XB-ART-16528Mech Dev May 1, 1997; 63 (2): 211-25.
Show Gene links Show Anatomy links
A role for Xenopus Gli-type zinc finger proteins in the early embryonic patterning of mesoderm and neuroectoderm.
Gli-type zinc finger proteins play important regulatory roles in vertebrate and invertebrate embryogenesis. In Xenopus, the Gli-type proteins XGli-3 and XGli-4 are first expressed in earliest stages of mesoderm and neural development. Transient transfection assays reveal that XGli-3 and XGli-4 can function as transcription repressors. Counteracting the Gli-protein repressor activity by ectopic expression of a fusion protein that contains the Gli-zinc finger cluster connected to the E1A activator domain in Xenopus embryos results in specific morphological alterations in the developing somites and in the central nervous system. Altered expression characteristics for a broad set of molecular markers highlighting specific aspects of mesodermal and neural differentiation demonstrate an important role for Gli-type zinc finger proteins in the early mesodermal and neural patterning of Xenopus embryos.
PubMed ID: 9203143
Article link: Mech Dev
Species referenced: Xenopus
Genes referenced: actl6a egr2 gli3 h4c4 lgals4.2 mst1 myod1 myt1 neurod1 pax6 slc5a5 tbx2 tub tubb2b tyro3
Article Images: [+] show captions
|Fig. 1. Structural comparison of Xenopus XGli-3 and Xgli-4. (A) cDNA structure of XGli-3 and Xgli-4. Composite cDNA sequences were generated from sets of overlapping partial clones as indicated. XGli-3 and XGli-4 cDNAs contain multiple in frame termination sites upstream of the putative initiator methionine in the 5¢-untranslated region as well as a polyadenylation signal and a poly(A) tail at their 3¢-ends. (B) Protein sequence comparison of XGli-3 and XGli-4 (EMBL accession numbers are U42461 and U42462, respectively). Boxed regions (1–7) indicate the conserved sequence elements as defined by Ruppert et al. (1990) from a comparison of the human Gli and Gli-3 protein sequences. ZF denotes the zinc finger domain. Boxed sequence elements A, B and C denote a set of novel highly conserved sequence elements in vertebrate and invertebrate Gli-type proteins, as defined in this study. Each one of these contains a perfect copy of a PKA recognition site (Kennely and Krebs, 1991), which is in bold and underlined. Other conserved PKA consensus sequences within the same structural domain of the vertebrate Gli-type proteins (box 4) are also in bold and underlined.|
|Fig. 2. Primary sequence conservation in Gli-type and Gli-related proteins from vertebrates and invertebrates. The position of conserved sequence elements as defined by Ruppert et al. (1990) and in this study are indicated (see also legend of Fig. 1) for different Gli-type proteins. In Gli-related proteins, primary sequence conservation is restricted to the zinc finger domain.|
|Fig. 3. Temporal pattern of XGli expression in embryonic development and distribution of XGli-3/XGli-4 mRNA in adult tissue. RT-PCR reactions were carried out with equal amounts of total RNA preparations from the different embryonic stages and adult tissues (as indicated) with sets of primers specific for either XGli-3 or XGli-4, and, as a control for RNA integrity, with primers specific for histone H4 mRNA.|
|Fig. 4. Spatial distribution of XGli-3 and XGli-4 encoding mRNAs in Xenopus embryos. XGli-3 (A–H) and XGli-4 (I–S) gene transcripts were detected by whole mount in situ hybridization. XGli-3 is first detected in the anteriormost portion of the prospective neural plate in stage 12.5 (advanced gastrula) embryos (A). In the process of neurulation this primary domain of expression is maintained and extends in two lines along the anterioposterior-axis as the neural tube starts to close (stage 17). Neural tube stage embryos (stage 20) reveal strong XGli-3 expression in the area of the prospective brain, as well as weaker expression in the dorsal portion of the caudal neural tube (A–C). Xenopus tadpoles (stage 34) exhibit XGli- 3-specific signals in forebrain, midbrain and hindbrain, but also in the branchial arches, which are derived from the cephalic neural crest (D,E). Within the neural tube, XGli-3 transcripts are restricted to the ventricular zone (F–H, transverse section at the level fore-, mid- and hindbrain). XGli- 4-specific transcripts are first detected in the involuting mesodermal mantle of early and late gastrula (stage 11) Xenopus embryos (I). Axial mesoderm is excluded from the expression domain (J, transverse section). Upon folding of the neural tube (stage 16/17) a second phase of XGli-4 expression becomes apparent within the anterior portion of the neuroectoderm, whereas mesodermal expression appears to be reduced, first on the ventral side and later on the dorsal side (K, parasagittal section, and L). An area of strong mesodermal expression becomes concentrated in the posterior portion of the embryo (M). Late neurula stage (stage 18) Xenopus embryos start to develop a segmented mesodermal pattern of XGli-4 gene transcription corresponding to somites formation. Posterior and somitic mesoderm signals persist in later tailbud stage (stage 25) embryos, which also maintain XGli-4 expression in the anterior neuroectoderm (N). In stage 34 tadpole embryos, the somitic staining pattern has become most obvious, and the posterior mesodermal expression now defines the proliferating tail tip. XGli-4 signals are also visible in the anterior neural system, i.e. midbrain, forebrain and, more strongly, the rhombencephalon, as well as in the branchial arches and in the otic vesicle (O,P). The most anterior portion of tadpole stage embryos reveals an area of XGli-4 expression that is anterior to the forebrain and dorsal relative to the cement gland. This area is also visible in the anteriormost transverse section (Q), and we assume that it may correspond to the mandibular neural crest. Staining is also very pronounced in the branchial arch structures (R, transverse section at the level of the midbrain) as well as in the otic vesicles (S, transverse section at the level of the hindbrain).|
|Fig. 5. Xenopus Gli-type proteins function as transcription repressors. Various effector plasmids and combinations hereof were transfected into two different cell lines (COS-1 and HeLa) and the activity of proteins encoded by these constructs in transcription regulation measured as CAT-activity driven by the corresponding reporter plasmids. The structure of the various effector and reporter constructs is illustrated schematically in the top left of the figure. The zinc finger domains are represented by small transversally hachured boxes. The length of the various functional domains combined in these artificial fusion proteins is not to scale (see Section 4 for details). (A) XGli-3 and XGli-4 function as transcription repressors; the isolated zinc finger cluster stimulates transcription. The activity of either full-length XGli-3, XGli-4 or the Gli zinc finger cluster was tested on a promoter that contains five copies of the Gli protein DNA binding element in tandem repeat. XGli-3 and XGli-4 repress basal levels of expression in a dose-dependent manner; the zinc finger cluster alone mediates a significant stimulation of transcription. Cotransfection of increasing amounts of competitor binding sites with the XGli-3 expression vector counteracts the repressor activity of Xgli-3. (B) The repressor activity of Xenopus Gli-type proteins is mediated directly via the Gli-recognition elements. A reporter plasmid that contains multiple Gal4 binding sites is negatively regulated by a fusion protein carrying the Gal4 DNA binding domain connected to the erb-A repressor domain, and positively regulated by a fusion protein containing the Gal4 DNA binding site connected to the VP16 activator domain. XGli-3 has no effect on this reporter plasmid and similarly no effect on a reporter plasmid that contains binding sites for a different Xenopus zinc finger protein (XMyT1). (C) The Gli zinc finger cluster alone or connected to the E1A activator domain can counteract repression of transcription by Xenopus Gli-type proteins. The XGli-3 protein was coexpressed with increasing amounts of the Gli-ZF-E1A fusion protein. Transcription levels from a Gli-binding sitedependent promoter were stimulated in a dose-dependent manner. The zinc finger cluster alone exhibits an anti-repressor effect similar to the one obtained with XGli-3. The artificial repressor fusion Gli-ZF-engR leads to increased levels of repression if coinjected with XGli-3.|
|Fig. 6. Ectopic expression of the Gli-ZF-E1A activator interferes with the proper development of mesoderm- and neuroectoderm-derived structures in Xenopus embryos. (A) Representative embryos expressing the different Xenopus Gli-derived protein variants (as indicated). Microinjection of the corresponding mRNAs into one cell of a two-cell stage Xenopus embryo was performed in a mixture with lacZ mRNA in order to identify the treated cells (injected). Embryos were stained for b-gal activity (sky blue) and for expression of the pan-neural marker XMyT1 (reddish brown). Only the injected side is shown. Non-injected sides of embryos appeared to develop normally. (B) Horizontal section of Gli-ZF-E1A activator- and Gli-ZF-engR repressor-injected embryos. The b-gal staining in the injected half of the embryo highlights somitic segmentation. (C) Expression of different molecular markers for muscle development in Gli-ZF-E1A activator-injected embryos. MyoD expression or cardiac actin expression was analyzed by whole mount in situ hybridization (as indicated). IS is the injected side of the embryo, NIS is the non-injected side of the same embryo.|
|Fig. 7. Ectopic expression of the Gli-ZF-E1A activator affects neural differentiation. Xenopus embryos were injected with Gli-ZF-E1A-encoding mRNA at the two-cell stage and stained for expression of various neural markers (as indicated). NIS is the non-injected side of the embryo, IS is the injected side. Pax-6 expression in early embryos follows eye development. The number of Pax-6-expressing cells is severely reduced in the injected side of neurula stage embryos. The transverse section is at the level of the eye. cont, for control, is a non-injected embryo. MyT1/N-tub/NeuroD/Delta are molecular markers for primary neuronal differentiation. MyT1/N-tub/NeuroD expression is severely reduced in the injected half of neurula stage embryos. Transverse sections of MyT1-stained embryos are at the level of the eye and otic vesicle, respectively (as indicated). Delta mRNA levels are significantly increased in the injected half of the embryo (as identified by coinjection of lacZ mRNA). Delta-stained embryos are shown from the anterior, dorsal and posterior view, respectively. Krox-20 gene expression identifies the developing rhombomeres 3 and 5 in the hindbrain area, as well as neural crest cells originating from rhombomere 5. The transverse section is at the level of rhombomere 5. twi serves as a molecular marker for cephalic neural crest. Gli-ZF-E1A expression maintains the rhombomeric segmentation of the hindbrain, but inhibits neural crest cell migration and twi expression.|