Inhibition of striatal dopamine release by the L‐type calcium channel inhibitor isradipine co‐varies with risk factors for Parkinson's

Abstract Ca2+ entry into nigrostriatal dopamine (DA) neurons and axons via L‐type voltage‐gated Ca2+ channels (LTCCs) contributes, respectively, to pacemaker activity and DA release and has long been thought to contribute to vulnerability to degeneration in Parkinson's disease. LTCC function is greater in DA axons and neurons from substantia nigra pars compacta than from ventral tegmental area, but this is not explained by channel expression level. We tested the hypothesis that LTCC control of DA release is governed rather by local mechanisms, focussing on candidate biological factors known to operate differently between types of DA neurons and/or be associated with their differing vulnerability to parkinsonism, including biological sex, α‐synuclein, DA transporters (DATs) and calbindin‐D28k (Calb1). We detected evoked DA release ex vivo in mouse striatal slices using fast‐scan cyclic voltammetry and assessed LTCC support of DA release by detecting the inhibition of DA release by the LTCC inhibitors isradipine or CP8. Using genetic knockouts or pharmacological manipulations, we identified that striatal LTCC support of DA release depended on multiple intersecting factors, in a regionally and sexually divergent manner. LTCC function was promoted by factors associated with Parkinsonian risk, including male sex, α‐synuclein, DAT and a dorsolateral co‐ordinate, but limited by factors associated with protection, that is, female sex, glucocerebrosidase activity, Calb1 and ventromedial co‐ordinate. Together, these data show that LTCC function in DA axons and isradipine effect are locally governed and suggest they vary in a manner that in turn might impact on, or reflect, the cellular stress that leads to parkinsonian degeneration.

K E Y W O R D S Calb1, dopamine D2 receptor, dopamine release, dopamine transporter, L-type calcium channel, Parkinson's disease, sex differences, striatum, synuclein

| INTRODUCTION
Ca 2+ entry via voltage-gated Ca 2+ channels (VGCCs) is important for pacemaker activity in dopamine (DA) neurons within the substantia nigra pars compacta (SNc) and critical for release of DA in the striatum (Brimblecombe et al., 2015;Guzman et al., 2009).However, intracellular Ca 2+ also generates metabolic stress that promotes vulnerability of SNc DA neurons to degeneration in Parkinson's disease (PD) (Surmeier et al., 2010).SNc DA neurons and their axons in dorsolateral striatum (DLS) are particularly vulnerable to parkinsonian degeneration, whereas DA neurons in ventral tegmental area (VTA) and their axons in ventral striatum (nucleus accumbens, NAc) are relatively spared.L-type VGCCs (LTCCs) support pacemaker activity in SNc predominantly via the Cav1.3 subtype (Guzman et al., 2009;Philippart et al., 2016) and by facilitating ATP production (Zampese et al., 2022) and also support DA release from axons in DLS (Brimblecombe et al., 2015).By contrast, in DA neurons in VTA and DA axons in NAc, LTCCs do not operate strongly (Brimblecombe et al., 2015;Guzman et al., 2009;Philippart et al., 2016;Zampese et al., 2022).However, this regional difference is not reflected by LTCC gene expression at the level of the soma at least: the dominant subtype of LTCCs in DA soma in SNc and VTA is thought to be the Cav1.3 rather than Cav1.2 subtype (Philippart et al., 2016), but Cav1.3 gene expression is more enriched in VTA than in SNc (Poulin et al., 2014).The identity of the subtype operating in axons is not yet known and cannot be determined pharmacologically.The potential discrepancies between expression and function suggests that regulatory mechanisms locally govern function.Here, we explored factors that locally regulate the LTCC support of DA release from striatal DA axons in DLS vs NAc core.
Several biological factors and mechanisms that have direct relevance to the biology of DA neurons and/or to PD are candidates for LTCC regulators.On a population level, after ageing, the biggest risk factor for PD is biological sex: men are twice as likely as women to develop PD (Wooten et al., 2004), but the biology underlying this difference is poorly understood.It has been established in cardiac myocytes that LTCC function is greater in males than females (Curl et al., 2008), but sex differences in LTCC function in central DA systems have not previously been explored.Here, we explored how LTCCs support DA release in both sexes of mice and define sex differences.
Several genetic mutations and polymorphisms confer an increased risk to PD (Bekris et al., 2010).These range from rare, high penetrance mutations, for example, in the genes for α-synuclein (SNCA) and PINK-1/Parkin, to more common, less penetrant single-nucleotide polymorphisms, discovered by GWAS including GBA (gene encoding glucocerebrosidase, GCase) as the most commonly associated genetic risk factor.Interestingly, variants in SNCA occur as both rare, high penetrance and common, low penetrance forms.Here, we tested whether α-synuclein or GCase levels could impact on LTCC support of DA release.
We also tested whether several known physiological regulators of DA output from SNc and VTA DA neurons might support a differential contribution of LTCCs to DA release.We focussed on calbindin-D28K (Calb1), the DAT and DA D 2 -receptors (D 2 Rs), each of which have been reported to interact with LTCCs in different systems.Calb1 is enriched in PD-resistant VTA neurons (Gerfen et al., 1987;Poulin et al., 2014), limits DA release in NAc (Brimblecombe et al., 2019) and, in cellular expression systems, can bind to and limit function of LTCCs (Lee et al., 2006).We report here that Calb1 limits LTCC support of DA release in a regionally and sexually divergent manner.DATs are enriched in SNc and VTA DA neurons, with highest expression levels reported in SNc, in both mouse and human brains (Ciliax et al., 1995;Poulin et al., 2014), and differential DAT expression has long been hypothesised to be a driving factor in sensitivity to PD on cellular and population levels (Lohr et al., 2017;Ritz et al., 2009).In cellular expression systems, DATs and other electrogenic monoamine transporters promote LTCC currents (Cameron et al., 2015).We report here that DATs play a role in facilitating how LTCCs support axonal DA release.Lastly, D2Rs are described as more enriched in SNc than VTA DA neurons (Greene et al., 2005;Poulin et al., 2014) and can inhibit LTCC function (via βγ-inhibition) in striatal GABAergic projection neurons (Hern andez-L opez et al., 2000;Olson et al., 2005).Despite D2Rs not being associated with either risk or protection in PD (not least due to their broad expression pattern), they are well placed to modulate LTCC support of DA release.We report here that D2Rs operate a limiting factor on LTCC function in NAc.Overall, we reveal a network of factors that, by either inhibiting or facilitating LTCC function in DA axons, contribute to a greater role for LTCCs in supporting DA release in DLS than NAc, and in males than females, and a correspondingly varying outcome of LTCC inhibitor isradipine on DA release.

| Electrical stimulation
DA recordings were obtained from dorsolateral quadrant of dorsal striatum, the dorsolateral striatum (DLS), or nucleus accumbens core (NAc).DA release was evoked by a local bipolar concentric Pt/Ir electrode (25 μm diameter; FHC inc.ME, USA) placed approximately 100 μm from the recording electrode.Stimulus pulses (200 μs duration) were given at .6 mA (perimaximal in drug-free control conditions).Stimulations were single pulses (1p) and were repeated at 2.5 min intervals.A stable pre-drug baseline of six stimulations before isradipine or CP8 was washed on for 12 stimulations.A stable baseline was determined during the experiment when the average peak oxidation for stimulus 5 and 6 was 95-105% of stimulus 1 and 2. All data were obtained in the presence of the nAChR antagonist, dihydro-β-erythroidine (DHβE, 1 μM) to exclude the strong effects of nAChRs on DA release (Cachope et al., 2012;Rice & Cragg, 2004;Threlfell et al., 2012;Zhou et al., 2001).

| Drugs and solutions
Reagents were purchased from Torcis or Sigma Aldrich (cocaine, water-soluble cholesterol).Stock solutions were made to 1,000-2,000Â final concentrations in H 2 O (DHβE, cocaine), DMSO (isradipine, L-741,626, CP8), .1 M HCl (nomifensine) or ethanol (lidocaine) and stored at À20 C. Drugs were diluted to their required concentrations in aCSF immediately prior to use.Drug concentrations were chosen in accordance with previous studies (Acevedo-Rodriguez et al., 2014;Brimblecombe et al., 2015).Water soluble cholesterol or vehicle control Me-ß-cyclodextrin were dissolved in 10 ml aCSF, and slices were incubated and oxygenated for 30 min before transferring to recording chamber as previously described (Threlfell et al., 2021).

| Data and statistical analysis
FCV data were acquired and analysed using Axoscope 10.5 (Molecular devices) and Excel macros written locally.Data are expressed as mean ± standard error of the mean (SEM), and N = number of slices (typically 1 slice used per animal for any given experiment) from a minimum of 3 animals.Data from each slice were obtained by averaging four recordings prior to drug wash-on (control) versus average of four recordings after 20 min.Typically, data are shown normalised to control conditions to standardise for variation in initial evoked extracellular DA concentrations ([DA] o ).Population means were compared using one-or two-way ANOVA with Dunnett's post-test and unpaired t-test where appropriate using GraphPad Prism (9.3.1).Curve fits were performed in GraphPad Prism.Data for all figures are available at Zenodo (10.5281/zenodo.7801867).

| Classification tree
A classification tree was performed in R(4.2.1) (10.5281/ zenodo.7801867).The normalised effect of isradipine for each individual experiment was annotated with the experimental condition: sex and genotype of mouse and presence of drug prior to washing on isradipine (fig.5 data tab in xlsx file 10.5281/zenodo.7801867).These data were sorted using a classification tree using Gini impurity algorithm with a cut off of .91 normalised effect of isradipine, which corresponds to 2 s.d.above the mean effect of isradipine in DLS of male WT mouse (i.e., the original recording conditions where we identify a significant effect of isradipine on decreasing DA; Brimblecombe et al., 2015).It should be noted that .91 is also 2 s.d.below mean normalised effect of isradipine in male WT NAc (2.4 mM [Ca 2+ ] o ), which was our original recording conditions when we did not identify an effect of isradipine.Due to the structure of the classification tree whereby, factors were effectively sorted with no outliers (i.e., the leaves [final nodes] were mostly pure with probabilities of 0 or 1), we determined it was unsuitable to run a logistic regression analysis on these data.

| Western blots
Mice were culled by cervical dislocation, brains extracted and dissected on ice.The striatum and midbrain were extracted from both hemispheres and snap frozen.Samples were prepared for western blot as previously described (Connor-Robson et al., 2019).Briefly, tissue was homogenised on ice in RIPA buffer containing protease inhibitors (Roche) and protein concentration determined by BCA assay.Following dilution, Laemmli buffer was added, and samples were boiled for 5 min at 95 C. Samples were loaded and run on 4-15% criterion-TGX gels and transferred to PDVF membranes (BioRad).Membranes were blocked for 1 h at room temperature and then probed with primary antibody (calb1: 1:1,000 Cell Signalling #13176) overnight at 4 C.After incubation with HRP conjugated secondary antibodies, the blot was developed using chemiluminescent HRP substrate.

| Sex difference in the LTCC support of DA release
We and others have to date principally explored LTCC function in males rather than females (Brimblecombe et al., 2015;Guzman et al., 2009Guzman et al., , 2018;;Ilijic et al., 2011).There is an $2:1 prevalence of PD in males and females, and it is therefore important to understand the biology of LTCCs in both sexes, in order to gain a fuller understanding of disease aetiology but also drug action.Here, we tested whether LTCCs support DA release in DLS of female mice, which they do in male mice (Brimblecombe et al., 2015).We first assessed DA release and uptake parameters.We found no difference in peak evoked extracellular concentrations of DA ([DA] o ) evoked by single electrical stimulus pulses (1p) (Figure 1a,b, t-test t 14 = .14P = .89)but did find faster DA clearance rates in females (Figure 1a,c, T-test t 10 = 4.15 P = .002),consistent with greater DAT levels in the striatum of female mice (Costa et al., 2021).We then tested the effect of LTCC inhibitor isradipine (5 μM) on 1p-evoked [DA] o in DLS and confirmed a decrease in evoked [DA] o in male mice (Figure 1d, comparison of normalised peak [DA] o , one-sample t-test t 2 = 5.6, P = .03,N = 3) as published previously (Brimblecombe et al., 2015) but, by contrast, not in female mice (Figure 1d, comparison of normalised peak [DA] o , one-sample t-test t 4 = .87,P = .44,N = 5), revealing a statistically significant difference between the two sexes (Figure 1e, two-way ANOVA: sex X drug interaction, F 16,67 = 3.0, P < .001;main effect of sex, F 1,67 = 44.9,P < .001).
We did not see an LTCC role in DA release in females in WT mice (see Figure 1c) but tested whether LTCC support of DA release in females could be established in either of two different PD-related mouse lines.First, we explored whether overexpression of α-synuclein could result in LTCC support of DA release.However, in female SNCA-OVX mice, isradipine did not decrease [DA] o in DLS (Figure 1i, comparison of normalised peak [DA] o t 2 = 2.2 P = .16),indicating a significant difference between male and female SNCA-OVX mice (Figure 1j, two-way ANVOA: genotype X drug interaction, F 17,102 = 1.89,P = .027;main effect of genotype, F 1,6 = 7.85 P = .03).Second, since mutations in the GCase GBA gene are associated with PD and seen more commonly in women with PD than men (Li et al., 2021), we explored whether LTCC support of DA release varied with GCase levels in females, by comparing mice that were heterozygous for the mouse Gba gene (Gba +/À ) with wild-type littermate homozygous controls (Gba +/+ ).Isradipine did not modify 1p-evoked [DA] o in DLS Gba +/+ female controls, as seen for other wild-type females, but strikingly, isradipine did inhibit 1p-evoked [DA] o in female Gba +/À mice (Figure 2k, comparison of normalised peak [DA] o , Gba +/+ (WT), t 2 = .33P = .77;Gba +/À , t 4 = 12.03 P = .0003),showing a significant difference in drug effect between these two genotypes (Figure 1l: two-way ANVOA: genotype X drug interaction F 16,96 = 6.4 P < .0001;main effect of genotype, F 1,6 = 16.2P = .007;absolute levels of evoked [DA] o for these recording sites were similar between genotypes pre-isradipine, at 1.09 ± .09for Gba +/+ and .98 ± .15μM for Gba +/À ).These observations indicate that a role for LTCCs in supporting DA release is exposed when GCase function is compromised.
We note that across all experiments and variables, we found no correlation between 1p-evoked [DA] o levels and the subsequent effect of isradipine (Figure 1m, simple linear regression, slope does not significantly differ from zero F 1,28 = .049P = .83),confirming that effect size was not a function of different initial release probability in different conditions.

| Sex differences and calb1 interact to govern LTCC function
VTA and SNc DA neurons differ in a number of key aspects that might contribute to their differing sensitivities to parkinsonian degeneration.One notable difference is the greater expression of calbindin-D28K (Calb1) in VTA neurons, which has been suspected to be protective against PD, although evidence is mixed (Airaksinen et al., 1997;Yuan et al., 2013).We used cre-lox P recombination to selectively knockdown Calb1 in DA neurons in a cross of DAT-Cre mice with floxed Calb1 mice as previously (Brimblecombe et al., 2019), which does not affect evoked [DA] o or DA uptake in DLS, but it has profound impact in NAc, consistent with the regional endogenous expression profile of Calb1.We found here that in control mice (DAT-Cre hets, referred to as CalbWT), isradipine decreased 1p-evoked [DA] o in males in DLS but not NAc and did not modify evoked [DA] o in females in DLS (Figure 2a) as per wild-type mice (see Figure 1).By contrast, in CalbKD mice, isradipine decreased 1p-evoked [DA] o in male DLS and NAc, and in female DLS (Figure 2b).Overall, Calb1 knockdown significantly increased the effect of isradipine on DA release (Figure 2c, two-way ANOVA effect of genotype F 1,14 = 31.17P < .001)and enabled LTCCs to support DA release in both male NAc and female DLS, even at lower [Ca 2+ ] o in NAc (Figure 2c).We tested whether there was a gross upregulation of Calb1 levels in the striatum of wild-type female mice relative to males, leading to the occlusion of LTCC function seen in WT female mice.However, western blot analysis of protein levels in whole striatal tissue did not detect a difference in Calb1 expression in striatum of females versus males (Figure 2d,e, unpaired t-test T 4 = .18P = .86).

| DAT function promotes LTCC support of DA release
Translocation of DA across the membrane by DAT involves the co-transport of 2 Na + and 1 Cl À , an electrogenic process that can depolarise DA neuron membranes, independently of DA translocation (Ingram et al., 2002) and in cellular expression systems can activate LTCCs (Cameron et al., 2015).We tested whether DATs operate in DA axons to support LTCC function.We inhibited DATs using cocaine (5 μM) or nomifensine (10 μM), monoamine uptake inhibitors, which are also known to prevent the electrogenic effects of DAT (Sonders et al., 1997), and tested the impact of isradipine on 1p-evoked [DA] o in DLS in male wild-type mice.We found that in the presence of cocaine or nomifensine, the effect of LTCC inhibitor isradipine in the DLS was no longer apparent (Figure 3a-c, two-way ANOVA: isradipine X pre-condition interaction F 32,192 = 3.57, P < .001;comparison of peaks one-sample t-test: control t 3 = 6.48,P = .0074,cocaine t 4 = .37P = .73,nomifensine t 4 = 1.39 P = .23).The loss of isradipine effect in the presence of cocaine was not due to any potential effects of cocaine as an inhibitor of voltage-gated Na + -channels (Na v ) because application of the Na v blocker and analogue lidocaine did not similarly prevent the effect of isradipine.Rather, in the presence of lidocaine, isradipine decreased 1p-evoked [DA] o as seen in standard conditions (Figure 3c, one-way ANOVA F 2,9 = 11.5 P = .03,Dunnett's post-test lidocaine vs. control P = .83,cocaine vs. control P = .003,nomifensine vs. control P = .014).These data show that inhibition of DATs occludes LTCCs from contributing to DA release and suggest that DATs support LTCC function in DLS.
We tested whether a potentiation of DAT function could facilitate LTCC support of DA release where it was not previously observable.We have previously shown that DAT function in Snca-null mice can be augmented by cholesterol (Threlfell et al., 2021) and exploited this action to see whether in Snca-null male mice an effect of isradipine could be brought about after cholesterol incubation and its allied increase in DAT function.Indeed, after incubation in cholesterol, isradipine significantly decreased 1p-evoked [DA] o (Figure 3d, control: t 3 = 4.2 P = .025).To confirm that this outcome depended on DAT function, we repeated in the presence of cocaine to inhibit DAT.In the presence of cocaine, cholesterol incubation no longer supported an effect of isradipine on DA release (Figure 3d, cocaine: t 3 = .17P = .87,Figure 3e, one-way ANOVA F 3,12 = 6.29 P = .0083,Dunnett's posttest Snca-null control vs. cholesterol P = .0034,vs. chol +cocaine P = .54),suggesting that a DAT-dependent increase in LTCC function underlay this impact of cholesterol.
We excluded an alternative hypothesis that DAT function promotes the effect of isradipine not by increasing LTCC function but by changing the ability of isradipine to bind to LTCCs.It is known that at hyperpolarised potentials, isradipine does not effectively bind to LTCC (Helton et al., 2005;Rey et al., 2020).Since inhibitors of DATs will reduce DAT-associated depolarising electrogenic currents and maintain DA axons in a more hyperpolarised state, they might prevent isradipine from binding to LTCC.We therefore tested whether the actions of an alternative LTCC antagonist to isradipine, CP8, which binds LTCCs in a voltage-independent manner (Kang et al., 2013;Rey et al., 2020), were also prevented in the presence of cocaine.We found first that CP8 decreased 1p-evoked [DA] o in DLS of male mice, but not in male NAc (Figure 3f,g, t-test comparison on peak DLS: t 2 = 5.19 P = .035,DLS + cocaine t 3 = 1.39 P = .26,NAc: t 2 = 1.54 P = .26;two-way ANOVA post-test time point 10 vs. 40 min DLS: P = .0052.NAc [not shown]: P = .78,DLS + cocaine P = .99),paralleling the effects of isradipine.Further in parallel with the effects of isradipine, CP8 did not reduce 1p-evoked [DA] o in DLS in the presence of cocaine (Figure 3g, two-way ANOVA post-test time point 10 vs. 40 min DLS: P = .0052.NAc: P = .78,DLS + cocaine P = .99).These data indicate that DATs indeed support LTCC function in contributing to DA release and do not simply support the voltage-sensitive binding and action of isradipine.
Given this apparent role for D2Rs in limiting LTCC support of DA release in NAc, we tested a hypothesis that the mechanism through which cocaine limits LTCC function in DLS might be elevation of DA levels leading to an enhancement of D2R-activation.We inhibited D2Rs with L-741,626, to see if this allowed isradipine to decrease evoked [DA] o in the presence of cocaine.However, isradipine did not decrease DA release in DLS in the presence of cocaine and L-741,626, indicating that the control of LTCC support of DA release by DATs in DLS is independent of D2R regulation (Figure 4f, t-test t 7 = 1.65 P = .14).
We tested finally whether D2Rs or DATs interact with the effects of Calb1 in NAc on LTCC support of DA release.In NAc of CalbKD mice where LTCCs support DA release (see Figure 2), D2R-inhibition did not augment the impact of isradipine on DA release compared to without D2R inhibition (Figure 4g-i, Dunnets post-test t 9 = .3P = .94)indicating either that the impact of CalbKD and D2R inhibition of LTCC role has reached a ceiling or that the mechanism through which D2Rs are governing LTCC impact involves Calb1.The inhibition of DATs in NAc of CalbKD mice with cocaine occluded the effect of isradipine on DA release (Figure 4g-i, t 2 = .73P = .53).These data indicate that LTCC support of DA release in NAc in CalbKD mice can be facilitated by DATs (Figure 4i, one-way ANOVA Sidak post-test CalbKD vs cocaine t 9 = 3.9 P = .007CalbKD vs. L471 t 9 = .3p = .94).These features of DAT support but no D2R control of LTCC function in NAc in CalbKD mice mirror the regulation of LTCC function in DLS of wildtype mice.Together, they also reveal that while Calb1 and D2Rs each limit LTCC function in wild-type NAc they do not do so in DLS (in the presence of normal DAT function).
Here, we have shown that the apparent function of LTCCs in supporting DA release varies with a large range of biological conditions, some of which show apparent co-dependencies.To better understand how these factors, of genotype, sex, region and molecular regulators interact to govern whether LTCCs support DA release, and by extension set the conditions when isradipine will decrease DA release, we employed a classification tree algorithm with Gini impurity, for each experimental replicate, to understand the relative influence of each factor (Figure 5a).The classification tree illustrated that DAT activity (cocaine data) is a key means of supporting L-function; the presence of cocaine is a robust limiter of isradipine's effect.Conversely, Calb1 and Gba strongly limit LTCC function; knockdown of either promoted isradipine action.In contrast, in wild-type, SNCA-OVX and Snca-null mice, LTCC function co-varied with sex, striatal region, D2-receptor state or cholesterol level (Figure 5a).The branching of the tree into multiple pure leaves (each end node had probability of 0 or 1, no outliers) facilitated a straightforward classification of factors into either facilitators or inhibitors of LTCC function (Figure 5b).

| DISCUSSION
We reveal here that LTCC support of DA release from DA axons is regulated by multiple, endogenous factors in a regionally and sexually divergent manner.In DLS, LTCCs operate more evidently in males, and their support of DA release is facilitated by α-synuclein and DATs and permitted by the absence of Calb1.In NAc, LTCC support of DA release is prevented by Calb1 and D2Rs.Intriguingly, regulators that promote LTCC function are associated with increased risk of degeneration in PD, while some of those that limit function are associated with reduced neuronal vulnerability, suggesting that intersecting mechanisms regulating DA handing and LTCC function might together contribute to the metabolic load of vulnerable neurons.We believe that the pattern of isradipine-sensitive DA release reflects a vulnerable state of the axons, and therefore, the sensitivity of DA release to isradipine is a phenotype of vulnerable DA axons.

| Sex differences interact with multiple factors to govern LTCC function
LTCCs have been suggested to confer vulnerability to degeneration of SNc DA neurons in PD (Surmeier, 2018), a disease with a higher incidence in men than women  (Wooten et al., 2004).Yet the relative roles of LTCCs in DA neurons in males versus females has been unknown.We reveal there is a greater role for LTCCs in supporting DA release in male DLS than in female, and this raises the speculation that differential LTCC function might contribute to the greater incidence of PD in men than women.Future studies should assess function in both sexes if we are to better understand how to exploit Ca 2+ biology for a neuroprotective strategy.We believe that including both sexes in pre-clinical research could provide powerful insights into the aetiology of Parkinson's.
Alpha-synuclein is well known to be a contributing factor in PD; however, its precise physiological and pathological functions remain incompletely characterised (Alegre-Abarrategui et al., 2019).Here, we identify in male mice that axonal LTCC support of DA release is dependent on α-synuclein.It should be noted that some commercially available wild-type mouse strains including the Harlan C57Bl6/6S (also listed as C57BL/6JOlaHsd) are null for α-synuclein due to a spontaneous deletion (Specht & Schoepfer, 2001), which likely limits LTCC function in that strain also.The mechanism is yet to be determined, but we note a cascade of potentially interacting factors: We previously found that α-synuclein impacts on brain cholesterols and other lipids that shape DAT activity (Threlfell et al., 2021) and found here that cholesterol reinstates an LTCC function lost after α-synuclein deletion in a DAT-dependent manner.
Mutations in the GBA gene are the most common genetic factors in PD (Smith & Schapira, 2022) and appear especially important in female patients with PD (Li et al., 2021).Here, we show that GBA knockdown led to isradipine decreasing DA release in the DLS of Gba +/À female mice, indicating that intact GCase activity normally limits how LTCCs support DA release in female mice.Although we have not identified the cellular pathway responsible for GCase limiting LTCC control of DA, there are a number of possibilities (Do et al., 2019).A downregulation of GCase (via Gba +/À ) might alter the lipid composition of the membrane, allowing LTCC to contribute to release, in a mechanism complementary to that of α-synuclein regulation of LTCC described above; alternatively following a decrease in GCase levels, Ca 2+ might become rate-limiting in the fusion of lysosome to autophagosomes, leading to a compromise in axonal function and DA release following LTCCinhibition (Tian et al., 2015); potentially, dysregulation of lysosomal function might promote ER stress and mitochondrial stress, which may lead to an increase in LTCC signalling to increase ATP production (Zampese et al., 2022).These candidate mechanisms could be tested in future research.
The fast Ca 2+ buffer calbindin-D28k (Calb1) has been considered previously as potentially neuroprotective, but evidence for a neuroprotective role has been mixed.It remains possible that the toxin models typically used are too aggressive to reveal neuroprotection (Airaksinen et al., 1997;Yuan et al., 2013).In any event, we found that Calb1 was important for both regional and sex differences in LTCC regulation of DA release.Under our standard experimental conditions, LTCCs do not contribute to DA release in male NAc (Brimblecombe et al., 2015), and we show here that Calb1 plays a key limiting role.We found also that Calb1 prevents LTCCs from supporting DA release in DLS of females, despite finding previously that Calb1 does not limit the levels of DA released (Brimblecombe et al., 2019).The mechanisms by which Calb1 limits how LTCCs support DA release have not been fully elucidated and are likely to depend on a number of factors, due to the global signalling nature of Ca 2+ and the ability of Calb1 to act as a sink and a source for Ca 2+ (Schwaller, 2020;Schwaller et al., 2002).Potential mechanisms include Calb1 buffering Ca 2+ after entry via LTCCs; Calb1 acting as a Ca 2+ shuttle to facilitate Ca 2+ -dependent inactivation (CDI) of LTCC (Lee et al., 2006); or Calb1 reducing activation of LTCC via inhibition of DAT function (see below).The mechanisms through which Calb1 and LTCC interact in a sex-specific manner are also likely to be complex.Sexual dimorphism in LTCC function has been documented in cardiac tissue (Curl et al., 2008;Machuki et al., 2019;Prabhavathi et al., 2014), while in the brain, Calb1 function can be regulated by sex-linked genes and hormones at transcriptional and protein level (Abel & Rissman, 2012).The mechanisms operating in DA neurons are not yet known but could have important clinical implications for sex-specific incidence of disease and responsiveness to medication.

| DAT promotes and D2R limits LTCC support of DA release
We found that LTCC function is supported by DATs, in males in DLS in wild-types and in NAc after calb1 knockdown.Correspondingly, electrogenic monoamine transporters including the DAT are documented to promote Ca 2+ entry via LTCCs in expression systems (Cameron et al., 2015).We speculate that a similar electrogenic mechanism is responsible in DA axons.It is intriguing to note that in NAc, DA uptake rates are greater after calb1 knockdown than in controls (Brimblecombe et al., 2019), supporting a hypothesis that LTCC support of DA release in NAc after Calb1 knockdown might result from the associated impact of elevated DAT on depolarisation, in addition to a more simple explanation that there is reduced Ca 2+ buffering.The greater LTCC function in DLS than NAc (Brimblecombe et al., 2015) parallels the higher DAT levels in source DA neurons in SNc than VTA (Poulin et al., 2014).However, DAT function alone is not the sole predictor of LTCC function, as LTCC support of DA release in DLS was not different between wild-type and DAT-Cre +/À mice (CalbWT), despite DAT-Cre +/À mice having lower DAT levels and function (Bäckman et al., 2006;Brimblecombe et al., 2019).Alternative explanations for DAT regulation of LTCC function could yet include that LTCCs might regulate DA release in a DAT-dependent manner, so when DAT is inhibited, LTCC dependence is occluded.Our findings together with other literature add to the growing body of evidence that DATs serve more functions than simply clearing extracellular DA, as they participate in the underlying control of DA release (Brimblecombe et al., 2019;Condon et al., 2019;Kile et al., 2010;Siciliano et al., 2018;Venton et al., 2006).It remains to be understood why females appear to have enhanced DAT function yet limited LTCC control of DA release.Understanding this discrepancy in the future could offer important mechanistic insights.
We also found that LTCC support of DA release is inhibited by D2Rs, in male NAc.LTCCs have vice versa been shown to regulate D2Rs (Dragicevic et al., 2014;Duda et al., 2016;Zhang & Sulzer, 2012), and therefore, together these data show that there is reciprocal regulation between LTCCs and D2Rs.D2Rs have been shown previously to inhibit LTCC function in striatal GABAergic projection neurons and cultured midbrain neurons (Hern andez-L opez et al., 2000;Olson et al., 2005;Yasumoto et al., 2004), but this effect is somewhat surprising for DA release because D2Rs do not appear to be activated sufficiently in this preparation to be limiting DA release for these stimuli (Condon et al., 2019;Cragg, 2003;Kennedy et al., 1992), suggesting that there is limited tonic activity at D2Rs.An explanation for this disparity could be that for 1p-evoked DA release, there is no net effect of inhibiting D2Rs on [DA] o due to the wide range of ion channels and intracellular cascades affected by D2Rs, including voltage-gated K + channeldependent mechanisms (Fulton et al., 2011), N-and PQtype Ca 2+ channels (Cardozo & Bean, 1995) and LTCCdependent mechanisms (Hern andez-L opez et al., 2000) (reviewed in Ford, 2014).These multiple outcomes of D2R activity might also provide an explanation for why D2Rs were not detected to limit LTCC function in DLS.From these collected observations, it is evident that roles of LTCCs in DA release, and the effects of isradipine, are plastic; they are variable between DA neuron types; they vary with the activity of a range of biological regulatory mechanisms, and experimental conditions, and whether the subjects are males or females.
The mechanisms by which LTCCs are supporting DA release, and differently with diverse factors, are not yet resolved.Incoming Ca 2+ via LTCCs could directly contribute to the Ca 2+ required for exocytosis or could contribute to other processes that modify axonal physiology and DA release probability.LTCCs lack the synprint site possessed by other types of VGCCs, which tethers VGCCs to the exocytotic machinery (Mochida et al., 2003), suggesting LTCCs might not be located in sufficient proximity to catalyse exocytosis directly.Besides catalysing vesicular exocytosis, there are other potential mechanisms through which LTCCs might support DA release including (but not limited to) supporting AP propagation, intracellular signalling cascades or metabolic support of DA axons.
How APs propagate throughout the vast, branched, unmyelinated and varicose DA axonal arbour (Matsuda et al., 2009) is not well understood but is likely to be highly energy consuming (Bolam & Pissadaki, 2012) and subject to branch point failure.LTCCs have been shown able to modulate AP conductance in optic nerves prior to myelination (Alix et al., 2008) and might modulate AP shape, axonal excitability and AP propagation, and consequently, recruitment of DA release sites in DA axons.A change in AP propagation might also couple to Ca 2+dependent signalling cascades including those regulated by GPCRs that regulate DA release (Fulton et al., 2011).In cardiac myocytes, LTCC function is integrated with physiological need by adrenergic regulation: upon ßadrenergic stimulation, activation of PKA leads to phosphorylation of Cav1.2 and enhanced calcium-dependent inactivation (CDI) of Ca 2+ channels, limiting influx of Ca 2+ during APs (Morales et al., 2019).LTCCs might similarly undergo CDI in DA axons, with resulting loss of effect of isradipine on DA release.
There is the possibility that Ca 2+ entry via LTCCs and downstream activation of RyRs drives mitochondrial generation of ATP (Zampese et al., 2022) that in turn might support axon function and DA release.ATP production, downstream of Ca 2+ entry via LTCCs, is an attractive potential explanation for the detrimental effects of isradipine on DA release, particularly given that we hypothesise that the situations where isradipine decreases DA release reflects a physiologically vulnerable state of the axon.The different factors we report that modulate LTCC support of DA release might change axonal demand for ATP, thus driving differing dependencies on Ca 2+ entry via LTCCs to signal to mitochondria to support the required production of ATP.LTCC function might be a reporter for axons with higher metabolic demand or, since this process contributes a metabolic stress (Pacelli et al., 2015), might drive axonal metabolic stress.Female rodents have been found to have decreased ageing-related ROS production and better mitochondrial function in the brain, compared to males (Gaignard et al., 2015;Guevara et al., 2011), which may be linked to the difference in LTCC function we report here.These hypotheses are not mutually exclusive but are yet to be tested for DA axons.In turn, if a putative LTCC-driven ATP generation is causing a state of vulnerability in PD, for example, via generation of oxidative stress, then the inhibition of LTCCs might be neuroprotective.However, if LTCC function is a readout of axonal vulnerability, by reporting high demand for energy supply, then the inhibition of LTCCs might add a further strain to an already exhausted axonal population.Understanding the energetic demands of dopaminergic axons is non-trivial, but interventions to decrease their energetic demand rather than limit their energetic supply seems more likely to be neuroprotective.
In summary, we show that the factors that govern the variable role of axonal LTCCs in supporting DA release in DLS versus NAc include α-synuclein, DATs, calb1 and D2Rs.These factors can vary between striatal regions and support LTCC function in DLS while preventing LTCC function in NAc, thereby accounting for the disparity between LTCC expression and function in different DA neurons.It is further noteworthy that factors that promote LTCC support of DA release such as male sex, α-synuclein, DAT and the absence of Calb1 correspond to risk factors for parkinsonian degeneration, understanding how these factors interact may allow the stratification of patient cohorts to identify those who would most benefit from targeting LTCC as a neuroprotective strategy with isradipine (Parkinson Study Group STEADY-PD III Investigators, 2020).By facilitating LTCC function these factors might drive metabolic damage to vulnerable neurons, corroborating the hypothesis that LTCCs might then offer a target for a neuroprotective therapy.Alternatively, LTCC support of DA release might be only a readout of vulnerability, a 'canary in the coalmine', when in turn manipulation of LTCC function would not protect vulnerable neurons.The utility of inhibiting LTCCs as a neuroprotective strategy against Parkinson's requires there to be a net benefit from a number of competing mechanisms.The reduction in a Ca 2+ -associated oxidative stress in somatodendritic or axonal compartments might be beneficial, but only if this outweighs the detrimental effects of loss of DA function associated with a decrease in evoked striatal DA release and potential loss of ATP generation.A fuller understanding of the interconnected underlying biology of these Ca 2+ -dependent mechanisms in DA axons will better inform neuroprotective strategies for PD targeting LTCCs and beyond.

F
I G U R E 1 Male sex, presence of α-synuclein and knockdown of Gba promote LTCC support of DA release in DLS.(a) Mean [DA] o ± SEM versus time evoked by a single pulse (1p, arrow) in DLS of male (black) versus female (grey) wild-type mice.(b) Summary of mean peak evoked [DA] o ± SEM, N = 6 mice per sex.(c) Decay constants of falling phases of mean 1p-evoked DA transients in DLS.N = 6 mice per sex.(d) Mean transients for [DA] o ± SEM versus time evoked by 1p (arrow) in control conditions (black) versus isradipine (5 μM, red) in male (left) and female (right) normalised to own control.(e) Peak [DA] o ± SEM as % of pre-isradipine control levels versus time.N = 3 male, 5 female.(f) Summary of peak [DA] o evoked by 1p in DLS from male SNCA-OVX vs Snca-null mice.N = 8 mice.(g) Mean [DA] o ± SEM versus time evoked by 1p (arrow) in control versus isradipine in SNCA-OVX (left) and Snca-null (right), normalised to control.(h) Peak [DA] o ± SEM as % of pre-isradipine control levels versus time.N = 4 mice.(i) Mean [DA] o ± SEM versus time evoked by 1p (arrow) in control and isradipine in female SNCA-OVX mice, normalised to control.N = 3 mice.(j) Peak [DA] o ± SEM as % of preisradipine control levels versus time of female vs male SNCA-OVX mice.(k) Mean [DA] o ± SEM versus time evoked by 1p (arrow) in DLS in control and isradipine in female wild-type (Gba +/+ ) versus Gba-het (Gba +/À ) mice, normalised to control.(l) Peak [DA] o ± SEM as % of preisradipine control levels versus time for female Gba +/+ (N = 3) versus Gba +/À mice (N = 5).(m) Peak [DA] o in isradipine (as % of preisradipine control levels) versus peak [DA] o (μM) in pre-isradipine control conditions plotted for all data shown in conditions in A-L.Male = squares; female = triangles; genotypes indicated by colour.*<.05, ** <.01, *** <.001.

F
I G U R E 2 Calb1 limits LTCC function in NAc in male mice, and in DLS in female mice.(a, b) Mean [DA] o ± SEM versus time evoked by single pulses (1p, arrow) in control and isradipine (5 μM, red) in (a) CalbWT mice or (b) CalbKD mice, in male DLS (left), male NAc (centre), and female DLS (right) normalised to their own control conditions.** P < .01comparison of peak [DA] o .(c) Summary of peak [DA] o remaining after isradipine (% of pre-isradipine control) in CalbWT and CalbKD striatum.** P < .01effect of genotype 2-way ANOVA.N = 3 pairs DLS male and female N = 4 NAc.(d) Representative examples of Western blots for three males and three females showing actin (42 kDa) and Calb1 (25 kDa).(e) Calbindin-D28K/β-actin protein ratios in male and female striatum.Mean ± SEM; colours group technical replicates (i.e., blue points correspond to blot shown in c and are samples from which stats correspond to, N = 3 brains, repeated for three blots, illustrated by red, green and blue).

F
I G U R E 4 LTCC support of DA release is silenced by D2R.(a, c) Mean [DA] o ± SEM versus time evoked by single pulses (1p, arrow) in wild-type male mice in DLS (a) or NAc (c) in control versus isradipine (5 μM, red) alone (left) or in the presence of D2R-antagonist L-741,626 (2 μM, right).(b, d) Peak [DA] o ± SEM (normalised to pre-isradipine level) versus time in control (black) or L-741,626 (blue) in DLS (B) or NAc (D).(e) Summary of peak [DA] o remaining after isradipine (% of pre-isradipine control), in NAc and CPu without (red) or with L-741 (blue) (N = 4).(F) Mean [DA] o ± SEM versus time evoked by 1p (arrow) in DLS in the presence of cocaine + L-741,626 (blue dashed) alone, or with isradipine (red).(g) Mean [DA] o ± SEM versus time evoked by 1p (arrow) in NAc from CalbKD mice with and without isradipine in the prior presence of either L-741,626 (left) or cocaine (right).(h) Peak [DA] o ± SEM (% of pre-isradipine level) versus time during isradipine application in NAc from CalbKD mice in control conditions (black), or the presence of L-741,626 (blue), or cocaine (unfilled).(i) Summary of peak [DA] o remaining after isradipine (% of pre-isradipine control) in NAc in CalbWT (N = 3) or CalbKD mice, N = 4 with L-741 (N = 3) or cocaine (N = 3).*P < .05,** P < .01.

F
I G U R E 5 Segregation of factors to facilitators or attenuators of LTCC support of DA release.(a) Decision tree segregating experimental conditions of sex, genotype and drug condition using the Gini index.Blue to green background shading shows predictive power of a FALSE (blue) or TRUE (green) outcome, respectively.White area shows high uncertainty.Labels on branches indicate recording condition.Left branches for 'isradipine decreases DA release' are FALSE.Labels on final boxes (leaves) indicate outcome: TRUE/FALSE (top), probability of outcome being correct, that is, 0 or 1, indicate all data from that branch has been correctly sorted (middle) and % of datapoints contained within each leaf (bottom).*Cocaine is the presence of cholesterol or L-741,626.(b) Cartoon illustrating which endogenous factors limit (downward arrows) or promote (upward arrows) LTCC support of DA release.