17 November 2017 – “The real voyage of discovery consists not in seeking new landscapes, but in having new eyes,” wrote French author Marcel Proust in his novel Remembrance of Things Past. And while Proust was, in fact, describing art, he might just as well have been referring to science. Often, innovative technologies provide researchers with ‘new eyes’ to look deeper into well-studied matters.
High-throughput sequencing played a central role in understanding the richness of the non-coding RNA world. “Until recently, we did not have the methodology to investigate these molecules,” says EMBO Member Jörg Vogel of the University of Würzburg, Germany. But now researchers are uncovering large regulatory networks consisting of non-coding RNAs that had hitherto escaped detection.
This network of sRNAs, microRNAs, siRNAs, piRNAs, lncRNAs and more was the focus of the EMBO | EMBL Symposium “The non-coding genome” that took place in Heidelberg from 13–16 September 2017. The conference, which was co-organized by Jörg Vogel, EMBO Member Elisa Izaurralde, David Bartel and John Rinn, brought together scientists working on all aspects of non-coding RNAs.
Ubiquitous regulators
There are thousands of different non-coding RNAs in each cell. But while they are numerous, their mode of action is usually subtle, and this could be, according to Vogel, the other reason they have been overlooked for so long. “They often lack the strong phenotypes that we see when we knock out, for example, a transcription factor,” he explains. “The way they regulate gene expression is very different, they are involved in fine-tuning.”
The molecular mechanisms for many of the small non-coding RNAs of plants and animals is very similar: They associate with a protein of the Argonaute family to form a silencing complex, and guide it towards a target mRNA. “It is a two-component system where Argonaute provides the molecular function, and the non-coding RNA is responsible for target specificity,” says David Bartel of the Massachusetts Institute of Technology, USA. Moreover, base pairing with the target sequence is usually short, between about 6–20 base pairs. In particular, microRNAs in most animal cells primarily use a short seed sequence of six nucleotides. Accordingly, they each have many targets – about 400 on average and provide a whole additional layer of regulation that affects almost every process in the cell.
Tuning a complex system
The ability of non-coding RNAs to subtly downregulate the expression of multiple genes at once makes them particularly well suited for fine-tuning regulatory transitions. For example, in a recent study Vogel showed that small RNAs play a crucial role during Salmonella infections, helping the pathogen transit from an invasive state to a state of intracellular replication. “These sRNAs modulate response curves through feedback or feed-forward loops. They can induce threshold values for switching on genes or allow genes to be switched off more quickly. Depending on how the feedback loops are constructed, you will get different time curves of gene expression,” says Vogel.
Open questions remain
Much has been learned about non-coding RNAs in the past decade, but there is more to be understood about their exact function, their target genes and their networks. “We do by now have a pretty good idea of how microRNAs work, but other classes have proven to be more challenging,” says Elisa Izaurralde of the Max Planck Institute for Developmental Biology in Tübingen. “piRNAs, for example, have been more difficult to understand and many mechanistic questions remain open. There may be more than one mechanism by which they regulate genes.”
“For some of the RNA classes, an important future challenge will be to separate the signal from noise. Not everything that is transcribed will also have a function,” says Vogel. In addition, the diversity of non-coding RNAs is far from understood. “Research thus far has concentrated on a few model organisms.”
“Start to learn the language”
The greatest challenge, however, is to understand long non-coding RNAs (lncRNAs). There are many thousands of them and they are part of a very heterogeneous group, unified only by two features as their name suggests – they are long and non-coding. “It took about 10 years to do a cartography and to find out which RNAs to study. Now we are ready to investigate how they work,” says John Rinn from the BioFrontiers Institute in Boulder, USA. According to Rinn, an important next step – admittedly, a very large one – is to understand what he calls the syntax of non-coding RNAs – the link between structure and function. “We have a lexicon for the language of proteins, we understand words like ‘Zinc-finger domain’ or ‘kinase domain’. We can predict what they are doing. The RNA language, in contrast, is still hieroglyphic to us. A big task for the future will be to translate these hieroglyphs into meaning,” says Rinn.
High-throughput technologies will play a major role in this endeavor. “We can play with the sequence through an evolutionary approach and change it a little and then see if it is still functions,” says Rinn. He and colleagues have used this approach to identify sequences and structures required for nuclear localization of non-coding RNAs. “The key is to have a functional assay,” he adds. “Once you have that you can start to learn the language”.
Towards clinical applications
For decades, researchers have been using non-coding RNA as a research tool to silence genes in a technology called RNAi. Today, this approach is also being used therapeutically. Although the start was slow – there are major problems in drug delivery and toxicity – there has been some success. Only recently, pharmaceutical companies Alnylam and Sanofi announced positive phase 3 clinical data for their RNAi-based drug patisiran that targets a rare genetic disease known as hereditary ATTR amyloidosis with polyneuropathy.
Aside from treating human cells with RNAi, Jörg Vogel sees another option in targeting bacteria. The advantage of RNAi over classic antibiotics is that RNAi can act in a more refined way. This is a great advantage when it comes to treating dysbiosis, a microbial imbalance that most often affects a person’s digestive tract. “One of our goals is to specifically manipulate or eliminate some bacteria of the microbiome,” says Vogel.
