XB-ART-2409Development February 1, 2005; 132 (4): 829-39.
Show Gene links Show Anatomy links
bHLH-dependent and -independent modes of Ath5 gene regulation during retinal development.
In a wide range of vertebrate species, the bHLH transcription factor Ath5 is tightly associated with both the initiation of neurogenesis in the retina and the genesis of retinal ganglion cells. Here, we describe at least two modes of regulating the expression of Ath5 during retinal development. We have found that a proximal cis-regulatory region of the Xenopus Ath5 gene (Xath5) is highly conserved across vertebrate species and is sufficient to drive retinal-specific reporter gene expression in transgenic Xenopus embryos. Xath5 proximal transgene expression depended upon two highly conserved bHLH factor binding sites (E-boxes) as well as bHLH factor activity in vivo. However, we found that bHLH activity was not required for expression of a longer Xath5 transgene, suggesting that additional mechanisms contribute to Xath5 expression in vivo. Consistent with this, we showed that a more distal fragment that does not include the conserved proximal region is sufficient to promote transgene expression in the developing retina. In mouse, we found that a longer fragment of the cis-regulatory region of either the mouse or Xenopus Ath5 gene was necessary for transgene expression, and that expression of a mouse Math5 (Atoh7) transgene was not dependent upon autoregulation. Thus, despite extensive conservation in the proximal region, the importance of these elements may be species dependent.
PubMed ID: 15677728
Article link: Development
Species referenced: Xenopus laevis
Genes referenced: atoh7 fos
Article Images: [+] show captions
|Fig. 1. Identification of a Xath5 regulatory fragment that functions in vivo. (A) The pG1X5 construct contains 3.3 kb of 5′ Xath5a genomic DNA cloned into the promoterless vector pG1 with a GFP reporter. The TATAA element is at –226 bp, the putative transcription start site is at– 205 bp and the translation start site is at +1 bp. (B) In situ hybridization showing that endogenous Xath5 mRNA is expressed in the retina, the pineal gland (arrowhead), and the olfactory placodes (arrows) of a stage 28 embryo (frontal view). (C,D) Frontal and lateral views of a stage 28 transgenic pG1X5 embryo showing expression of the transgene in the same tissues as the endogenous Xath5 mRNA (compare with B).|
|Fig. 3. The proximal regulatory region mimics endogenous Xath5 retinal expression in the ciliary marginal zone. (A-C) Double in situ hybridization on retinal sections from stage 42 embryos comparing expression of the GFP transgene mRNA (red; A,B) as driven by the pG1X5-proximal construct to endogenous Xath5 mRNA (purple; B,C). (D) Unlike the RNA, GFP protein is found throughout the central retina and in many cell types at stage 42. Bracket marks the CMZ, which lacks GFP fluorescence. (E-G) Higher magnification images of the ciliary marginal zone (CMZ) from A-C showing cell-by-cell correlation of GFP (red) and Xath5 (purple) mRNAs.|
|atoh7 (atonal homolog 7) gene expression in Xenopus laevis embryos, NF stage 28, as assayed by in situ hybridization, anterior view, dorsal up. Expression bilaterally in retina of the eye, in olfactory placode ( arrows) and pineal gland (arrowhead).|
|Fig. 2. Deletion analysis isolates a proximal Xath5 cis-regulatory region. (A) HindIII, DraII and PstI sites were used to delete regions of Xath5 genomic sequence in the pG1X5 construct. All three constructs drove strong transgene expression in the retina. (B) The pG1X5-proximal construct containing sequences from 27 to drove robust retinal expression of the transgene. The pG1X5-TATAA construct containing sequences from 26 to showed no expression. Xath5 sequence from 27 to 26, when fused to the Fos heterologous basal promoter (pG1-cfos-201bp-X5), drove retinal expression, but more weakly. The basal Fos promoter alone (pG1-cfos) did not confer retinal GFP expression. (C) A pG1X5-proximal transgenic embryo (stage 33) shows retinal GFP expression. (D) In situ hybridization of pG1-cfos-201bp-X5 (stage 32) transgenic embryo shows weak GFP expression in the retina.|
|Fig. 5. Proximal transgene expression is dependent upon conserved E-boxes and bHLH activity, but E-boxes alone are not sufficient for retinal expression. (A) Sequential deletions of the Xath5 proximal region: pG1X5-proximal-401 provided robust retinal GFP expression while pG1X5-proximal-389, -349 and -337 drove weaker retinal expression and at reduced frequency. (B) Mutation of either E1 or E2 (δE) reduced the percentage of embryos expressing GFP while mutation of both eliminated all transgene expression. (C-E) Injection of RNA for the dominant-negative Xath5-EnR suppressed expression of the pG1X5-proximal GFP transgene on the injected side (D) compared with the uninjected side (C). RNA encoding RFP (red) was co-injected to mark the injected side (E). (F) Multimerized E-boxes (pG1X5-TATAA+2xE1E2) were not sufficient to promote expression, while the E-boxes and adjacent α-box (pG1X5-TATAA+33 bp) promoted non-specific GFP expression throughout the CNS and head musculature (GFP+ overall), but did not promote specific retinal expression. The pG1X5-TATAA+48 bp transgenic construct promoted GFP expression in a Xath5-like pattern, but also in the axial somites, with a small percentage only showing expression in axial somites (contributing to the increase in overall GFP+ embryos). The robustness of transgene expression was reduced by mutation of the α-box with the pG1X5-TATAA+48 bp transgenic construct. (G) pG1X5-TATAA+33 bp is expressed non-specifically in CNS and muscle (H) pG1X5-TATAA+48 bp transgene is expressed in Xath5-like pattern and in axial somites. Asterisk indicates gut autofluorescence.|
|3.3 kb Xath5 transgene expression can be activated by bHLH factors, but is not dependent upon conserved E-boxes or bHLH activity. (A-C) Injection of RNA encoding RFP alone did not alter transgene expression on the injected side. (D-E) Injection of RNA for Xath5 ectopically activated transgene expression on the injected side (bracket in E) compared with the uninjected side (D). RFP marks the region targeted on the injected side (bracket in F). (G) Mutation of E1 and E2 (δE1,2) in the pG1X5-3.3 kb transgene did not eliminate retinal transgene expression. Mutation of two additional conserved E-boxes, E3 and E4, also did not abolish transgene activity. (H-J) Injection of RNA for the dominant-negative Xath5-EnR did not suppress expression of the pG1X5-3.3 kb transgene on the injected side (I) compared with the uninjected side (H). RNA encoding RFP (red) was co-injected to mark the injected side (J).|
|Fig. 7. The distal cis-regulatory region of Xath5 alone is sufficient to promote transgene expression in the developing retina. (A) A PstI fragment that lacks the conserved proximal region promotes retinal transgene expression when coupled to either the pG1X5-TATAA basal promoter (pG1X5 distal + TATAA) or to the Fos basal promoter (pG1X5 distal + Fos). (B) Fluorescent image of a stage 30 pG1X5distal+TATAA transgenic embryo showing expression of the transgene in the retina.|
|Fig. 8. Cross-species analysis of Ath5 transgene expression. (A) The proximal 600 bp of the Math5 cis-regulatory region drives weak GFP expression in the retina of transgenic frog embryos, as shown by in situ hybridization on a pG1M5-0.6 kb transgenic Xenopus embryo. (B) GFP in situ hybridization in a pG1M5-2.3 kb transgenic embryo. The 2.3 kb Math5 fragment drives strong transgene expression in the retina, cranial ganglia, midbrain and hindbrain regions in transgenic Xenopus embryos. (C) In situ hybridization on retinal sections from stage 41 pG1M5-2.3 kb transgenic embryos shows that the domain of GFP mRNA is restricted to the CMZ (bracket). (D) E13.5 whole embryo image demonstrating retinal expression of the pG1M5-2.3 kb transgene in mouse embryos at E13.5 (arrow). (E) The Xenopus 3.3 kb Xath5 transgene (pG1X5-3.3 kb) also shows retinal expression at E13.5. (F,G) Math5 is not required for expression of the pG1M5-2.3 kb transgene as equivalent fluorescence from the transgene was observed between wild-type (F) and Math5(G) embryos. Scale bar: 500 μm in D.|