Double labelling of NG2-glia and astrocytes or microglia in the striatum of 6-OHDA-lesioned parkinsonian rats treated with L-DOPA and/or doxycycline
We performed a double immunofluorescence reaction to determine the presence and/or co-localization of the astroglial protein GFAP, the microglial protein OX-42 and the NG2-antigen in the striatum from 6-OHDA-injected animals and those presenting LID. The yellow pseudo-color for these structures visible in microscopic images identified the superimposition of NG2-glia immunoreactivity (green) on either GFAP or OX-42 positive immunoreactivity (red) cell body/process and vice versa .
Similar to Bortolanza et al., (2015), the dopamine-depleted striatum of rats under L-DOPA treatment (6OHDA+Vehicle+L-DOPA , p<0.001, Table 1) had a prominent GFAP and OX42 immunoreactivity (6OHDA+Vehicle+L-DOPA , p<0.05, Table 1). Doxycycline administered before L-DOPA reduced LID and also GFAP and OX42 immunoreactivity in the depleted striatum (6OHDA+DOXY+L-DOPA , p<0.001, Table 1).
High resolution in single plane image showed an extensive interdigitation of the processes deriving from astrocytes/microglia and NG2-glia. The fine processes of NG2 (green) and astrocytes (red) or microglia (red), characteristically shown as irregular shaped dots, closely associated in tiny points (yellow) of all subdivisions of striatum (Fig. 4, 5A-D). There was no effect of the lesion or the treatment in the measured parameters of the microglia and astrocytes markers.
DISCUSSION (1.002 words)
Our data revealed (i) a slight increase in the immunoreactivity of NG2-glia in the lesioned dorso- and ventrolateral striatum 36 days after the 6-OHDA-microinjection; (ii) a robust decrease of NG2-glia immunoreactivity in the dorsolateral, dorso- and ventromedial striatum of rats presenting LID with NG2-glia density negatively correlated with LID score; (iii) a robust increase of NG2-glia density in the dorsomedial, dorso- and ventrolateral striatum following doxycycline antidyskinetic therapy, in contrast to a decrease in the GFAP and OX-42 immunoreactivity. The morphometric analysis of the striatum of the animals expressing LID revealed (iv) an increase in the NG2-glia indicators of activation in the dorso- and ventromedial striatum with a decrease in the dorsolateral one. The aforementioned results revealed: (v) a further increase in the activation cells parameters only in the dorsolateral striatum, associated to doxycycline antidyskinetic effect; (vi) a decrease in the NG2-glia indicators of activation in the dorso- and ventromedial striatum. The no-lesion animals and the only 6-OHDA lesioned rats receiving vehicle or doxycycline exhibited: (vii) a NG2-glia morphological features of a typical resting cell; (viii) a NG2-glia immunoreactivity decrease in the dorsal and ventromedial striatum. The results provide evidence of the dynamic involvement of NG2-glia in the dyskinesia induced by L-DOPA chronic treatment of parkinsonian rats.
One of the remarkable findings of the present study is that NG2-glia expression decreases in the dorsal striatum of animals’ disclosing LID. The dorsal striatum receives dense dopaminergic innervation from the SNc and the lateral portion of the ventral tegmental area (Haber et al., 2000; Ikemoto, 2007; Lerner et al., 2015), strong projections from somatomotor cortical areas and lateral thalamic motor nuclei (Smith, et al., 2004). Independently, controlled information flows from the SNc to the dorsolateral and dorsomedial striatum (Lerner et al., 2015). The dorsolateral striatum presents a sensorimotor role, controlling voluntary movement severely compromised in PD (Flaherty and Graybiel, 1994).
The doxycycline antidyskinestic action increased NG2-glia cell density, with activated phenotype. In this study, we did not measure L-DOPA plasmatic concentrations. Therefore, a potential doxycycline interference on L-DOPA gastrointestinal absorption cannot be ruled out. However, this is an unlikely possibility because previous research from our group (Bortolanza et al., 2020) provide evidences of no drug interference with the positive motor effects of L-DOPA and the anti-dyskinetic effects of doxycycline were still present when L-DOPA was administered subcutaneously. The large majority of NG2-glia in the adult brain is maintained in a quiescent state under physiological conditions (Hughes et al., 2013). Phenotypically, NG2 glia provides a stereotypic reaction with increased NG2 expression, retraction of cell processes, cell body swelling, cell proliferation, and migration toward the lesion site to almost all kind of injury, independently of the extent of myelin loss (Bedner et al., 2020; Jin et al., 2018; Valny et al., 2018). A neuromodulatory mechanism has been proposed to be related to NG2-glia reactivity (Levine et al., 2016). A co-culture of primary NG2-glia with damage hippocampal slices activated NG2-glia, produced neurotrophic factors and immunomodulatory action by creating anti-inflammatory cytokines as interleukin-10 and transforming growth factor (TGF) β2 (Sypecka and Sarnowska, 2014).
PD preclinical studies evidenced NG2-glia as a protector factor against neuroinflammation and dopaminergic neuron apoptosis (Kitamura et al., 2010; Zhang et al., 2019). The NG2-glia ablated mouse brain exhibited increases in pro-inflammatory response subsequent exposure to endotoxin lipopolysaccharide (Zhang et al., 2019). An enriched environment with physical activity induces increased numbers of newborn NG2-positive and GFAP-positive cells in the adult SN and improves motor behavior function in the 6-OHDA rat model of PD (Steiner et al., 2006; Kirby et al., 2019). There are however, contradictory observations. Nakano et al. (2017) and Zhang et al. (2019) found that NG2-glia ablation exacerbated dopaminergic neuronal cell loss in a mouse PD model, induced neurodegeneration, microglia activation, and neuroinflammation in the adult hippocampus. In the post-mortem brain specimens from patients with Alzheimer’s disease, there is a reduction in NG2-glia immunoreactivity, which is negatively correlated to microglial immunoreactivity (Nielsen et al., 2013), but associated with the amyloid-β plaques (Fiedorowicz et al., 2008; Zhang et al., 2019). As shown in amyotrophic lateral sclerosis models NG2-glia has been observed either to limit central nervous system damage or to actively contribute to neuroinflammation or neurotoxicity (Kang et al., 2013).
Consistent with the notion that NG2-glia, microglia, and astrocytes are distinct glial cells subpopulations NG2/GFAP or NG2/OX42 staining was not once found to be co‐localized on the same cell. Noteworthy, the number of NG2/GFAP or NG2/OX42 staining superimposition points was not altered by the factors lesion or treatment. An interaction was demonstrated between NG2 glia and astrocytes (Hamilton et al., 2010; Xu et al., 2014), microglia (Nishiyama et al., 1997), and neurons (Maldonado and Angulo, 2015). This anatomical relationship is a basis for the coupling and functional communication between the glial subtypes. Further, NG2-glia is a source of neurotrophic factors in the central nervous system (Bankston et al., 2013). Liu and Aguzzi (2020) demonstrated that NG2-glia and microglia interaction played pivotal roles in regulating microglia states in the adult mouse brain in Alzheimer model disease. Astrocytes signal to NG2-glia via the release of ATP and glutamate that evokes a Ca2+ rise in NG2-glia (Hamilton et al., 2010). Considering that the TGFβ pathway is essential for normal microglia development (Butovsky et al., 2014; Krasemann et al., 2017), TGFβ pathway suppression may be associated with microglia dysfunction. Whether the NG2-glia function contributes to the disturbing microglia and astrocyte signature in the dorsal striatum of rats presenting LID is not determined yet.
A limitation of the present study was the performance only in male rats to avoid gender influence in the results. It is important to highlight that the analysis of dyskinesia induced by L-DOPA was originally standardized in female rats (Cenci et al., 1999).
As aforementioned, there is an association between damaged dopamine neurons, L-DOPA-treatment, and dysregulated inflammation (Bortolanza et al., 2015b; Del-Bel et al., 2016; Mulas et al., 2016; Teema et al., 2016). The hypothesis of NG2-glia decrease inflammatory reaction is compatible with the observation of the NG2-glia immunoreactivity decrease in LID, and the increase in the microglia/astrocytes. The implications of this discovery are far-reaching and needs more investigation.
SIGNIFICANCE: Our findings provide the first description of the distribution and morphological changes of NG2-glia in the striatum of parkinsonian rats presenting LID and the effect of doxycycline therapy. They support an inverse link between the transformation of NG2 glia to the reactive form and microglial/astrocyte activation/recruitment in a specific brain region, directly enrolled in PD and the dyskinesia manifestation. These data could reflect the possibility that NG2-glia cells promote striatal plasticity as a form of dyskinesia recovery. Besides, this study may also contribute to establishing NG2-glia as a novel therapeutic target for LID and so, we believe that our findings are of great interest to the neuroscience and medical community.
ACKNOWLEDGMENTS The authors would like to thank Célia A. da-Silva, Sara Saltareli, and Vitor Castania for their technical assistance.
CONFLICT OF INTEREST EADB is listed as co‐inventor on a U.S. patent application covering doxycycline and related compounds (Patent No. EP18306400.5. PCT: 1054). All other co‐authors declare that no conflict of interest exists.
AUTHOR CONTRIBUTIONS GCN, MB, AB, GCLL performed the experiments. EDB, LLM, GCN, MB AB contributed to the experimental design. EDB, LLM contributed with resources. GCN, MB, AB, EDB, LLM wrote and edited the manuscript. GCN, MB performed data acquisition and analysis. All authors have approved the final manuscript.
DATA AVAILABILITY STATEMENT: The data that support the findings of this study are available from the corresponding author upon reasonable request.
FUNDING STATEMENT: The study was supported by the São Paulo State Foundation for the Support of Research (FAPESP, Brazil; Grants 2014/25029-4 and 2017/24304-0). EDB is a recipient of grants from the National Council for Scientific and Technological Development (CNPq, Brazil). EDB is a CNPq research fellow. MB was recipient of FAPESP, Brazil, fellowship 2016-06602-0). GCN was recipient of FAPESP, Brazil, fellowships 2015/03053-3 and 2018/05146-7 and PNPD CNPQ, 88882.317597/2019-01. LLM: Research Grants from the MICINN (PID2019-105218RB-I00).