Combining fibril‐induced alpha‐synuclein aggregation and 6‐hydroxydopamine in a mouse model of Parkinson's disease and the effect of cerebral dopamine neurotrophic factor on the induced neurodegeneration

The existent pre‐clinical models of Parkinson's disease do not simultaneously recapitulate severe degeneration of dopamine neurons and the occurrence of alpha‐synuclein (aSyn) aggregation in one study system. In this study, we injected aSyn pre‐formed fibrils (PFF) and 6‐hydroxydopamine (6‐OHDA) unilaterally into the striatum of C57BL/6 wild‐type male mice at an interval of 2 weeks to induce aggregation of aSyn protein and trigger the loss of dopamine neurons simultaneously in one model and studied the behavioural effects of the combination in these mice. 6‐OHDA was tested at three different doses, and 2 μg of 6‐OHDA combined with PFF‐induced aSyn aggregation was found to produce the most optimal disease phenotype. At 14 weeks timepoint, mice injected with a combination of PFF and 6‐OHDA sustained significant damage to the nigrostriatal pathway and exhibited aSyn‐positive aggregation. Our data suggest that the neurons that formed large aSyn aggregates were particularly vulnerable to 6‐OHDA‐induced degeneration. We also demonstrate the manifestation of a relatively aggressive pathology in 2‐ to 4‐month‐old mice, as compared to younger 7‐ to 9‐week‐old ones. Furthermore, cerebral dopamine neurotrophic factor (CDNF) administered intrastriatally rescued dopamine neurons and motor behaviour of the animals to some extent from 6‐OHDA toxicity. However, no such effect could be seen in the novel 6‐OHDA + PFFs combination model. For the first time, we demonstrate the combined effect of PFF and 6‐OHDA simultaneously in one model. We further discuss the scope for further optimizing this combination model to develop it as a promising pre‐clinical platform for drug screening and development.

for further optimizing this combination model to develop it as a promising pre-clinical platform for drug screening and development.

| INTRODUCTION
Parkinson's disease (PD) is the second most common neurodegenerative disease after Alzheimer's disease, with an increasing prevalence primarily attributed to today's globally ageing population.Pathologically, the degeneration of dopamine (DA) neurons in the nigrostriatal pathway leads to loss of DA in the striatum, resulting in motor deficits in PD patients.In these patients, the onset of motor symptoms occurs at an estimated loss of about 30-50% of the dopaminergic neurons in the substantia nigra (SN) and 50-60% of their axon terminals in the striatum (reviewed in Cheng et al., 2010).The degeneration of neurons along the nigrostriatal tract seems to follow a dying-back mechanism where the loss of axons in the striatum is followed by the loss of the cell bodies in the SN (Cheng et al., 2010).Over the years, this dying-back theory has formed the basis of many therapeutic pursuits targeting the restoration of the DA content and signalling in the striatum by preserving the intact DA axons and regenerating the impaired axon terminals.
In 1912, Fritz Jacob Heinrich Lewy described peculiar intraneuronal inclusion bodies in the brain of PD patients, which we now know as Lewy bodies (LBs) (Holdorff et al., 2013).LBs have been found to comprise of misfolded and aggregated alpha-synuclein (aSyn) protein (Spillantini et al., 1997) along with other vesicles and damaged membranous organelles (Shahmoradian et al., 2019).Pathological protein aggregation has been suggested as the core mechanism of several synucleinopathy-and tauopathyrelated diseases (Lashuel, 2021;Vasili et al., 2019).In its physiological form in neurons, aSyn is mostly found in the presynaptic terminals where it is involved in vesicle trafficking, fusion of synaptic vesicles to the presynaptic membrane and neurotransmitter release (reviewed by Sulzer & Edwards, 2019).Aggregation of aSyn protein has been found to disrupt several biological functions along the length of a neuron (Bernal-Conde et al., 2019), playing a significant role in PD pathogenesis (Hijaz & Volpicelli-Daley, 2020).While there has been intense research in the field, the mechanisms causing PD pathology remain obscure.This complex pathology calls for more intricate models and methods to investigate the multifactorial aspects of PD.
There is an abundance of research models aiming to study aSyn pathology in PD.The majority of these models utilize either transgenic rodent models, viral overexpression of aSyn, or aSyn fibril-induced misfolding and aggregation (reviewed by Gomez-Benito et al., 2020).First developed by Virginia M.-Y. Lee and her team, recombinant aSyn pre-formed fibrils (PFFs) neuronal seeding models have gained immense attention over the last 10 years.The model functions by PFF-induced recruitment of the endogenous aSyn to form aggregates, forming phosphorylated, ubiquitinated and proteinase K-resistant LBlike inclusions (Luk et al., 2009).Since then, PFFs have been used to model PD by intracerebral injections of PFFsalone, or in combination with AAV-mediated overexpression of aSyn (Hoban et al., 2020;Thakur et al., 2017) or by injecting PFFs into different peripheral sites such as muscles of the gut of wild-type mice (Kim et al., 2019) or hindlimbs of M83 transgenic mice (Sacino et al., 2014).
The currently available pre-clinical models for PD research are limited in mirroring PD pathology and symptoms found in patients.The aSyn aggregation models exhibit neurodegeneration that often proceeds relatively slowly, not providing a sufficiently robust model for drug development.On the other hand, the classic neurotoxin models of PD usually do not produce any aSyn pathology, and the neuron death occurs relatively fast, thus not representing the progressive nature of PD pathology well.Drawbacks in the current PD models not only hinder the discovery and development of pharmacological therapies for PD but also limit revealing the mechanisms behind neurodegeneration.
Cerebral dopamine neurotrophic factor (CDNF) is a protein with neurotrophic factor properties that has reached its safety and tolerability endpoint in its first phase I/II clinical trials in PD patients (Huttunen et al., 2023;ClinicalTrials.gov identifier: NCT03295786, NCT03775538).CDNF is a 161 amino acid long protein, endogenously present in the brain and various nonneuronal tissues such as skeletal muscle and testis of both humans and mice (Lindholm et al., 2007).In the pre-clinical studies, CDNF has shown restoration of DA neurons in the SNpc and the striatal axon terminals, along with locomotor recovery in the toxin-injected rodent models of PD (Airavaara et al., 2012;Back et al., 2013;Lindholm et al., 2007;Voutilainen et al., 2011Voutilainen et al., , 2017)).Additionally, CDNF promoted DAT binding activity in toxin-lesioned marmoset monkeys (Garea-Rodriguez et al., 2016) and was recently reported to interact with aSyn and induce behavioural recovery in a fibril-induced mouse model of PD (Albert et al., 2021).
In this study, we have utilized intrastriatal injections of aSyn PFFs in combination with a classic neurotoxin 6-hydroxydopamine (6-OHDA) to merge two different pathological mechanisms of PD-the LB-like inclusion formation (Volpicelli-Daley et al., 2011) and mitochondrial stress, respectively.We aimed to obtain a PD model that would present both the LB-like pathology and a progressive neurodegeneration in a sufficient time scale and tested it as a platform to evaluate the neurorestorative effects of CDNF.

| MATERIALS AND METHODS
This study was conducted in two experimental parts: Part 1: Experiments were conducted to optimize the dose of 6-OHDA to establish the novel combination model of PD.Part 2: Experiments were conducted to analyse the amount of DA neuron survival and initiation of phosphorylated aSyn (paSyn) aggregation at early timepoints of the combination model of PD, and to test the effect of CDNF in this model of PD over an extended period.

| Animals
Two to four months old wild-type C57BL/6 male mice were used for conducting Part 1 of the study.For Part 2, 7-9 weeks old wild-type C57BL/6 male mice were used due to issues with availability of animals at the time (covid restrictions).All the animals were allowed to acclimatize to the new housing for 3 weeks before being handled by the experimenters.The animals were housed in groups of four until the first set of injections and were moved to separate cages for the remaining duration of the experiment.The cages were maintained in a temperature-and humidity-controlled facility, at 12 h light/dark cycle and ad libitum access to food and water.All the surgeries and behavioural tests were carried out at the Laboratory Animal Center core facility at the Viikki campus of University of Helsinki, Finland.

| Unilateral intrastriatal injections of PFFs, 6-OHDA and CDNF in animals
All stereotaxic injections were performed unilaterally to the right striatum of the animals under isoflurane anaesthesia (3% for induction and 2.5% for maintenance), at a 10 angle and relative to the bregma and dura (Table 1).The aSyn PFFs were produced by our collaborator Prof. Kelvin C. Luk (University of Pennsylvania) (protocol in Volpicelli-Daley et al., 2014).On the day of the injection, the PFFs were diluted in phosphate-buffered solution (PBS) and sonicated in a Bioruptor water bath sonicator (Diagenode, Belgium) at high power mode settings for 20 cycles of 30 s on and 30 s off.The animals received a single injection of 5 μg mouse aSyn PFFs in 2.5 μl PBS at the flow rate of 0.2 μl/min (Albert et al., 2021;Luk et al., 2012).The needle was kept in place for 5 min after the completion of the injection to avoid any backflow of the solutions.
Two weeks later, animals were injected with 6-OHDA (Sigma #H4381; calculated as base-free) in 0.02% ascorbic acid solution (Sigma-Aldrich #A92902; diluted in 0.9% NaCl).The toxin was injected in two sites, amounting to a total of either 1, 2 or 4 μg per animal.
Two weeks after 6-OHDA injections, a total of 10 μg in 4 μl CDNF (human recombinant CDNF, 9.6 μg/μl, Biovian, Turku) was administered unilaterally to these animals to the same coordinates as 6-OHDA.The flow rate for both 6-OHDA and CDNF injections was 0.5 μl/min.The animals in the control groups received artificial cerebrospinal fluid (aCSF), ascorbic acid and PBS as controls for PFF, 6-OHDA and CDNF injections, respectively.
The animals were administered 5 mg/kg carprofen (Rimadyl, Pfizer) subcutaneously, as a painkiller after the surgery.All the details for the stereotaxic injections have been summarized in Table 1.

| Behavioural analysis
The weight of the animals was recorded every week until the day of perfusion.Cylinder and rotarod tests were conducted as baseline behavioural tests (Week À2) before the PFF injections.The behavioural tests were performed every 2 weeks until the day of perfusion.In the graphs, the timepoints for behavioural tests are expressed relative to the 6-OHDA injections: Week À2 (just before PFF injection), Week 0 (just before 6-OHDA injection/2 weeks post-PFF injections), Week 2 (just before CDNF injections/4 weeks post-PFF injections) and so on.

| Cylinder test
The cylinder test was performed to evaluate the usage of the ipsilateral paw of the lesioned and CDNF-treated animals in comparison to their contralateral paw usage.The mice were placed in a plexiglass cylinder (dimensions: 11 cm diameter Â 15 cm height), raised from a transparent platform.A video camera was attached to the bottom of the cylinder and the behaviour of the animal placed inside the cylinder was recorded for the first 5 min.The videos were quantified frame by frame for the number of paw touches with their ipsilateral, contralateral and both paws while rearing to the cylinder wall.Animals that did not rear at all were excluded from the cylinder test analysis.The number of excluded animals for each group: V The test was conducted under red light conditions, and the analysis was done blinded to the injection groups.

| Amphetamine-induced rotations
Mice were injected intraperitoneally with 10 mg/kg D-amphetamine sulphate (Tocris Bioscience, 2813), dissolved in saline and placed in a transparent cylinder (dimensions: 11 cm diameter Â 15 cm height).The 360 rotations of these mice were recorded for intervals of 5 min over a period of 90 min, in both clockwise and counterclockwise directions using RotoRat software (Med Associates Inc.).The data obtained were reported as the number of ipsilateral turns minus the number of contralateral turns made by the mice over the studied interval of time.

| Rotarod
Before performing the baseline and consecutive tests, the mice were habituated to the rotarod (Ugo Basil, Italy).The balance of the mice on the rotarod was tested at 4-40 rpm (revolutions per minute) for a maximum of 4 or 5 min.

| Coat hanger
Coat hanger test was performed to test the grip strength of the mice.The mice were made to hang from the centre of the coat hanger from their fore paws and were scored from 0 to 5 based on their ability to stay on the coat hanger and their agility.The test was conducted for a maximum of 60 s.The scoring system for the test is as follows: 0 = mouse falls off in less than 10 s. 1 = mouse stays on for more than 10 s and hangs there. 2 = mouse puts one hind limb onto the hanger.3 = mouse puts two hind limbs onto the hanger.4 = mouse wraps its tail around the hanger (and both hind limbs).5 = mouse climbs to the top of the hanger.

| Open field test
Mice were placed in the centre of an open field chamber (30 Â 30 Â 20 cm, Med Associates), where their movement within the chamber was recorded over a period of 30 min.The results were analysed as the distance travelled by the mice in the central zone and in the periphery, closer to the walls.The data were plotted for the total trial time of 30 min as well as the first 5 min of the trial.

| Immunohistochemistry
The mice were anaesthetized with pentobarbital (Orion Pharma, i.p.; 90 mg/kg, diluted in saline) and perfused using 4% paraformaldehyde (PFA; dissolved in PBS).The brains were isolated and post-fixed overnight in 4% PFA.The brains were paraffinized, and the paraffin-embedded brains were sliced into 10 μm thick coronal sections using a microtome (Leica RM2235, Leica Biosystems, Germany), collected on slides and stored at +4 C until immunostaining (refer to Table 2 for list of reagents and antibodies).The sections were immunostained in multiple batches.Inevitably, the background staining intensities vary slightly between the batches, so brain sections from all study groups were included in each staining batch to avoid any possible bias in analysis.

| TH and DAT stainings
Slides containing three to four sections each were incubated in a series of xylene and ethanol solutions to carry out deparaffinization and rehydration of the sections.Antigen retrieval was performed by immersing and heating the sections in 10 mM sodium citrate buffer (Sigma-Aldrich, 6132-04-3) containing 0.05% tween-20-(Sigma-Aldrich, P1379) in a microwave at 600 W for 5 min and then at 400 W for 6 min.The sections were cooled down and incubated in 0.3% H 2 O 2 (Sigma-Aldrich, 18304), diluted in TBS, for 30 min and rinsed with TBS-T (TBS with 0.05% tween-20).The sections were then blocked with 1.5% NGS (Vector Laboratories, S1000) in TBS-T for 30 min at room temperature and incubated with either rabbit anti-TH antibody (1:1,000, Sigma-Aldrich, AB152) or rat anti-DAT (1:400, Santa Cruz, sc-32258) in blocking solution (1.5% NGS in TBS-T) overnight at +4 C in a humidified chamber.The next day, the sections were incubated with either biotinylated-anti-rabbit IgG antibody (1:400, Vector Laboratories, BA-1000) for TH stainings or biotinylated-anti-rat IgG antibody (1:400, Vector Laboratories, BA-4000) for DAT stainings, for 1 h at room temperature.The sections were rinsed again in TBS-T and avidin/biotin complex solution (Vector Laboratories, PK-4000), pre-incubated for 20-30 min, was added to the sections for another 30 min.At the end of the incubation, the sections were rinsed in TBS-T, and the signal was visualized using 3,3 0diaminobenzidine (DAB) (Vector Laboratories, SK-4100).

| TH-and paSyn-double staining
The sections were deparaffinized and stained with the first primary rabbit anti-paSyn antibody [EP1536Y] (1:10,000, Abcam, ab51253) and biotinylated goat anti-rabbit IgG antibody (1:400, Vector Laboratories, PK-6101).The paSyn staining was visualized using DAB substrate, as explained earlier.The DAB reaction was stopped by rinsing with TBS, and then, sections were incubated in H 2 O 2 , followed by 15 min incubation in avidin blocking solution and then 15 min in the biotin blocking solution (Vector Laboratories, SP-2001).The sections were then incubated with the second primary mouse anti-TH antibody (1:1,000, Sigma-Aldrich, MAB318) overnight at +4 C in a humidified chamber.The next day, the sections were incubated with biotinylated-anti-mouse IgG antibody (1:400, Vector Laboratories, BA-9200) for 30 min, followed by incubation in avidin/biotin complex solution (Vector Laboratories, PK-4000) for another 30 min.The TH signal was visualized using HistoGreen solution (POD substrate kit, Linaris, E109).

| TH and DAT fibre density measurement
For each mouse, striatal coronal sections between A/P + 0.62 mm (relative to bregma; Paxinos, 2001) and A/P + 1.10 mm were used for the analysis of fibre density.The sections for TH (n = 4 slices per animal) and DAT stainings (n = 4 slices per animal) were imaged with 20X magnification using 3DHistec scanner (Pannoramic P250, Budapest, Hungary).The fibre optical density from the striatum was quantified using Fiji ImageJ 1.53 (Schneider et al., 2012).The optical density from the corpus callosum was used as a background and was subtracted from the striatal densities.The data are represented as the percentage of the contralateral (control) hemisphere.

| TH, DAT counts in SN and pS129aSyn count
For each mouse, nigral coronal sections between A/P -2.92 mm (relative to bregma; Paxinos, 2001) and A/P À3.64 mm were used for the analysis of TH, DAT and paSyn count.To stain for DAT (n = 4) and dual staining of TH and paSyn (n = 4), the sections were imaged at 20X magnification and extended focus, using 3DHistec scanner and uploaded to the Aiforia image platform (Aiforia Technologies Plc.Helsinki, Finland).
The TH-and DAT-positive cells, along with the pS129aSyn-positive aggregates, were counted manually using the cell counting tool in Aiforia.The 'ventral cortex' areas were analysed from the same sections that were used for SNpc analysis, and the region ventral to the perirhinal area of the cortex was considered as 'ventral cortex' for quantification for paSyn-positive aggregates.The TH-and DAT-positive cell counts are represented as a percentage of the contralateral (control) hemisphere, and the data for paSyn-positive aggregates are represented as absolute count in the ipsilateral hemisphere.

| Statistical analysis
The statistical comparison was performed with GraphPad Prism 9.2.0.One-way analysis of variance (ANOVA) was used to analyse the statistical difference between three or more groups based on one factor, and two-way ANOVA was used in case of two factors under study.A significant ANOVA was followed by Tukey's multiple comparisons test to identify pairwise differences between groups.Šid ak's multiple comparisons test was used to compare the ipsilateral and contralateral paws in cylinder tests, TH and DAT quantifications in histology samples, and the different brain regions for the paSyn-positive aggregate counts.The statistical difference was considered significant when the p value was less than .05.The data have been expressed as mean ± SEM.

| RESULTS
3.1 | 6-OHDA caused nigrostriatal DA neuron degeneration and motor impairment in a dose-dependent manner, whereas PFFs initiated robust aSyn aggregation pathology without neuron loss 14 weeks post-injection

| DA neuron survival
In order to optimize our model to induce progressive and adequate DA neuron loss and paSyn-positive aggregate formation, we tested the combination of aSyn PFFs (intrastriatal injection of 5 μg) and three different doses of 6-OHDA 1, 2 and 4 μg (Figure 1a) in 2-4 months old mice.The mortality rate for mice receiving 6-OHDA 4 μg injections was extremely high within the first 2 weeks of the surgery; we started with n = 7 mice for 6-OHDA 4 μg and n = 9 mice for PFF + 6-OHDA 4 μg groups at week À2 and were left with n = 3 and n = 5 mice from each of the groups at week 12 (Figure S1m).We quantified the amount of TH-and DAT-positive DA fibres in the striatum and TH-and DAT-positive DA neurons in the SNpc at the timepoint of 14 weeks after PFF injection (and 12 weeks after 6-OHDA injections) in the remaining mice.Although in PFF studies, neurodegeneration is observed at much longer timescale, we hypothesized that in the combination model, the neurodegeneration would occur already at this earlier timepoint.PFFs and 6-OHDA, alone and in combination led to a significant reduction in the density of DA terminals in the striatum (Figure 1b-f; n = 3-8 per group) and the number of DA cell bodies in the SNpc (Figure 1g-k; n = 3-8 per group), when compared to the vehicle-treated animals.The highest dose of 6-OHDA (4 μg) led to an approximately 75% reduction in DA neurons, and in combination with PFFs, almost 85% DA neuron loss in comparison to the vehicleinjected group.Since the injection with the lowest dose of 6-OHDA (1 μg) did not significantly affect the DA neuron survival and caused only an approximate 3% THpositive fibre loss and 15% TH-positive cell loss along the nigrostriatal pathway, we decided to report these data as supporting information (Figure S3a-f).Based on TH and DAT stainings, PFFs, 6-OHDA or their combined administration showed loss of DA neuron fibres in the striatum.Interestingly, in the SNpc, while TH and DAT stainings showed reduced DA neuron count after 6-OHDA or the combination model, for the only PFF-injected mice, the TH-positive cell count was significantly reduced, but the DAT-positive cell count remained unchanged (Figure 1i, n = 5 per group, TH vs. DAT for P + V p = .0016).This suggests that 14 weeks after the PFF injection, without 6-OHDA, the DA cell bodies in the SNpc were not lost; instead, there was possibly a downregulation of TH in the SNpc DA neurons.

| paSyn aggregates
The brain sections were stained against aSyn phosphorylated at S129 (paSyn) to quantify the hyperphosphorylated aSyn aggregates in these mice (Figure 2a-c).Aggregation pathology was not detected in the vehicle/ non-PFF-injected animals (Figure 2).There were a large number of aggregates in the SNpc region in the mice injected with only PFFs, whereas the mice injected with PFFs in combination with either 2 or 4 μg 6-OHDA showed significantly reduced aggregates as compared to PFF-only mice-approximately 86% reduction in PFF + 6-OHDA 2 μg animals and about 93% reduction in PFF + 6-OHDA 4 μg animals (Figure 2a; n = 5-8 per group; p < .0001).This suggested that intrastriatal 6-OHDA injection 2 weeks after PFF injection either caused a selective loss of DA neurons that had aSyn aggregation pathology or the toxin affected the large paSyn aggregates.To clarify the role of 6-OHDA, which is selectively taken into neurons by DA and norepinephrine transporters, as the culprit for the reduction of aSyn aggregation in the combination model, we analysed the ventral cortical regions near SNpc in these animals (Figures 2b and 2a,b).This area is known to also exhibit the aSyn aggregates after intrastriatal PFF injections (Luk et al., 2012), but the neurons in this region should not be affected much by the intrastriatal 6-OHDA injection.As expected, the level of aSyn aggregation in the ventral cortex was not affected by 6-OHDA (Figure 2b; n = 5-8 per group, ns).Injection with the lowest dose of 6-OHDA (1 μg) did not have any effect on the paSyn-positive aggregate count in any of the analysed regions of the brain tissue (Figure S3g,h).

| Motor behaviour
The motor behaviour of these mice injected with PFFonly, 6-OHDA-only, combination of both and their control-injected groups was monitored using various tests throughout the duration of the experiment.As expected, when administered with d-amphetamine, the mice that received unilateral intrastriatal injection of 6-OHDA 4 μg rotated significantly more than the mice with 2 μg dose throughout the experiment (Figure 3a, n = 3-8 per group).The difference in ipsilateral rotations was statistically significant at 2 weeks post-6-OHDA injections (Figure 3a; V + V vs V + O4 p = .0003,V + O2 vs V + O4 p = .0264).Clearly, the 6-OHDA dose of 4 μg was too high for the animals and therefore also induced apparent motor deficits in these mice.Similarly, the mice receiving 2 μg 6-OHDA rotated more than the 1 μg 6-OHDA group (Figure S4a), suggesting a clear dosedependent response of these mice to the increasing concentrations of 6-OHDA.During the rotarod test, at Week 8, the mice injected with PFF + 6-OHDA 4 μg dropped off the rotarod significantly sooner than the PFF + 6-OHDA 2 μg group (Figure 3b, n = 3-8 per group; P + O2 vs. P + O4 p = .0195).In the cylinder test, out of the total number of wall touches (Figure 3c), mice with 6-OHDA 4 μg, with and without PFFs, reared significantly more with their ipsilateral paw than with their contralateral paw at Weeks 2, 8 and 12.As for the total number of only-first touches (Figure 3d), in addition to the 6-OHDA 4 μg group at all tested timepoints, the mice with 6-OHDA 2 μg only also used their ipsilateral paw significantly more than the contralateral paw at Week 2. Interestingly, by Week 12, the PFF + 6-OHDA 2 μg mice also developed a paw usage bias and used more of their ipsilateral paw, suggesting a developing pathology.
We also performed coat hanger (Figure S1a-c) and open field (Figure S1d-k) tests with these mice, where they showed no significant difference in their motor behaviours, except for reduced vertical counts for PFF + 6-OHDA 4 μg mice, compared with PFF + 6-OHDA 2 μg mice in the first 5 min of open field test at week 12 (p = .0298).The weights of mice in different injection groups did not change significantly compared to the vehicle-treated group (Figures S1l and S4e).The mice that received the lowest dose of 6-OHDA (1 μg) had no motor deficits compared to the vehicle-injected mice (Figure S4a-d).
On comparing the immunohistochemistry and behavioural data obtained from mice injected with either 1, 2 or 4 μg 6-OHDA, 6-OHDA at 2 μg concentration led to a significant nigrostriatal DA degeneration that was mirrored as deficits in the motor function.The highest dose caused an overly extensive lesion in the DA system, while the lowest dose was insufficient to induce robust pathology.Importantly, when 2 μg of 6-OHDA was administered in the presence of PFF-induced aSyn aggregation, this combination model exhibited both the degeneration of the nigrostriatal DA tract and the paSyn-positive aggregate pathology, although a reduced number of aggregates was visible in SNpc due to 6-OHDA-induced toxicity.
3.2 | Some PFF-seeded paSyn aggregates were visible already 4 weeks after seeding and 6-OHDA induced robust neurodegeneration 2 weeks after its injection Before testing our candidate therapeutic agent in this combination model, we wanted to evaluate the extent of neuron degeneration and aggregation caused by aSyn PFF, 6-OHDA and their combination 2 weeks after each of the injections.For this study, 7-9 weeks old mice were perfused in two sets-one at 2 weeks after PFF injections and the rest 2 weeks after 6-OHDA injections (4 weeks after PFF injections) (Figure 4a).

| DA neuron survival
To confirm the extent of the lesion at these two timepoints, the striatal and SN brain tissues from these mice were stained for TH and DAT.At 2 or 4 weeks after PFF injection (without or with vehicle injections, respectively), neither the striatal fibre density (Figure 4b,c), nor the neuron counts in the SNpc (Figure 4d,e) were affected.On the contrary, 6-OHDA 2 μg caused between 70% and 80% loss of striatal DA fibres, both with and without PFFs (Figure 4b,c; n = 3-4 per group, p < .0001).Similarly, 6-OHDA, irrespective of the presence of PFFs, caused a significant reduction of about 45% TH-and 55-65% DAT-positive DA cell counts in the SNpc region compared to the vehicle injection (Figure 4d,e; n = 3-4 per group).

| paSyn aggregate pathology
The paSyn aggregates in SNpc and ventral cortex were analysed at both the timepoints of Weeks 0 and 2. At Experiment design for combination model with PFF and different doses of 6-OHDA.The mice were injected with sonicated PFFs (at Week À2) into a single site, followed by either 2 or 4 μg 6-OHDA into two sites, 2 weeks after PFFs (Week 0).All the injections were performed intrastriatally to the right hemisphere.The baseline behavioural tests were conducted just before PFF injections (shown as Week À2 in graphs), followed by behavioural tests just before 6-OHDA injections (shown as Week 0 in graphs), and continued every 2 weeks until perfusion at Week 12. (b-k) Impact on the DA neurons along the nigrostriatal pathway of mice injected with PFF and different doses of 6-OHDA.Tyrosine hydroxylase (TH)-and dopamine transporter (DAT)-positive (b-f) fibres in the striatum and (g-k) cells in the SNpc are quantified as percentages of the control (uninjected) sides.(d, i) The graphs present a comparison between the TH-and DAT-positive signals for both striatum and SNpc.The values are presented as mean ± SEM.Ordinary one-ANOVA and two-way ANOVA, followed by Tukey's or Šid ak's multiple comparisons test, were used for statistical analysis (statistics summarized in Table S1).n = 3-8 mice per group (2-4 months old), *p < .05,** p < .01,*** p < .001and **** p < .0001.The '#' symbol represents the statistical difference between the specified group and the vehicle-treated group.In graphs, V = Vehicle, P = PFF, O2 = 6-OHDA 2 μg and O4 = 6-OHDA 4 μg.(e, f, j, k) Representative images for TH-and DAT-positive stainings in striatum and SN.Scale bars are (e, f) 500 μm and (j, k) 200 μm, 50 μm insets.
2 weeks after the PFF injection, very few clear aggregates of paSyn were found in these regions (Figure 4f-h).Interestingly, we could see diffused paSyn-positive stainings in PFF-injected mice, compared to the vehicle-injected sections, suggestive of early events of aggregate formation.At 4 weeks after fibril injection, the aggregates in both 6-OHDA-lesioned and non-lesioned mice increased in number, though this increase was not statistically significant when compared to the aggregate number in PFFinjected mice from week 0 (Figure 4f,g).

| Motor behaviour
The mice injected with 6-OHDA, with and without PFFs, exhibited significantly increased amphetamine-induced ipsilateral rotations compared to the non-6-OHDA-lesioned mice at 2 weeks after 6-OHDA injection (Figure 5a; n = 3-4 per group), suggesting a successful lesion already at this timepoint of the experiment.There was no difference in the motor imbalance of these mice on rotarod throughout the experiment (Figure 5b; n = 3-4 per group).There was also no difference in the weights of these mice between the different injection groups at the studied timepoints (Figure S5).
Overall, while at Week 0, the PFF-induced pathology did not damage the nigrostriatal pathway and only presented diffused paSyn-positive signal, at Week 2, the The graphs represent the ipsilateral, contralateral and both paw touches to the cylinder wall as a percentage of (c) the total wall touches and of (d) the total first paw touches to the wall for 5 min of test duration.The values are presented as mean ± SEM.Two-way ANOVA, followed by Tukey's or Šid ak's multiple comparisons test, was used for statistical analysis (statistics summarized in Table S1).n = 3-8 mice per group (2-4 months old), *p < .05,** p < .01,*** p < .001and **** p < .0001.The '#' symbol is used to represent the statistical difference between the specified group and the vehicle-treated group.The green and blue dots on the graphs (x-axis; time) represent the approximate timepoints of PFF and 6-OHDA injections, respectively, and V = vehicle, P = PFF, O2 = 6-OHDA 2 μg and O4 = 6-OHDA 4 μg.
PFFs induced a higher amount of aggregation and 6-OHDA caused significant damage to the DA cell bodies and their fibres in the striatum.The resulting combination model showed promise as a platform for testing the restorative effects of CDNF on the PFF-induced aggregation and 6-OHDA-induced neurodegeneration when administered at Week 2 (2 weeks after adding 6-OHDA to the model) in this experimental paradigm.

| Testing the possible neurorestorative effect of CDNF in the combination model
Based on the previously published evidence on the neuroprotective and restorative effects of CDNF and its interaction with aSyn, we tested CDNF in our novel combination model of PD.Four weeks after PFF injection Experiment design for analysis of neuron survival and paSyn aggregates at earlier timepoints.The mice were injected with sonicated PFFs (at Week À2) into a single site, followed by 6-OHDA 2 μg into two sites, 2 weeks after PFFs (Week 0).All the injections were performed intrastriatally to the right hemisphere.Some mice were perfused 2 weeks after PFF injection (at Week 0), and others were perfused 2 weeks after 6-OHDA injections (at Week 2, 4 weeks with the PFFs).Tyrosine hydroxylase (TH)-and dopamine transporter (DAT)-positive (b, c) fibres in the striatum and (d, e) cells in the SNpc are quantified as percentages of the control (uninjected) sides.paSynpositive aggregates in (f) SNpc and (g) ventral cortex from mice injected with PFFs and PFF + 6-OHDA.The values are presented as mean ± SEM.Ordinary one-and two-way ANOVA, followed by Tukey's or Šid ak's multiple comparisons test, were used for statistical analysis (statistics summarized in Table S1).n = 3-5 mice per group ( 7 Ordinary one-and two-way ANOVA, followed by Tukey's multiple comparisons test, were used for statistical analysis (statistics summarized in Table S1).n = 3-5 mice per group (7-9 weeks old), *p < .05 and ** p < .01.The green and blue dots on the graphs (x-axis; time) represent the approximate timepoints of PFF and 6-OHDA injections.respectively, and V = vehicle, P = PFF and O2 = 6-OHDA 2 μg.
and 2 weeks after 6-OHDA injections, we injected a total of 10 μg CDNF to the same sites as 6-OHDA (Figure 6a).

| Effect on DA neuron survival
After 12 weeks with 6-OHDA and 14 weeks with PFFs, a significant loss of striatal DA neuron fibres could be seen in the 6-OHDA only-injected and in 6-OHDA + PFFsinjected animals (Figure 6b,c; n = 10-17 per group).CDNF treatment of the 6-OHDA-only-injected animals increased the TH-positive fibre density to the same level as the vehicle treatment group (Figure 6b; V + O2 + C vs. V + V + V p = .3963,ns) and showed a trend towards a rescue of the DAT-positive fibre density as well, although the latter was not statistically significant (Figure 6c; V + O2 + C vs. V + O2 + V p = .1085,ns).Intriguingly, the administration of CDNF did not affect the density of DA neuron fibres in the combination model.
In scoring of the DA neuron counts in SNpc, a high level of variation was found in all treatment groups (Figure 6d,e).In 6-OHDA-only-injected animals, the TH staining revealed reduced neuronal counts regardless of the CDNF treatment, whereas the DAT staining suggested a trend towards CDNF reducing the loss of neurons (Figure 6e; V + O2 + V vs. V + V + V p = .0013;V + O2 + C vs. V + V + V p = .2957,ns).However, in the combination model, based on the TH and DAT stainings, CDNF treatment could not rescue the neuron count.Instead, there was an even more significant reduction in the neuronal counts in the animals that received CDNF after the combination of PFFs and 6-OHDA.The animals that were injected with PFFs-only showed neither a loss of DA neurons, nor their fibres.

| Effect on paSyn aggregate pathology
The PFF-induced aggregates were quantified at the experimental endpoint of 14 weeks from the day of PFF injection.Aggregation pathology was not detected in any of the analysed regions for the vehicle-injected (non-PFF) animals (data not shown).In SNpc, the PFF-only animals had the highest number of paSyn-positive aggregates (Figure 6f; one-way ANOVA; F[3, 55] = 2.655, p = .0574).Although the difference was not statistically significant, in part due to the high variability of data, the combination model had considerably fewer aggregates than the PFF-only group (n = 14-15 per group, p = .1132,ns).CDNF administration did not have a statistically significant effect on the aggregate count in these PFF-injected animals.

| Effect on motor behaviour
At Week 2 (before CDNF administration), the mice with 6-OHDA had significantly more amphetamine-induced rotations than the vehicle-injected mice (Figure 7a; n = 10-17 per group) or PFF-injected mice (V + O2 + C vs. P + V + V p = .0372).For the combination model, the rotations on average were not statistically different from the vehicle-or PFF-injected groups, but many of those mice did show high numbers of rotations.However, unexpectedly, the lesioned animals, including the ones that exhibited a high number of amphetamineinduced rotations earlier, went on to rotate lesser over the course of the experiment, ending with almost as less rotations as the vehicle-injected animals at Week 12. CDNF treatment had no effect on these outcomes.However, there was a trend of increased latency to fall for the 6-OHDA-lesioned animals injected with CDNF (Figure 7b; n = 10-17 per group), compared to their control groups.In cylinder test, the 6-OHDA-only animals developed a statistically significant bias for using the ipsilateral paw more than the contralateral paw, and treating them with CDNF diminished this bias (Figure 7c; Week 12: V + O2 + V n = 10, ipsi vs. contra p = .0088and V + O2 + C n = 15, ipsi vs. contra p = .9991,ns).Also, the animals in the combination model developed increased usage of the ipsilateral paw, exhibiting statistically significant difference bias at week 12 (Figure 7c; P + O2 + V: n = 11, ipsi vs. contra p = .0012).Interestingly, CDNF injections could not rescue the motor imbalance for the combination model mice (Figure 7c; P + O2 + C: n = 14, ipsi vs. contra p = .0011).These effects were similar when analysing the first paw touches only (Figure 7d; Week 8: V + O2 + V: n = 10, ipsi vs. contra p = .0221,V + O2 + C: n = 15, ipsi vs. contra p = .4087,ns, P + O2 + V: n = 11, ipsi vs. contra p = .1461,ns, P + O2 + C: n = 14, ipsi vs. contra p = .0130;Week 12: V + O2 + V: n = 10, ipsi vs. contra p = .0097,V + O2 + C: n = 15, ipsi vs. contra p > .9999,ns, P + O2 + V: n = 11, ipsi vs. contra p = .0006,P + O2 + C: n = 14, ipsi vs. contra p = .0018).The weights of the 6-OHDA-lesioned mice dropped slightly after the injection, but the animals recovered quickly (Figure S6).Overall, there was no difference in the weights of these mice between the different injection groups at the studied timepoints.
We also tested CDNF treatment in mice with PFF and a lower dose of 6-OHDA (1 μg) to test if the neurorestorative effects of CDNF would be more pronounced in a less acute model (Figures S3 and S4).However, since the combination model with such a low dose of 6-OHDA did not incur any observable pathology, we were unable to test this hypothesis further.
Disease modelling in pre-clinical studies is a crucial step towards ensuring an efficient drug development process.PD, like many other neurodegenerative diseases, is multifaceted, with a complex interaction of genetic, cellular and environmental factors leading to the onset of various motor and non-motor symptoms in patients (Poewe et al., 2017).While the degeneration of DA neurons, formation of aSyn-positive LBs, the occurrence of mitochondrial dysfunction and many such events have been identified in PD, we still do not have a setup that combines these different features in one model.This has in turn resulted in the development of models that have Experiment design for combination model, with and without CDNF.The mice were injected with sonicated PFFs (at Week À2) into a single site, followed by 6-OHDA 2 μg into two sites, 2 weeks after PFFs (at Week 0).CDNF was administered at sites same as 6-OHDA, 4 weeks after PFF injection (at Week 2).All the injections were performed intrastriatally to the right hemisphere.The baseline behavioural tests were conducted just before PFF injections (shown as Week À2 in graphs), followed by behavioural tests just before 6-OHDA injections (shown as Week 0 in graphs), before CDNF injections (shown as Week 2 in graphs), and continued every 2 weeks until perfusion at week 12. (b-f) Effect of CDNF on the survival of DA neurons along the nigrostriatal pathway in different treatments.Tyrosine hydroxylase (TH)-and dopamine transporter (DAT)-positive (b, c) fibres in the striatum and cells in (d, e) SNpc are quantified as percentages of the control (uninjected) sides.(f) paSynpositive aggregates in SNpc of PFFinjected mice, with or without CDNF.The values are presented as mean ± SEM.Ordinary one-way ANOVA and two-way ANOVA, followed by Tukey's or Šid ak's multiple comparisons test, were used for statistical analysis (statistics summarized in Table S1).n = 10-17 mice per group (7-9 weeks old), *p < .05,** p < .01,*** p < .001and **** p < .0001.The '#' symbol represents the statistical difference between the specified group and the vehicle-treated group.
F I G U R E 7 Legend on next page.either face validity or construct validity, and even these to a limited extent.Additionally, only a few of these models are robust enough to be reproduced among different research groups.
At present, the mitochondrial toxin-induced models of PD, using 6-OHDA and 1-methyl-4-phenyl-1,-2,3,6-tetrahydropyridine (MPTP), are considered the gold standard in the field.When administered either locally (6-OHDA) or systemically (MPTP), these toxins produce a loss of DA neurons by causing oxidative stress and resulting in quantifiable motor deficits.The overall effect of these toxins is acute in damaging the nigrostriatal pathway, inducing a rapid loss of striatal fibres and neurons, but is unable to reproduce the slow-emerging progressive nature of the pathology, as seen in patients.This hinders the assessment of not only the early stages of PD pathology but also of the effects of candidate therapeutics along the different stages of PD.Additionally, while the chronic MPTP/probenecid model has been reported to cause aSyn accumulation (Han et al., 2020), the acute models with only MPTP injections do not produce LBlike inclusions, which are the hallmark pathology found in postmortem analysis of most PD patients' brains.In 2012, injection of aSyn PFF into the striatum of mice was shown to recruit endogenous aSyn, and promote its misfolding and aggregation (Luk et al., 2012).After 180 days of the PFF seeding, in addition to formation of phosphorylated and ubiquitinated aSyn aggregates, the PFFs caused loss of DA neurons as well as observable motor deficits in these mice.However, in another study, Albert et al. (2021) injected aSyn PFFs intrastriatally to 3-month-old mice and did not see loss of DA fibre density in the striatum even after 180 days, although progressive aSyn pathology was evident.
In this study, we attempted to find a middle ground between the robust acute loss of DA neurons without paSyn aggregates caused by mitochondrial toxins, and the progressive occurrence of paSyn aggregation, but minimal to no loss of DA neurons in sufficient time as in the PFF seeding model.To pursue this, we injected wildtype mice with both aSyn PFFs and 6-OHDA, 2 weeks apart.We hypothesized that combining a low dose of 6-OHDA, which by itself would not induce robust neuron death, with PFF seeding would lead to neurodegenerative pathology that exhibits LB-like aSyn aggregates and progressive DA neuron degeneration in sufficient timescale.
We saw dose-dependent toxicity with 6-OHDA 1, 2 and 4 μg that, respectively, caused a loss of approximately 3%, 60% and 75% of TH-positive fibres in the striatum and 15%, 55% and 88% of TH-positive DA cell bodies in the SNpc.These values correlated well with the cylinder test and amphetamine-induced rotations, where the mice with the highest dose suffered severe motor imbalance.Interestingly, a recent study characterized the effect of 6-OHDA in a unilateral lesion after injecting a high toxin dose of 18 μg into the dorsolateral striatum of mice with more than 70% survival rate (Mendes-Pinheiro et al., 2021).These animals lost about 50% of their dorsal striatal fibres and approximately 70% of the SNpc THpositive neurons after 14 weeks of toxin injection.An earlier study showed that unilateral injection of 4 μg 6-OHDA into the striatum of mice led to a loss of approximately 20% of TH-positive cells in SN after 3 days of injections, and the numbers progressively declined over the next few weeks until 56 days (8 weeks, experimental endpoint) leading to 80% loss of SN TH-positive cells (Alvarez-Fischer et al., 2008).We, on the other hand, saw extensive fatality for mice injected with 6-OHDA 4 μg, ending up with a 40% survival rate.We paid utmost attention to maintaining the unoxidized state of 6-OHDA during the intrastriatal injection procedure, which can be otherwise challenging and can cause high variability in the 6-OHDA results, between experiments and between different research groups.If 6-OHDA gets oxidized before reaching the striatum, the final effective dose would be much less.It should be noted that we did not inject desipramine, a noradrenergic reuptake inhibitor, before the 6-OHDA injection.We injected 6-OHDA into a limited area in the striatum where it causes a relatively specific DA lesion, as only a few noradrenergic axons are present there (Masini et al., 2021).Thus, in our study, we expected the outcomes originating only from the lesion to the DA system, making the possible minor effects from noradrenergic neurons insignificant.Additionally, we  c, d) Cylinder tests at Week À2 (baseline) and consecutive tests until Week 12 (before perfusion).The graphs represent the ipsilateral, contralateral and both paws touches to the cylinder wall as a percentage of (c) the total wall touches and of (d) the total first paw touches to the wall for 5 min of test duration.The values are presented as mean ± SEM.Two-way ANOVA, followed by Tukey's or Šid ak's multiple comparisons test, was used for statistical analysis (statistics summarized in Table S1).n = 11-17 mice per group (7-9 weeks old), *p < .05,** p < .01 and *** p < .001.The green, blue and red dots on the graphs (x-axis; time) represent the approximate timepoints of PFF, 6-OHDA and CDNF injections, respectively, and V = vehicle, P = PFF, O2 = 6-OHDA 2 μg and C = CDNF.avoided desipramine to eliminate its possible cardiotoxic effects in these mice (Bianchetti et al., 1977;Pentel et al., 1995).We performed 6-OHDA lesion intrastriatally to obtain a progressive neurodegeneration model as opposed to injecting 6-OHDA to the medial forebrain bundle, which yields a more severe model with a near complete loss of DA neurons and fibres (Lundblad et al., 2004;Rentsch et al., 2019).
Intrastriatal injection of PFFs to mice in our experiments, alone and in combination with 6-OHDA 2 μg, led to neurodegeneration and formation of large paSynpositive aggregates in the SNpc and certain cortical areas as seen at 4 and 14 weeks after PFF injections.The combination model did not induce a progressive neuron loss, and instead, it mostly mirrored the acutely developing 6-OHDA-induced DA neuron degeneration that was evident already at 2 weeks after 6-OHDA administration.Interestingly, the SNpc aggregate counts at 14 weeks were clearly lower in the PFF-seeded mice injected with 6-OHDA.Our straightforward speculation is that the DA neurons that form these aSyn aggregates were more vulnerable to 6-OHDA-induced toxicity.We think it will be interesting to also assess the possible effects of the 6-OHDA-induced oxidative stress on the dynamics of aSyn aggregation.For example, the oxidative stress could hinder the formation of these aggregates or affect their phosphorylation, which could have compromised their detection with the antibody used in our study.At all timepoints, 6-OHDA 2 μg (with and without PFFs) showed behavioural phenotypes, especially in amphetamine-induced rotation and cylinder tests.These correlated well with the histological quantification of the neurodegeneration caused by 6-OHDA compared to PFFonly injections.Ultimately, to achieve a reliable neurodegenerative model over a time course of 14 weeks, the combination of 5 μg of PFF and 2 μg of 6-OHDA allowed us to initiate aSyn aggregation in the neurons and produce a robust loss of DA neurons in these animals.
The combination model proposed here couples the two important features of PD-occurrence of oxidative stress and protein aggregation-that we originally wanted to establish in our model.In 2021, Merghani et al. developed a model catering to a similar paradigm of additional stress where they combined human PFFs with multiple injections of MPTP 6 weeks after PFF injections in mice and demonstrated an increased loss of THpositive neurons and paSyn aggregation with 5 μg PFF + 10 mg/kg.b.wt MPTP (Merghani et al., 2021).We saw a similar effect in our PFF + 6-OHDA-injected mice.However, MPTP injected systemically leads to bilateral degeneration of DA neurons.Our unilateral combination model facilitates the use of the uninjected hemisphere as a control within the injected animal and allows use of motor asymmetry behavioural tests for further assessments.
The drawbacks of the study were the high variability in the pathology within treatment groups and between experiments.With some further optimization, our combination model could constitute a more suitable research and drug testing platform.In the first pilot experiments, where we tested the different 6-OHDA doses, we used older mice (2-4 months old) and observed a more aggressive pathology in them, as compared to the second part of the experiment where slightly younger animals (7-9 weeks old) were used (this was due to the availability of animals at that time).This aligns with the idea that age is one of the prime risk factors for PD.Thus, continuing with the 2-4 months old mice for the later part of the experiment could have possibly resulted in better pathological outcomes in the model.Second, instead of injecting the 6-OHDA after 2 weeks of PFF seeding, a longer incubation period with PFF, for instance, 4-6 weeks, would lead to a more robust paSyn aggregation at the time of 6-OHDA lesion, possibly resulting in the manifestation of further PD pathology in this model.Third, a higher concentration of PFFs could be tested to try to induce faster and more extensive aggregation pathology that could promote faster progression of neurodegeneration.Finally, this combination of PFF and 6-OHDA was tested only in male mice, which indeed is a limitation of our study.Acknowledging the sex-based differences in the resulting susceptibility of animals to neurotoxins (Masini et al., 2021;Tamas et al., 2005), it would be interesting to test the effect of the combination in male versus female mice.
Lastly, we tested whether CDNF could restore the nigrostriatal dopamine tract and affect the paSyn aggregate pathology in our combination PD model.The restorative action of CDNF has been observed previously in 6-OHDA-lesioned rats by either a single injection after 4 weeks of the lesion (Lindholm et al., 2007;Voutilainen et al., 2017), 2 weeks of chronic infusion post-lesion (Voutilainen et al., 2011), AAV-mediated CDNF overexpression after 6 weeks of the lesion (Ren et al., 2013;Wang et al., 2017) or grafting mesenchymal stem cells overexpressing CDNF after 1 week of 6-OHDA lesion (Jiaming & Niu, 2015).Nadella et al. (2014) also reported that CDNF when transfected into the SN of rats after 15 days of the 6-OHDA-induced striatal lesion, led to a reduction in neuroinflammatory markers in these animals.In our study, we administered CDNF after 2 weeks of 6-OHDA injections because the animals sustained acute damage to the DA striatal fibres at this timepoint and presented spontaneous partial recovery by Week 12.This encouraged us to test whether CDNF could provide further neurorestorative support during this time.With CDNF treatment, although we were able to promote the restoration of DA neurons to some extent, as well as rescue the motor performance in the cylinder test in mice injected with 6-OHDA only, we did not observe any neurorestoration in the combination model where both 6-OHDA-induced toxicity and PFF-seeded aSyn aggregation were causing the pathology.The CDNF treatment dose and its timing might have been suboptimal in our study.6-OHDA (2 μg) led to approximately 70-80% loss of DA axons 2 weeks after its injection.Matlik et al. (2017) have shown that the retrograde transport of CDNF is hampered in rats with 6-OHDA-lesioned striatum, resulting from damage to their nigrostriatal tracts.Subsequently, this leads to reduced availability of CDNF in SN, where it potentially has therapeutic targets.In the striatum of 6-OHDA-only animals, where CDNF was injected, it did rescue the DA neuron fibre loss from the 6-OHDAinduced neurodegeneration.CDNF had no effect on neurodegeneration or motor dysfunction in the combination model, suggesting that the resulting aSyn aggregation pathology combined with 6-OHDA neurotoxicity was too extensive, at least with this dose and timing.Instead of injecting CDNF to the striatum only once, it might be necessary to administer multiple injections over time or use a viral vector-mediated overexpression of CDNF or perform chronic infusion of the protein to allow for effective availability along the dopaminergic pathway.

| CONCLUSION
In this study, we emphasize the importance and challenges of developing research models that would encompass many of the pathological features of PD to be able to investigate the disease mechanisms and therapeutic approaches more efficiently.With our combination model, we strived to recreate the two central pathological elements of PD.Overall, our findings indicate that with further optimization, this kind of combination models hold promise to serve as an effective tool to study the complexity of a multifactorial disease such as PD.

F
I G U R E 3 Motor function of mice injected with PFFs and two different doses of 6-OHDA.(a) 10 mg/kg of D-amphetamine was injected intraperitoneally into mice, and the rotations were recorded over 90 min.(b) Rotarod test-the latency of fall recorded on the rotarod at 4-40 rpm for 4 min.(c, d) Cylinder tests at Week À2 (baseline) and consecutive tests until Week 12 (before perfusion).
-9 weeks old), *p < .05,** p < .01 and **** p < .0001.In graphs, V = vehicle, P = PFF and O2 = 6-OHDA 2 μg.(h) Representative images for TH-and paSyn-positive stainings in the SN and ventral cortex of mice.The scale bar is 100 μm.F I G U R E 5 Motor function of mice injected with PFFs and/or 6-OHDA 2 μg.The baseline behavioural tests were conducted just before PFF injections (shown as Week À2 in graphs), followed by behavioural tests just before 6-OHDA injections (shown as Week 0 in graphs), and 2 weeks after that (Week 2).(a) 10 mg/kg of D-amphetamine was injected intraperitoneally into mice, and the rotations were recorded over 90 min.(b) Rotarod test-the latency of fall recorded on the rotarod at 4-40 rpm for 5 min.The values are presented as mean ± SEM.
List of antibodies and reagents T A B L E 2 Saku Reunanen: Data curation; formal analysis; investigation; writing-review and editing.Kelvin C Luk: Conceptualization; resources; writing-review and editing.Merja H. Voutilainen: Conceptualization; data curation; formal analysis; funding acquisition; investigation; methodology; project administration; resources; supervision; writing-review and editing.