The developing fetus or embryo is mostly sheltered from environmental impacts through layers of protection both on the side of the gestational parent and the fetus. However, it is becoming increasingly clear that some environmental factors such as pharmaceuticals, pathogens, and maternal immune system molecules can reach the developing fetal brain. To do so, the molecular mediators need to cross several barriers including the liver of the gestational parent, the placenta, and the developing blood-brain barrier of the fetus.
Upon severe infections, the immune system of the gestational parent induces the so-called maternal immune activation or MIA. MIA has been associated with increased risks for neurodevelopmental disorders, NDDs, in humans, including autism spectrum disorder, schizophrenia, and others (Figure 1a). Our understanding of causal mechanisms underlying this association is still incomplete due to the lack of an appropriate model system. Studies in rodents showed a major role of several cytokines, including interleukin 6 (IL-6), in MIA. However, we do not understand how this translates to the enlarged human brain, which has special adaptations during fetal brain development when neurons are being born, differentiate and mature. Recent advances in stem cell technologies have allowed us to overcome some of these hurdles since we can now recapitulate key features of human brain development in a dish in 3D cell cultures called brain organoids.
Progress over the last years has also allowed to generate brain region-specific organoids. We worked with dorsal forebrain organoids, which recapitulate the development of human neocortex, because neocortex is thought to be the most vulnerable brain region for NDDs. In dorsal forebrain organoids, we investigated the effects of IL-6 on neural progenitor cells and neurons. We first aimed to understand which cell types respond to IL-6. For that, we analyzed the expression of receptor molecules. Two receptor molecules are needed for a cell to respond to IL-6, IL-6 receptor, IL6R, and IL-6 family signal transducer, IL6ST. When the cell expresses both receptors, the cascade is called classic, and when IL6R is provided from neighboring cells, the cascade is called trans (Figure 1b). We found that most cells in the human neocortex do not express IL6R, and therefore cannot respond to IL-6 following the classic mechanism. However, IL6R is expressed in microglia, the immune cells of the brain, and can be released by these cells into the intercellular space. In this case, other cells in the brain can respond to IL-6 signaling through the trans mechanism. To model trans signaling, we used Hyper-IL-6, a chimeric molecule that consists of IL-6 and soluble IL6R, s-IL6R, bound together through a molecular linker.
Upon binding to Hyper-IL-6, IL6ST activates a cascade of events which finally leads to the phosphorylation of the transcription factor STAT3 (Figure 1c). In dorsal forebrain organoids, STAT3 was phosphorylated after 5 days of Hyper-IL-6 treatment (Figure 2a for experimental design). Interestingly, using immunohistochemistry, we showed that STAT3 phosphorylation occurs predominantly in SOX2-positive ventricular radial glia cells (vRG), the primary progenitor cell pool of the neocortex at midgestation (Figure 2b). Additionally, we show that IL6ST is associated with Nestin, the cytosolic marker of radial glia cells, in dorsal forebrain organoids (Figure 2c). Together, these findings demonstrate that the primary cell type responsive to IL-6-dependent signaling in dorsal forebrain organoids is radial glia cells.
Since phosphorylated STAT3 regulates transcription of its target genes, we hypothesized that radial glia cells should have the most altered transcriptional landscape among cell types present in the organoids. To check this hypothesis, we performed single-cell transcriptomic analysis in Vehicle- and Hyper-IL-6-treated organoids. As expected, we found the highest number of differentially expressed genes in radial glia cells. Among genes with a decreased expression upon Hyper-IL-6 treatment, there were genes involved in protein translation. This finding corroborates recent results in a murine model of MIA. There, Here, we show that reduced protein translation may be apparent already in radial glia cells, the precursors of neurons. Among the genes with increased expression in the radial glia cells of Hyper-IL-6-treated organoids, we find genes involved in the innate immune response, which validates our model since these are known transcriptional targets of STAT3.
In murine models of MIA, the differences in the number and localization of different cell types have been previously reported. Therefore, we used several approaches to assess the cell type composition of brain organoids upon Hyper-IL-6 treatment. Using single-cell transcriptomics, we analyzed the distribution of cell types along the differentiation path from progenitor cells to neurons. We found a higher proportion of radial glia cells at the start of this differentiation path in Hyper-IL-6 treated organoids. We validated this finding by performing immunostaining against the progenitor cell marker SOX2. Next, we wanted to understand whether short-term treatment and the ensuing changes in cell type composition can have long-term consequences for organoid development. For that, we treated the organoids with Hyper-IL-6 for 10 days and waited for 35 more days before analyzing the number of different types of neurons. We found that, while the number of cortical deep-layer excitatory neurons (dExN) remained stable, the number of cortical upper-layer excitatory neurons (uExN) was elevated in Hyper-IL-6-treated organoids. Moreover, both deep- and upper-layer neurons were mislocalized in the cortical plate which may be caused by deficiencies in differentiation or migration.
Our most intriguing finding was the activation of NR2F1, a transcription factor with multiple functions in brain development following Hyper-IL-6 treatment. NR2F1 may be involved in synapse organization and cell migration and may thus explain the abnormal laminar positioning of neurons in the cortical plate. This question, however, remains to be explored in further studies.
Taken together, in this study we established dorsal forebrain organoid model of maternal immune activation using Hyper-IL-6 treatment (Figure 3). We found that radial glial cells within dorsal forebrain organoids activate STAT3 upon Hyper-IL-6 treatment. It results in changes on both cellular and molecular levels. On the cellular level, we found an elevated number of ventricular radial glia in the short term. In the long term, we found an elevated number of upper-layer excitatory neurons and mispositioning of both deep- and upper-layer excitatory neurons. On the molecular level, we found transcriptional changes, predominantly in radial glia cells, that corroborated findings in rodent models of MIA. We also found transcriptional changes not reported in previous studies, including elevated NR2F1 expression, which could have species-specific effects and may be especially relevant to link to human epidemiological findings.
In the future, our model system would benefit from adding microglia into brain organoids to assess brain-immune interactions in the context of MIA. Specifically, microglia can release soluble IL6R into the intercellular space and thus allow us to use IL-6 but not Hyper-IL-6 for eliciting trans signaling in radial glia. We further expect that using a similar treatment paradigm in older organoids may be useful to reveal how MIA affects later progenitor cell populations, for example, outer radial glia, which is abundant in humans but almost absent in mice. Next, our study only focused on male organoids, and further research is needed to understand the effects of MIA in female organoids and the differences between sexes in terms of susceptibility and resilience to environmental impacts. Finally, our model may serve as a platform for further investigating the effects of individual molecules involved in MIA as well as gene-environment interactions contributing to the emergence of neurodevelopmental disorders.