Control of embryonic development occurs at several biochemical/molecular levels, including transcription, RNA processing, translation, and post-translational processing. Using the frog Xenopus laevis as an experimental model organism, our project aims at identifying factors and mechanisms responsible for the control of early vertebrate development. We focus on the ectoderm, which gives rise to epidermis, the central nervous system, and the neural crest (NC) and its many derivatives. We base our work on the hypothesis that the appearance and differentiation of ectoderm-derived tissues is regulated primarily at the level of transcription and that such regulation can be elucidated by identifying and testing the function of specific transcription factors and the genes regulated by them (including but not limited to other transcription factors). We work primarily on transcription activator AP2alpha and its relatives and on downstream target gene candidates identified by microarray analysis.
The TFAP2 gene family in Xenopus
Sargent, Luo, Zhang
In mammals, the transcription factor AP2 (TFAP2) family has five members (alpha-epsilon or TFAP2a, b, c, d, and e). Until recently, only a single TFAP2 had been characterized in Xenopus, the homologue of TFAP2a. We showed that TFAP2a was critical in both epidermal and neural crest development in the frog. Other laboratories have investigated the function of this gene in mouse and zebrafish, with results similar to ours in Xenopus but pointing toward some redundancy and interaction with other TFAP2 genes. We performed an exhaustive search of Xenopus databases and other resources to identify other members of the TFAP2 family in Xenopus. As a result, we discovered Xenopus TFAP2b (or TFAP2beta) and TFAP2c (or TFAP2gamma). Based on analysis of the Xenopus tropicalis genomic sequence, which is completed to about eight-fold coverage, we have observed no other TFAP2 genes in the Xenopus genome. As is the case in mammals, all three genes are expressed in neural crest, with different boundaries and different expression patterns in other tissues such as epidermis and pronephros. Loss-of-function studies suggest that both TFAP2a and TFAP2b are important in the neural crest but that TFAP2c plays less of a role in this tissue. Thus, in the frog as in other vertebrates, TFAP2 family members display overlapping expression and function in early development.
Downstream regulatory targets of TFAP2 in Xenopus
Sargent, Luo, Rangarajan, Zhang; in collaboration with Fallon, Lim, Schilling, Williams
We discovered that if Wnt-beta catenin signaling was activated and BMP signaling completely quenched in Xenopus ectoderm, tissue identity could be switched from central nervous system to neural crest by supplying the factor TFAP2 (Luo et al., 2003). We exploited this phenomenon in order to generate microarray probes that we then used to identify a collection of about 40 genes that could be activated by TFAP2 under the above experimental conditions, thus potentially representing a battery of downstream targets for the factor. In addition to revealing unexpected features of neural crest induction, such as a consistent link to the epidermal differentiation program, our investigation yielded a few novel genes with potentially interesting developmental functions. The laboratory is now focusing on two of the genes, PCNS and Inca.
Luo T, Lee YH, Saint-Jeannet J-P, Sargent TD. Induction of neural crest in Xenopus by transcription factor AP2alpha. Proc Natl Acad Sci USA 2003;100:532-537.
A novel protocadherin required for somite and neural crest development
Rangarajan, Luo
Protocadherins are a large subfamily (about 70 genes in mammals) of the cadherin superfamily of calcium-dependent cell adhesion molecules. We discovered a novel protocadherin strongly upregulated by AP2alpha and have named it PCNS (protocadherin in neural crest and somites). The gene is transiently expressed in somites in an anterior-posterior wave correlating with the condensation of somites from paraxial mesoderm and strongly expressed in pre-migratory and migratory neural crest. Late in development, PCNS mRNA vanishes in derivatives of both of these embryonic tissues but appears in the heart and ear vesicle. Achieved via anti-sense oligonucleotides as well as by using a dominant negative approach, loss of PCNS function leads to striking phenotypes in neural crest and somites. Neural crest cells lacking PCNS are induced normally but fail to migrate. The defect appears fairly early, probably in the epithelial/mesenchymal transition. In the somite, loss of PCNS prevents the orchestrated rotation of somite cells into an orderly periodic array. PCNS does not appear to be a particularly strong adhesive molecule, and we hypothesize that it functions in one or more signaling pathways to control cytoskeleton or cell polarity.
Inca, a novel factor in craniofacial development
Luo, Xu; in collaboration with Schilling
Another interesting regulatory target of AP2alpha is a novel gene we named Inca (induced in neural crest by AP2). Beginning after gastrulation and continuing throughout development, Inca is intensely expressed in the neural crest. Its expression is dependent on TFAP2 activity in frog and zebrafish. Inca is also expressed in mesoderm during gastrulation and, as development proceeds, in additional tissues such as heart. Homologues of Inca exist in all vertebrates, including mouse, human, and zebrafish, but not in invertebrates such as Drosophila or in other phyla. The Inca protein sequence is entirely novel and without distinguishing features that would allow its assignment to existing protein families. The early expression pattern of Inca is conserved in fish and mouse embryos, reinforcing the hypothesis that Inca plays an important developmental function. We confirmed our hypothesis with loss-of-function experiments that used anti-sense morpholino oligonucleotides, resulting in severe defects in neural-crest–derived craniofacial bone and cartilage in both Xenopus and zebrafish. In a collaborative effort with the laboratory of Trevor Williams, a mouse knockout project is currently underway.
Using yeast two-hybrid screening of a mouse embryo cDNA library with mouse Inca, we identified a p21-activated kinase (PAK4) as a candidate interaction partner with Inca. Xenopus Inca and XPAK5, the Xenopus homologue of mouse PAK4, form a stable complex when co-expressed in frog cells. With PAK proteins implicated in the regulation of cytoskeletal dynamics, we were interested to observe that overexpression of Inca in this context disrupts the cytoskeleton and cell-cell adhesion in the early embryo. Furthermore, Inca and PAK5 are strongly synergistic in this effect. Inca may also interact with 14-3-3, a scaffolding protein that has been implicated in cytoskeletal control mechanisms and other functions. We expect Inca to be a useful tool in understanding cell migration and cell polarity control in neural crest, with more general implications as well.
Luo T, Zhang Y, Rangarajan J, Khadka D, Cho K, Sargent TD. Regulatory targets for transcription factor AP2 in Xenopus embryos. Dev Growth Differ 2005;47:403-413.
Regulation of Fmr1 by TFAP2
Luo; in collaboration with Fallon, Lim
In a collaborative project with Jae Lim, a former predoctoral fellow and now an MD/PhD student in Justin Fallon’s laboratory, we identified and characterized the Xenopus homologue of the human Fragile X Mental Retardation-1 gene (Fmr1). Loss-of-function mutations in this gene are the basis for most cases of Fragile X syndrome, which is the predominant form of hereditary mental retardation and is accompanied by other neuropathological problems and craniofacial dysmorphology. Based on sequence analysis of the human Fmr1 gene and the observation that, in Xenopus, Fmr1 is expressed in neural crest and branchial arch mesenchyme, which are tissues whose induction requires TFAP2a, we hypothesized that Fmr1 might be partly regulated by TFAP2a in mammalian and amphibian cells. Overexpression of a dominant negative TFAP2a led to reduction of Fmr1 expression in the Xenopus embryo while wild-type TFAP2 could rescue Fmr1 expression in ectoderm. Furthermore, chromatin immunoprecipitation assays showed that TFAP2a binds to the human Fmr1 promoter in HeLa cells. The results corroborated findings in TFAP2a-null knockout mice in which Fmr1 expression is significantly reduced, supporting the conclusion that TFAP2a regulates Fmr1 in the vertebrate embryo.
Lim JH, Booker AB, Luo T, Williams T, Furuta Y, Lagutin O, Oliver G, Sargent TD, Fallon JR. AP-2alpha selectively regulates fragile X mental retardation 1 gene transcription during embryonic development. Hum Mol Genet 2005;14:2027-2034.
Lim JH, Luo T, Sargent TD, Fallon JR. Developmental expression of Xenopus Fragile X mental retardation-1 gene. Int J Dev Biol 2005;49:981-984.
Sargent TD. Patterning non-neural ectoderm by organizer-modulated homeodomain factors. In: Grunz H, ed. The Vertebrate Organizer. Berlin: Springer, 2004;219-231.
Sargent TD. Transcriptional regulation at the neural plate border. In: Saint-Jeannet J-P, ed. Neural Crest Induction and Differentiation. Georgetown TX: Landis Bioscience, 2005 (in press).
1Left laboratory in 2005.
Collaborators
Justin Fallon, PhD, Brown University, Providence, RI
Jae Lim, MD, Brown University, Providence, RI
Thomas F. Schilling, PhD, University of California Irvine, Irvine, CA
Trevor Williams, PhD, University of Colorado, Denver, CO
For further information, contact tsargent@nih.gov.