Poster L12 - A bioinformatics analysis of human thalamic-like organoids and their comparison to the developing human and mouse brains

Introduction: Methods to replicate structures of the human brain through the generation of organoids have been explored extensively. In this study, we compared the gene expression data of human Thalamic-like Organoids (hThO) and human Embryonic Stem Cells (hESCs) generated in vitro to that of current human and mouse brain data, that are accessed through the Neuroscience Multi-Omic Data (NeMO) Portal, and both human and mouse brain data being accessed through BrainSpan and the Allen Brain Atlas (ABA). Validating hThO’s similarities to the thalamus and broader brain structures is a vital step in allowing researchers to use these organoids in future studies with full confidence that they truly represent that of in vivo thalamus structure, function, and gene expression. Single-Cell RNA sequencing (scRNA-seq) analysis was performed on both hThO and NeMO portal data and comparisons of these datasets to existing bulk-RNA sequencing data derived from the developing human brain and in-situ hybridized data derived from the developing mouse brain were facilitated through structure similarity mapping to reveal similarities to structures that exist in the human and mouse brain at different stages of development.

Materials and

Methods: Study on data from human thalamic-like organoids from Early (day 34) time point that was accessed through the NCBI GEO portal. Human thalamus from Carnegie Stage (CS) 19 to Gestation Week (GW) 18, human diencephalon from CS12, and human prosephalon from CS 13 were used, with this data being accessed through the NeMO portal. Human and mouse developing brain data from Post Conception Week (PCW) 8-9 and Embryonic Day 11 (E11) to Embryonic Day 13 (E13) accessed through the Allen Brain Atlas and BrainSpan through use of the VoxHunt package. Statistical Packages from R (v4.3.1) programming language were used for the analysis of each of our data. scRNA-seq analysis and integration were done using R package Seurat (v5.0.1), comparison to BrainSpan and ABA data using Voxhunt (v1.0.1) package, data manipulation using Dplyr (v1.1.4), data visualization using ggplot2 (v3.4.4) and cell clustering done using umap (v0.2.10.0). The structures of the developing human brain that exist after PCW 8 & 9 are as follows: NCx- Neo Cortex; HIP – Hippocampus; AMY- Amygdala; GE – Ganglionic Eminence; DTH – Dorsal Thalamus; URL – Upper Rhombic Lip. Larger structures were also used for this comparison, and the structures of the developing mouse brain at E13 are as follows: Pall- Pallium; POTel- Preoptic Telencephalon; D- Diencephalon; PPH – Prepontine Hindbrain; MH- Medullary Hindbrain; Spall – Subpallium; Hyp – Hypothalamus; M – Midbrain; PH – Pontine Hindbrain.

Results, Conclusions, and Discussions: Figure 1 a) shows the comparison of early timepoint data to data from developing human brain structures that exist in post conception weeks 8 and 9, alongside NeMO portal data from the CS12 Diencephalon, CS13 Procephalon and CS19 Thalamus. We see strong similarities between our data and the Dorsal Thalamus, with the strongest existing between Early, CS19 Thalamus and CS13 Procephalon and the Dorsal Thalamus at PCW 9 with similarity scores of 0.5. We see a less prevalent similarity between the Dorsal Thalamus and the CS12 Diencephalon. In addition to this, we see weaker, yet consistent similarities between our data and the Upper Rhombic Lip. Our early data similarities most closely resemble the similarities exhibited by the CS19 Thalamus and the CS13 Procephalon at PCW 8 & 9.
Figure 1 b) shows the same datasets as 1a) and their comparison to the developing mouse brain structures alongside each other. In this comparison we see strong similarities between the Early, CS19 Thalamus and CS13 Procephalon and the substructures of the Diencephalon, being prosomere 1 & 2, with weaker similarities to the Prosomere 3. The trend continues that our Early data and the CS19 Thalamus data have the closest resemblance in their similarities to the developing mouse brain data. Surprisingly, we see the weak similarity between the CS12 Diencephalon data and the Diencephalon. This may be due to the difference in developmental stage between the data, or due to the fact that they come from different species.
In summary, our study suggests that there exhibits a close similarity between human Thalamic Organoids and Diencephalic and Dorsal Thalamic structures, showing consistent similarities when compared alongside in vivo human thalamus data. Further investigations could provide insights into the molecular mechanisms underlying thalamic development and function. We believe that these Bioinformatics Analysis have the potential to understand neurodegenerative diseases and/or neurodegenerative disorders pertaining to the Thalamus.

Acknowledgements (Optional):