported on AD (40, 41), because TNFα and p75NTR are up-regulated in AD patients (75, 76) and p75NTR inhibition prevents cognitive and neurite dysfunction in an AD mouse model (77). Chronic administration of FTY720 raised cAMP levels and promoted CREB phosphorylation in the hippocampus of R6/1 mice. Consistent with our behavioral data, increased cAMP and pCREB hippocampal levels have been previously shown to improve spatial and object recognition memories in R6/1 mice (78). Moreover, up-regulation of cAMP/CREB signaling in the CA1 region reversed memory deficits and attenuated NF-κB activity in an AD mouse model (79). FTY720 treatment in R6/1 mice also promoted phosphorylation of RhoA, a GTPase that functions as a negative modulator of dendritic spine formation and maintenance (80, 81). RhoA phosphorylation by PKA facilitates its binding to RhoGDI (57), which leads to displacement of RhoA from the plasma membrane and precludes it from interacting with downstream effectors (82). Therefore, the increased pRhoA levels observed in the hippocampus of R6/1 mice after FTY720 treatment may be a consequence of cAMP/PKA pathway activation. Furthermore, rise of cAMP lowers TNFα levels (83), and TNFα has been involved in the negative regulation of neuritic branching by promoting RhoA activity in cultured hippocampal neurons (84). In addition, we recently showed that p75NTR contributes to synaptic dysfunction and memory decline in HD by enhancing RhoA activity
Downloaded from http://hmg.oxfordjournals.org/ at CSU Fresno on July 22, 2015
activated astrocytes and the up-regulation of p75NTR in hippocampal neurons. Remarkably, our
12 (12). Altogether, these data support the notion that FTY720 preserves dendritic spines through negative regulation of both p75NTR and RhoA. In conclusion, we provide compelling evidence that FTY720 administration improves hippocampal-dependent cognitive function and structural synaptic plasticity in R6/1 mice. We propose the reduction of NF-κB activity and astrocyte-derived TNFα as the main mechanism of action of the drug in the R6/1 hippocampus, having a particular impact on p75NTR and cAMP signaling within CA1 pyramidal neurons (Fig. 8). These data argue in favor of a scenario where FTY720 treatment at the early stages of HD should help restoring normal BDNF signaling and preventing synaptic and memory impairments. However, before starting clinical trials, it would
mHtt. Altogether, our findings take a significant step in understanding how FTY720 acts on brain cells, underscoring the pleiotropic nature of this drug.
MATERIALS AND METHODS Mice Male R6/1 transgenic mice expressing exon-1 of mHtt were obtained from Jackson Laboratory (Bar Harbor, ME, USA) and maintained in a B6CBA background. WT littermate animals were used as controls. Genotypes were determined by polymerase chain reaction (PCR). CAG repeat length was determined as previously described (1) and our R6/1 colony has 145 CAG repeats (16). Experiments were conducted in a blind-coded manner respect to genotype and treatment condition of mice. Experimental procedures were approved by the Animal Experimentation Ethics Committee of the University of Barcelona, in compliance with Spanish (RD 53/2013) and European (2010/63/UE) regulations for the care and use of laboratory animals.
Drug administration FTY720 was obtained as a powder from Cayman Chemicals and dissolved in EtOH 10% in distilled water (vehicle). For chronic pharmacological treatment, intraperitoneal injections of FTY720 were given every 4 days at a dose of 0.3 mg/kg during 12 weeks, starting at 8 weeks of
Downloaded from http://hmg.oxfordjournals.org/ at CSU Fresno on July 22, 2015
be recommendable to test the drug in a less aggressive HD mouse model expressing full-length
13 age. Animals were weighted weekly in order to determine the appropriate dose. Mice were divided into two cohorts and distributed into four experimental groups (n=10-12 each): WT + FTY720, WT + vehicle, R6/1 + FTY720 and R6/1 + vehicle. Last dose of FTY720 was administered 24h before killing the animals for histological and biochemical analysis.
Behavioral tests Novel object recognition test (NORT). NORT was performed at 17 weeks of age as previously described (12). Briefly, after 2 days of habituation to the open field, two similar objects were presented to each mouse and 24h later one of them was changed by a new object. Object
software (Panlab). T-maze spontaneous alternation task test (T-SAT). T-SAT was performed at 18 weeks of age as described elsewhere (10). During the training trial, one arm of the T-maze was closed. After 4h, mice were allowed to freely explore both the familiar and the novel arm. The arm preference was determined by calculating the percentage of time spent in each arm. Animals were tracked and recorded with SMART software.
Golgi-Cox staining Fresh brain hemispheres were processed following the Golgi-Cox method, as previously described (85, 86). Bright-field images of Golgi-impregnated hippocampal CA1 neurons were captured with a Leica epifluorescence microscope (×100 oil objective). Image Z stacks were taken every 0.5 μm and analyzed with Image J software. Dendritic segments (>10 μm long) were traced and spine density was quantified in 30-40 second order apical dendrites, from at least 10 different neurons per animal (n=3 mice per genotype and treatment).
Primary hippocampal culture Primary hippocampal neuronal cultures were performed as described elsewhere (87). Brains from E18.5 WT and R6/1 mice embryos were excised and placed in sterile phosphate-buffered
Downloaded from http://hmg.oxfordjournals.org/ at CSU Fresno on July 22, 2015
preference was measured as the percentage of time exploring each object, using SMART junior
14 saline (PBS). The hippocampus was dissected and gently dissociated with a fire-polished Pasteur pipette. Cells were plated onto 6-well plates pre-coated with poly-D-lysine (0.1 mg/ml; Sigma-Aldrich) at a density of 50000 cells/cm2. Neurobasal medium supplemented with B27 and Glutamax (Gibco) was used to grow the cells in serum-free conditions. Cell cultures were maintained in an incubator with 5% CO2 at 37°C. Neuronal cells represented more than 95% of the cell population. Media changes were performed weekly and cells were treated after 21 DIV. TNFα (Life Technologies) was added to the culture at a concentration of 10 ng/ml and FTY720P (provided by CHDI Foundation) was used at 10 nM. 8 hours after treatment, cells were washed with cold PBS and lysed for WB analysis.
For immunohistochemical analysis, brains were fixed by inmersion in paraformaldehyde (Sigma-Aldrich) 4% in PBS, cryoprotected in PBS/Sucrose 30% and frozen in methyl-butane (Sigma-Aldrich). 30 µm thick coronal sections of the brain were obtained using a cryostat (Microm) and kept in PBS as free-floating sections. Immunohistochemistry was performed as previously described (Giralt et al., 2010). Tissue was incubated with the following primary antibodies: rabbit anti-p75NTR (1:100, Promega), rabbit anti-Iba1 (Wako, 1:500), rabbit antiNFкB p65 (1:50, Santa Cruz Biotechnology), mouse anti-MAP2 (1:200, Sigma-Aldrich), mouse anti-GFAP (1:500, Sigma-Aldrich), mouse anti-cAMP (1:100, Abcam), followed by the corresponding secondary antibodies. Images were acquired with a Leica SP5 laser scanning confocal microscope (Leica). cAMP immunoreactivity in the CA1 area of the hippocampus was quantified by analysis of integrated optical density with ImageJ software, as previously described (78). Automated quantification of astrocyte number, size and GFAP intensity was performed using Cell Profiler v2.8 software. Briefly, images from the CA1 region were acquired at 10x magnification and Hoescht fluorescence was used in a first segmentation to delimitate CA1 area. Colocalizing pixels from GFAP images were identified with primary object detection modules and further measured to compare relevant properties using appropriate
Downloaded from http://hmg.oxfordjournals.org/ at CSU Fresno on July 22, 2015
Immunohistochemistry
15 modules of this software. The pipeline file containing the specific parameters of the automated quantification is available upon request (http://www.cellprofiler.org/citations.shtml).
Western blot After killing the animals by cervical dislocation, striatum and hippocampus were rapidly removed and homogenized. Protein extraction and western blotting were performed as described elsewhere (35). Blots were incubated with the following antibodies (1:1000, unless stated otherwise): anti-EDG1 (Abcam), anti-BDNF (Santa Cruz Biotechnology), anti-TrkB (Santa Cruz Biotechnology), anti-pTrkB (817) (Abcam), anti-p75NTR (Promega), anti-RhoA (Santa
pCREB (Abcam), anti-NFкB p65 (Santa Cruz Biotechnology), anti-IкBα (1:250, Santa Cruz Biotechnology) and anti-PSD95 (1:2000, Affinity BioReagents). Loading control was performed by reincubating the membranes with either anti--tubulin (1:100000; Sigma-Aldrich) or anti-β-actin (1:20000; Sigma-Aldrich). Membranes were incubated with the corresponding horseradish peroxidase-conjugated antibody (1:2000; Promega). Immunoreactive bands were visualized using the Western Blotting Luminol Reagent (Santa Cruz Biotechnology) and quantified by a computer-assisted densitometer (Gel-Pro Analyzer). Discontinued bands indicate that samples were run in the same gel but were not contiguous.
RNA isolation and quantitative real-time PCR mRNA expression was determined in 20 week-old hippocampal samples from vehicle- and FTY720-treated R6/1 and WT mice. Total RNA was isolated with Tri-Reagent (SigmaAldrich), following product instructions. 2 µg of RNA for each condition were reverse transcribed using PrimeScript RT reagent kit (Takara). cDNA was diluted (1/25) and 2 µl were used to perform qRT-PCR. PrimeTime qRT-PCR assays (see table) were used as recommended by provider (IDT technologies). Rn18s and GAPDH mRNA levels were not altered by treatments and were used as reference genes. qRT-PCR was carried out with Premix Ex Taq (Takara) in 6 µl of final volume using CFX384-C1000 Thermal Cycler equipment (Bio-Rad).
Downloaded from http://hmg.oxfordjournals.org/ at CSU Fresno on July 22, 2015
Cruz Biotechnology), anti-pRhoA (Santa Cruz Biotechnology), anti-CREB (Abcam), anti-
16 Samples were run for 40 cycles (95ºC for 5 sec, 60ºC for 20 sec). Relative gene expression values were calculated with -2∆Ct (88), using Bio-Rad CFX manager software (Bio-Rad). Gene Name GAPDH Rn18s BDNF TNF NOS2
Assay Name Mm.PT.39a.1 Control 18s Mm00432069_m1 Mm.PT.56a.12575861 Mm.PT.56a.14178818
RefSeqNumber NM_008084 NR_003286 NM_001048141 NM_013693 NM_010927
Statistical analysis Statistical analyses were performed as appropriate using GraphPad Prism software and are
considered statistically significant.
ACKNOWLEDGEMENTS We thank Ana López and María Teresa Muñoz for technical assistance, and Teresa Rodrigo and the staff of the Animal Care Facility (Facultat de Psicologia, Universitat de Barcelona) for their help. We are grateful to the Confocal Microscopy Facility (Campus Casanova) of the Centre Científic i Tecnològic (Universitat de Barcelona) for their advice with microscopy techniques. This work was supported by an IDIBAPS Postdoctoral Fellowship-BIOTRACK, supported by the European Community’s Seventh Framework Program (EC FP7/2007-2013) under the grant agreement number 229673 and the Spanish Ministry of Economy and Competitiveness (MINECO) through the grant COFUND2013-40261 to A.M.; Ministerio de Economía y Competitividad [SAF2012-39142 to S.G., RETICS (RD06/0010/0006; Red de Terapia Celular) and SAF2012-37417 to J.M.C., CIBERNED and SAF2011-29507 to J.A]; Cure Huntington’s Disease Initiative (CHDI) to S.G. and J.M.C.; and Generalitat de Catalunya (2014SGR- 968 to J.A.).
CONFLICT OF INTEREST STATEMENT The authors declare that there are no conflicts of interest.
Downloaded from http://hmg.oxfordjournals.org/ at CSU Fresno on July 22, 2015
indicated in the figure legends. All data are expressed as mean ± S.E.M. Values of p
1 Fingolimod (FTY720) enhances hippocampal synaptic plasticity and memory in Huntington’s disease by preventing p75NTR up-regulation and astrocyte-mediated inflammation
Andrés Miguez1,2,3, Gerardo García-Díaz Barriga1,2,3, Verónica Brito1,2,3, Marco Straccia1,2,3, Albert Giralt1,2,3, Silvia Ginés1,2,3, Josep M. Canals1,2,3, Jordi Alberch1,2,3,* 1
Departament de Biologia Cel·lular, Immunologia i Neurociències, Facultat de Medicina,
2
Institut d'Investigacions Biomèdiques August Pi i Sunyer (IDIBAPS), 08036 Barcelona, Spain
3
Centro de Investigación Biomédica en Red sobre Enfermedades Neurodegenerativas
(CIBERNED), Spain
*
To whom correspondence should be addressed at: Departament de Biologia Cel·lular,
Immunologia i Neurociències, Facultat de Medicina, Universitat de Barcelona, Casanova 143, 08036 Barcelona, Spain; Tel: +34 934035285, Fax: +34 934021907; E-mail: [email protected]
© The Author 2015. Published by Oxford University Press. All rights reserved. For Permissions, please email: [email protected]
Downloaded from http://hmg.oxfordjournals.org/ at CSU Fresno on July 22, 2015
Universitat de Barcelona, 08036 Barcelona, Spain
2
ABSTRACT Huntington’s disease (HD) is a hereditary neurodegenerative disorder characterized by motor and cognitive impairments, involving striatum, cortex and hippocampus. Synaptic and memory dysfunction in HD mouse models have been related to low levels of brain-derived neurotrophic factor (BDNF) and imbalance between TrkB and p75NTR receptors. In addition, astrocyte overactivation has also been suggested to contribute to HD cognitive deficits. Fingolimod (FTY720), a modulator of sphingosine-1 phosphate (S1P) receptors, has been shown to increase
inflammatory response. In this view, we have investigated whether FTY720 improves synaptic plasticity and memory in the R6/1 mouse model of HD, through regulation of BDNF signaling and astroglial reactivity. Chronic administration of FTY720 from pre-symptomatic stages ameliorated long-term memory deficits and dendritic spine loss in CA1 hippocampal neurons from R6/1 mice. Furthermore, FTY720 delivery prevented astrogliosis and over-activation of nuclear factor kappa beta (NF-κB) signaling in the R6/1 hippocampus, reducing tumor necrosis factor alpha (TNFα) and induced nitric oxide synthase (iNOS) levels. TNFα decrease correlated with the normalization of p75NTR expression in the hippocampus of FTY720-treated R6/1 mice, thus preventing p75NTR/TrkB imbalance. In addition, FTY720 increased cAMP levels and promoted phosphorylation of CREB and RhoA in the hippocampus of R6/1 mice, further supporting its role in the enhancement of synaptic plasticity. Our findings provide new insights into the mechanism of action of FTY720 and reveal a novel therapeutic strategy to treat memory deficits in HD.
Downloaded from http://hmg.oxfordjournals.org/ at CSU Fresno on July 22, 2015
BDNF levels and to reduce astrogliosis, proving its potential to regulate trophic support and
3
INTRODUCTION Huntington’s disease (HD) is an autosomal dominant neurodegenerative disorder caused by the expansion of a CAG tract in the exon-1 of the huntingtin gene (1). Clinical diagnosis of HD relies on the manifestation of motor abnormalities. However, cognitive and behavioral impairments are evident at least 15 years before motor diagnosis (2–4), and are currently considered the greatest burden on HD families (5). The primary regions of neurodegeneration in HD have long been considered the striatum and cerebral cortex (6, 7), but other structures involved in cognition, particularly the hippocampus, are affected in early stages of the disease
hippocampal dysfunction (10–12). Neurotrophins promote the development and survival of neurons. Among them, brainderived neurotrophic factor (BDNF) stands out for its ability to regulate synaptic plasticity and cognitive function (13). BDNF is initially synthesized as a precursor protein called pro-BDNF, which is then cleaved by proteases to generate the mature form (14). Reduction in BDNF levels affects the onset and severity of HD in mouse models (15, 16). Despite its therapeutic potential, BDNF delivery is hampered by its poor pharmacokinetic properties and its inability to cross the blood-brain barrier (17). TrkB and p75NTR receptors can bind to both BDNF and pro-BDNF (18, 19). TrkB activation promotes long-term potentiation, whereas p75NTR is regarded as a negative modulator of synaptic plasticity (20, 21). Imbalance of TrkB/p75NTR expression has been reported in the striatum and hippocampus of HD patients and mice, affecting cell survival and synaptic plasticity (12, 22). Therefore, current research is aimed at developing strategies to promote the release of endogenous BDNF and to modulate TrkB and p75NTR receptors (12, 23, 24). Fingolimod (FTY720), a compound used as an immunomodulator in Multiple Sclerosis patients (Brinkmann et al., 2010), has recently been shown to improve BDNF release in the striatum and to rescue motor deficits in mouse models of Rett’s Syndrome and HD (25, 26). FTY720 is a structural analogue of sphingosine that can cross the blood-brain barrier and reach
Downloaded from http://hmg.oxfordjournals.org/ at CSU Fresno on July 22, 2015
(8, 9). Notably, deficits in spatial and recognition memory have been recently associated to
4 the brain, where it is converted to FTY720-phosphate (FTY720-P). FTY720-P initially acts as an agonist of sphingosine-1 phosphate (S1P) receptors, but chronic exposure to the drug leads to receptor internalization and functional antagonism (27). Most of FTY720 therapeutic effects within the brain are mediated by S1P1 receptor, which is highly expressed in astrocytes (28) and to a lesser extent in neurons (25), oligodendrocytes (29) and microglia (30). Conditional deletion of S1P1 receptor from astrocytes mimics the effects of FTY720 in experimental autoimmune encephalomyelitis (EAE), suggesting a neuroprotective effect through astrocytes modulation (28). Indeed, FTY720 treatment reduced astrogliosis, nuclear factor kappa beta (NFκB) activity and nitric oxide production in EAE mice, further supporting the relevance of S1P
HD (32–35) and associated with increased NF-κB activity (36). Furthermore, expression of mutant huntingtin (mHtt) in astrocytes is sufficient to induce neurological symptoms characteristic of HD mice (37, 38), strongly suggesting a critical role of astroglial dysfunction in HD pathogenesis. Besides its capacity to modulate astrocyte reactivity, FTY720 was found to reduce the severity of dendritic spine loss in striatal neurons of EAE mice (39) and to improve cognitive function in rodent models of Alzheimer’s disease (40, 41), suggesting additional effects on structural and functional synaptic plasticity. However, the mechanism through which this compound may enhance hippocampal synaptic plasticity remains unclear, and no studies to date have examined its ability to counteract memory deficits in HD. The present work was designed to investigate the potential of FTY720 to improve synaptic and memory function in R6/1 mice through regulation of hippocampal BDNF signaling and astrocyte activation.
RESULTS Chronic administration of FTY720 ameliorates long-term memory deficits in R6/1 mice Since the most prominent effects of FTY720 within the brain are related to its action on S1P1 receptor (27), we first examined by Western blot (WB) whether S1P1 expression was altered in the hippocampus of R6/1 mice. At 20 weeks of age, S1P1 protein levels were similar in wild-
Downloaded from http://hmg.oxfordjournals.org/ at CSU Fresno on July 22, 2015
antagonism in astroglia for drug efficacy (31). Interestingly, astrogliosis has been reported in
5 type (WT) and R6/1 mice (Supplementary Material, Fig. S1), suggesting a similar sensitivity to the drug in both genotypes. To analyze whether FTY720 improves long-term memory deficits in HD, either FTY720 or vehicle solution were delivered intraperitoneally in R6/1 and WT mice starting at pre-symptomatic stages (8 weeks) and continuing until 20 weeks of age, when R6/1 mice exhibit an overt motor and cognitive phenotype. Recognition memory was evaluated at 17 weeks of age by performing the novel object recognition test (NORT). We found that, whereas WT mice spent more time exploring the new object than the old one, this preference was lost in R6/1 mice. Notably, chronically treated R6/1 mice showed a similar preference for the new object than WT mice (Fig. 1A). To further confirm this cognitive improvement, we next
During the test, vehicle-injected R6/1 mice did not show any preference for the new arm respect to the old one, whereas FTY720-treated R6/1 mice explored more the new arm (Fig. 1B). No differences were detected between vehicle- and FTY720-treated WT mice. Therefore, FTY720 administration ameliorates spatial and recognition memory deficits in R6/1 mice.
FTY720 chronic treatment prevents dendritic spine loss and PSD-95 down-regulation in the hippocampus of R6/1 mice Learning and memory impairments are often associated with dendritic spine pathology and most HD animal models show lower spine density (12, 42). To analyze whether the memory improvement observed after chronic delivery of FTY720 correlated with structural synaptic changes, we quantified the number of dendritic spines in hippocampal neurons of R6/1 mice at 20 weeks of age. Following Golgi-Cox staining, the number of second order apical dendrites was counted in pyramidal neurons of the CA1 area (Fig. 2A). Spine density was decreased by 17% in R6/1 mice, compared to WT animals. FTY720 treatment prevented spine loss in the R6/1 hippocampus (Fig. 2A and B), while it had no effect in WT mice. Maintenance and stabilization of dendritic spines can be regulated by scaffolding protein post-synaptic density-95 (PSD-95), which is known to be reduced in HD patients and animal models (43–45). Therefore, we examined the expression of PSD-95 by WB in the hippocampus at 20 weeks of age. R6/1
Downloaded from http://hmg.oxfordjournals.org/ at CSU Fresno on July 22, 2015
performed the T-maze spontaneous alternation task (T-SAT), which measures spatial memory.
6 mice showed a decrease in PSD-95 protein levels that was prevented by FTY720 treatment (Fig. 2C), indicating a role for FTY720 in modulating structural synaptic plasticity.
FTY720 attenuates astrocyte-mediated inflammation in the hippocampus of R6/1 mice Astrocyte over-activation has been reported in HD patients (34) and mouse models, where it has been suggested to contribute to poor cognitive function (36). We thus examined if there was any evidence of astrogliosis in the hippocampus of R6/1 mice at 20 weeks of age. Increased astrocyte infiltration and activation was detected in the CA1 region of the R6/1 hippocampus (Fig. 3A), with no apparent changes in astrocyte proliferation. Within the CA1 area, astrocytes
(Fig. 3D). We found no significant differences in the size and density of Iba1+ cells between WT and R6/1 mice, suggesting no obvious microgliosis (data not shown). Remarkably, chronic administration of FTY720 reduced the number, size and GFAP intensity of astrocytes within the CA1 region of R6/1 mice, proving the ability of the drug to prevent astrogliosis (Fig. 3A-D). Reactive astrocytes have been directly associated with aberrant NF-κB activity in R6/2 mice (36). In order to learn whether NF-κB activity was altered in the hippocampus of R6/1 mice, we performed a WB analysis for NF-κB and inhibitor of kappa beta α (IкBα), an inhibitory protein that sequesters NF-κB in the cytoplasm (46). Vehicle-injected WT and R6/1 mice showed similar NF-κB protein levels (Supplementary Material, Fig. S2B), but IкBα expression was strongly reduced in the R6/1 hippocampus (Fig. 3E), suggesting increased NFκB translocation to the nucleus. Immunohistochemical analysis confirmed the presence of reactive astrocytes surrounding the CA1 area displaying increased nuclear localization of NFκB p65 (Supplementary Material, Fig. S2A), indicating an over-activation of NF-кB signaling in R6/1 mice. Interestingly, FTY720 chronic treatment prevented IкBα down-regulation (Fig. 3E) and subsequent NF-κB p65 translocation to the nucleus in astroglial cells (Supplementary Material, Fig. S2A), revealing a role for FTY720 in the reduction of hippocampal NF-κB activity. Since NF-κB regulates the expression of a large number of genes associated with neuroinflammation (47), we next examined whether FTY720-induced normalization of NF-κB
Downloaded from http://hmg.oxfordjournals.org/ at CSU Fresno on July 22, 2015
displayed a significant increase in number (Fig. 3B), size (Fig. 3C) and GFAP immunoreactivity
7 signaling correlated with decreased levels of pro-inflammatory mediators in the R6/1 hippocampus. As shown by quantitative real-time PCR (qRT-PCR) analysis, FTY720-treated R6/1 mice displayed decreased mRNA levels of tumor necrosis factor alpha (TNFα) (Fig. 3F) and inducible nitric oxide synthase (iNOS) (Fig. 3G), with no significant changes in IL-1β and IL-6 expression (data not shown). Altogether, these data support a role for FTY720 in diminishing astrocyte-mediated inflammation in the R6/1 hippocampus.
FTY720 increases BDNF mRNA in the hippocampus of R6/1 mice The effects of pro-inflammatory factors on synaptic plasticity and memory have been associated
been shown to increase BDNF levels in the hippocampus of WT mice (25, 51). To find out whether FTY720 chronic treatment was able to improve BDNF synthesis in the hippocampus of R6/1 mice, we performed qRT-PCR and WB analysis at 20 weeks of age to assess mRNA and protein levels, respectively. R6/1 mice displayed a decrease in BDNF mRNA, which was prevented by FTY720 delivery (Fig. 4A). However, we found no differences in BDNF protein, neither in WT nor in R6/1 mice (Fig. 4B).
FTY720 chronic treatment prevents p75NTR/TrkB imbalance in the hippocampus of R6/1 mice Imbalance of BDNF receptors, p75NTR and TrkB, has been reported in the hippocampus of HD patients and mice and associated to synaptic and memory deficits (12). Immunohistochemical and WB analysis of the R6/1 hippocampus at 20 weeks of age revealed abnormally high levels of p75NTR (Fig. 5A and B), which correlated with decreased TrkB expression in R6/1 mice (Fig. 5C). Chronic delivery of FTY720 was sufficient to prevent up-regulation of p75NTR expression in the hippocampus of R6/1 mice (Fig. 5A and B). Moreover, TrkB protein levels were not significantly different between WT and R6/1 mice after treatment (Fig. 5C). Interestingly, FTY720 promoted activation of TrkB in the R6/1 hippocampus, as showed by the increase in TrkB phosphorylation (Fig. 5D).
Downloaded from http://hmg.oxfordjournals.org/ at CSU Fresno on July 22, 2015
with altered neurotrophic support, particularly with BDNF (48–50). Moreover, FTY720 has
8 p75NTR expression can be regulated by TNFα in hippocampal cells (52). To determine the potential contribution of TNFα to the up-regulation of p75NTR in the hippocampus of R6/1 mice, we purified primary hippocampal neurons from WT and R6/1 animals. After 21 DIV, we examined p75NTR expression under basal conditions and following stimulation with TNFα (10 ng/ml), in the presence and in the absence of FTY720-P (10 nM). p75NTR protein levels were similar in WT and R6/1 primary hippocampal neurons under basal conditions (Fig. 6). Nevertheless, incubation with TNFα significantly increased p75NTR expression in R6/1 hippocampal neurons, whereas it had no effect on WT neurons. Remarkably, addition of FTY720-P to R6/1 hippocampal cultures was sufficient to counteract the effect of TNFα on
robust effect of FTY720 on the reduction of p75NTR expression in hippocampal neurons, likely acting through inhibition of TNFα.
FTY720 chronic treatment enhances cAMP signaling in the hippocampus of R6/1 mice Activation of cyclic AMP (cAMP) signaling improves synaptic plasticity and hippocampusdependent long-term memory (53). Notably, S1P inhibits cAMP pathway (54), whereas FTY720-P has been suggested to induce sustained cAMP signaling in vitro (55). In order to determine whether FTY720 treatment modulated the levels of cAMP and its downstream targets in a HD mouse model, we examined cAMP expression in the hippocampal CA1 region by immunohistochemical analysis. Both WT and R6/1 mice showed an evident increase in cAMP immunoreactivity after FTY720 chronic administration (Fig. 7A and B). To validate the activation of cAMP downstream signaling, we next examined two well-known targets whose phosphorylation is mainly mediated by protein kinase A (PKA). First, levels of the cAMPresponse element binding (CREB) and its phosphorylated form (pCREB) were assessed in the hippocampus of both mouse genotypes. FTY720 treatment increased pCREB protein in R6/1 mice, albeit not in WT mice, without significant changes in total CREB levels (Fig. 7C). Next, we determined the protein levels of RhoA, a negative modulator of spine density and complexity of hippocampal neurons (56) whose activity can be down-regulated by cAMP
Downloaded from http://hmg.oxfordjournals.org/ at CSU Fresno on July 22, 2015
p75NTR levels (Fig. 6). These results, together with the in vivo data described above, point to a
9 through PKA-mediated phosphorylation (57). No changes were detected in RhoA expression, but phospho-RhoA (pRhoA), the inactive form of the protein, augmented in the hippocampus of FTY720-treated R6/1 mice (Fig. 7D). Overall, these data suggest that enhancement of cAMP signaling by FTY720 could contribute to improve synaptic and cognitive deficits in R6/1 mice.
DISCUSSION The present work provides the first evidence that FTY720 prevents memory decline and enhances hippocampal synaptic plasticity in a mouse model of HD. Furthermore, we uncover an unexpected role for FTY720 in preventing imbalance of BDNF receptors in the hippocampus of
In HD patients, many lines of evidence demonstrate that cognitive disturbances precede the onset of motor symptoms and cell death (4, 58). This supports the view that early deficits in learning and memory are likely caused by synaptic dysfunction or cellular atrophy, rather than being a consequence of neuronal loss. In fact, human postmortem HD brain samples and most animal models display lower dendritic spine densities on cortical, striatal and hippocampal neurons (12, 42, 59, 60). As a consequence of spine loss, reduced levels of PSD-95 have also been found in several human postmortem samples and mouse models (43–45). Our results show that FTY720 chronic treatment from pre-symptomatic stages leads to a significant amelioration of long-term memory deficits in symptomatic R6/1 mice. Consistently, the effect of FTY720 on memory improvement was accompanied by prevention of PSD-95 down-regulation and preservation of dendritic spines in CA1 pyramidal neurons. Since dendritic spine pathology underlies learning and memory impairments (61) and the CA1 area is required for acquisition and retrieval of long-term memory (62), FTY720-induced structural changes correlate with the prevention of cognitive dysfunction in R6/1 mice. Chronic neuroinflammation involving glial activation is a common feature of several neurodegenerative diseases (47) and it has been associated with deficits in neural plasticity and memory (48). On the other hand, the hippocampal CA1 region has been suggested to be particularly vulnerable to neuroinflammation and aging (63, 64). We detected evidences of a
Downloaded from http://hmg.oxfordjournals.org/ at CSU Fresno on July 22, 2015
R6/1 mice, acting through the down-regulation of TNFα and p75NTR.
10 localized inflammatory response in the CA1 area of 20 week-old R6/1 mice, characterized by astrocyte over-activation and infiltration. The presence of reactive astrocytes correlated with upregulation of NF-κB signaling and increased production of pro-inflammatory mediators, namely TNFα and iNOS. Remarkably, this astrocyte-mediated inflammation was completely prevented by FTY720 chronic treatment. Consistent with these data, FTY720 administration in EAE mice showed a neuroprotective effect by reducing astrocyte activation and nitric oxide production (28, 31). In agreement with our findings, TNFα levels positively correlate with disease progression in HD patients (65), and targeting NF-κB and TNFα up-regulation has recently been shown to be an effective therapeutic approach in R6/2 mice (36, 66). Interestingly, NF-κB,
cognitive impairment (36, 67–69). In particular, TNFα has been suggested to inhibit long-term potentiation and mediate long-term depression of synaptic transmission in the CA1 hippocampal region (70, 71). Excessive release of pro-inflammatory factors can lead to reduced production of neurotrophins, which underlies detrimental effects on synaptic plasticity, learning and memory (48). BDNF, an important neurotrophin in HD pathophysiology (15–17, 72), has been particularly linked to inflammation-induced cognitive dysfunction (49, 50). We observed that FTY720 chronic administration was able to increase BDNF mRNA in the WT hippocampus, in accordance with previous reports (25, 51). Compared to WT animals, the R6/1 mice analysed in this study displayed a decrease in BDNF mRNA, which was prevented by FTY720 delivery. However, we could not detect significant changes in BDNF protein in the hippocampus of R6/1 mice treated with FTY720. Instead, FTY720 likely exerted its main therapeutic effects by preventing imbalance of BDNF receptors in the hippocampus of R6/1 mice. In line with these findings, recent research suggests that BDNF deficits may be preceded by dysfunctional signaling of TrkB and p75NTR, which modulate BDNF-dependent plasticity at both striatal and hippocampal levels (12, 73). Consistent with this hypothesis, FTY720 delivery normalized p75NTR expression and promoted phosphorylation of TrkB, despite no obvious increase in BDNF protein levels. Overall, these data suggest that FTY720 can modulate hippocampal balance of
Downloaded from http://hmg.oxfordjournals.org/ at CSU Fresno on July 22, 2015
TNFα and iNOS have been proposed as critical mediators of neuroinflammation-induced
11 p75NTR/TrkB receptors in R6/1 mice, which we previously identified as a critical mechanism for memory and synaptic function in HD (12). We detected increased levels of p75NTR in primary hippocampal neurons from R6/1 mice, but only after TNFα treatment. Accordingly, previous research has shown that proinflammatory cytokines, such as IL-1β and TNFα, can regulate expression of p75NTR in hippocampal neurons (52, 74). In particular, TNFα has been suggested to induce p75NTR expression via NF-κB (52). Therefore, the decreased NF-κB activity observed in the hippocampus of FTY720-treated R6/1 mice might also be associated with the reduction of p75NTR levels. Taken together, these data suggest a correlation between TNFα secreted by
results provide a possible mechanistic explanation for FTY720 cognitive benefits reported on AD (40, 41), because TNFα and p75NTR are up-regulated in AD patients (75, 76) and p75NTR inhibition prevents cognitive and neurite dysfunction in an AD mouse model (77). Chronic administration of FTY720 raised cAMP levels and promoted CREB phosphorylation in the hippocampus of R6/1 mice. Consistent with our behavioral data, increased cAMP and pCREB hippocampal levels have been previously shown to improve spatial and object recognition memories in R6/1 mice (78). Moreover, up-regulation of cAMP/CREB signaling in the CA1 region reversed memory deficits and attenuated NF-κB activity in an AD mouse model (79). FTY720 treatment in R6/1 mice also promoted phosphorylation of RhoA, a GTPase that functions as a negative modulator of dendritic spine formation and maintenance (80, 81). RhoA phosphorylation by PKA facilitates its binding to RhoGDI (57), which leads to displacement of RhoA from the plasma membrane and precludes it from interacting with downstream effectors (82). Therefore, the increased pRhoA levels observed in the hippocampus of R6/1 mice after FTY720 treatment may be a consequence of cAMP/PKA pathway activation. Furthermore, rise of cAMP lowers TNFα levels (83), and TNFα has been involved in the negative regulation of neuritic branching by promoting RhoA activity in cultured hippocampal neurons (84). In addition, we recently showed that p75NTR contributes to synaptic dysfunction and memory decline in HD by enhancing RhoA activity
Downloaded from http://hmg.oxfordjournals.org/ at CSU Fresno on July 22, 2015
activated astrocytes and the up-regulation of p75NTR in hippocampal neurons. Remarkably, our
12 (12). Altogether, these data support the notion that FTY720 preserves dendritic spines through negative regulation of both p75NTR and RhoA. In conclusion, we provide compelling evidence that FTY720 administration improves hippocampal-dependent cognitive function and structural synaptic plasticity in R6/1 mice. We propose the reduction of NF-κB activity and astrocyte-derived TNFα as the main mechanism of action of the drug in the R6/1 hippocampus, having a particular impact on p75NTR and cAMP signaling within CA1 pyramidal neurons (Fig. 8). These data argue in favor of a scenario where FTY720 treatment at the early stages of HD should help restoring normal BDNF signaling and preventing synaptic and memory impairments. However, before starting clinical trials, it would
mHtt. Altogether, our findings take a significant step in understanding how FTY720 acts on brain cells, underscoring the pleiotropic nature of this drug.
MATERIALS AND METHODS Mice Male R6/1 transgenic mice expressing exon-1 of mHtt were obtained from Jackson Laboratory (Bar Harbor, ME, USA) and maintained in a B6CBA background. WT littermate animals were used as controls. Genotypes were determined by polymerase chain reaction (PCR). CAG repeat length was determined as previously described (1) and our R6/1 colony has 145 CAG repeats (16). Experiments were conducted in a blind-coded manner respect to genotype and treatment condition of mice. Experimental procedures were approved by the Animal Experimentation Ethics Committee of the University of Barcelona, in compliance with Spanish (RD 53/2013) and European (2010/63/UE) regulations for the care and use of laboratory animals.
Drug administration FTY720 was obtained as a powder from Cayman Chemicals and dissolved in EtOH 10% in distilled water (vehicle). For chronic pharmacological treatment, intraperitoneal injections of FTY720 were given every 4 days at a dose of 0.3 mg/kg during 12 weeks, starting at 8 weeks of
Downloaded from http://hmg.oxfordjournals.org/ at CSU Fresno on July 22, 2015
be recommendable to test the drug in a less aggressive HD mouse model expressing full-length
13 age. Animals were weighted weekly in order to determine the appropriate dose. Mice were divided into two cohorts and distributed into four experimental groups (n=10-12 each): WT + FTY720, WT + vehicle, R6/1 + FTY720 and R6/1 + vehicle. Last dose of FTY720 was administered 24h before killing the animals for histological and biochemical analysis.
Behavioral tests Novel object recognition test (NORT). NORT was performed at 17 weeks of age as previously described (12). Briefly, after 2 days of habituation to the open field, two similar objects were presented to each mouse and 24h later one of them was changed by a new object. Object
software (Panlab). T-maze spontaneous alternation task test (T-SAT). T-SAT was performed at 18 weeks of age as described elsewhere (10). During the training trial, one arm of the T-maze was closed. After 4h, mice were allowed to freely explore both the familiar and the novel arm. The arm preference was determined by calculating the percentage of time spent in each arm. Animals were tracked and recorded with SMART software.
Golgi-Cox staining Fresh brain hemispheres were processed following the Golgi-Cox method, as previously described (85, 86). Bright-field images of Golgi-impregnated hippocampal CA1 neurons were captured with a Leica epifluorescence microscope (×100 oil objective). Image Z stacks were taken every 0.5 μm and analyzed with Image J software. Dendritic segments (>10 μm long) were traced and spine density was quantified in 30-40 second order apical dendrites, from at least 10 different neurons per animal (n=3 mice per genotype and treatment).
Primary hippocampal culture Primary hippocampal neuronal cultures were performed as described elsewhere (87). Brains from E18.5 WT and R6/1 mice embryos were excised and placed in sterile phosphate-buffered
Downloaded from http://hmg.oxfordjournals.org/ at CSU Fresno on July 22, 2015
preference was measured as the percentage of time exploring each object, using SMART junior
14 saline (PBS). The hippocampus was dissected and gently dissociated with a fire-polished Pasteur pipette. Cells were plated onto 6-well plates pre-coated with poly-D-lysine (0.1 mg/ml; Sigma-Aldrich) at a density of 50000 cells/cm2. Neurobasal medium supplemented with B27 and Glutamax (Gibco) was used to grow the cells in serum-free conditions. Cell cultures were maintained in an incubator with 5% CO2 at 37°C. Neuronal cells represented more than 95% of the cell population. Media changes were performed weekly and cells were treated after 21 DIV. TNFα (Life Technologies) was added to the culture at a concentration of 10 ng/ml and FTY720P (provided by CHDI Foundation) was used at 10 nM. 8 hours after treatment, cells were washed with cold PBS and lysed for WB analysis.
For immunohistochemical analysis, brains were fixed by inmersion in paraformaldehyde (Sigma-Aldrich) 4% in PBS, cryoprotected in PBS/Sucrose 30% and frozen in methyl-butane (Sigma-Aldrich). 30 µm thick coronal sections of the brain were obtained using a cryostat (Microm) and kept in PBS as free-floating sections. Immunohistochemistry was performed as previously described (Giralt et al., 2010). Tissue was incubated with the following primary antibodies: rabbit anti-p75NTR (1:100, Promega), rabbit anti-Iba1 (Wako, 1:500), rabbit antiNFкB p65 (1:50, Santa Cruz Biotechnology), mouse anti-MAP2 (1:200, Sigma-Aldrich), mouse anti-GFAP (1:500, Sigma-Aldrich), mouse anti-cAMP (1:100, Abcam), followed by the corresponding secondary antibodies. Images were acquired with a Leica SP5 laser scanning confocal microscope (Leica). cAMP immunoreactivity in the CA1 area of the hippocampus was quantified by analysis of integrated optical density with ImageJ software, as previously described (78). Automated quantification of astrocyte number, size and GFAP intensity was performed using Cell Profiler v2.8 software. Briefly, images from the CA1 region were acquired at 10x magnification and Hoescht fluorescence was used in a first segmentation to delimitate CA1 area. Colocalizing pixels from GFAP images were identified with primary object detection modules and further measured to compare relevant properties using appropriate
Downloaded from http://hmg.oxfordjournals.org/ at CSU Fresno on July 22, 2015
Immunohistochemistry
15 modules of this software. The pipeline file containing the specific parameters of the automated quantification is available upon request (http://www.cellprofiler.org/citations.shtml).
Western blot After killing the animals by cervical dislocation, striatum and hippocampus were rapidly removed and homogenized. Protein extraction and western blotting were performed as described elsewhere (35). Blots were incubated with the following antibodies (1:1000, unless stated otherwise): anti-EDG1 (Abcam), anti-BDNF (Santa Cruz Biotechnology), anti-TrkB (Santa Cruz Biotechnology), anti-pTrkB (817) (Abcam), anti-p75NTR (Promega), anti-RhoA (Santa
pCREB (Abcam), anti-NFкB p65 (Santa Cruz Biotechnology), anti-IкBα (1:250, Santa Cruz Biotechnology) and anti-PSD95 (1:2000, Affinity BioReagents). Loading control was performed by reincubating the membranes with either anti--tubulin (1:100000; Sigma-Aldrich) or anti-β-actin (1:20000; Sigma-Aldrich). Membranes were incubated with the corresponding horseradish peroxidase-conjugated antibody (1:2000; Promega). Immunoreactive bands were visualized using the Western Blotting Luminol Reagent (Santa Cruz Biotechnology) and quantified by a computer-assisted densitometer (Gel-Pro Analyzer). Discontinued bands indicate that samples were run in the same gel but were not contiguous.
RNA isolation and quantitative real-time PCR mRNA expression was determined in 20 week-old hippocampal samples from vehicle- and FTY720-treated R6/1 and WT mice. Total RNA was isolated with Tri-Reagent (SigmaAldrich), following product instructions. 2 µg of RNA for each condition were reverse transcribed using PrimeScript RT reagent kit (Takara). cDNA was diluted (1/25) and 2 µl were used to perform qRT-PCR. PrimeTime qRT-PCR assays (see table) were used as recommended by provider (IDT technologies). Rn18s and GAPDH mRNA levels were not altered by treatments and were used as reference genes. qRT-PCR was carried out with Premix Ex Taq (Takara) in 6 µl of final volume using CFX384-C1000 Thermal Cycler equipment (Bio-Rad).
Downloaded from http://hmg.oxfordjournals.org/ at CSU Fresno on July 22, 2015
Cruz Biotechnology), anti-pRhoA (Santa Cruz Biotechnology), anti-CREB (Abcam), anti-
16 Samples were run for 40 cycles (95ºC for 5 sec, 60ºC for 20 sec). Relative gene expression values were calculated with -2∆Ct (88), using Bio-Rad CFX manager software (Bio-Rad). Gene Name GAPDH Rn18s BDNF TNF NOS2
Assay Name Mm.PT.39a.1 Control 18s Mm00432069_m1 Mm.PT.56a.12575861 Mm.PT.56a.14178818
RefSeqNumber NM_008084 NR_003286 NM_001048141 NM_013693 NM_010927
Statistical analysis Statistical analyses were performed as appropriate using GraphPad Prism software and are
considered statistically significant.
ACKNOWLEDGEMENTS We thank Ana López and María Teresa Muñoz for technical assistance, and Teresa Rodrigo and the staff of the Animal Care Facility (Facultat de Psicologia, Universitat de Barcelona) for their help. We are grateful to the Confocal Microscopy Facility (Campus Casanova) of the Centre Científic i Tecnològic (Universitat de Barcelona) for their advice with microscopy techniques. This work was supported by an IDIBAPS Postdoctoral Fellowship-BIOTRACK, supported by the European Community’s Seventh Framework Program (EC FP7/2007-2013) under the grant agreement number 229673 and the Spanish Ministry of Economy and Competitiveness (MINECO) through the grant COFUND2013-40261 to A.M.; Ministerio de Economía y Competitividad [SAF2012-39142 to S.G., RETICS (RD06/0010/0006; Red de Terapia Celular) and SAF2012-37417 to J.M.C., CIBERNED and SAF2011-29507 to J.A]; Cure Huntington’s Disease Initiative (CHDI) to S.G. and J.M.C.; and Generalitat de Catalunya (2014SGR- 968 to J.A.).
CONFLICT OF INTEREST STATEMENT The authors declare that there are no conflicts of interest.
Downloaded from http://hmg.oxfordjournals.org/ at CSU Fresno on July 22, 2015
indicated in the figure legends. All data are expressed as mean ± S.E.M. Values of p
