Dysfunctional Presynaptic Regulation Of Striatal Medium Spiny Neurons In A Novel Mouse Model Of Parkinson’s Disease

Parkinson’s disease (PD) is the second most common neurodegeneration disorder. Clinically PD is characterized by tremor, rigidity, bradykinesia and postural instability. A pathological hallmark of PD is the progressive loss of dopaminergic neurons in the substantia nigra pars compacta (SNc). Though its etiology is not completely understood, it is generally thought that PD results from a complex interaction between environmental and genetic factors. Aging also plays an important role in the incidence of PD. The prevalence of PD for those aged65years or older is1.7%in China, similar to the incidence in the developed western countries. The estimated number of individuals over age50with PD in China was1.99million in2005. And this number is expected to be4.99million in2030. As the life expectancy increases, the economical and societal burdens of PD will likely also grow in China.The current pathophysiological model to explain the symptoms of PD is based on the imbalance between the activities of direct pathway and indirect pathway of the basal ganglia circuit. The basal ganglia are composed of the striatum, the globus pallidus (GP), the subthalamic nucleus (STN) and the substantia nigra (SN). The GP in primates includes the external segment of the GP (GPe) and the internal segment of the GP (GPi), equivalent to the GP and the entopeducular nucleus (EP) in rodents, respectively. The substantia nigra can be divided into two parts, the substantia nigra pars compacta (SNc) and the substantia nigra pars reticulata (SNr). Those nuclei of the basal ganglia are highly interconnected but can be simplified into two main pathways, direct pathway and indirect pathway. The striatum is viewed as the major input structure of the basal ganglia, while the GPi and the SNr are output structures. The input and output nuclei are linked through a monosynaptic direct pathway and a polysynapitc indirect pathway. The indirect pathway involves GABAergic projections from the striatum to GPe and from GPe to the STN, as well as excitatory glutamatergic projections from the STN to GPi and SNr. According to this scheme, under physiological conditions, striatal dopamine released from the nigrastriatal dopaminergic terminals enhances the activities of the direct pathway via D1dopamine receptors and reduces the activities of the indirect pathway via D2dopamine receptors. The net outcome is inhibiting the activities of GPi and SNr, which in turn disinhibit the targets of the basal ganglia and facilitate the movement initiation. In PD, the dopaminergic neurons in the SNc degenerate progressively and selectively and lead to dopamine denervation of the dorsal striatum. The consequence of dopamine depletion in the dorsal striatum is decreased activities of direct pathway and increased activities of indirect pathway. And the net outcome is increasing the activities of GPi and SNr, which in turn inhibit the targets of the basal ganglia and inhibit the movement initiation. As the anatomical data about basal ganglia cumulating, it is generally believed that this classic scheme is too simplistic and the connectome is far more complex than this.Indeed, even the medium spiny neurons of the direct pathway (also D1MSNs) and the medium spiny neurons of the indirect pathway (also D2MSNs) in the striatum are interconnected by recurrent collateral axon terminals. And the GABAergic neurons in the GP project axon terminals to all parts of the basal ganglia. Moreover, dopaminergic neurons in the SNc project axon terminals to the striatum, GPe, STN and GPi and release dopamine to SNr from their dendrites. Thus in PD, the progressive loss of dopaminergic neurons in SNc will lead to dopamine depletion in the whole basal ganglia, including the striatum and the GPe. It has long been believed that dopamine depletion in the dorsal striatum was the main pathological change in PD, while dopamine denervation in the extrastriatum sites of the basal ganglia might also contribute to the symptoms of PD. The consequences of dopamine depletion in the striatum and the GPe are not entirely understood. In this study, we set out to unravel the consequences of dopamine depletion on the recurrent collateral connections between MSNs in the dorsal striatum and on the striatapllidal GABAergic transmission. Previous studies have shown that after dopamine depletion in the striatum, both D1and D2dopamine receptors became supersensitive to dopamine modulation. Previous studies have also demonstrated that both the recurrent collateral GABAergic synaptic transmission between MSNs within the striatum and the GABAergic transmission from the striatum to GPe were modulated by dopamine. Thus, our hypothesis of this study is that after dopamine depletion the recurrent collateral GABAergic synaptic transmission between MSNs within the dorsal striatum and the GABAergic synaptic transmission from the striatum to GP are more sensitive to dopamine modulation.To test this hypothesis, we used a new genetic mouse model of PD, Pitx3deficient aphakia mice (Pitxhomo mice). The dopaminergic neurons in SNc of Pitxhomo mice were constitutionally lost. And the dopamine loss in the dorsal striatum of those mice is close to90%. Those characteristics of the nigrostriatal dopamine pathway of Pitxhomo mice are similar to late stages of PD. Moreover, Pitxhomo mice also showed motor deficits that resembling the movement disorder of PD. Compared with the traditional toxin-induced mice models of PD, Pitxhomo mice as a PD model has several advantages and complementary features. First, the lesion of nigrostriatal dopamine neuron by neurotoxin was usually partial and had variations. Furthermore, in some cases the nigrostiatal pathway could recovery from a partial SNc lesion. In contrast, the loss of nigrostiatal dopamine system is more complete, consistent and irreversible in the Pitxhomo mice. Second, in the traditional neurotoxin-induced PD model, the loss of dopaminergic terminal occurred in the whole striatum, whereas Pitxhomo mice have more selective depletion of nigrostiatal dopamine projections and the loss of dopaminergic terminal occurred selectively in the dorsal striatum with relative sparing of ventral striatum. Third, the lack of nigrostriatal dopamine projection is constitutional in Pitxhomo mice and exists throughout the development, which may favor the condition of L-dopa induced dyskinesia (LID) induction. Fourth, Pitxhomo mice are suitable for electrophysiological studies, since most traditional neurotoxin-induced PD models use adult animals and to perform patch clamp experiments on adult animals is very difficult. Compared with other genetic mice PD models, Pitxhomo mice also have unique advantage. Among all the genetic mice PD models, only Pitxhomo mice displayed constitutional selective and almost complete dopamine loss in SNc, whereas other genetic only display mild dopaminergic neuron loss at very late stage. Nurr1knock out homo mouse also did not develop dopaminergic neurons in SNc, but they died soon after birth thus limits its use. Thus we took the advantages of this PD model and performed the following experiments.The first sets of the experiments were to study the dopamine modulation of the recurrent collateral connection between MSNs within the dorsal striatum. MSNs account more than90%of the neurons of the striatum. They are divided into two functionally segregated populations:direct pathway MSNs or D1MSNs and indirect pathway MSNs or D2MSNs. These two populations of MSNs are intermingled together and morphologically indistinguishable. We took the advantage of transgenic mice in which D2receptor-expressing MSNs are labeled with enhanced green fluorescent protein (EGFP)(D2EGFP mice hereinafter). We backcrossed the Pitxhomo mice with D2EGFP mice and produced Pitxhomo/D2EGFP phenotype and PitxWT/D2EGFP. We did paired patch clamp recording to study the recurrent collateral connections between MSNs in the dorsal striatum in the PitxWT/D2EGFP mice and Pitxhomo/D2EGFP mice. We first compared the connection rates among MSNs between PitxWT/D2EGFP mice and Pitxhomo/D2EGFP mice. Next, we compared the peak amplitude of the recurrent collateral synaptic inhibitory postsynaptic currents (IPSCs) among MSNs between PitxWT/D2EGFP mice and Pitxhomo/D2EGFP mice. Then we compared the dopamine modulation of the recurrent collateral synapses from D1to D1MSNs and from D2to D2MSNs between PitxWT/D2EGFP mice and Pitxhomo/D2EGFP mice. Last, we also treated the Pitxhomo/D2EGFP mice with low dose L-dopa began at P10, since dopamine replacement therapy with L-dopa remains the most effective drug therapy for PD. The reason we chose this dose and time point for treatment is based the behaviour data of our lab. We made the following findings:1. Both in the PitxWT/D2EGFP mice and in the Pitxhomo/D2EGFP mice, the recurrent collateral connections were coupled to D1and D2MSNs, with D1MSNs and D2MSNs preferentially formed recurrent collateral connections to D1MSNs and D2MSNs respectively. The connection rates of recurrent collateral synapse from D1to D1MSNs and from D1to D2MSNs were42.55%and8%in PitxWT/D2EGFP mice respectively. And the connection rates of recurrent collateral synapse from D2to D2MSNs and from D2to D1MSNs were44.29%and24%in PitxWT/D2EGFP mice respectively. Similar to the connection rates in PitxWT/D2EGFP mice, the connection rates of recurrent collateral synapse from D1to D1MSNs and D1to D2MSNs were41.86%and5%in Pitxhomo/D2EGFP mice respectively. And the connections rates of recurrent collateral synapse from D2to D2MSNs and from D2to D1MSNs were63.41%and35%in Pitxhomo/D2EGFP mice respectively. Both in PitxWT/D2EGFP mice and in Pitxhomo/D2EGFP mice, the connection rates of recurrent collateral synapse from D1to D1MSNs were significantly higher than that of the recurrent collateral synapse from D1to D2MSNs (Chi-square test, P<0.001), and the connection rates of recurrent collateral synapse from D2to D2MSNs were significantly higher than that of recurrent collateral synapse from D2to D1MSNs (Chi- square test, P<0.01).2.The connection rates of recurrent collateral connections and the peak amplitudes of the recurrent IPSCs in the dorsal striatum of Pitxhomo/D2EGFP mice were similar to those of PitxWT/D2EGFP mice (42.55%vesus41.86%for connection rates from D1to D1MSN in PitxWT mice and Pitxhomo mice respectively, chi-square test, P=0.6700.05;54.29%vesus63.41%for connection rates from D2to D2MSN in PitxWT mice and Pitxhomo mice respectively, chi-square test, P=0.290>0.05;8%vesus5%for connection rates from D1to D2MSN in PitxWT mice and Pitxhomo mice respectively, chi-square test, P=0.446>0.05;24%vesus35%for connection rates from D2to D1MSN in PitxWT mice and Pitxhomo mice respectively, chi-square test, P=0.273>0.05), which indicated that dopamine depletion in the dorsal striatum didn’t disrupt the recurrent collateral connections between MSNs.3. Dopamine and D1dopamine receptor agonist could facilitate the peak amplitudes of the IPSCs of the recurrent collateral synapses from D1to D1MSNs in PitxWT/D2EGFP mice.50μM dopamine increased the peak amplitude of the recurrent collateral synaptic IPSCs by117.3±40.8%(paired t-test, P=0.044<0.05) and1μM SKF81297, a D1dopamine receptor agonist, facilitated the peak amplitude of the recurrent collateral synaptic IPSCs by63.87±40.40%(paired t-test, P=0.049<0.05) in PitxWT/D2EGFP mice. And those effects were through acting on the presynaptic D1dopamine receptors, as evidenced by decreased paired pulse ratio (PPR) by around30%during bath application of dopamine. However, dopamine and D1dopamine receptor agonist failed to facilitate the peak amplitudes of the IPSCs of the recurrent collateral synapses from D1to D1MSNs in Pitxhomo/D2EGFP mice.50μM dopamine increased the peak amplitude of the recurrent collateral synaptic IPSCs by4.20±3.20%(paired t-test, P=0.462>0.05) and1μM SKF81297only facilitated the peak amplitude of the recurrent collateral synaptic IPSCs by1.30±4.50%(paired t-test, P=0.362>0.05) in Pitxhomo/D2EGFP mice. There was significant difference in the effect of dopamine and D1agonist on the peak amplitude of the recurrent collateral synaptic IPSCs between PitxWT/D2EGFP mice and Pitxhomo/D2EGFP mice (unpaired t-test, P=0.022<0.05).4. After low dose L-dopa treatment at P10to Pitxhomo/D2EGFP mice, dopamine could significantely facilitate the amplitudes of the IPSCs of the recurrent collateral synapses from D1to D1MSNs by37.10±14.10%(paired t-test, P=0.012<0.05).5. Dopamine and D2dopamine receptor agonist could inhibit the peak amplitudes of the IPSCs of the recurrent collateral synapses from D2to D2MSNs both in PitxWT/D2EGFP mice and in Pitxhomo/D2EGFP mice.50μM dopamine had similar inhibitory effects on the peak amplitudes of the recurrent collateral synaptic IPSCs from D2to D2MSNs in both PitxWT/D2EGFP mice and Pitxhomo/D2EGFP mice (79.00±3.10%versus77.90±5.90%, One-way ANOVA, P=0.318>0.05). And those effects were through acting on the presynaptic D2dopamine receptors, as evidenced by increased PPR during bath application of dopamine. However the inhibitory effects of25μM dopamine on the peak amplitudes of the IPSCs of the recurrent collateral synapses from D2to D2MSNs were significantly larger in Pitxhomo/D2EGFP mice that those in the PitxWT/D2EGFP mice (83.00±2.20%versus34.50±8.80%, One-way ANOVA, P<0.001).6. Low dose L-dopa treatment at P10to Pitxhomo/D2EGFP mice could decrease the sensitivities of the presynaptic D2dopamine receptors at the recurrent collateral synapses from D2to D2MSNs to dopamine modulation.25μM dopamine inhibited the peak amplitudes of the recurrent collateral synaptic IPSCs from D2to D2MSNs by67.20±8.20%(One-way ANOVA, P<0.001). Conclusions:1. Both in the PitxWT/D2EGFP mice and in the Pitxhomo/D2EGFP mice, the recurrent collateral connections were coupled to D1and D2MSNs, with D1MSNs and D2MSNs preferentially formed recurrent collateral connections to D1MSNs and D2MSNs respectively. And dopamine depletion in the dorsal striatum didn’t affect the recurrent collateral connections.2. After dopamine depletion, the D1dopamine receptors at the collateral axon terminals from D1to D1MSNs became less sensitive to dopamine modulation, while the D2dopamine receptors at the collateral axon terminals from D2to D2MSNs became supersensitive to dopamine modulation.3. Low dose L-dopa treatment at the early stage after dopamine depletion could reverse the sensitivities of the D1and D2dopamine receptors at the collateral axon terminals, with increased sensitivities of D1dopamine receptors at the collateral axon terminals to dopamine modulation and decreased sensitivities of D2dopamine receptors at the collateral axon terminals to dopamine modulation. The functional roles of those findings are not clear. Since the main function of the recurrent collateral synapses in the striatum is to control the dendritic excitability, we believe that the change of the sensitivities to dopamine modulation of the presynaptic D1and D2dopamine receptors at the recurrent collateral axon terminals might increase the dendritic excitabilities both in D1and D2MSNs after dopamine depletion, which might contribute to the motor deficits of Pitxhomo/D2EGFP mice.The second sets of experiments were to study the dopamine modulation of the striatopallidal IPSCs on GABAergic neurons in GP after dopamine depletion in Pitxhomo/D2EGFP mice. GP in rodents plays a pivotal role in controlling the activities of virtually the whole basal ganglia by projection axon terminals to all parts of the basal ganglia and providing tonic inhibition to all parts of the basal ganglia. The major GABAergic input to GP is striatopallidal input from the indirect pathway D2MSNs. The GABAergic neurons in GP also inhibited each other by recurrent collateral axon terminals. To avoid evoking IPSCs from recurrent collateral synapses, we stimulated the striatum and recorded evoked IPSCs (elPSCs) in GP, which was exclusively from striatopallidal axon terminals. We first characterized the striatopallidal elPSCs in both PitxWT mice and Pitxhomo mice. Next we compared the sensitivities to dopamine modulation of the striatopallidal elPSCs between PitxWT mice and Pitxhomo mice. Then we treated the Pitxhomo mice with low dose L-dopa began at P10or P14. The dose of L-dopa and the time point to begin the treatment were based on the behavior data in our lab. We found out that:1. Both in the PitxWT mice and Pitxhomo mice, there were two major types of GABAergic neurons in GP according to their electrophysiological properties. Dopamine depletion did not affect the cell compositions in GP and their basic electrophysiological properties.2. Dopamine and D2dopamine receptor agonist could inhibit the peak amplitudes of the striatopallidal elPSCs in both PitxWT mice and Pitxhomo mice. And those effects was through presynaptic D2dopamine receptors at the striatopallidal axon terminals, as evidenced by increased PPR during bath application of dopamine.3. Higher concentrations of dopamine (10-50μM) had similar inhibitory effects on the peak amplitudes of the striatopallidal elPSCs in both PitxWT mice and Pitxhomo mice.50μM,20μM, and10μM dopamine significantly inhibited the peak amplitudes of elPSCs by72.80±10.90%,71.90±10.70%and71.50±11.70%respectively in PitxWT mice and by69.40±2.66%,74.60±8.10%and71.80±4.90%respectively in Pitxhomo mice. There was no significant difference in the effects of dopamine among those doses in both PitxWT mice and Pitxhomo mice (One-way ANOVA, P>0.05). And there also was no difference in the effects of dopamine at each of those doses between PitxWT mice and Pitxhomo mice (unpaired t-test, P>0.05). However, lower concentrations of dopamine (1-5μM) had significantly larger inhibitory effects on the peak amplitudes of the striatopallidal elPSCs in Pitxhomo mice, compared with PitxWT mice.5μM,3μM and1μM dopamine reduced68.70±8.90%,66.40±7.80%and64.46±11.50%of the peak amplitude of the striatopallidal elPSC in Pitxhomo mice respectively, while they only inhibited54.00±5.30%,33.80±6.00%and25.20±3.70%of the peak amplitude of the striatopallidal elPSC in PitxWT mice respectively. There were significant differences in the effects of dopamine at each of those doses between PitxWT mice and Pitxhomo mice (unpaired t-test, P<0.005). The dose response curve was left shifted in Pitxhomo mice compared with that of PitxWT mice, which indicated that the D2dopamine receptors at the striatopallidal axon terminals were more sensitive to dopamine modulation in Pitxhomo mice than those in PitxWT mice.4. Low dose L-dopa treatment at P10(early treatment group) instead of at P14(late treatment group) could reduce the sensitivities of the D2dopamine receptors at the striatopallidal axon terminals in Pitxhomo mice, as evidenced by right shifted dose response curve compared with untreated Pitxhomo mice (control group). Higher concentrations of dopamine (10-50μM) had similar inhibitory effects on the peak amplitudes of the striatopallidal elPSCs among control group, early treatment group and late treatment group.50μM,20μM and10μM dopamine significantly reduced the peak amplitudes of the striatopallidal elPSCs by66.70±4.40%,65.10±7.10%and62.40±5.60%respectively in early treatment group and by68.30±3.50%,72.00±6.10%and74.30±6.90%respectively in the late treatment group. And there were no significant differences in the effects of dopamine at each dose of those concentrations among the three groups (One-way ANOVA, P>0.05). However,5μM dopamine could inhibit the peak amplitudes of the striatopallidal elPSCs by74.40±2.10%,54.50±4.50%and68.70±8.90%in control group, early treatment group and late treatment group respectively. Even though there were no significant differences in the effects of5μM dopamine among the three groups (Duncan’s Test), the effects of5μM dopamine in the early treatment group was significantly smaller than that of the control group (Duncan’s Test) and that of the late treatment group (Duncan’s Test).3μM dopamine could inhibit the peak amplitudes of the striatopallidal elPSCs by66.40±7.80%,44.50±12.10%and67.80±7.80%in control group, early treatment group and late treatment group respectively. There were significant differences in the effects of3μM dopamine among the three groups (Duncan’s Test).1μM dopamine also could inhibit the peak amplitudes of the striatopallidal elPSCs by64.60±11.50%,17.30±3.10%and67.50±7.50%in control group, early treatment group and late treatment group respectively. And there were significant differences in the effects of1μM dopamine among the three groups (Duncan’s Test).5. The striatopallidal evoked inhibitory postsynaptic potentials (elPSP) were powerful enough to reset the autonomous firing activities of the GABAergic neurons in GP in both PitxWT mice and Pitxhomo mice, since they could prolong the inter-spike intervals (ISI) in both types of mice.6. High concentration of dopamine (50μM) had similar effects to reduce the capabilities of the striatopallidal elPSPs to reset the firing activities of GABAergic neurons in GP in PitxWT mice and Pitxhomo mice, since50μM dopamine could reduce the|S|ipsp by51.11±6.73%and60.51±4.90%in PitxWT mice and Pitxhomo mice respectively and there was no significant difference in the effects of50μM dopamine on the|S|ipsp between PitxWT mice and Pitxhomo mice (P=0.091>0.05). However, low concentration of dopamine (3μM) could still reduce the capabilities of the striatalpallidal elPSPs to reset the firing activities of GABAergic neurons in GP in Pitxhomo mice, whereas it failed to do so in PitxWT mice. Low concentration of dopamine (3μM) only reduced|S|ipsp by13.19±6.57%in PitxWT mice. In contrast, during bath application of3μM dopamine, the|S|ipsp was significantly reduced by68.15±10.10%in Pitxhomo mice. There was significant difference in the effect of3μM dopamine on|S|ipsp between PitxWT mice and Pitxhomo mice (unpaired t-test, P=0.001<0.01). Conclusions:1. We mainly found two types of GABAergic neurons in GP both in PitxWT mice and Pitxhomo mice. Dopamine denervation of GP didn’t change the cell composition of GP and the basic electrophysiological properties of those GABAergic neurons.2. Dopamine and D2dopamine receptor agonist could inhibit the peak amplitudes of the striatopallidal elPSCs in both PitxWT mice and Pitxhomo mice through presynaptic D2dopamine receptors at the striatopallidal axon terminals.3. D2dopamine receptors at the striatopallidal axon terminals became supersensitive to dopamine modulation after dopamine depletion in GP.4. Low dose L-dopa treatment at the early stage instead of the late stage could postpone the appearance of the supersensitivities of the D2dopamine receptors at the striatopallidal axon terminals to dopamine modulation in Pitxhomo mice.5. The supersensitivities of the D2dopamine receptors at the striatopallidal axon terminals to dopamine modulation in Pitxhomo mice are functionally important. Even very low dose of dopamine could reduce the capability of the striatopallidal elPSPs to reset the autonomous firing activities of GABAergic neurons in the GPe, which could disinhibit the GABAergic neurons in the GPe. The reduced capabilities of the striatopallidal elPSCPs to reset the autonomous spiking of the GABAergic neurons could facilitate movement initiation by two mechanisms. First, the inhibitory inputs to the STN will increase and the firing frequency of the glutamaterigc neurons in the STN will decrease, which in turn lead to decreased excitatory glutamatergic inputs to the output nuclei of the basal ganglia, GPi and SNr and decrease the firing frequency of the GABAergic projection neuron in those nuclei. Second, GABAergic neurons in the GP provide tonic inhibition to the output nuclei of the basal ganglia. Increased firing frequency of the GABAergic neurons in GPe will lead to increase inhibition input to the output nuclei and decrease the firing frequency of the GABAergic neurons in those nuclei. Decreased inhibition from the output nuclei of the basal ganglia to their targets, thalamus, superior colliculus and pedunclepontine nucleus, will facilitate the movements initiation. This might be part of mechanisms of the hyperactivity of Pitxhomo mice to L-dopa treatment.In summary, we made the following important findings:1. After dopamine depletion, the D2dopamine receptors both at the recurrent collateral axon terminals from D2to D2MSNs and at the striatopallidal axon terminals, which were from D2MSNs to the GABAergic neurons in GP, became supersensitive to dopamine modulation.2. After dopamine depletion, the D1dopamine receptors at the recurrent collateral axon terminals from D1to D1MSNs were insensitive to dopamine modulation.3. Low dose L-dopa treatment at early stage could reverse the sensitivities of both D1and D2dopamine receptors at the recurrent collateral axon terminals and the D2dopamine receptors at the striatopallidal axon terminals. The functional role of the findings of the first sets of experiments is that the excitability of the dendrites of both D1and D2MSNs were increased after dopamine depletion, which might facilitate the transmissions through both direct pathway and indirect pathway. The physiological relevance of the findings of the second sets of experiments is that dopamine can disinhibit the GABAergic neurons in the GPe much easier after dopamine depletion and facilitate movement initiations, which might be a part of the mechanisms of L-dopa treatment and L-dopa induced dyskinesia. These important findings might shed some new lights on our understanding the pathological changes in the dorsal striatum and GP in PD. Moreover, those findings might also expand our knowledge about treating PD with L-dopa and L-dopa induced dyskinesia.