Whole Transcriptome Sequencing Unveils the Genomic Determinants of Putative Somaclonal Variation in Mint (Mentha L.)

Abstract

Mint (Mentha L., Lamiaceae) is a strongly scented herb of the family Lamiaceae that is grown mostly by clonal propagation, making it a valuable species for the study of somaclonal variation and its phenotypic consequences. The recent introduction of a few species of mint in South America, followed by a presumably rampant propagation, make this region particularly ideal for studying the extent of somaclonal genetic diversity. Hence, the objective of this work was to offer a preliminary characterization of somaclonal genetically coding diversity of the mint in the northern Andes in order to address the question of whether somaclonal variants may have emerged despite relatively recent introductions in a region where mint is not native. A total of 29 clonally propagated specimens, collected in mint export farms in the province of Antioquia, a major region for mint production in the northwest Andes of Colombia, were genotyped using RNA sequencing (RNA-Seq). SNP calling was carried out from the leaves' transcriptome profiles of each plant by combining the GATK4 and TRINITY protocols, obtaining a total of 2033 loci across 912 transcripts with a minimum read depth of 20X and 4% of missing data. Unsupervised machine learning algorithms considered the K-means, AGNES and UPGMA approaches, all of which suggested three genetic clusters for M. spicata and a unique cluster for M. × piperita. The results indicate that at least two different origins of M. spicata reached the eastern region of the Antioquia province, clonally propagated in the locality ever since for local consumption and export. One of these ancestries had more population structure, possibly due to environmental or anthropological pressures that intervened in the fragmentation of this genetic group or to a higher somaclonal mutation rate. This work offers a first step into the study of the accumulation and transmission of presumably quasi-neutral somatic mutations at coding regions in an herbaceous clonally propagated scented species such as mint, likely favored by an expected population expansion after its Andean introduction. These ad hoc hypotheses warrant further study as part of future research.

Keywords: M. spicata; M. × piperita L.; RNA-Seq; comparative transcriptomics; crop biodiversity.

Figure 1

Overall diversity patterns in mint from the northwest Andes. (A) PCA analysis from 2033 loci distributed in 912 transcripts. The first component explained 58.56% of the variance, and the second component explained 22.27% of the variance (both totaling 80.83% of explained variance). Clustering validation was performed using the algorithms NbClust and optCluster, which suggested a total of four clusters: three of M. spicata (spicata A, spicata B and spicata C) and one of M. × piperita. All clusters were recovered by the first two components. (B) Dendrogram carried out by UPGMA analysis from 2033 variants distributed in 912 transcripts using Nei’s distance and bootstrap as resampling method with 10,000 replicates. These results also suggested a total of four clusters: three from M. spicata (spicata A, spicata B and spicata C) and one for M. × piperita. Clusters under both approaches in (A) and (B) are fully concordant.

Figure 2

GO pathway to analyze polymorphic biological processes via GO codes from Blast2GO. GO terms ranged from 35,822 to 1471.06 and the number of associated transcripts ranging from 365 to 428 sequences. Main GO terms related to carbohydrate and energy metabolism (GO:0044237 and GO:0044238) and cysteine and methionine synthesis (GO:0008152 and GO:0009987). We also explored the enzyme codes of the 52 KEEG pathways associated with all polymorphic transcripts (Table 2, extended in Table S5). Of the KEEG pathways, 25% were related to carbohydrate metabolism, 19.23% of the KEEG pathways were related to amino acid metabolism, 11.54% of the KEEG pathways were related to energy metabolism and the other pathways were associated with less than 10% of the target queries. Main pathways of carbohydrate/amino acid synthesis were linked to glycolysis/gluconeogenesis (Figure S6), and cysteine/methionine metabolism (Figure S7) [14].