Nucleic Acids Res
November 1, 2008;
CRX controls retinal expression of the X-linked juvenile retinoschisis (RS1) gene.
X-linked juvenile retinoschisis is a heritable condition of the retina
in males caused by mutations in the RS1
gene. Still, the cellular function and retina
-specific expression of RS1
are poorly understood. To address the latter issue, we characterized the minimal promoter driving expression of RS1
in the retina
. Binding site prediction, site-directed mutagenesis, and reporter assays suggest an essential role of two nearby cone-rod homeobox (CRX
)-responsive elements (CRE) in the proximal
promoter. Chromatin immunoprecipitation associates the RS1
promoter in vivo with CRX
, the coactivators CBP
and acetylated histone H3. Transgenic Xenopus laevis expressing a green fluorescent protein (GFP) reporter under the control of RS1
promoter sequences show that the -177/+32 fragment drives GFP expression in photoreceptors and bipolar cells. Mutating either of the two conserved CRX
binding sites results in strongly decreased RS1
expression. Despite the presence of sequence motifs in the promoter, NRL
appear not to be essential for RS1
expression. Together, our in vitro and in vivo results indicate that two CRE sites in the minimal RS1
promoter region control retinal RS1
expression and establish CRX
as a key factor driving this expression.
Nucleic Acids Res
Wheat Germ Agglutinin
[+] show captions
References [+] :
Figure 1. Analysis of conserved transcription factor binding sites in the RS1 promoter and delineation of the proximal region. (A) Sequence alignment of the putative RS1 promoter regions from human, dog and mouse. Conserved sequences are highlighted in gray and canonical binding sites for CRX (CRE1 to 3), NR2E3 and NRL (NRE) are boxed. (B) DNA sequence of the putative human RS1 promoter. Potential cis-elements are indicated. An interspersed ALU repeat is italicized. (C) Basal activity of three RS1 promoter–luciferase constructs (−703/+32, −419/+32, −177/+32) and of a PDE6A positive control (−300) transfected into Y79 retinoblastoma cells. RS1 promoter activities are comparable to the PDE6A promoter, which is known to be active in Y79 cells. A schematic of the RS1 promoter is shown with transcription factor binding sites and the ALU repeat (box). Each transfection was repeated three times with duplicate wells analyzed. Error bars represent the standard deviation of the mean from protein normalized luciferase activities.
Figure 2. RS1 promoter induction levels by CRX, OTX2, NR2E3 and NRL. HEK293 cells were cotransfected with the −177/+32 RS1 promoter–reporter construct and transcription factor expression vectors. X-fold stimulation values were calculated by normalizing CRX, OTX2, NRL and NR2E3 cotransfected cells to mock-vector transfections. CRX (black bars) specifically induces the RS1 promoter in a dose-dependent manner. OTX2 (dark gray bars), NR2E3 (light gray bars) and NRL (white bars) do not induce or suppress the RS1 promoter. Each transfection was repeated three times with duplicate wells analyzed. Error bars represent the standard deviation of the mean from protein normalized luciferase activities.
Figure 3. CRX, HATs and AcH3 are associated with the RS1 promoter region. (A and B) EMSAs with Y79, mouse retina, BV-2 microglia nuclear extracts or in vitro translated transcription factors. The sequences used for EMSA analysis are shown below with the mutated nucleotides underlined and conserved motifs printed in bold letters. Each gel was loaded with 5 µg of nuclear extracts or 4 µl of in vitro translated proteins, as indicated. Arrows represent specific binding identifiable by shifted bands that were inhibited by excess unlabeled weight or consensus oligonucleotides, but not by their mutant counterparts. Incubation with BV-2 microglia cells nuclear extract (M) did not result in a specific band shift. Addition of 1 µl anti-CRX antibody did not produce a supershift, but specifically inhibited CRX binding to the oligonucleotide. Asterisks indicate unspecific bands. (A) Interaction of RS1 CRE1 and (B) CRE3 with proteins in Y79 cells, mouse retina, BV-2 cells (M) and in vitro translated CRX (IVT). (C) ChIP assays using wild-type Y79 cells or Y79 cells transfected with CRX, NR2E3 or NRL. Immunoprecipitation was carried out with antibodies against CRX, NR2E3 or NRL. Input DNA served as positive control and IP with rabbit IgG antibody served as negative control. The RS1 promoter fragment −177/+130 was analyzed for in vivo CRX, NR2E3 and NRL binding by ChIP–PCR. (D) ChIP assays using wild-type or CRX-transfected Y79 cells with antibodies against CRX, CBP, P300, AcH3 and GCN5. Input DNA served as positive control and IP with rabbit IgG antibody served as negative control. DNA fragments were analyzed by PCR for two different RS1 promoter fragments and the red-cone opsin promoter. The RS1 promoter fragment −177/+130 and the red-cone opsin control promoter displayed ChIP-PCR positive signals whereas the RS1 promoter region −703/−530 lacking CRX sites did not show specific CRX binding.
Figure 4. Delineation of the critical sequences required for CRX binding and RS1 promoter activity. Luciferase assays were performed with the RS1 wild-type promoter construct −177/+32, and CRE single-, double- and triple mutant promoters in Y79 cells (A) and HEK293 cells cotransfected with CRX (B). The basal luciferase activity of the wild-type contruct −177/+32 (A) or the CRX-induction of the wild-type contruct −177/+32 was set to 100%. Each transfection was repeated three times with duplicate wells analyzed. Error bars represent the standard deviation of the mean from protein normalized luciferase activities.
Figure 5. In vivo analysis of RS1 promoter–GFP constructs. Confocal micrographs of transgenic X. laevis retinae expressing GFP (green) under the control of various promoters. Cryosections were counterstained with wheat germ agglutinin (red) and Hoechst nuclear stain (blue). (A) Nontransgenic control retina (NON-TG); (B) Xenopus opsin promoter (XOP); (C) RS1 promoter wild type (RS1-WT); (D) RS1 promoter with mutated CRE1 (mCRE1); (E) RS1 promoter with mutated CRE2 (mCRE2); (F) RS1 promoter with mutated CRE1 and CRE3 (mCRE13); (G) RS1 promoter with mutated NRE (mNRE); (H) RS1 promoter with mutated NR2E3 (mNR2E3). GFP was not expressed in the nontransgenic retina but was highly expressed in photoreceptors in the XOP retina. The wild-type RS1 promoter (NON-TG) drives GFP expression in photoreceptors albeit at lower levels. Abolishing NRL and NR2E3 binding did not significantly alter GFP expression (G and H). Mutating the CRE1 site (D and F) reduced GFP expression to below detectable limits, but no obvious effect resulted from mutating the CRE2 site (E). The GFP signals in panels A and C-H are directly comparable as the same laser intensity and amplifier setting were used, and identical image processing was applied. In panel B, the amplifier gain was reduced due to the intensity of the GFP signal. The inset shows an image acquired with settings equivalent to the other panels. Photoreceptor outer segments (os), outer nuclear layer (onl), inner nuclear layer (inl) and inner plexiform layer (ipl). Scale bar, 50 µm.
Figure 6. RS1 promoter activity in bipolar cells. (A–D) Confocal micrographs of transgenic retinas expressing GFP. (A) Within the retina, the X. laevis opsin promoter drives expression exclusively in rod photoreceptors. In contrast, the RS1 promoter (B) drives expression in both photoreceptors and bipolar cells (asterisks). Mutant RS1 promoters lacking NR2E3 (C) or CRE2 (D) sequences were also capable of driving expression in bipolar cells. Photoreceptor outer segments (os), inner segments (is), outer nuclear layer (onl), outer plexiform layer (opl), inner nuclear layer (inl) and inner plexiform layer (ipl). Scale bar, 20 µm.
Targeting of GFP to newborn rods by Nrl promoter and temporal expression profiling of flow-sorted photoreceptors.