Advancing age and sex modulate antidyskinetic efficacy of striatal Ca_V1.3 gene therapy in a rat model of Parkinson’s disease

Abstract

We previously demonstrated that viral vector-mediated striatal Ca_V1.3 calcium channel downregulation in young adult (3mo) male parkinsonian rats provides uniform, robust protection against levodopa-induced dyskinesias (LID). Acknowledging the association of PD with aging and incidence in male and female sexes, we have expanded our studies to include rats of advancing age of both sexes. The current study directly contrasts age and sex, determining their impact on efficacy of intrastriatal AAV-Ca_V1.3-shRNA to prevent LID induction, removing the variable of levodopa-priming. Considering both sexes together, late-middle-aged (‘aged’; 15mo) parkinsonian rats receiving AAV-Ca_V1.3-shRNA developed significantly less severe LID compared control AAV-scramble(SCR)-shRNA rats, however therapeutic benefit was significantly less robust than observed in young males. When considered separately, females showed significantly less therapeutic benefit than males. Furthermore, aged non-cycling/proestrous-negative female rats were refractory to LID induction, regardless of vector. This study provides novel insight into the impact of age and sex on the variable antidyskinetic responses of Ca_V1.3-targeted gene therapy, highlighting the importance of including clinically relevant age and sex populations in PD studies.

1.1. INTRODUCTION

Parkinson’s disease (PD) is the most common movement disorder, affecting 1% of the global population over the age of 60 (Kryskowiak et al., 2013). Age is the greatest risk factor for PD, and disease incidence increases exponentially in individuals of both sexes over 60 (reviewed in (Collier et al., 2017; Pang et al., 2019; Reeve et al., 2014)).

The dopamine (DA) precursor levodopa remains the gold-standard therapeutic for PD. However, levodopa-induced dyskinesias (LID) are a common side effect with up to 90% of individuals with PD developing LID within a decade of treatment (Bastide and Bezard, 2015; Bove and Calabresi, 2022; Huot et al., 2013; Huot et al., 2022). Although the dyskinetic side effect of long-term levodopa use have been documented for more than 50 years (Fahn, 2015; Fahn and Poewe, 2015), only extended-release amantadine is FDA-approved for treatment of LID (Hauser et al., 2017; Rascol et al., 2022); however it is only partially effective, working best in cases of mild LID and resulting in significant adverse events in ~25% of patients (Hauser et al., 2017). Thus, preventing or delaying the onset of LID without interfering with the motor benefits of levodopa remains a significant unmet clinical need (Cenci et al., 2020).

Based on their involvement in the aberrant synaptic plasticity associated with LID, L-type voltage gated calcium channels (i.e., Ca_V1.3) located on striatal medium spiny projection neurons (SPNs) are a promising antidyskinetic target (Bastide et al., 2015; Day et al., 2006; Schuster et al., 2009; Soderstrom et al., 2010; Steece-Collier et al., 2019; Zhang et al., 2013). Initial proof-of-concept studies attempted pharmacological Ca_V1.3 antagonism using the antihypertensive drugs isradipine (Schuster et al., 2009) and nimodipine (Soderstrom et al., 2010), demonstrating partial and/or temporary antidyskinetic benefit. Similarly, a recent PD clinical trial employing isradipine to antagonize Ca_V1.3 channels in substantia nigra (SN) dopamine (DA) neurons as a neuroprotective therapy reported modest success (Parkinson Study Group, 2013; Parkinson Study Group, 2020; Surmeier et al., 2022). The lack of potent therapeutic benefit in both preclinical and clinical studies is presumed to be due to inadequate target engagement by systemically administered pharmacologic agents (e.g.,(Maiti and Perlmutter, 2020; Surmeier et al., 2022)).

To overcome the limitations of pharmacologic Ca_V1.3 antagonists as well as potential off-target side effects of any systemically administered small molecule, we developed a messenger (m)RNA-level silencing approach involving stereotaxic striatal delivery of an adeno-associated virus (AAV) expressing a short-hairpin RNA (shRNA) targeting Ca_V1.3 (AAV-Ca_V1.3-shRNA). In our initial proof-of-concept gene therapy study, we reported that in young male parkinsonian rats this therapeutic completely prevented LID induction and reversed established severe LID by ~60% (Steece-Collier et al., 2019).

Acknowledging the increasing prevalence of PD in the aging global population and disease incidence in both male and female sexes, we have expanded our studies to include rats of advancing age (referred hereon as ‘aged’) of both sexes. Here we report for the first time contrasting efficacy of mRNA-level striatal Ca_V1.3 silencing to proactively prevent LID induction in young versus aged male, and in aged male versus female parkinsonian rats. Importantly, the current study removed the variable of levodopa-priming, which is known to induce additional aberrant striatal plasticity in the aged, parkinsonian striatum, potentially compounding interpretation of our previous study in aged rats (Caulfield et al., 2023b). This translational study provides crucial new details on the capacity and limitations of striatal Ca_V1.3-targeted gene therapy for treatment of LID in an aging population. It also emphasizes the need for careful age and biological sex considerations, as well as consideration of disease relevant symptomatic interventions in animal-model-based studies to guide this, or any promising therapeutic into effective clinical therapeutic development.

1.2. METHODS

1.2.1. Experimental Animals and Design

All procedures were performed in accordance with the Association for Assessment and Accreditation of Laboratory Animal Care International guidelines and following approval of the Michigan State University Institutional Animal Care and Use Committee. Experimental animals were late-middle aged (“aged”) male and female Fischer (F344) rats (15–18 months old; 15 m.o. at time of vector injection, 18 m.o. at sacrifice) (National Institute on Aging, Division of Aging Biology, Bethesda MD). For simplicity we use the term ‘aged’ in this report. The numbers of experimental rats were: female CAV-shRNA, n=7; male CAV-shRNA, n=8; female SCR-shRNA, n=6; male SCR-shRNA, n=8. In the current study, there was a total of six female rats lacking detectable proestrus (n=4 SCR-shRNA, illustrated only in Fig. 5; n=2 CAV-shRNA, excluded from all figures and behavioral analyses based on details below).

To allow direct comparison of the capacity of striatal CAV-shRNA to prevent the onset of LID between young and aged rats, we employed the identical protocols in the current study employing late-middle aged male rats (15 mos; Fig. 1A) as we previously employed in young male rats (3 mos)((Steece-Collier et al., 2019); John Wiley and Sons, License: 5560220212888). The timing of the vector and 6OHDA surgeries, and the intervals, timing, and dose escalation paradigm for levodopa administration were all identical between experiments involving the young and aged rats.

Figure 1. Aged parkinsonian rats receiving intrastriatal CAV-shRNA are protected from severe LID with increasing levodopa doses.

(A) Experimental timeline. Animals were administered intrastriatal AAV vector (CAV-shRNA or control SCR-shRNA), one week later rendered unilaterally parkinsonian via 6-OHDA, and four weeks post-vector subjected to a levodopa escalation paradigm with two weeks at each dose. Doses range from low (6mg/kg) to high (18mg/kg).

(B) Total dyskinesia severity at each rating day and across all doses of levodopa in rAAV-SCR-shRNA (SCR) and rAAV-Ca_V1.3-shRNA (CAV) rats. CAV rats had significantly lower total LID than control SCR rats at days 1 and 6 of 9mg/kg and days 1 and 6 of 12mg/kg (one-tailed Mann-Whitney U test at each rating day).

(C) Total dyskinesia severity on day 1 of each levodopa dose showing significantly lower LID severity in the CAV vector rats compared to SCR vector rats on each day (one-tailed Mann-Whitney U test at each rating day).

(D) Peak-dose dyskinesia severity at each rating day. CAV rats had significantly lower peak-dose LID than SCR rats (all days were significant: p<0.01). Group difference for CAV vs SCR were compared at each rating day (one-tailed Mann-Whitney U tests).

(E-H) Daily time course of LID severity (20-220 mins post-levodopa) on day 1 of each levodopa dose, demonstrated (Kruskal-Wallis tests with Dunn’s multiple comparisons test).

(I, J) A modified ‘cylinder task’ was used to explore rats’ motor behavior responses in the absence and presence of low dose levodopa. Ambulatory contralateral/clockwise (c.w.) rotations (I) and exploratory rearing behavior (J) were quantified before (“pre-”) and 50 min after (“post-”) an injection of 6 mg/kg levodopa. (I) There was a significant increase in ambulatory rotational response to levodopa only in the CAV-shRNA subjects demonstrated by a significant increase in the number of contralateral rotations in these subjects post-levodopa as compared to pre-levodopa; this was also observed post-levodopa in the CAV compared to SCR subjects (2-way ANOVA’s with uncorrected Fisher’s LSD, time: p=0.0005, vector: p=0.0442; CAV pre vs. post: p=0.0007; SCR pre vs. post: p=0.147; CAV post vs SCR post: p=0.0231). (J) There was a similar modest level of exploratory rearing behavior in both vector groups, with no significant effect of vector, or low dose levodopa (pre- vs post-levodopa time points within vector groups; 2-way ANOVA’s with uncorrected Fisher’s LSD comparisons tests, time: p=0.1454, vector: p=0.0796; SCR pre vs. post: p=0.3967; CAV pre vs. post: p=0.6304). Animals who did not have both a pre- and post-levodopa treatment score were excluded from this analysis. Included number of animals: n=13 SCR; n= 14 CAV.

B - I: lines or bars represent group mean ± SEM.

1.2.2. Vector Design

Ca_V1.3-specific shRNA (Ca_V1.3 (5’-GAAGAGGCGCGGCCAAGAC-3’) or scrambled control shRNA (5’-CAACAAGATGAAGAGCACC-3’) were inserted into a recombinant AAV2/9 genome and titer was determined as previously described (Benskey and Manfredsson, 2016; Benskey et al., 2016; Caulfield et al., 2023b; Sandoval et al., 2019; Steece-Collier et al., 2019). Our vectors employ a separate reporter cassette for GFP, being expressed by the CAG promoter, and the shRNA expressed in another cassette behind an H1 promoter.

1.2.3. Vector Administration and Lesion Surgery

Experimental animals first received unilateral stereotaxic infusion of the rAAV-shRNA (1 × 10¹³ vg/mL) into two dorsolateral sites in the left striatum (2 μLs/site, 0.5 μL/min) per (Caulfield et al., 2023b; Steece-Collier et al., 2019). One week later, rats were rendered unilaterally parkinsonian via stereotaxic administration of 6-hydroxydopamine (6-OHDA) into the left SN and left medial forebrain bundle as routinely done in our dyskinesia studies to ensure sufficient nigrostriatal DA depletion necessary for LID induction (Caulfield et al., 2023b; Steece-Collier et al., 2019). This is based on the fact that while often underappreciated, near complete loss of striatal DA innervation is generally required for LIDs to manifest, regardless of species (e.g., (Cenci and Crossman, 2018; Cotzias et al., 1969; Konradi et al., 2004; Morin et al., 2014; Zhang et al., 2013)). Vector injection and 6-OHDA lesion were in the same brain hemisphere.

1.2.4. Levodopa Administration and Dyskinesia Rating

Four weeks after vector administration, a levodopa dose-escalation paradigm was initiated (6, 9, 12, 18 mg/kg levodopa (L-3,4-dihydroxyphenylalanine methyl ester hydrochloride (HCl) plus 12 mg/kg benserazide HCl, Sigma-Aldrich; subcutaneous (s.c.); Figure 1A). Each dose was administered once daily for 2 weeks (Monday–Friday), and LID severity rated at 50-minute intervals over 220 minutes beginning 20 minutes post injection on days 1, 6, and 10 per (Steece-Collier et al., 2019). The term ‘LID’ is used here to refer to abnormal involuntary behaviors in parkinsonian rats in the presence of levodopa, including dystonia, hyperkinesia, and/or stereotypies (Caulfield et al., 2021; Maries et al., 2006). LID severity was evaluated as ‘total’ LID severity (cumulative LID scores per animal combined across a 220 min rating period) and ‘peak’ LID severity (maximal LID severity score obtained during a 220 min rating period) (Caulfield et al., 2021; Maries et al., 2006). For tables containing detailed rating criteria and video descriptions of rating methods, please see (Caulfield et al., 2021; Maries et al., 2006). Rats were withdrawn from levodopa one week prior to euthanasia per (Caulfield et al., 2023b).

1.2.5. Cylinder Motor Testing

To provide insight into whether CAV-shRNA gene therapy impacted the motor effects of levodopa, we employed a modified cylinder task to explore rats’ motor behavior (i.e., rotational asymmetry & rearing) in the absence (pre-levodopa; “pre-”) and presence (post-levodopa; “post-”) of low dose (6 mg/kg) levodopa (Fig. 1, I,J). This low dose was employed because it results in minimal/less severe LID that can be “overridden” by ambulatory (i.e., rotational) and exploratory rearing behavior observed when rats are placed in a novel environment (i.e., cylinder). Rats were placed in a clear plexiglass cylinder (diameter: 16 cm, height: 25 cm) and videotaped for 5 minutes prior to and 50 minutes following levodopa injection. The number of 360° contralateral rotations and rears were quantified by a blinded investigator as previously detailed (Steece-Collier et al., 2019).

1.2.6. Estrus Determination

Saline vaginal lavage was performed on all female rats for three consecutive days one week prior to sacrifice. Lavage fluid was spread on charged slides and allowed to dry per (McLean et al., 2012). Slides were stained with 0.1% filtered cresyl violet and imaged using a Nikon Eclipse 80i microscope (brightfield, 10x objective). The stage of the rat’s cycle was determined by microscopic evaluation of the proportion of cornified or nucleated epithelial cells and leukocytes (Marcondes et al., 2002; McLean et al., 2012). Based on experimental findings detailed below, only rats with a detectable proestrus smear (‘Proestrus+’) were included in combined (female and male) behavioral analyses.

1.2.7. Euthanasia and Immunohistochemistry

Rats were deeply anesthetized with pentobarbital (250 mg/kg i.p.; Beuthanasia-D Special, VetOne, Boise, ID, USA) and intracardially perfused with heparinized room temperature 0.9% saline followed by cold 4% paraformaldehyde. Brains were removed and post-fixed in 4% paraformaldehyde at 4°C for 24 hours, submersed and stored in 30% sucrose solution until time of sectioning at 4°C. Brains were sectioned coronally at 40 μm thickness using a sliding microtome and stored in a cryoprotectant solution at −20°C.

Tissue sections (40 μm) were processed (1-in-6; 240 μm spacing) for the DA synthetic enzyme tyrosine hydroxylase (TH) (Millipore-AB152b rabbit anti-TH, 1:4,000) and the vector reporter tag green fluorescent protein (GFP) (Millipore-Ab290 rabbit anti-GFP, 1:20,000) immunohistochemistry (IHC) per (Caulfield et al., 2023b; Steece-Collier et al., 2019). The degree of SN DA neuron depletion was quantified using total enumeration of TH-positive neurons (Caulfield et al., 2023b). Striatal volume of GFP-positive vector transduction was determined using the Cavalieri Estimator function in Stereo Investigator^® (MBF Bioscience, Williston, VT) as previously detailed (Caulfield et al., 2023b).

1.2.8. Cacna1d Transcript Analyses

CaV1.3 protein and mRNA (Cacna1d) transcripts localize principally to neuronal cell bodies, however they also are abundant in proximal dendrites and synaptic regions including within dendritic spines (Olson et al., 2005; Zhang et al., 2006; Zhang et al., 2013). For cell-specific quantification of Ca_V1.3 mRNA (Cacna1d), sections were fluorescently labeled for Cacna1d using RNAScope^® in situ hybridization (ISH) (ACD, Newark, CA; Probe-RnCacna1d; REF 409361; NM_017298.1, nucleotides 5401-6474), then co-stained via IHC for GFP and neuron-specific proteins HuC/D (Thermo-Fisher-A21271 mouse anti-HuC/HuD, 1:2000) for identification of transduced vs. non-transduced neurons. The number of Cacna1d transcripts in GFP+ and HuC/D+ cell bodies were quantified using Imaris^® (v. 9.8.0, Oxford Instruments) per (Caulfield et al., 2023b). The fold change in Cacna1d mRNA expression between vector-transduced and non-transduced cells is expressed as the average number of Cacna1d transcripts per GFP+ (transduced) cell in ipsilateral striatum normalized to the average number of Cacna1d transcripts per HuC/D+ (GFP−, non-transduced) cell in contralateral striatum (Caulfield et al., 2023b).

For tissue-level Cacna1d transcript determination (including soma and dendrites, both transduced and non-transduced), Cacna1d mRNAs were 3,3′-diaminobenzidine (DAB) labeled via ISH (identical RNAScope^® probe as above). The DAB-positive area in the dorsolateral striatum was quantified in ImageJ (v. 1.53t, National Institutes of Health) per (Steece-Collier et al., 2019). Total striatal Cacna1d fold change is expressed as a ratio of DAB-labeled transcript area in ipsilateral vs. contralateral dorsolateral striatum.

For dorsal lateral heat mapping of GFP expression, Z-stacks images of fluorescent GFP-immunoreactivity (GFP-ir) in a single striatal section per animal were taken at 20x using a Zeiss Axioscan Z.1. The stacks were processed with Zeiss Zen Blue for export. Striatal areas of interest (Figure 2B) were selected and GFP-ir density analyzed with heat maps created using the open-source platform FIJI (ImageJ) (Schindelin et al., 2012). The heat map (also called color map) is a single channel gray 8-bit image that has color assigned to it via a lookup table (LUT). A LUT is a predefined table of gray values with matching red, green, and blue values so that shadows of gray are displayed as colorized pixels. Thus, differences in color in the pseudo-colored image reflect differences in intensity of the object rather than differences in color of the specimen that has been imaged. A low value is represented by the blue end of the spectrum where a high expression value is represented by a red coloring. White, represents an absence of signal.

Figure 2. CAV-shRNA rats have decreased levels of striatal Cacna1d transcript confirming successful Ca_V1.3 target engagement.

(A) Percentages of striatal AAV-GFP transduction between SCR-shRNA and CAV-shRNA cohorts. In all rats, the percent area of GFP-immunoreactivity in the striatum was greater than 80% and equivalent between vector groups (one-tailed Mann-Whitney U tests; p=0.6678).

(B) Representative micrographs showing dorsolateral striatal vector transduction indicated with GFP-reporter IHC; adjacent heatmaps show the intensity of GFP-immunoreactivity of the corresponding striatal areas. The top images (green outline) show a representative example of “high GFP coverage/low LID” rats. The bottom images (blue outline) show a representative example of a “low GFP coverage/high LID” rats. They are both highlighted as enlarged color-coded data points in panel C as well. Scale bars in GFP IHC micrographs = 500μm.

(C) XY plot of percent of GFP+ dorsolateral striatum against LID severity showing a trend that higher LID severity is associated with relatively lower vector coverage of the dorsolateral striatum (simple linear regression, two-tailed Pearson; p=0.1169). Enlarged blue and green data point correspond to examples in panel B.

(D) Representative brightfield scans of a coronal section processed with DAB-ISH for Cacna1d (Cav1.3). Left: striatal hemisphere showing dorsolateral region of quantification (red contour; scale bar = 1000μm). Center: Zoomed image of brown (DAB+) punctate Cacna1d transcripts in the dorsolateral quadrant (scale bar = 50μm). Right: Zoomed image showing DAB+ Cacna1d mRNA pseudo-colored red in ImageJ using the “Threshold” function (scale bar = 25μm).

(E) Tissue-level (e.g., cell soma and dendrites, transduced and non-transduced cells) striatal Cacna1d. Fold change calculated as the ratio of DAB+ area (Cacna1d mRNA) in the ipsilateral to the contralateral striatum. Striatal Cacna1d is significantly decreased in CAV-shRNA as compared to SCR-shRNA rats (unpaired two-tailed t-test).

(F) Quantification in vector-transduced cells (GFP+) reveals significantly fewer Cacna1d transcripts in the cells of the CAV-shRNA striatum as compared to SCR-shRNA (unpaired two-tailed t-test). Cacna1d fold change is expressed as the ratio of Cacna1d mRNA in transduced (GFP+), ipsilateral to non-transduced (GFP−, HuC/D+), contralateral cells.

(G) Representative confocal images showing vector-transduced cells (GFP+, green) and Cacna1d transcripts (cyan puncta) in the striatum of a SCR rat. Left image scale bar = 40μm; Right image scale bar = 20μm

(H) Representative confocal images showing vector-transduced cells (GFP+, green) and Cacna1d transcripts (cyan puncta) in the striatum of a CAV rat, illustrating fewer Cacna1d mRNA per transduced cell relative to those in the SCR striatum (G). Left image scale bar = 40μm; Right image scale bar = 20μm.

A, E, F: represent group mean ± SEM.

1.2.9. Statistical Analyses

LID data are created using ordinal rating scales and thus analyzed with non-parametric statistics including Mann-Whitney U-test (for between-subject contrasts; one-tailed per apriori hypotheses to test for specific directional differences between groups) and Friedman or Kruskal-Wallis with Dunn’s multiple comparison test (for within subjects tests). Parametric data including cylinder behavior data i.e., (number of rotations or rears) were compared using 2-way ANOVA with Šídák’s multiple comparisons tests, with vector and levodopa status as sources of variation. SCR vs CAV parametric comparisons of Cacna1d fold change and percent transduction were analyzed using unpaired, two-tailed t-tests. All analyses were performed with Prism GraphPad.v9.4.1 for MacOSX.

1.3. RESULTS

1.3.1. Striatal CAV-shRNA prevents severe LID in aged parkinsonian rats.

All parkinsonian rats included in this study had a nigral lesion of >99% (average ± SD: 99.37% ± 0.7923). Group averages were not significantly different (male SCR 99.98% ± 0.0071; male CAV 99.97% ± 0.0245; female SCR + Proestrus 99.97% ± 0.0097; female CAV + Proestrus 99.98% ± 0.0145; female SCR – Proestrus 99.99% ± 0.0095; female CAV − Proestrus 99.10% ± 1.273; Kruskal-Wallis with Dunn’s multiple comparison test).

An experimental timeline using post-vector levodopa dose escalation for LID induction (Figure 1A), recapitulating our previous work in young adult (3 mo) male rats (Steece-Collier et al., 2019) was used to allow for direct comparison of striatal CAV-shRNA efficacy in young versus aged male rats, and for aged female versus male comparisons. When considering male and female rats together, the current data demonstrate that in control rats receiving intrastriatal SCR-shRNA, there was a progressive escalation in the severity of total LID severity over time with increasing doses (Figure 1B). In contrast, rats receiving intrastriatal CAV-shRNA exhibited significantly lower total LID severity compared to SCR-shRNA rats, significant at days 1 and 6 of 9 mg/kg and days 1 and 6 of 12 mg/kg total LID (Figure 1B). Interestingly, differences in total LID severity between SCR-shRNA and CAV-shRNA rats were most evident on the first day of each increasing dose of levodopa (Figure 1C). When considering peak LID severity (Figure 1D), rats receiving intrastriatal CAV-shRNA exhibited significantly lower LID severity compared to SCR-shRNA across the entire timeline. Similarly, examination of the daily time course of LID expression (20, 70, 120, 170, 220 mins post levodopa) on day 1 of each dose revealed that LID severity in SCR-shRNA rats was significantly more pronounced compared to CAV-shRNA rats at multiple time points across all levodopa doses (Figure 1E-H).

1.3.2. CAV-shRNA does not interfere with levodopa motor response in aged rats.

Both CAV-shRNA and SCR-shRNA recipients displayed similar low levels of ambulatory behavior noted as contralateral rotational behavior in the absence of levodopa using a modified cylinder task (Figure 1I). However, in response to low dose (6mg/kg) levodopa, there was a significant increase in contralateral rotations in CAV-shRNA rats post-levodopa compared to pre-levodopa, and post-levodopa between CAV-shRNA vs SCR-shRNA rats (Figure 1I, 2-way ANOVA’s with uncorrected Fisher’s LSD comparisons tests, statistics in graph). There was a trend of increased ambulatory rotational behavior after levodopa treatment in the SCR-shRNA rats, however this did not reach significance at this low dose in these aged control rats. This is similar to what was seen in young male SCR-shRNA rats (Steece-Collier et al., 2019) suggesting enhanced motor response to levodopa in CAV-shRNA recipients.

When examining motor rearing behavior in a novel environment (i.e., plexiglass cylinder), both vector groups showed similar active, albeit modest levels rearing in the absence of levodopa, demonstrating no negative impact of either vector on this locomotor behavior. In contrast to a significant enhancement of this rearing behavior that was observed in response to low-dose levodopa in young male CAV-shRNA recipients (Steece-Collier et al., 2019), there was no further activation of this exploratory behavior in response to low dose levodopa with either vector in these aged rats (Figure 1J, 2-way ANOVA’s with uncorrected Fisher’s LSD comparisons tests, stats in figure legend).

For further evidence that our CaV1.3-shRNA therapeutic does not negatively impact motor behavior and activation following levodopa, the reader is referred to our previously published video ((Steece-Collier et al., 2019), mds27695-sup-0001-VideoS1.mp4) that demonstrates significant activation of motor behaviors (sniffing, exploring, contralateral walking around the cage perimeter) for approximately 2 hours post-levodopa injection -- in the absence of LID. This contrasts the severe LID profiles and tight rotational behavior in response to the Scr-shRNA control vector.

1.3.3. Striatal Cacna1d is downregulated in CAV-shRNA-injected rats.

We compared the volume of vector transduction indicated by GFP-positive immunoreactivity (GFP+) in the full ipsilateral striatum between CAV-shRNA and SCR-shRNA groups demonstrating no significant difference between vector groups (Figure 2A). We next restricted our examination to the vector coverage in only the dorsolateral striatum (Figure 2B-D)—the striatal area associated with LID in rats (e.g., (L et al., 2017)). This region-specific analysis revealed a trend, albeit not statistically significant, suggesting a positive relationship between the degree of vector coverage (GFP+ area) and LID severity in individual CAV-shRNA animals, i.e., the greater the vector coverage the lower the LID severity (Figure 2C). There was no relationship between the percentage of striatal volumes covered by GFP and LID severity when the entire striatum was examined (data not shown).

To determine the degree to which CAV-shRNA administration downregulated Ca_V1.3 (Cacna1d), mRNA transcript levels were measured using total parenchymal quantification and cell-specific and methods. Parenchymal quantitation (i.e., soma and dendrites, transduced and non-transduced cells) of Cacna1d transcripts in the dorsolateral striatum revealed significant downregulation in CAV-shRNA compared to SCR-shRNA striata (Figure 2E). On average, SCR-shRNA rats showed approximately equal levels of striatal Cacna1d mRNA between hemispheres, while CAV-shRNA rats had significantly fewer transcripts in the injected striatum. Within vector-transduced (GFP+) cells, intrastriatal CAV-shRNA resulted in a significant reduction in Cacna1d transcripts in the ipsilateral striata in CAV-shRNA rats as compared to SCR-shRNA (Figure 2F-H). There was also a decrease in Cacna1d transcripts in the striatum of these aged rats in response to SCR-shRNA, albeit highly variable and significantly less than in the CAV-shRNA recipients.

1.3.4. Protection from LID induction is less robust in aged compared to young male CAV-shRNA rats.

Using our previous experimental design (Steece-Collier et al., 2019) that allowed for a direct comparison of young and aged male parkinsonian rats, as expected, aged male rats receiving SCR-shRNA exhibited an increase in both total and peak dose LID severity over the time course of levodopa escalation. However, aged rats developed a lower level of total LID severity compared to that observed in young male parkinsonian rats at the highest doses (12 and 18mg/kg; Figure 3A,D); peak dose LID were roughly equivalent between ages at these higher doses. In contrast to the SCR group, aged male CAV-shRNA rats demonstrated significant dampening of LID severity for both total and peak LID scores (Figure 3A, B, D, E). When comparing the LID score for each individual animal on day 1 of 6mg/kg levodopa dose versus day 1 of 18mg/kg levodopa, in young rats, there was a significant increase only in the SCR-shRNA group, with CAV-shRNA rats uniformly showing sustained resistance to LID induction (Figure 3F). In the aged parkinsonian rats, there was an increase in LID severity in both the SCR-shRNA and CAV-shRNA groups, albeit to a lesser degree and with noticeable variation in the protection from individual to individual in the aged CAV-shRNA rats (Figure 3C). Notably, the degree of LID protection was greater in young male rats, i.e., young CAV-shRNA rats exhibited no LID escalation, with no significant difference in LID severity between day 1 of low-dose, 6 mg/kg and day 1 of the highest dose, 18 mg/kg levodopa (Figure 3D-F). While there is clearly significant therapeutic benefit of CAV-shRNA in aged rats, the degree of variability dramatically contrasts the highly uniform antidyskinetic response in young adult male rats (e.g., Figure 2C vs. Figure 2F).

Figure 3. Intrastriatal AAV-CaV1.3-shRNA protects aged parkinsonian male rats from severe LID, but less effectively than in young male parkinsonian rats.

(A) Total dyskinesia severity (the sum of all individual attributes of dyskinesias that were scored in rats over the 220-minute post-levodopa rating period) shown over time with increasing levodopa doses in aged male SCR and CAV parkinsonian rats (one-tailed Mann-Whitney U test at each rating day).

(B) Peak-dose dyskinesia severity (dyskinesia score at the 70-minute rating time point when LID severity is generally maximal) in aged male SCR and CAV rats. (one-tailed Mann-Whitney U tests).

(C) The difference in total dyskinesia severity between day 1 of 6 mg/kg and day 1 of 18 mg/kg for individual aged male SCR and CAV rats. Both vector cohorts show an increase in LID severity between these time points, though SCR rats demonstrate more severe LID elevation (one-tailed Mann-Whitney U tests).

(D) Total dyskinesia severity over time with increasing levodopa doses in young adult male SCR and CAV rats. Groups were compared at each rating day via one-tailed Mann-Whitney U tests (p<0.01 for all days except day 1 of 6 mg/kg). Modified with permission (John Wiley and Sons, License: 5560220212888).

(E) Peak-dose dyskinesia severity in young adult male SCR and CAV rats. Groups were compared at each rating day via one-tailed Mann-Whitney U tests (p<0.01 for all days except day 1 of 6 mg/kg). Modified with permission (John Wiley and Sons, License: 5560220212888).

(F) The difference in total dyskinesia severity between day 1 of 6 mg/kg and day 1 of 18 mg/kg for individual young adult male SCR and CAV rats. SCR rats show a significant increase in LID severity between time points; CAV rats maintain consistent low-level LID (one-tailed Mann-Whitney U tests).

(G) The percent reduction in total dyskinesia severity was calculated by comparing individual CAV rats’ Total LID scores to the average SCR age group Total LID score on each rating day ((1 — CAV/SCR)*(−1)). Generally, young rats show less variability and a greater percent reduction in Total LID when compared to aged rats; significant on 9, 12, and 18 mg/kg levodopa rating days (two-tailed Mann-Whitney U tests).

(H) The percent reduction in peak-dose dyskinesia severity was calculated as described in H. Compared to aged rats, young rats show less variability and a greater percent reduction in Peak LID, significant on day 10 of 6mg/kg and all rating days at 9, 12, and 18 mg/kg levodopa (two-tailed Mann-Whitney U tests).

A-H: error bars represent group mean ± SEM. Portions of the data from young adult male rats (D, E) were modified with permission (John Wiley and Sons, License: 5560220212888).

As above, total dyskinesia scores in the aged SCR-shRNA rats were lower overall as compared to young SCR-shRNA rats (Figure 3A,D). To visualize a comparison of CAV-shRNA efficacy between young and aged male rats, the “percent reduction” in LID was calculated: a ratio of each individual CAV rat’s LID severity to the SCR age group average LID for each rating day ((1-CAV/SCR)*(−1)) (Figure 3G,H). Specifically, there were significant differences in percent reduction in both total and peak-dose LID between young and aged cohorts on most 9, 12, and 18 mg/kg rating days (Figure 3G,H). These comparisons demonstrate, again, a maintained but diminished therapeutic efficacy and increased variability of CAV-shRNA in aged rats.

1.3.5. The antidyskinetic benefit of striatal CAV-shRNA is greater and more uniform in aged males than aged females.

While significant protection against LID induction was observed in aged male CAV-shRNA rats (Figure 3A), total LID severity in aged female CAV-shRNA is statistically indistinguishable from their SCR-shRNA counterparts except on two ratings days; day 1 of 6mg/kg levodopa and day 1 of 9mg/kg levodopa (Figure 4A). The continued escalation of LID severity and individual-to-individual variability is evident when comparing total LID severity on day 1 of 6mg/kg levodopa dose versus day 1 of 18mg/kg levodopa for both vector groups (Figure 4B), with both vector groups showing significant escalation in total LID severity. When total LID scores for days 1, 6, 10 for each dose are averaged for each vector group, these group comparisons demonstrate that at the lower levodopa doses (i.e., 6, 9, 12 mg/kg) the female CAV-shRNA group compared to SCR-shRNA group has significantly lower LID scores, an effect is lost at the highest dose (18mg/kg; Figure 4C). In contrast, the aged male CAV-shRNA group has significantly lower LID scores compared to SCR-shRNA group, and the benefit is maintained throughout all the levodopa doses (Figure 4D). As before, the percent reduction in LID was calculated: a ratio of each individual CAV rat’s LID severity to the SCR age group average LID for each rating day (Figure 4E). These comparisons demonstrate, again, a diminished therapeutic efficacy and increased variability of CAV-shRNA in aged female compared to male rats, particularly at higher levodopa dosing (Figure 4E).

Figure 4. Intrastriatal AAV-CaV1.3-shRNA in aged parkinsonian female rats are only modestly protects against the development of severe LID, and only at low to moderate doses of levodopa.

(A) The total LID score of individual aged female SCR (grey-dashed) and CAV (purple) rats across the experimental time course. Only days 1 of doses 6mg/kg and 9 mg/kg levodopa are significantly different between vector groups (one-tailed Mann-Whitney U test at each rating day).

(B) The difference in total dyskinesia severity between day 1 of 6 mg/kg and day 1 of 18 mg/kg for individual aged female SCR and CAV rats (Figure, split by sex vector cohorts show an increase in LID severity between time points, though SCR rats as a group demonstrate a slightly more severe LID elevation (one-tailed Mann-Whitney U tests).

(C) When data from the three rating days at each levodopa dose are combined to examine group differences, female SCR and CAV group averages differ significantly at 6, 9, and 12 mg/kg doses (two-tailed Mann-Whitney U tests).

(D) When data from the three rating days at each levodopa dose group are combined, male SCR and CAV group averages differ significantly at each dose (one-tailed Mann-Whitney U tests).

(E) The percent reduction in total dyskinesia severity was calculated by comparing individual CAV rats’ total LID scores to the average SCR age group total LID score on each rating day ((1 - CAV/SCR)*(−1)). Generally, female parkinsonian rats show greater individual variability and less LID reduction in total LID when compared to male aged rats, although without reaching statistical significance (daily LID compared with two-tailed Mann-Whitney U tests; p>0.2000 at each time point, except the lowest p-values at D6 of 18mg/kg (p=0.0846) and D10 of 18 mg/kg (p=0.0753).

(F) Despite a distinct trend of decreased Cacna1d transcripts in CAV-shRNA compared to SCR-shRNA striatum, the tissue-level (e.g., cell soma and dendrites, transduced and non-transduced cells) transcript data (from Figure 2E), when split by sex demonstrates no significant effect of sex or vector (ordinary two-way ANOVA; sex: F(1, 25)= 0.1491, p=0.7026; vector: F (1, 25)= 4.121, p= 0.0531), presumably due to high variability; and no interaction (F(1,25)= 0.0006, p=0.9804). Data was calculated as the ratio of DAB+ area (Cacna1d mRNA) in the ipsilateral to the contralateral striatum.

(G) When Cacna1d transcripts in vector-transduced cells (GFP+) of the CAV-shRNA compared to SCR-shRNA striatum were split by sex, ordinary two-way ANOVA indicated a significant effect of sex and vector (sex: F(1, 23)= 13.93, p=0.0011; vector: F (1, 23)= 17.55, p=0.0004), with no interaction (F (1, 23)= 2.019, p=0.1688). Post-hoc (Tukey’s multiple comparisons) revealed significant difference between F CAV vs F SCR (p=0.0052), and F SCR vs M SCR (p=0.0061). Cacna1d fold change was expressed as the ratio of Cacna1d mRNA in transduced (GFP+), ipsilateral to non-transduced (GFP−, HuC/D+), contralateral cells.

(H) Percent of striatal AAV-GFP transduction volume between male and female CAV-shRNA cohorts. In all rats, the percent area of GFP-immunoreactivity in the striatum was greater than 80% (two-tailed Mann-Whitney U tests, p=0.8569).

(I) Percent of dorsolateral striatal area of AAV-GFP transduction area between male and female CAV-shRNA cohorts (two-tailed Mann-Whitney U tests).

A, C-I: error bars represent group mean ± SEM.

To determine whether a differential degree of downregulation Cacna1d was the underlying cause of male vs female differences, we first compared total parenchymal/tissue level Cacna1d (e.g., cell soma and dendrites, transduced and non-transduced cells) in aged male and female vector groups. Despite a distinct trend of decreased Cacna1d transcripts in CAV-shRNA compared to SCR-shRNA striatum in both male and female striatum, when this data (i.e., Figure 2E) was split by sex, statistical analyses revealed no significant effect of sex or vector (ordinary two-way ANOVA, statistics in Figure 4F legend). Surprisingly, when examined at the cellular level, there was no significant difference in the expression of Cacna1d in GFP+ striatal cells between the aged male CAV-shRNA vs SCR-shRNA groups (Figure 4G); however, there was a significant difference in the aged female CAV-shRNA compared to SCR-shRNA groups (statistics in Figure 4G legend).

Although we did not observe a marked difference between the sexes in the percent total striatal volume of GFP as a measure of vector transduction (Figure 4H), the percent of the dorsolateral striatum immunoreactive for GFP was significantly lower in the female CAV-shRNA cohort (Figure 4I). These data suggest that the degree of CAV-shRNA vector transduction in the dorsolateral striatum is paramount for protection from severe LID in this rat model of PD.

1.3.6. Aged female rats undergoing reproductive senescence are resistant to developing severe LID.

We examined estrous cycle status to provide insight into whether it correlated with the highly variable results in our aged female rats. Histological examination of vaginal lavage smears indicated that 6 female rats (4 SCR, 2 CAV) no longer had a regular estrous cycle. Comparing individual females’ behavior, split by estrus status (‘proestrus+’ vs. ‘proestrus−’), we demonstrate here for the first time that aged female rats, specifically those receiving a control SCR-shRNA vector, which are proestrus− show lower total and peak-dose LID across the dose escalation experimental timeline compared to proestrus+ SCR-shRNA females (Figure 5A,C). Indeed, on day 1 of levodopa doses 9, 12, and 18 mg/kg there was significantly less severe total LID in proestrus− as compared to proestrus+ control SCR-shRNA rats (Figure 5B). These differences are also evident in the daily time course of each rating day, with proestrus+ SCR-shRNA females showing consistently higher LID than proestrus− SCR-shRNA females for each levodopa dose (Figure 3D-G). As a result, proestrus− SCR-shRNA females were excluded from all group analyses as were proestrus− CAV-shRNA since it was not possible to determine whether low-level LID was a result of therapeutic protection or reproductive senescence.

Figure 5. Aged, proestrus-negative (non-cycling) female rats are resistant to developing severe LID.

(A) Total dyskinesia severity over time with increasing levodopa dose in proestrus+ (orange, cycling) and proestrus− (gray, non-cycling) female SCR-shRNA control rats. Proestrus+ SCR-shRNA females exhibited higher total LID than proestrus− females on all rating days, with significance on rating day 1 of 9 mg/kg, day 1 of 12 mg/kg, days 1, 6 of 18 mg/kg (one-tailed Mann-Whitney U tests). Of these SCR rats, n=4 were proestrus− (not included in previous figures) and n=6 were proestrus+ (these 6 animals were included in the SCR group in previous figures).

(B) Total dyskinesia severity between proestrus+ (P+) and proestrus− (P−) female SCR rats on day 1 of each levodopa dose (one-tailed Mann-Whitney U tests).

(C) Peak-dose dyskinesia severity in proestrus+ and proestrus− female SCR-shRNA rats. Proestrus+ rats had significantly higher peak-dose LID than non-cycling rats on rating days 1, 6, 10 of 6 mg/kg; days 1, 6, 10 of 9 mg/kg; days 1, 10 of 12 mg/kg; days 1, 10 of 18 mg/kg (one-tailed Mann-Whitney U tests).

(D-G) Daily LID rating time courses (20-220 mins post-levodopa) on day 1 of each levodopa dose for individual female SCR rats (top) and by group averages (bottom) (Kruskal-Wallis tests with Dunn’s multiple comparisons test).

A-G: error bars represent group mean ± SEM.

1.4. DISCUSSION

The current study is part of our preclinical series of investigations using rat models of PD aimed at defining the capacity and limitations of striatal Ca_V1.3 RNA-interference (RNAi) gene therapy to effectively ameliorate LID (Caulfield et al., 2023b; Steece-Collier et al., 2019), a remaining unmet need in the clinical management of PD. Our initial investigations in young adult male parkinsonian rats demonstrated the ability of this therapeutic to provide some of the most profound antidyskinetic efficacy reported to date (Steece-Collier et al., 2019). Acknowledging the association of PD and aging, as well as sex-related differences in PD and LID (Colombo et al., 2015; Eusebi et al., 2018; Gomez-Mancilla and Bedard, 1992; Zappia et al., 2005), we have crucially expanded our translational investigations into rat models that include advancing age and male and female sexes.

In our first study employing male and female rats of advanced middle age (Caulfield et al., 2023b), rats were levodopa-primed to express stable, mild LID prior to receiving intrastriatal vector delivery. This study revealed that only about half of the aged CAV-shRNA vector rats with pre-existing LID demonstrated protection from LID escalation with subsequent levodopa dose escalation (Caulfield et al., 2023b) -- unlike the highly uniform efficacy in young rats. Under the pretense of fully understanding variables that will be important for clinical development of this promising therapeutic, the current study was designed to provide the first direct comparison antidyskinetic potency of AAV-Ca_V1.3-shRNA in parkinsonian male rats of advanced age to that observed in their young counterpart in levodopa-naïve cohorts; removing the variable of levodopa priming and focusing on aging. In addition, while our first study in aged rats (Caulfield et al., 2023b) provided the first insight that there was more variability in aged female compared to male rats, prior to the current report, there was no information on how estrous cycle status in aged females might have impacted this variability.

Cumulatively our translational series of work highlights the increased complexity of including aged animals of both male and female sexes in preclinical studies, as well as the import of doing so. The current study also emphasizes additional variables, for example, the degree of vector coverage in the dorsolateral striatum that may be tied to anatomical size differences between male and female, and young and aged rats. It has been recently shown that rat brain volume increases with age, and that the proportion of tissue varieties (i.e., grey matter, white matter) also vary across adulthood (MacNicol et al., 2021). Historically, our experimental systems (including our stereotaxic coordinates) are based on measurements from young adult male rats (Swanson, 2018). These studies reinforce that in order to achieve optimal transduction of vector-mediated therapeutics, technical alterations need to be made to target the selective, brain region specific area(s) of interest in a strain, age, and sex-specific manner. Recent technological advancements are improving brain visualization for preclinical (and clinical) stereotaxic delivery of desired therapeutic and should be used in future studies (Kleven et al., 2023).

An additional complexity of gene therapy, including the current CAV-shRNA, is the cellular specificity of transduction. In the current study, the AAV2/9 serotype and optimized capsid design of our viral vector efficiently transfects neurons, and while it does so with greater volumetric spread than other serotypes in the CNS (Haery et al., 2019; Jackson et al., 2015; Kanaan et al., 2017; Lukashchuk et al., 2016; Watakabe et al., 2015), it is not specific for D1-receptor (D1-r) containing ‘direct pathway’ versus D2-r containing ‘indirect pathway’ SPNs; and it likely transfects smaller populations of interneurons. In the rodent striatum, approximately 95% of all neurons are SPNs (for review (Mao et al., 2019; Tepper et al., 2010)). In a previous publication we provide extensive discussion on the interplay of Ca_V1.3 channels and D1 and D2 receptors, and the potential differential impact of our CAV-shRNA on these neuron populations in the transduced striatum (for details please see (Caulfield et al., 2023c)). Additionally, and beyond the scope of this report, the remaining approximately 5% of non-SPNs are a variety of interneuron subtypes that play important roles in striatal plasticity. Most striatal interneurons are sensitive to antagonism of Ca_V1 L-type calcium currents (e.g., (Goldberg et al., 2009; Rendon-Ochoa et al., 2018)), and two of these populations, cholinergic interneurons (ChI) and low-threshold spike interneurons (LTSI), have been implicated in LID (for review (Paz and Murer, 2021; Shen et al., 2022)). Future studies are required to determine the extent to which Cacna1d transcript is impacted in the various subpopulations of interneurons in our CAV-shRNA recipients, and the extent to which this might impact the antidyskinetic efficacy of vectored Ca_V1.3 downregulation.

1.4.1. Ca_V1.3 Gene Therapy in the Aging and Parkinsonian Brain

Increased intracellular Ca²⁺ is a feature of neuronal aging, mediated by enhanced Ca²⁺ influx and/or reduced Ca²⁺ buffering (reviewed in (Griffith et al., 2000; Marambaud et al., 2009)). Ca²⁺ elicits wide-ranging effects in its role as an intracellular messenger and metabolic regulator, and disruptions in neuronal Ca²⁺ homeostasis are associated with various age-related neuronal pathologies including PD (e.g., (Ge et al., 2022; Marambaud et al., 2009; Xu and Lipscombe, 2001)).

Indeed, there has been extensive interest in voltage-gated Ca_V1.3 Ca²⁺ channels as therapeutic targets in PD; specifically in SN DA neurons to forestall disease progression and in striatal SPNs to address therapy dysfunction including LID (for review (Caulfield et al., 2023a)). In SN DA neurons, activity of Ca_V1.3 channels and alterations in Ca²⁺ homeostasis are linked to cellular stress thought to underlie their vulnerability to neurodegeneration, supporting rationale for clinical trials aimed at reducing Ca²⁺ transients to slow disease progression (Caulfield et al., 2023a; Liss and Striessnig, 2019). In the striatum, alterations in Ca²⁺ homeostasis occurs secondary to dysfunctional Ca_V1.3 channel activity that occurs in the DA-depleted parkinsonian striatum (Day et al., 2006), resulting in SPN structural and circuit pathology. In the presence of dyskinogenic levodopa, this SPN striatopathology is exacerbated giving rise to the idea that Ca_V1.3 channel antagonism would ameliorate LID (Caulfield et al., 2023a; Liss and Striessnig, 2019).

In keeping with the outcomes of clinical trials in PD and other age-related neurodegenerative diseases (Lang and Espay, 2018; Leonard et al., 2020), the responses of aged rats to Ca_V1.3-targeted gene therapy were more variable than the responses of young rats. In both young and aged rats, although Cacna1d is downregulated to a greater degree in the CAV-shRNA rats compared to those receiving the control SCR-shRNA vector, the degree of downregulation is less consistent in aged striatum. Age-related dysfunction in Ca²⁺ homeostasis may render the aged parkinsonian striatum less responsive to therapeutic intervention than the young brain, in which viral vectors do not have to compete with age-associated challenges to homeostatic cellular processes.

Interestingly, we found no significant correlation between the level of Cacna1d downregulation and final LID severity scores in either young or aged rats (young: (Steece-Collier et al., 2019); aged: day 10, 18 mg/kg ratings; Spearman correlation, p=0.4368; data not shown). Cumulatively, these data provide evidence that even partial silencing of overactive Ca_V1.3 channels in the parkinsonian striatum is sufficient for functional efficacy and prevention of severe LID induction. It further suggests that factors other than Cacna1d silencing may underlie the differential functional efficacy in aged vs. young rats.

The current study importantly corroborates that antidyskinetic effects of Ca_V1.3 gene therapy persists in aged rats, though its protective benefits are reduced and more variable between rats. Specifically, on the final rating day, in response to 18 mg/kg levodopa, young male CAV-shRNA rats showed an average 91.8% reduction in total LID severity compared to the young SCR-shRNA rats. In aged animals, this percent reduction in total LID was 63.8%. While the efficacy is dampened in aged rats, this degree of LID abatement would have a profound clinical benefit if carried into individual with PD suffering from intractable LID.

1.4.2. Age-Related Differences in LID Severity and Motoric Activation

Our preclinical data demonstrate that in the control groups (i.e., SCR-shRNA vector), young parkinsonian rats compared to their aged counterparts develop more severe LID, which is especially notable for ‘total’ LID scores. This finding of more severe LID is analogous to dyskinesia development in individuals with young onset PD (Kumar et al., 2005). Given that LID are considered a form of aberrant motor learning, associated with synaptic plasticity and miswiring of basal ganglia neural circuits (Fieblinger and Cenci, 2015; Fieblinger et al., 2014), it is not surprising that less severe LID are associated with aged animals in which neural plasticity is diminished and/or aberrant (Sikora et al., 2021).

We previously reported enhanced motor activation of levodopa in young male parkinsonian rats receiving CAV-shRNA compared to SCR controls (Steece-Collier et al., 2019). These results largely align with the current study, where we observed a modest but significant increase in contralateral rotations in the CAV-shRNA cohort but not the SCR-shRNA cohort in response to low dose levodopa. However, again toward elucidating factors that are important for the timing of this gene therapy, in our previous report where aged parkinsonian rats received vector after LID induction (Caulfield et al., 2023b), motor activation following low-dose levodopa was lost in CAV-shRNA rats, but interestingly was observed in that SCR cohort. Maintaining motor response to DA agonists including levodopa is paramount for any antidyskinetic therapy. These discrepancies suggest a complicated relationship between levodopa-related modifications in the aged striatum and the timing of Ca_V1.3 silencing.

1.4.3. Age and Sex as Covariates

In the current study, we report the first data linking LID variability to estrus status in aging female rats, specifically in our control SCR-shRNA rats. As female F344 rats transition into acyclicity through stages of constant estrus and pseudopregnancy, they eventually display the lack of a detectable proestrus stage. We therefore used the presence or absence of proestrus to determine cycling status (Ajayi and Akhigbe, 2020; Cora et al., 2015; Cruz et al., 2017; Koebele and Bimonte-Nelson, 2016). Remarkably, in the SCR-shRNA control cohort, proestrus-negative female rats were resistant to developing severe LID.

Estrogen has been reported to regulate L-type Ca²⁺ channels in the brain (Vega-Vela et al., 2017), and the age-related decline of estrogen is associated with increased channel expression in the hippocampus (reviewed in (Foster, 2005)). In SPNs, membrane estrogen receptors are expressed on axon terminals, somas, and dendritic spines (Almey et al., 2014; Almey et al., 2012; Almey et al., 2015, 2022), and the electrophysiologic characteristics of SPNs have been reported to change with the estrous cycle (reviewed (Krentzel and Meitzen, 2018)). With respect to LID in individuals with PD, females tend to have a higher risk of developing LID and a shorter time to onset compared to males. In addition, females with PD have been reported to show a significant reduction in peak-dose dyskinesia post-menopause (Nicoletti et al., 2007), like our findings in proestrus-negative female rats. Further, ovariectomized female parkinsonian rats have intensified LID when treated with estradiol (Kolmancic et al., 2022), though there is contradictory evidence regarding the benefit of hormone replacement therapy in female PD patients (reviewed in (Meoni et al., 2020)).

Cumulatively, our findings suggest that gene therapy-mediated silencing of Ca_V1.3 is less effective for LID prevention in the aged female parkinsonian rats than in aged males. These findings are in keeping with those of the recent clinical trial, STEADY-PD III, in which pharmacological silencing of nigral Ca_V1.3 with the Ca_V1 antagonist isradipine was trialed as a neuroprotective therapy. It was initially reported that long-term isradipine treatment did not slow progression of early-stage PD (Grp, 2020); however, follow up analyses revealed a modest yet significant correlation between indicators of disease progression and plasma concentration of isradipine (Venuto et al., 2021). Compellingly, this correlation was driven by male trial participants; the correlation amongst female participant was, although not statistically significant, in the opposite direction. Taken together, both this clinical report and our preclinical findings suggest that there are sex-specific factors influencing the efficacy of Ca_V1.3-targeted therapies.

1.4.4. Translatability and Promise of Targeted Ca_V1.3 Silencing.

Despite intense interest in, and strong biological rationale for Ca_V1.3 channels as promising therapeutic targets for PD (Caulfield et al., 2023a; Kang et al., 2012; Liss and Striessnig, 2019), critical information gained from clinical (Maiti and Perlmutter, 2020; Surmeier et al., 2022) and preclinical (Schuster et al., 2009; Soderstrom et al., 2010) studies involving Ca_V1/1.3 pharmacological antagonists has demonstrated that dosing required for effective target engagement is currently not achievable without risk of impacting peripheral systems (e.g.: cardiovascular) and/or unintended brain regions (e.g., hippocampus, suprachiasmatic nucleus) resulting in negative off-target side-effects (Filippini et al., 2023; Lauffer et al., 2022; Marcantoni et al., 2007; McNally et al., 2021; Mesirca et al., 2016). Indeed, any systemically administered small molecule regardless of target efficacy risks peripheral and central off-target effects. In contrast, anatomically targeted gene therapy, such as our Ca_V1.3 RNA interference (RNAi)-based vector approach offers a potent and brain region specific alternative for targeted delivery to the parkinsonian brain for a variety of indications. Our work thus far with this gene therapy has been aimed principally at exploring the utility of striatal Ca_V1.3 channel as a first-in-class antidyskinetic target, with these studies being instrumental in validating target engagement and strong efficacy of our vectors (current study, (Caulfield et al., 2023c; Steece-Collier et al., 2019)). We posit that direct stereotaxic delivery of RNAi gene therapy is currently the only viable approach for achieving optimized and anatomically selective target engagement of Ca_V1.3 channels for addressing not only therapy mediated side-effects in PD, but also providing much needed neuroprotective therapy targeting SN DA neurons. Current trends in PD and LID research are migrating more towards gene therapies, with evidence of long-term efficacy and safety (Bulaklak and Gersbach, 2020; Buttery and Barker, 2020; Chu and Kordower, 2023; Wirth and Yla-Herttuala, 2006).

1.5. CONCLUSIONS

The current study provides important validation in rats of advancing age that silencing of dysfunctional Ca_V1.3 channels in the parkinsonian striatum provides significant protection against the induction and escalation of LID. It also highlights challenges inherent to therapeutic intervention in aged-related diseases. Our systematic investigative series of studies have been able to elucidate that advanced age, sex and levodopa priming as covariates introduces increased variability in responses of this RNAi therapeutic that has potential application to clinical PD. Our studies reinforce the importance of utilizing clinically relevant populations in animal models of PD and highlight challenges intrinsic to the study of aging animals. As recently reviewed (Kohn et al., 2023), there continues to be advances in the science of gene therapy as a means of providing enduring treatments for increasing number of diseases including PD. As our Ca_V1.3 gene silencing therapeutic continues to show promise for clinical development, further studies are warranted to identify the mechanisms which underlie age-related and sex-specific variability in response to Ca_V1.3-targeted therapeutics.

HIGHLIGHTS.

• Striatal AAV-Ca_V1.3-shRNA prevents induction of severe dyskinesia in aged rats

• Aged rats display greater variability and diminished therapeutic benefit

• Age-related diminution of therapeutic benefit is most notable in female rats

• Aged female rats in reproductive senescence are refractory to developing severe LID

Data Availability

Data will be made available for no less than 3 years after publication upon reasonable request.