Brain-specific NRSF deficiency aggravates dopaminergic neurodegeneration and impairs neurogenesis in the MPTP mouse model of Parkinson’s disease

Brain-specific NRSF deletion potentiates PD-associated behavioral deficits in mice challenged with MPTP

We previously showed that injuries in the nigrostriatal pathway induced by the neurotoxin MPTP were more severe in neuronal NRSF deficient mice than in WT mice [19, 20]. In this study, we extend those findings through a comprehensive evaluation of neuronal and behavioral alterations resulting from acute MPTP treatment in cKO mice. Mice with lox P-flanked alleles of NRSF (NRSF^fl/fl) were bred to Nestin-Cre transgenic mice to induce deletion of NRSF in the brain [23, 24]. The expected NRSF fragment lacking exon 2 was detected by PCR in the cortex, hippocampus, and striatum of Nestin-Cre:NRSF^fl/f mice (cKO mice), but not in littermate NRSF^fl/fl mice (WT) (Supplementary Figure 1A). Brain-specific ablation of NRSF in cKO mice was confirmed at the transcriptional and translational level. A reduction in NRSF transcripts was detected in the brain but not in peripheral tissues of cKO mice by quantitative real-time PCR (Supplementary Figure 1B), while double immunofluorescence against NRSF and NeuN further confirmed NRSF depletion in the brains of cKO mice (Supplementary Figure 1C). Of note, a striking reduction of body weight was observed in cKO mice compared with WT littermates (Supplementary Figure 2).

Nissl staining revealed no obvious morphological abnormalities in the brains of cKO mice compared to WT littermates (Supplementary Figure 3). Striatal levels of amino acid neurotransmitters were next characterized by HPLC. Results showed increased levels of GABA in cKO mice (Supplementary Figure 4A), whereas glycine, aspartate, asparagine, glutamate, and glutamine levels did not differ from WT mice (Supplementary Figure 4B–4F). No differences regarding locomotor ability and motor coordination were detected between genotypes in rotarod and open field tests (Supplementary Figure 4G–4I).

To assess the impact of brain-specific NRSF silencing on the behavioral, cellular, and molecular alterations induced by MPTP, 12-16 weeks old cKO and WT control mice received 4 intraperitoneal injections of either 20 mg/kg MPTP-HCl or saline solution, spaced by 2 h intervals. Early and late PD-associated behavioral deficits were assessed by applying the pole and the wire hanging tests 7 and 42 days after MPTP administration. Pole test results showed obvious motor dysfunction in both WT and cKO mice exposed to MPTP, and this deficit was significantly enhanced in cKO mice (Figure 1A, 1B). In the wire hanging test, MPTP-treated WT and cKO mice both showed shorter fall latency time than saline-treated controls; here again, cKO mice challenged with MPTP scored also significantly lower than MPTP-challenged WT mice (Figure 1C, 1D). To assess whether the behavioral impairment was enduring, mice were tested 42 days after MPTP injection. Results showed that motor deficits were still present in all MPTP-intoxicated animals (Figure 1E), and were again more severe in cKO mice (Figure 1F). Consistently, MPTP-treated cKO mice also scored significantly worse than both MPTP-treated WT and saline-treated mice (Figure 1G, 1H).

MPTP-induced neurodegeneration is aggravated by NRSF deletion in brain cells

In previous studies we found that the number of TH- positive (+) neurons in the SN and the density of TH (+) nerve fibers in the striatum were dramatically higher in neuron-specific NRSF knockout mice compared to WT mice [19, 20]. These phenotypes reappeared in the brain of the cKO mice used in the present study, as determined by both immunohistochemistry and stereological cell counting (Figure 2). Damage to the nigrostriatal axis, including reductions in the number of TH (+) neurons in the SNpc, striatal TH protein levels, and density of TH (+) nerve fibers, were detected in both WT and cKO mice 7 days after MPTP administration (Figure 2). However, reductions in both total and dopaminergic neuron numbers (Figure 2A) and striatal TH protein levels (p=0.012, vs MPTP-treated WT mice, by unpaired t test) (Figure 2B) were more severe in MPTP-treated cKO mice. Meanwhile, HPLC analyses of dopamine (DA) and its metabolites DOPAC and HVA showed higher striatal concentrations of DA and DOPAC in cKO compared to WT mice (Table 1). After MPTP injection, striatal levels of DA, DOPAC and HVA were reduced to 24.8%, 63.3% and 75.8% of respective baseline levels in WT mice, while more marked decreases to 9.4%, 29.8% and 47.2% respectively were measured in cKO mice. In addition, the DOPAC to DA ratio was higher in both WT and cKO mice challenged with MPTP than in saline-treated control mice. Meanwhile, in MPTP-treated cKO mice HVA levels and the HVA to DA ratio were significantly decreased and increased, respectively, compared to saline-treated cKO mice. In turn, the ratio of HVA to DA was significantly increased in MPTP-treated cKO mice compared to MPTP-treated WT mice (Table 1).

Late MPTP effects were analyzed at 42 days post-injection. At this time point, sustained damage to the dopaminergic system was detected in both cKO and WT mice (Figure 3). However, compared with the results at 7 days post-injection, WT mice exhibited significant increases in dopaminergic cell numbers (4,541 ± 91 vs 5,536 ± 248, p = 0.002), striatal TH protein levels (50.04% ± 4.03% vs 60.82% ± 14.9%, p = 0.046) and TH (+) nerve terminals (39.97% ± 1.71% vs 52.04% ± 2.04%, p = 0.004). cKO mice also showed partial but significant upregulation in dopaminergic cell numbers (3,028 ± 98 vs 4,014 ± 182, p = 0.001) and striatal TH protein levels (33.83% ± 2.1% vs 45.28% ± 2.3%, p = 0.01). Thus, although there was a recovery of dopaminergic neurons in cKO mice, loss of both total and dopaminergic neurons, reductions in striatal TH protein levels and TH (+) fiber density (p=0.0013 and p=0.0011, respectively, vs MPTP-treated WT mice, by unpaired t test) remained higher (Figure 3A–3C). These results suggest that despite having higher DA levels in the striatum and more dopaminergic neurons in the SN under baseline conditions, cKO mice are more sensitive to the toxicity of MPTP.

NRSF deficiency exacerbates astrocyte activation in mice challenged with MPTP

Microglia and astrocytes are key modulators of the inflammatory responses that contribute to the progression of PD [25]. Using immunostaining and western blot, we investigated whether NRSF affects glial activation in the MPTP model of PD. One day after MPTP injection, microglial cells in the nigrostriatal pathway were activated in both WT and cKO mice, with no significant differences observed between the two genotypes (Supplementary Figure 5). Meanwhile, higher GFAP expression and more GFAP (+) astrocytes were detected in the nigrostriatal system of untreated cKO mice relative to WT (Figure 4). Seven and 42 days after MPTP injection, astrocytes were obviously activated in the striatum and SN of both WT and cKO mice. Notably, astrocyte activation was more intense in cKO mice, indicated by dramatically elevated levels of striatal GFAP (Figure 4A, 4C), and increased numbers of GFAP (+) astrocytes in both the striatum (Supplementary Figure 6) and the SN (Figure 4B, 4D).

Chronically activated astrocytes are believed to contribute to PD through production of various pro-inflammatory molecules including tumor necrosis factor alpha (TNF-α), inducible nitric oxide synthase (iNOS), interleukin-1α (IL-1α), IL-1β and IL-6 [26]. Using quantitative real-time PCR, we found that the expression of TNF-α and IL-1α in the striatum was increased in MPTP-treated cKO mice compared to MPTP-treated WT mice (p=0.088 and p=0.035, respectively) (Figure 5A, 5C). Striatal transcripts of iNOS and IL-1α were also elevated in MPTP-treated cKO mice compared to their saline-treated counterparts (p=0.062 and p=0.068, respectively) (Figure 5B, 5C). Meanwhile, no changes were observed in IL-1β and IL-6 expression in mice treated with MPTP (Figure 5D, 5E). These results suggest that genetic ablation of NRSF triggered more severe neuroinflammation in the MPTP-injured nigrostriatal pathway.

The apoptosis-related genes Bax, Bid, Mapk11 and Fas are NRSF/REST targets [27], and their transcript levels were comparable between untreated cKO mice and WT littermates. However, upregulated striatal expression of Bax, Bid and Mapk11, but not Fas, was observed after MPTP treatment in both WT and cKO mice, and this effect was higher in the latter genotype (Figure 5F–5I). On the other hand, lower BDNF protein levels were found in the striatum of MPTP-treated cKO mice compared to saline-treated cKO counterparts (p=0.087) (Figure 5J).

Brain-specific NRSF deletion transiently accelerates neurogenesis after exposure to MPTP

The presence of adult neurogenesis in PD is much debated. To investigate the impact of NRSF deficiency on neurogenesis in the MPTP mouse model of PD, BrdU was injected over a 5-day period beginning on day 3 following MPTP treatment, and then sacrificed. Microscopic examination showed that the number of BrdU (+) cells in the SVZ was significantly higher in saline-treated cKO mice than in WT mice (Figure 6A). MPTP treatment further increased the number of BrdU (+) cells in the SVZ of both WT and cKO mice, but the increase was much larger in the latter (Figure 6A). To investigate the identity of BrdU (+) cells, parallel detection of the neural stem cell markers nestin and Sox2, and GFAP, was carried out by immunofluorescence (Figure 6B–6D). To confirm the above results, a second regimen of BrdU incorporation was adopted. BrdU was injected four times daily at 2 h intervals 7 days after the last MPTP injection, and mice were sacrificed for BrdU expression analysis 2 h later. MPTP insult increased the incorporation of BrdU into cells in the SVZ (Supplementary Figure 7A). BrdU (+) cells in the SVZ compartment showed nestin, Sox2, or GFAP immunoreactivity (Supplementary Figure 7B–7D).

NRSF knockout decreases long-term survival of proliferating neural stem/progenitor cells and severely impairs adult neurogenesis

Next, we investigated if NRSF played a role in the survival of newborn cells in the SVZ of the PD mouse brain. Beginning on day 3 following the last MPTP injection, mice were injected with BrdU over a 5-day period and sacrificed 42 days after MPTP treatment. We observed that the number of BrdU (+) cells in the SVZ of saline-treated cKO mice were reduced compared to saline-treated WT mice (p=0.084, by unpaired T test). MPTP treatment further reduced the number of BrdU (+) cells in both genotypes, although the effect was larger in cKO mice (Figure 7A). In addition, BrdU (+) cells in the SVZ showed GFAP immunoreactivity (Figure 7B).

Animal models

All experimental protocols were approved by the Institutional Animal Care and Use Committee of Fudan University, Shanghai Medical College (IACUC Animal Project Number: 20110307-055). All surgeries were performed under general anesthesia, and all efforts were made to minimize suffering.

WT mice and brain-specific conditional NRSF knockout (Nestin-Cre:NRSF^fl/fl, cKO) mice (12 to 16 weeks old) were obtained from Shanghai Model Organisms Center, INC. and maintained in a temperature-controlled room under a 12 h light/ dark cycle with ad libitum access to food and water. All animal experiments were performed in accordance with the Guidelines of the Fudan University Animal Care and Use Committee.

MPTP lesion

Both WT and cKO mice were divided into MPTP-treated groups and saline-treated groups. Mice in MPTP-treated groups received four intraperitoneal injections of 20 mg/kg MPTP-HCl at 2 h intervals in a single day, while control mice were injected with equal volumes of normal saline. Mice were sacrificed 1 day, 7 days or 42 days later.

Neurogenesis studies

To assess neurogenesis, three intraperitoneal BrdU injection regimens were adopted: 1) 3 days after MPTP/saline treatment, BrdU (20 mg/kg) was injected twice daily for 5 days and then sacrificed; 2) 7 days after MPTP/saline treatment 4 BrdU (50 mg/kg) injections were applied at 2 h intervals, and mice were sacrificed 2 h later; and 3) 3 days after MPTP/saline treatment, BrdU (20 mg/kg) was injected twice daily for 5 days, and mice sacrificed 34 days later (i.e. 42 days after MPTP/saline treatment).

Western blot analysis

Mouse tissues were lysed in T-PER reagent with a protease inhibitor cocktail (Thermo Scientific). Equal amounts of proteins were fractionated by SDS-PAGE, transferred to polyvinylidene fluoride (PVDF) membranes. Blots were probed overnight at 4°C with the following primary antibodies: rabbit anti-TH (1:4000, Sigma), mouse anti-β-actin (1:2000, Abgent), rabbit anti-GFAP (1:4000, Millipore), mouse anti-GAPDH (1:5000, Santa Cruz). Next, the membranes were incubated for 1 h with IRDye 680LT goat anti-mouse IgG (H+L) (926-68020, Li-Cor, USA) or IRDye 800CW goat anti-rabbit IgG (H+L) (926-32211, Li-Cor, USA) (1:10,000 dilutions in TBST with 0.02% SDS) and protein bands detected by an Odyssey infrared imaging system (Li-Cor, USA). Protein levels were quantified by densitometry using Quantity One 4.5.2 software (Bio-Rad, Hercules, USA).

Immunohistochemistry and immunofluorescence

Mouse brains were fixed in 4% PFA at 4 °C overnight and cryosectioned at a thickness of 30 µm. The sections were rehydrated, blocked by 10% goat serum, and incubated with primary antibodies: mouse anti-tyrosine hydroxylase (1:1000, Sigma), mouse anti-nestin (1:500, Abcam), rabbit anti-Sox2 (1:500, Abcam), rabbit anti-Iba1 (1:500, Wako), rabbit anti-GFAP (1:1000, Millipore), and rabbit anti-BrdU (1:500, Abcam). For immunohistochemistry staining, the sections were washed and incubated with biotinylated anti-mouse or anti-rabbit secondary antibodies (1:200, Vector Laboratories, USA), and then with AB peroxidase (1:200, Vector Laboratories, USA). The peroxidase reaction was detected with 0.05% DAB (Sigma, USA) in 0.1 M Tris-HCl buffer and 0.03% H₂O₂. For immunofluorescence staining, the sections were then washed and labeled with fluorescent secondary antibodies (1:500, Invitrogen). Sections were mounted with ProLong Gold Antifade Reagent with DAPI (1:5000, Invitrogen), and images taken under a Leica confocal microscope (TCS SP-2, Leica, Germany).

Densitometric analysis and cell counting

Densitometric analysis of TH-positive fibers in the striatum was performed as previously described [37]. An average of 8 sections from the striatum (bregma +1.18 mm to +0.02 mm; anteroposterior axis) were examined using Image-Pro Plus (vision 6.0, Media Cybernetics) on a computer attached to a light microscope (Leica, Germany).

To measure the density of TH-positive cells in the SNpc we performed stereological cell counting as previously described [19, 37]. Total numbers of TH-positive neurons and Nissl-stained neurons in the SNpc were counted using the optical fractionator method on a Stereo Investigator system (Micro Brightfield, USA) attached to a Leica microscope. A total of 8 sections (one every five 30 μm-thick sections collected −2.80 to −3.80 mm from bregma) were analyzed. Assays were performed in a double-blind fashion by two operators.

Gene expression analysis using RT-PCR

Total RNA from brain tissues was extracted using the Total RNA Kit (TIANGEN, China). First-strand cDNA synthesis was performed using the FastQuant RT Kit (TIANGEN, China). The primers used for real-time PCR are shown in the Supplementary Table 1.

Behavioral studies