Deciphering the Contribution of Known cis-Elements in the Mouse Cone Arrestin Gene to its Cone-Specific Expression

Abstract

purpose. Cone arrestin (CAR) is highly expressed in all cone photoreceptors of the retina and a subset of pinealocytes in the pineal gland. This study was initiated to examine the cis-elements responsible for the cell-specific expression pattern of CAR.

methods. Mutagenesis and specific deletions of known cis-elements in the proximal promoter of the mouse CAR (mCAR) gene were introduced and analyzed in vitro and in vivo. A series of mCAR promoter-luciferase reporter constructs were transiently transfected into COS-7 or Weri-Rb-1 retinoblastoma cells and tested in in vitro promoter assays. Transgenic Xenopus laevis were created with deletional or mutated promoter fragments driving an enhanced green fluorescent protein (EGFP) reporter gene. The resultant EGFP expression pattern in the transgenic animals was analyzed by fluorescence microscopy and immunocytochemistry.

results. A significant decrease in in vitro transcriptional activities was observed when the minimal 215-bp promoter fragment was mutated in each of the four cone-rod homeobox (CRX)-binding elements (CBEs) or in either of the two TATA-elements. The 215-bp mCAR proximal promoter drove EGFP expression to cone photoreceptors and pinealocytes in transgenic Xenopus laevis; however, the truncated 147-bp fragment drove expression to both cone and rod photoreceptors. Transgenic tadpoles carrying a single mutation in either the TATA-box, the TATA-element or the proximal CBE had undetectable EGFP expression in the retina. However, when one of the other three CBEs was mutated, EGFP expression was observed in muscle and brain tissues, in addition to the eyes. Also, when both TATA elements were mutated, transgenic animals had EGFP expression in all photoreceptors. Because no reporter activity was observed when either a 3.2-kb 5′ extended region or the first intron of the mCAR gene was tested in Weri-Rb-1 cells, neither construct was examined in vivo.

conclusions. The data demonstrate that the regulatory functions of the known cis-elements in the mCAR promoter are highly dependent on location and nucleotide sequence conservation. The TATA elements and CBEs are crucial for driving both basal transcriptional activity and tissue specificity to cone photoreceptors and pinealocytes.

Although many aspects of cone photoreceptors have been extensively explored, the molecular mechanisms that control cone photoreceptor-specific gene expression remain elusive. The regulatory 5′ flanking region that controls the cellular expression of rhodopsin, ¹ interphotoreceptor retinoid-binding protein (IRBP), ^{2 3} β-phosphodiesterase (PDE), ⁴ arrestins, ^{5 6 7 8} and cone opsins ⁹ have been characterized in vitro. Several cis-elements in the upstream regions of photoreceptor-specific genes and a hierarchy of photoreceptor-specific factors regulating gene expression and cell fate determination have been identified, including cone-rod-homeobox (CRX), ^{10 11} neural retina leucine zipper (NRL), ¹² thyroid hormone receptor β2 (TRβ2), ¹³ and photoreceptor-specific nuclear receptor (NR2E3). ^{14 15}  

Because cone arrestin (CAR) is highly expressed in all cone photoreceptor cells and a subset of pinealocytes, we explored the mechanisms of cellular enriched expression. A 215-bp proximal promoter region, located within the region from −178 to +37 bp in the mouse CAR gene, is necessary and sufficient to drive cone photoreceptor-specific expression of an EGFP reporter gene in transgenic Xenopus laevis. ⁵ Within this short 215-bp fragment, four CRX-binding elements (CBEs), one TATA box, one TATA-like element, and one Sp1 site were noted. Many photoreceptor-specific or enriched genes have CBEs in their promoter regions, such as rhodopsin, rod and cone arrestin, phosducin, transducin-α, PDE-β, IRBP, guanylyl cyclase 1 and 2, ¹⁶ and CRX itself. ¹⁷ We propose that multiple CBEs combined with a TATA-box are the key elements for directing appropriate cone-specific expression. ⁵  

In this study, transgenic Xenopus laevis carrying the mutated mCAR promoter–EGFP reporter constructs were created to evaluate the contribution of each known cis-element to targeting cone-specific expression. Immunohistochemistry was undertaken with an antibody recognizing the specific Xenopus CAR (xCAR) protein to confirm the location and pattern of EGFP cellular expression. Paralleling the in vivo experiments, in vitro transient transfection and luciferase assays were performed to verify the promoter activities in cultured retinoblastoma cells. Our results demonstrate that one of the TATA elements and the most proximal CBE are required for both basal activation and cone-specific expression of the mCAR gene. Also, one of the other three CBEs is required for optimum promoter activity. 

Materials and Methods

Plasmid Construction

A 3.2-kb mCAR 5′-flanking fragment (−3211/+37) was amplified from the mCAR genomic BAC clone 16505, ⁵ by using a pair of mCAR-specific primers with restriction endonuclease sites (italic): +mCAR(−3170 to −3148)XhoI: 5′-GGCTCGAGAGTTTCTGCTTGTTTAAAGTCCC-3; and −mCAR(37-17)EcoRI: 5′-CCGAATTCGAGGGAAGAGATGAAGCTCGC-3′. 

The amplified 3.2-kb mCAR promoter fragment contains the basic promoter region (−523/+37) ⁵ and 2.6-kb farther upstream. A second 1.6-kb fragment (−523/+1091) containing the basic promoter region (−523/+37) and the first intron of the mCAR gene ⁵ were also amplified from the mCAR BAC16505 clone ⁵ with the following primer pairs: +mCAR(−484 to −461)XhoI: 5′-GGCTCGAGCTTTTTTTTTACCTTTTTGGTTCC-3′; and −mCAR(1091-1068)EcoRI: 5′-CCGAATTCGCTGGCCAGTTGAATCTTTCC-3′. 

The polymerase chain reaction (PCR) products were directly digested with the restriction endonucleases EcoRI and XhoI, and unidirectionally ligated into the pRL-Null Renilla luciferase reporter vector (Promega, Madison, WI) through its EcoRI and XhoI sites. 

The 5′-progressive deletion promoter fragments were amplified from the 215-bp mCAR promoter (−178/+37) ⁵ by PCR with the following primer pairs: +mCAR(−144 to −120): 5′-GTTTTCCAAACTCTTGGCCCCTGAG-3′; +mCAR(−110 to −86): 5′-GCAGTGCTTCATGCCACCCACCCCA-3′; and −mCAR(37-12): 5′-GAGGGAAGAGATGAAGCTCACATCAG-3′. 

The 181-bp (−144/+37) and 147-bp (−110/+37) PCR products were ligated into a cloning vector (TOPO TA; Invitrogen, Carlsbad, CA) and then subcloned into the pEGFP-1 (BD Biosciences-ClonTech, Palo Alto, CA) or pRL-Null vectors (Promega) at their EcoRI sites. 

The predicted cis-elements within the 215-bp promoter fragment were mutated by site-directed mutagenesis PCR, as previously published. ¹⁸ The nucleotide sequences of the primers used in mutagenesis PCR are shown in Table 1 . The predicted CBEs were mutated from CTAATCT and CTAATTA to CTggaCT and CTggaTA, respectively. According to published data, ² the core nucleotides of the CBE are “TAAT”, and mutation of these core nucleotides effectively suppresses the promoter activity. The TATA-like element (TTATA), TATA-box (TATAA), and GC-rich Sp1 site (GAGGCTGGG) were mutated to TtgcA, TgcAA, and GaaatTGaG, respectively. All the promoter fragments with mutant cis-element(s) were subcloned into either the pEGFP-1 or pRL-Null vector at its EcoRI site. All the constructs were sequenced on an automatic sequencer (model 310; Applied Biosystems, Foster City, CA) to confirm correct orientation and appropriate introduction of each deletion or mutation. 

Transient Transfection Assays

Weri-Rb-1 retinoblastoma and COS-7 cells (American Type Culture Collection, Manassas, VA) were maintained as described previously. ^{19 20} Transfection was performed with cationic lipid (Lipofectamine; Invitrogen). Weri-Rb-1 cells grown in six-well plates were transiently transfected with 3 μg of the mCAR promoter-luciferase reporter construct or the promoterless pRL-Null vector and 1 μg of the pGL3-P plasmid (Promega) carrying the Firefly luciferase reporter gene under the SV40 basic promoter as an internal control for transfection efficiency. ⁵ COS-7 cells in six-well plates were cotransfected with 1 μg of mCAR promoter-pRL-Null construct, 1 μg of the pGL3-C plasmid (same as pGL3-P except has an additional SV40 enhancer; Promega) and 2 μg of either the control pcDNA3 vector or the vector containing the bovine CRX (bCRX) expression construct. ⁵ Cells were incubated for 45 hours before being harvested. Both Firefly and Renilla luciferase activities were assayed with 20 μL of crude cell lysate (Dual-Luciferase Reporter Assay System; Promega) and a luminometer (TD-20/20; Turner Designs, Sunnyvale, CA). 

Production of Transgenic Xenopus laevis

All procedures involving animals were approved by the institutional animal care and use committee and were performed according to the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. Adult Xenopus laevis were purchased from NASCO (Fort Atkinson, WI), and eggs and sperm were obtained from the adult frogs, as previously described. ²¹ Transgenic tadpoles harboring mCAR promoter-EGFP were produced using restriction enzyme-mediated integration (REMI) in which the transgene is stably inserted into the sperm genome, followed by fertilization of eggs with the sperm nuclei. ^{21 22} Each mCAR promoter-EGFP construct was linearized at their XhoI site. The REMI reaction was prepared by mixing 1 μL of linearized plasmid (250 ng/μL) with 4 μL of sperm nuclei (∼4 × 10⁴ nuclei). After a 5-minute incubation at room temperature, 1 μL of 100 mM MgCl₂, 0.5 μL of a 1:25 dilution of XhoI (12 U/μL), 2 μL oocyte extract (preheated at 80°C for 10 minutes, centrifuged at 10,000 rpm for 2 minutes), and 9 μL of sperm dilution buffer (250 mM sucrose, 75 mM KCl, 0.5 mM spermidine trihydrochloride, and 0.2 mM spermidine tetrahydrochloride) was added, and the mixture was incubated for 10 minutes at room temperature. The reaction mixture was diluted in sperm dilution buffer to a final concentration of approximately three sperm nuclei in 5 nL, and then injected into dejellied Xenopus laevis eggs at a rate of 5 nL/sec, with a glass needle and a syringe pump (model 11; Harvard Apparatus, Holliston, MA). 

Normal developing embryos were maintained in 12-hour light–dark cycles until they reached 2 weeks of age. Xenopus laevis carrying the transgene were identified by genomic PCR, by using a sense primer in the mCAR promoter sequence (+mCAR[−52 to −32] 5′-ACACATTCAGAGGATTGTGGA-3′), and an antisense primer in the EGFP sequence located in the pEGFP-1 vector from 163 to 143 (5′-CGTCGCCGTCCAGCTCGACCAG-3′). 

Antisera Generation and Characterization

Rabbit antisera against a peptide of Xenopus cone arrestin (xCAR, residues 372-385, RQKLQGAEGEDDKD; Luminaires Orange County [LUMIoc]) ²³ were generated for our research project by Zymed Laboratories (South San Francisco, CA) and affinity purified against the peptide with a kit (SulfoLink; Pierce, Rockford, IL), as previously described. ¹⁹ The purified xCAR peptide antibody was tested by immunoblot analysis of selected Xenopus tissue homogenates and by immunohistochemistry of Xenopus retinal sections, as previously described. 1D4 monoclonal antibody is from Chemicon (Temecula, CA). 

Immunohistochemistry

To analyze the EGFP expression pattern in the retinas of transgenic Xenopus laevis, tadpoles (1–2 months of age) or froglets (∼4 months of age) were fixed in 4% paraformaldehyde in phosphate buffer (PBS) for 2 hours and transferred to 30% (wt/vol) sucrose in PBS for overnight incubation. The whole head was cut and embedded in tissue-freezing medium (OCT; Sakura Finetek, Torrance, CA). Frozen sections of 10 μm thickness were prepared. To visualize Xenopus cones, sections were subjected to double-immunofluorescence sequential staining. ^{24 25} After blocking, the sections were incubated with the first primary antibody (anti-GFP antibody, BD Living Colors Aequorea victoria Peptide Antibody; BD Biosciences-ClonTech, Palo Alto, CA) at 1:100 dilution and then incubated with a fluorescein anti-rabbit IgG. An incubation with anti-rabbit IgG F(ab) fragments (20 μg/mL in PBS) for 2 hours at 37°C, followed by incubation in 4% paraformaldehyde for 20 minutes at room temperature was performed to block the anti-GFP antibody to prevent cross-reaction with subsequently applied reagents. After the treatment, the second primary antibody LUMIoc (anti-xCAR) at 1:500 dilution was added, followed by a Texas red anti-rabbit IgG. Sections without either the first or the second primary antibody or both were used as the negative control. Fluorescence images were obtained with a digital camera (Spot; Diagnostic Imaging, Sterling Heights, MI) mounted on a fluorescence microscope (Leica Microsystems, Bannockburn, IL). The illustrations in the figures were prepared on computer (Photoshop, ver. 5; Adobe Systems, San Jose, CA). 

Results

Promoter Activities of the mCAR 5′ Extended Region and First Intron

Previously, transcriptional activity was observed in Weri-Rb-1 retinoblastoma cells with the mCAR promoter of up to 1003 bp, and the highest level is in the mCAR p-523 (−484/+37) construct. ⁵ To search for other enhancer or inhibitory elements in the 5′ extended region and the first intron, further upstream areas and the first intron were analyzed for promoter activities in transient transfection studies. In Weri-Rb-1 cells, neither the p-mCAR(−3170/+37) nor the p-mCAR(−484/+1091) had detectable promoter activities, whereas both the p215 (−178/+37) and the p523 (−484/+37) had significant reporter expression (Fig. 1A) . In COS-7 cells with CRX cotransfection, p-mCAR(−3170/+37) drove reporter expression at the same level as p215 (−178/+37) and p523 (−484/+37), but p-mCAR(−484/+1091) lacked detectable transcriptional activity (Fig. 1A) , suggesting that no additional CBE exists in the farther upstream region or the first intron of the mCAR gene. None of the mCAR promoter constructs drove detectable luciferase expression in COS-7 cells without CRX cotransfection, demonstrating that CRX is essential for mCAR gene transcription. These results suggest that no significant enhancer elements exist in the 3.2-kb 5′ extended region, but there may be transcriptional inhibitory sites within the first intron. 

Contribution of Each cis-Element to Promoter Activities In Vitro

The p215 mCAR promoter contains a TATA box, a TATA-like element, four CBEs (designated from 5′ to 3′ as 1-, 2-, 3-, and 4CBE), and an Sp1 site. ⁵ To evaluate the contribution of each predicted cis-element to cone-specific transcriptional activity, promoter-luciferase reporter constructs harboring either deleted or mutated cis-elements were created and transiently transfected into Weri-Rb-1 retinoblastoma cells. The 181-bp deletion construct lacking the two most distal CBEs completely lost promoter activity, whereas the 147-bp fragment, which lacks the distal three CBEs and the TATA-like element, had promoter activity comparable to the p215 construct (Fig. 1B) . With a single mutation of each of the CBEs and the TATA elements, decreased transcriptional activities were observed, compared to the transcriptional activities of the 215-bp promoter. Promoter activities were totally abolished when one of the first, third, or fourth CBE; the TATA-box; and the TATA-like element was mutated (Fig. 1B) . 

We also tested promoter activities of double or multisite mutation fragments. The results demonstrated that the transcriptional activities were totally lost when the 215-bp mCAR promoter was simultaneously mutated at multiple sites (Fig. 1B , double-TAm, 12CBEm, 123CBEm, 123CBE&TA-like, and 3CBE&TA-like). 

In Vivo mCAR Promoter Activities in Transgenic Tadpoles

To evaluate whether the in vitro mCAR promoter activities were reflections of in vivo activities, we tested the contribution of each cis-element to the mCAR promoter activities in transgenic Xenopus laevis harboring the mCAR promoter-EGFP reporter constructs. Transgenic tadpoles (at 4 months of age) were screened for EGFP expression and subsequently genotyped. 

The expression pattern and levels of EGFP in various tissues (eyes, pineal, brain, nervous system, and muscle) in the transgenic animals are summarized in Figure 2A . Representative EGFP transgenic Xenopus images are shown in Figure 2B . The wild-type p215 mCAR promoter drove medium levels of EGFP expression in eyes and the pineal gland. ⁵ Weak EGFP expression was observed in jaw muscles in tadpole stages but disappeared as the tadpoles developed into froglets. This nonspecific pattern of expression in tissues other than retina and pineal has been reported and may be due to cross-species mismatch. ⁵ The 181-bp truncated promoter construct (n = 6) drove low levels of EGFP expression in the eyes, optic nerve (Fig. 2Ba) , and higher levels in tail muscles (Fig. 2Bb) . The 147-bp truncated promoter (n = 10) drove low to high levels of EGFP expression in the eyes, tail muscles, brain (Fig. 2Bc) , and spinal cord ganglia (Fig. 2Bd) . 

With a single mutation of one of the three distal CBEs (1CBEm, n = 10; 2CBEm, n = 4; 3CBEm, n = 3), weak to medium EGFP expression was retained in the eyes but was also detected in muscles and brain. When the most proximal CBE was mutated (4CBEm, n = 13), EGFP expression was not detectable in any tissues. When either the TATA-like element (n = 9) or the TATA box (n = 8) was mutated, EGFP expression was not observed in the eye and the pineal gland, although weak expression was detected in the brain, tail muscles, and heart. However, when both the TATA-like element and the TATA-box were mutated (Double-TAm, n = 3), EGFP expression appeared again in the eyes but not in the pineal gland (Fig. 2Be) , and there was weaker EGFP expression in heart, brain, and tail muscle. When the Sp1 site was mutated (Sp1m, n = 5), four of five tadpoles observed had no EGFP expression in the eye and the pineal gland but had low expression in tail muscles. 

When the distal two CBEs were simultaneously mutated (12CBEm, n = 4), tadpoles had low to medium levels of EGFP expression in the eye, jaw muscle (Fig. 2Bf) , brain, and tail muscles. When the distal three CBEs were all mutated (123CBEm, n = 12), no detectable EGFP expression was observed in any tissues in 8 of the 12 transgenic tadpoles examined, whereas the other 4 tadpoles had low EGFP expression in muscles only. When the first three CBEs plus the TATA-like element were all mutated (123CBE&TA-like, n = 4), only weak EGFP was observed in tail or jaw muscles. When the third CBE and TATA-like element were mutated (3CBE&TA-like, n = 11), low to medium levels of EGFP expression was observed in the eye, tail muscles, brain, and spinal cord ganglia. 

Retinal Immunohistochemistry

A unique xCAR-specific peptide polyclonal antibody, LUMIoc, was generated to visualize the immunoreactive cone arrestin in Xenopus cone photoreceptors and pinealocytes. The specificity of the xCAR antibody was tested by immunoblot analysis of various Xenopus tissue homogenates and immunohistochemistry of adult Xenopus retinal frozen sections. A major immunoreactive protein band of approximately 44 kDa was detected only in the retina (and pineal after longer exposure, data not shown) but below detectable levels in other tissues examined (Fig. 3A) . Immunohistochemical analysis showed that all the CAR-immunoreactive cells had a colorless oil droplet in their myoid regions, confirming they were cone photoreceptors (Fig. 3B) . The anti-bovine rhodopsin monoclonal antibody 1D4 labeled Xenopus rod outer segments, as well as small single cones, as previously published. ²⁶  

To monitor the cellular expression pattern of EGFP in transgenic Xenopus laevis retina, cryosection slides were prepared from transgenic animals, and immunofluorescence double labeling was performed using the anti-EGFP and anti-xCAR primary antibodies (Fig. 4) . In transgenic tadpoles harboring the wild-type 215-bp mCAR promoter fragment, EGFP expression completely overlapped with xCAR expression (Fig. 4B) , suggesting cone-specific EGFP expression as previous published. ⁵ The 147-bp truncated promoter fragment drove EGFP expression in both rod and cone photoreceptor cells (Fig. 4C) . The 181-bp truncated promoter fragment and the single CBE mutation promoter fragments (1, 2, and 3 CBEm) did not drive detectable levels of EGFP expression in the retina (data not shown), although weak green fluorescence was observed in the eyes of most of the transgenic animals carrying these promoter constructs (Fig. 2A) , probably because of the sensitivity limit of the EGFP antibody or the extraretina EGFP expression in the eye. When the fourth CBE (Fig. 4D) , the TATA-like element (Fig. 4E) , or the TATA-box (Fig. 4F) was mutated, EGFP expression was not detected in the retina. When both the TATA box and the TATA-like element were mutated, EGFP expression was observed in both rods and small single cones (Fig. 4G) . 

Discussion

Potential Regulatory Role of the TATA Elements in the mCAR Promoter

Among photoreceptor-specific genes characterized to date, all cone-specific genes, such as human, bovine and murine red/green pigments, ⁹ ground squirrel green opsin gene (Yan W, et al. IOVS 2003;44:ARVO E-Abstract 3525) and cone transducin-α subunit, ²⁷ have a typical TATA box, whereas other photoreceptor-specific genes including rod-specific genes, such as rod arrestin, ²⁸ IRBP, ² and β-PDE, ²⁹ have no typical TATA box. A TATA-like element, in addition to a typical TATA box, exists in the mCAR proximal promoter region. Mutation of either TATA element results in a decrease of luciferase reporter expression, whereas mutation of both TATA elements causes complete loss of reporter expression in Weri-Rb-1 cells (Fig. 1B) . In transgenic Xenopus laevis, mutation of either TATA element leads to undetectable EGFP expression in the eye (Fig. 2) . This may be due to the much more sensitive luciferase reporter assay used in the in vitro experiments compared with the EGFP fluorescence detection in vivo. However, double TATA element mutations result in high levels of EGFP expression in rod photoreceptors, in addition to a low level of cone expression (Fig. 4G) . These data suggest that both TATA elements are functional and complement each other in controlling the cone-specific expression of the CAR gene. When both TATA elements are mutated, the mCAR promoter becomes a TATA-less promoter, similar to rod-specific gene promoters, and, indeed, this TATA-less promoter drives high levels of reporter expression in rod photoreceptors in transgenic Xenopus. Because Weri-Rb-1 is a cone-like cell line, ^{20 30 31} the TATA-less mCAR promoter does not drive reporter expression in these cells. 

It has been reported that the TATA-binding protein (TBP) and associated factors (TAF) selectively activate promoters. ³² In in vitro experiments, overexpression of exogenous TFIID strikingly enhanced the expression of a TATA-containing promoter. In contrast, a TATA-less promoter showed no response with the addition of TFIID. ³³ Cone photoreceptor cells may have higher basal levels of TBP, TAF, and TFIID than rods, which selectively activate TATA-containing promoters but not TATA-less promoters. 

Contribution of the CBEs to the Regulation of mCAR Gene Expression

CRX regulates both rod- and cone-specific gene expression. ^{10 11 17} Four potential CBEs exist in the mCAR proximal promoter region, and addition of exogenous CRX activates the mCAR proximal promoter in COS-7 cells. ⁵ Although all four CBEs in the mCAR promoter region were individually mutated, complete loss of reporter expression both in vitro and in vivo occurred only when the most proximal CBE was mutated. This provides clear evidence that the most proximal CBE is the critical element controlling the transcription of the CAR gene. 

Because the proximal CBE is only 21 bp upstream of the TATA box, the binding of CRX to this site may facilitate the recruitment of other essential transcriptional coactivators working in concert to drive the basal transcription machinery. Retina-specific transcription of CRX-regulated genes correlates with hyperacetylation of histones on their promoter/enhancer region, which requires CRX and its interaction with histone acetyl-transferases-containing coactivator complexes (Peng G, et al. IOVS 2004;45:ARVO E-Abstract 649). Direct binding of CRX to the mCAR promoter was recently confirmed by chromatin immunoprecipitation with the use of an anti-CRX antibody followed by PCR with a pair of mCAR promoter-specific primers (Chen S, et al. IOVS 2004;45:ARVO E-Abstract 2253). 

Mutation of each of the three remaining CBEs results in promoter activity decrease both in vitro and in vivo, suggesting a dosage effect of the multiple CBEs. When all three CBEs are mutated simultaneously, EGFP expression is completely aborted in the eyes. These results suggest that at least one of the three more distal CBEs is required, in addition to the proximal one, for activation of the mCAR gene in vivo. 

Of note, when the distal two or three CBEs were either deleted or mutated, reporter expression was observed in brain, muscle, and other tissues, as well as rod photoreceptors, in addition to cones, suggesting that these distal CBEs may function to repress gene expression in cell types other than cones. DNA footprint and gel-shift assays using wild-type and mutated promoter fragment and nuclear extracts from different tissues are currently underway to confirm these observations. 

Contribution of the Sp1 Site to mCAR Gene Regulation

In the 215-bp mCAR promoter, the Sp1 site is located downstream of the TATA box and is essential for full promoter activities of the mCAR 5′ flanking region both in vitro and in vivo. The Sp family of transcription factors contains eight members (Sp1 to -8), each containing conserved zinc finger DNA-binding domains near their C termini and glutamine-rich domains adjacent to the serine/threonine stretches in the N-terminal regions. Among them, Sp1, -3, and -4 are closely related and have highly conserved DNA-binding domains. They bind GC boxes and are required for the expression of housekeeping, tissue-specific, and viral genes. ³⁴ Their DNA-binding domains allow them to bind with identical affinity to the consensus GC box. ³⁵ GC boxes (Sp sites) have been reported to be involved in the regulation of retinal expression of many genes, including the rod PDE α-, ³⁶ β-, ^{4 29} and γ-subunits ³⁷ and the dominant drusen gene Fibulin 3. ³⁸ Further studies are needed to identify the specific Sp family member responsible for the mCAR gene regulation. 

Other Potential Elements

Transcription regulatory elements have been identified in distal upstream regions, introns, or downstream of the coding sequence. ³⁹ For example, the mammalian red/green pigments have an essential distal region beyond 4 kb upstream of the transcription start site for expression. ^{9 40} In our studies, the basic mCAR promoter plus the region approximately 3 kb farther upstream drove reporter expression in COS-7 cells with CRX cotransfection at a similar level to the 215- and 523-bp promoter fragments, suggesting that no further CBEs exist in this upstream region. This was confirmed by sequence alignment analysis (data not shown). It is intriguing that this construct did not drive reporter expression in Weri-Rb-1 cells, implying the existence of potential retina-specific inhibitory elements in the upstream region, although sequence alignment with upstream regions of other known retinal genes ^{9 40} did not reveal any significant conserved region in the mCAR upstream region. The basic mCAR promoter with the first intron did not drive detectable reporter expression in either COS-7 cells cotransfected with CRX or Weri-Rb-1 cells, suggesting that a universal inhibitory element may be present in the first intron of the mCAR gene. Future studies with progressive deletion constructs of the extended regions will identify the potential inhibitory elements. 

In this study, we successfully used transgenic Xenopus to study the mCAR promoter elements. In contrast to the paucity of cone photoreceptors in mouse, 45% of photoreceptors in Xenopus laevis are cones. Moreover, the transgenic Xenopus is faster and less expensive to produce than the transgenic mouse, making the Xenopus an excellent model for the initial screening of cone-specific transcription elements. However, due to the cross-species nonspecific binding between promoter elements and transcription factors, as observed in this study and previously reported studies, the data obtained with transgenic Xenopus need to be verified in mice. We are currently setting up the in vivo electroporation technique in our laboratory, a technique that has been successfully used to study promoter activities, and to introduce short interfering RNA (siRNA) into retinal photoreceptors to silence gene expression. ⁴¹ The same wild-type and mutant mCAR promoter-EGFP constructs will be tested by in vivo electroporation into the mouse retina to verify our results observed in the Xenopus laevis. 

 Supported by National Eye Institute (NEI) Grants EY00395 (CMC), and NEI Core Grant to Doheny Eye Institute EY03040. CMC is the Mary D. Allen Chair in Vision Research, Doheny Eye Institute.

 Submitted for publication June 7, 2004; revised July 29, 2004; accepted July 30, 2004.

 Disclosure: S.W. Pickrell, None; X. Zhu, None; X. Wang, None; C.M. Craft, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Table 1.

 

Sense and Antisense Primers for Site-Directed Mutagenesis of mCAR

Target Element(s)	Sense (5′) Primers	Antisense (3′) Primers
1CBE CTAATCT	5′-GATCCCTGGCCACTggaCTGTACTCACTAATC-3′	5′-GATTAGTGAGTACAGtccAGTGGCCAGGGATC-3′
2CBE CTAATCT	5′-CTAATCTGTACTCACTggaCTGTTTTCCAAACTC-3′	5′-GAGTTTGGAAAACAGtccAGTGAGTACAGATTAG-3′
3CBE CTAATTA	5′-CTTGGCCCCTGAGCTggaTATAGCAGTGCTTC-3′	5′-GAAGCACTGCTATAtccAGCTCAGGGGCCAAG-3′
TATA-like TTATA	5′-GCCCCTGAGCTAATTgcAGCAGTGCTTCATGC-3′	5′-GCATGAAGCACTGCTgcAATTAGCTCAGGGGC-3′
4CBE CTAATCT	5′-CTCTGACTCCCACTggaCTACACATTCAGAGG-3′	5′-CCTCTGAATGTGTAGtccAGTGGGAGTCAGAG-3′
TATA-box TATAA	5′-CAGAGGATTGTGGATgcAAGAGGCTGGGAGGC-3′	5′-GCCTCCCAGCCTCTTgcATCCACAATCCTCTG-3′
Sp1 GAGGCTGGG	5′-GGATATAAGAaatTGaGAGGCCAGC-3′	5′-GCTGGCCTCtCAattTCTTATATCC-3′
1- and 2CBE	5′-GATCCCTGGCCACTggaCTGTACTCACTggaCT GTTTTCCAAACTC-3′	5′-GAGTTTGGAAAACAGtccAGTGAGTACAGtccAGTGGCCAGGGATC-3′
3CBE&TATA-like	5′-CTTGGCCCCTGAGCTggaTgcAGCAGTGCTTCATGC-3′	5′-GCATGAAGCACTGCTgcAtccAGCTCAGGGGCCAAG-3′

Figure 1.

 

Transient transfection analysis of the mCAR promoter activities in Weri-Rb-1 and COS-7 cells. The cells were processed and assayed for luciferase activity 45 hours after transfection. The promoterless pRL-Null vector was used as a control, and its luciferase activity was set at 1. Data are expressed as the mean ± SEM of results in three independent experiments performed in duplicate. (A) Four different promoter fragments in the pRL-Null vector were transfected. Exons 1 and 2 are represented with an empty box and are numbered 1 and 2, respectively. (B) The nucleotide sequence of the mCAR p215 proximal promoter fragment is shown with the predicted transcriptional regulatory elements. The CRX binding elements (CBEs) are designated from 5′ to 3′ as 1-, 2-, 3-, and 4CBE. The TATA-like element and TATA box are in italics, and the transcription start site is designated +1. Various truncated or site-directed mutant p215 promoter fragments were linked upstream of the Renilla luciferase reporter gene in the pRL-Null vector and transfected into Weri-Rb-1 retinoblastoma cells. Filled diamonds, rectangles, and ovals represent mutated elements. The promoter activities of mutant or truncated promoters were compared with that of the p215 by one-way ANOVA: *P < 0.05, ***P < 0.001.

Figure 2.

 

Tissue distribution of EGFP in transgenic Xenopus. (A) EGFP expression patterns in transgenic tadpoles. Anesthetized transgenic tadpoles were observed under a fluorescence dissection microscope. Tadpoles carrying the same construct were grouped and numbered. The EGFP expression levels in four types of tissues (eyes, pineal, brain, and muscle) in each animal are shown. Dark, medium, and light green fillings represent high, medium, and weak EGFP expression, respectively. (B) Representative images of EGFP expression in transgenic Xenopus. Anesthetized animals were observed and photographed with a fluorescence dissection microscope. The 181-bp truncated fragment drives EGFP expression in the optic nerve (Ba) and tail muscle (Bb), in addition to the eyes. The 147-bp shorter fragment drives EGFP expression in the brain (Bc) and spinal cord ganglia (Bd), besides the eyes. The 215-bp fragment with both TATA-elements mutated drives EGFP expression in the eyes but not in the pineal gland (Be). The 215-bp fragment with the distal two CBEs simultaneously mutated drives EGFP expression in eyes and jaw muscle (Bf). ON, optic nerve; TM, tail muscle; B, brain; SCG, spinal cord ganglia; P, pineal.

Figure 3.

 

Characterization of the anti-xCAR polyclonal antibody (Lumioc) by immunoblot and immunohistochemistry. (A) Immunoblot analysis was performed after electrophoresis of selected tissue homogenates from adult Xenopus laevis in an 11.5% SDS-PAGE. (B) Immunohistochemistry was performed on adult Xenopus retinal frozen sections. All xCAR-expressing cells were labeled with the xCAR antibody, LUMIoc, and visualized with Texas red (xCAR). Xenopus rod outer segments were labeled with the anti-bovine rhodopsin monoclonal antibody 1D4 and visualized with fluorescein (green).

Figure 4.

 

EGFP expression in transgenic Xenopus laevis retina. Xenopus cone photoreceptors were labeled with the xCAR LUMIoc polyclonal antibody and visualized with Texas red. EGFP was labeled with anti-EGFP polyclonal antibody visualized with fluorescein (green). Yellow was produced by overlap of red and green staining, showing EGFP expression in cone photoreceptors. Arrows: EGFP expression in cone photoreceptors. (A) No EGFP expression was detected in either rods or cones in control tadpoles (genotyped negative). (B) The wild-type 215-bp mCAR promoter fragment drove EGFP expression only in cone photoreceptors. (C) When the 215-bp mCAR promoter was truncated to 147-bp, the shorter fragment drove EGFP expression in both cone and rod photoreceptors. When the fourth CBE (D), the TATA-like element (E), or the TATA-box (F) was individually mutated, EGFP expression was not detectable in retina. (G) With the TATA-like element and TATA-box both mutated, EGFP was expressed in both cones and rods. RPE, retinal pigment epithelium; OS, photoreceptor outer segment; IS, photoreceptor inner segment.

The authors thank Mary D. Allen for her generous continued support of vision research for over a decade; the Mary D. Allen Laboratory for Vision Research, especially senior biochemist Bruce Brown, former research associate Aimin Li, the undergraduate (supported in part by the Provost’s USC Undergraduate Award) and high school students Katie Bourzac, Risha Patel, Deborah Kim, Kristine Kay, Rebekah Coss and Terrie Richman for contributions to the study; and Carla B. Green and research group, especially laboratory manager Silvia I. LaRue, for lending time and expertise in teaching the transgenic Xenopus techniques; and the Final Year Dissertation Fellowship from USC College of Letters, Arts, and Sciences.