Some wisterias have flowers with a fragrance reminiscent of tropical fruit. A 2026 study investigated the American wisteria cultivar ‘Amethyst Falls’. A sulfur compound with a strong tropical-fruity character was one of the major elements contributing to its green floral scent.
The compound was 4-mercapto-4-methyl-2-pentanone, abbreviated 4MMP. The study also identified linalool and eugenol as major contributors, linking three different odor characters to the flower’s distinctive fragrance.
Investigating a fragrance unlike familiar wisteria
The floral scent components of familiar Japanese wisterias, such as noda-fuji and yama-fuji, have been studied. Knowledge of American wisteria flowers, however, was limited, and this study asked what creates the distinctive fragrance of ‘Amethyst Falls’.
The researchers studied a single plant growing in a field at Ashikaga Flower Park in mid-May 2025, when many of its flower clusters were approximately 80–90% in bloom. One author described the scent as green floral, with a character reminiscent of fruity sulfur compounds.
Branches at a similar flowering stage were brought back for assessment by four people with more than ten years of fragrance-development experience. They perceived a green floral scent with citrus and tropical impressions. The researchers then connected these sensory characteristics with chemical analysis.
Two collection methods to find trace components
The researchers used two methods to collect components from the air around the flowers. One placed a sorptive device called a Twister in a small enclosure around a flower cluster for 60 minutes. The other, dynamic headspace sampling, enclosed two flower clusters in a bag and circulated air through another sorbent for 140 minutes.
The collected components were analyzed by gas chromatography–mass spectrometry. Two authors also performed GC-O analysis on the sample collected with circulating air, smelling the components as the instrument separated them. Using two columns with different properties, they compared odor characteristics and elution positions with reference standards.
This approach can establish not only whether a component is present, but also which components have distinctive odors detectable by the nose.
Three characteristics: tropical fruit, citrus, and spice
Three components stood out in the olfactory analysis:
- 4MMP: A very strong tropical-fruity character.
- Linalool: A citrus-floral character.
- Eugenol: A spicy-floral character.
Linalool and eugenol were also major constituents in the mass-spectrometric analysis of the circulating-air sample. Linalool has two mirror-image forms, but only the S form was detected in this sample.
Although 4MMP was smelled strongly during olfactory analysis, it could not be confirmed by mass spectrometry in that sample. The researchers therefore examined more closely the Twister sample with the larger overall analytical signal and found signals characteristic of 4MMP at the expected position. A reference standard was also collected and analyzed using the same procedure to confirm the match.
This is an example of finding an important trace component by changing the collection method and following sensory clues, even when one analysis fails to detect it.
Sulfur-containing molecules can also smell fruity
The phrase “sulfur compound” may bring unpleasant odors to mind. Yet the odor perceived at the position corresponding to 4MMP in this study was strongly tropical and fruity. The name of an element alone does not determine how a molecule will smell.
Citing earlier literature, the authors explain that 4MMP is also known as a trace aroma component contributing to the character of foods. The present finding, however, is that 4MMP occurs in this cultivar’s flowers and contributes to their odor. It does not mean that the overall fragrances of the flowers and fruits are identical.
The four experienced assessors also compared their remembered impression of the flowers with serially diluted 4MMP solutions and estimated that the perceived intensity was close to that of solutions at 1–10 ppm. Those numbers are the concentrations of solutions used for sensory comparison, not measured concentrations in the air around the flowers, and cannot be treated as emission quantities.
Abundant components are not necessarily the main odor contributors
The analysis also detected substantial amounts of other components, including 4-(2-propenyl)phenol. However, their odors did not reach noteworthy intensity in olfactory analysis, suggesting limited contributions to the flowers’ distinctive character. This contrasts with 4MMP, which could be difficult for the instrument to detect yet conspicuous to the nose.
The identification of major contributors was based on matching separated odors with reference standards. The researchers did not reconstruct the floral scent by mixing the three components or test differences after omitting individual components. Because the study examined one plant at one flowering period, further work is needed to establish how broadly the results apply to other plants or American wisteria cultivars.
Summary
In American wisteria ‘Amethyst Falls’, tropical-fruity 4MMP, citrus-floral linalool, and spicy-floral eugenol were identified as major elements of the floral fragrance. The presence of 4MMP, in particular, was confirmed by combining clues from olfactory analysis with chemical analysis using a different collection method.
A flower’s character is not shaped solely by components that produce large analytical signals. Tracking distinctive trace molecules with both the human nose and instruments reveals an unexpected aspect of fragrance among wisterias.
At publication, two authors were affiliated with Shiono Koryo and one with Ashikaga Flower Resort. The authors declared no matters related to conflicts of interest requiring disclosure. No separate research-funding declaration section was identified.
Reference
Endo H, Okada A, Muraoka N. (2026). Identification of odor-active compounds in the flower of American wisteria (Wisteria frutescens ‘Amethyst Falls’). Journal of Japan Association on Odor Environment, 57(2), 125–128. doi:10.2171/jao.2025.12 (Original article’s public page)
