Can Yeast Adjust the Constituent Ratios of Sandalwood’s Aroma?

Imagine producing compounds associated with sandalwood’s aroma in yeast rather than extracting them from trees. Is producing a large quantity enough to obtain the fragrance material you want? For materials comprising multiple constituents, not only quantity but also the proportions of those constituents matter.

In 2022, Wenlong Zha and colleagues modified an enzyme that produces compounds associated with sandalwood oil and adjusted production levels and constituent ratios in yeast. They used calculations to investigate reactions inside the enzyme and connected those insights to the design of actual yeast strains.

Sandalwood’s aroma involves more than one constituent

Sandalwood oil contains multiple constituents, including santalenes and santalols. The researchers focused particularly on Z-α-santalol and Z-β-santalol, considered important to its aroma. The labels α and β distinguish structurally different forms.

The paper targeted the ranges of 41–55% Z-α-santalol and 16–24% Z-β-santalol in the sandalwood oil standard ISO 3518:2002 cited at the time. Here, those two constituent proportions served as design benchmarks for the research.

In yeast production, one enzyme first makes santalenes, and another converts them to santalols. Both the amounts of different santalenes made by the first enzyme and the extent of the subsequent conversion affect the final constituent ratios.

Focusing on the space available to reaction intermediates

The researchers compared SaSSy, a sandalwood enzyme that produces multiple constituents, with SanSyn from another plant, Clausena lansium, which mainly produces α-santalene. The latter enzyme was the principal target of modification.

Combining predictions of three-dimensional protein structure with calculations of atomic movement and chemical reactions suggested that differences in space inside the enzymes help determine the products. When a reaction intermediate fits into a confined space, its possible shapes and reaction routes are restricted. A little more room can allow another shape and a route to a different product.

These calculations are not direct recordings of the reactions. The researchers therefore also replaced amino acids that the calculations suggested were important and examined how the products changed.

Changing one amino acid changed the product proportions

The focus was amino acid 441 in SanSyn. Replacing phenylalanine at this position with the smaller valine produced the enzyme SanSyn F441V. The changes in products matched the computational explanation that the internal space expanded, giving reaction intermediates greater freedom of movement.

With the modified enzyme, α-santalene accounted for approximately 57% and β-santalene for approximately 29%, with two other constituents produced in small amounts. An enzyme that had mainly made one product now produced the desired combination of constituents in a different balance.

These proportions still refer to the santalene stage. They are distinct from the final proportions of santalols valued as fragrance ingredients, so the subsequent conversion also needed adjustment.

Keeping precursors from being diverted into other yeast reactions

The researchers not only changed the enzyme but also reengineered yeast metabolism. They increased the supply of precursors for the target compounds, reduced pathways that consumed the same materials in other reactions, and suppressed conversion of the resulting santalols into other by-products.

Isolating by-products from yeast cultures and determining their structures showed that the yeast’s own enzymes were further modifying the santalols produced. Removing genes thought to be involved reduced some by-products and increased the target constituents. It did not eliminate every side reaction, however.

After these adjustments established a yeast production host, the researchers introduced the modified santalene synthase and the enzyme that converts santalenes to santalols. This combined control of precursor flow with control of how products branch into different forms.

The highest-production conditions fell short of the target proportions

After flask experiments, fed-batch cultivation was conducted in a five-liter fermenter, adding sugar and ethanol during cultivation. Results at this stage were averages of two independent cultures.

Among the conditions examined, a high-producing condition with strain SZ22 yielded approximately 843 milligrams per liter of santalenes and 1,076 milligrams per liter of santalols, totaling approximately 1.92 grams per liter. These figures express target compound production relative to culture volume.

However, Z-α-santalol accounted for 29.4% and Z-β-santalol for 15.0% of the combined target santalene and santalol mixture. These did not reach the paper’s target ranges for the two constituents. Considerable amounts of the intermediate santalenes remained.

Increasing conversion improved the proportions but reduced total production

The researchers therefore increased the copy number of the gene for CYP736A167, the enzyme responsible for converting santalenes to santalols. In the adjusted strain SZ24, Z-α-santalol reached 43.4% and Z-β-santalol 22.0%, within the ranges cited in the paper for these two constituents.

Meanwhile, combined santalene and santalol production fell to 704.2 milligrams per liter, or approximately 0.70 grams per liter. This value is neither the amount of santalols alone nor a yield of unmodified natural sandalwood oil.

Because yeast growth was also lower under the increased-conversion-enzyme-copy conditions, the authors suggest that cellular burden may have reduced production. They identify improving the conversion enzyme’s activity, rather than simply increasing its gene copy number, as a next challenge.

These results show the difficulty of optimizing quantity and proportions simultaneously. At least under the conditions tested, the design that maximized total quantity did not coincide with the design that delivered the desired constituent balance.

Matching constituent ratios does not necessarily mean an identical aroma

Bringing the two major constituent proportions into the ranges cited in the paper is an achievement in controlling chemical composition. However, it is not certification of conformity with every aspect of the standard, nor a result showing that people judged the aroma identical to natural sandalwood.

The study did not compare the development, persistence, or pleasantness of finished perfumes. Evaluating the entire aroma, including minor constituents and by-products, requires work separate from measuring constituent ratios.

Nor can this study alone establish that fermentation necessarily has lower environmental impacts or costs than extraction from trees. It did not conduct a commercial-scale comparison encompassing feedstocks, energy, and purification. Its contribution is foundational technology for controlling production methods and constituent composition.

The research was supported by funding from the Chinese Academy of Medical Sciences, a Guangdong Province research and development program, the National Natural Science Foundation of China, and other sources. The authors declared no competing interests.

Summary

The researchers used calculations to understand reactions inside an enzyme and adjusted santalene ratios by changing one amino acid. They then modified yeast metabolism and the conversion enzyme to bring the two major santalol constituents closer to the proportions targeted in the paper.

The authors propose evaluating fermentation-based fragrance production by more than the number of grams produced. Which constituents form, and in what proportions, can be designed through the enzyme’s mechanism and adjusted alongside precursor supply and cellular burden. Illustrated here with sandalwood, this approach may extend to production of other multicomponent natural products.

References

Zha, W., Zhang, F., Shao, J., Ma, X., Zhu, J., Sun, P., Wu, R., & Zi, J. (2022). Rationally engineering santalene synthase to readjust the component ratio of sandalwood oil. Nature Communications, 13, 2508. DOI: 10.1038/s41467-022-30294-8. Full original article. © The Author(s) 2022. The original article is available under CC BY 4.0.