Efficient and Scalable Generation of Human Ventral Midbrain Astrocytes from Human-Induced Pluripotent Stem Cells

Parkinson's disease affects 2%-3% of people over 65 years of age, making it the most prevalent neurodegenerative movement disorder¹. It is caused by degeneration of ventral midbrain dopamine neurons within the substantia nigra, resulting in debilitating motor symptoms, as well as frequent cognitive and psychiatric issues². Parkinson's pathology is typified by aggregates of the protein, α-synuclein, which are toxic to neurons and result in their dysfunction and death^1,2,3. As the dopaminergic neurons are the degenerating population in Parkinson's, they were historically the focus of research. However, it is apparent that another cell type in the brain, the astrocytes, also demonstrate abnormalities in Parkinson's, and are believed to contribute to degeneration in models of Parkinson's^4,5,6,7.

Astrocytes are a heterogenous cell population that can transform both physically and functionally as required. They support neuronal function and health via a plethora of mechanisms, including the modulation of neuronal signaling, shaping of synaptic architecture, and trophic support of neuronal populations via secretion of specific factors^6,8,9,10. However, astrocytes also have a substantial immunomodulatory role, integral to the development and propagation of neuroinflammation^10,11. Neuroinflammation is observed in the brains of Parkinson’s patients, and significantly has recently been shown to pre-empt the onset of Parkinson's symptoms^12,13,14,15, thereby taking the center stage in Parkinson's research.

At a cellular level, astrocytes are said to become reactive in response to injury, infection, or disease, as an attempt to facilitate neuroprotection^9,6,10,16. Reactivity describes a shift in astrocyte phenotype characterized by changes in gene expression, secretome, morphology, and mechanisms of clearance of cell debris and toxic byproducts^9,10,11,17. This reactive shift occurs in response to inductive signals from microglia, which are the immune cells of the CNS and the first responders to injury and disease⁹. Both astrocytes and microglia respond to inflammatory signals by moderating their own function and can transduce inflammatory signals and thus directly influence neuroinflammation^9,10. However, the chronic nature of Parkinson's results in a transition where reactive astrocytes become toxic to neurons, and themselves promote degeneration and disease pathology^6,9,10,18,19. Significantly it was recently demonstrated that blocking the transformation of astrocytes into the reactive neurotoxic phenotype prevents the progression of Parkinson's in animal models¹¹. Astrocyte reactivity in the paradigm of neuroinflammation has therefore become a major focus of Parkinson's research, and similarly relates to a wide spectrum of diseases of the CNS. Together these findings build a picture of significant astrocytic involvement in the etiology of Parkinson's, emphasizing the need for accurate research models that recapitulate the phenotype of the human astrocyte populations that are involved in Parkinson's.

In the embryonic brain, neurons appear first, with the astroglial lineage, namely, the astrocytes and oligodendrocytes, appearing later in development⁶. In vivo and in vitro studies have highlighted a number of signaling pathways that appear to control the potency of neural progenitor cells from neuronal to astroglial derivatives. In particular, JAK/STAT, EGF, and BMP signaling play roles in the proliferation, differentiation, and maturation of astroglia^20,21. These pathways have been the focus of in vitro protocols for the generation of astrocytes from pluripotent cells, including hiPSC^6,22,23. There have been many successful examples of generating astrocytes from hiPSCs^6,24,25. However, it is apparent that in vivo astrocytes in the CNS possess specific regional identities, which relate directly to their function, in accordance with the specific requirements of those astrocytes in relation to their specialized neuronal neighbors^17,24,25,26. For example, relating specifically to the ventral midbrain, it has been demonstrated that astrocytes in this region express specific sets of proteins, including receptors for dopamine enabling communication with the local population of midbrain dopamine neurons²⁶. Furthermore, ventral midbrain astrocytes demonstrate unique signaling properties²⁶. Therefore, to study the role of ventral midbrain astrocytes in Parkinson's, we require an in vitro model that reflects their unique set of characteristics.

To address this, we have developed a protocol to generate ventral midbrain astrocytes (vmAstros) from hiPSCs. The resulting vmAstros exhibit characteristics of their in vivo ventral midbrain counterparts such as expression of specific proteins, as well as immunomodulatory functions. The results presented are from the differentiation of the NAS2 and AST23 hiPSC lines, which were derived and gifted to us by Dr. Tilo Kunath²⁷. NAS2 was generated from a healthy control subject whereas AST23 is derived from a Parkinson's patient carrying a triplication in the locus encoding α-Synuclein (SNCA). These hiPSC lines have been previously characterized and used in a number of published research papers, including for the generation of various neural cell types^{27,28,29,30,31}.