The concept of fitness landscape is a commonly used metaphor in evolutionary biology: evolving populations are represented as sets of points that navigate a genotype space in search of peaks of high fitness. Although the concept has been used for decades, it has long been unclear what an actual fitness landscape looks like. Recent advances in synthetic biology and next-generation sequencing allow experimental investigation of fitness landscapes, and open the way to answering long-standing, fundamental questions in molecular evolution: Which mutations influence function? How do mutations influence function? How do effects of mutations depend on environmental conditions and genetic background?

We are studying these questions using yeast U3 snoRNA as a model system. U3 is an abundant, evolutionarily conserved noncoding RNA, which plays an essential role in ribosome biogenesis. By measuring the effects of 60,000 mutated variants of U3 on yeast growth, we found that the effects of individual mutations were correlated with evolutionary conservation and structural stability. Many mutations had no measurable effect in an otherwise wild-type background, but were deleterious in combination with additional mutations in U3. We also found pairs of compensatory mutations, and used these to predict the secondary structure of the RNA. I will discuss our findings in the context of the recent cryo-electron microscopy structural studies of yeast U3 ribonucleoprotein.