Nanosecond-scale kinetics of nematocyst discharge

dispersed genes in different lineages. First, the cluster may have been broken up in organisms with elevated rates of genome evolution, such as C. intestinalis, C. elegans and Drosophila melanogaster. This implies that cluster integrity in other lineages may not be due to selection. Conversely, clusters may have been maintained in some lineages due to selection. This would imply a relaxation of selection in lineages that have undergone cluster breakup. In this context it is notable that those lineages that have undergone Fox cluster breakup (C. intestinalis, C. elegans and D. melanogaster) also show Hox cluster breakup (Supplemental Data). Gene expression has previously provided important insight into the nature of selective forces acting on gene clusters [1] and amphioxus has proven useful for identifying the primitive role of developmental genes [9]. Consequently, we compared the expression of FoxL1, FoxC, FoxF and FoxQ1 in amphioxus embryos [10]. Our results show that all four genes are activated sequentially in the developing endo-mesoderm (Figure 1B). FoxL1 and FoxC are activated in the mesoderm of the early neurula, and later come to mark different dorsal mesoderm compartments. FoxF is activated at the end of neurulation, and marks the ventral mesoderm. FoxQ1 is activated later still in the developing endoderm [10]. These data highlight coordinated endo-mesodermal expression as a potential selective force for maintaining Fox gene clustering. This is supported by functional studies in vertebrates and Drosophila, in which manipulation of FoxF, FoxC, FoxL1 and FoxQ1 gene function has resulted in a variety of mesoderm-associated phenotypes. Importantly, a feature often observed in such manipulations is a transformation in the identity of mesoderm compartments, suggesting a role for the Fox genes in regulating compartment identity (Supplemental data). Our data demonstrate that a small cluster of Fox genes evolved prior to the radiation of crown bilaterians and has been at least partially maintained in several lineages, including amphioxus, vertebrates and the honeybee, for over 500 million years. Furthermore, regulation of endo-mesoderm development may be a potential selective constraint underlying Fox gene cluster maintenance. Thus, preservation of ancient gene clustering may be more widespread than previously recognised.