Miguel Ramirez is a PhD candidate in the laboratory of Dr. Dan Goldowitz at the Centre for Molecular Medicine and Therapeutics at BC Children’s Hospital. Miguel’s research uses a combination of bioinformatic and molecular biology approaches to examine changes in enhancer activity throughout development of the cerebellum. We sat down with Miguel to discuss the first steps of his project, and why understanding regulation of cerebellar development is so important.

Since starting as a graduate student in the Goldowitz lab, you’ve been studying development of the cerebellum. Why were you interested in this particular area of the brain?

Good question! I’ll start with some background. The cerebellum is located just above the hindbrain and below the midbrain. Its most well known function is motor coordination in response to environmental stimuli. Many of the processes that occur within the cerebellum have been well described, and it has a relatively small number of cell types compared to other regions of the brain, making it a good model system for brain development.

Although the anatomical and histological aspects of the cerebellum are well described, the genetics and regulation of cellular differentiation in its development are relatively unknown. I’m interested in discovering which areas of the genome are active at different points during cerebellar development, and to do this, I’m looking at enhancers. These genetic elements are tissue specific, cell type specific, and temporally specific, which makes them perfect for studying the genetics of cerebellar development. Development is a long process, so it would be interesting to determine which parts of the genome are important for regulating early, middle, and late development all the way into the postnatal ages.

How do you plan to identify active enhancers throughout cerebellar development?

Fortunately there are a few molecular properties that have been associated with enhancer elements. The most well known properties are DNA hypersensitivity due to an open chromatin conformation, and specific histone modifications, such as H3K27ac and H3K4me1. A less well known property of enhancers are enhancer RNAs (or eRNAs). These are bidirectional transcripts derived from enhancer sequences, and they are known as a signal of active enhancers. I therefore want to use eRNAs as a marker of active enhancers throughout development.

How did you go about identifying eRNAs?

Before I started the project, the Goldowitz lab invested in describing the cerebellar transcriptome throughout 12 days of development in collaboration with the FANTOM5 (Functional Annotation of the Mammalian Genome) research consortium. The consortium developed a way to capture transcripts based on the presence of a 5’ cap, called Cap Analysis of Gene Expression (CAGE). They then mapped those sequences back to the genome, and quantified them throughout time.

To specifically identify eRNAs, they eliminated transcripts that mapped to transcription start sites of known genes or non-coding RNAs. Then they looked for balanced bidirectional transcription to specifically identify the eRNAs. Once they had that dataset, they compared it to datasets for H3K27ac, H3K4me1, and DNA hypersensitivity to validate the method.

How many enhancers were identified by the consortium using this method?

They identified a catalogue of approximately 45,000 potential enhancers. We then created a bioinformatic pipeline to filter out any transcripts not being expressed with good enough quality in our own samples. That brought us down to around 10,000 possible enhancers that are being transcribed. We also took advantage of the fact that enhancer usage is typically tissue specific. The FANTOM5 consortium gives us access to expression levels not only in the cerebellum, but other tissues as well, which allowed us to narrow down our dataset even further to those enhancers that are only active in the cerebellum. That brought the number down to around 2000 enhancers.

Now that you have this dataset, what are the next steps?

Similar to how FANTOM5 validated their datasets by comparing them to those of other enhancer markers, we now have to do the same. I’m currently doing ChIP-seq for histone modifications (H3K4me1 and H3K27ac), and trying to optimize my next experiments. I’m going to need to determine whether our putative eRNAs can be detected using qPCR. Then I want to validate the enhancer activity itself. The gold standard for this is to clone the sequence into a reporter plasmid downstream of a reporter gene under the control of a promoter that isn’t typically used in the system you’re studying. If you introduce that into the genome and you see reporter activity, it indicates that there’s enhancer activity from the sequence itself.

We can also utilize our time course to do a co-expression analysis of the eRNAs and the genes the enhancers are working on. The logic being that if eRNAs upregulate enhancer transcription, then throughout time their expression should be correlated with those genes. Doing this will give us insight into the developmental processes that these enhancers are influencing.

What are the big picture goals in aiming to understand development of the cerebellum at the genetic level?

The overall big picture goal is to further the annotation of the genome. This is especially important when you consider the heredity of complex neurological diseases like schizophrenia, which have typically been associated with non-coding genetic regions. Once the relationship between loci and function has been identified, it gives us perspective on what’s happening neurodevelopmentally, and how that might lead to neurodevelopmental disorders.

Do you have any advice for undergraduate students thinking about pursuing a PhD in science?

Take your time in choosing a lab for grad school, and try to get some experience working in the lab before you commit to it. You may be interested in a particular lab because of the research, but you need to consider whether or not you enjoy the types of experiments that you’ll be performing and the people you work with, including the supervisor. It contributes to the overall experience. Graduate school is like having a job. It’s not just school. And you’re going to be there every day for at least 5 years! If you’re not comfortable with a particular lab, you probably shouldn’t work in it for graduate school.

Thank you for taking the time to discuss your research, Miguel! We wish you the best of luck as you continue on in your studies!