The MPTP administration pattern induces different glial activation in the striatum and in the midbrain

The mechanism of neuronal death induced by MPTP involves mitochondrial dysfunction and oxidative stress [33]. In this study we questioned whether the same neurotoxic stimuli, MPTP, would induce different glial activation profiles depending on the administration pattern. For this purpose, we selected two widely used MPTP administration regimens to generate experimental PD in mice. In the subacute regimen (sMPTP), 5 doses of MPTP (30 mg/kg) are injected daily for 5 consecutive days and in the chronic regimen (cMPTP), 10 doses of the neurotoxin (20 mg/kg) are co-administered with probenecid to prevent renal clearance, twice a week along 5 weeks (Fig. 1A). The different administration patterns would induce dopaminergic neuron degeneration at different rates. Upon sacrifice of the animals, the striatum and the ventral midbrain were dissected out to prepare a cell suspension from each region. CD11b+ cells and ACSA2+ astrocytes were subsequently separated for bulk mRNA sequencing. Due to MPTP-induced degeneration, glial cells presented different activation profiles in each region that varied with the MPTP intoxication regimen (Fig. 1B). CD11b+ cells showed increased differentially displayed transcripts in the midbrain compared to the striatum and these differences were exacerbated with the chronicity of MPTP administration (Fig. 1B). By contrast, ACSA2+ astrocytes showed profound rearrangements in gene expression in the striatum that decreased with the chronicity of MPTP administration (Fig. 1B). The low number of commonly triggered genes upon damage for each cell type indicates that glial cells react differently to the type or the intensity of damage.

Fig. 1

Differentially displayed transcripts in CD11b+ and ACSA2+ cells purified from the striatum and the midbrain after subacute or chronic MPTP intoxication. (A) Scheme of the two MPTP intoxication protocols used. In the subacute regimen (sMPTP), mice received 5 consecutive MPTP doses and in the chronic administration (cMPTP) animals received 10 MPTP doses along 5 weeks. Control mice were injected with the MPTP vehicle or MPTP vehicle with probenecid following the same MPTP administration pattern. sMPTP mice were sacrificed 24 h and cMPTP 48 h after the last MPTP injection. CD11b+ and ACSA2+ were purified from the striatum and the ventral midbrain for bulk RNA sequencing. (B) Venn diagrams representing differentially displayed genes (p < 0.01) in sMPTP and cMPTP animals with respect to their corresponding controls. (N = 3 animals/group). Str: striatum; Mdb: ventral midbrain

Next, we questioned whether the transcriptomic profile of glial cells in the context of nigrostriatal damage presented common features with other pathological conditions. Different signatures were selected for a gene set enrichment analysis (GSEA). The activation state of CD11b+ cells was compared with the neurodegenerative microglia (MGnD, ) which was established from different mouse models of neurological diseases including ALS, AD and multiple sclerosis (MS) [34]; with disease-associated microglia (DAM) described in an AD and ALS transgenic mouse models [35]; and with a specific subset of phagocytic microglia described in the same AD model [36](Table 1). The differentially displayed genes in CD11b+ cells from parkinsonian animals failed to show a MGnD profile except for the downregulated transcripts in the midbrain of sMPTP mice. DAM microglia were identified exclusively in striatal CD11b+ cells from cMPTP mice. Striatal and midbrain CD11b+ cells from sMPTP mice showed downregulation of genes that are also reduced in DAM, which may be attributed to the decreased expression of homeostatic microglial genes associated with neuronal loss [37] (Table 1). Phagocytic CD11b+ cells similar to that described in AD models were identified in the midbrain of cMPTP mice. In the striatum, CD11b+ cells shared upregulated, but not downregulated, transcripts with the microglial phagocytic phenotype (Table 1). The absence of a phagocytic profile in CD11b+ cells in the midbrain of sMPTP mice would reflect different activation patterns in response to the type of neuronal death [33, 38]. To delve into the phagocytic phenotype, we used an additional gene set obtained from gene ontology (GOBP_Phagocytosis, GO0006909, MM4904). Only striatal CD11b+ cells of cMPTP mice shared transcripts with this signature, the low correlation with the rest of CD11b+ cell subsets suggests that specific phagocytic processes are triggered by neurodegeneration or that these cells are at different stages of the process.

Table 1 p values corresponding to the GSEA analysis performed to compare striatal and midbrain differentially displayed genes in CD11b+ cells with signatures that represent different microglial states

The activation profile of ACSA2+ cells obtained from parkinsonian animals was compared to an LPS-induced neurotoxic phenotype and to a beneficial phenotype identified in ischemia [39]. Midbrain astrocytes in sMPTP mice presented the neurotoxic phenotype not detected in the striatum or in cMPTP mice. By contrast, the expression profile of striatal astrocytes correlated strongly with the GOBP phagocytosis gene set (Table 2).

Table 2 p values corresponding to the GSEA analysis performed to compare striatal and midbrain differentially displayed genes in ACSA2+ cells with signatures that represent different astroglial states

Selection of genes from the phagocytic microglia and the GOBP Phagocytic gene sets expressed in CD11b+ and ACSA2+ cells allowed us to elaborate a heatmap that confirms the phagocytic profile except for CD11b+ cells in sMPTP mice (Suppl. Figure 1). These results suggest that degeneration of the nigrostriatal pathway polarizes ACSA2+ cell activity towards a neuroinflammatory and phagocytic phenotype. CD11b+ cells are involved in the removal of striatal terminals and neuronal cell bodies in the cMPTP.

To gain further insight into the activation state of glial cells under parkinsonian conditions, we generated a network of biological functions, pathways and upstream regulators predicted to be altered for each differentially displayed gene set with Ingenuity Pathway Analysis (IPA). The network obtained for the striatum of sMPTP mice showed that most of the predicted pathways activated in CD11b+ cells are related to phagocytosis/removal of pathogens (Daam1, Agpat1, Siglec8), and to the recruitment of immune cells/production of pro-inflammatory signals (Ilr1, Cxcl10, HSP90AA1, Map3k8, Grk2, Ptgs1) (Fig. 2A). ACSA2+ cells in this region are also involved in inflammation, the downregulation of pro-inflammatory pathways such as IL7, IL17, IL18 or toll like receptor (TLR), together with the activation of TNFR1 and inhibition of TGFβ signaling, suggest that these cells are playing a relevant role in containing the pro-inflammatory reaction (Fig. 2B). Additional activated pathways revealed that astrocytes are experiencing profound morphological changes (Actb, Actn1, Pak1) and are adapting to changes in dopamine metabolism (MaoA, MaoB, Adora1) (Fig. 2B). In the midbrain, the predicted decrease in Ilr2, Cxcr2, Ccl6, Cd209b, Itgal, Clec4D, Ccnd3, Coro1A and the increase in Bcl6 signaling in CD11b+ cells suggests that they are involved in the downregulation of pro-inflammatory signals and adaptive immune cell infiltration (Fig. 2C). In the same context, ACSA2+ cells undergo a reprogramming process that affects astrocyte-neuron interactions (Gria2, Map2, Syngap1, Acap2), activates a phagocytic program (Mertk, Tyro3), promotes pro-inflammatory responses (S100A4) and protective inflammatory pathways mediated by IL15 and regulated by Aire. Concomitantly, classical neuroprotective signaling pathways such as the nerve growth factor and the WNT/β-catenin are downregulated (Fig. 2D).

Fig. 2

Graphical summary of networks predicted to be altered in glial cells of the sMPTP mice. Ingenuity pathway analysis yielded networks of pathways, upstream regulators and biological functions predicted to be altered (z < 1 or z > 1) and differentially expressed genes (p < 0.01) in: (A) striatal CD11b+, (B) striatal ACSA2+, (C) midbrain CD11b+ and (D) midbrain ACSA2+ cells

An equivalent analysis carried out in cMPTP mice revealed that, although the mediators are different, striatal CD11b+ cells also exhibit a phagocytic (increased Fgcr3, decreased Birc6 signaling) and pro-inflammatory (increased Ccrl2, Cd86, Fgr, S100a8) functionality (Fig. 3A). ACSA2+ cells expressed mainly pathways related to astrocyte/neuron communication (Grinc2c, Adcy5, Aph1a, Aifm1, Bace1, Adora1) and an increase of IL6 and CXCR4 signaling pathways (Fig. 3B). In the midbrain, immune modulating pathways that involve Cd300a, Cd93, Cd22 or Lag3 together with increased Cd44, Cxcl13, Ccr1, Hso90aa1, Vcam1, Gna13 and decreased Infgr1 or Ilr1l2 pro-inflammatory signals are altered in CD11b+ cells. Overall, the predicted signaling suggests that these cells are contributing to the inflammatory reaction and immune cell infiltration in this region (Fig. 3C). In the same context, ACSA2+ cells decreased pathways involved in synaptic plasticity and activated a pro-inflammatory profile through the increase in IL17 signaling and Ccl2, Il17, Chga, Cebpb, Dapp1 (Fig. 3D). Altogether, these results suggest that in the striatum CD11b+ cells are the main drivers of the pro-inflammatory reaction, independently of the MPTP intoxication pattern. In the midbrain of sMPTP mice, CD11b+ cells show an anti-inflammatory response that turns into pro-inflammatory upon damage chronification. Through different mediators, ACSA2+ cells exhibit a loss of homeostatic functions and a pro-inflammatory profile in the midbrain of sMPTP and cMPTP mice.

Fig. 3

Graphical summary of networks predicted to be altered in glial cells of the cMPTP mice. Ingenuity pathway analysis yielded networks of pathways, upstream regulators and biological functions predicted to be altered (z < 1 or z > 1) and differentially expressed genes (p < 0.01) in: (A) striatal CD11b+, (B) striatal ACSA2+, (C) midbrain CD11b+ and (D) midbrain ACSA2+ cells

Involvement of glial cells in the phagocytosis of dopaminergic cell debris

To delve into the differential phagocytic profile of CD11b+ cells in the SNpc, we prepared a new set of animals which was processed for histological techniques. Mice receiving the sMPTP experienced motor impairment measured in the rotarod, bar test and pole test (Fig. 4A). A decrease in TH+ terminals that could reflect the loss of dopaminergic terminals was detected in the striatum of sMPTP animals (Fig. 4B). Phagocytosis, quantitated as the fraction of TH+ signal measured inside Iba1+ cells, increased significantly in this region (Fig. 4B). In the SNpc, the significant decrease in TH immunostaining was not accompanied by an increase in the phagocytic activity (Fig. 4C). This data correlate with the absence of a phagocytic transcriptomic profile in CD11b+ cells (Table 1). The increased number and morphological changes of Iba1+ cells in the SNpc (Fig. 4D) indicate that these cells are reacting to dopaminergic damage. Iba1+ were classified based on their shape as ramified, hypertrophic and bushy (Suppl. Figure 2). Control animals presented a predominant subset of ramified Iba1+ cells and a minor population of hypertrophic cells. Upon sMPTP administration, Iba1+ cells polarized towards a hypertrophic phenotype accompanied by a decrease of the ramified morphology and a new population of bushy cells was detected mainly in the SNpc (Fig. 4D).

Fig. 4

Phagocytic assessment of Iba1+ cells in sMPTP mice. Animals were sacrificed at 10 days after the first MPTP injection. (A) Motor behavior was evaluated in the rotarod, bar and pole test. (B and C) Representative images of a single plane and the corresponding orthogonal projection of the stack of TH/Iba1 double immunofluorescence in (B) the striatum and (C) the SNpc of control and sMPTP mice. Quantitation of TH+ volume and percentage of TH+ signal in Iba1+ cells. (D) Quantitation of the number of ramified, hypertrophic and bushy Iba1+ cells in the striatum and SNpc from control and sMPTP mice. Data represent the mean ± 95% CI from 7 animals per group. Statistical analysis: (A, B and C) t-test with Welch correction when variances differed significantly, (D) 2-way ANOVA followed by Bonferroni post hoc test. Statistical significance in, (*) cell number, (#) cell morphology. */#p < 0.05, **/##p < 0.01, ***/###p < 0.001. Magnification bars: 10 μm

Mice receiving the cMPTP regimen exhibited motor impairment in the rotarod, bar test and pole test (Fig. 5A), and a decreased TH immunostaining in the striatum (Fig. 5B) and in the SNpc (Fig. 5C) corresponding to a damaged nigrostriatal pathway. Phagocytic activity of Iba1+ cells increased in the striatum (Fig. 5B) and in the SNpc (Fig. 5C). At this stage, the number of Iba1+ cells remained constant in both regions, independently of the parkinsonian conditions, and no morphological changes were appreciated in the striatum (Fig. 5D). By contrast, an increased fraction of Iba1+ cells in the SNpc showed a hypertrophic and bushy morphology together with a reduction of the ramified phenotype (Fig. 5D). These results suggest that phagocytosis of neuronal cell bodies in the SNpc is a late event.

Fig. 5

Phagocytic assessment of Iba1+ cells in cMPTP mice. Animals were sacrificed at 5 weeks after the first MPTP injection. (A) Motor behavior was evaluated in the rotarod, bar and pole test. (B and C) Representative images of a single plane and the corresponding orthogonal projection of the stack of TH/Iba1 double immunofluorescence in (B) the striatum and (C) the SNpc of control and cMPTP mice. Quantitation of TH+ volume and percentage of TH+ signal in Iba1+ cells. (D) Quantitation of the number of ramified, hypertrophic and bushy Iba1+ cells in the striatum and SNpc from control and cMPTP mice. Data represent the mean ± 95% CI from 7 animals per group. Statistical analysis: (A, B and C) t-test with Welch correction when variances differed significantly, (D) 2-way ANOVA followed by Bonferroni post hoc test. Statistical significance in (#) cell morphology. #p < 0.05. Magnification bars: 10 μm

Next, we questioned whether the different phagocytic activity of Iba1+ cells in the SNpc was determined by the MPTP administration pattern or by the time of sacrifice. Thus, we prepared a new set of animals following the cMPTP regimen that was sacrificed 10 days after the first MPTP injection (3 MPTP/probenecid doses). We termed this group “early chronic MPTP” (ecMPTP). At this point, mice experienced a mild motor impairment that was detectable in the rotarod and in the pole test (Fig. 6A). The lack of differences in TH immunostaining in the striatum (Fig. 6B) and in the SNpc (Fig. 6C) indicated that the nigrostriatal pathway was still preserved. Phagocytic Iba1+ cells were detected in the striatum (Fig. 6B), but not in the SNpc (Fig. 6C), corroborating the results obtained in the sMPTP suggesting that Iba1+ cells do not activate an early phagocytic program in the SNpc after MPTP damage. The increased number of cells in both regions and the polarization of Iba1+ cell morphology towards a hypertrophic phenotype (Fig. 6D) indicate that these cells sense alterations in the nigrostriatal pathway even in the absence of a noticeable decrease in TH expression. An additional Sholl analysis performed to determine Iba1+ cell ramification showed that in the striatum, microglia from cMPTP mice presented the highest degree of ramification (Suppl. Figure 3A). By contrast, in the SNpc the highest extension of microglia branches was detected in the ecMPTP condition (Suppl. Figure 3B). These results agree with the appearance of a new bushy phenotype with a low level of ramifications. Altogether, these results indicate that morphological changes in microglia reflect a change in the basal activation state, but do not correlate with the phagocytic activity.

Fig. 6

Phagocytic assessment of Iba1+ cells in ecMPTP mice. Animals were sacrificed at 10 days after the first MPTP injection following the cMPTP intoxication pattern. (A) Motor behavior was evaluated in the rotarod, bar and pole test. (B and C) Representative images of a single plane and the corresponding orthogonal projection of the stack of TH/Iba1 double immunofluorescence in (B) the striatum and (C) the SNpc of control and ecMPTP mice. Quantitation of TH+ volume and percentage of TH+ signal in Iba1+ cells. (D) Quantitation of the number of ramified, hypertrophic and bushy Iba1+ cells in the striatum and SNpc from control and ecMPTP mice. Data represent the mean ± 95% CI from 7 animals per group. Statistical analysis: (A) Mann-Whitney test for the rotarod and bar test, t-test for the pole test; (B and C) t-test, (D) 2-way ANOVA followed by Bonferroni post hoc test. Statistical significance in, (*) cell number, (#) cell morphology. */#p < 0.05, **/##p < 0.01. Magnification bars: 10 μm

Finally, the phagocytic activity of astrocytes was studied under the different MPTP administration/time regimens. Since striatal GFAP immunoreactivity is almost absent in control animals, only parkinsonian mice were selected for this analysis. Astrogliosis, measured as the volume of GFAP+ signal, was similarly observed in the striatum of the different models of parkinsonism (Fig. 7A). An active phagocytic process was detected at early stages of dopaminergic terminal impairment, in the sMPTP and ecMPTP mice, but not when damage was established (cMPTP) (Fig. 7A). The lack of GFAP immunoreactivity in the SNpc of ecMPTP and cMPTP mice (Fig. 7B) did not allow to carry out a similar analysis in this region. In summary, our data show that both, Iba1+ and GFAP+ cells, participate in the removal of impaired striatal terminals at the onset of the neurodegenerative process (sMPTP and ecMPTP). As the process chronifies, only Iba1+ cells maintain their phagocytic activity. In the SNpc the scenario is different, a lack of phagocytic microglia is observed in the sMPTP (in agreement with the GSEA analysis) and ecMPTP mice. The transcriptomic profile of ACSA2+ cells shows an increased expression of the phagocytic receptor Mertk suggesting that astrocytes may initiate the process of removal of damaged cells in this region. Polarization of microglial activity towards a phagocytic phenotype is a late event in the SNpc.

Fig. 7

Phagocytic assessment of astrocytes in MPTP mice. Animals were sacrificed at 10 days (sMPTP and ecMPTP) or at 5 weeks (cMPTP) after the first MPTP injection. (A) Representative images of a single plane and the corresponding orthogonal projection of the stack of TH/GFAP double immunofluorescence in the striatum of sMPTP, ecMPTP and cMPTP mice. Quantitation of GFAP+ volume and percentage of TH+ signal in GFAP+ cells. Dashed line corresponds to the mean of GFAP+ volume in control animals. (B) Representative images of a single plane and the corresponding orthogonal projection of the stack of TH/GFAP double immunofluorescence in the SNpc of sMPTP, ecMPTP and cMPTP mice. Quantitation of GFAP+ volume. Dashed line corresponds to the mean of GFAP+ volume in control animals. Data represent the mean ± 95% CI from 7–8 animals per group. Statistical analysis: (A) One-way ANOVA test. **p < 0.01, ***p < 0.001. Magnification bars: 10 μm

Immune cell infiltration in the striatum and in the ventral midbrain

Next, we questioned whether the inflammatory profile of CD11b+ cells in the striatum and ventral midbrain could be related to immune cell infiltration. In parallel we investigated the source of the increased number of Iba1+ cells. Thus, we prepared cell suspensions from the striatum and the ventral midbrain that were analyzed by flow cytometry. CD11b+ cells were classified based on the cell surface expression of CD11b and CD45, the CD11b+CD45low gate was ascribed to resident microglia and the CD11b+CD45high to infiltrated myeloid cells and resident microglia (Fig. 8A) that, under pathological conditions, are known to upregulate CD45 expression [40, 41]. In line with the Iba1+ cell counts, the proportion of CD11b+CD45low, but not CD11b+CD45high, was significantly increased exclusively in the midbrain of sMPTP animals. The significant difference in Ki67+ cells in this region suggests that cell proliferation, rather than monocyte/macrophage infiltration from the periphery, may account for this increase (Fig. 8B). The increase in Ki67+ observed in striatal CD11b+CD45low and in CD11b+CD45high midbrain cells, that was not accompanied by an alteration in cell subsets (Fig. 8B), may indicate an accelerated turnover. However, a significant increase in the fraction CD11b+CD45low in the striatum and CD11b+CD45high in the midbrain of ecMPTP mice could be due to cell proliferation of CD11b+CD45high subset (Fig. 8C). The subpopulations of CD11b+ and Ki67+ cells remained constant in cMPTP animals (Fig. 8D). These results validate our histological observation of an expansion in the number of Iba1+ cells in SNpc in the early stages of the degeneration, which could be explained by the proliferation of CD11b+CD45low or CD11b+CD45high cells.

Fig. 8

Flow cytometry analysis of CD11b+ cells from the striatum and in the ventral midbrain after MPTP intoxication. (A) Gating strategy for CD11b+ cell subsets differentiated based on the level of CD45 expression, CD11b+CD45low and CD11b+CD45high, and Ki67 expression. The fraction of CD11b+CD45low and CD11b+CD45high cells out of viable cells and frequency of Ki67+ cells in these populations were analyzed in the striatum and midbrain of (B) sMPTP, (C) ecMPTP and (D) cMPTP mice. Data represent the mean ± 95% CI from 6 to 8 animals/group. Statistical analysis: (B-D) t-test for data following a normal distribution and Mann-Whitney test for data not following a normal distribution. *p < 0.05, **p < 0.01

Many of the biological functions predicted in the IPA analysis to be activated in glial cells were related to immune system infiltration and inflammation. To determine the output of these inflammatory signals, we examined T-cell infiltration within each region. The gating strategy is illustrated in Fig. 9A. Predicted pathways in striatal ACSA2+ cells and in midbrain CD11b+ cells of sMPTP mice pointed towards a downregulation of pro-inflammatory signals that could be responsible for the decreased infiltration of CD4 and CD8 T cells at this stage of degeneration (Fig. 9B). An increased CD4 T cell infiltration was present in the ecMPTP mice (Fig. 9C). The pro-inflammatory pathways activated in the midbrain CD11b+ cells after cMPTP administration may underly the significant increased CD4 T cell infiltration observed in this region (Fig. 9D), suggesting that lymphocyte infiltration is an early event under mild degenerative conditions. Altogether, these results indicate that the inflammatory signals generated in the context of MPTP intoxication correlate with the intensity of neuronal injury. Acute and intense dopaminergic impairment in the sMPTP mice would generate an anti-inflammatory response, while mild and chronic damage observed in cMPTP animals would induce a pro-inflammatory response detectable even at the early stages of the neurodegenerative process.

Fig. 9

Flow cytometry analysis of lymphocyte infiltration in the striatum and in the ventral midbrain after MPTP intoxication. (A) Representative gating strategy to select CD11b−CD45high population, and then CD4+ and CD8+ cells in cells suspension prepared from the ventral midbrain. Fraction of CD4+ and CD8+ cells out of viable cells in the striatum and midbrain of (B) sMPTP, (C) ecMPTP and (D) cMPTP mice. Data represent the mean ± 95% CI from 6 animals/group. Statistical analysis: (B-D) t-test for data following a normal distribution with Welch correction when variances differed significantly. Mann-Whitney test for data not following a normal distribution. *p < 0.05

Activation state of microglia and astrocytes in the human brain with Parkinson´s disease

The subacute and chronic MPTP administration protocols induce different glial reactivity, inflammatory profile and immune cell infiltration into the midbrain parenchyma. To understand which activation pattern reflects better the state of these cells in PD, we took advantage of single nuclei RNA sequencing (snRNA-seq) published dataset obtained from the midbrain of 6 idiopathic PD patients with severe neuronal loss and 5 aged-matched control subjects [31]. Differentially displayed transcripts from microglia and astrocytes were analyzed as data obtained from MPTP mouse models. The IPA network for the differentially displayed genes expressed by cells annotated as microglia (CD74) and astrocytes (AQP4) was generated following the same criteria used for the MPTP pre-clinical models. Predicted activated pathways in microglia showed a clear pro-inflammatory profile related to Th1 immune pathways and TNF, IL1, IL6, IL8, CCL20, TLR or CXCR4 signaling and downregulation of IL10 (Fig. 10A). Astrocytes in the midbrain also presented a pro-inflammatory state with activation of TNF, IFNG, IL1, IL8, IL17, CD44, CD40 or CXCR4 pathways and inhibition of IL10 or CTLA4 signaling (Fig. 10B). Next, we compared the upstream regulators and biological functions predicted to be modulated in the midbrain of MPTP models and in PD patients. Human and cMPTP CD11b+ cells shared most of their upstream regulators and functions related to the activation of the immune system and the unfolded protein response (Fig. 10C). The inflammatory activation state of sMPTP CD11b+ cells differed from their counterparts in PD and cMPTP. Common pathways shared by PD and MPTP mice were related to phagocytosis (Fig. 10C). Predicted upstream regulators in human astrocytes showed a good correlation with the cMPTP astrocytes and a poor correlation with sMPTP cells (Fig. 10D). The biological functions with the best Z-score in human astrocytes were related to cytoskeleton and morphology. However, they were absent or underscored in the MPTP experimental models (Fig. 10D). The high density of dopaminergic cell bodies in the mouse SNpc compared to the equivalent human region may account for these differences. Human and cMPTP astrocytes shared activated pathways related to immune cell infiltration into the brain, and human and sMPTP astrocytes shared functions related to nuclear factor of activated T cells (NFAT) and IL15 (Fig. 10D). CD11b+ and ACSA2+ cell activation in the experimental models failed to upregulate pathways related to B lymphocytes that were present in the human samples (Fig. 10C and D). Since human samples correspond to advanced stages of the disease (Braak stages 5–6) it could be expected a higher similarity with the cMPTP model. Therefore, these results validate the cMPTP as a representative model of a late stage PD.

Fig. 10

Summary of the networks predicted to be altered in human microglia and astrocytes in the midbrain. Differentially displayed genes in human glia obtained from the midbrain of PD patients were subjected to Ingenuity Pathway Analysis (IPA). The graphical summary represents the pathways, upstream regulators and biological functions predicted to be altered (A) in microglia and (B) astrocytes. The results of IPA analyses of human, sMPTP and cMPTP glial cells in the midbrain were subjected to Ingenuity Comparison Analysis to compare the z-scores of upstream regulators, pathways and biological functions in (C) microglia and (D) astrocytes