3. Discussion
At the first stage of this study, which aimed to assess the expression of genes encoding proteins specific for DAergic neurons as well as proteins involved in the molecular mechanisms of neurodegeneration in the SN as affected by PD, it was necessary to select a PD model that would reproduce the progressive degradation of the nigrostriatal system. Such a subchronic model of PD was recently developed in our laboratory in mice with systemic MPTP administration [
6]. In this model, progressive degradation of the nigrostriatal DAergic system, which reproduced the continuous development of PD at the preclinical and clinical stages, was induced via repeated administration of MPTP at gradually increasing doses. MPTP was injected at intervals of 24 h, which corresponds to the time of neurodegeneration in response to a single injection of this neurotoxin. Two consecutive MPTP injections at doses of 8 mg/kg and 10 mg/kg reproduced the development of PD at the preclinical (asymptomatic) stage. Seven consecutive MPTP injections at doses of 8, 10, 12, 16, 20, 26, and 40 mg/kg, in that order, imitated the development of the clinical (symptomatic) stage of PD. In mice used in the subchronic PD model, as in patients with PD, the preclinical stage continuously passed to the clinical stage with the loss of 70–80% of DA in the striatum and the death of 38% nigral DAergic neurons. This was accompanied by the appearance of motor disorders [
6]. In this study, we have accurately reproduced this model. The difference in the concentration of DA in the striatum in repeated experiments did not exceed 5%.
The value of the results obtained from a model of any disease depends on how well the model reproduces the pathological processes that develop in patients. Therefore, in the second stage of this study, we compared the expression of genes encoding proteins specific for DAergic neurons, as well as selected proteins involved in intercellular signaling and neurodegeneration in the SN, in our mouse model of subchronic PD (our results)to that found in patients in the clinical stage of PD (published literature data). In the absence of a preclinical PD diagnosis, postmortem material for such an analysis cannot be obtained from patients at the preclinical stage.
According to the research literature, analysis of gene expression in postmortem material was carried out both in the whole SN [
7,
8,
9,
10,
11,
12,
13,
14,
15,
16,
17,
18,
19] and in individual DAergic neurons isolated by neuromelanin content [
8,
20,
21,
22,
23,
24]. Our data on gene expression in the SN in mice in the PD model were compared with the published data obtained from the SN of patients but not with those obtained from individual neuromelanin-containing neurons (
Table 2) because pathological material from patients can be obtained only many years after diagnosis and treatment. Obviously, the treatment itself can change the gene expression of proteins involved in pathological processes. Indeed, Tiklova and co-authors [
25] have shown that gene expression is significantly different between the early and late clinical stages of PD. In addition, researchers often do not provide information on patient treatments that can affect gene expression in the nigrostriatal system [
26,
27]. However, despite the difficulties in interpreting changes in gene expression in PD patients, this is the only approach that can be used to validate animal models, although with great caution.
Of all of the genes we analyzed in the SN of mice in the clinical stage model, only 20 genes were also analyzed in the SN of patients with PD (
Table 2). For 7 out of 20 genes in mice and patients, we found either a unidirectional change in expression or no change (
Th,
Ddc,
Map2,
Snca,
Psmc3,
Casp3, and
Mapk8). For 13 genes, opposite changes in gene expression were observed (
Maoa,
Dynll1,
Mapt,
Syt11,
Syt1,
Nsf,
Park2,
Ube2n,
Psmd4,
Ubb,
Gfap,
Vps35, and
Ctsb). As for the genes with unidirectional expression in the SN in the mouse model of the clinical stage of PD and in patients with PD, they are present in almost all of the selected gene clusters except for “Oxidative stress” and “Inflammation and glial activation” (
Table 2).
Changes in the expression of genes for proteins of the DAergic phenotype (
Th and
Ddc of the “DA synthesis and degradation” cluster) in the SN in our model of the clinical stage of PD coincided with changes in the expression of the same genes in patients with PD. In both cases, the expression of these genes was reduced [
7,
14,
28]. It should be noted that in mice in the acute model of the clinical stage of PD and in patients with PD, a direct relationship was shown between the expression of the tyrosine hydroxylase gene and the level of the tyrosine hydroxylase protein in the SN [
29,
30]. The content of tyrosine hydroxylase, as well as its activity in surviving neurons, remained at the control level [
29,
30,
31]. It is possible that the decrease in the gene expression of proteins characteristic of DAergic neurons that we found in this study is associated with a loss of DAergic neurons.
It is believed that the death of neurons in PD begins with a degradation of axonal terminals and followed by a retrograde spread of degradation up to the cell bodies, which is accompanied by impaired axonal transport [
32]. However, it is still not clear what the primary cause of neuronal death in PD is, whether it is the impaired functioning of axonal terminals or impaired axonal transport [
33,
34,
35]. Considering that analysis of gene expression in PD patients is carried out on pathological material obtained many years after the diagnosis of PD (by the appearance of motor symptoms) and the start of pharmacotherapy, it is impossible to determine which gene expressions change at an early clinical stage before the start of patient treatment. However, in the SN both in our mouse model of the clinical stage of PD (see Results) and in postmortem PD patients, a decrease in the expression of
Map2, a microtubule stabilizing protein, was found [
16]. These data suggest that the assembly/disassembly of microtubules is impaired at the initial stage of neurodegeneration, which leads to a disorganization of axonal transport and of the functioning of synaptic terminals. The decreased expression of genes for microtubule stabilizing proteins such as β-tubulin, α-tubulin, synapsin-1, and parkin [
36] found in the SN in our mouse model of the clinical stage of PD supports this assumption. Despite the fact that analysis of the gene expression of microtubule stabilizing proteins has not been performed in PD patients, our experimental data are a strong argument in favor of the assumption that neurodegeneration begins with a degradation of microtubules.
An important characteristic of the pathogenesis of PD is the formation of Lewy bodies, which contain proteins such as aggregated α-synuclein, tau protein, and ubiquitin [
37,
38]. Despite the accumulation of α-synuclein in patients with PD, most studies have shown that the expression of the α-synuclein gene does not change or even decrease compared with age-related controls [
14,
16,
17]. This result is confirmed in our mouse model of the clinical stage of PD. Indeed, we did not find any changes in the expression of the
Snca gene in these animals. It was shown that the aggregation of α-synuclein and its transformation first into oligomeric neurotoxic complexes and then into Lewy bodies is the result of impaired degradation of α-synuclein by the ubiquitin–proteasome system [
39,
40,
41]. This is confirmed by our data, which details a decrease in the expression of the
Psmc3 gene that encodes one of the proteasome subunits in the SN in our model of the clinical stage of PD. This has also been shown in patients with PD [
15]. It is important to note that one of the targets for the toxic action of oligomeric α-synuclein is constituted by microtubules, and that their degradation disrupts axonal transport. Although oxidative stress is believed to be the trigger for α-synucleinopathy [
42], we have not found any unidirectional changes in the expression of genes for antioxidant system enzymes that regulate the expression of these genes in the SN in mice in our model of the clinical stage of PD (see Results) and in patients with PD.
Despite a great interest in the death mechanisms of DAergic neurons of the nigrostriatal system in PD (apoptosis, necrosis, or autophagy), this question still remains open. Research literature data on this issue are contradictory. However, some studies claim that DAergic neurons die mainly due to apoptosis [
43,
44,
45,
46,
47,
48,
49]. Evidence that apoptosis is the main death mechanism of DAergic neurons is constituted by an increased expression of active caspase 3, the main inducer of apoptosis [
45,
46]. An increase in
Casp3 gene expression was confirmed in our subchronic model of the clinical stage of PD, which is consistent with the data obtained in other subchronic and chronic models that show neuronal death through apoptosis [
50,
51,
52]. In addition, both in patients with PD [
16] and in our model of the clinical stage of PD, no changes in the expression of the
Mapk8 gene, a protein that triggers autophagy, were found in the SN.
There are only a few studies evaluating gene expression in the SN when modeling PD in non-human primates using MPTP [
53]. Of the seven genes evaluated by Ohnuki et al. (2010) and in our study (
Th,
Ddc,
Ubb,
Gfap,
Drd2,
Mapt, and
Tubb3), the same changes in the expression of only four genes (
Th,
Ddc,
Ubb, and
Gfap) were observed. In addition, it was shown that changes in the expression of the
Mapt gene in the SN of non-human primates in the PD model and in the SN postmortem in PD patients coincide, whereas the changes in the expression of
Ubb and
Gfap genes differ. It follows that both of the above models do not fully reproduce the neurodegenerative changes in the SN that are characteristic of PD patients. Nevertheless, both models of PD are useful and complement each other in studying the mechanisms of the neurodegeneration and neuroplasticity of the nigrostriatal dopaminergic system.
Considering that the SN contains both neurons and glial cells, we have evaluated in the mouse model of the clinical stage of PD genes that are predominantly expressed in neurons. This has made it possible to compare our data with those obtained on neuromelanin-containing cells isolated via microdissection from the SN of PD patients (
Table 3).
It was shown that the expression of the five genes characteristic of neurons did not change in the SN in our mouse model of the clinical stage of PD (
Table 3). These data are partly consistent with the study of Zaccaria [
22], who did not find changes in the expression of three neuronal genes (
Drd2,
Kif1a, and
Kif5a) in the SN of patients, but they do not agree with the data obtained in other similar studies [
20,
21,
24]. As noted earlier, discrepancies in the published literature data on gene expression in patients with PD may be due to the use of pathological material obtained from patients with varying degrees of pathology as well as with different histories of pharmacotherapy [
25,
54]. Our failure to detect changes in the expression of kinesin genes (kinesins 1a and 5a), axonal transport proteins, and synaptotagmin 1, a vesicular cycle protein, in the SN of mice in our model of the clinical stage of PD supports our assumption that neurodegeneration is associated with impaired axonal transport and neurotransmission. It is possible that a disruption of the vesicular cycle in neurodegeneration predominates over a disruption of axonal transport. This is evidenced by the fact that all available studies show a decrease in the expression of the synaptotagmin 1 and 11 genes [
20,
21,
22,
23,
24], whereas this is not always the case for kinesin gene expression [
20,
22,
24].
Given that there is no data on gene expression in the nigrostriatal system at the preclinical stage in patients with PD due to a lack of preclinical diagnoses, the only way to conduct such studies is through the use of experimental models.
The development of a subchronic model of PD, which reproduces the progressive degradation of the nigrostriatal system, has made it possible to evaluate the expression of genes encoding DA-synthesizing enzymes and proteins involved in the molecular mechanisms of neurodegeneration in the SN in mice in models of the preclinical and clinical stages of PD (
Table 4).
According to the expression of genes in the SN of mice in our models of the preclinical and clinical stages, these genes can be divided into two groups. The first group includes genes whose expression does not change in the model of the preclinical stage but changes in the model of the clinical stage of PD compared with the control. The second group includes genes whose expression is downregulated in the model of the preclinical stage but does not change in the model of the clinical stage of PD compared with the control.
In the first case, changes in gene expression indicate the progression of the neurodegenerative process during the transition from the preclinical to the clinical stage. This is characteristic of genes encoding proteins involved in axonal transport (
Map2,
Tubb3, and
Tuba1a), synaptic vesicle cycle and neurotransmission (
Syn1), inflammation and glial activation (
Gfap), and cell death mechanisms (
Casp3, and
Parp1). It should be noted that the decrease in the gene expression of some neuronal proteins in the SN (e.g.,
Th) in mice in the clinical stage model could be due to the death of DAergic neurons rather than downregulation of gene expression [
54]. In the second case, a change in gene expression could indicate early neurodegenerative processes that develop in the SN at the preclinical stage. An example is the decrease in the expression of the
Drd2 (DA reception),
Syt11 (synaptic vesicle cycle of neurotransmission), and
Gpx1 (antioxidant system) genes in the model of the preclinical stage of PD. Considering that D2 receptors are autoreceptors [
55] and that synaptotagmin 11 (
Syt11) is a non-Ca2+-dependent protein that regulates endocytosis [
56], the detected changes may suggest a reorganization of the chemical machinery of DA release and uptake.