Fruits of wild (a) and domesticated (b) tomatoes. Picture courtesy of Dani Zamir, and reproduced with permission from Zamir, D. Nature Rev. Genet. 2, 983–989 © (2001) Macmillan Magazines Ltd.

Fruit weight and size are agriculturally important traits, but little is known of their genetic and molecular bases. Many genetic studies on these traits have been done in the tomato, owing to the disparity in size between the fruits of wild and domesticated tomatoes (see picture). Quantitative trait loci (QTL) mapping studies in this plant have identified nearly 30 tomato QTL that affect fruit weight and size.

One such QTL is fw2.2, which accounts for ∼30% of the difference in fruit weight between wild and domesticated tomatoes. Earlier studies have strongly indicated that altered gene regulation underlies the effects of the large- and small-fruit alleles of fw2.2 on fruit weight. It has long been believed that such mutations, especially when they affect the timing of development ('heterochronic' mutations), might be a natural force of evolutionary change in plants. In a detailed study of fw2.2, Steven Tanksley's group now provide the first experimental evidence to support this theory.

Because previous studies in plants and Drosophila have shown that both cell division and expansion are essential factors that determine organ and fruit size, Cong et al. analysed cell size and mitotic index (MI) in two nearly isogenic tomato lines in which either a large- or small-fruit fw2.2 allele was present. Differences in MI were found between the fruits of these two lines, but not in cell size. In the small-fruit fw2.2 line (TA1144), a rapid but brief rise in MI occurs immediately after fertilization. By contrast, a more gradual and sustained rise in MI occurs in the large-fruit allele line (TA1143), indicating that an extended period of cell division might underlie larger fruit size in this line. Next, the authors found that the fw2.2 alleles differ in the timing of their peak expression by around one week. This difference in expression timing inversely correlated with changes in mitotic activity during early fruit development, indicating that fw2.2 might negatively regulate cell division. Moreover, by ∼12 days post-fertilization, fw2.2 levels in TA1144 were more than double those in TA1143. However, only subtle differences in expression patterns were evident between the two lines.

These results show that the differences in transcript levels between the small- and large-fruit alleles of fw2.2 are both quantitative (with the small-fruit allele being more abundantly expressed) and qualitative (as evident from the difference in their expression timing). Importantly, these findings provide empirical evidence that heterochronic regulatory changes in gene expression can bring about phenotypic, and probably evolutionary, change in plants. But how fw2.2 actually modulates cell division remains unknown.