Interview: Improving the polar bear transcriptome using long-read sequencing and Cas9 depletion

Date: Thursday 7th October

Time: 4pm GMT

Speaker: Ashley L. Byrne, University of California, Santa Cruz

Ashley is a graduate student researcher working in the Vollmers Lab at University of California, Santa Cruz where she focuses on developing new tools that harness long-read sequencing to identify RNA isoforms and make a better snapshot of the transcriptome. She has developed a new tool to improve transcriptome profiling in the polar bear using Cas9 depletion for abundant transcripts. We caught up with her to discuss her current research, how she became interested in genomics and transcriptomics, and how long-read sequencing is benefiting her work.

Ashley will be presenting her webinar ‘Improving the polar bear transcriptome using long-read sequencing and Cas9 depletion’ with Technology networks on Thursday 7th November, 4pm GMT, 8am PST, 5pm CET.

What are your current research interests?

I’ve just started a postdoc position at Chan Zuckerberg BioHub, where some of my current research interests are applying some of the full-length long-read sequencing methods that I helped developed at UCSC with Chris Vollmers to better define transcriptomes within the tumour microenvironment. My expertise has been in single-cell methods - however, I’m also interested in performing whole transcriptome analysis to not only analyse the complex tumour microenvironment but also try to make better predictions as to what type of treatment regimen patients may need.

I also have a soft spot for immunology and I think we need to start looking outside of humans and mouse models in order to get a full understanding of this complex system. Not only would this allow us to broaden our knowledge of the immunology field but given the circumstances of climate change and human impact on many animals within the animal kingdom, they are struggling to fight off infectious diseases making them more vulnerable to extinction. By looking at the immune system of other organisms, we could potentially start developing vaccines for those that are most vulnerable.

What first ignited your interest in transcriptomics?

I actually have a Masters in Forensics from UC Davis, so I started off in the genetics field where I focused on finding mutations and attributing them to phenotypic traits. However, I really started to get interested in RNA when I came to UCSC, where it’s kind of hard to avoid working in the RNA field. I wanted to understand more about the functional consequences of these mutations - do they alter the alternative splicing landscape, or do they create alternative cis-elements? Although mRNA is what I mainly focus on, I think the whole RNA field is always changing and I hear about a new type of RNA species every couple of years.

Can you tell us more about how long-read sequencing is changing your field?

I think long-read sequencing has really changed how we can look at the transcriptome. Before, we had a vague idea of what RNA isoforms were being made and this was particularly hard to predict with complex RNA isoforms, and it becomes even more challenging for non-model organisms that have a poor genome annotation or lack one altogether. But now, not only can we say this gene is turned on or off, but we can define transcript features furthering our understanding of transcriptional and translational regulation and can start to quantify RNA isoform expression as well. Overall, we can now define transcriptomes much better with little guess work since we can sequence full-length cDNA from end-to-end using long reads.

How has it benefited your work?

When I first joined the Vollmers Lab at UCSC we were interested in B cell genetics, specifically antibody transcripts. Using short-read sequencing for this can be very challenging, as you have to assemble these reads which are fairly complex and each B cell has a unique antibody transcript, so there is no real reference. By using long reads we can not only determine the mRNA landscape but ultimately identify full-length antibody transcripts as well.

What impact could a deeper understanding of the transcriptome from non-model organisms have?

Currently, the genomics field has been expanding rapidly and several tools for creating “centromere-to-centromere” genomes of many non-model organisms have been developed. It is truly amazing how we can assemble these high-quality genomes, but many of the tools used for generating these do not translate well in the transcriptomics field. Some genome annotations are just lifted from another species that is evolutionarily closely related - but these annotations might be incomplete or there may be differences, and thus it is possible that certain information gets missed or even mis-represented altogether. Annotations may also be predicted using ab initio tools using modeling to look for conserved genes - most annotations are based on short-read RNAseq data, but again this fails to identify all transcript features. If we can get accurate full-length cDNA reads, we can potentially eliminate using these tools altogether. If we want a better understanding of the differences within these comparative transcriptome studies, we need a better snapshot of the transcriptome to begin with.

Why the polar bear?

This was just a small start of a much larger project looking at different polar bear populations across the Arctic, collaborating with Megan Supple of the Shapiro lab at UCSC and the Laidre Lab at the University of Washington. They mainly focus on the population dynamics within Arctic animals and their adaptations, given that they live in extreme environments, studying these changes as a result of environmental impact such as climate change. The Shapiro lab has always been keen on using genomics for determining strategies for conservation efforts, and here I’ve just made a better tool to look at the transcriptomes of the polar bears to get a better understanding of these dynamic changes.

What have been the main challenges in your work, and how have you approached them?

There have been many challenges within the transcriptomics field and in the Vollmers lab at UCSC we were always good at thinking, “how can we fix this?”, and so I have spent most of my graduate career developing tools to fix problems. We had struggled getting complete antibody transcripts from single B cells so we went ahead and developed a full-length cDNA RNA sequencing method using long-reads to be able to get these complex transcripts. The next challenge we came across was - how can we get more accurate reads with long-read sequencing? As a result, we developed R2C2, which uses Phi29 to make a circular consensus. This allowed us to increase base accuracy, making it much easier to resolve isoforms without using short-read RNAseq. Finally, we were performing transcriptomic studies for annotating genomes for many different types of animals and I found we were unable to get the throughput needed to create these annotations. This was due to highly abundant transcripts, such as hemoglobin which can comprise a vast majority of my sequencing reads. The best approach was to develop a method that could be applied for long-read sequencing and be easily implemented across different species; thus, I developed the Long-DASH approach using the CRISPR-Cas9 system.

What’s next for your research?

I think one of the things I really would like to work on next is tweaking some of the methods I helped develop at UCSC and broadening it to look at more of the RNA landscape. Most of my work has been focused on mRNA but the RNA world is crazy, with so many different RNA species, including many types of non-coding RNA, and some of the evolutionary differences between organisms can be contributed to the regulatory landscape. Unfortunately, some regulatory RNA species like long-noncoding RNAs (lncRNAs) are poorly characterised compared to protein-coding genes because they are weakly conserved. As a result, lncRNA annotation has been mainly based upon short-read RNAseq data which has its limitations. I really think using a long-read approach will be very useful in identifying certain RNA species, and I would hope to further adapt these methods to better understand the RNA landscape within the context of looking at the tumour microenvironment in patient samples.

Sign up for Ashley's webinar on 7th November, here: ‘Improving the polar bear transcriptome using long-read sequencing and Cas9 depletion’.