Staining and single-cell RNA-seq analysis of cerebral organoids were used to show that the cerebral organoids we grew are models of the cortical section of the developing brain . Single-cell RNA-seq analysis was also used to evaluate the effects of METH treatment . Cerebral organoids contain cortical markers, neural stem cell markers, astrocyte markers, and proliferation markers which are also found in the developing brain, further supporting cerebral organoids as a model for the developing brain . The analysis of the single-cell RNA-seq data found that there were 10,135 cells from control organoids and 10,613 cells from METH treated organoids. These cells were clustered together using a clustering algorithm that groups cells with similar gene expression together in the same cluster. After the clustering algorithm was performed, these cells clustered into 16 distinct clusters, and the cells were plotted using the tSNE statistical method to visualize the clusters that are found . The clustering also shows that all clusters contain similar proportions of cells from both experimental groups, indicating that the data has been corrected for batch effects, a phenomenon in which cells cluster into groups based solely on their experimental condition . Cells should cluster based on their cell type rather than due to batch effects to decrease bias when comparisons between the experimental conditions are made. Furthermore, although we do see heterogeneity in the organoids, when clustered together we find that all organoids are expressed in each cell type,growing indoor cannabis which may suggest that the transcriptomes are conserved between the organoids.Known neural marker genes were used to identify the cell types of the distinct clusters, where clusters were labeled a specific cell type if they had a high expression of the cell type specific neural marker genes . The expression of these neural marker genes indicates that cerebral organoids contain most of the neural cells found in the developing brain.
These results demonstrate the neural progenitor cells, neurons, and glial cellular diversity within cerebral organoid models, thus showing that they contain neural cells found in the brain and may be used as a model for the developing brain.The next step in the analysis pipeline is to address organoid heterogeneity. Cerebral organoids are self-mapping and as a result, there is variability in each organoid. Thus, it is important to make sure that the results we see from comparing the drug treatments are due to the effects of the drugs and not just due to the organoid’s internal variability. In order to address this issue, we confirmed that our cell types are present in all of the organoids and that they represent fairly even cell counts per condition. Here are examples of three of our cell types and the percentage of cells from each organoid for each of the cell types . All of the organoids represent some proportion of our cell types. The percentages of each organoid represented vary and this is likely due to the fact that each organoid had different total cell counts. For example, the METH organoid M1 had a cell count of 811 vs another METH organoid had a cell count of 3602. We can see based on these numbers alone it is not possible to see a completely even representation of the organoids in each cell type because the total cell counts of the organoids themselves vary. Although the variability in our cell counts may still be an issue, when we compare overall gene expression, we compare groups of cells. We compare all the cells that were from METH treated organoids to all the cells that were from control organoids, which are more even cell counts, with 10,135 cells from control organoids and 10,613 cells from METH treated organoids.The changes in gene expression as a result of METH treatment were analyzed after determining the validity of our organoids as a model for the fetal brain. The top up regulated and down regulated genes as a result of METH treatment were found using Seurat integration analysis. There were only 48 down regulated genes, all of which are shown in Figure 3A. There were 89 up regulated genes and the top 50 of genes are shown in Figure 3B and these genes had a higher expression in METH treated organoids. The changes in expression can be clearly seen through the variations in colors, with yellow being up regulated and purple and light purple having a lower about of expression. These genes can be used to provide valuable insights into the pathways that are affected by METH treatment. Alone these genes do not serve much meaning, but pathway analysis can be used to translate these gene lists into biological insight.
Integration analysis comparing control and experimental organoids was used to show which pathways were affected as a result of METH treatment. Pathway analysis of the differentially expressed genes found to be up regulated as a result of METH treatment showed stress pathways were enriched . Furthermore, a total of 216 pathways were found to be up regulated , however, some notable ones include those related to immune response, inflammation, and apoptosis suggesting that there is neuroinflammation within our organoids . The pathways found to be down regulated by METH treatment, or with a lower expression in METH treated organoids, had primarily to do with development and neurogenesis suggesting that neural development is hindered . Next, we wanted to focus on the genes in specific pathways and examine their gene expression between the METH and control sets of data. Since some of the key pathways that were up regulated associated with apoptosis and stress, the next step was to focus on the specific genes found within these pathways and confirm their expression. When examining apoptotic genes, a subset of METH cells has greater expression of those genes in comparison to the control cells. METH is known to induce apoptosis via genes such as NUPR1, so these results are consistent with previous findings and suggest that METH may induce apoptosis in the developing brain as well. Furthermore, stress-related genes were also found to be up regulated as a result of METH treatment . These genes may lead to some of the neurotoxicity that is known to be associated with METH consumption. Furthermore, in the brain, astrocytes are responsible for maintaining homeostasis. The astrocyte clusters are clusters 1, 3, 6, and 7 and they all appear to have higher expression of the apoptotic and stress genes in comparison to the other cell types . Again, the characteristic side effects of METH consumption are present in the organoids, suggesting that METH may also have a detrimental effect on fetal brain development.Based on the expression of these immune response genes, we see the up regulation of immediate early response genes such as JUN, FOS, IER2, and B2M and A2M which are known to play a role in the immune system as they are part of pathways related to the immune system. In addition, we also see the up regulation of cytokines which include CXLCL8 . These results combined may show that the organoids are immunocompetent, and further suggests that METH treatment induces neuroinflammation within the organoids. In the brain, astrocytes have an immune function and can express MHC compatibility and act as a type of immune effector cells40. Again, we see that the astrocyte clusters 1, 3, 6, and 7 tend to have higher expression of these markers, especially clusters 1 and 3 . Since, we saw a greater expression of VIM, NES,indoor cannabis growing and GFAP it is possible that these astrocytes are also reactive, which may be further contributing to the neuroinflammation that may be occurring in our cerebral organoids. This further confirms that our labeling of astrocytes is accurate and our cerebral organoids can express a variety of brain cell types. The next analysis performed was looking at the effects of THC treatment on the cerebral organoids. This dataset consisted of 6 THC organoids and 5 control organoids for a total of 4421 THC organoid cells and 8455 control organoid cells. There is a clear discrepancy in the cell counts and as a result we took a slightly different approach in the analysis.
This was our attempt at bypassing the bias that we found in our datasets based on cell counts and ensure that the results that are not as skewed. After looking at the astrocyte and proliferative neural progenitor pathways up and down regulated by treatment, no major pathways were found to be up regulated. The proliferative neural progenitor pathways did show that after THC treatment, pathways related to gliogenesis were up regulated. When examining neurons, we decided to go further and identify neural sub-types in the brain to see if there were any changes in these sub-types caused by THC treatment. The brain contains a variety of neural sub-types including glutamatergic and GABAergic neurons. Moreover, investigating the effects of THC on the neural sub-types may provide new insights.The neuronal subset was clustered again and the markers for glutamatergic and GABAergic neurons were used to label the subsets 42. One cluster of cells, cluster 0 did not exhibit any of the known neural markers to a large degree. It did however express SOX11+ and MEIS2+ . Therefore, this cluster may potentially be a new neural cell type that has not previously been identified. After examining the enriched pathways determined by the gene lists, we found that glutamatergic neurons had pathways associated to generation of neurons up-regulated in the THC dataset, this was due to the expression of genes such NEUROD2 and NEUROD6 . When examining these genes closely, we can see that they are clearly expressed in THC glutamatergic neurons more heavily . The THC glutamatergic neurons also exhibit CNR1 or cannabinoid receptor 1 more heavily than control glutamatergic neurons which still express the markers GRIA, SNAP, and MAPT . The expression of NEUROD6 has been shown to be correlated with glutamatergic pyramidal fate and neurogenesis during embryogenesis. The overexpression of NEUROD6 in THC glutamatergic neurons may indicate that THC is leading to preferential differentiation into glutamatergic neurons and thereby altering glutamatergic neurons in the developing brain. Moreover, since the SOX2+ MEIS2+ cluster also expresses a similar phenotype to the glutamatergic neurons, it is possible that this cluster will preferentially differentiate to glutamatergic neurons. CB1 which is encoded CNR1 is typically shown to be decreased as a result of chronic THC exposure because THC acts as an antagonist for the receptor44. In our results, we see that CNR1 is up-regulated as a result of THC exposure in especially in glutamatergic neurons. At this time is not clear if overexpression of CNR1 will have a detrimental effect, but our data further suggests that THC may be altering glutamatergic neurons in the developing brain. The METH figures in the results section and the methods section for organoid generation are reprinted and paraphrased from Glial cell diversity and methamphetamine-induced neuroinflammation in human cerebral organoids as it appears in Mol Psychiatry , Dang, Jason; Tiwari, Shashi Kant; Agrawal, Kriti; Hui, Ashley; Qin, Yue; and Rana, Tariq. The thesis author was a primary investigator and author of this paper. The THC figures in the results section and the methods section are currently being prepared for submission for publication of the material. Dang, Jason; Tiwari, Shashi Kant; Agrawal, Kriti; and Rana, Tariq. The thesis author was a primary investigator and author of this paper.Cerebral organoids are a model for the developing brain3 . Single-cell RNA-seq analysis of control and METH treated organoids showed that organoids contain many of the major neural cell types found in the developing brain . We addressed organoid heterogeneity and found at all our organoids were present in each cluster at some percentage . Moreover, after single-cell RNA-seq differential gene expression was performed on the METH treated and control organoids, genes associated with neurogenesis and neuronal development were found to be down regulated in the METH treated organoids . These results suggest that the development of neurons in the brain and thus brain development as a whole may be hindered as a result of METH treatment. Furthermore, genes associated with inflammation, apoptosis, cytokines, and stress pathways were found to be a part of the 216 pathways up-regulated in METH treated organoids .