Plant Molecular Biology 43: 179-188 (2000)
Epigenetic aspects of somaclonal variation in plants
Shawn M. Kaeppler, Heidi F. Kaeppler and Yong Rhee


Somaclonal variation is manifested as cytological abnormalities, frequent qualitative and quantitative phenotypic mutation, sequence change, and gene activation and silencing. Activation of quiescent transposable elements and retrotransposons indicate that epigenetic changes occur through the culture process. Epigenetic activation of DNA elements further suggests that epigenetic changes may also be involved in cytogenetic instability through modification of heterochromatin, and as a basis of phenotypic variation through the modulation of gene function. The observation that DNA methylation patterns are highly variable among regenerated plants and their progeny provides evidence that DNA modifications are less stable in culture than in seed-grown plants. Future research will determine the relative importance of epigenetic versus sequence or chromosome variation in conditioning somaclonal variation in plants.

Somaclonal mutant phenotypes as qualitatively and quantitatively inherited mutations

Somaclonal mutant phenotypes segregate as qualitatively and quantitatively inherited mutations Somaclonal variation was first detected by the high frequency of qualitatively segregating phenotypes observed among progeny of plants that were expected to be genetically identical (reviewed in Larkin and Scowcroft, 1981, 1983; Orton, 1984; Ahloowahlia, 1986; Larkin, 1987; Sun and Zheng, 1990; Peschke and Phillips, 1992; Kaeppler and Phillips, 1993a). This was especially true in diploid species such as maize where mutations could be easily observed and were not obscured by genetic buffering that is prevalent in polyploid species. For example, a study of maize grown for 8 months in culture found that, on average, every regenerated plant contained 1.32 mutants that produced a visible phenotype (Lee and Phillips, 1987b). These qualitative mutants were stably inherited for several subsequent seed-derived generations.

Detailed phenotypic analyses in later studies showed that quantitative variation is also frequently found among regenerant-derived progeny (reviewed in Duncan, 1997; Veilleux and Johnson, 1998). Quantitative variation has been described for many phenotypes including plant height, plant biomass, grain yield, and agronomic performance (e.g. Earle and Gracen, 1985; Zehr et al., 1987; Lee et al., 1988; Carver and Johnson, 1989; Dahleen et al., 1991; Bregitzer et al., 1998). A generalization of studies that have assessed quantitative variation is that quantitative variation is frequent and inheritance studies indicate alteration of numerous loci.

Global methylation changes have also been studied in tissue culture. LoSchiavo et al. (1989) showed that global methylation levels changed in response to hormone concentration in the media of carrot cultures. Methylation levels decreased with increasing concentration of kinetin, but increased with increasing amounts of the auxin 2,4-D. In this study, methylation levels were also developmentally regulated, being "reset" as cells were induced to embryogenesis. Arnholt-Schmitt et al. (1995) also reported that global methylation levels of carrot culture changed in a growth-phase-dependent manner. A companion study (Arnholt-Schmitt, 1995) found that the copy number of repeated sequences also changed through development, concurrent with the methylation change. Therefore, the combined data suggest that methylation reduction may have occurred due to genome diminution rather than changes in the frequency of methylated target sites.

Studies of both global methylation levels and methylation of specific sites show that variation in DNA methylation occurs frequently in the culture process. Global methylation studies support the idea that developmental timing may play a role in effecting variation in methylation levels and patterns.