Growth differentiation factor 5 exerts neuroprotection in an α-synuclein rat model of Parkinson’s disease

The concept of neurotrophic factor therapy holds significant promise as a disease-modifying therapy for Parkinson’s disease (Paul and Sullivan, 2019). Whone et al. (2019) recently reported the results of a randomized, double-blind, placebo-controlled trial of intermittent intraputamenal convection-enhanced delivery of glial cell line-derived neurotrophic factor (GDNF) in Parkinson’s disease patients. Despite an extensive body of excellent work and significant improvements to dosing and delivery, this trial failed to reach its primary end point. This outcome is largely consistent with an earlier randomized, placebo-controlled trial of GDNF in Parkinson’s disease (Lang et al., 2006). Preclinical studies have shown that delivery of adeno-associated virus (AAV)-GDNF to either the striatum or substantia nigra failed to protect dopaminergic neurons or their terminals in a rat model of Parkinson’s disease generated through viral-mediated overexpression of human wild-type α-synuclein (AAV-αSyn) (Decressac et al., 2011) or A30P mutant α-synuclein (LV-αSyn) (Lo Bianco et al., 2004). This has been suggested to be due to αSyn-induced downregulation of Nurr1 and Ret in the midbrain, measured at 2 weeks post-surgery (Decressac et al., 2012). Similar αSyn-induced downregulation of Ret was reported in the substantia nigra of Parkinson’s disease patients (Decressac et al., 2012). This may explain the lack of clinical efficacy of GDNF, as the RET receptor has been reported to be critical for the dopaminergic neurotrophic effects of GDNF in vivo (Drinkut et al., 2016). However, others found that transgenic overexpression of human α-synuclein in mice, or intranigral administration of AAV-αSyn in rats, did not downregulate Nurr1 or tyrosine hydroxylase (Th) expression after 8 weeks (Su et al., 2017). Therefore, the effect of α-synuclein on RET expression remains an open and important question. More generally, the lack of clinical efficacy of GDNF has raised the larger question of whether neurotrophic factor delivery as a therapeutic approach remains viable, or if this approach requires further testing and evaluation in clinical settings (Whone et al., 2019). Another consideration is that if RET receptors are downregulated or not functioning optimally in the Parkinson’s disease brain, then RET-independent neurotrophic factors may have therapeutic benefit, assuming that α-synuclein does not also affect the expression of their key signalling mediators. Thus, a critical requirement to justify the development of RET-independent neurotrophic factor therapy, is that these factors must display therapeutic benefit in the α-synuclein model of Parkinson’s disease, which more closely mimics the disease pathology than other rodent models. Our aims in the present study were 3-fold. First, we investigated whether α-synuclein overexpression reduced Ret expression in the rat substantia nigra in vivo. Second, we examined if a RET-independent neurotrophic factor, growth differentiation factor 5/bone morphogenetic protein 14 (GDF5/BMP14, hereafter called GDF5), could protect dopaminergic neurons and their axon terminals in an AAV-αSyn rat model of Parkinson’s disease, similar to that used in the Decressac et al. (2011) GDNF study. Like GDNF, GDF5 is a member of the transforming growth factor (TGF)-β protein superfamily (Hegarty et al., 2014b), and it exerts neuroprotective effects similar to those of GDNF in neurotoxin models of Parkinson’s disease (Sullivan et al., 1997, 1998). GDF5 signals through a receptor complex consisting of bone morphogenetic protein receptor (BMPR)2, BMPR1B and the transcription factor, SMAD1 (Hegarty et al., 2013, 2014a). Third, we investigated whether α-synuclein overexpression affected the expression of these three key signalling mediators of GDF5’s neurotrophic actions. Adult Sprague-Dawley rats received unilateral intranigral injection of either AAV2/6-αSyn (wild-type human α-synuclein under the control of synaptophysin promoter) or the corresponding null control vector, and were sacrificed 20 weeks later (Supplementary material). AAV-αSyn induced the expression of wild-type human α-synuclein throughout the midbrain and in axonal terminals throughout the ipsilateral striatum (Fig. 1A). Additionally, there was strong pSer-129-α-synuclein expression in the ipsilateral substantia nigra, within cell bodies and processes (Fig. 1B). Intranigral AAV-αSyn led to significant reductions in the expression of transcripts for Th (Fig. 1C) and dopamine transporter protein (Dat) (Fig. 1D), but not NeuN (Fig. 1E). AAV-αSyn also induced a significant reduction in Ret expression (Fig. 1F), but no change in the expression of Bmpr2, Bmpr1b and Smad1, all of which are required for the dopaminergic neurotrophic effects of GDF5 (Fig. 1G). Given these findings, and since previous studies have shown that LV- or AAV-GDNF did not protect nigrostriatal neurons in the AAV-αSyn rat model (Decressac et al., 2011), we next investigated whether AAV-GDF5 could protect nigrostriatal neurons against α-synuclein-induced toxicity, in this Parkinson’s disease model. α-Synuclein downregulates Ret but not key mediators of BMP signalling in vivo. Representative images showing (A) wild-type α-synuclein (αSyn) immunostaining in rat striatum and midbrain, and (B) pSer129-αSyn staining in substantia nigra, at 20 weeks after unilateral stereotaxic injection of AAV-αSyn vector into the substantia nigra. Blue arrows indicate dopaminergic neuronal cell bodies and red arrows indicate their processes. Real-time PCR data showing expression of transcripts for (C) Th (**t = 4.211, df = 8, P = 0.0029), (D) Dat (***t = 5.28, df = 8, P = 0.0007), (E) NeuN, (F) Ret (*t = 2.45, df = 8, P = 0.0399), and (G) Bmpr2, Bmpr1b and Smad1 (key mediators of BMP signalling) (n = 5 per group). All data are presented as mean ± standard error of the mean (SEM) and analysed by Student’s t-test. α-Synuclein downregulates Ret but not key mediators of BMP signalling in vivo. Representative images showing (A) wild-type α-synuclein (αSyn) immunostaining in rat striatum and midbrain, and (B) pSer129-αSyn staining in substantia nigra, at 20 weeks after unilateral stereotaxic injection of AAV-αSyn vector into the substantia nigra. Blue arrows indicate dopaminergic neuronal cell bodies and red arrows indicate their processes. Real-time PCR data showing expression of transcripts for (C) Th (**t = 4.211, df = 8, P = 0.0029), (D) Dat (***t = 5.28, df = 8, P = 0.0007), (E) NeuN, (F) Ret (*t = 2.45, df = 8, P = 0.0399), and (G) Bmpr2, Bmpr1b and Smad1 (key mediators of BMP signalling) (n = 5 per group). All data are presented as mean ± standard error of the mean (SEM) and analysed by Student’s t-test. We modelled our approach on the study by Decressac et al. (2011), using a similar protocol with some modifications. Adult rats received stereotaxic unilateral intranigral injection of α-synuclein vector (AAV2/6-αSyn) and concomitantly either control vector (AAV2/5-Cont) or vector carrying the human GDF5 transgene (AAV2/5-GDF5) (Fig. 2A). Rats were sacrificed 20 weeks later for analysis of transgene expression and nigrostriatal integrity. There were no differences in α-synuclein expression levels between control and GDF5-treated animals, in the striatum (Fig. 2B and C) or substantia nigra (Fig. 2D). AAV-GDF5 induced expression of human GDF5 within dopaminergic neurons in the substantia nigra (Fig. 2E). To evaluate the neurodegeneration induced by overexpression of human wild-type α-synuclein and the potential neuroprotective effect of AAV-GDF5, we measured the numbers of TH-positive and DAT-positive neurons in the ipsilateral and contralateral substantia nigra. We found that, at 20 weeks after injection of AAV-αSyn, there was ∼36% loss of TH-positive (Fig. 2F and G) and ∼30% loss of DAT-positive (Fig. 2H) neurons, compared to the intact side. AAV-mediated delivery of GDF5 to the substantia nigra prevented this loss of dopamine neurons (P < 0.01 compared to AAV-Cont group) (Fig. 2G and H). AAV-GDF5 maintains nigrostriatal integrity and striatal dopamine levels in the rat AAV-αSyn model of Parkinson’s disease. (A) Animals received unilateral injection of AAV-αSyn with either AAV-Cont (blue) or AAV-GDF5 (red) into the substantia nigra (SN). (B) Representative images of α-synuclein immunostaining in the striatum (Str). Quantification of α-synuclein immunostaining in (C) striatum and (D) substantia nigra. (E) Representative photomicrographs showing expression of the human (h) GDF5 transgene (green) in TH-positive neurons (red; white arrows) in the substantia nigra at 20 weeks post-injection of AAV-GDF5. (F) Representative photomicrographs of TH-immunostained sections through the substantia nigra on the ipsilateral side. Numbers of (G) TH-immunopositive (**t = 3.46, df = 12, P = 0.0047) and (H) DAT-immunopositive neurons (**t = 3.76, df = 12, P = 0.0027) in the ipsilateral substantia nigra, expressed as percentage of the intact side (n = 7 per group). (I) Representative photomicrographs of TH-immunostained sections through the ipsilateral and contralateral striatum. (J) Quantification of TH immunoreactivity in the ipsilateral striatum, expressed as percentage of the intact side (**t = 3.86, df = 12, P = 0.0023) (n = 7 per group). (K) HPLC analysis of dopamine (DA) levels in the ipsilateral striatum, expressed as percentage of the intact side (*t = 2.397, df = 17, P = 0.0283) (n = 9–10 per group). All data are presented as mean ± SEM and analysed by Student’s t-test. AAV-GDF5 maintains nigrostriatal integrity and striatal dopamine levels in the rat AAV-αSyn model of Parkinson’s disease. (A) Animals received unilateral injection of AAV-αSyn with either AAV-Cont (blue) or AAV-GDF5 (red) into the substantia nigra (SN). (B) Representative images of α-synuclein immunostaining in the striatum (Str). Quantification of α-synuclein immunostaining in (C) striatum and (D) substantia nigra. (E) Representative photomicrographs showing expression of the human (h) GDF5 transgene (green) in TH-positive neurons (red; white arrows) in the substantia nigra at 20 weeks post-injection of AAV-GDF5. (F) Representative photomicrographs of TH-immunostained sections through the substantia nigra on the ipsilateral side. Numbers of (G) TH-immunopositive (**t = 3.46, df = 12, P = 0.0047) and (H) DAT-immunopositive neurons (**t = 3.76, df = 12, P = 0.0027) in the ipsilateral substantia nigra, expressed as percentage of the intact side (n = 7 per group). (I) Representative photomicrographs of TH-immunostained sections through the ipsilateral and contralateral striatum. (J) Quantification of TH immunoreactivity in the ipsilateral striatum, expressed as percentage of the intact side (**t = 3.86, df = 12, P = 0.0023) (n = 7 per group). (K) HPLC analysis of dopamine (DA) levels in the ipsilateral striatum, expressed as percentage of the intact side (*t = 2.397, df = 17, P = 0.0283) (n = 9–10 per group). All data are presented as mean ± SEM and analysed by Student’s t-test. We next investigated whether nigral GDF5 overexpression could prevent the loss of striatal dopaminergic terminals that is known to occur in this model (Decressac et al., 2011). Quantitative analysis of TH-positive striatal fibre innervation revealed that nigral overexpression of α-synuclein induced ∼45% reduction (Fig. 2I and J). In contrast, nigral overexpression of GDF5 by AAV-GDF5 had a protective effect on striatal dopaminergic terminal density (P < 0.01 compared to the AAV-Cont group) (Fig. 2I and J). Because of this significant protection of axonal terminals in the striatum, we repeated the experiment using a second cohort of animals to determine if this protection of axonal terminals was translated into maintenance of striatal dopamine levels. Animals received intranigral injection of AAV-αSyn and concomitantly either AAV-Cont or AAV-GDF5. Analysis of striatal dopamine levels using high performance liquid chromatography (HPLC) after 20 weeks revealed that nigral overexpression of α-synuclein induced ∼45% reduction in striatal dopamine levels (Fig. 2K). AAV-mediated nigral overexpression of GDF5 had a protective effect on striatal dopamine levels (P < 0.05 compared to the AAV-Cont group) (Fig. 2K). Collectively, these data show that intranigral injection of AAV-GDF5 protected dopaminergic neurons and their terminals, maintaining striatal dopamine levels in the AAV-α Syn rat model of Parkinson’s disease. The first implication of these findings is that, in agreement with Decressac et al. (2012), we found that viral-mediated overexpression of α-synuclein led to a significant reduction in Ret expression in the adult rat substantia nigra at 20 weeks post-surgery. However, it is important to note that the reduction of 25% in our study was observed at 20 weeks, in contrast with the much greater (∼65%) reduction reported by Decressac et al. after 2 weeks. As we found that ∼75% of Ret expression remained in the α-synuclein-overexpressing brain, these data suggest that the signalling capacity of GDNF may be retained, at least partially, in these animals. This is consistent with the significant putamenal increase in 18F-DOPA uptake in GDNF-treated patients at Week 40 reported in the Whone et al. (2019) study. The key determinant of clinical efficacy of GDNF may therefore be the level of α-synuclein burden in the host brain. High levels of α-synuclein may lead to an early and strong downregulation of Ret, similar to that reported by Decressac et al. (2012), whereas modest α-synuclein levels may lead—in the short term—to minimal reductions in Ret, as observed by Su et al. (2017) after 8 weeks, but more substantial impairment of Ret expression in the longer term, as seen in our study. The second key finding of our study is that the RET-independent neurotrophic factor, GDF5, exerted neuroprotective effects in the AAV-αSyn rat model, in which GDNF was not effective (Decressac et al., 2011). Moreover, and unlike RET, the expression of GDF5’s receptors and downstream transcription factor SMAD1 was not affected by α-synuclein overexpression, at least at the 20-week time-point examined. The observed neuroprotective effects of GDF5 in this preclinical model support the theory that Ret-independent neurotrophic factors may have therapeutic benefit in Parkinson’s disease and are worthy of further exploration. Further evidence that supports this proposal comes from an independent study by Vitic et al. (2021). They show that lentiviral delivery of BMP5/7 prevented A53T-α-synuclein-induced loss of dopamine neurons, motor impairment and associated gliosis in a mouse model of Parkinson’s disease. They also demonstrated that loss of BMP/SMAD signalling led to the accumulation of α-synuclein (Vitic et al., 2021). This does not mean that RET-dependent neurotrophic factors, such as GDNF, are not worthy of further testing (in fact our findings on RET suggest that they are); on the contrary, collectively these data suggest that neurotrophic factor therapy remains a viable therapeutic approach for protecting dopaminergic neurons against the effects of α-synuclein in Parkinson’s disease. The data that support the findings of this study are available from the corresponding author, upon reasonable request. This work was supported by a research grant from Science Foundation Ireland (SFI) under the grant number 15/CDA/3498 (G.W.O.) and a RISAM PhD Scholarship from Cork Institute of Technology (R00094948) (S.R.G.). The authors report no competing interests.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.