Johns Hopkins School of Medicine

Transcriptional and post-transcriptional mechanisms of neuronal diversification

Nervous systems perform complex functions that rely on the activities of vastly diverse neuron types. Distinct neuronal morphologies, physiologies and connectivities are largely defined by neuron-type-specific gene expression, which in turn is controlled by specific regulators. Our goals are to uncover the basic regulatory principles that instruct neuronal diversity during development, and to illuminate how neurons have diversified during evolution. 

We have thus far focused on two main areas: 1) the effect of transcriptional histories along development on the terminal identity of a neuron, and 2) the contribution of post-transcriptional regulators (specifically miRNAs) to neuronal specialization. To address these questions, we draw from the molecular biology, genetics and RNA biochemistry toolsets, also developing novel, necessary tools ourselves (e.g. a method to sequence miRNAs from individual cell types within complex tissues). To extract fundamental concepts in cellular differentiation, we use the nematode C. elegans as our primary model system. This is mainly because it allows us to design experiments with unparalleled cellular resolution, but also because the vast body of work amassed by the community working on this animal provides a superb background enabling deep mechanistic understanding. At the same time, the mechanisms that we study are basic molecular mechanisms that also operate in other vertebrates and invertebrates.

1) Temporal combinations of transcription factors as a mechanism for neuronal diversification. Transcription factors (TFs), play indisputable roles in the expression of different terminal gene sets, typically acting in a combinatorial manner. Such combinatorial or intersectional use of TFs is a widespread strategy for specifying distinct cell types. Most cases of combinatorial activity represent spatial intersections, in which two or more TFs are co- expressed and jointly required to activate transcription of a given locus. However, while studying the development of two sensory neurons in C. elegans, we discovered a novel type of combinatorial activity, that we termed temporal intersection (Cochella & Hobert, Cell 2012; Charest et al. Dev. Cell 2020) . In this mechanism, the transient binding of an early TF leaves a “memory” of activation that can then be boosted by a later acting TF. This relies on changes in chromatin states and TF binding abilities that we are beginning to unravel using newly tailored means of manipulation and molecular readouts.

The combinatorial action of temporally separated TFs has two important implications: i) not only the terminally expressed genes, but also the transcriptional history of a cell defines its identity, and ii) because cells are specified through progenitors that go through different transcriptional states, a vast number of TF combinations over time is possible, and may contribute to diversifying cell types during development. We thus propose that temporal intersections, like spatial intersections, broadly contribute to neuronal diversity.

2) MicroRNA contributions to the diversification of neuronal function. MicroRNAs (miRNAs) are enriched in animal brains. Although we know the roles of some individual miRNAs, we lack comprehensive insight into their contribution to neuronal diversity. MiRNAs are fascinating regulators to study in this context, because as repressors, they can not only sharpen spatial and temporal boundaries of gene expression but also, they provide insight into what a cell needs to keep at low levels or off - information we miss by looking only at genes that are expressed in a cell. For example, we found that miR-791 is exclusively expressed in the three pairs of CO₂-sensing neurons of C. elegans, and is required in these cells for worms to respond to CO₂ (Drexel et al., Genes & Dev 2016). Interestingly, miR-791 acts by repressing two otherwise ubiquitous genes, including a carbonic anhydrase. This revealed a unique physiological requirement of these neurons. Another reason why miRNAs are interesting in this context, is that they appear and disappear with fast rates during evolution, and thus have the potential to contribute to the generation of neuronal complexity at this scale.

MiRNA expression profiling in various animals has been performed mainly at the tissue/organ level. We developed an innovative approach that allows us to sequence mature miRNAs from rare cell types, without the need for cell sorting (Alberti et al., Nat Methods 2018). This approach, called mime-seq, utilizes a plant enzyme to methylate miRNAs. Animal miRNAs are not endogenously methylated, so cell-specific expression of this enzyme leads to cell-specific labeling of miRNAs. Subsequently, a methylation-specific cloning protocol enables sequencing of the methylated, cell-specific miRNAs. We used mime-seq and a deconvolution strategy to generate an atlas of miRNA expression for the complete C. elegans nervous system (in progress). We are currently refining this map but our initial analysis revealed that miRNAs are enriched in sensory neurons relative to motor or interneurons. Sensory systems rely on large cell diversity to respond to multiple different stimuli, but also flexibility over evolutionary time to allow adaptation to new niches. MiRNAs are excellent candidates to contribute to sensory-neuron diversification during development and evolution.