Molecular Vision 2005; 11:1041-1051 <http://www.molvis.org/molvis/v11/a123/> Received 28 February 2005 | Accepted 22 November 2005 | Published 7 December 2005

Download Reprint

Conserved structure and spatiotemporal function of the compact rhodopsin kinase (GRK1) enhancer/promoter

Joyce E. Young,^1,2 Kenneth W. Gross,³ Shahrokh C. Khani^1,2
 
 

Departments of ¹Ophthalmology and ²Biochemistry, State University of New York, Buffalo, NY; ³Department of Molecular and Cell Biology, Roswell Park Cancer Institute, Buffalo, NY

Correspondence to: Shahrokh C. Khani, State University of New York at Buffalo, 462 Grider Street-ECMC, Buffalo, NY, 14215; Phone: (716) 898-5216; FAX: (716) 898-3343; email: skhani@buffalo.edu

Abstract

Purpose: To demonstrate that the crucial elements responsible for the spatial and temporal expression patterns of rhodopsin kinase (Rk) are contained within a narrow conserved segment immediately flanking the Rk transcription start sites.

Methods: Sequences upstream of the mouse Rk gene were compared to the human sequence to identify areas of conservation. Transgenic mice carrying a segment of the conserved human DNA sequence linked upstream of the green fluorescent protein (GFP) gene were examined by fluorescence microscopy and RT-PCR to localize GFP expression in retina and pineal gland. Rk and GFP temporal expression patterns were further compared by immunostaining and real-time RT-PCR in transgenic eyes during development.

Results: Comparison of the mouse and human 5' flanking sequence revealed only a small island of conserved sequence upstream of the respective Rk start sites. Uniform GFP expression was supported by a 0.2 kb fragment of the conserved human sequence in the transgenic mouse rods, cones, and pinealocytes. Developmental studies revealed an exponential rise in Rk and GFP transcripts in the first ten day postnatal period followed by a plateau later extending to adulthood. Rk and GFP proteins were first detected after postnatal day 10 and rose in parallel afterwards, overlapping in time with the maturation of photoreceptor outer segments and eye opening.

Conclusions: The conserved short enhancer/promoter immediately upstream of the Rk gene contains the key elements required for appropriate response to spatial and temporal cues during photoreceptor cell differentiation and fate determination. The above studies narrow the core sequences that govern gene expression in photoreceptors in vivo.

Introduction

Rhodopsin kinase is a photoreceptor-specific G protein-dependent receptor kinase (GRK) [1] whose expression and activity is essential for proper adaptation, recovery, and protection of light sensing neurons [2-5]. A recent study [6] on Rk gene transcription has been the first to examine the role of transcriptional regulation in the GRK family in significant mechanistic detail, and additionally has provided unique insight into the core requirements for basal expression of this gene and potentially others in photoreceptors. Unlike most other photoreceptor-specific genes whose promoter activity depends on interactions of a complex array of elements scattered across several kilobases of DNA [7-11], the Rk promoter appears more focally organized with active sequences concentrated within a discrete 0.2 kb zone located immediately near the transcription start site. Enhanced activity of this segment is apparent, in isolation, both in vitro and in vivo in Rk-GFP transgenic mouse outer retinas where the segment drives uniform expression of GFP [6]. Further studies evaluating the extent of conservation and contribution of this enhancer/promoter to the overall spatial and temporal function of the promoter will be important in assessing the physiological significance of this discrete region and its interaction with the core transcriptional complex that directs the expression of Rk and potentially other photoreceptor-specific genes in response to endogenous and exogenous cues.

Rk expression patterns and mechanisms have only been partially characterized to date. The presence of Rk in terminally differentiated rods, cones [12,13], and pinealocytes [14,15] is well-known, paralleling other photoreceptor-specific proteins in spatial expression profile. Immunostaining studies in primate and ferret retinas during development [16,17] have suggested an increase in Rk immunoreactive protein in response to terminal differentiation of photoreceptors, however, neither the spatial nor the developmental mechanisms have been studied in great detail in these or other species. Recent in vitro and in vivo functional studies on the human Rk 5' flanking sequence, as mentioned previously, have identified the highly active human enhancer/promoter (-112 to +87) by in vitro transient transfection assays and further confirmed the uniform activity of this segment in outer layer of Rk-GFP transgenic mouse retinas. The architecture and spatial activity of the human segment was found to resemble most closely the interphotoreceptor retinoid-binding protein (IRBP) promoter [18,19] with an Otx-like bicoid homeodomain binding element, H1, and G-rich sequences among the shared features contributing to the robust activity of both these short promoters despite absence of longer stretches of intergenic conservation [6,20-23]. Additional important parallels in transcriptional regulatory mechanism within the GRK family are likely given the collective role of the entire class in short and long term receptor desensitization and moderation of the signal relay in various systems [24], however paucity of data on Rk and GRK transcriptional mechanisms to this point has limited the exploration of potential similarities.

In this report, we provide additional new evidence in support of the physiological significance of the enhancer/promoter region by demonstrating preferential evolutionary conservation of this segment, the exclusive activity of the human segment in mouse photoreceptors and pinealocytes, and the accurate temporal functioning of the human segment in mouse. These findings suggest that the core elements required for exclusive expression of genes in photoreceptors, can be relatively few and condensed into a small segment of genome.

Methods

Structural analysis of the Rk 5' flanking region

Rapid amplification of cDNA ends (RLM-RACE kit, Ambion, Austin, TX) was used to identify the position of the transcription initiation sites for the mouse Rk gene. Total C57BL/6 mouse ocular or splenic RNA (5 μg), prepared by Trizol method, was sequentially treated with calf intestinal phosphatase, tobacco alkaline phosphatase and RNA ligase to selectively replace the G-caps of mRNA with an anchoring oligomer sequence. The product was then annealed with random primers and reverse-transcribed with M-MLV reverse transcriptase (RT) at 42 °C for 1 h [6]. Two sequential 35-cycle amplifications of the resultant cDNA were performed using a hot start/touch-down program with a peak annealing temperature of 65 °C. The first was in the 1X PCR mix supplied with the RACE kit together with the outer set of primers (forward: anchor-specific outer oligomer, reverse: Rk-specific 5'-TCA CGG AGA CCC TCA CAC T-3' oligomer complementary to positions 287-269 encoding residues 45-52). The second was in a previously described PCR buffer [6] supplemented with 1.25 mM MgCl₂, 8% dimethylsulfoxide with the inner primers (forward: anchor-specific inner oligomer, reverse: Rk-specific 5'-TCT TGT CTC GGG AAG ACG GT-3' oligomer complementary to positions 232-213 encoding residues 27-34). The amplified fragments were then cloned into pGEM-T vector (Promega, Madison, WI) and sequenced by automated dideoxy method and compared to the sequences in the public genome database. Comparison of the mouse and human genomic sequences was performed using the pairwise BLAST or BestFit algorithm (GCG).

Animals

All protocols were approved by IACUC and adhered to the standard conventions on animal use for scientific studies. Generation of the Rk-GFP transgenic mouse line carrying 0.11 kb of the human Rk enhancer/promoter upstream of GFP has previously been detailed [6]. Animals were genotyped by amplification of tail genomic DNA with GFP amplimers. Globes and other tissues were harvested after euthenization of animals reared in a 12 h light/dark cycle in the light-adapted state. Identity of the dissected pineal tissue was further verified histologically with hematoxylin and eosin stain.

For developmental studies, GFP-negative C57BL/6 female mice were crossed with GFP-positive males. The females were checked daily and pregnancies were dated from the time the copulation plug was first detected (E0). The embryonic gestational ages were verified by counting the number of somites. GFP-positive embryonic eyes from the same pregnancy or the same gestational age were harvested then pooled for RNA preparation after verification of the genotype.

Microscopy and immunofluorescence

Primary antibodies or other labeling reagents were obtained from several sources. Polyclonal antibodies against GFP, mouse Rk (8585), mouse Chx10 were kindly provided by Drs. C. W. Smith, R. Lefkowitz, and R. McInnes, respectively. Monoclonal antibodies against human Rk (D11 and G8) [25] and human CRALBP [26] were gifts from Drs. K. Palczewski and J. Saari, respectively. Peanut agglutinin (PNA) and monoclonal antibodies against syntaxin (HPC-1) and calbindin D-28k were obtained from Sigma.

Fluorescence labeling of 6 μm-thick tissue sections was carried out as previously described [6]. Whole globes, eyecups or pineal glands from light-adapted mice were fixed in 4% paraformaldehyde in 0.1 M sodium phosphate pH 7.4, frozen-sectioned, and repeatedly exposed to blocking solution prior to being placed in one or more of the above primary antibodies/reactants at 1:500-1:1000 dilution. After several washes, Alexa-green (488) and red (568) fluorescent secondary antibodies/reactants (Invitrogen, Carlsbad, CA) were added in combination to complete the immunostaining. Stained sections were then scanned by confocal microscope Bio-Rad 1024 at the default pinhole settings with adjustments made automatically depending on magnification. 1 μm optical slices collected in both red and green channel, were then projected, and merged using Image J and Confocal Assistant software to generate the final figures. Controls lacked one or more of the primary antibodies/reactants. In experiments requiring primary incubation with PNA, Alexa-green (488) streptavidin was substituted in the secondary reaction.

Conventional and real-time RT-PCR

Total RNA was isolated from globes of various gestational age mice or adult pineals, treated with DNA-free resin (Ambion), and reverse-transcribed into cDNA as described above. The product was then further amplified either by conventional PCR using hot start/touch down conditions described above or in real-time with one of the following gene-specific primer pairs in Table 1. Primer selection was based on specificity and efficiency. With the exception of GFP amplimers, forward and reverse primers were from separate exons, at least 1 kb apart, to minimize background genomic DNA amplification signal. Conditions and primer sequences were additionally optimized to achieve comparable efficiencies [27] among the primer pairs with resolution over a four orders of magnitude dynamic range in real-time PCR amplifications.

Real-time amplifications were performed in 96 well plates in a Bio-Rad i-cycler. Each 25 μl amplification contained an aliquot of RT reaction or control without RT corresponding to 30 to 300 ng of RNA, 1X iQ SYBR Green Supermix (Bio-Rad, Hercules, CA) and 4 pmol of each forward and reverse primer for each selected gene. Fluorescence emission from the reactions was monitored in real time over a 50-cycle range with alternating denaturation (94 °C for 30 s) and annealing/extension (60 °C for 30 s) steps. The number of cycles needed to reach threshold (C_t) detection limit for each sample was subtracted from the threshold values for GAPDH and actin to obtain ΔC_t. Use of actin RNA levels as the normalization constant for studying relative variations in photoreceptor-specific transcript levels during development was based on a comparable study in the literature [28]. A modification of the published formula R=100(2^-ΔΔCt) was further used to represent relative transcript levels in terms of percentage of the adult value. For each time point and primer set, the average R value (R_avg) and its range were obtained from at least three animals of a gestational age, three quantitative amplifications, and three separate RT reactions and plotted against gestational age.

Results

A unique enhancer/promoter region immediately upstream of the human Rk gene was previously described and shown to contain elements capable of confering uniform GFP expression in the outer retina of Rk-GFP transgenic mice [6]. To further evaluate the role of this segment in determining the overall spatial and temporal Rk expression pattern, the region was examined with respect to the degree of structural conservation and its ability to drive GFP expression in Rk-GFP transgenic mice in a cell-specific and developmentally-regulated pattern matching that of endogenous mouse Rk gene.

Structural conservation of the DNA sequence upstream of the Rk genes

Comparison of 5' flanking sequences was carried out to delineate areas of evolutionary conservation. Products of the 5' primer extension/RACE reaction from mouse eye RNA defined the position of the mouse Rk transcription initiation sites relative to a downstream inner RACE primer (data not shown). Amplified cDNA-based products up to 230 bp in length were only seen with the complete reaction containing ocular RNA and reverse transcriptase but were absent from reactions with splenic RNA as template or no reverse transcriptase controls. Multiplicity of the observed bands was consistent with the heterogeneity of the mouse transcription start sites as seen with the human Rk gene [6]. Sequence analysis of the amplified fragments defined the extent of untranslated sequences and the positions of the mouse Rk transcript cap sites, the most distal of which was arbitrarily designated as position +1 (Figure 1). These experimentally determined sequences matched the mouse genomic sequences in the public database precisely indicating colinearity and contiguity of the transcript sequence in this region. Further pairwise BLAST and BestFit comparison of the human and mouse DNA (eight to nine kbs) in the interval between the Rk and the adjacent H,K-ATPase β-subunit gene [29,30] revealed a single, highly conserved 216 bp island of 78% sequence identity immediately flanking the transcription start sites (Figure 1). The boundaries of the conserved region closely matched and included those previously defined by deletion mutagenesis and in vitro transfection assays of the human Rk promoter [6]. A T-rich stretch and a putative homeodomain binding sequence (H1, Figure 1, sequence in red text), previously found to be crucial to the human promoter activity in vitro, were also included in the mouse sequence. No distinct TATA box, Inr [31], H2 [6], PCE/RET 1-like element [32,33], or definitive Nrl binding sequence [34] were detected in the conserved segment of the mouse sequence, however a putative inverted AP-1-like site was present at -64 to -69. The data above point to the selective evolutionary conservation and potential functional significance of the noncoding DNA segment in the vicinity of the Rk transcription start sites in at least the higher vertebrates. Additional comparisons to the human cone opsin kinase (GRK7) [35-38] (Genbank accession number AC112504, positions 18348-215000) and drosophila Rk (GRK1) [39] (Genbank accession number NW_632775, positions 69961-72019) 5' flanking genomic sequences revealed overall architectural similarities and multiple homeodomain binding elements present in the vicinity of the transcription start sites in GRK7 but no longer stretches of sequence identity.

Photoreceptor-specific expression of Rk-GFP transgene in retina

The 0.2 kb human enhancer/promoter, (-112 to +87), used previously to construct the Rk-GFP transgenic mice consists of the portion of the conserved 5' directly upstream of the major transcription start sites (Figure 1). The mice carrying this segment upstream of GFP exhibited uniform GFP fluorescence in the outer layers of fresh whole mounted retinas [6]. To further demonstrate photoreceptor-restricted localization driven by this DNA segment, additional double-labeling studies were performed on Rk-GFP transgenic globe sections. A combination of anti-GFP polyclonal antibody together with red fluorescent secondary antibody was used to maximize the GFP detection sensitivity in fixed tissue sections. As seen from Figure 2A-C, GFP immunoreactivity (red) was distributed throughout photoreceptor layer and colocalized with cone marker PNA (green) in the outer segments. An identical outer segment colocalization signal (yellow) was also encountered (Figure 2D) with anti-Rk (red) and PNA (green) combined, further validating this method of colocalization. The observed difference in Rk and GFP subcellular distribution in photoreceptors reflects the near full partitioning of Rk into outer segments in the light-adapted state as opposed to unbiased distribution of GFP throughout the cytosol [40]. Additional studies with M- and S-cone antibodies supported GFP colocalization in both cone types [41] (data not shown). In contrast, other cell types including amacrine (anti-syntaxin-labeled), Müller glial and retinal pigment epithelial (anti-CRALBP-labeled), horizontal (anti-calbindin-labeled), and bipolar (Chx 10-labeled) cells did not contain GFP epitopes or fluorescence as evident from nonoverlapping spatial distribution of GFP and cell-specific markers (Figure 2E-H). Even in the densely packed outer nuclear layer where red-labeled photoreceptor cell bodies are tightly juxtaposed against the green-labeled Müller cell processes (Figure 2F), the red and green remained distinct indicating absence of GFP from these glia or their processes. Similarly in Figure 2H, direct GFP fluorescence from the outer nuclear layer, seen at higher laser intensities, showed no spatial overlap with Chx10 expressing (red) bipolar cells. No other ocular cell type examined besides retina showed any GFP fluorescence or labeling (data not shown). Exclusive localization of the GFP in cells carrying cone and rod markers and its absence elsewhere is consistent with the photoreceptor-restricted activity of the conserved enhancer/promoter segment in the globe.

Expression of Rk-GFP transgene in the pineal gland

Extraocular transcriptional activity of the human segment was examined next in pinealocytes. As seen in Figure 3A, both Rk and GFP cDNA bands could be amplified from pineal RNA in the presence of reverse transcriptase. Dual immunostaining studies further confirmed Rk and GFP colocalization in at least a subset of pinealocytes (Figure 3B). Interestingly the preferential localization of GFP reporter in the pineal periphery as opposed to the central region was precisely as encountered in a transgenic mouse carrying 2 kb or 12 kb Crx promoter upstream of LacZ reporter [11]. Specific immunocolocalization was absent in merged images of GFP positive tissue in the absence of one or both primary antibodies (Figure 3C,D). Prior RT-PCR studies excluded expression of GFP in other transgenic mouse tissues including nearly the entire brain [6]. Taken together the above findings clearly demonstrate that the full spatial pattern attributed to the Rk gene can be reproduced in transgenic model with the use of the 0.2 kb human Rk enhancer/promoter linked upstream of GFP reporter.

Developmental changes in the expression of Rk and the transgene

To determine whether the narrow conserved human genomic segment flanking the Rk transcription start site also mediated appropriate developmental regulation, temporal GFP expression pattern was compared to that of endogenous mouse Rk gene by real-time PCR and immunostaining. As seen from Figure 4, GFP and Rk transcripts were both readily detectable by real-time RT-PCR in the immediate postnatal period and showed an exponential rise for the first 10 days followed by a plateau in levels extending to adulthood. Although the exponential phase appeared somewhat earlier for GFP than for Rk, the overall temporal patterns of both were similar qualitatively and appear tied to temporal landmarks of terminal photoreceptor differentiation, closely mimicking rodopsin pattern. The relative transcript levels were based on C_t's that ranged between 14 and 30 in the presence of RT, and >35 in the absence of RT. All real time RT-PCR yielded single products with a sharp melting peak. No product bands were detectable in reactions without RT. Earliest RT-PCR detectable Rk transcriptional activity was on or after E10 albeit at very low level, as also seen with opsin and other photoreceptor-specific genes that are activated upon photoreceptor birth or cell fate determination [10,42] (data not shown).

Additional immunostaining experiments during eye development in Figure 5, showed a parallel gradual rise in GFP and Rk protein levels first detected after P10, near the eye opening time frame. As previously noted, the complete partitioning of Rk to outer segments in the full light adapted state rendered the simultaneous monitoring of Rk and GFP in stained globe sections feasible. Immunolabeling of prenatal globes from E10, E14, and E17 was also performed but revealed no detectable GFP or Rk expression (data not shown). The above findings support the presence of developmentally sensitive transcriptional regulatory elements within the conserved segment in the immediate vicinity of the Rk transcription start sites.

Discussion

The majority of photoreceptor-specific promoters are complex and consist of a series of conserved functional elements scattered over several kilobases of DNA extending upstream of the gene [7,8,10,43]. Despite excellent definition of the overall boundaries of these promoter by transgenic studies, identification of a core collection of crucial elements responsible for the activity of the promoter in vivo has been difficult with many isolated sequences losing their physiologic activity, uniformity or specificity when isolated from the larger genomic context [7,32]. Rk promoter is one of a few exceptions to this rule. In this study we provide additional structural and functional evidence to support the crucial contribution of the highly active 0.2 kb enhancer/promoter to the overall function of the Rk promoter. In addition to establishing the selective conservation of this region among the 5' flanking sequences, we have also demonstrated the ability of the enhancer/promoter to drive expression excludively to rods, cones and pinealocytes and to induce downstream gene expression tied with terminal differentiation of photoreceptors as seen for Rk gene. These findings provide new insight into the minimal prerequisites for expression of genes shared in common between terminally differentiated rods and cone photoreceptors.

Rk enhancer/promoter shares its compact configuration and composition and activity profile most closely with IRBP and to a lesser extent with visual arrestins. Despite absence of long stretches of conservation, all basal promoters are approximately 0.2 kb in length in the immediate vicinity of the start sites and capable of supporting the uniform and specific spatial expression patterns of the respective genes in transgenic models [32,44,45]. These promoters like those of many photoreceptor-specific genes all contain a collection of one or more homeodomain binding site together with other gene-specific elements [46,47]. The putative Otx/bicoid-type homeodomain binding site in the conserved Rk enhancer/promoter region is designated H1, which is adjacent to a conserved T-rich and a partially conserved G-rich segment similar to the IRBP promoter [22]. A clear role has been established for the H1 site by mutagenesis in transient transfection assays by mutagenesis suggesting its interaction with a potential homeodomain cognate [6]. Studies in the mice lacking the major photoreceptor-specific homeodomain protein Crx [48-50] however, show no alterations in the Rk transcript levels despite marked decline in the mRNA of other photoreceptor-specific genes suggesting an alternative homeodomain molecule as the cognate for H1. Recently Otx2 has been suggested as a possible alternative homeodomain protein for this target especially given its recently identified role in photoreceptor-fate determination [22], however the levels of this homeodomain protein are barely detectable in differentiated photoreceptors [51]. A separate putative PCE/RET1-type homeodomain binding site, H2, encountered in the human 5' flanking sequence distal to the enhancer/promoter region is missing from the mouse sequence suggesting species specific differences. However, interestingly, the removal of H2 had no effect in the human Rk promoter activity in vitro [6]. No Nrl responsive elements [34] were present in either species in the conserved zone consistent with the expression of Rk in both rods and cones and the absence of any major effect on Rk levels resulting from the targeted disruption of Nrl in mouse [13]. Thus, the key core elements for general activity in photoreceptors and pineal appear to be limited to the conserved homeodomain elements together with the T- and possibly G-rich regions described above.

The transgene used here only included portions of the conserved human genomic sequences that extended upstream of the Rk start sites; hence the role of the remainder of the contiguous conserved segments in this region remains unknown. It is possible that some of these sequences may play a redundant or quantitative role to further enhance the overall promoter activity in the native Rk gene as compared to the transgene promoter. GFP expression levels in our transgenic line appeared lower than Rk expression in our transgenic line consistent with the possibility that the excluded segment might have role as a quantitative or regulated enhancer. Recent transient transfection data are also consistent with this hypothesis showing modest gain in activity resulting from the inclusion of the additional conserved sequences (unpublished data). Despite inability to draw a clear conclusion regarding the excluded segments, our data, from the single transgenic line, nevertheless provide validation for the crucial role of the upstream conserved segment in the overall Rk promoter activity. It is possible that additional lines generated with this or full-length conserved sequence could have higher levels of GFP expression comparable to the levels of Rk, however any change in levels of expression among different lines could be a reflection of a number of factors besides promoter strength including transgene copy number, genomic position, and epigenetic modifications.

Our mouse mRNA leader sequences were in part different than the 5' untranslated sequence reported by Chen et al. [4] The cDNA sequences here were based on sequencing the multiple RACE products and later the same precise sequence was retrieved from genome data bank on a BAC clone from chromosome 8. The previously reported sequence (NM011881) by Chen et al. [4] matched ours precisely downstream of position 314 (Figure 1) but diverged completely upstream of this point. In fact, according to a BLAST analysis, the published divergent sequence localized it to a chromosome 10 BAC (Genbank accession number AC126242 positions 77462-77583) suggesting that a cross ligation of sequences or rearrangement may have occurred during library construction. The possibility of an actual divergence in Rk noncoding sequences among different strains is unlikely but cannot be ruled out since the sequences here were obtained from C57BL/6 mice as opposed to 129SvJ strain libraries used for construction of Rk-deficient mouse.

The spatial and temporal activity profile of the conserved enhancer/promoter are typical for the genes abundantly expressed in photoreceptors, and hence provide a broader perspective of the mechanisms that underlie photoreceptor-specific expression. Like most other photoreceptor-specific genes, abundant Rk enhancer/promoter activity is restricted to fully differentiated photoreceptors and a subset of pinealocytes. The data are also consistent with observed increase in Rk expression late in ferret and primate eye development [16,17]. This pattern is however unlike that of IRBP that is expressed in substantial quantities by the undifferentiated photoreceptors early in development [28,52]. Rk transcripts were also detected in very low levels at E10 coinciding with the emergence of first cone photoreceptors and first evidence of photoreceptor fate determination. In Rk-GFP transgenic mice, we observed significant difference in the activity levels between mouse and human Rk promoter during development despite remarkable qualitative similarities in the cellular distribution pattern for endogenous mouse Rk and human promoter/enhancer-driven GFP. The lower level of human enhancer/promoter activity in mouse is not unexpected and could be attributed to the limited divergence of the human and mouse promoter or alternatively to the lack of contribution from missing distant elements outside the enhancer/promoter region.

A transcriptional basis for the observed expression patterns is highly likely. The conserved genomic sequence included in the Rk-GFP transgenic construct, serendipitously falls in a region not included in even the longest of mouse transcripts according to our RACE mapping data and hence it is highly unlikely to contribute to posttranscriptional regulation of Rk expression. The parallel rise seen in mouse Rk and GFP transcripts cannot easily be attributed to Rk-specific changes in RNA stabilization. However, the possibility that a more global posttranscriptional regulation might affect the levels of multiple transcripts and proteins including Rk and GFP with emergence and maturation of outer segments cannot be ruled out. The lag in protein accumulation by approximately 10 days could reflect the presence of such a mechanism.

The above findings demonstrate that the assembly of the core photoreceptor-specific complex can be initiated in vivo on a discrete regulatory DNA platform. Given the prominent role of Rk in regulating the photoreceptor response to light, it is possible that additional light-dependent and independent regulatory pathways may ultimately interface with these sequences. Previously another major shut-off pathway gene encoding arrestin has been shown to be transcriptionally regulated by light [53,54]. There is also increasing evidence that GRKs levels in other systems may be regulated both transcriptionally and posttranscriptionally in response to chronic stimulus exposure or overload [24,55,56]. At this point, Rk is one of the only GRK loci for which detailed promoter function studies are available [24]. Additional future studies will focus the molecules that interact with the conserved DNA sequences to regulate the transcriptional activity of this GRK locus.

Acknowledgements

The authors are grateful to Dr. W. C. Smith (University of Florida, Gainsville, FL) for the GFP antibody, Drs. K. Palczewski (University of Washington, Seattle, WA) and R. Lefkowitz (Duke University, Durham, NC) for Rk antibodies, Dr. J. Saari (University of Washington, Seattle, WA) for CRALBP antibody, Dr. C. Craft (University of Southern California, Los Angeles, CA) for M- and S-cone opsin antibodies, and Dr. R. McInnes (University of Toronto, ON) for Chx10 antibody. M. K. Ellsworth provided expert technical assistance and help with animal care.

Supported by National Eye Institute R01 EY13600 (SCK), a challenge grant from Research to Prevent Blindness to the Department of Ophthalmology, R01 HL48459 (KWG) and National Cancer Institute funded Cancer Center Support Grant CA16056

References

1. Pitcher JA, Freedman NJ, Lefkowitz RJ. G protein-coupled receptor kinases. Annu Rev Biochem 1998; 67:653-92.

2. Maeda T, Imanishi Y, Palczewski K. Rhodopsin phosphorylation: 30 years later. Prog Retin Eye Res 2003; 22:417-34.

3. Cideciyan AV, Zhao X, Nielsen L, Khani SC, Jacobson SG, Palczewski K. Null mutation in the rhodopsin kinase gene slows recovery kinetics of rod and cone phototransduction in man. Proc Natl Acad Sci U S A 1998; 95:328-33.

4. Chen CK, Burns ME, Spencer M, Niemi GA, Chen J, Hurley JB, Baylor DA, Simon MI. Abnormal photoresponses and light-induced apoptosis in rods lacking rhodopsin kinase. Proc Natl Acad Sci U S A 1999; 96:3718-22.

5. Hao W, Wenzel A, Obin MS, Chen CK, Brill E, Krasnoperova NV, Eversole-Cire P, Kleyner Y, Taylor A, Simon MI, Grimm C, Reme CE, Lem J. Evidence for two apoptotic pathways in light-induced retinal degeneration. Nat Genet 2002; 32:254-60.

6. Young JE, Vogt T, Gross KW, Khani SC. A short, highly active photoreceptor-specific enhancer/promoter region upstream of the human rhodopsin kinase gene. Invest Ophthalmol Vis Sci 2003; 44:4076-85.

7. Zack DJ, Bennett J, Wang Y, Davenport C, Klaunberg B, Gearhart J, Nathans J. Unusual topography of bovine rhodopsin promoter-lacZ fusion gene expression in transgenic mouse retinas. Neuron 1991; 6:187-99.

8. Wang Y, Macke JP, Merbs SL, Zack DJ, Klaunberg B, Bennett J, Gearhart J, Nathans J. A locus control region adjacent to the human red and green visual pigment genes. Neuron 1992; 9:429-40.

9. Nathans J, Davenport CM, Maumenee IH, Lewis RA, Hejtmancik JF, Litt M, Lovrien E, Weleber R, Bachynski B, Zwas F Klingaman R, Fishman G. Molecular genetics of human blue cone monochromacy. Science 1989; 245:831-8.

10. Chen J, Tucker CL, Woodford B, Szel A, Lem J, Gianella-Borradori A, Simon MI, Bogenmann E. The human blue opsin promoter directs transgene expression in short-wave cones and bipolar cells in the mouse retina. Proc Natl Acad Sci U S A 1994; 91:2611-5.

11. Furukawa A, Koike C, Lippincott P, Cepko CL, Furukawa T. The mouse Crx 5'-upstream transgene sequence directs cell-specific and developmentally regulated expression in retinal photoreceptor cells. J Neurosci 2002; 22:1640-7.

12. Lyubarsky AL, Chen C, Simon MI, Pugh EN Jr. Mice lacking G-protein receptor kinase 1 have profoundly slowed recovery of cone-driven retinal responses. J Neurosci 2000; 20:2209-17.

13. Mears AJ, Kondo M, Swain PK, Takada Y, Bush RA, Saunders TL, Sieving PA, Swaroop A. Nrl is required for rod photoreceptor development. Nat Genet 2001; 29:447-52.

14. Lorenz W, Inglese J, Palczewski K, Onorato JJ, Caron MG, Lefkowitz RJ. The receptor kinase family: primary structure of rhodopsin kinase reveals similarities to the beta-adrenergic receptor kinase. Proc Natl Acad Sci U S A 1991; 88:8715-9.

15. Zhao X, Haeseleer F, Fariss RN, Huang J, Baehr W, Milam AH, Palczewski K. Molecular cloning and localization of rhodopsin kinase in the mammalian pineal. Vis Neurosci 1997; 14:225-32.

16. Sears S, Erickson A, Hendrickson A. The spatial and temporal expression of outer segment proteins during development of Macaca monkey cones. Invest Ophthalmol Vis Sci 2000; 41:971-9.

17. Johnson PT, Williams RR, Reese BE. Developmental patterns of protein expression in photoreceptors implicate distinct environmental versus cell-intrinsic mechanisms. Vis Neurosci 2001; 18:157-68.

18. Bobola N, Hirsch E, Albini A, Altruda F, Noonan D, Ravazzolo R. A single cis-acting element in a short promoter segment of the gene encoding the interphotoreceptor retinoid-binding protein confers tissue-specific expression. J Biol Chem 1995; 270:1289-94.

19. Boatright JH, Buono R, Bruno J, Lang RK, Si JS, Shinohara T, Peoples JW, Nickerson JM. The 5' flanking regions of IRBP and arrestin have promoter activity in primary embryonic chicken retina cell cultures. Exp Eye Res 1997; 64:269-77.

20. Bobola N, Briata P, Ilengo C, Rosatto N, Craft C, Corte G, Ravazzolo R. OTX2 homeodomain protein binds a DNA element necessary for interphotoreceptor retinoid binding protein gene expression. Mech Dev 1999; 82:165-9.

21. Liou GI, Matragoon S, Yang J, Geng L, Overbeek PA, Ma DP. Retina-specific expression from the IRBP promoter in transgenic mice is conferred by 212 bp of the 5'-flanking region. Biochem Biophys Res Commun 1991; 181:159-65.

22. Fong SL, Fong WB. Elements regulating the transcription of human interstitial retinoid-binding protein (IRBP) gene in cultured retinoblastoma cells. Curr Eye Res 1999; 18:283-91.

23. Borst DE, Boatright JH, Si JS, Stodulkova E, Remaley N, Pallansch LA, Nickerson JM. Structural characterization and comparison of promoter activity of mouse and bovine interphotoreceptor retinoid-binding protein (IRBP) gene 5' flanking regions in WERI, Y79, chick retina cells, and transgenic mice. Curr Eye Res 2001; 23:20-32.

24. Penela P, Ribas C, Mayor F Jr. Mechanisms of regulation of the expression and function of G protein-coupled receptor kinases. Cell Signal 2003; 15:973-81.

25. Zhao X, Huang J, Khani SC, Palczewski K. Molecular forms of human rhodopsin kinase (GRK1). J Biol Chem 1998; 273:5124-31.

26. Nawrot M, West K, Huang J, Possin DE, Bretscher A, Crabb JW, Saari JC. Cellular retinaldehyde-binding protein interacts with ERM-binding phosphoprotein 50 in retinal pigment epithelium. Invest Ophthalmol Vis Sci 2004; 45:393-401.

27. Tichopad A, Dilger M, Schwarz G, Pfaffl MW. Standardized determination of real-time PCR efficiency from a single reaction set-up. Nucleic Acids Res 2003; 31:e122. Erratum in: Nucleic Acids Res 2003; 31:6688.

28. Liou GI, Wang M, Matragoon S. Timing of interphotoreceptor retinoid-binding protein (IRBP) gene expression and hypomethylation in developing mouse retina. Dev Biol 1994; 161:345-56.

29. Song I, Brown DR, Yamada T, Trent JM. Mapping of the gene encoding the beta-subunit of H+,K(+)-ATPase to human chromosome 13q34 by fluorescence in situ hybridization. Genomics 1992; 14:1114-5.

30. Morley GP, Callaghan JM, Rose JB, Toh BH, Gleeson PA, van Driel IR. The mouse gastric H,K-ATPase beta subunit. Gene structure and co-ordinate expression with the alpha subunit during ontogeny. J Biol Chem 1992; 267:1165-74.

31. Smale ST, Baltimore D. The "initiator" as a transcription control element. Cell 1989; 57:103-13.

32. Kikuchi T, Raju K, Breitman ML, Shinohara T. The proximal promoter of the mouse arrestin gene directs gene expression in photoreceptor cells and contains an evolutionarily conserved retinal factor-binding site. Mol Cell Biol 1993; 13:4400-8.

33. Morabito MA, Yu X, Barnstable CJ. Characterization of developmentally regulated and retina-specific nuclear protein binding to a site in the upstream region of the rat opsin gene. J Biol Chem 1991; 266:9667-72.

34. Rehemtulla A, Warwar R, Kumar R, Ji X, Zack DJ, Swaroop A. The basic motif-leucine zipper transcription factor Nrl can positively regulate rhodopsin gene expression. Proc Natl Acad Sci U S A 1996; 93:191-5.

35. Weiss ER, Ducceschi MH, Horner TJ, Li A, Craft CM, Osawa S. Species-specific differences in expression of G-protein-coupled receptor kinase (GRK) 7 and GRK1 in mammalian cone photoreceptor cells: implications for cone cell phototransduction. J Neurosci 2001; 21:9175-84.

36. Hisatomi O, Matsuda S, Satoh T, Kotaka S, Imanishi Y, Tokunaga F. A novel subtype of G-protein-coupled receptor kinase, GRK7, in teleost cone photoreceptors. FEBS Lett 1998; 424:159-64.

37. Weiss ER, Raman D, Shirakawa S, Ducceschi MH, Bertram PT, Wong F, Kraft TW, Osawa S. The cloning of GRK7, a candidate cone opsin kinase, from cone- and rod-dominant mammalian retinas. Mol Vis 1998; 4:27 <http://www.molvis.org/molvis/v4/a27/>.

38. Chen CK, Zhang K, Church-Kopish J, Huang W, Zhang H, Chen YJ, Frederick JM, Baehr W. Characterization of human GRK7 as a potential cone opsin kinase. Mol Vis 2001; 7:305-13 <http://www.molvis.org/molvis/v7/a43/>.

39. Lee SJ, Xu H, Montell C. Rhodopsin kinase activity modulates the amplitude of the visual response in Drosophila. Proc Natl Acad Sci U S A 2004; 101:11874-9.

40. Inglese J, Koch WJ, Caron MG, Lefkowitz RJ. Isoprenylation in regulation of signal transduction by G-protein-coupled receptor kinases. Nature 1992; 359:147-50.

41. Zhu X, Brown B, Li A, Mears AJ, Swaroop A, Craft CM. GRK1-dependent phosphorylation of S and M opsins and their binding to cone arrestin during cone phototransduction in the mouse retina. J Neurosci 2003; 23:6152-60.

42. Cepko CL, Austin CP, Yang X, Alexiades M, Ezzeddine D. Cell fate determination in the vertebrate retina. Proc Natl Acad Sci U S A 1996; 93:589-95.

43. Chiu MI, Nathans J. Blue cones and cone bipolar cells share transcriptional specificity as determined by expression of human blue visual pigment-derived transgenes. J Neurosci 1994; 14:3426-36.

44. Mani SS, Besharse JC, Knox BE. Immediate upstream sequence of arrestin directs rod-specific expression in Xenopus. J Biol Chem 1999; 274:15590-7.

45. Zhu X, Ma B, Babu S, Murage J, Knox BE, Craft CM. Mouse cone arrestin gene characterization: promoter targets expression to cone photoreceptors. FEBS Lett 2002; 524:116-22.

46. Freund C, Horsford DJ, McInnes RR. Transcription factor genes and the developing eye: a genetic perspective. Hum Mol Genet 1996; 5 Spec No:1471-88.

47. Furukawa T, Kozak CA, Cepko CL. rax, a novel paired-type homeobox gene, shows expression in the anterior neural fold and developing retina. Proc Natl Acad Sci U S A 1997; 94:3088-93.

48. Chen S, Wang QL, Nie Z, Sun H, Lennon G, Copeland NG, Gilbert DJ, Jenkins NA, Zack DJ. Crx, a novel Otx-like paired-homeodomain protein, binds to and transactivates photoreceptor cell-specific genes. Neuron 1997; 19:1017-30.

49. Furukawa T, Morrow EM, Cepko CL. Crx, a novel otx-like homeobox gene, shows photoreceptor-specific expression and regulates photoreceptor differentiation. Cell 1997; 91:531-41.

50. Furukawa T, Morrow EM, Li T, Davis FC, Cepko CL. Retinopathy and attenuated circadian entrainment in Crx-deficient mice. Nat Genet 1999; 23:466-70.

51. Nishida A, Furukawa A, Koike C, Tano Y, Aizawa S, Matsuo I, Furukawa T. Otx2 homeobox gene controls retinal photoreceptor cell fate and pineal gland development. Nat Neurosci 2003; 6:1255-63.

52. O'Brien KM, Schulte D, Hendrickson AE. Expression of photoreceptor-associated molecules during human fetal eye development. Mol Vis 2003; 9:401-9 <http://www.molvis.org/molvis/v9/a52/>.

53. Bowes C, van Veen T, Farber DB. Opsin, G-protein and 48-kDa protein in normal and rd mouse retinas: developmental expression of mRNAs and proteins and light/dark cycling of mRNAs. Exp Eye Res 1988; 47:369-90.

54. McGinnis JF, Austin BJ, Stepanik PL, Lerious V. Light-dependent regulation of the transcriptional activity of the mammalian gene for arrestin. J Neurosci Res 1994; 38:479-82.

55. Lombardi MS, van den Tweel E, Kavelaars A, Groenendaal F, van Bel F, Heijnen CJ. Hypoxia/ischemia modulates G protein-coupled receptor kinase 2 and beta-arrestin-1 levels in the neonatal rat brain. Stroke 2004; 35:981-6.

56. Iaccarino G, Tomhave ED, Lefkowitz RJ, Koch WJ. Reciprocal in vivo regulation of myocardial G protein-coupled receptor kinase expression by beta-adrenergic receptor stimulation and blockade. Circulation 1998; 98:1783-9.