Accumulation of NM in Nigrostriatal Dopaminergic Neurons Induces a Parkinson’s Disease Phenotype in Mice

Given the significant recent advances in modelling PD through the use of rodents producing NM (Carballo-Carbajal et al. 2019), we employed an AAV9-hTyr vector, which was directly injected in the SNpc of C57BL6/6J mice to mimic a PD phenotype. The AAV-hTyr or empty AAV-null (control) vector was infused into both sides of the SNpc and 5-weeks later their motor behavior was evaluated in the rotarod and catalepsy tests. As expected, AAV-hTyr injected mice exhibited a shorter time latency to fall from the accelerating rotarod compared to their control (AAV-null) (P < 0.0001, Fig. 1A). Similarly, the time to start movement in the catalepsy test increased significantly in the AAV-hTyr injected mice relative to those that received AAV-null (P < 0.0001, Fig. 1A), indicating that AAV-hTyr injected mice developed a clear PD-like phenotype. Prior to analyzing the nigrostriatal pathway in these animals, hTyr expression in the substantia nigra pars compacta (SNpc) was confirmed by immunohistochemistry (Fig. 1B) and NM production was evidenced both macroscopically (Fig. 1C) and microscopically using neutral red staining (Fig. 1D). Next, we compared the amount of intraneuronal NM pigment stained with neutral red in the brains of mice at 3- and 5- weeks post AAV-Tyr injection. Figure 1D shows that mice injected with hTyr for 5 weeks had significantly more intracellular NM accumulation in the catecholaminergic neurons of the SNpc than those injected for 3 weeks.

Fig. 1

AAV9-mediated hTyr overexpression in the substantia nigra (SN) leads to motor deficits, DA neuronal loss and an inflammatory response in C57BL/6J mice three weeks after injection. (A) Motor function was evaluated 3 weeks after the AAV9-null or AAV9-hTyr injection in the SN of C57BL/6J mice (n = 14 − 13 per group). Motor coordination was assessed by measuring the latency to fall in the accelerating rotarod test (Rotarod test). Muscular rigidity was evaluated by measuring the time required for the mice to withdraw their forepaws from a bar (Catalepsy test). (B) Representative images of a SNpc section from an AAV-hTyr-injected mouse (3 weeks after AAV injection) immunostained for hTyr. NM was visualized macroscopically (C) and microscopically (D) in the SNpc of an AAV-hTyr-injected mouse 3 weeks after AAV9-hTyr injection. Quantification of intracellular NM optical density in DA neurons of mice at 3 and 5 weeks after AAV9-hTyr injection (n = 130–135 neurons). (E) Representative photomicrographs showing TH + dopaminergic neurons and quantification in the midbrain of C57BL/6J mice injected with AAV9-null and AAV9-hTyr (n = 6–8 per group). (F) Representative photomicrographs showing TH + fibers in the striatum and quantification measured by optical densitometry of C57BL/6J mice injected with AAV9-null and AAV9-hTyr (n = 6–8 per group). (G) Representatives images of Iba 1 immunostaining in the midbrain of C57BL/6J mice injected with AAV9-null or AAV9-hTyr. Magnification bar (B), (D), (E), (F) (G) 1 mm, and 200 μm in sets with high magnification. Data are presented as mean ± standard error of the mean (SEM). Unpaired two tailed Student´s t-test was used. ****p ≤ 0.0001

Next, to determine whether overexpression of hTyr affected nigrostriatal dopaminergic neurons, an unbiased stereology method was used to estimate the number of tyrosine hydroxylase-positive neurons (TH+) in the SNpc. Representative photomicrographs of TH + neurons are shown in Fig. 1E. Stereological cell counts revealed a significant loss of TH-positive neurons of SNpc in AAV-hTyr-injected mice compared to AAV-null-injected group (P < 0.0001, Fig. 1E). A Nissl staining was also achieved to confirm that stereological cell counts of TH + neurons really represent a significant neuronal loss rather than a decrease in TH expression (Suppl Fig. 1).

Likewise, striatal tissue sections were also processed for immunohistochemistry to determine whether the loss of nigrostriatal neurons resulted in decreased TH + axons in their target region. Interestingly, a reduction of striatal DA TH-positive fibers measured by optical densitometry, was also patent in AAV-hTyr injected mice compared to controls (AAV-null) (Fig. 1F, P < 0.001), that correlates with the significant motor deficits that these animals displayed. Finally, considering that neuroinflammation and microglia activation are considered neuropathological hallmarks in PD, AAV-hTyr and AAV-null mice brains were immune-stained for Iba-1. As expected, the SNpc of hTyr injected animals showed an important microglial response, whereas the injection of the null vector produced minimal microglial activation (Fig. 1G). The increase of Iba-1 + cells in the SNpc of AAV-hTyr- injected mice suggests that the NM accumulation induces a robust inflammatory response. This finding is consistent with previous studies indicating that NM released from dying neurons can activate microglia, thereby promoting the degeneration of neighboring neurons (Zecca et al. 2003).

Dexamethasone Ameliorates Behavioral Deficits Exhibited by NM-PD Mice

Given the significant neuroinflammation observed in our model and the consensus among several authors that inflammation may contribute to dopaminergic neuron degeneration in PD, and thus to its progression, we propose to investigate the potential benefits of an anti-inflammatory therapy. Thus, leveraging the NM PD mouse model described in Fig. 1, we selected dexamethasone, a potent glucocorticoid, as a potential candidate to reduce brain inflammation and protect dopamine-producing neurons from degeneration.

Considering that NM levels are detectable at 3 weeks, with a significant increase observed at 5 weeks, where we have confirmed the establishment of a severe PD phenotype (Fig. 1) two schedules of dexamethasone or vehicle treatment (2- and 4-week treatment starting 1-week after AAV9-hTyr injection9), were implemented (Fig. 2A). Two groups of mice injected with an empty AAV (AAV-null) and scarified 3- or 5 weeks post injection were used as controls. AAV9-hTyr mice, whether treated with vehicle or dexamethasone, exhibited comparable body weights at baseline and experienced anticipated weight gain over the subsequent 3 weeks of treatment. By the end of this period, mice receiving the vehicle displayed slightly higher weights compared to those receiving dexamethasone, though no significant differences between both groups were detected (Suppl. Figure 2A). Two behavioral tests (rotarod and catalepsy) were employed to measure motor function of control and AVV9-hTyr mice after 2- or 4-weeks of dexamethasone or vehicle treatment. In the rotarod test, no significant differences were observed among the groups at 3 weeks post-injection. However, by 5 weeks post-injection, the vehicle-AAV9-hTyr mice showed a significantly shorter latency to fall from the accelerating rotarod compared to the control group (P < 0.01) and compared to dexamethasone-treated AAV9-hTyr animals (P < 0.05, Fig. 2B). Most notably, the time to descend or initiate movement in the catalepsy test significantly increased progressively over time in the AAV-hTyr vehicle-treated mice compared to controls (P < 0.001 at 3 weeks and P < 0.0001 at 5 weeks post-injection, Fig. 2B). Interestingly, the AAV-hTyr injected mice that received 2 or 3 weeks of dexamethasone treatment showed significant improvements compared to the AAV-hTyr vehicle group at both time points (P < 0.05 at 3 weeks and P < 0.001 at 5 weeks post-injection, Fig. 2B). These data indicate that dexamethasone can alleviate the behavioral deficits exhibited by the NM PD mouse model.

Fig. 2

Dexamethasone (DXM) significantly ameliorates the behavioral deficits exhibited by NM-PD mice. (A) Timeline of AAV Administration, motor testing, and procedures in different experimental groups of C57BL/6J mice over 5 weeks. (B) Motor coordination was analyzed by measuring the latency to fall in accelerating rotarod test (Rotarod test) at 3- and 5- weeks post-injection. Muscular rigidity was evaluated by measuring the time required for the mice to withdraw their forepaws from a bar (Catalepsy test) at 3- and 5- weeks post-injection. (n = 5–9 per group). Data are presented as mean ± standard error of the mean (SEM). One-way ANOVA followed by Tukey post hoc test was used. *p ≤ 0.05, **p ≤ 0.01, better ***p ≤ 0.001****p ≤ 0.0001

Dexamethasone Prevents the loss of Dopaminergic Neurons in the SNpc of NM-PD mice

Dopaminergic cell death in the SNpc was measured in the experimental groups using unbiased stereology to estimate the number of TH + neurons in the SNpc (see Fig. 3A for representative photomicrographs). A one-way ANOVA analysis indicated that, at 3- and 5 weeks AAV9-hTyr injection, vehicle-treated mice exhibited a notable and progressive reduction in TH + cells within the SNpc compared to their respective control group (P < 0.05 and P < 0.0001 in AAV-null vs. AAV-Tyr at 3- and 5- weeks post-inj respectively). Remarkably, as shown in Fig. 3B, both 2- (P < 0.05) and 4-week (P < 0.001) dexamethasone treatments effectively and significantly prevented the decline of DA neurons caused by the overexpression of hTyr. To ensure that the neuroprotective effect was not due to dexamethasone reducing the AAV-mediated expression of hTyr, we performed a real-time PCR analysis. The results confirmed that the enzyme levels were similar in both the vehicle-treated and dexamethasone-treated groups (Suppl. Figure 2).

Fig. 3

Dexamethasone prevents the loss of dopaminergic neurons in the SNpc of NM-PD mice. (A) Representative photomicrographs showing TH + dopaminergic neurons in the midbrain of C57BL/6J mice at 3 and 5 weeks following AAV9-null and AAV9-hTyr injections and treated with Dexamethasone (DXM) or vehicle. (B) Quantification of TH + neurons in the midbrain of C57BL/6J mice at 3 and 5 weeks following AAV9-null and AAV9-hTyr injections and treated with DXM or vehicle. (C) Representative photomicrographs showing TH-immunoreactivity in the striatum of C57BL/6J mice at 3 and 5 weeks following AAV9-null and AAV9-hTyr injections and treated with DXM or vehicle. (D) TH + fibers measured by optical densitometry in striatum of C57BL/6J mice at 3 and 5 weeks following AAV9-null and AAV9-hTyr and treated with DXM or vehicle (n = 5 per group). Magnification bar (A) 1 mm. Data are presented as mean ± standard error of the mean (SEM). One-way ANOVA followed by Tukey post hoc test was used.***p ≤ 0.001****p ≤ 0.0001

Next, striatal sections (see Fig. 3C for representative photomicrographs) were examined by immunohistochemistry, finding a significant reduction in striatal DA TH + fibers measured by optical densitometry in vehicle AAV-hTyr injected mice compared to control group at both time points (P < 0.001 and P < 0.0001 in AAV-null vs. AAV-Tyr at 3- and 5- weeks post-inj respectively Fig. 3D). Interestingly, 3-weeks after the injection of AAV-hTyr, dexamethasone is able to almost completely preserve DA fibers from neurodegeneration (P < 0.001). Significantly, at 5 weeks post-injection, the intensity of TH + immunoreactivity in the Tyr-dexamethasone group was notably higher compared to the Tyr-vehicle group. However, it is important to note that the disease has progressed despite the treatment, probably indicating that the anti-inflammatory effect had less impact on the terminals compared to the cell bodies (Fig. 3D). These findings, while highlighting the potential impact of anti-inflammatory treatment on modulating the degenerative cascade, underscore the complexity of neurodegenerative processes and the challenges in maintaining therapeutic effects with anti-inflammatory drugs over prolonged time periods.

Dexamethasone Inhibits AAV9-hTyr-induced Microglia Activation in the Substantia Nigra pars Compacta

Considering that dexamethasone reduces microglial activation associated to neuroinflammation (Hinkerohe et al. 2010; Hui et al. 2020a), microglia were next analyzed by immunohistochemistry using Iba-1 as the classical microglial marker in the experimental groups (Fig. 4A). As expected, at 3-weeks post injection, a marked increase in immunoreactivity for Iba-1 in the SNPc (Fig. 4B) was observed in the AAV9-hTyr-vehicle group compared to control group (AAV-null), indicative of the microglial response to damage and disease (Gao et al. 2014). The inflammatory response was significantly higher at 3 weeks post-viral injection compared to 5 weeks, likely corresponding to the peak of neurodegeneration, when a substantial amount of extracellular NM is released, triggering microglial reactivation. Interestingly, dexamethasone significantly mitigated Iba1 immunoreactivity at 3 and 5 weeks post-AAV9-hTyr injection (Fig. 4B).

Fig. 4

Dexamethasone inhibits AAV9-hTyr-induced microglia activation in the SNpc of C57BL/6J mice at 3- and 5- weeks post-injection. (A) Representative images of Iba1 immunostaining in the midbrain of C57BL/6J mice 3 and 5 weeks after AAV9-null or AAV9-hTyr injection and treated with Dexamethasone (DXM) or vehicle. (B) Iba1-immunoreactivity in the SN of C57BL/6J mice 3 and 5 weeks after the AAV9-null and AAV9-hTyr injection and treated with DXM or vehicle (C) Representative images of iba1 + cells with different phenotypes (ramified, dystrophic, ameboid and bushy) in C57BL/6J mice injected with AAV9-null, AAV9-hTyr and AAV9-hTyr treated with DXM at 3 weeks post-injection (upper panel). Ramified phenotype is characterized by long and thin ramifications with apparent cell integrity, dystrophic cells exhibit thicker and less clear ramifications, ameboid microglia are distinguished by the absence of ramifications and bushy cells posse a large nucleus and shorter ramifications (upper panel). Classification and quantification of Iba + cells in the midbrain of C57BL/6J mice 3 and 5 weeks after the AAV9-null and AAV9-hTyr injection and treated with DXM or vehicle (lower panel) (n = 5 per group). Magnification bar (A) 1 mm (C) 100 μm. Data are presented as mean ± standard error of the mean (SEM). One-way ANOVA followed by Tukey post hoc test was used. ***p ≤ 0.001****p ≤ 0.0001

We subsequently assessed the extent of Iba1-expressing microglial cells within the striatum; however, no significant differences were observed between groups at any time point (Suppl. Figure 3). Astroglial activation, characterized by GFAP expression, showed an increase in the number of immunoreactive cell bodies in the AAV-Tyr group at 3 weeks post-injection, which was attenuated with Dexamethasone treatment (Suppl Fig. 4). At 5 weeks post-injection, astrocyte activation was less pronounced, and no significant differences were detected between the treated and untreated AAV-Tyr groups.

Next, considering that distinctive alterations in microglial morphology, such as changes in cell shape, branching patterns, and soma size, can signify different stages of microglial activation and may correlate with disease progression and severity, we explored the morphology of these Iba1 + cells. A meticulous analysis of different forms of microglia was counted at the SNpc of the experimental groups based in a recent study (Basurco et al. 2023). Basurco et al., identified four distinct morphological types of microglia: ramified, hypertrophic, bushy, and ameboid. As displayed in Fig. 4C, the ramified phenotype is characterized by long and thin ramifications, in contrast, to dystrophic cells, which exhibit thicker and less clear ramifications. Ameboid microglia, are distinguished by the absence of ramifications. Conversely, bushy cells possess a large nucleus and shorter ramifications. Specifically, in a total of 100 cells counted per group, our results indicate that control animals predominantly exhibit ramified microglia, representing 53% of the phenotype, followed by 33% dystrophic, and 13% ameboid phenotypes (Fig. 4B). In contrast, vehicle-treated animals, at both 3 and 5 weeks post-hTyr injection, showed only 3–5% ramified cells, higher percentage of dystrophic (46–53%) and ameboid (37–39%) phenotype, and 8–10% bushy cells. Interestingly, Dexamethasone-treated animals display a microglial morphology pattern more similar to that of control (AAV-null) animals, with a greater number of ramified cells (39–41%) and less dystrophic (36–37%) and ameboid (15–20%) cells than the vehicle groups. It is noteworthy that bushy microglia are observed solely in hTyr-injected animals treated with vehicle. Overall, the results suggest that vehicle-treated animals with PD-like pathology display a significant shift towards dystrophic and ameboid phenotypes, indicating an activated and potentially neurodegenerative state. Dexamethasone, on the other hand, appears to mitigate this microglial activation, maintaining a microglial phenotype closer to that of healthy controls.

To validate and confirm these findings, we conducted an additional analysis using CD68 to label activated phagocytic microglia (Walker and Lue 2015). Similar to Iba-1 expression, AAV9-hTyr injection induced a marked CD68 immunoreactivity in the substantia nigra that again was more pronounced at 3 weeks post-viral injection compared to 5 weeks (Fig. 5A). Interestingly, dexamethasone significantly mitigated CD68 immunoreactivity at 3 weeks post-AAV9-hTyr injection, but no significant differences were observed between the vehicle and dexamethasone groups at 5 weeks (Fig. 5B). Next, a double immunofluorescence staining of Iba1 and CD68 was performed at the earliest time point. The results showed that in Tyr-injected animals treated with the vehicle, a higher CD68 immunoreactivity was detected in the cell body of Iba1 + cells, indicating a phagocytic microglial phenotype. In contrast, the immunoreactivity for CD68 and the co-localization of both markers was scarce in dexamethasone-treated animals (Fig suppl 5 and Fig. 5C).

Fig. 5

Dexamethasone inhibits AAV9-hTyr-induced activated phagocytic microglia in the SNpc of C57BL/6J mice at 3 weeks post-injection. (A) Representative photomicrographs showing CD68 + cells in the midbrain of C57BL/6J mice 3 and 5 weeks after the AAV9-null and AAV9-hTyr injection and treated with Dexamethasone (DXM) or vehicle. (B) Quantification of CD68 + cells in the midbrain of C57BL/6J mice 3 and 5 weeks after the AAV9-null and AAV9-hTyr injection and treated with DXM or vehicle. (C) Representative images of double labelling of Iba1 (red) and CD68 (green) of C57BL/6J mice 3 weeks after AAV9-null and AAV9-hTyr injection and treated with DXM or vehicle. Magnification bar (A) 1 mm in sets with low magnification and 200 μm in sets with high magnification. Data are presented as mean ± standard error of the mean (SEM). One-way ANOVA followed by Tukey post hoc test was used. **p ≤ 0.01 ***p ≤ 0.001****p ≤ 0.0001

Finally, considering that under pathological conditions, microglial cells may contribute to the recruitment of peripheral leukocytes (mostly T cells) into the brain (McGeer et al. 1993) (McGeer et al., reference), we investigated whether CD3 + T cells were also observed in the brains of hTyr-injected animals. Interestingly, significant brain extravasation of lymphocytes, evidenced by CD3 + immunostaining, was observed in the SNpc of hTyr-injected animals, mainly at 3 weeks (Fig. 6A) and much less evident at 5 weeks (Fig. 6B) post injection. In contrast, the SNpc of control animals and other brain regions of hTyr-injected mice were devoid of T cell infiltration (data not shown). Notably, Dexamethasone treatment resulted in a substantial decrease in CD3 + cells in the SNpc at 3 weeks post injection. Surprisingly, at 5 weeks post-injection, a significantly higher number of CD3 + cells were detected in the Dexamethasone group compared to the vehicle group (Fig. 6C). This suggests that while the neurodegenerative process persists, it progresses more slowly.

Fig. 6

Dexamethasone inhibits AAV9-hTyr-induced brain extravasation of lymphocytes in the SNpc. Representative images of CD3 immunostaining in the midbrain of C57BL/6J mice 3 (A) and 5 (B) weeks after the AAV9-null and AAV9-hTyr injection and treated with Dexamethasone (DXM) or vehicle. (C) Quantification of CD3 + cells in the midbrain of C57BL/6J mice 3 and 5 weeks after the AAV9-null and AAV9-hTyr injection and treated with DXM or vehicle. (D) Quantification of CD3+, CD4 + and CD8 + cells in peripheral blood of C57BL/6J mice after treatment with DXM (1 mg/kg) or vehicle during 15 days (n = 4 per group). Data are presented as mean ± standard error of the mean (SEM). One-way ANOVA followed by Tukey post hoc test was used.*p ≤ 0.05, **p ≤ 0.01. Magnification bar 1 mm and 200 μm in sets with high magnification

Given the immunosuppressive effects of glucocorticoids, the reduced CD3 + infiltration observed in the brain at 3 weeks post-injection (following 15 days of treatment) may result from treatment-induced depletion of peripheral CD3 + cells. To determine whether the dexamethasone dose we used impacted immune cells in the blood, we administered 1 mg/kg of dexamethasone to C57BL/6 mice for 15 consecutive days and assessed the abundance of peripheral T cells by flow cytometry. Notably, as shown in Fig. 6D, we observed a significant reduction in CD3 + and CD4 + cells in the blood of dexamethasone-treated C57BL/6J mice compared to the untreated group. These findings indicate that Dexamethasone attenuates microglial activation and slows the infiltration of peripheral immune cells, particularly CD3 + T cells, into the brain, thereby slowing the progression of the disease.