Layer-specific genetic variation unlocks secondary metabolite diversity in long-lived clonal peppermint

Mutations that arise in the shoot apical meristems can become fixed, but typically only in one of the meristem layers. Therefore, in long-lived, clonally propagated species, polymorphic genomes coexist in the form of periclinal chimeras. Given their evolutionary and economic impact, it is critical to understand the dynamics and phenotypic implications of layer-specific variation. Here, we γ-irradiated axillary buds from an elite peppermint clone and obtained 261 independent mutants carrying large indels. We produced a haplotype-aware, high-continuity assembly of this sterile allohexaploid and, using short-read sequencing, detected, on average, six large indels per mutant. Importantly, most of these mutants were periclinal chimeras: comparison of mutation frequency in root (derived solely from the L2/3 layer) and leaves (which contain cells from all three layers) demonstrated that the indels are confined to either the outer, L1-derived layer, or the inner L2/3 layers. We observed that the L1 layer was more often mutated, confirming that mutation rate in the shoot apical meristem is potentially optimized to each meristematic layer. To assess whether deletion of a single haplotype in a single meristematic layer could affect plant function, we characterized mutants under field conditions, detecting variation in secondary metabolite production. Two mutants produced an oil with very low (-)-menthol levels, associated with the loss of a single haplotype of the menthone-menthol reductase gene in the epidermal layer. These results highlight the evolutionary relevance of layer-specific genetic variation and present opportunities for improvement of clonally propagated crops that suffer from genetic diversity bottlenecks.