Abstract
Keywords
INTRODUCTIONS
Activity-dependent neuroprotective protein (ADNP), a vasoactive intestinal peptide (VIP) regul-ated gene, is essential for brain formation [1, 2], neuronal/glial cell differentiation [3]; cell survival [4], and axonal transport [5]. ADNP contains an eight amino acid peptide sequence (NAPVSIPQ), termed NAP (generic name: davunetide), which has been proposed as a potential therapy for neurological disorders [5]. Mutations in the ADNP gene cause an autism syndrome [6] and are associated with cognitive and social deficits [6, 7]. ADNP knockout mice have only partial cranial neural tube closure and die between embryonic days 8.5–9.5 [1]. Heterozygous ADNP+/– mice are viable, but show cognitive deficits associated with tau hyperphosphorylation, neurofibrillary-like tangle formation, and aging-associated neurodegeneration [8]. During development ADNP regulates the tyrosination of alpha-tubulin and increases neuronal growth cone surface [9] and neurite outgrowth [10–12]. Accordingly, knock down of ADNP in cultured neurons results in reduced neurite growth [2]. Interestingly, multiple independent lines of evidences suggest that NAP has astounding neuroprotective effects in a variety of neurodegenerative disease models including amyotrophic lateral sclerosis [13], traumatic brain injury [14], tauopathies [15, 16], and synucleinopathies [17, 18]. Additional studies demonstrate that NAP treatment significantly protects cognitive functions and decreases tau hyperphosphorylation in the triple transgenic mouse model of Alzheimer’s disease [8] and in the mutated tau mouse model of frontotemporal dementia [19]. This reduction of tau hyperphosphorylation by NAP treatment has been shown to involve reduced protein misfolding via enhanced breakage of beta pleated sheets [20]. A recent study revealed that NAP treatment reduced proteinase K-resistent alpha-synuclein inclusions in the substantia nigra of mice that overexpress wild-type human α-synuclein and improves motor performance and coordination [18]. The effect of NAP-preventing α-synuclein accumulation and aggregation appears to occur by stabilizing microtubule function and enhancing axonal transport [18].
We recently found that reductions in axonal transport markers, conventional kinesin heavy and light chains, precede alterations in dopaminergic phenotypic markers (i.e.; somatic tyrosine hydroxylase) in early stages of PD [21]. This reduction was significantly greater in nigral neurons containing α-synuclein inclusions, suggesting that impairments in axonal transport contribute to PD pathogenesis. However, whether nigral neuron degeneration in PD is associated with down-regulation of ADNP expression and alterations in molecular motor proteins involved in the execution of bidirectional axonal transport needed further investigation. The objectives of this study were: (1) to determine whether ADNP protein levels are altered in nigral dopaminergic neurons in PD; (2) to establish whether PD-related alterations in ADNP are associated with α-synuclein accumulation, and (3) to evaluate potential correlations between ADNP levels and reductions in axonal transport motor proteins in PD. Results from these studies suggest that reductions in ADNP levels represents an early degenerative event in both experimental and sporadic PD, further linking this event to reductions in levels of axonal transport motor proteins.
MATERIALS AND METHODS
Human tissue acquisition and processing
Human tissues were obtained at autopsy from 20 subjects with clinical and neuropathological diagnoses of sporadic PD (n = 12) and age-matched controls (n = 8). There were no differences in age at the time of death (P > 0.05) or post-mortem interval (P > 0.05) between two groups (Table 1). PD patients were diagnosed by movement disorder specialists in the Section of Movement Disorders in the Department of Neurological Sciences at Rush University Medical Center. Post-mortem, a board-certified neuropathologist at Rush University Medical Centre confirmed the clinical diagnosis. For PD, inclusion criteria included a history compatible with idiopathic PD and at least two of the four cardinal motor signs (rest tremor, rigidity, akinesia/bradykinesia and gait disturbance/postural reflex impairment). The Unified PD Rating Scale 3 (UPDRS3 ON and OFF medication) and Hoehn and Yahr (ON and OFF medication) were recorded. Pathological PD diagnosis was based on finding Lewy bodies in catecholamine nuclei such as the substantia nigra. Exclusion criteria included familial PD, the Lewy body variant of Alzheimer’s disease or the combination of PD and Alzheimer’s disease. Age-matched control subjects were all participants in the Rush University Religious Order Study, a longitudinal clinical–pathological study of ageing and Alzheimer’s disease, which comprised older Catholic nuns, priests and brothers. Each participant received a clinical evaluation that included an assessment for movement disorders. Details of the clinical evaluation were previously reported [22]. Subjects without psychiatric illnesses during life and neurological abnormalities at post-mortem were included in the control group. The Human Investigation Committee at Rush University Medical Center approved this study.
At autopsy, the brains were removed from the calvarium and processed as described previously [22]. Briefly, each brain was cut into 1 cm coronal slabs using a Plexiglas brain slice apparatus and then hemisected. The slabs were fixed in 4% paraformaldehyde for 5 days at 4°C. The one side brain slabs were used for pathological diagnoses. The other side brain slabs were cryoprotected in 0.1 M phosphate buffered saline (PBS; pH 7.4) containing 2% dimethyl sulphoxide, 10% glycerol for 2 days followed by 2% dimethyl sulphoxide and 20% glycerol in PBS for at least 2 days before sectioning. The fixed slabs were cut into 18 adjacent series of 40μm thick sections on a freezing sliding microtome. All sections were collected and stored in a cryoprotectant solution before processing.
Injection of adeno-associated virus into rat brain
To further assess the relationship between ADNP and α-synuclein, young adult Sprague–Dawley rats (both male and female; Charles River Laboratories, Wilmington, MA) were housed two to a cage with ad libitum access to food and water during a 12 h light/dark cycle; following guidelines established by the Rush University Institutional Animal Care and Use Committee. Recombinant adeno-associated virus 2 serotype 6 vector encoding human wild type α-synuclein gene (rAAV-α-synuclein) or empty vector (control) were prepared and titered as described previously [23]. Under xylazine/ketamine anesthesia, 2μl of the vector suspension was injected stereotaxically into the left nigral region (5.3 mm posterior and 2.3 mm lateral to bregma; 7.7 mm ventral to dura). The needle was kept in place for an additional 5 min before slowly being withdrawn. At 6 weeks after injection, animals (rAAV-α-synuclein, n = 8; empty vector control, n = 8) were perfused through the ascending aorta with physiological saline, followed by 4% ice-cold paraformaldehyde. The brains were post fixed in the same solution for 2 h, progressively transferred through 10%, 20%, and 30% sucrose, and sectioned on a freezing microtome at 40μm in the coronal plane. All sections were collected and stored in order in a cryoprotectant solution before processing.
Single-label immunofluorescence
Several antibodies were used for this study (Table 2). A single-label immunofluorescence method was employed to visualize ADNP. Endogenous autofluorescence were quenched by 20-minute incubation in 0.1 percent of hydrogen peroxide [24], and background immunostaining prevented by 1-hour incubation in a solution containing 2% bovine serum albumin and 5% normal goat or horse serum. One series of sections through the rostrocaudal substantia nigra were immunostained for ADNP overnight at room temperature (1:500 for LS-B61 antibody; 1:1000 for AF5919 antibody). After 6 washes, sections were sequentially incubated for 1 hour in biotinylated goat anti-rabbit (for LS-B61) or horse anti-goat (for AF5919) secondary antibodies followed by incubation with DyLight 488-coupled streptavidin (1:500) for 60-minutes. Sections were mounted on gelatin-coated slides, dehydrated through graded alcohol, cleared in xylene, and coverslipped with DPX.
Stereological quantification of ADNP immunoreactive neurons
An optical fractionator-based unbiased sampling design was used to estimate the total number of ADNP-immunoreactive neurons within the substantia nigra [25, 26]. For each human subject, the substantia nigra pars compacta extending from the caudal level of the mammillary bodies to the decussation of the superior cerebellar peduncle were examined. Seven equispaced sections along the substantia nigra were sampled from each brain. The section sampling fraction (ssf) was 1/0.027. The distance between sections was approximately 1.44 mm. In cross-section, the substantia nigra is located in the ventral midbrain. The substantia nigra pars compacta was outlined according to neuromelanin (NM) distribution using a 1.25× objective. A systematic sample of the area occupied by the substantia nigra pars compacta was made from a random starting point (StereoInvestigator v10.40 software; Micro-BrightField, Colchester, VT). Counts were made at regular predetermined intervals (x = 313μm, y = 313μm), and a counting frame (70×70μm = 4900μm2) was superimposed on images obtained from tissue sections. The area sampling fraction (asf) was 1/0.05. These sections were then analyzed using a 60× Planapo oil immersion objective with a 1.4 numerical aperture. The section thickness was empirically determined. Briefly, as the top of the section was first brought into focus, the stage was zeroed at the z-axis by StereoInvestigator v10.40 software. The stage then stepped through the z-axis until the bottom of the section was in focus. Section thickness averaged 17.21±2.3μm in the midbrain. The disector height (counting frame thickness) was 12μm. This method allowed for 2μm top guard zones and at least 2μm bottom guard zones. The thickness sampling fraction (tsf) was 1/0.58. Care was taken to ensure that the top and bottom forbidden planes were never included in the cell counting. The human midbrain dopaminergic neurons contain NM. NM provides an easily discernible endogenous marker for dopaminergic neurons, allowing for an easy assessment of co-localization with ADNP-immunofluorescent products in dopaminergic neurons. ADNP-immunofluorescent-positive/NM-laden or NM-laden only (ADNP undetectable) nigral neurons were separately counted with different markers. Using stereological principles, ADNP-immunofluorescent-positive/NM-laden or NM-laden neurons in each case were sampled using a uniform, systematic, and random design. The total numbers of ADNP-immunofluorescent-positive /NM-laden neurons, NM-laden neurons only, and ADNP-immunofluorescent-positive/NM-laden plus NM-laden neurons within the substantia nigra pars compacta were calculated separately using the following formula: N = ΣQ–·1/ssf · 1/asf · 1/tsf, where ΣQ was the total number of raw counts. The coefficients of error (CE) were calculated according to the procedure of Gunderson and colleagues as estimates of precision [27]. The values of CE were 0.13±0.05 (range 0.10 to 0.15) in PD and 0.10±0.02 (range 0.08 to 0.12) in age-matched control. As NM-laden neurons degenerate in PD cases, the percentage of ADNP immunofluorescent-positive neurons in total NM-laden nigral neurons was calculated and compared between PD cases and age-matched controls.
Double-label immunofluorescence
A double-label immunofluorescence procedure was employed to determine whether ADNP expression was altered in neurons that co-expressed tyrosine hydroxylase, α-synuclein, kinesin heavy chain, and dynein light chain Tctex type 3. Midbrain sections from each brain were incubated overnight in the first primary antibody (anti-ADNP; LS-B61 antibody) and the goat anti-rabbit antibody coupled to DyLight 488 (1:200, Jackson ImmunoResearch) for 1 hour. After blockade for 1 h, the sections were then incubated in the second primary antibodies (α-synuclein, 1:500; tyrosine hydroxylase, 1:5000; kinesin heavy chain, 1:500; dynein light chain Tctex type 3. 1:500) overnight, and the goat anti-mouse antibody coupled to DyLight 649 (1:200) for 1 hour. The sections were mounted on gelatin-coated slides and allowed to air dry overnight. To block autofluorescence, the sections were rinsed in distilled water, dehydrated in 70% alcohol for 5 min, incubated in the autofluorescent eliminator reagent (2160; Millipore) for 5 min, and immersed in three changes of 70% alcohol. After rinsing in distilled water, the sections were cover slipped using polyvinyl alcohol with DABCO (Sigma-Aldrich).
Optical density measurements
Optical density measurements were performed according to our previously published procedures [22, 26]. All immunofluorescence double-labeled images were scanned with an Olympus Confocal Fluoroview microscope equipped with argon, helium-neon lasers, and transparent optics. With a 20× magnification objective and a 488 and 633 nm excitation source, images were acquired at each sampling site in the substantia nigra pars compacta and were saved to a Fluoroview file. Following acquisition of an image, the stage moves automatically to the next sampling site to ensure a completely non-redundant evaluation. Once all images were acquired, optical density measurements were performed on individual nigral neurons at the nuclear level. To maintain consistency of the scanned image for each slide, the laser intensity, confocal aperture, photomultiplier voltage, offset, electronic gain, scan speed, image size, filter and zoom were set for the background level whereby autofluorescence was not visible with a control section. These settings were maintained throughout the entire experiment [28]. The intensity mapping sliders ranged from 0 to 4095; 0 represented a maximum black image and 4095 represented a maximum bright image. The ADNP-immunoreactive neurons with or without α-synuclein-immunoreactive inclusions were identified and outlined separately. Quantitative optical density of immunofluorescence was performed on individual ADNP-immunoreactive neurons with or without α-synuclein-immunoreactive inclusions in different channels. Five equispaced sections across the entire length of the substantia nigra were sampled and evaluated. The number of cells per case was analyzed as follows: >100 nigral cells in normal cases, 50–70 nigral cells per PD case that contained inclusions and >100 nigral cells per PD case that did not contain inclusions. To account for differences in background staining intensity, five background intensity measurements lacking immunofluorescent profiles were taken from each section. The mean of these five measurements constituted the background intensity that was then subtracted from the measured optical density of each individual neuron to provide a final optical density value. To evaluate the optical density of tyrosine hydroxylase, kinesin heavy chain and dynein light chain immunofluorescence in ADNP-immunoreactive nigral neurons, we performed similar analyses in all human and rat tissues. To confirm co-localization of the double labeling, optical scanning through the neuron’s z-axis was performed at 1-μm thickness and neurons suspected of being double labeled were confirmed with confocal cross-sectional analyses.
Data analyses
Demographic and clinical characteristics were compared with Mann-Whitney tests. Optical density measurements were compared across groups with one-way Kruskal–Wallis test followed by Dunn’s post hoc tests for multiple comparisons and with Mann-Whitney test for two groups (Prism 4, GraphPad Software, Inc.). Correlations between optical density measurements were performed by using Spearman’s rank correlation. The level of significance was set at 0.05 (two-tailed).
Digital illustrations
Confocal images were exported from the Olympus laser-scanning microscope with Fluoview software and stored as tif files. Conventional light microscopic images were acquired using an Olympus microscope (BX61) attached to a Nikon digital camera DXM1200 and stored as tif files. All figures were prepared using Photoshop 8.0 graphics software. Only minor adjustments of brightness were made.
RESULTS
Morphological analysis and stereological estimate of ADNP-immunoreactive nigral neurons in age-matched control and PD brains
Immunofluorescent staining revealed that ADNP-immunoreactive neurons were widely distributed throughout the substantia nigra of age-matched control brains (Fig. 1A). Strong ADNP-immunoreactive profiles were observed within nucleus and perikarya in the substantia nigra, with little to no staining in neuronal processes. The vast majority of NM-laden neurons (Fig. 1B) displayed ADNP-immunoreactive in control brains (Fig. 1C). Intense ADNP immunoreactivity was also observed in scattered non-NM-laden small cells (Fig. 1A, C). In contrast, immunoreactivity ADNP was markedly reduced in age-matched PD brains (Fig. 1D, G) in which NM-laden neurons (Fig. 1E, H) showed variable levels of ADNP immnostaining. Some NM-laden neurons displayed light ADNP-immunoreactive somata, whereas others exhibited undetectable ADNP-immunoreactive soma staining (Fig. 1F, I). Although there were abundant NM-laden nigral neurons in cases with Hoehn and Yahr 2 PD (Fig. 1E), only a subset of them were ADNP immunopositive (Fig. 1F). To estimate the number of ADNP immureactive neurons in remaining nigral NM-laden neurons, stereological methods were employed and they revealed a significant decrease in the number of ADNP-immunoreactive nigral neurons in all PD cases. This finding is not surprising since ADNP extensively co-localizes within nigral NM-laden neurons and these are lost in PD (see Table 3). More relevant, the percentage of ADNP-immunoreactive/NM positive neurons in the total remaining nigral NM-laden neurons was significantly decreased (49.71%; P < 0.01) in PDs as compared with age-matched controls (82.07%), indicating that the reductions in ADNP is part of the nigral pathogenic process in PD.
Co-localization and quantitative analysis of ADNP and dopaminergic markers in nigral neurons in PD
To evaluate whether the decrease in ADNP levels precedes well-established PD-related alterations in dopaminergic neuron markers [26]; double immunostaining for tyrosine hydroxylase and ADNP was performed in midbrain sections of cases with sporadic PD and age-matched control brains. In early Hoehn and Yahr stage 2 PD cases, co-localization analyses revealed that remaining nigral somata featured intense tyrosine hydroxylase immunostaining (Fig. 2E), yet these neurons displayed a marked reduction in ADNP-immunoreactivity (Fig. 2D) compared with age-matched controls (Fig. 2A). Both ADNP and tyrosine hydroxylase immunoreactivities (Fig. 2G, J, H, K) were reduced in late Hoehn and Yahr 3–5 PD cases. Optical density measurements of tyrosine hydroxylase (Fig. 2M) and ADNP (Fig. 2N) immunoreactive levels were performed in remaining NM-laden neurons. A Kruskal–Wallis test revealed significant reductions of tyrosine hydroxylase levels in PD (P < 0.001). Post hoc analyses revealed normal optical density measures of tyrosine hydroxylase immunoreactivities in early Hoehn and Yahr stage 2 PD cases (P > 0.05), as previously described [21]. In contrast, remaining nigral neurons in Hoehn and Yahr Stages 3–5 PD cases displayed diminished tyrosine hydroxylase optical density (P < 0.01), compared with age-matched controls. Thus the pattern of ADNP staining in these samples was different than what was seen for tyrosine hydroxylase. Kruskal–Wallis test revealed significant reduction of ADNP levels in PD (P < 0.001). Post hoc analyses demonstrated that in Hoehn and Yahr stage 2 PD cases, when tyrosine hydroxylase staining was comparable to age-matched controls, optical density values for ADNP were greatly diminished (Fig. 2N; P < 0.01). Further, this reduction in ADNP levels was not exacerbated when PD progressed to Stages 3–5 (Fig. 2N; (P < 0.01), compared to age-matched controls. Taken together, results from qualitative and quantitative immunofluorescence analyses demonstrated significant reduction in ADNP levels that precede tyrosine hydroxylase downregulation in remaining nigral neurons of sporadic PD cases.
We next wanted to determine whether the decrease in ADNP immunofluorescence intensity seen in PD was specific for dopaminergic neurons. In this regard, the ADNP expression in oculomotor nuclei was examined. We chose this brain region because it is located cytoarchitectonically close to the substantia nigra and co-resides in many of the same sections. Intense ADNP immunoreactivities were observed in both PD and age-matched control brains (Fig. 3A-C). Quantitative ADNP immunofluorescence intensity was performed on oculomotor nuclei in all human subjects. Kruskal–Wallis test revealed that the level of ADNP immunofluorescence intensity within the oculomotor nucleus in early and late PD cases was similar to age-matched controls (P > 0.05; Fig. 3D), suggesting that ADNP expression is normal in the area which PD pathology does not occur.
Co-localization and quantitative analyses of ADNP and α-synuclein in the substantia nigra in PD
Results from experiments above indicated that a reduction in ADNP levels in tyrosine hydroxylase immunopositive neurons in sporadic PD, compared to other neurons. Based on results from our previous work [22, 28], we hypothesized that the selective reductions in ADNP levels was associated with α-synuclein aggregation. To evaluate this possibility, we performed double labeling with anti-ADNP and α-synuclein antibodies and obtained the optical density measurements for ADNP from neurons with or without α-synuclein immunoreactive inclusions in sporadic PD. Co-localization studies revealed that both neurons with or without α-synuclein-immunoreactive inclusions had significantly lower density of ADNP immunoreactivity in PD brains, compared with age-matched controls (Fig. 4F, I, L). Although cytoplasmic non-aggregated α-synuclein (Fig. 4C) was present within most nigral neurons in age-matched controls, ADNP immunoreactivity levels were similar among α-synuclein-positive and -negative neurons (Fig. 4C). To unequivocally determine whether decreases in levels of ADNP were associated α-synuclein inclusions in PD, we analyzed the relative intensities of ADNP in nigral neurons that did or did not contain α-synuclein-positive inclusions. Kruskal–Wallis test revealed a statistically significant difference in optical density of ADNP-immunoreactive intensity across these groups (Fig. 4M; P < 0.001). Post hoc analyses revealed a significant decrease of ADNP-immunoreactive optical density in both nigral neurons with (P < 0.001) and without (P < 0.05) α-synuclein-immunoreactive inclusions compared with controls. There was some differences but not significant between neurons with and without α-synuclein-immunoreactive inclusion. These data indicate that reductions in ADNP process independently of α– synuclein aggregation in nigral neurons of sporadic PD.
Co-localization and quantitative analysis of ADNP and axonal transport markers in nigral neurons in PD
ADNP plays an important role on stabilizing microtubule function [5, 17] and protects against impairments in axonal transport [29]. Our recent studies documented alterations in levels of motor proteins responsible for axonal transport as an early pathogenic feature of sporadic and experimental PD [21]. Based on these precedents, we evaluated whether the reduction of ADNP levels was associated with these alterations. To this end, a double immunostaining for ADNP and kinesin heavy chain (anterograde axonal transport) or dynein light chain tctex-type 3 (retrograde axonal transport) was performed in nigral neurons. Optical density measurements corresponding to ADNP and these motor proteins were performed within the same NM-laden neurons. This co-localization analyses revealed that NM-laden neurons with low intensity of ADNP staining (Fig. 5E, I) displayed reduced kinesin heavy chain immunoreactive intensity (Fig. 5F, J) in all PD cases, relative to age-matched controls (Fig. 5A, B). Similarly both ADNP (Fig. 6E, I) and dynein light chain tctex-type 3 (Fig. 6F, J) immunoreactivities were significantly lower in all PD cases relative to age-matched controls (Fig. 6A, B). However some non-NM-laden nigral neurons displayed intense of ADNP (Figs. 5E, 5I, 6E, 6I), kinesin heavy chain (Fig. 5F, J), and dynein light chain tctex-type 3 (Fig. 6F, J) immunoreactivity, suggesting that alterations in ADNP and axonal transport motor proteins mainly affect nigral dopaminergic neurons in sporadic PD. To confirm these observations, we performed quantitative measurement of kinesin heavy chain/ADNP or dynein light chain tctex-type 3/ADNP immunoreactive intensity on NM-laden neuron from all 20 subjects. When all cells were analyzed together, the levels of ADNP (P < 0.001; Figs. 5M, 6M), kinesin heavy chain (P < 0.001; Fig. 5N), and dynein light chain tctex-type 3 (P < 0.005; Fig. 6N) were significant reduced in PD as compared with age-matched controls. A regression analysis demonstrated that there was a positive correlation between ADNP and kinesin heavy chain levels (r = 0.66; P < 0.001; Fig. 5O) across groups. Similarly, the decrease of ADNP levels was also correlated with alterations of dynein light chain tctex-type 3 across groups (r = 0.85; P < 0.001; Fig. 6O). Collectively, these data indicated a concomitant reduction in levels of ADNP and motor protein declined in nigral neurons of sporadic PD.
Viral over-expression of a-synuclein in rat brain-induced alterations in ADNP
To further study the role of α-synuclein expression in ADNP laden cells, we analyzed a well-studied rat PD model based on the intranigral viral (rAAV) over-expression of wild-type α-synuclein within the nigrostriatal system [26, 30]. Control rats received intranigral injections of empty viral vector. The LB509 antibody, which selectively recognizes human α-synuclein protein [26], was used to examine the expression of viral-encoded human α-synuclein. Intense LB509 immunoreactivity was observed in both the striatum (Fig. 7A) and substantia nigra (Fig. 7B, C) of brain hemispheres injected with rAAV-α-synuclein, compared to contralateral side. As observed in cases with sporadic PD, loss of tyrosine hydroxylase-immunoreactive terminals in striatum (Fig. 7D) and neurons in substantia nigra (Fig. 7E, F) was demonstrated in this PD model [21, 31]. Interestingly, tyrosine hydroxylase-immunoreactivity was dramatically decreased in axonal fibers featuring high levels of α-synuclein expression (Fig. 7D, A). α-synuclein aggregation was observed in nigral neurons (Fig. 7C’). However, some of these nigral neurons displayed intense tyrosine hydroxylase immunoreactivity (Fig. 7F). On the basis of results from our previous work [21, 26], we hypothesized that the levels of tyrosine hydroxylase were related to α-synuclein accumulation. To evaluate this possibility, we performed double labeling with anti-α-synuclein and tyrosine hydroxylase antibodies and obtained the optical density measurements for tyrosine hydroxylase from nigral neurons with or without α-synuclein immunoreactivity. Co-localization analyses revealed that nigral neurons with co-expression α-synuclein displayed a marked diminution of tyrosine hydroxylase reactivity while neurons expressing low α-synuclein levels exhibited intense tyrosine hydroxylase reactivity (Fig. 8C) relative to controls (Fig. 8F). Quantitative fluorescence intensity measurements revealed that the levels of tyrosine hydroxylase were significantly reduced in the nigral neurons with increased levels of α-synuclein, compared to those lacking human α-synuclein expression (Fig. 8G). Together, pathological alterations in rats that received rAAV-α-synuclein mimicked the features seen in cases with sporadic PD [22].
Next, we further examined the levels of ADNP in rAAV-α-synuclein-injected rats. In substantia nigra of empty vector-injected control rats, robust ADNP immunoreactivities (Fig. 9E) were observed mainly in neuronal soma. In contrast, the intensity of ADNP immunoreactivity was severely reduced in nigral neurons of rAAV-α-synuclein-injected rats (Fig. 9B). Double labelling of ADNP with anti-α-synuclein antibody further revealed that the ADNP immunofluorescent intensity was reduced in both nigral neurons with α-synuclein immuno-positive or -negative profile (Fig. 9C). Qualitative observations were confirmed by quantitative immunofluorescence studies. The optical density measurements of ADNP were performed on the cell body on a per neuron basis in nigral neurons with or without human α-syn expression. Kruskal-Wallis test revealed a statistically significant difference in the level of ADNP immunoreactive intensity (P < 0.001, Fig. 9G). Post hoc analyses revealed that, compared with controls, the level of ADNP was significantly lower in nigral neurons with (p < 0.001) or without (p < 0.05) α-synuclein-immunoreactivity in rats that received rAAV-α-synuclein. The reduction of ADNP immunoreactivity was significantly more pronounced in nigral neurons displaying α-synuclein-immunoreactivity compared with neurons with undetectable α-synuclein (p < 0.05). This data further demonstrated that the ADNP expression is selected impaired molecule in synucleinopathies.
To evaluate the association of ADNP with the axonal transport marker, colocalization analysis revealed that intensities of ADNP/kinesin heavy chain (Fig. 10B, C) or ADNP/dynein light chain tctex-type 3 (Fig. 11B, C) staining were severely reduced in the substantia nigra of rAAV-α-synuclein-injected rats relative to controls (Figs. 10F, 10G, 11F, 11G). Qualitative observations were confirmed by quantified fluorescence intensity measurements for kinesin heavy chain and dynein light chain tctex-type 3. Measurements were performed on the cell body of kinesin heavy chain or dynein light chain tctex-type 3 immunofluorescent staining nigral neurons which were ADNP immunopositive. Mann-Whitney test revealed a statistically significant decreased in optical densities of ADNP (P < 0.001, Figs. 10I, 11I), kinesin heavy chain (P < 0.001, Fig. 10J) and dynein light chain (P < 0.001, Fig. 11J) immunoreactive intensity compared with control. A regression analysis demonstrated that there were positive correlations between ADNP/kinesin heavy chain levels (r = 0.61; P < 0.01; Fig. 10K) and ADNP/dynein light chain tctex-type 3 (r = 0.91; P < 0.001; Fig. 11K) across groups. Together, these data verified that target α-synuclein overexpression results in parallel reductions of ADNP and axonal transport motor protein levels in the nigral neurons of rats, similar to what was observed in PD brains (Figs. 5, 6).
DISCUSSION
This study is the first to examine alterations of ADNP in sporadic PD. The major findings of the current work are: 1) the ADNP protein levels were reduced in remaining NM-laden nigral neurons of PD vs the age-matched controls. The decrease in ADNP expression occurred at early PD stages, preceding the loss of the dopaminergic phenotypic marker (tyrosine hydroxylase). These data indicate that ADNP deficits represent an early pathological change in sporadic PD. 2) ADNP levels were decreased in nigral NM-laden neurons in a manner independent of α– synuclein aggregations. Additional data indicate that ADNP deficits are also associated with experimental synucleinopathies and these studies suggest that elevated levels of monomeric or tetrameric α– synuclein influence ADNP levels; Furthermore, 3) decreases in ADNP expression were positively correlated with decreased levels of kinesin heavy chain and dynein light chain Tctex type 3, important motor proteins responsible for axonal transport. This observation, and prior studies showing a role of ADNP in axonal transport regulation [5, 21]; suggest that ADNP deficits may contribute to axonal transport defects previously documented in PD; 4) reductions in ADNP levels are recapitulated in a rat model of PD based on over-expression of human wild-type α-synuclein, furthering the conception that ADNP reductions in PD are a direct result of α-synuclein overexpression but not necessarily associated with α-synuclein aggregation. Together, these findings support the concept that down-regulation of ADNP expression might represent a critical pathogenic event contributing to dopaminergic neurodegeneration in PD.
Multiple lines of evidence demonstrated that ADNP modulates axonal transport functions [8, 32]. Recent studies demonstrated that the ADNP interacts with the microtubule end-binding protein family (EB1-3) [33] and the autophagy regulating microtubule-associated protein 1 light chain 3 (LC3) [34]. The ADNP/EB interaction regulates synaptic plasticity and axonal transports whereas the ADNP/LC3 binding can modulate autophagy function, explaining the neuroprotective and neurotrophic capacities. In vitro studies demonstrated that NAP motif of ADNP regulated the tubulin pool and affected the α-tubulin tyrosination in cycle in PC12 cells and neuronal differential model [9] and protected against cyanide-related microtubule distraction [35] and colchicine-related microtubule breakdown [29]. Present studies demonstrated that decrease in ADNP expression paralleled to the alterations of axonal transport markers. Supported by the previous and present data, it is therefore tempting to speculate that the reduction of ADNP expression is a critical event which interferes with axonal transport functions.
ADNP is widely distributed in brain and likely plays a role of neuroprotection [2, 37]. Decreased ADNP synthesis by antisense oligodeoxynucletides reduces the cellular viability [4]. Our data revealed that the ADNP level was significantly reduced (Fig. 2N) in nigral neurons of PD relative to age-matched control. Interestingly, overexpression of wild type α-synuclein in rat leaded to reduction of ADNP expression (Fig. 9G). We hypothesize that the reduction of ADNP expression by abnormal α-synuclein accumulation may be a critical event for the dopaminergic neurodegeneration. Several studies have indicated that ADNP down-regulation result in cell death that was associated with an increase of the proapoptotic protein p53 expression [4, 39]. P53 has been postulated to contribute to PD pathogenesis. Several PD toxins such as MPTP, rotenone, and 6-hydroxydopamine activate p53 [40]. In addition, PD-linked proteins; α-synuclein [41], parkin [42], DJ-1 9 [43], and LRRK2 [44] all effect effecting p53 expression. Inhibition of ADNP protein expression also results in increasing the tumor suppressor p53 expression [4]. These studies suggest that there is an interrelation of ADNP and p53. Further studies are needed to explore the mechanism of ADNP deficit involving in PD pathogeneses.
Qualitative and quantitative observations revealed reductions of ADNP levels in both nigral neurons with and without α-synuclein aggregations in PD brains (Fig. 4). In order to confirm this result, we further analyzed ADNP levels in rats overexpressing human wild-type α-synuclein. Similar to what is seen in the PD brain, both nigral neurons with or without α-synuclein immunoreactivity exhibited reduction of ADNP levels (Fig. 9) but not tyrosine hydroxylase (Fig. 8). Interesting the reductions in the preclinical model, as opposite what was seen in PD, was exacerbated in cells with α-synuclein aggregates.
Our previous studies in sporadic PD indicated that the expressions of neuronal bio-markers in substantia nigra have different response to synucleinopathy. Myocyte enhancer factor-2D [28] and voltage-dependent anion channel 1 [26] displayed a selectively decreased in nigral neurons with α-synuclein inclusions, whereas cytochrome c oxidase subunit 2 (MT-CO2) exhibited unchange to α-synuclein inclusions [26]. Interestingly, the axonal transport markers and ADNP were reduced in nigral neurons independently to α-synuclein inclusions. We hypothesize that decline of ADNP may associate with the expression of axonal transport motors. To this regard, further co-localization and correlative analyses were performed revealing that both ADNP and axonal transport motor proteins were decreased in both PD and rats with nigral α-synuclein over-expression. Our data further indicated that deficiencies of the microtubule-stabilizer ADNP and axonal transport motor proteins are early pathological alterations during Parkinson’s disease progression. Whether the ADNP deficiency affects axonal transport marker expression or vice versa is still unknown. ADNP is a 9 zinc finger, homeodomain containing protein that has a transcription factor activity [4] and regulates multiple gene families including lytic vacuoles and lipid transport [45]. Detail studies are needed to clarify how ADNP regulates the axonal transport motor protein expression.
Post-mortem studies must be interpreted with caution as factors such as disease heterogeneity, drug treatments, and post-mortem interval can influence the results. To support our findings, we injected rats with viral vectors inducing the over-expression of human wild-type α-synuclein. rAAV-α-synuclein caused reductions in ADNP and axonal transport motor proteins accompanied by decrease of tyrosine hydroxylase immunoreactivity in striatum. The data from rAAV-α-synuclein-injected animals exhibited reductions in ADNP and axonal transport motors levels were similar to those observed in early stage human PD brains. These pre-clinical data support the accuracy of the human post-mortem findings.
In summary, the present data indicate that reduction in ADNP levels occurs during early stages in nigrostriatal pathogenesis in PD. The specificity of this effect was confirmed by the absence of changes in these same cases in the oculomotor nucleus. Decreases of ADNP were associated with reductions of axonal transport motor proteins. These data were recapitulated in rats with targeted viral over-expression of α-synuclein. Taken together, these data suggest that ADNP or its motif NAP is worth to be further investigated in the PD models.
CONFLICTS OF INTEREST
The authors have not conflict of interest to report.
Footnotes
ACKNOWLEDGMENTS
We would like to thank Yinzhen He and Katie Nice for histological assistance.