Dissertation committee:
Φωτεινή Στυλιανοπούλου, Καθηγήτρια, Τμήμα Νοσηλευτικής, ΕΚΠΑ
Ρεβέκκα Μάτσα, Διευθύντρια Ερευνών, Ελληνικό Ινστιτούτο Παστέρ
Αντώνιος Σταματάκης, Αναπληρωτής Καθηγητής, Τμήμα Νοσηλευτικής, ΕΚΠΑ
Σπύρος Ευθυμιόπουλος, Καθηγητής, Τμήμα Βιολογίας, ΕΚΠΑ
Λεωνίδας Στεφανής, Καθηγητής, Ιατρική Σχολή, ΕΚΠΑ
Μαρία Παναγιωτακοπούλου, Καθηγήτρια, Ιατρική Σχολή, ΕΚΠΑ
Δήμητρα Θωμαΐδου, Διευθύντρια Ερευνών, Ελληνικό Ινστιτούτο Παστέρ
Summary:
Parkinson’s disease (PD) is the second most common neurodegenerative disorder characterized by motor and non-motor symptoms arising from loss of striatal-projecting dopaminergic neurons of the substantia nigra pars compacta as well as of other types of neurons throughout the brain. Even though it is still unknown whether dopamine neuron degeneration is an initial disease feature or the inevitable consequence of multiple dysfunctions throughout the brain, it represents a common pathological manifestation in PD and is responsible for many of the clinical symptoms. A major neuropathological hallmark of PD is the presence of intracellular protein aggregates in the cell bodies and neurites of affected neurons, respectively termed Lewy bodies and Lewy neurites, which are mainly composed of α-synuclein (αSyn). This is a small pre-synaptic protein whose physiological function is still under investigation, yet its pathological involvement in PD is widely accepted. αSyn is the major sporadic PD linked gene, whereas point mutations and multiplications of the locus cause an autosomal dominant form of the disease, often characterized by early onset and a generally severe phenotype. The best-studied αSyn mutation is p.A53T (G209A in the SNCA gene), first identified in families of Italian and Greek ancestry. Although the majority of PD cases are sporadic, studies on familial forms that are clinically and neuropathologically similar to sporadic PD have assisted in gaining insights into PD etiopathology.
Although 200 years have elapsed since the disease was first described and despite intensive research efforts towards understanding the disease using animal models and post-mortem human brain tissues, the causes that lead to the appearance and progression of PD remain largely unresolved. Moreover, there is an unmet need for the development of effective therapies since currently available treatments address the symptoms, but do not cure the disease. A major drawback has been the lack of appropriate models simulating efficiently the human disease. As a consequence several therapeutic approaches that appeared promising at a preclinical level, failed to deliver the expected results when tested in clinical trials. Nowadays the advent of cell reprogramming technologies and the generation of induced pluripotent stem cells (iPSCs) have opened up new prospects for understanding PD pathogenesis and progression in a patient-specific setting. These approaches allow the generation of patient-derived disease models by directed differentiation of iPSCs to the desired cell types of the brain and also offer the possibility for drug discovery or repositioning in a human setting.
In recent years several studies have used iPSC-based cellular systems generated from patient cells as a valuable means for modeling PD in vitro. These investigations revealed a number of disease-associated phenotypes, including increased sensitivity to oxidative and nitrosative stress, mitochondrial deficits, and axonal defects and synaptopathy. In a 2017 collaborative study led by Kouroupi et al. at the Laboratory of Cellular and Molecular Neurobiology – Stem Cells of the Hellenic Pasteur Institute, a disease-in-a-dish model for familial PD was developed using induced pluripotent stem cells (iPSCs) from two patients carrying the p.A53T α-synuclein (αSyn) mutation. By directed differentiation, a PD model was generated that displays protein aggregation, compromised neurite outgrowth, axonal neuropathology and synaptic defects [420].
In this work we aimed to answer whether the vulnerability of these p.A53T (designated PD) iPSC-derived neurons is also retained in an in vivo setting after transplantation in the rodent brain. Towards this, we investigated the in vivo phenotypes of iPSC-derived cells from one p.A53T patient in comparison to control iPSCs derived from a healthy donor, after transplantation in a lesion mouse model established by unilateral intrastriatal 6-hydroxydopamine (6-OHDA) injection in the immunosuppressed NOD/SCID strain. To direct the differentiation of control and PD iPSC lines towards the dopaminergic lineage, we applied a floor plate induction protocol that involves the use of a cocktail of small molecules that either inhibit the BMP/ TGF-β/ Activin pathways or activate the sonic hedgehog pathway and WNT signaling followed by neuronal differentiation-promoting factors. At the end of floor plate induction (11 days in vitro; DIV), practically all cells were LMX1A-positive floor plate neuroepithelial cells and about 50% were LMX1A/FOXA2-positive dopaminergic precursors. Engraftable neurons were obtained with this protocol after 25 DIV, and in order to avoid cellular overgrowth, further enrichment in PSA-NCAM-positive neuronal cells was achieved at 28 DIV by isolation on magnetic beads covered with an antibody against PSA-NCAM. Sorted cells comprised primarily of immature neurons (approximately 70%) as assessed by expression of the neuronal lineage markers doublecortin (DCX) and βΙΙΙ-Tubulin while a proportion were nestin-positive neuronal precursors, with no statistically significant differences between control and PD cells.
Because the differentiation protocol used in this work was different from that previously reported in the Kouroupi study [420], before proceeding to in vivo transplantation we addressed the phenotype of PD cultures maintained for longer periods of time in vitro, in terms of cell morphology and functionality that included electrophysiological recordings and calcium imaging. In PD cells analyzed between 45-70 DIV, degeneration signs became apparent. DCX-positive immature PD neurons exhibited aberrant neuritic growth whilst intracellular protein aggregates were detected in both DCX- and tyrosine hydroxylase-positive (TH) dopaminergic PD neurons. Additionally TH-positive PD cells, which formed a dense network at 70 DIV, displayed dystrophic neurites with swollen varicosities that quite often ended up in fragmented processes. Up-regulation of αSyn protein, indicative of pathology, was also noted in PD neurons as compared to controls. In terms of functionality, electrophysiological recordings did not reveal statistically significant differences on active and passive membrane properties between control and PD cells. Interestingly, however, calcium imaging demonstrated a higher frequency of spontaneous calcium transients in PD cells with a significantly larger mean amplitude. As calcium dynamics regulate neurite growth and synaptic connectivity, the observed alterations in calcium signaling should impact on, and explain, the morphological phenotypes of PD neurons.
To investigate the in vivo phenotype of PSA-NCAM-enriched iPSC-derived cells at 30 DIV, we established a 6-hydroxydopamine (6-OHDA) lesioned mouse model in the NOD/SCID strain that supports xenograft survival. A unilateral lesion was induced by intrastriatal injection of 6-OHDA in 9-10 week old mice and the resulting functional deficit was confirmed 2 weeks later using drug-induced and drug-free behavioral analysis. However, functional analysis in longer time periods up to 15 weeks revealed a spontaneous behavioral recovery in the lesioned animals. Although a 60% loss of TH-positive neurons was verified by immunohistochemistry in the substantia nigra which remained stable over 15 weeks, striatal dopaminergic reinnervation was observed in agreement with the concurrent behavioral recovery. This phenomenon could be explained by sprouting of remaining undamaged dopaminergic fibers within the ipsilateral lesioned striatum and/or by a cross-hemispheric compensatory mechanism by which TH fibers originating in the contralateral substantia nigra are induced to project into the ipsilateral striatum.
Next, we proceeded to cell transplantation and immunohistochemical analysis of the graft and the surrounding host environment. To investigate their phenotypic characteristics in vivo, control and PD iPSC-derived PSA-NCAM-enriched cells (30 DIV) were transplanted 3 weeks after 6-OHDA injection and immunohistochemical analysis followed after another 12 weeks. The analysis revealed that despite the disease-related characteristics that PD cells displayed when maintained up to 70 DIV, they could survive and differentiate in vivo over a 12-week period. However, interesting differences were noted between patient-derived and control grafts. First, a significant rise in αSyn immunoreactivity was noted in PD grafted cells indicative of a first step towards pathology. Second, control-derived grafts appeared to integrate better than PD grafts within the host tissue extending projections that formed more contacts with host striatal neurons. Third, significantly more DCX-positive immature neurons were found in PD derived grafts as compared to controls, suggesting that PD neurons are stalled at this immature differentiation state. This observation is likely to account for their compromised ability to extend neurites and form contacts with host neurons.
Overall our data suggest that the distinct disease-related characteristics which p.A53T cells develop in vitro, may be attenuated or take longer to emerge in vivo after transplantation within the mouse brain. Considering the limited numbers of TH-positive neurons present in both control and PD grafts it is desirable to examine in future studies longer time points, exceeding 6 months after transplantation, which is a challenging task given the increased mortality rate of the 6-OHDA lesioned NOD/SCID mice. Long-term studies are also needed to clarify whether the elevation in αSyn immunoreactivity seen in PD grafts - a phenotype that cannot be attributed to their more immature state - would eventually result in a pathological phenotype with formation of protein aggregates in mature neurons. Another issue is whether p.A53T pathology can spread from the graft to the host environment. Cell to cell seeding of αSyn and transmission of pathology from patient to healthy human neurons has been recently observed in vitro. Whether this may also occur in vivo remains to be seen.
To conclude, further analysis of the phenotypes that patient cells acquire over longer periods of time as well as the use of multiple iPSC clones from different patients should extend our current proof-of-concept study and provide additional evidence for in vivo disease modeling.
