Insights into watermelon genetic diversity

There’s no doubt that most of the time when we non-experts think of genetic diversity in DNA terms, we think about alterations in the actual DNA sequence. Change the genetic code and there’s a chance you change the nature of a gene and as a result change the outward appearance, the phenotype, of an organism. There have been lots of studies relating the differences among species to differences in the DNA code of this sort, but far fewer looking at the differences among varieties of a single species. Amnon Levi and his colleagues looked at the sequence differences among heirloom watermelon varieties and were surprised by the lack of diversity. Despite the fact that heirloom watermelons differ in highly complex ways in a whole slew of traits, ((Parris, 1949, is the motherlode on this, if you can get it.)) Levi and colleagues concluded that “there is a very narrow genetic diversity at the DNA sequence level”.

There are, however, other sources of variation that are not reflected in the DNA sequence. Once such is methylation, the attachment of a methyl group to two of the four letters that make up the genetic code. Methylation is closely involved in whether a gene is active or not, and the same gene with two different methylation patterns in two different individuals can be expressed differently, resulting in different phenotypes despite genotypes that are similar in sequence (though obviously not in methylation). Methylation patterns are inherited, but they are not strictly speaking genetic; they are one kind of epigenetic inheritance.

Levi and his group turned their attention to the methylation pattern of 47 watermelon varieties. ((Not including, alas, the famed Moon and Stars watermelon.)) The epigenetic diversity varied from 16-43%, while diversity measured by conventional DNA sequence markers ranged from 3-20%. “Diversity at the methylation level is three times higher than the genetic diversity revealed by DNA markers on the same set of heirloom DNAs,” the authors conclude. ((Nimmakayala, P., Vajja, G., Gist, R., Tomason, Y., Levi, A., & Reddy, U. (2010). Effect of DNA methylation on molecular diversity of watermelon heirlooms and stability of methylation specific polymorphisms across the genealogies Euphytica DOI: 10.1007/s10681-010-0259-z)) The vast majority of methylation patterns are inherited stably from the variety’s parents, with very few arising fresh.

The bit I really do not understand in all this is the extent to which the relationships among varieties deduced from methylation patterns match those derived from DNA sequence differences. The new paper and one from 2001 both contain tree diagrams of the relationships, but there is no direct comparison. ((I’m not even sure if that is doable.)) Eyeballing the trees, and looking only at the nearest neighbours, the overlap does not seem all that impressive. I’d like to know more.

That 2001 paper also uses the low genetic diversity (at the DNA sequence level) to argue for “the need to broaden the genetic base of cultivated watermelon”. That idea seems to have fallen by the wayside as the diversity in epigenetic factors has emerged. Is there a more general conclusion to be drawn?

This observation of greater methylation than DNA sequence differences adds to the growing importance of epigenetics in studies of diversity, and may become important in breeding new varieties with specific desired traits. If the methylation patterns can be linked to phenotypic traits, as has already been shown in Arabidopsis, it could be possible to alter methylation without needing to do crosses and selection.

Bonus factoid: “Watermelon is the fifth most economically important vegetable crop and is grown in 44 states in the United States.”

Bonus hint: Don’t go searching Google images for “watermelon diversity”. You’ll likely be disgusted.