Eucalyptus Notes in Kiwifruit: Tracing the Genes Behind the Aroma

When you imagine the aroma of kiwifruit, sweet fruit or green, grassy notes may come to mind. It also contains a substance reminiscent of eucalyptus or mint.

A 2020 study used the human nose to investigate the character that one such compound, 1,8-cineole, adds to the aroma, then traced the enzyme and gene responsible for producing it in the fruit. It connected the perception of aroma differences with the mechanism by which plants make those aromas.

Beyond measuring compounds: testing differences that matter to the nose

Zeng and colleagues aimed to use sensory evaluation to identify compounds contributing to kiwifruit’s characteristic aroma and establish the genetic and biochemical basis of their biosynthesis. They also sought indicators that could eventually help breeders take aroma into account.

They examined 1,8-cineole and the sulfur-containing compound dimethyl sulfide (DMS), both thought to contribute to the character of the yellow-fleshed cultivar Hort16A. After testing their sensory roles, the researchers focused on the mechanism of cineole synthesis.

Adding compounds one at a time to an aroma made from four ingredients

Rather than having people eat the fruit itself, the sensory tests used a model mixture of four common kiwifruit aroma compounds: fruity ethyl butanoate and methyl butanoate, and green-leaf-like hexanal and (E)-2-hexenal. The researchers compared this base aroma with versions containing added cineole or DMS.

Fifty-one people recruited in New Zealand smelled two aromas presented in bottles and reported whether they were the same or different. Seventy-three percent distinguished the base aroma from the cineole version, and 96% distinguished it from the DMS version. When two identical base aromas were presented, 78% said they were the same.

A separate group of 45 people selected all applicable aroma descriptions from a list of 14. Adding cineole increased selections of eucalyptus, mint, pine, and floral notes. For example, the estimated frequency of the eucalyptus description rose from 2% for the base aroma to 25% with cineole, and the floral description from 16% to 31%. These were estimates obtained by analyzing responses with a statistical model.

Adding DMS increased the blackcurrant description. Against the same fruity background, small amounts of individual compounds shifted the direction of the aroma.

However, the cineole version was not the one most often described as “kiwifruit.” That description was selected for an estimated 63% of the base-aroma responses and 67% of the DMS responses, compared with 37% of the cineole responses. The experiment established contributions to particular aroma notes; whether increasing the amount improves liking or makes real fruit taste better requires separate evaluation.

Searching 87 individuals for regions associated with cineole content

The researchers next examined a cross population whose female parent was Hort22D, which contains high levels of cineole in ripe fruit. Comparing genetic variation in 87 individuals with volatile compounds in their ripe fruit identified two chromosome regions associated with cineole content.

One region, corresponding to chromosome 29, contained a cluster of candidate terpene synthase genes. Terpene synthases are enzymes that produce various plant aroma compounds from shared starting materials. Similar sequences do not necessarily make the same products, so the researchers isolated the candidates and tested their functions.

Four genes obtained from Red5 fruit and leaves were named AcTPS1a–AcTPS1d. Of these, AcTPS1b emerged as the leading candidate for cineole production.

Testing function with purified enzymes, leaves, and fruit

The researchers combined experiments using enzymes produced in bacteria and purified with experiments that temporarily expressed the genes in plant leaves. The enzyme encoded by AcTPS1b produced several terpenes, including cineole, from the substrate geranyl diphosphate. Its product profile closely resembled the terpene profiles of Red5 and Hort22D fruit.

Temporarily expressing AcTPS1b in the flesh of another kiwifruit species, Actinidia eriantha, also produced a cineole peak that was absent from the control. In this experiment, DXS, which helps supply starting material, and p19, which helps maintain expression of introduced genes, were co-introduced; the control included the same supporting conditions.

The agreement between reactions with the isolated enzyme and those within plants strengthened the conclusion that AcTPS1b is a gene responsible for cineole production in kiwifruit.

Closely related genes can produce different aromas and act in different places

The other candidates had different functions. AcTPS1a mainly produced sabinene, AcTPS1c mainly produced geraniol, and expressing AcTPS1d in plant leaves yielded springenes. AcTPS1a could also produce higher-carbon sesquiterpenes from another substrate.

Examination of their locations within cells showed that AcTPS1a–1c were targeted mainly to chloroplasts and AcTPS1d mainly to the cytosol. Their tissue-expression patterns also differed: AcTPS1d was expressed in leaves and buds, while AcTPS1a and AcTPS1b were expressed in tissues including fruit. Within this cluster of similar genes, products, usable substrates, and sites of activity had diverged.

AcTPS1a and AcTPS1b were particularly striking: their main products differed even though only 15 of their 603 amino acids differed. A model based on known enzyme structures highlighted differences in amino acids at the reaction site as candidates. This structural explanation was based on modeling and earlier research in other plants, rather than an experiment in this study that replaced the individual residue to test its effect.

Not just whether a gene can make cineole, but how strongly it is expressed

Following fruit development and postharvest ripening showed that cineole content varied with cultivar, stage, and whether flesh or skin was examined. In Red5 and Hort22D, flesh content was high during development and subsequently declined. Cineole levels in Red5 skin were generally two to three times those in the flesh at the same stage.

Among 12 lines selected from the cross population, higher AcTPS1b expression was associated with higher cineole content in ripe fruit, with R² = 0.90. Thus, differences in compound abundance may depend importantly on how strongly the gene is expressed, not merely on whether the enzyme can be produced.

However, the causal gene in the other associated region was not established. Several regulatory processes may influence cineole levels, and the relationship observed in these 12 lines cannot automatically be applied to every cultivar.

Summary

In a model kiwifruit aroma, 1,8-cineole increased impressions such as eucalyptus, mint, and flowers, while DMS increased the impression of blackcurrant. Genetic analyses, purified-enzyme studies, and plant experiments together identified AcTPS1b as a major gene for cineole production.

Related genes produced different terpenes and differed in where and when they were expressed. Starting with human perception and connecting aroma differences to compounds, enzymes, and genes offers a route toward aroma-informed breeding.

The research was supported by New Zealand’s Ministry of Business, Innovation and Employment and internal PFR funding, which included income derived from kiwifruit cultivars and royalties. Travel support from Chinese research programs and other sources was also acknowledged. The authors declared no conflicts of interest.

Reference

Zeng Y, et al. Sensory-Directed Genetic and Biochemical Characterization of Volatile Terpene Production in Kiwifruit. Plant Physiology. 2020;183(1):51–66. doi:10.1104/pp.20.00186. Open-access full paper.