MULTI-OMICS ANALYSIS OF DIFFERENTIATING DOPAMINERGIC NEURONS IDENTIFIES NOVEL FACTORS NECESSARY FOR LINEAGE SPECIFICATION

Midbrain dopaminergic neurons (mDANs) control voluntary movement, cognition, and reward behavior and are implicated in human diseases such as Parkinson’s disease (PD). Many transcription factors (TFs) controlling human mDAN differentiation have been described but much of the regulatory landscape remains undefined. The location and the low number of these cells in the brain have limited the application of epigenomic assays, as they usually require a high number of cells. Thanks to the emergence of induced pluripotent stem cell (iPSC) technology, differentiation protocols for the derivation of mDANs were developed, making access to this neuronal subtype easier, facilitating its study. However, current protocols for the differentiation of human iPSC towards mDANs produce a mixture of developmentally immature and incompletely specified cells together with more physiological cells. Differentiation protocols are based on the developmental knowledge generated from animal studies and the translation of this knowledge to humans appears not to be completely compatible. Therefore, a better understanding of human development is needed, encouraging the use of human-based models. A proper understanding of the epigenetic landscape of human mDAN differentiation will have direct implications for uncovering gene regulatory mechanisms, disease-associated variants (as most of them are in the non-coding regions of the genome), and cell identity.
In this study, a human tyrosine hydroxylase (TH) reporter line of iPSC was used for the generation of time series transcriptomic and epigenomic profiles from differentiating mDANs. TH is the rate-limiting enzyme for dopamine production and therefore a specific marker for mDANs. In the reporter line, mCherry was expressed under the control of the TH promoter, which allowed to isolate mDANs from the cultures by FACS. Integration of time-point-specific chromatin accessibility and associated TF binding motifs with paired transcriptome profiles across 50 days of differentiation was performed using an adapted version of the EPIC-DREM pipeline. Time-point-specific gene regulatory interactions were obtained and served to identify putative key TFs controlling mDAN differentiation. Low-input ChIP-seq for histone H3 lysine 27 acetylation (H3K27ac) was performed to identify and prioritize key TFs controlled by super-enhancer regions. LBX1, NHLH1, and NR2F1/2 were found to be necessary for mDAN differentiation. Overexpression of either LBX1 or NHLH1 was also able to increase mDAN numbers. LBX1 was found to regulate cholesterol biosynthesis and translation possibly via mTOR signaling. NHLH1 was found to be necessary for the induction of miR-124, a potent neurogenic microRNA. Interestingly, miR-124 and NHLH1 appear to be part of a positive feedback loop. Thus, the results from this study provide novel insights into the regulatory landscape of human mDAN differentiation.
In addition, as the identified candidates from EPIC-DREM did not show selective expression in mDANs, the data produced was further explored for the identification of novel expression selective TFs in these cells. ZFHX4 was selected as a relevant TF for mDANs that was also downregulated in PD patients. It presented a high and specific expression during development and in adult mDANs from human brains. Depletion of ZFHX4 during differentiation affected mDAN neurogenesis. However, CRISPR-mediated overexpression of ZFHX4 during differentiation did not affect mDAN numbers. Transcriptomic analysis revealed a role of ZFHX4 in controlling cell cycle and cell division on mDANs. ZFHX4 seems to be regulating cell cycle control by interaction with E2F TFs and the NuRD complex, as these proteins have also been associated with this function and appeared in the analysis performed.
Overall, the present study provides a novel profile of mDANs during differentiation that can be used for many other applications apart from the one presented here, like the identification of disease-associated variants affecting these neurons. Incorporating epigenetic information into the current transcriptomic knowledge increased the understanding of this neuronal subtype and uncovered important pathways involved in the biology of these cells and most probably with implications to disease.