Towards physiologically relevant human pluripotent stem cell (hPSC) models of Parkinson's disease

Abstract

The derivation of human embryonic stem cells followed by the discovery of induced pluripotent stem cells and leaps in genome editing approaches have continuously fueled enthusiasm for the development of new models of neurodegenerative diseases such as Parkinson's disease (PD). PD is characterized by the relative selective loss of dopaminergic neurons (DNs) in specific areas of substantia nigra pars compacta (SNpc). While degeneration in late stages can be widespread, there is stereotypic early degeneration of these uniquely vulnerable neurons. Various causes of selective vulnerability have been investigated but much remains unclear. Most studies have sought to identify cell autonomous properties of the most vulnerable neurons. However, recent findings from genetic studies and model systems have added to our understanding of non-cell autonomous contributions including regional-specific neuro-immune interactions with astrocytes, resident or damage-activated microglia, neuro-glia cell metabolic interactions, involvement of endothelial cells, and damage to the vascular system. All of these contribute to specific vulnerability and, along with aging and environmental factors, might be integrated in a complex stressor-threshold model of neurodegeneration. In this forward-looking review, we synthesize recent advances in the field of PD modeling using human pluripotent stem cells, with an emphasis on organoid and complex co-culture models of the nigrostriatal niche, with emerging CRISPR applications to edit or perturb expression of causal PD genes and associated risk factors, such as GBA, to understand the impact of these genes on relevant phenotypes.

Keywords: CRISPR; GBA; Human pluripotent stem cells (hPSCs); Neurodegenerative disease modeling; Parkinson’s disease.

Fig. 1

Modeling the nigrostriatal niche. Schematic representation of nigrostriatal multilineage co-culture models. a Human pluripotent stem cells (hPSC) undergo specific protocols of differentiation and generate region-patterned organoids (midbrain in yellow; forebrain in blue), as well as microglia (green). b Organoids dissociation allows for the isolation of neuronal populations and patterned astrocytes for further analysis. Dopaminergic neurons rise from the midbrain organoid (DNs, yellow), medium spiny neurons from the forebrain organoid (MSNs, blue), and region-patterned astrocytes from either one (purple). c hPSC-derived components are engineered into co-culture models. Midbrain and forebrain organoids are assembled and integrated with microglia to generate a multilineage assembloids model, which allows for neuronal circuit formation and neuro-immune communication in a 3D environment. Dissociated components are combined in 2D co-culture, a model in which each component can undergo differential genomic editing

Fig. 2

Pinpoint disease-relevant genes and cell types. a Samples collected from unaffected controls and patients are analyzed through genome-wide association studies (GWAS). Genetic variants determination and network analysis reveals genetic candidates. b To validate candidates’ effects, genome editing is carried out in disease-relevant cell types with isogenic background, as vulnerable and resilient population of neurons and glial cells. c Phenotypes are characterized and compared, in order to link functional consequences to specific genes and cell types. OMICs techniques allow to link candidate editing to a differential expression set of genes, overall generating additional candidates. Comparisons among different cell types elucidate genes responsible for cell-type-specific response and cell population vulnerability. Convergent phenotypes of differential genes pinpoint to pathology-relevant pathways. Multiple gene perturbations are useful to detect phenotype modifiers, that can either alleviate, therefore be considered potential therapeutic targets, as well as aggravate phenotypes

Fig. 3

Modeling GBA biology in the nigrostriatal niche. Schematic representation of co-culture modeling for GBA study. The model presented can be used to determine cell-type-specific phenotypes associated with GBA misexpression. Astrocyte (purple) and microglia (green) strictly control neuronal homeostasis. Co-culture models allow for the assessment of neuroimmune interactions, metabolic support, and inflammatory response (left panel) as well as phagocytic and aggregates-clearance function (right panel). The model opens the possibility to perform the study in a physiologically relevant environment and evaluate how GCase dysfunction in a specific cell type impacts other components of the model