Deletion of an MMR Gene Copy in Peppermint’s Epidermal Layer Linked to a Lower Menthol Proportion
Background: Genomes Can Differ Within a Clone
A study has linked gene-copy deletions confined to peppermint’s epidermal layer with a lower proportion of menthol in its essential oil. Published in PNAS, the research tracked the relationship between tissue-layer-specific genetic variation and essential-oil composition in peppermint plants treated with gamma radiation to induce mutations.
Even plants propagated vegetatively do not necessarily have identical genomes in every cell. When a mutation arising in the shoot apical meristem becomes established in a particular cell layer, a periclinal chimera can form, with tissue layers carrying different genomes. Peppermint’s oil-producing glandular trichomes originate in the outermost L1 layer, so the layer in which a mutation occurs can help explain aroma composition.
The study examined Black Mitcham, a sterile, allohexaploid peppermint cultivar with genomes derived from different species. Because conventional genetic analysis based on crossing is difficult in this crop, the researchers combined induced mutagenesis and genome analysis to look for variants that alter essential-oil composition.
Key Findings: Epidermal Deletions Linked to Low-Menthol Oil
The researchers irradiated axillary buds with gamma rays, then carried out six rounds of propagation by cuttings and replanting to obtain 261 independent lines. They assembled a reference genome using PacBio HiFi, which reads long DNA sequences, and Hi-C, which captures chromosomal contact information. They then analyzed relative sequence copy number in short-read DNA sequencing data from each line. They detected 1,406 large insertions and deletions, an average of 5.38 per line; 96.3% were classified as deletions.
Most variants appeared to be restricted to particular tissue layers rather than present throughout the plant. In two low-menthol lines, DNA from leaves, which contain cells from all layers, was compared with DNA from roots derived from L3, supporting the layer-specific pattern. Across the full population, L1-specific deletions were 1.65 times as numerous as L2/3-specific deletions, although the mechanism behind this difference remains unresolved.
The researchers preselected 78 lines for aroma-related traits and evaluated them in field trials over three years, with different lines assessed each year. P11 and P28 showed a low menthol proportion. In P11 essential oil, the proportion of (−)-menthol was 2.6%, compared with 42% in the control for that year. This is a change in the proportion of oil components; it does not indicate a decrease in total essential-oil production.
Both lines shared a deletion region in the chromosome 11 D haplotype, a particular chromosome sequence form, and the deletion was confined to the L1-derived epidermal layer. The region contained a menthone–menthol reductase (MMR) gene, which converts menthone to menthol. In bulk RNA sequencing of whole mature leaves, expression of the affected MMR copy was about one-tenth that of the control in both lines. Low menthol and high menthone levels were also observed under controlled growing conditions, consistent with the field-trial results.
Implications and Outlook: Tissue Layers Matter in Crop Improvement
The findings support the possibility that, even in polyploid crops with multiple gene copies, a key copy in a particular tissue layer can strongly affect compound composition. Other copies may remain but fail to compensate sufficiently for its function. In the control, expression of the D-haplotype MMR copy was higher than that of the other copies.
However, the shared deletion region in the two lines contains 295 genes. The researchers did not precisely remove or restore MMR alone, so the observed changes cannot be conclusively attributed to MMR as the sole cause. The reference genome may also be missing some regions, and the low sequencing depth limited definition of variant boundaries to a resolution of 100 kb. The study did not assess the long-term accumulation rate of naturally occurring mutations or the permanent stability of essential-oil quality.
The study design and findings are available in the full PNAS article.
Proceedings of the National Academy of Sciences, Volume 123, Issue 21, May 2026. DOI: 10.1073/pnas.2606214965
Summary: Challenging the historical dogma of genetic uniformity within vegetatively propagated clonal lineages, this landmark study elucidates the structural mechanisms where somatic mutations drive essential oil diversity in long-lived peppermint (Mentha × piperita). By integrating single-cell RNA-sequencing (scRNA-seq) with long-read genome phasing across the distinct histological layers (L1, L2, and L3) of the shoot apical meristem (SAM), the framework isolates persistent mosaicism established inside founder stem cell niches. The genomic metadata proves that layer-specific somatic mutations programmatically rewire down-stream monoterpene and terpenoid secondary metabolite pathways. This intra-clonal micro-evolutionary drift alters the catalytic kinetics of rate-limiting biosynthetic enzymes to dictate overall volatile chemistry, delivering a scalable computational baseline for targeting metabolic hyper-variants and engineering high-fidelity molecular farming platforms.
When selecting peppermint lines for fragrance use, checking which tissue layer carries a variant alongside measuring essential-oil composition can help guide propagation strategies. A trait caused by a mutation in the epidermis may not persist in tissue-culture regenerants derived from other layers. For example, maintaining a low-menthol line could involve repeatedly measuring oil composition and testing whether the propagation and regeneration route preserves the variant. This study provides a basis for using layer-specific variation to modify compound profiles in sterile, polyploid crops. A lower menthol proportion does not necessarily mean improved quality, however. Whether the composition suits its intended use, and whether cultivation performance and propagation stability can also be maintained, require separate evaluation.