Evolution of scent genes in roses

Rosa is a complex taxon with more than 150 intertwined species. Only few have been domesticated by humans since Antiquity, reaching today more than 30,000 cultivars. One of the major traits that have been selected is scent. However, rose scent is a complex trait comprised of dozens of volatile molecules. Some of these molecules originate from a specific and uniquely evolved biosynthetic pathway in the genus Rosa, which arose from acquisition of the duplication and neofunctionalization of genes to become involved in the production of scent compounds. Examples include NUDX1, a gene involved in geraniol biosynthesis, specifically in roses (Magnard et al., 2015). We have shown that multiple trans- and cis-duplications of NUDX1 during the evolution of Rosaceae and Rosa, have led to the specialization of the paralog NUDX1-1a toward geraniol production (Conart et al., 2022). This paralog is not present in the more ancient wild roses making them unsuitable for crosses to obtain fragrant roses. Previously we showed that some hybrid cultivars with R. wichurana as one parent, have a different specialization: NUDX1-1a is inactive, while NUDX1-2c, another paralog, is active and involved in (E,E)-farnesol production (Sun et al., 2020). Furthermore, some genes well-known to be involved in scent production encode enzymes that are functional in vitro, but are not always highly expressed in planta. Examples include LIS and PAAS genes, respectively involved in linalool and 2-phenylethanol biosynthesis (Magnard et al., 2018; Roccia et al., 2019). Taken together, these results indicate different evolutionary scenarios in different rose species. A better understanding of the genes and alleles involved in the production of fragrant molecules is thus needed to help the selection of new, scented rose cultivars. This paper focuses on the NUDX1 gene evolution as an example of what knowledge of a gene family can bring to the breeding of roses.