Advanced Complexity and Short-Term Plasticity of Neural Activity in Reciprocally Connected Human Cerebral Organoids

An inter-regional cortical tract is one of the most fundamental architectural motifs that integrates neural circuits to orchestrate and generate complex functions of the human brain. To understand the mechanistic significance of inter-regional projections on development of neural circuits, we investigated an in vitro neural tissue model for inter-regional connections, in which two cerebral organoids are connected with a bundle of reciprocally extended axons. The connected organoids produced more complex and intense oscillatory activity than conventional or directly fused cerebral organoids, suggesting the inter-organoid axonal connections enhance and support the complex network activity. In addition, optogenetic stimulation of the inter-organoid axon bundles could entrain the activity of the organoids and induce robust plasticity of the macroscopic circuit. These results demonstrated that the projection axons could serve as a structural hub that boosts functionality of the organoid-circuits. This model could contribute to further investigation on development and functions of macroscopic neuronal circuits in vitro. To access the functional role of the inter-organoid connection through an axon bundle, we compared the ‘connected’ organoids with that of conventional cerebral organoids (‘single’ organoids). The connected organoids contained approximately twice the number of cells of a single organoid, which could affect the neuronal characteristics. We thus also assessed the tissues generated by directly fusing two cerebral organoids (‘fused’ organoids) as an additional control. To fuse organoids, two organoids were cultured together in a low-adhesion surface culture vessel to induce their spontaneous fusion. We assessed how the inter-organoid connections altered the gene expression profiles in the organoids. Single, fused, and connected organoid at 7 weeks of culture on PDMS-MEA chip were first extracted from PDMS device in warmed maintenance medium and centrifuged at 100 xg for 30 seconds. To get single-cell suspensions, the organoids were dissociated by AccuMax for 10-30 min at 37°C and then centrifuged 200 xg for 5 min. Then, the pelleted cells were re-suspended in DMEM containing 10% FBS to be subjected to 10x Genomics Chromium single-cell RNA-seq library preparation, according to the manufacturer’s protocol. Finally, the library was sequencing with 150-bp paired-end reads on DNBSEQ. The sequencing data were processed using Cell Ranger analysis pipeline v3 with default parameters. Reads were aligned to human reference genome (GRCh38). Cell Ranger output “filtered gene-barcoded” count matrix was loaded into Scanpy (Wolf et al., 2018) with other python packages (scanpy==1.8.1 anndata==0.7.6 umap==0.5.1 numpy==1.19.5 scipy==1.4.1 pandas==1.1.5 scikit-learn==0.22.2.post1 statsmodels==0.10.2 python-igraph==0.9.6 pynndescent==0.5.4) for downstream analysis. Poor-quality cells were excluded based on the following criterion: min_genes > 200 and min_cells < 3, mitochondrial gene percentage < 10%, and nFeatures <4000. Cells with percentage of hemoglobin reads > 5% were discarded. In total, 17,636 cells remained for subsequent analysis. Batch effect correction was performed.