Neuroinflammation is a common feature of many neurodegenerative diseases, including Huntington's disease (HD). HD is an autosomal dominant genetic disease caused by an expanded CAG repeat in exon 1 of the huntingtin (HTT) gene. Previous studies demonstrated that levels of several proinflammatory cytokines, including tumor necrosis factor (TNF)-α, were higher in the plasma and brain tissues of mice and patients with HD, suggesting that inflammation may contribute to HD progression. To evaluate the pathological role of TNF-α in HD pathogenesis, we blocked TNF-α signaling using a dominant negative inhibitor of soluble TNF-α (XPro1595). XPro1595 effectively suppressed the inflammatory responses of primary astrocytes-enriched culture isolated from a transgenic mouse model (R6/2) and human astrocytes-enriched culture derived from induced pluripotent stem cells (iPSCs) of HD patients evoked by lipopolysaccharide and cytokines, respectively. Moreover, XPro1595 protected the cytokine-induced toxicity of primary R6/2 neurons and human neurons derived from iPSCs of HD patients. To assess the beneficial effect of XPro1595 in vivo, an intracerebroventricular (i.c.v.) infusion was provided with an osmotic minipump. ELISA analyses showed that i.c.v. infusion of XPro1595 decreased elevated levels of TNFα in the cortex and striatum, improved motor function, reduced caspase activation, diminished the amount of mutant HTT aggregates, increased neuronal density and decreased gliosis in brains of R6/2 mice. Moreover, reducing the peripheral inflammatory response by a systemic injection of XPro1595 improved the impaired motor function of R6/2 mice but did not affect caspase activation. Collectively, our findings suggest that an effective and selective anti-inflammatory treatment targeting the abnormal brain inflammatory response is a potential therapeutic strategy for HD.
Neurodegenerative diseases, including Alzheimer's disease (AD), Parkinson's disease (PD), multiple sclerosis (MS) and Huntington's disease (HD), are associated with chronic neuroinflammation and increased levels of various cytokines (1–4). An altered immune profile in HD patients is known to correlate with disease progression (1). Several proinflammatory cytokines, including interleukin (IL)-6, IL-8 and tumor necrosis factor (TNF)-α, are elevated both centrally (in the striatum and cerebrospinal fluid) and peripherally (in the plasma) (1). The adaptive immune responses of monocytes, macrophages and microglia of HD patients are also altered (1, 5). Our recent findings indicate that enhancement of the p65-mediated inflammatory response in HD astrocytes-enriched culture and microglia contributes to HD pathogenesis (6). Inflammation has thus become a potential target for treating neurodegenerative diseases. However, many therapeutic strategies aimed at suppressing inflammation in HD were shown to be ineffective and controversial (7–10).
TNF-α is classified into two forms: transmembrane (tmTNF) and soluble TNF (solTNF) (11). TNF is expressed as a transmembrane protein and exists in a stable homotrimeric form, and tmTNF can be cleaved by the TNF-α-converting enzyme on the cell surface and released as the homotrimeric solTNF (11). TNF elicits various cellular functions through TNF receptors (TNFRs) (12). SolTNF preferentially binds to TNFR1 and mediates responses of apoptosis and chronic inflammation, whereas tmTNF preferentially binds to TNFR2 and maintains the functions of cell survival, anti-inflammation and myelination (13–15). Both TNFR1 and TNFR2 are widely expressed in many tissues, including the brain, under normal and pathological conditions (16, 17). In the central nervous system (CNS), it is well documented that TNF can be synthesized by neurons, astrocytes-enriched culture and microglia (18–20).
The elevation of TNF is a common feature of many neurodegenerative diseases (1–4), and TNF inhibitors are potential drugs for treating neurodegeneration (3, 21, 22). However, in addition to this role in inflammatory responses, TNF signaling is responsible for maintaining a number of cellular physiological responses, including cellular proliferation, survival and differentiation (12). Due to the important and complex functions of TNF, the therapeutic effects of non-selective TNF inhibitors are controversial and often accompanied by side effects, such as an increased risk of granulomatous infections and demyelinating disorders (23–25). In the present study, we employed a solTNF-specific inhibitor (XPro1595) (26) to assess whether TNF plays a critical role in the pathogenesis of HD. XPro1595 is an anti-TNF biologic that forms heterotrimers with solTNF monomers and reduces the binding affinity of the resultant complex to TNFRs without disturbing the signaling and function of tmTNF (27, 28). In a well-characterized mouse model of HD (R6/2, (29)), an inhibitor of solTNF was administered via two different routes, including an intracerebroventricular (i.c.v.) infusion by osmotic minipumps and an intraperitoneal (i.p.) injection. We provide evidence that suggests that administration of XPro1595, particularly in the brain, effectively improved functional outcomes in R6/2 mice.
XPro1595 reduced the hypersensitive inflammatory response of R6/2 glia in vitro
To evaluate the pathological role of TNF-α in HD pathogenesis, we first assessed the anti-inflammatory effect of XPro1595 in vitro using primary glia (astrocytes-enriched culture and microglia) isolated from a well-characterized mouse model [R6/2; (29)] of HD and its littermate control (wild type, WT). XPro1959 was chosen as a TNF-α inhibitor to test our hypothesis because of its selectivity for the proinflammatory solTNF (26). We demonstrated previously that due to abnormal activation of nuclear factor (NF)-κB, both astrocytes-enriched culture and microglia harvested from HD mice are hypersensitive and produce more proinflammatory factors during inflammation (6). Primary astrocytes-enriched culture (days in vitro, DIV 30) were first pretreated with XPro1595 (20 or 200 ng/ml) for 1 day and then stimulated with lipopolysaccharide (LPS) (0.5 μg/ml) for 72 h. As previously shown (6), LPS evoked a higher level of inducible nitric oxide synthase (iNOS) and caused higher production of cytokines from HD astrocytes-enriched culture than from WT astrocytes-enriched culture (Fig. 1). Treatment with XPro1595 dose-dependently suppressed LPS-induced expression of iNOS (Fig. 1A) and cytokines (TNF-α, IL-1β and IL-6; Fig. 1B–D) in both WT and HD astrocytes-enriched culture. Similar results were also observed in primary microglia. XPro1595 significantly decreased the iNOS expression (Fig. 2A) and cytokines production (Fig. 2B–D). The effects of XPro1595 on HD glia were more significant than those on WT glia (Figs 1 and 2). Collectively, these results demonstrated that solTNF plays a critical role in the initial inflammatory response of HD glia.
XPro1595 protected primary neurons from cytokine-induced toxicity in vitro
To evaluate the neuroprotective function of XPro1595, primary cortical neurons (DIV 4) harvested from indicated mice were pretreated with XPro1595 (200 ng/ml) for 1 day, incubated with proinflammatory cytokines for 2 days and fixed for immunostaining with βIII-tubulin (a neuronal marker, red) and activated caspase 3 (an apoptotic marker, green, Fig. 3A and B). We also determined cell death using a TUNEL assay (Fig. 3C) and assessed neuronal survival by a MAP2 assay (Fig. 3D) (30). Similarly, the effects of XPro1595 on HD neurons appeared more significant than those on WT neurons. A possible explanation for this difference is that mutant huntingtin (mHTT) sensitizes the NK-κB/IKK-mediated pathway in both glia and neurons (6, 31) and, thus, causes HD cells to be more sensitive to XPro1595, which targets the TNF-α/NK-κB/IKK pathway. Alternatively, XPro1595 might have additional effects on the mHTT-mediated abnormalities and toxicities.
Infusion of XPro1595 to the brain by a minipump decreased TNF-α levels in brains of R6/2 mice
We next assessed whether XPro1595 blocked the hyper-inflammatory response in brains of HD mice. WT and R6/2 mice were i.c.v. infused with XPro1595 (0.08 mg/kg) at the ages of 7.5–11.5 weeks and were sacrificed at the age of 15 weeks for further analyses. Tissue lysates (serum, liver, cortex and striatum) were collected to determine levels of TNF-α. Infusion of XPro1595 into the brain by a minipump significantly reduced levels of TNF-α in the cortex and striatum (Fig. 4C and D), but not in the serum or liver (Fig. 4A and B), of HD mice compared with saline-infused HD mice.
Infusion of XPro1595 into the brain by a minipump reduced gliosis and enhanced neuronal survival in brains of R6/2 mice
Because neuroinflammation is usually associated with gliosis (32, 33), and gliosis occurs in mice and humans with HD (34, 35), we next assessed the effect of XPro1595 on gliosis of HD mice at the age of 15 weeks. As shown in Figure 5, levels of glial fibrillary acidic protein (GFAP) (an astrocytic marker) were markedly elevated in the cortex and striatum of HD mice, whereas chronic treatment with XPro1595 reduced GFAP levels.
We next assessed the effect of Xpro1595 on microglia. The extent of microglial activation was quantified by measuring the number and size of Iba1-positive microglial cells. Quantitative analyses of Iba1-positive immunoreactivity suggested that chronic treatment with XPro1595 significantly decreased both the number (Fig. 6A and B) and size (Fig. 6C and D) of microglia in the cortex of HD mice. In the striatum of HD mice, although XPro1595 did not markedly affect the number of microglia, it statistically reduced the size of striatal microglia (Fig. 6). Collectively, these observations suggest that a chronic infusion of XPro1595 in the brain ameliorated gliosis of HD mice and might reduce the detrimental role of overactivated glial cells in the brain of R6/2 mice.
Similar to what was observed in WT and HD primary neurons (Fig. 3), XPro1595 treatment in the brain also reduced caspase activation as monitored by SR-FLIVO, which binds activated caspases in vivo as an indication of apoptosis (36), in the cortex and striatum of HD mice (Fig. 7). We next quantified the number of NeuN-positive neurons in brains of HD mice and found that chronic infusion of XPro1595 into the brain rescued the reduced neuronal density in the cortex (Fig. 8A). Although XPro1595 also moderately enhanced the striatal neuronal density in HD mice, the difference between the control and XPro1595-treated groups did not reach statistical significance (Fig. 8B).
Infusion of XPro1595 into the brain by a minipump ameliorates symptoms of R6/2 mice
We next evaluated the effects of XPro1595 on disease progression in mice by measuring rotarod performance, beam walking and foot clasping, which were used as indicators of motor coordination (Fig. 9B–D), and by measuring performance in a T-maze, which was used as an indicator of cognitive function (Fig. 9E). The body weight (BW) and survival were also monitored at 7–15 weeks of age (Fig. 9A and F). Although the decline in the BW of HD mice was not affected (Fig. 9A and F), administration of XPro1595 into the brain of R6/2 mice improved the progressive deterioration of motor coordination, as measured using rotarod performance (Fig. 9B), beam walking (Fig. 9C) and the foot-clasping test (Fig. 9D). In addition, chronic infusion of XPro1595 into the brain of R6/2 mice also improved the decline in cognitive function, as assessed using a T-maze (Fig. 9E). Collectively, our data suggest that chronic infusion of XPro1595 into the brain ameliorates symptoms of R6/2 mice.
Notably, we found that chronic treatment with XPro1595 in the brain also markedly reduced the level of mHTT aggregates. Immunocytochemical staining (Fig. 10A), filter-retardation assays (Fig. 10B) and western blotting (Fig. 10C) using an anti-mHTT antibody (EM48) showed that the amounts of mHTT aggregates in the cortex and striatum of HD mice that were chronically treated with XPro1595 via i.c.v. injection were lower than those of saline-treated HD mice.
An i.p. injection of XPro1595 moderately improved the impaired motor function of R6/2 mice
To assess the importance of peripheral inflammation in HD pathogenesis, we delivered XPro1595 to the peripheral system by an i.p. injection. Mice were injected with XPro1595 (30 mg/kg) or phosphate-buffered saline (PBS) as the control twice per week at the ages of 7–15 weeks. As expected, the 8-week i.p. injections of XPro1595 significantly reduced the level of TNF-α level in the serum and liver (Fig. 11A and B), but not in the cortex or striatum of R6/2 mice (Fig. 11C and D), indicating that an i.p. injection of XPro1595 mainly affected the peripheral system. A slight decrease in BW with an i.p. injection of XPro1595 was noted in HD mice (Fig. 11E). Consistent with the finding that an i.p. injection of XPro1595 did not normalize elevated levels of TNF-α in the cortex or striatum of HD mice, XPro1595 administered via the i.p. route did not improve the reduced caspases activation in either the cortex or striatum of HD mice (Supplementary Material, Fig. S1) nor did it affect the neuronal density in the cortex or striatum of HD mice (Supplementary Material, Fig. S2). In line with this lack of an effect in the brain, an i.p. injection of XPro1595 did not affect the amounts of mHTT aggregates in the cortex and striatum, which were analyzed by immunocytochemical staining (Supplementary Material, Fig. S3A) or filter-retardation assay (Supplementary Material, Fig. S3B) using an anti-mHTT antibody (EM48). Insoluble aggregates retained on the filters or stacking gel were detected using EM48. Nonetheless, we found that an i.p. injection of XPro1595 moderately ameliorated the inferior motor function of HD mice. Although only a limited beneficial effect was observed, this finding suggests that the hyperactive inflammatory response in the peripheral system might somewhat contribute to the pathogenesis of HD.
XPro1595 protected inflammation-induced toxicity in iPSCs derived from skin fibroblast of HD patients
To explore the potential clinical application of XPro1595, we examined the neuroprotective and anti-inflammatory functions of XPro1595 in HD iPSCs-derived astrocytes and neurons, respectively. HD iPSCs-derived neurons were pretreated with XPro1595 (200 ng/ml) for 24 h, then stimulated with TNFα (10 ng/ml) plus IL-1β for 24 h, and fixed for immunostaining with βIII-tubulin (red) and cleaved caspase 3 (green, Fig. 12A, Supplementary Material, Fig. S4A and B). Western blot analysis also showed that treatment with XPro1595 significantly decreased the expression of cleaved caspase 3 in HD iPSCs-derived neurons (Fig. 12B). In addition, HD iPSCs-derived astrocytes were also pretreated with XPro1595 (200 ng/ml) for 24 h and then stimulated with TNFα (10 ng/ml) plus IL-1β for 24 h. Treatment with XPro1595 significantly suppressed cytokine-induced expression of iNOS (Fig. 12C; Supplementary Material, Fig. S4C) in HD iPSCs-derived astrocytes. These results demonstrated that XPro1595 effectively protected both neurons and astrocytes derived from HD iPSCs against cytokine-induced toxicity.
XPro1595 is a selective inhibitor of solTNF (26) and was previously shown to improve functional outcomes in multiple animal models of neuroinflammatory diseases, including MS, PD and AD (28, 37–39). In experimental allergic encephalomyelitis (EAE), XPro1595 decreased production of inflammatory mediators and increased levels of neuroprotection mediators in the spinal cord via an NF-κB-dependent mechanism. Mice with selective deletion of IKKβ (an upstream regulator of NF-κB) in CNS neurons were not protected by XPro1595 in myelin oligodendrocyte protein (MOG)-induced EAE (39). This finding is of particular interest because the IKK/NF-κB pathway is abnormally sensitized in HD and is believed to be responsible for the aberrant activation of immune responses in HD (1, 31, 40). In the present study, we observed that blocking TNF signaling by XPro1595 effectively suppressed the LPS-induced inflammation of HD astrocytes-enriched culture and microglia (Figs 1 and 2) and protected HD neurons from cytokine-induced toxicity (Fig. 3). Infusion of XPro1595 into the brain using an osmotic minipump decreased elevated levels of TNFα in the cortex and striatum of R6/2 mice (Fig. 4), improved motor functions (Fig. 9B–D), reduced caspase activation (Fig. 7), diminished the amount of mHTT aggregates (Fig. 10), increased neuronal density (Fig. 8) and decreased gliosis (Figs 5 and 6) in brains of R6/2 mice. In addition, XPro1595 showed an obvious beneficial effect in protecting HD iPSCs from cytokine-induced toxicity (Fig. 12; Supplementary Material, Fig. S4). An effective and selective anti-inflammatory treatment targeting the abnormal brain inflammatory response is thus a potential therapeutic strategy for HD. Although such i.c.v. infusion of XPro1595 did not decrease body weight loss or increase the life span of HD mice, XPro1595 did improve deteriorating motor and cognitive activities in these mice (Fig. 9). Given that motor and cognitive function impairments greatly impact the life quality of patients (41), XPro1595 (or other anti-inflammatory agents) could be an attractive therapeutic agent for a devastating disease that disables patients almost entirely in its late stage.
Our results show that infusion of XPro1595 into the brain reduced the elevated inflammatory response in the brain (Fig. 4), whereas systemic administration of XPro1595 by an i.p. injection only normalized the abnormal inflammatory responses in peripheral tissues (Fig. 11). No effect of XPro1595 on the behaviors of the HD mice was observed (Fig. 11E and F; Supplementary Material, Fig. S3). These data suggest that XPro1595 does not readily pass through the Blood–brain barrier (BBB) of HD mice. However, it is surprising to find that chronic i.p. injection of XPro1595 elevated the TNF-α level in the cortex of the HD mice (Fig. 11C). We assessed the amount of XPro1595 in mouse brains to evaluate whether i.p. injected XPro1595 could penetrate the BBB. The levels of XPro1595 were measured using a quantitative human TNF ELISA assay that specifically detects human XPro1595 protein without cross-reacting with mouse TNF-α (38). Our results showed that i.p. injection of XPro1595 into mice for 1–3 h resulted in the existence of XPro1595 in the serum and liver but not in the cortex and striatum, suggesting that Xpro1595 cannot effectively pass through the BBB in WT mice (Supplementary Material, Fig. S5). Consistent with the above observation, XPro1595 was detected in the cortex and striatum of mice that were chronically treated with XPro1595 via i.c.v. injection (Supplementary Material, Fig. S6C and D), while no XPro1595 was found in the brain of mice that were chronically treated with XPro1595 via i.p. injection (Supplementary Material, Fig. S6A and B). Taken together, these data suggest that, although XPro1595 itself might not readily pass through the BBB, its metabolic product(s) may.
Our data also showed that an i.p. injection of XPro1595 further decreased the BW of R6/2 mice (Fig. 11E), whereas an i.c.v. infusion of a dominant negative inhibitor of soluble TNF-α (XPro1595) did not affect the BW (Fig. 9A). These effects could all be a function of dose because 30 mg/kg (used in i.c.v. infusion), versus 0.08 mg/kg (used in i.p. injection), could have very different effects. Additionally, R6/2 mice are fragile and may be more susceptible to frequent injections. Because HD patients commonly suffer from weight loss, such an effect of XPro1595 administered through the i.p. route might worsen the energy dysfunction of HD. Although XPro1595 administered to the CNS and peripheral tissues both improved motor functions (Figs 9 and 11), i.c.v. delivery of XPro1595 appeared to be safer and more effective for treating HD. Such direct brain administration of XPro1595 (or other anti-inflammatory agents) is feasible using devises that are currently available in clinics (such as intrathecal pumps or intraventricular catheters) and might be considered for future clinical application. In patients with other degenerative diseases where the BBB permeability is enhanced (e.g. PD) (42, 43), the peripheral administration of XPro1595 might be sufficient to produce a beneficial effect.
Our finding that blocking abnormal peripheral inflammation by an i.p. injection of XPro1595 ameliorated the motor impairment of HD mice (Fig. 11) is consistent with the results of a recent study suggesting that the expression of mHTT fragments occurs in immune cells and can be used as a non-invasive biomarker for HD (44). In addition, recent studies showed that replacing diseased bone marrow cells with healthy WT cells or blockage of a hyper-reactive immune system moderately delayed disease progression in HD mice (45, 46). Therefore, the abnormal peripheral immune system in HD is a recognized, but relatively minor, cause of HD pathology.
Many beneficial treatments for HD are associated with reduced levels of TNF-α. For example, treatment with a peroxisome proliferator-activated receptorγ activator (pioglitazone) lowered the TNF-α level and protected neurons from quinolinic acid-induced neurotoxicity in rats (47). Chronic treatment with a cyclooxygenase-2 inhibitor (celecoxib) attenuated TNF-α, IL-6 and caspase-3 levels in quinolinic acid-induced experimental models of HD (48). The beneficial effects of a tetracycline antibiotic (minocycline) on HD were associated with inhibition of microglial activation, attenuation of apoptosis and suppression of inflammation (49–52). To date, the effects of minocycline on HD remain controversial and require further investigation (53, 54). One possible reason for this controversy may be due to the off-target actions of anti-inflammatory drugs used in different studies. Given the importance and complex regulation of TNF-α, overall inhibition of TNF-α signaling in the body is expected to disturb the protective functions of TNFR2 signaling mediated by tmTNF (39). This effect is a critical issue because activation of TNFR2 provides neuroprotection under many pathological conditions, such as ischemia reperfusion and glutamate- and β-amyloid-induced cytotoxicity (55–57). In treating EAE, XPro1595 exerted an anti-inflammatory function by selectively blocking solTNF signaling without impairing the neuroprotective ability of tmTNF (37, 39). In HD, we found that XPro1595 exhibits anti-inflammatory and neuroprotective effects and delays disease progression. An effective and selective anti-inflammatory treatment targeted at abnormal solTNF-mediated brain inflammatory responses may therefore be a good therapeutic strategy for HD.
MATERIALS AND METHODS
LPS prepared from Escherichia coli serotype O111:B4 was purchased from Sigma–Aldrich (St Louis, MO, USA). One nanogram of endotoxin is equivalent to 5 endotoxin units (Limulus lysate assay) (58). Mouse-TNF-α and mouse- IL-1β were purchased from Sigma (St Louis, MO, USA). An anti-cleaved caspase 3 antibody was purchased from Cell Signaling (Danvers, MA, USA). An anti-GFAP antibody and an anti-βIII-tubulin antibody were obtained from Merck Millipore (Billerica, MA, USA). An anti-Iba1 antibody was purchased from Wako Pure Chemical Industries (Chuo-Ku, Osaka, Japan). An anti-iNOS antibody was purchased from BD Biosciences (San Jose, CA, USA). SR-FLIVO was obtained from Immunochemistry Technologies (Bloomington, MN, USA). XPro1595 was a generous gift from Xencor (Monrovia, CA, USA).
Primary astrocytes-enriched culture were isolated from the cortex of postnatal 1- to 2-day-old R6/2 mice and their littermate controls as previously described (59). Briefly, the cortices were washed three times with Hank's balanced salt solution buffer (5.33 mm KCl, 0.441 mm KH2PO4, 4.17 mm NaHCO3, 137.9 mm NaCl and 0.338 Na2HPO4) and digested with 0.05% trypsin–Ethylenediaminetetraacetic acid (EDTA) (Invitrogen, Grand Island, NY, USA) at 37°C for 10 min. After removing the trypsin, cortical tissues were then mechanically dissociated by gentle pipetting in Dulbecco's modified Eagle medium (DMEM; Invitrogen) supplemented with 10% (v/v) heat-inactivated fetal bovine serum (FBS) and 1% penicillin/streptomycin and passed through a 70-μm-pore nylon mesh. Cells were plated onto poly-l-lysine-coated dishes and grown in DMEM (supplemented with 10%, v/v, heat-inactivated FBS and 1% penicillin/streptomycin) at 37°C in a humidified 5% CO2-containing atmosphere. The purity of the primary astrocyte cultures was determined by immunocytochemical staining using an antibody against an astrocyte-specific marker (GFAP, dilution 1: 1000; Sigma) or a microglia-specific marker (anti-CD11b, dilution 1: 200; Serotec, Oxford, UK). At 30 DIV, most of the primary astrocytes-enriched cultures were GFAP positive. No detectable Iba1-positive cells (i.e. microglia) were found.
Primary microglia were detached from the astrocytic monolayer at 14 DIV by gentle agitation based on the distinct adhesive properties of microglia and astrocytes-enriched culture (6) and grown in DMEM (Invitrogen) supplemented with 10% FBS and 1% penicillin/streptomycin at 37°C in a humidified 5% CO2-containing atmosphere. Approximately 99% of the primary microglia were Iba1 positive, which was determined by immunocytochemical staining.
Primary neuronal cultures were cultured as previously described (59). Briefly, the cortices were purified from R6/2 mice and their littermate controls on embryonic day 17 (E17) to E18 and digested with 0.25% trypsin–EDTA for 10 min at 37°C. After removing the trypsin, cortical tissues were mechanically dissociated by gentle pipetting in modified Eagle's medium supplemented with 5% (v/v) FBS, 5% (v/v) horse serum, 0.6% (v/v) glucose, 0.5 mm glutamine, 1% penicillin/streptomycin and 1% insulin–transferrin–sodium selenite media supplement (ITS mixture; Sigma). Cells were plated on poly-l-lysine-coated coverslips. The culture medium was replaced with neurobasal medium (supplemented with 0.5 mm glutamine, 12.5 μm glutamate, 2% B27 and 1% penicillin/streptomycin) after 2 h of attachment. The purity of the neuronal cultures was determined by immunocytochemical staining using an antibody against a neuron-specific marker and class III β-tubulin (diluted 1: 1000; Promega, Madison, WI, USA). Approximately 90% of the primary neurons were βIII-tubulin positive.
Human-induced pluripotent stem cells (iPSCs) were prepared from two HD patients with 43 CAG repeats as described previously (60). Patient A is 52-year-old female, and patient B is 41-year-old male. Briefly, fibroblasts were reprogrammed using a combination of VSV-G-coated retroviruses expressing human OCT3/4, NANOG, SOX2, KLF4 and MYC. Genomic DNA and RNA were extracted from HD iPSCs clones and polymerase chain reaction (PCR) analysis was performed to confirm the integration of retroviral transgenes. RT–PCR was performed to analyzed the expression of viral transgenes of OCT4, SOX2, KLF4 and MYC. Embryoid bodies generated from HD iPSCs were differentiated into neural stem cells (NSCs) in the neural proliferation medium (DMEM/F12 supplemented with 1× N2, 2 mml-glutamine, 100 U/ml penicillin, 100 μg/ml streptomycin and 25 ng/ml bFGF). NSCs were further differentiated into striatal neurons by replacing the neural proliferation medium to the neural differentiation medium I (neural proliferation medium: neural differentiation medium II = 1 : 1) for 4 weeks, followed by the neural differentiation medium II (neural-basal medium supplemented with 1× B27, 2 mml-glutamine, 100 U/ml penicillin and 100 μg/ml streptomycin) for another 4–6 weeks. HD iPSCs-derived neurons were dissociated, replated on the Matrigel®-coated coverslips or plates and incubated in the neural differentiation medium II for 3 days for further experiments. To differentiate HD-iPSC into astrocytes-enriched culture, HD iPSCs were maintained in the neural differentiation medium II supplemented with ciliary neurotrophic factor (20 ng/ml, R&D Systems, Minneapolis, MN, USA) for 4 weeks. The purity of HD iPSCs-derived astrocytes-enriched culture was determined by immunocytochemical staining using an antibody against GFAP. Nearly 95% of the differentiated cells were GFAP positive (Supplementary Material, Fig. S7).
Sodium dodecylsulfate–polyacrylamide gel electrophoresis and western blotting
Protein concentrations were determined using the Bio-Rad Protein Assay Dye Reagent Concentrate (Bio-Rad, Hercules, CA, USA). Equal amounts of protein were heat denatured in sample treatment buffer by boiling for 5 min and resolved on 10% sodium dodecylsulfate (SDS)– polyacrylamide gel electrophoresis (PAGE). After electrophoresis, gels were transferred onto nitrocellulose membranes, blocked by 5% bovine serum albumin (BSA) in Tris-buffered saline and tween 20 (TBST) (0.2 m Tris–base, 1.37 m NaCl and 0.05% Tween 20), and incubated with the indicated primary antibody at 4°C overnight. After washing three times with TBST, membranes were incubated with a peroxidase-conjugated secondary antibody for 1 h at room temperature (RT). After washing three times with TBST, immunoreactive bands were stained using a light-emitting nonradioactive method [enhanced chemiluminescence (ECL); PerkinElmer, Waltham, MA, USA] and recorded using X-ray film (Kodak, Rochester, NY, USA).
Enzyme-linked immunosorbent assay
The levels of TNFα, IL-1β and IL-6 were determined using a mouse TNFα, mouse IL-1β and mouse IL-6 DuoSet ELISA Development System (R&D Systems, Minneapolis, MN, USA) following the manufacturer's protocol. Briefly, 96-well microplates were coated overnight with the capture antibody at RT and blocked with 1% BSA in PBS. After washing three times with wash buffer, samples or standards (prepared in Reagent Diluent) were added to the microplates and incubated for 2 h at RT. The microplates were then incubated with the detection antibody for 2 h at RT after washing three times with the wash buffer, followed by a 30-min incubation with streptavidin-horseradish peroxidase (HRP) plus substrate. The optical density at 450 nm was detected using a microplate reader (ELISA Reader: spectraMAX340PC; Molecular Devices, Union City, CA, USA). Concentrations of cytokines were calculated based on the standard curve performed in the same experiment.
Mice were anesthetized by sodium pentobarbital (i.p. injection, 80 mg/kg) before perfusion with 4% paraformaldehyde in 0.1 m phosphate buffer (Na–PB, pH 7.4). Brains were carefully isolated after perfusion, postfixed in 4% paraformaldehyde for 24 h, and immersed in 30% sucrose for 2 days. Serial coronal sections (20 μm) were immunohistochemically stained as previously described (36). Briefly, brain sections were permeabilized with 0.1% Triton X-100 and blocked with 2% normal goat serum in Na–PB for 2 h. After washing, sections were incubated with the indicated primary antibody in Na–PB for 2 days at 4°C and then incubated with the corresponding secondary antibody for 2 h at RT. Nuclei were stained with Hoechst 33258. Slides were mounted with Mounting Media (Vector Laboratories, Burlingame, CA, USA).
Cells were fixed with 4% paraformaldehyde plus 4% sucrose in PBS (pH 7.4) at RT for 30 min and permeabilized with 0.1% Triton X-100 at RT for 15 min. Samples were blocked by incubating cells with 2% normal goat serum plus 2% BSA in PBS for 1 h at RT to decrease nonspecific binding. After blocking, samples were incubated with the indicated primary antibody at 4°C overnight, followed by incubation with the corresponding Alexa secondary antibody for 2 h at RT. Nuclei were stained with Hoechst 33258. Patterns of immunostaining were analyzed with the aid of ImageJ software (National Institutes of Health, Washington DC, USA) and a CCD microscope (Zeiss, Göttingen, Germany) or a confocal microscope LSM 780 (Zeiss).
The TUNEL assay was determined using the DeadEndTM Fluorometric TUNEL System and conducted following the manufacturer's protocol (Promega). Briefly, DIV-4 neuronal cultures were plated on poly-l-lysine-coated coverslips and stimulated with the indicated cytokines for 3 days. Slides were fixed with 4% methanol-free formaldehyde in PBS and then permeabilized with 0.1% Triton X-100 in PBS. After pre-equilibration, slides were labeled for DNA strand breaks with fluorescein-12-dUTP in the incubation buffer, which contained equilibration buffer, the nucleotide mix and the rTdT enzyme, at 37°C for 1 h. The reaction was terminated by dipping the slides in 2× saline sodium citrate buffer (0.3 m NaCl and 34.2 mm sodium citrate, pH 7). The localized green fluorescence of apoptotic cells was detected by a confocal microscope (LSM780).
The MAP2 assay was performed as described elsewhere (30). Briefly, cells were fixed with 4% paraformaldehyde and 4% sucrose in PBS and incubated with an anti-MAP2 antibody (1: 1000, Sigma) at 4°C overnight. After washing three times with PBS, cells were incubated with HRP-conjugated mouse immunoglobulin G (1: 5000, PerkinElmer, Boston, MA, USA) in PBS containing 1% BSA for 30 min at RT. After washing three times with PBS, cells were incubated with the substrate solution (50 μm 10-acetyl-3,7-dihydroxyphenoxazine and 200 μm hydrogen peroxide in 50 mm sodium phosphate buffer) for 30 min at RT. After reacting, resorufin fluorescence in the mixture was detected using a fluorescence reader (with excitation at 550 nm and emission at 590 nm; Molecular Devices).
Animals and treatments
Male R6/2 mice and littermate controls were initially obtained from Jackson Laboratory (Bar Harbor, ME, USA) and mated to female control mice (B6CBAFI/J). Offspring was verified by PCR genotyping of genomic DNA extracted from tail tissues using primers (5′-CCGCTCAGGTTCTGCTTTTA-3′ and 5′-GGCTGAGGAAGCTGAGGAG-3′) located in the transgene. The number of CAG repeats of R6/2 mice used in the present studies was 239 ± 7.8 (mean ± SEM, n = 60). Mice were housed at the Institute of Biomedical Sciences Animal Care Facility (Taipei, Taiwan) under a 12-h light–dark cycle. Animal experiments were performed under protocols approved by the Academia Sinica Institutional Animal Care and Utilization Committee (Taipei, Taiwan).
For i.c.v. delivery of XPro1595 (Xencor), mice were anesthetized with ketamine/xylazine HCl (87.56 and 7.5 mg/kg; i.p.) and implanted with osmotic minipumps (micro-osmotic pump 2004, brain infusion kit 3, Alzet, Cupertino, CA, USA) which delivered XPro1595 or saline at a rate of 0.08 mg/kg/day into the lateral ventricle for 28 days starting at the age of 7 weeks. The dorsoventral coordinates were AP −0.6 mm, lateral + 1 mm, and dorsoventral −2 mm relative to the bregma and dural surface. For the i.p. injection of XPro1595, mice were injected with 30 mg/kg twice weekly at the ages of 7–15 weeks.
BWs of mice were recorded twice weekly. Rotarod performance was measured three times per week at the ages of 7–15 weeks to assess motor coordination using a rotarod apparatus (UGO BASILE, Comerio, Italy) at a constant speed (12 rpm) as previously described (61). Mice were assessed in three trials for a maximum of 2 min for each trial. Animal experiments were performed under protocols approved by the Academia Sinica Institutional Animal Care and Utilization Committee, Taiwan. Beam walk analysis was conducted as described in previous studies (62) and used to assess motor coordination. Briefly, mice were first trained on a 17-mm-diameter beam one time, followed by a weekly test on an 11-mm-diameter beam. The latency of a mouse in walking across the 11-mm-diameter beam was recorded once weekly. To quantify limb clasping (63), mice were suspended by the tail for 30 s. Limb clasping was determined using the following scoring rule: 0, no clasping; 1, clasping by either the fore- or hind limbs; and 2, clasping by both the fore- and hind limbs. T-maze analysis (spontaneous alteration) was used to assess the cognitive function of the animals and was conducted as described in previous studies (6, 64). Briefly, each mouse was placed at the start area of the T-maze and allowed to walk to the intersection, where it could choose the right or left arm. When a mouse is subjected to T-maze analysis repeatedly after an initial trial, it usually explores the arm opposite to the previous selection. Each animal was tested for 10 consecutive trials. The number of alternative choices out of 10 trials was recorded and presented as the ratio of correct choices for each animal.
In vivo apoptosis assay
The SR-FLIVO in vivo apoptosis assay kit (ImmunoChemistry Technologies, LLC, Bloominton, MN, USA) was used to label apoptotic cells by forming covalent bonds with active caspases in vivo (65). SR-FLIVO was intravenously injected into the lateral tail vein of each mouse. Sixty minutes after the injection, animals were subjected to a fixation procedure using 4% paraformaldehyde as previously described (36). The brain was carefully removed and used to prepare serial 20-μm-thick coronal sections to analyze activation of caspases by detecting the red signal of SR-FLIVO using a laser confocal microscope (LSM780; Zeiss, Göttingen, Germany) with excitation at 565 nm and emission at 600 nm. Quantitative analysis of caspase activation was conducted using ImageJ software. The total intensity of red fluorescence representing activated caspases was quantified in different brain areas, including the cortex and striatum.
Filter assays were performed as previously described (61). Briefly, brain tissues were lysed in RIPA buffer (50 mm Tris–HCl, 0.25% sodium deoxycholate, 1% Triton X-100, 150 mm NaCl, 1× protease inhibitor and 1× phosphatase inhibitor) on ice. Protein concentrations were determined using the Bio-Rad Protein Assay Dye Reagent Concentrate. Samples were prepared with 2% SDS in PBS and loaded onto OE66 membrane filters (0.2-μm pore size; GE Healthcare, Chalfont St Giles, Buckinghamshire, United Kingdom) through a slot-blot manifold (Bio-Rad Laboratories). Membranes were blocked with 5% skim milk in PBS and incubated overnight with an anti-HTT antibody (EM48; Millipore Corporation, Billerica, MA, USA) at 4°C. After washing three times with TBST, membranes were incubated with the corresponding secondary antibody for 1 h at RT. The immunoreactive bands were detected by ECL.
This work was supported by grants from the National Science Council (NSC97-2321-B-001-030, NSC98-2321-B-001-017, NSC99-2321-B-001-012, NSC100-2321-B-001-009 and NSC101-2321-B-001-047) and Academia Sinica (AS-100-TP2-B02), Taiwan.
We thank David E. Szymkowski and Xencor (Monrovia CA, USA) for providing XPro1595. We are grateful to Daniel P. Chamberlin for editing the manuscript and Ms Hsing-Lin Lai for technical support.
Conflict of Interest statement. None declared.