In triploblastic animals, LSF/Grainyhead (GRH) transcription factors perform a number of functions essential to both development and homeostasis. They are involved in regulation of the cell cycle, cell division, and cellular differentiation in a range of developmental and non-developmental contexts [1–14].
The LSF/Grainyhead family is split into the LSF/CP2 subfamily and the Grainyhead (GRH) subfamily, which can be distinguished by their distinctive oligomerization domains and differences in their oligomerization behavior [10, 15, 16]. GRH binds to DNA as a dimer, whereas LSF binds as a tetramer [17, 18]. The DNA binding regions in both protein subfamilies show a large amount of conservation [17, 18], but each has distinct transcriptional targets. GRH binds to the DNA sequence: (A/T)C(A/C/T)(G/T)GTT(C/G/T), whereas LSF binds to a direct repeat with the consensus sequence of N(C/G/T)N(C/G/T)(C/G)N(C/T)N(C/G/T)NN(C/G/T)(C/G/T)N(A/C/G)N [15, 16, 18, 19]. LSF proteins can also be distinguished from GRH by the possession of a sterile alpha motif (SAM) . Members of both LSF and GRH subfamilies were previously identified in vertebrates, arthropods, and nematodes, so the origin of the family and the diversification into subfamilies is known to predate the evolutionary split between protostomes and deuterostomes . Recently, a common origin for the LSF/GRH family and the p53 family has been proposed based on similarities in the folding of their DNA-binding domains .
The differences in the molecular functions of LSF and GRH are accompanied by important differences in their biological roles. In both vertebrates and protostome invertebrates, GRH proteins are involved in the development and maintenance of epithelial integrity . For example, in mice, grh is required during embryogenesis where it is expressed exclusively in the developing ectodermal epithelium . Furthermore, embryonic mice lacking grhl-3 exhibit insufficient wound repair and abnormal skin barrier formation leading to excessive postnatal water loss. The water loss is associated with reduced expression of the gene encoding TGase1, an enzyme that promotes cross-linking of parts of the stratum corneum, thus preventing the movement of water and solutes . Likewise, in Xenopus, a Grh-like gene (Xgrh1) has been implicated in the development of the epidermis . One of its primary targets is epidermal keratin. In morpholino studies, knockdown of Xgrhl led to loss of surface structures and pigmentation as well as neck and eye defects associated with epidermal instability . In Drosophila, GRH plays a critical role in epithelial integrity that is analogous to and perhaps homologous with the role played in vertebrates--GRH maintains the tension of the Drosophila cuticle, and it induces cuticle development and cuticle repair following injury [23, 24]. Similarly, the CeGrh1 protein of C. elegans appears to be required for proper cuticle formation during development, as its knockdown leads to soft, malformed cuticles and embryonic lethality .
In addition to its widely conserved role in maintaining epidermal integrity, Grh is also involved in the specification and development of the CNS in both Drosophila and mice [9, 25]. Additionally, in mice, Grh mutants exhibit defects of the salivary and kidney ducts and eyelid closure [26–28], and in humans, a single nucleotide polymorphism found in GRHL2 is associated with age-related hearing impairment .
The biological roles of LSF are diverse and they have clearly diverged from those of GRH, at least in mammals, where the function of LSF has been well characterized. LSF is ubiquitously expressed . It appears to play a role in liver function, eye development, erythropoesis, neural and immune function, regulation of the cell cycle progression, and cell survival [8, 16, 31–42].
When the ancestors of the LSF and GRH subfamilies first originated via a gene duplication event from their common ancestor, they would presumably have had identical or largely overlapping functions. However, at least in extant mammals, LSF and GRH have diverged extensively with respect to their biological roles. The basis for this functional diversification is not clear. The common ancestral functional repertoire of LSF and GRH may have become "subfunctionalized" in the two descendants . Alternatively or in concert, LSF and GRH may have independently acquired novel functions since their split from a common ancestral gene ("neofunctionalization") [43, 44].
If we wish to reconstruct the initial functional diversification of LSF and GRH, it is necessary to identify the ancestor in which the original gene duplication occurred. This may permit us to infer the functional repertoire of the LSF/GRH ancestor, and to compare this ancestral condition with the function of LSF and GRH in a phylogenetic progression of extant taxa. By comparing vertebrates, arthropods, and nematodes, Venkatesan and co-workers previously showed that the origin of distinct LSF and GRH subfamilies predated the diversification of triploblasts into distinct protostome and deuterostome lineages . With the recent availability of sequenced genomes from several basal metazoans, a choanoflagellate, and more distantly related fungal outgroups, we can track the evolution of the LSF/GRH family into the much more distant past. In this study, we report the identification of LSF/GRH family members in 24 previously unreported species. Through a combination of genome prospecting and phylogenetic analysis, we show that the original gene duplication that produced the LSF and GRH subfamilies occurred prior to the evolutionary radiation of basal animal lineages (e.g., Bilateria, Cnidaria, Ctenophora, Porifera, and Placozoa). Interestingly, the GRH protein of the sea anemone Nematostella vectensis, a representative cnidarian, appears to have split into two distinct loci. We also identify six protein motifs that are widely shared between the LSF and GRH subfamilies of metazoans, all of which can be traced to the common ancestor of metazoans and fungi. In addition, there is a single motif that appears unique to the LSF subfamily.