Background
Parkinson’s disease (PD) is a progressive neurodegenerative disease which is characterized by the selective degeneration of dopaminergic (DA) neurons of the substantia nigra pars compacta (SN), whereas the neighboring DA neurons of the ventral tegmental area (VTA) are relatively spared [
1]. Previous studies have examined the DA neurons of these two regions to better understand why only one subset of DA neurons is significantly more vulnerable to PD than the other [
2‐
5]. Recently, greater focus has been placed on extrinsic mechanisms, such as the role of astrocytes, in the disease process suggesting that neuronal vulnerability results from both intrinsic and extrinsic influences. Indeed, our recent work (Kostuk et al. 2018,
in submission) has demonstrated a vast transcriptional difference between SN and VTA astrocytes such that VTA, but not SN, astrocytes release factors which mediate protection of VTA, SN, and induced pluripotent stem cell (iPSC) DA neurons. In addition to astrocytes, the role of microglia, another important glial cell type, in the pathogenesis of PD has become an area of great interest.
Microglia are considered to be the innate immune cells of the brain [
6]. Their role in the clearance of damaged cells, such as neurons, and other foreign contaminants of the brain is well established [
7‐
9]. It is well known that microglia derive from a myeloid lineage [
10] and that injury or disease can cause increased microglial activation within the brain [
11‐
14]. Indeed, in PD patients and animal models of PD, greater numbers of microglia [
15], as well as enhanced microglial reactivity [
16,
17], are observed in the SN. In addition to their role as scavengers in the brain, microglia also release pro- and anti-inflammatory molecules in response to insult or injury [
9,
18,
19]. Indeed, the canonical microglial activator lipopolysaccharide (LPS) has been used to stimulate microglial activation and the subsequent cytokine release in order to model the immune response in vivo and in vitro [
12,
20‐
22]. Thus, it has been reported that LPS stimulation of microglia within the SN is sufficient to produce DA neurodegeneration [
23], suggesting a role for microglia and their factors in disease progression.
Previous studies have additionally examined the interaction of microglia with PD mimetic toxins. For example, rotenone, a pesticide known to be a DA neurotoxin, was shown to require the presence of microglia in order to mediate toxicity of midbrain DA neurons in culture [
24]. This raises the possibility that the interaction of PD mimetic toxins with microglia could be a potential pathway by which neurodegeneration occurs. In addition to the potential role that toxins could have on microglia, leading to an exacerbation of toxicity, the ratio of microglia to neurons within the brain could potentially produce different reactions to inflammatory stimuli. Previous work has shown that the regional distribution of microglia throughout the brain is diverse [
25,
26], with the SN being populated by a high ratio of microglia to neurons. Furthermore, recent transcriptional profiling of discrete brain regions (cortex, hippocampus, cerebellum, striatum) has suggested that similar to astrocytes, microglia exhibit regionally distinct transcriptional profiles [
27]. However, the potential regionality of midbrain microglia has yet to be investigated.
Therefore, in this study, we sought to determine whether midbrain SN and VTA microglia exhibit regionally specific gene expression profiles and whether these regionally isolated microglia can contribute differentially to the susceptibility of SN DA neurons to MPP+-induced toxicity as compared to the VTA. Furthermore, we investigated whether the ratio of microglia to neurons to astrocytes within the midbrain plays a role in the selective vulnerability of the SN to PD mimetic toxins. We found that regionally isolated microglia from three distinct brain regions, the cortex, SN, VTA, exhibit differential basal cytokine profiles. Furthermore, when challenged with the PD mimetic toxin MPP+, all regionally isolated microglia respond with distinct cytokine profiles. Interestingly, the addition of microglia alone is not sufficient to induce DA neuron death in cultures of either SN or VTA neurons. However, when challenged with MPP+, all regionally isolated microglia similarly exacerbate MPP+ toxicity, and this exacerbation of toxicity can be alleviated by inhibiting the activation of microglia via pharmacological inhibition of the TLR4 receptor. Finally, we demonstrated that greater ratios of VTA, but not SN, astrocytes are able to protect both SN and VTA DA neurons from MPP+ toxicity and that increasing the microglial ratio counters the protective effect of VTA astrocytes. Together, these results suggest that the susceptibility of SN DA neurons to a PD mimetic toxin results from a regional sensitivity to multiple environmental factors, rather than simply from an intrinsic neuronal vulnerability. A greater appreciation of the role of extrinsic mechanisms, mainly the protective nature of astrocytes and deleterious functions of microglia, in disease pathogenesis is important for the consideration of possible therapeutics in PD.
Discussion
The regional vulnerability of SN DA neurons has been extensively studied for a number of years, with many studies focusing on the intrinsic underlying mechanisms and how these neurons differ from neighboring VTA neurons [
2‐
5]. Recently, however, greater emphasis has been placed on understanding potential non-cell autonomous mechanisms, such as the impact of microglial cells, which may underlie this selective vulnerability.
In this study, we demonstrate that regionally isolated microglia of midbrain SN and VTA subregions differ in both their basal and MPP+-induced cytokine profiles. Furthermore, we demonstrate that the addition of regionally isolated microglia themselves is not deleterious to DA neurons; however, their presence significantly exacerbates MPP+ toxicity in SN DA neurons. Conversely, VTA DA neurons co-cultured with astrocytes are robustly protected from MPP+ toxicity despite the presence of microglia, suggesting that VTA astrocytes may be responsible for protection of these neurons. We further showed that pharmacological inhibition of microglial activation diminished the exacerbatory effect of microglia. Finally, we demonstrated that VTA astrocytes, but not SN, are able to protect both SN and VTA DA neurons from MPP+ toxicity despite the presence of deleterious microglia but that it is possible to overburden protective astrocytes with increasing numbers of microglia, thus increasing the presumptive load of microglial released factors.
The role of microglia as the immune cells of the CNS has been appreciated for a number of years [
6]. Many studies have demonstrated an increased release of pro-inflammatory/neurotoxic species from microglia in response to inflammatory stimuli such as LPS [
18]. Additionally, some studies have postulated a role for microglia in the progression of PD [
11,
32,
33], though a debate continues as to whether microglia play a role in initiation of disease or are activated secondary to neuronal death. Supporting the latter proposition, there are a number of reports demonstrating a response of microglia to some PD toxins. Gao et al. [
24] showed that the effect of rotenone, a pesticide that has been used as a PD mimetic toxin in culture, requires the presence of microglia and that this effect is dependent on the number of microglia present within the culture. Additionally, Chien et al. [
34] demonstrated that microglial released cytokines and chemokines are responsible for the loss of nigrostriatal DA neurons in response to LPS. Furthermore, recent work by Smeyne et al. [
16] demonstrated that systemic injection of the PD mimetic toxin MPTP results in increased microglial reactivity within the SN, suggesting that PD mimetic toxins can directly cause microglial activation and subsequent cytokine release. These findings are similar to ours in which microglial addition to DA neuron cultures exacerbates the effect of the PD mimetic toxin MPP
+, suggesting that the release of pro-inflammatory molecules from microglia may have a vital role in the progression of PD DA neuron degeneration.
Previous studies have also demonstrated that the inhibition of microglial activation or knockdown of released cytokines from microglia may be neuroprotective to DA neurons. Pharmacological blockade of microglial activation [
35‐
37] or knockdown of cytokines, such as IL-1 [
38], demonstrates that prevention of microglial activation attenuates some of the degeneration seen in models of PD. Furthermore, the interaction of MPTP/MPP
+ with the microglial receptor TLR4, a key receptor in the microglial activation pathway, has previously been established [
39,
40]. As such, it is unsurprising that our findings demonstrate that the use of a new TLR4 inhibitor C34 [
31], prior to MPP
+ administration, results in decreased exacerbation of toxicity.
In addition to an effect of the PD mimetic toxins MPTP or MPP
+ on the activation of microglia, a number of studies have examined the impact of alpha synuclein (α-syn) on microglia and the subsequent activation of these cells [
33,
41‐
43]. Indeed, studies have demonstrated that α-syn is able to bind to surface receptors of microglia and result in their activation and subsequent upregulation of pro-inflammatory cytokines [
44‐
46]. Therefore, it is plausible that the interaction of α-syn and microglia of the midbrain may be exacerbating or contributing to DA neuronal degeneration similar to that seen by the effect of MPP
+ in our studies. However, it is still relatively unknown whether this interaction of α-syn and microglia is an initiating factor for neuronal death or secondary to the release of species from damaged or dying neurons [
43]. Furthermore, a number of studies have demonstrated that microglia can act as scavengers of extracellular α-syn [
47,
48], thereby clearing a potentially damaging species away from vulnerable neurons. The impact of microglial-sequestered α-syn is not fully understood, and it is possible that internalized α-syn could result in dysfunction of lysosomal pathways within microglia, similar to that seen in DA neurons [
49], impairing normal functions within these cells. Therefore, the effect that α-syn may have on microglia from the SN and VTA remains an important area of investigation.
The regional variability in the transcriptional profile of microglia has been investigated previously [
27]. However, potential differences of microglia within the midbrain have yet to be investigated. Our study is the first of our knowledge to sub-dissect the midbrain into the vulnerable SN and protected VTA to examine the effects of isolated microglia of these regions in a PD model. Though we demonstrate a number of basal and MPP
+-induced cytokine differences, it is interesting to note that we fail to find any regionally specific exacerbation of MPP
+ toxicity. A number of possible explanations exist for this fact.
First, our investigation of the transcriptional profile of these microglia was somewhat limited. We established a list of 21 canonical pro- and anti-inflammatory cytokines based on commercially available arrays. This is a very limited view of the transcriptional profile of these cells. As such, future studies utilizing more in-depth methods of investigation, such as RNAseq, are required. These methods have been used in other models, such as spinal cord injury [
50], and have demonstrated a vast array of subtle injury-induced changes to microglia that should likewise be appreciated in the context of the midbrain in PD.
Secondly, the deleterious effect of these regionally isolated microglia could simply be due to a combined CCL3, CCL4, and TNFα response, as these are the three pro-inflammatory cytokines upregulated in all three regionally isolated microglial populations. However, the role of these cytokines has been investigated extensively in the context of PD [
51‐
53], though the specific action of SN or VTA microglial-derived cytokines is still unknown. A possible course of investigation would be to selectively inhibit the production of these cytokines, or their downstream pathways, directly within these distinct midbrain nuclei and examine their effects in an in vivo PD model.
Finally, the response of microglia to inflammatory stimuli, such as LPS, is understood to be temporally dynamic. Indeed, previous studies have demonstrated that changes in the expression level of known pro-inflammatory cytokines can change even within 4 h of stimulus initiation, and these changes can be transient and disappear by 24 h after stimulus [
54,
55]. Our examination of expression levels of cytokines was limited to the 24-h time point, as this is when we typically see the greatest death of DA neurons after MPP
+ treatment. However, we cannot rule out the effect of earlier cytokine responses and their impact on DA neuronal survival. Further RNAseq studies examining the temporal gene expression changes of isolated SN, VTA, and CTX microglia are required to fully understand the role these cells may have in the initiation or exacerbation of DA neuronal death in response to MPP
+.
The role of astrocytes as neuroprotective cells has been well established for a number of years [
56‐
58]. Indeed, our recent work (Kostuk et al. 2018,
in submission) has demonstrated a vast transcriptional difference between the astrocytes of the SN and VTA, suggesting that VTA astrocytes are able to mediate protection of their regional DA neurons whereas SN astrocytes lack this ability. Furthermore, the crosstalk between astrocytes, neurons, and microglia has recently become more appreciated [
59‐
61]. Thus, the interaction of midbrain astrocytes and microglia in the context of PD pathogenesis is an important avenue of investigation. Consequently, it is unsurprising to us that greater ratios of SN astrocytes fail to mediate greater protection when deleterious microglia are present and that VTA astrocytes robustly protect both SN and VTA DA neurons from MPP
+ toxicity despite the presence of microglia. Interestingly, however, we were able to demonstrate that it is possible to overburden protective astrocytes with greater numbers of microglia by increasing the ratio of microglia to neurons and likely their released cytokines in our cultures. This is of vital importance, as SN microglial activation and proliferation is a key feature of both human PD pathogenesis [
15,
62] and animal models [
16] of the disease. Therefore, as the concentration of microglial factors increases during the course of disease, less protective astrocytes in the SN may become more overtaxed by significant increases of microglial-released cytokines. In contrast, VTA astrocytes may be better suited to balance against these deleterious factors as there are fewer microglia and greater proportions of protective astrocytes present basally [
25,
26,
63,
64], and thus, microglial activation or proliferation may never reach the critical threshold needed to cause VTA DA degeneration.