Mod. 23 Jan 2000
Trends in Plant Science, vol. 4, No. 111, Nov. 1999
An epigenetic mutation responsible for natural variation in floral symmetry (Nature 401: 157-161)
P. Cubas, C. Vincent, and E. Coen (1999)

As an extension to their work on flower symmetry in model species, the authors have turned their attention to a naturally occuring mutant in (Linaria vulgaris) that was first described by Linnaeus >250-years ago. Wild-type flowers of Linaria show dorsoventral asymmetry in both the shape and colour of the petals, whereas the naturally occurring peloric mutant is radially symmetrical. This mutant phenotype is similar to mutants of Anthirrhinum, which lack the functional genes, cycloidea and dichotoma. By isolating a genomic clone of a cycloidea-like gene from Linaria, it was possible to use RFLP analysis to show linkage to the peloric phenotype. However, comparison of the mutant and wild-type cycloidea-like genes revealed no significant sequence differences at restriction sites shown to be polymorphic. A possible explanation for this discrepancy is that the enzymes used in the RFLP analysis are sensitive to methylation and that methylation could be responsible for the mutant phenotype. This was tested by digesting wild-type and peloric genomic DNA using methylation-sensitive and -insensitive enzymes to reveal methylation of the cycloidea-like gene. Methylation did not extend to all the genes examined, but this was insufficient proof that methylation affected the activity of the cycloidea-like gene. Proof was obtained by analysing peloric mutants that showed somatic instability—cuttings from branches with nearly wild-type phenotypes gave digest patterns that were consistent with partial demethylation. Thus, this naturally occurring mutation is not caused by a change in the sequence, but is a consequence of a heritable epigenetic mutation involving methylation. The authors discuss the possible significance of this for plant evolution in natural populations.


Hugo de Vries coined the word "mutation" to describe the variations he found in Evening Primroses and other plants. He wrote Mutation Theory. Even so, biochemists have claimed that de Vries did not observe "true" mutations, since what he saw did not involve changes in the structure of genes. They redefined his word to exclude the original sense. It is nice to see de Vries' word again being used to describe natural variations that do not involve structural modifications of genes.

Gene silencing is obviously an important tool in evolution, since it allows a population to retain a gene that is no longer necessary, but which may become useful in some future environment. Gene-mutations, or structural changes, would have populations passing along a broken gene which might be re-mutated into a functioning form by a fortuitous future gene-mutation.

I have not yet read the original article, but I hope some mention is made of the fact that the epigenetic regulation described does not appear to be a "mutation", which was meant to represent a non-Darwinian leap. If the peloric trait can be unstable, there is presumably some means for stabilizing it. In other words:

  1. How can a methylated gene become partially methylated?
  2. Can the process work in reverse?
  3. Is there a methylation continuum of non-/partial-/full- ?
  4. What is responsible for the change and "remembers" the state?

It would be very interesting to re-examine some classic cases of dominance modification to see if methylation/demethylation is involved.

It would also be enlightening to explore gene silencing where the methylated gene is not structurally and functionally equivalent to its partner. Could the silencing be transferred?