Post-Mortem Studies of Neurturin Gene Therapy for Parkinson's Disease: Two Subjects with 10 Years CERE120 Delivery
Relevant conflicts of interest/financial disclosures: Nothing to report.
Abstract
Background
Neurturin is a member of the glial cell line-derived neurotrophic factor family of neurotrophic factors and has the potential to protectdegenerating dopaminergic neurons.
Objective
Here, we performed post-mortem studies on two patients with advanced Parkinson's disease that survived 10 years following AAV-neurturin gene (Cere120) delivery to verify long-term effects of trophic factor neurturin.
Methods
Cere120 was delivered to the putamen bilaterally in one case and to the putamen plus substantia nigra bilaterally in the second. Immunohistochemistry was used to examine neurturin, Rearranged during transfection(RET), phosphor-S6, and tyrosine hydroxylase expressions, inflammatory reactions, and α-synuclein accumulation.
Results
In both patients there was persistent, albeit limited, neurturin expression in the putamen covering 1.31% to 5.92% of the putamen. Dense staining of tyrosine hydroxylase-positive fibers was observed in areas that contained detectable neurturin expression. In substantia nigra, neurturin expression was detected in 11% of remaining melanin-containing neurons in the patient with combined putamenal and nigral gene delivery, but not in the patient with putamenal gene delivery alone. Tyrosine hydroxylase positive neurons were 66% to 84% of remaining neuromelanin neurons in substantia nigra with Cere120 delivery and 23% to 24% in substantia nigra without gene delivery. More RET and phosphor-S6 positive neurons were observed in substantia nigra following nigral Cere120. Inflammatory and Lewy pathologies were similar in substantia nigra with or without Cere120 delivery.
Conclusions
This study provides evidence of long-term persistent transgene expression and bioactivity following gene delivery to the nigrostriatal system. Therefore, future efforts using gene therapy for neurodegenerative diseases should consider means to enhance remaining dopamine neuron function and stop pathological propagation. © 2023 The Authors. Movement Disorders published by Wiley Periodicals LLC on behalf of International Parkinson and Movement Disorder Society.
Neurotrophic factors have neuroprotective and neurorestorative properties. Their therapeutic potential has been tested in various neurodegenerative diseases. Neurturin (NRTN) belongs to the glial cell line-derived neurotrophic factor (GDNF) family of neurotrophic factors and shares a 42% similarity in amino acid sequence with mature GDNF.1 GDNF binds to GDNF-family receptor-α1 (GFRα1), whereas NRTN ligand to GDNF-family receptor-α2 (GFRα2), then forms a complex with RET, to activates the mitogen-activated protein kinase (MAPK) and phosphoinositide 3-kinases (PI3K) signaling pathway, which plays an important role for neurite outgrowth and neuron survival.2, 3 NRTN might crosstalk with GFRα1 and GDNF with GFRα2 and GFRα34, 5 at high doses. NRTN gene or GFRα2 knockout mice have shown delayed growth of enteric, sensory, and parasympathetic neurons4-6 but have normal number of nigral dopaminergic neurons5 indicating that endogenous NRTN is not the survival factor for this cell type. Further study has revealed relatively low expression of NRTN in the striatum and substantia nigra.7 However, administrated NRTN on ventral mesencephalon cultures or embryonic dopaminergic neurons exhibit potent survival effect on both developing and mature dopaminergic neurons8 comparable with GDNF.9 These studies demonstrated that NRTN can protect and function and in a manner similar to GDNF.
Adeno-associated virus serotype 2 (AAV-2) vector encoding the human NRTN gene (Cere120) was designed to deliver the NRTN to the putamen and substantia nigra to protect or restore degenerating nigrostriatal dopaminergic neurons and to potentially provide clinical benefits for patients with Parkinson's disease (PD).10, 11 To extend our philosophy that the bravery shown by subjects by allowing us to operate on their brain in this manner, we believe that each case deserves to be published allowing for an eventual cohort to be collated from a series of case studies. In this regard, our series of post-mortem studies revealed that delivery of the NRTN gene has been continually expressing for 10 years with tyrosine hydroxylase (TH) upregulation within NRTN-expressing regions (Supplementary Table S1).
Presently, we examined two recent post-mortem PD brains that had Cere120 therapy. One subject (77 years old, male, Monttpelier/VT) underwent Cere120 therapy delivered bilaterally to both the putamen and substantia nigra 10 years before death (subject Cere120-21 SN/Pt). The second subject (76 years old, male, CarsonCity/NV) underwent Cere120 treatment delivered into the putamen bilaterally only and died 10 years post-surgery as well (subject Cere120-22 Pt). All PD subjects were diagnosed by a movement disorder specialist before entering the clinical trial, and the diagnosis was confirmed at post-mortem in each case by a board-certified neuropathologist. Importantly, neither of these subjects had other significant neuropathological changes and specifically, neither had any pathological change thought to be related to the gene therapy procedure. One PD subject with comparable Unified Parkinson's Disease Rating Scale (UPDRS) who did not undergo gene therapy was selected as disease control. One age-matched subject without psychiatric or neurological illnesses during life or neuropathological abnormalities at post-mortem was included as healthy control. An immunoperoxidase method was processed as described previously12, 13 to visualized NRTN, TH, RET, phosphor-S6 ribosomal protein (p-S6), phosphorylated α-synuclein (p-S129), and TMEM119 expression. The specificity of these antibodies used in this study has been verified in our previous studies.12-14 One series of sections was used for each immunostaining. Immunostained region and neurons were estimated using unbiased stereology.12 Briefly, the volume of whole putamen and area with NRTN staining or enhanced TH staining were separately estimated using the Cavalieri Estimator (StereoInvestigator v2021, MicroBrightField). The percent of NRTN or TH covered areas were calibrated by NRTN or TH stained areas/whole putamenal volume times 100. NRTN, TH, RET, p-S6 positive nigral neurons, and neuromelanin-laden (NM-Laden) neurons without any immunolabeling were separately estimated with Optical Fractionator (StereoInvestigator v2021, MicroBrightField) and nigral volume was estimated with Cavalieri Estimator. Densities of nigral neurons were calibrated by estimated NRTN-, TH-, RET-, p-S6-positive, and NM-laden neuronal number from Optical Fractionator/nigral volume from Cavalieri Estimator.12, 13
Immunohistochemistry revealed focal regions of diffuse NRTN staining within bilaterally putamen (Fig. 1A,E) in both subjects. Light NRTN staining was observed in substantia nigra of CER120-21SN/Pt (Fig. 2A,C). No NRTN immunoreactivity was detected in the caudate nucleus, global pallidus, or cortex in either case. Quantification of the NRTN-stained area was estimated to cover 2.12% (left) and 1.31% (right) of putamen (Fig. 1M) and 7.41% (left) and 4.26% (right) of substantia nigra (Fig. 2I) in subject Cere120-21 SN/PT. NRTN-positive perikarya and main processes were observed in substantia nigra (Fig. 2B,D). The density of NRTN positive neurons was 30.83/mm3 on left and 51.91/mm3 on right substantia nigra. NRTN labelling was detected in 11% of remaining nigral melanized neurons (Fig. 2J). In Cere120-22 Pt subject with putamenal Cere120 treatment only, the NRTN-stained area was estimated to cover 5.92% (left) and 3.54% (right) putamen (Fig. 1M). The NRTN-labeling was undetectable in substantia nigra (Fig. 2E–H,J).


TH-positive fibers, clearly above background, were observed in areas of the putamen (Fig. 1B,C,F,G) coincident with NRTN staining (Fig. 1A,E) in both subjects. Approximately 5.75% to 11.35% of putamenal area had obvious NRTN-mediated TH expression (Fig. 1N) in subject Cere120-21 SN/Pt. Optical density (OD) measurements of TH immunoreactive fiber density exhibited 23.96 OD in areas within NRTN expression and 5.49 OD in areas with absent NRTN expression (Fig. 1O). In contrast, the subject Cere120-22 Pt presented sparse TH-positive fibers in the area with NRTN expression (Fig. 1F,G). The enhanced TH expression was limited to 0.29% to 0.54% of putamenal area (Fig. 1N). The optical density was 6.67 in the area within NRTN expression and 0.46 in the area absent of NRTN expression (Fig. 2E-H). In the area absent of NRTN expression, the TH positive fiber were comprehensively low (Fig. 1D, 1H) in both subjects received Cere120 in a manner seen in untreated PD12 (Fig. 1K,L).
Scattered TH-positive neurons were distributed in substantia nigra of both subjects with Cere120 delivery (Fig. 3E–L). Subject Cere120-21SN/Pt displayed more TH positive neurons and processes than subject Cere120-22 Pt. Stereological estimates revealed that the density of TH positive neurons was 411.40/mm3 in left and 190.67/mm3 in right substantia nigra of Cere120-21 SN/Pt and 72.88/mm3 in left and 64.22/mm3 in right substantia nigra of Cere120-22 Pt. The percentage of TH positive neurons in remaining nigral neuromelanin neurons is 66% to 84% in subject Cere120-21SN/Pt and 23% to 24% in subject Cere120-22 Pt. Age-matched PD (UPDRS, 37; Hoehn and Yahr, 3) with non-gene delivery had 140/mm3 of TH positive neuronal density (Fig. 3M) that was similar to the PD with putamenal Cere120 delivery. However, substantia nigra with Cere120 delivery displayed a higher percentage (66%–84%) of TH positive neurons in neuromelanin laden neurons than PD (42.62%) (Fig. 3N) without gene delivery.

NRTN signaling is mediated through a multi-component receptor system including GFRα2, GFRα1, RET, and glycosyl phosphatidylinositol-link protein, which then initiates the signal pathway.3 RET and p-S6 play an important role in NRTN signaling pathway.15 Regarding this, we examined RET and p-S6 expression in these PD brains with Cere120 delivery.
Immunohistochemistry revealed more RET-positive neurons around the NRTN-expressing area of Cere120-21-SN/Pt (Fig. 4C,D), but less in non-NRTN expressing substantia nigra of Cere120-22-Pt (Fig. 4E,F). Density of RET positive neurons were 725.38/mm3 in substantia nigra of Cere120-21 SN/Pt and 275.03/mm3 in substantia nigra of Cere120-22 Pt (Fig. 4I). The percentage of RET positive neurons in remaining nigral neuromelanin neurons was 86.49% in Cere120-21SN/Pt and 52.52% in subject Cere120-22 Pt. Age-matched PD with non-gene delivery had 40.39% of RET positive neurons in remaining neuromelanin neurons (Fig. 4I). Similarly, there were more p-S6-pisitive nigral neurons (501.61/mm3) in Cere120-21-SN/Pt (Fig. 5C,D), but less (286.27/mm3) in Cere120-22-Pt (Fig. 5E,F). A total of 78.61% of nigral neuromelanin neurons were p-S6-positive in Cere120-21-SN/Pt and 36.06% in Cere120-22-Pt (Fig. 5I). PD with non-gene delivery displayed a much lower percentage of p-S6 positive neurons in remaining nigral neuromelanin neurons (15%) (Fig. 5G–I). This data indicated that nigral NRTN expression could keep more neurons expressing signaling protein in GDNF signal pathway.


To evaluate whether there was an inflammatory reaction in Cere120 treated brains, we examined a specific microglial marker, TMEM119, which is a transmembrane protein expressed by resident microglia. There was no obvious increase of microglial cells with Cere120 delivered to the putamen (Supplementary Fig. S1G,H,K,L) and was similar to the control brain (Supplementary Fig. S1C,D). Intense TMEM119-labeled cells were observed in substantia nigra (Supplementary Fig. S1E,F,I,J) in subjects with or without Cere120 delivery. Even more activated microglial cells were observed in PD nigra without gene delivery (Supplementary Fig. S1M,N).
To examine whether NRTN affects α-synuclein accumulation and aggregation, we examined phosphorylated α-synuclein in midbrain. Immunohistochemistry revealed α-synuclein inclusions existed similarly in substantia nigra whether or not the CERE120 was delivered (Supplementary Fig. S2).
These observations further confirmed that gene delivery of NRTN can induce long-standing transgene expression in Parkinson's disease subjects lasting for at least 10 years even though the breadth of NRTN distribution was too small to affect any clinical benefit.16, 17 This is likely due in part to the small volumes administeresd. Indeed, the NRTN expressed from Cere120 was distributed in these cases in a small portion of putamen and substantia nigra. As compared with previous studies (Supplementary Table S1), the putamenal area with NRTN expression was gradually reduced over time. Subject with both putamenal and nigral Cere120 deliveries displayed enhanced TH fibers relative to the case with putamenal delivery only.
More NRTN positive neurons were observed in subjects with both nigral and putamenal Cere120 delivery than the putamenal delivery alone. Similarly, TH positive neuron number was greater in subjects with nigral Cere120 delivery relative to the case with non-nigral CERE120 delivery. Present and previous studies suggest that Cere120 can enhance dopaminergic neuronal functions of PD cases for at least 10 years.
In the present study we observed many NRTN-positive neurons (41.32/mm3) and intense RET expression (725.38/mm3) around the area where Cere120 was directly delivered to the substantia nigra. When Cere120 was delivered to putamen, there was undetectable expression of NRTN in nigral neurons and lower RET expression (275.03/mm3) at anatomical levels similar to what was seen in a PD case without gene delivery (284.41/mm3). In addition, the number of p-S6-positive neurons was greater in Cere120 direct delivery to substantia nigra (501.61/mm3) when compared to Cere120 delivery to putamen (286.27/mm3) alone. These data suggest that direct nigral Cere120 could be more effective to activate it's signaling pathway than putamenal Cere120 delivery alone.
There was undetectable NRTN in nigral neuron where Cere120 was delivered to putamen in this study. Previous studies also demonstrated that there were very few NRTN-positive nigral neurons on the subjects with putamenal Cere120 delivery alone.18, 19 These data suggested that the remaining nigrostriatal dopaminergic neurons had lost the ability to transport NRTN retrogradely from putamen to substantia nigra, although there was intense local NRTN expressing from Cere120. The disruption of axonal transport20 should be considered as a challenge for gene therapy in PD.
Present and previous12, 18, 19 studies demonstrated that expressed NRTN from Cere120 was only diffused in a small portion of putamen or substantia nigra and the majority of putamen and nigra did not have any NRTN coverage. This limited NRTN diffusion could be associated with several factors. Several reports indicated that GNDF family, including NRTN, GDNF, and artemin, have high affinity for heparin that reduces tissue distribution.21-23 The AAV serotype is another factor to influence NRTN diffusion.24 The AAV2/2 capsid diffuses poorly. Modification of the capsid could improve AAV-mediated neuronal transduction to minimize the existing limitations of gene therapy.
Putamenal microglia cell density and distribution in subjects with Cere120 deliveries was similar to control brain suggesting that long term NRTN expression does not exacerbate the inflammatory reaction normally seen in PD. Indeed, the intense TMEM119-labeling seen in the substantia nigra with or without Cere120 delivery implies that this is part of the PD neurodegenerative process.
As expected, this study demonstrated similar α-synuclein inclusions in substantia nigra with or without Cere120 delivery indicating that NRTN was not involved in process of the evolving synucleinopathy in PD. GDNF therapy on rat α-synuclein models of PD has demonstrated that GDNF overexpression did not effect α-synuclein aggregation25, 26 whereas overexpression of α-synuclein results in negatively regulate Nurr1 and its downstream gene expression.26, 27
In summary, we demonstrate that NRTN gene delivery can induce long term (10 years) transgene expression and enhance TH expression and fibers in focal areas of putamen and substantia nigra with NRTN expression. Signaling pathway for NRTN such as RET and p-S6 is upregulated in remaining nigral neurons around NRTN-expressing area. However, the NRTN expression in this trial was distributed to a limited of putamen or nigra and its trophic effects are gradually reduced overtime. Inflammation and α-synuclein pathology are constantly contributed to nigral dopaminergic neurodegeneration.
Acknowledgments
We thank the patients and controls for their participation in this study. We also thank Yinzhen He in Arizona State University for histological assistance. This work was funded by The Michael J. Fox Foundation for Parkinson's Research (Grant ID, MJFF-007847) to J.H.K.
Author Roles
(1) Research project: A. Conception, B. Organization, C. Execution; (2) Statistical Analysis: A. Design, B. Execution, C. Review and Critique; (3) Manuscript: A. Writing of the First Draft, B. Review and Critique.
J.H.K.: contributed to the conception, design, and drafting text of the study.
Y.C.: contributed to the data collection, analysis of data, drafting the text, and preparing the figures.
Open Research
Data Availability Statement
The data from this study will be made available upon request.