Lily fragrance may bring a rich sweetness to mind. But some varieties have only a very faint scent, while others suggest coolness or fruit.
A 2019 study compared volatile compounds in 41 cultivars and organized them into six aroma types. Investigating the genes for enzymes that make aroma compounds also revealed that faint-scented lilies possess enzymes capable of producing them. The study explored both what creates the scent and how those substances are made.
Classifying aromas and finding their makers
Lilies include Asiatic, Oriental, and Trumpet groups, along with various cultivars produced by crossing them. Breeding often emphasizes flower color and shape, but scent is another important trait.
The researchers aimed to analyze aroma compounds across a wide range of cultivars, establish a basis for classification, and examine the diversity and function of genes that produce a family of aroma compounds called monoterpenes.
Monoterpene synthases are enzymes that convert the starting material geranyl diphosphate (GPP) into various aroma molecules. The study investigated cultivar differences in the sequences and expression of the TPS genes encoding these enzymes.
Forty-six compounds from 41 cultivars, grouped into six aroma types
For 41 lily cultivars grown at Shanxi Agricultural University in China, fully open flowers were collected in the morning. Flower samples weighing 1.5 g were placed in sealed containers, and compounds released into the headspace were collected at 50°C for 15 minutes and analyzed by gas chromatography–mass spectrometry. Each cultivar was measured in three replicates.
In total, 46 volatile compounds were identified, including 16 that the paper reports as being identified in lilies for the first time. Methyl benzoate was detected in every cultivar, while ocimene, myrcene, and other compounds were also widespread.
The classification used the proportions of major compounds together with odor descriptions from the literature. This produced the following six groups.
- Faint-scented: Fourteen cultivars, many from the Asiatic group and related groups. Although their compounds could be detected analytically, their fragrance was very weak to the human nose.
- Cool: Four cultivars, including Robert Swanson. Their high proportion of eucalyptol gave them a cooling impression.
- Fruity: Four cultivars, including White Heaven. Methyl benzoate predominated, with a light, fruity scent.
- Musky: Seven cultivars, including Pink Perfection. Eucalyptol and methyl benzoate were abundant, and the paper described their aroma as musky.
- Fruity-honey: Nine cultivars, including Triumphator. Other esters and terpenes accompanied methyl benzoate, producing an impression of sweet fruit and honey.
- Lily: Three cultivars: Sorbonne, Brasilla, and Miss Feya. Abundant ocimenes and methyl benzoate gave them a strong fragrance reminiscent of Oriental cut lilies.
Here, “musky” names an odor quality; it does not mean that musk fragrance molecules were detected. Nor was the classification based on a large consumer panel ranking its preferences. It was an attempt to organize scent differences using chemical composition.
Multiple forms in the genetic information, too
The study’s other major component was an analysis of TPS genes across 66 cultivars. By examining complementary DNA, or cDNA, copied from RNA expressed in flowers, the researchers obtained 32 candidate sequences. These fell into three broad types.
Comparing the sequences revealed 64 single-nucleotide differences, together with two regions of differing length: one form had an additional 12 nucleotides, while another lacked 114.
In Sorbonne, which contained all three types, the researchers also examined the underlying genomic DNA. They found that the length differences did not simply reflect insertions or deletions in the DNA itself. Instead, they resulted from differences in where RNA was joined during processing.
This mechanism is called alternative splicing. Even when starting from the same gene, different instructions can result depending on which RNA segments are retained and joined. This diversification mechanism also contributes to the enzymes involved in lily fragrance.
Enzymes from faint-scented cultivars could also make aroma compounds
The researchers tested not only the sequences but also whether the enzymes functioned. They selected five cDNAs: three from strongly scented Sorbonne and two from faint-scented Red Life. Each was introduced into Escherichia coli to produce its corresponding enzyme.
When supplied with GPP, all five produced (E)-β-ocimene, α-pinene, and limonene. A single enzyme could generate multiple products, and the enzymes from the faint-scented cultivar also retained synthetic activity.
This makes it difficult to explain faint fragrance solely by a complete lack of enzyme function. The researchers suggested that future work examine precursor supply, the level of gene activity, and the mechanisms that transport compounds made within cells to the outside.
However, the composition produced in E. coli did not fully match that of the flowers themselves. Activity under the experimental conditions was low, and treatment of sequences that direct enzymes to particular cellular locations may also have affected the measurements. Understanding how flower fragrance develops requires connecting single-enzyme tests to the whole plant.
Associations that may offer breeding leads
Comparing cultivars analyzed for both aroma compounds and gene sequences revealed associations between sequence differences and myrcene abundance. The position with the 12-nucleotide length variation was reported to be strongly associated with emissions of major monoterpenes, particularly (E)-β-ocimene.
These findings offer candidates for selecting cultivars with different aromas and investigating gene function more closely. However, direct tests are still needed to establish whether changing a particular sequence causes a desired scent, or what effects individual RNA forms have.
Compound abundance is also not simply proportional to perceived fragrance intensity. Odor detection thresholds differ among compounds, and their combinations matter. Because samples were collected in containers at 50°C, the measurements do not directly represent concentrations around flowers displayed indoors.
The study’s classification is therefore better viewed as a foundation for surveying cultivar diversity and treating fragrance as a breeding trait than as a finished standard for uniformly judging flowers in shops.
Summary
Based on their major volatile compounds, 41 lily cultivars were organized into six groups: faint-scented, cool, fruity, musky, fruity-honey, and lily. Genetic analysis of 66 cultivars additionally revealed diversity in TPS sequences and a mechanism for producing different RNAs from the same gene.
Another important result was that enzymes from both strongly and faintly scented cultivars could produce the same three monoterpenes. Lily scent differences may be better understood by examining a system that includes precursor supply and release, as well as the enzymes that make the compounds.
The research received support from natural science funding and agricultural key projects in Shanxi Province, China. The authors declared no conflicts of interest.
Reference
Du F, et al. (2019). Volatile composition and classification of Lilium flower aroma types and identification, polymorphisms, and alternative splicing of their monoterpene synthase genes. Horticulture Research, 6, 110. Original paper and open full text
