Identification of a Zebrafish Cone Photoreceptor–Specific Promoter and Genetic Rescue of Achromatopsia in the nof Mutant

Abstract

purpose. To identify in vivo a promoter fragment that specifically directs transgene expression in all zebrafish cone photoreceptors. This promoter subsequently enables GFP labeling of cones for facile morphologic analysis and purification and genetic rescue of achromatopsia.

methods. Promoter fragments of the zebrafish cone transducin α (TαC) gene were subcloned upstream of EGFP and microinjected into one- to two-cell–stage embryos. Promoter activity was assessed by fluorescence microscopy in wholemounts and retinal cryosections, and cone photoreceptors were purified by flow cytometry. Visual physiology was assessed by the optokinetic response (OKR) assay.

results. A 3.2-kb promoter fragment from zebrafish TαC specifically directed robust transgene expression in retinal cone photoreceptors and pineal photoreceptors. With this promoter, a stable transgenic line expressing EGFP in all zebrafish cone photoreceptors types was generated, and populations of cones were purified. Achromatopsia in the nof mutant was rescued using the identified promoter fragment to direct transgenic expression of wild-type cone transducin in mutant cones.

conclusions. A 3.2-kb TαC promoter fragment replicates the temporal and spatial pattern of endogenous TαC expression. The integrity of cones can be readily assessed in an EGFP transgenic line generated with this promoter, enabling downstream genetic and chemical screens for cone determinants.

Daylight vision is initiated by cone photoreceptors that are concentrated centrally in the macula of the human retina. The extraordinary ability of cones to adapt to ∼6 to 7 log units of illumination above dark-adapted threshold allows daylight vision, and cones with different spectral sensitivities enable color discrimination. ¹ In humans, loss of specific cone types causes partial color blindness, whereas general loss of cone function causes achromatopsia (total color blindness), cone dystrophy, or macular degeneration. Achromatopsia is characterized by difficulty seeing in bright light and complete loss of color discrimination. It can be the result of inherited mutations in genes encoding specific components of the cone phototransduction pathway: the α-subunit of cone transducin or the α- or β-subunit of the cone cGMP-gated cation channel. ^{2 3 4 5}  

The genetic and clinical heterogeneity of photoreceptor blindness suggests that numerous therapies are needed for treatment. Development of successful therapeutic approaches for rod-based blindness has been reported in animal models, including mice, rats, pigs, and dogs. ^{6 7} These approaches include pharmacological intervention (vitamin A supplementation, calcium channel blockers, antiapoptotics, and neuroprotectants), gene-based intervention (gene replacement, gene silencing or genetic expression of neuroprotectants) and cell-based intervention (cell transplant, encapsulated cell technology, and stem cells). However, the clinical utility of these approaches for humans remains unsubstantiated, and therapies for cone-based blindness have not been developed in animal models. 

In contrast to rod photoreceptors, ^{8 9 10 11} relatively little is known about the regulators of cone-specific expression. CRX is a transcription factor required for the expression of many rod and cone genes and for development of outer segments. ^{12 13} Mutations in the CRX gene can cause cone-rod dystrophy, an inherited retinal degeneration in which cone death precedes rod death. ¹⁰ Other factors affect the development of specific cone types. The transcription factors NR2e3 and NRL normally repress S-cone (blue) gene expression. Mutations in these genes cause excessive sensitivity to blue light. ^{14 15} Finally, targeted deletion of the thyroid hormone receptor beta 2 (TR beta 2) increases the number of S-cones at the expense of M-cones. ¹⁶  

A partial characterization of the mouse cone transducin promoter in conjunction with an enhancer from the IRBP gene has been reported ^{17 18} However, the scarcity of cones in normal mice limits their usefulness for studies of cone genetics. ¹⁹ Recently, transgenic Xenopus was used to investigate regulatory elements of the cone arrestin and phosphodiesterase genes ^{20 21} and regulatory elements that control UV cone-specific expression have been revealed by transient transgenesis in zebrafish. ^{22 23} We report herein that zebrafish can also be used to investigate factors that control gene expression in all subtypes of cones. 

Previously, we used a vision-dependent behavior, the optokinetic response (OKR), to isolate a zebrafish model of achromatopsia. The mutant, no optokinetic response f (nof) has a premature nonsense codon in the gene encoding cone transducin-α (TαC). ²⁴ Normally, expression of this gene is developmentally and spatially restricted to cone photoreceptors of the retina and to pineal photoreceptors. We report a characterization of the promoter for zebrafish TαC. We identified a ∼3.2-kb promoter fragment that directs cone-specific expression of EGFP in stable transgenic lines. This promoter fragment was also used to rescue achromatopsia in nof mutants by inducing expression of wild-type TαC cDNA in cones. 

Materials and Methods

Animal Use

The experiments were performed according to protocols approved by the institutional animal ethics committee, and adhered to the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. 

Isolation of Genomic Clones

“Down-To-The-Well” BAC pools (Incyte Genomics, St. Louis, MO) containing zebrafish genomic DNA were screened by PCR using primers specific for the zebrafish TαC cDNA. Four positive clones were isolated and characterized by Southern blot analysis, PCR, and DNA sequencing. 

Embryo Microinjections

Promoter fragments of the zebrafish TαC gene were subcloned upstream of the EGFP coding sequence in pEGFP-1 (BD-Clontech, Palo Alto, CA). Linearized plasmid was resuspended at 50 ng/μL in water plus 0.1% phenol red (Sigma-Aldrich, Poole, UK) and microinjected into one- to two-cell–stage embryos. 

Screening for EGFP Expression

Fish were anesthetized and placed on a depression slide for epifluorescence microscopy (Microphot-FX or Diaphot 300; Nikon, Tokyo, Japan). Retinal sections were analyzed by confocal microscopy (TCS SP/NT; Leica, Deerfield, IL; LSM 510; Carl Zeiss Meditec, Inc., Dublin, CA). 

Wholemount and Retinal Section Immunochemistry

Adult eyes and wholemount larval zebrafish were prepared for immunolabeling as described previously. ²⁵ Primary and secondary antibodies were diluted in blocking buffer (2% normal goat serum/1% BSA/1% Triton X-100/0.1 M phosphate buffer) and incubated overnight at room temperature. After washes in TBS/0.1% Tween, the samples were incubated at room temperature for 1 hour in secondary antibodies diluted in blocking buffer and mounted in antifade medium (Invitrogen-Molecular Probes, Eugene, OR). 

Flow Cytometry Purification of GFP Cones

Adult TG(3.2TαCP-EGFP) fish are euthanatized by lethal administration of benzocaine. Retinas are dissected in sterile PBS and dissociated for 10 minutes at 37°C with trypsin (1 mg/mL) and DNase I (10 U/mL). After 200 μL of trypsin inhibitor was added (10 mg/mL), the samples are vortexed for 5 seconds followed by centrifugation (5 minutes at 1200 rpm) and resuspension in PBS. The cell suspension was filtered through a 50-μm filter (Filcon GmbH, Taufkirckin/Munich, Germany) and sorted with a flow cytometer (FACSAria; BD Biosciences). Total RNA was isolated from the sorted cells (Qiashredder columns and RNeasy Extraction kit; Qiagen, Hilden, Germany). With the RT-PCR system (ThermoScript; Invitrogen) genes were amplified by using the following primer pairs: zfactin forward [F] 5′-CAA CGG CTC CGG CAT GTG-3′ and zfactin reverse [R] 5′-TGC CAG GGT ACA TGG TGG-3′; egfp F 5′-ATG GTG AGC AAG GGC GAG GAG CTG T-3′ and egfp R 5′-TAC AGC TCG TCC ATG CCG AGA GTG ATC C-3′; TαC-Ex1-F1 5′-AGAGGGGATAGAGCAACCAAAGG-3′ and TαC-Ex6-R1 5′-GCAAGTCACACCTTCGAAACAATG-3′; and 4439 F 5′-TTG AGC GCT GGA TGG TGG TC-3′ and 4440 R 5′-GAA GGA CTC GTT GTT GAC ACC-3′ for rhodopsin. 

Genetic Rescue of nof Mutants

A plasmid construct for the genetic rescue of blindness in nof mutants was created by subcloning the full-length coding region of the zebrafish TαC cDNA downstream of a zebrafish 3.2-kb cone-specific promoter fragment. Embryos from incrosses of nof heterozygotes were microinjected with the rescue construct and surviving larvae screened at 5 dpf for an optokinetic response. ²⁶ Genomic DNA was isolated from individual fish and genotyped for presence of the transgene and nof alleles by PCR, SSCP (single-strand conformational polymorphism), and DNA sequencing. ²⁶  

Results

Isolation of Zebrafish TαC Genomic Clones

Four positive zebrafish TαC genomic BAC clones (1A, 2B, 3D, 4F) were identified by PCR screening successive plate-pools with primers to the zebrafish TαC cDNA. The clones were confirmed to contain the zebrafish TαC locus by probing digested clones with radiolabeled zebrafish TαC cDNA on Southern blot analysis. Bioinformatic comparisons of the zebrafish TαC cDNA and the human TαC gene suggest that the zebrafish TαC gene contains eight exons and seven introns (Fig. 1A) . Sizes of the introns were approximated by analysis of genomic sequence and PCR amplification (data not shown). To subclone promoter fragments of the zebrafish TαC gene, BAC restriction fragments that hybridized with a zebrafish TαC exon 1–4 probe, were ligated into pZErO-2 (Invitrogen). DNA sequencing determined that subclones 1A10Zero and H5.2 contained ∼6 and ∼1 kb, respectively, of the 5′-flanking sequence (Fig. 1A) . 

Activity of Deletion Constructs in Transient G₀ Transgenics

To identify a promoter fragment of the zebrafish TαC gene that recapitulates its cone-specific expression, promoter deletions containing ∼0.7 to 6 kb of the TαC promoter (TαCP) upstream of the native translational start site were directionally subcloned into the promoterless pEGFP-1 reporter vector (Figs. 1A 1B) . The transcriptional activity of each reporter construct was evaluated by microinjecting embryos and analyzing EGFP expression at 5 dpf, by fluorescence microscopy (Figs. 1B 1C) . The 0.7-, 1.1-, and 2-kb promoter fragments directed only weak expression of EGFP in the eye, less than five positive cells per eye, however, retinal cryosections revealed that the weak expression was cone photoreceptor–specific (data not shown). In contrast, the 3.2- and 6-kb deletion constructs directed robust EGFP expression in the retina and pineal gland of G₀ transgenics (Figs. 1D 1E) . Transgene expression colocalized with cone markers, consistent with high-level cone-specific expression (data not shown). In summary, the proximal 0.7-kb fragment of the zebrafish TαC promoter contains the cis-elements necessary to direct cone-specific expression and a distal 1.2-kb fragment that distinguishes the 2- and 3.2-kb constructs contains regulatory elements that enhance this expression. 

Generation of a Stable TG(3.2TαCP-EGFP) Transgenic Line

To generate transgenic zebrafish lines that stably express EGFP in cone photoreceptors, zebrafish embryos were injected with linearized p3.2TαCP-EGFP-1. Approximately, 2000 embryos were injected and ∼200 survivors were raised to adulthood. Three founders from ∼200 screened fish stably transmitted the EGFP transgene to their G₁ offspring with transmission frequencies ranging from 3% to 10%. EGFP-positive G₁ transgenics were isolated and bred to generate homozygous G₂ fish that transmit the EGFP-expressing transgene to 100% of their offspring. 

Developmental Analysis of TG(3.2TαCP-EGFP) Expression

The temporal and spatial pattern of transgene expression was examined in wholemount 1- to 5-dpf G₂ transgenic embryos (Fig. 2) . Initially, EGFP expression was restricted to the pineal gland, with weak expression at ∼28 hpf and strong pineal-specific expression at ∼58 hpf (Figs. 2B 2C , respectively). EGFP expression was first observed in the retina at ∼70 hpf (Fig. 2D) , in the ventronasal patch where the first photoreceptors are known to differentiate. ²⁷ EGFP expression quickly spreads throughout the whole retina by ∼80 hpf (Figs. 2E 2F) . At 5 dpf, the 3.2-kb TαC promoter fragment specifically directs robust EGFP expression within the retina and pineal (Fig. 2G) . This developmental expression pattern recapitulates the known temporal and spatial pattern of endogenous zebrafish TαC expression and of cone photoreceptor differentiation. Thus, the 3.2 promoter fragment contains the cis-regulatory elements sufficient for retinal- and pineal-specific expression in vivo. 

Analysis of TG(3.2TαCP-EGFP) Expression in Adults

The activity of the 3.2-kb promoter fragment in adult transgenic fish was examined in cryosections. Robust EGFP expression is maintained in, and is specific to, the photoreceptor level of the adult retina and the pineal (Fig. 3) . Transverse and longitudinal sections demonstrate that the EGFP transgene is evenly expressed throughout the dorsal-ventral and temporal-nasal domains of the retina (Figs. 3B 3C) . Counterlabeling confirms that the EGFP expression in the retina colocalizes with known cone photoreceptor markers, including endogenous TαC, UV opsin, and zpr1, but not with rhodopsin (Fig. 3D 3E , and data not shown). Expression in the pineal is also seen to colocalize with zpr1 (Figs. 3F) . High-resolution confocal imaging demonstrates specific expression of EGFP in all four types of cone photoreceptor, short single cones (UV cones), long single cones (blue cones), and double cones (red-green cones). As anticipated, the untagged EGFP protein mostly localizes subcellularly to inner segments, nuclei and synaptic terminals (Fig. 3G) . 

Visual Function in TG(3.2TαCP-EGFP) Fish

There appear to be no deleterious effects resulting from transgenic expression of EGFP in cone photoreceptors. Three independent lines have been generated and transgene transmittance and expression have been stable in many generations for more than 4 years. Retinal morphology appears normal in embryonic and adult transgenics. Visual function, as assessed by the optokinetic and startle response, is normal (data not shown). 

Flow Cytometric Purification of Cone Photoreceptors

A better understanding of the molecular genetics of cone photoreceptors may be achieved if expression profiling can be performed on pure populations of cone photoreceptors. Specific and stable expression of EGFP in the cones of transgenic fish enabled us to purify cones by fluorescence-activated cell sorting. From five adult zebrafish retinas, we sorted 19,000 EGFP-positive cells, corresponding to ∼5.8% of all retinal cells analyzed. The cytometry histograms showed a significantly larger population of EGFP-positive cells in the transgenic retinas (10%) versus wild-type zebrafish retinas (<0.5%; Fig. 4A 4B ). Approximately 180 ng of cone photoreceptor RNA was purified from 19,000 sorted cells. 

Our initial findings that the 3.2-kb TαCP fragment drove cone-specific expression was based on morphologic analyses. To corroborate this finding, we used a more sensitive molecular genetic approach and analyzed the expression of rod, cone, and ubiquitous markers in the GFP-positive cells by PCR. Molecular analyses demonstrate that intact retinas or dissociated but unsorted retinas from transgenic fish expressed ubiquitous marker cDNAs (EGFP and actin), cone-specific markers (cone transducin) and rod-specific markers (rhodopsin; Fig. 4C ). In contrast the sorted GFP-positive cells expressed the ubiquitous- and cone-specific markers but not the rod-specific markers (Fig. 4C) . 

Genetic Rescue of Blindness in nof Mutants

We have isolated the visually compromised zebrafish mutant nof in OKR-based mutagenesis screens. ²⁴ A nonsense mutation in the zebrafish TαC gene was identified as causative for the recessive nof phenotype. We sought to rescue blindness in nof mutants by directing expression of wild-type zebrafish TαC cDNA under control of the 3.2-kb TαC promoter fragment. nof carriers were crossed, and offspring (25% homozygous for nof mutation) were injected at the one- to two-cell stage with the “rescue” DNA construct. Visual function was assessed in surviving embryos at 5 dpf by the OKR assay, and all animals that showed a positive visual response were genotyped. The nonsense mutation in the zebrafish TαC gene that results in the nof phenotype can be resolved from wild-type alleles by mobility differences on single-strand conformation polymorphism (SSCP) gels. ²⁴ The amplification primers used in this assay are complementary to intronic sequences that flank TαC exon2 and therefore amplify the endogenous TαC alleles but do not amplify the rescue transgene. Uptake of the transgene was verified by amplification of the TαC coding sequence using primers that are complementary to exonic sequences and that span an intron, thus resolving size differences between the endogenous alleles and the transgene. 

The blind nof phenotype was considered rescued in individual larvae that (1) genotyped as nof homozygotes, (2) typed positive for the transgene, and (3) exhibited an optokinetic response (Fig. 5) . Restoration of visual function was confirmed in 3 nof larvae. The rescue transgene was identified in 18 other nonrescued nof mutants (negative for OKR), corresponding to a rescue frequency of 14%. The low efficiency was expected due to the mosaic and variable expression directed by the injected transgene. However, these data provide proof that achromatopsia due to this single gene defect can be overcome in vivo by directing targeted expression of the wild-type protein. 

Discussion

In this study, we generated novel tools to help decipher the molecular genetics of cone photoreceptors. First, we identified a promoter fragment that specifically directs robust transgene expression in all cone types in vivo. Applying this cone-specific promoter we demonstrated that achromatopic blindness in the nof mutant can be reversed by directing transgenic expression of TαC. Second, we generated a stable transgenic line expressing EGFP in developing and mature cone photoreceptors. This line enables facile monitoring of the integrity of cone photoreceptors and the isolation of pure cone photoreceptor populations for expression profiling. 

Characteristics of the Cone Transducin Promoter

The 3.2-kb TαC promoter fragment that we identified was sufficient to direct transgene expression specifically to cone photoreceptors of the retina and to photoreceptors of the pineal gland. This expression pattern replicates the endogenous temporal and spatial expression of the endogenous zebrafish TαC gene. At 5 dpf, the promoter was active in all cone photoreceptors of the retina and continued to direct strong transgene expression in red-, green-, blue- and UV-sensitive cones of adult zebrafish. The 3.2-kb promoter also drove transgene expression in the pineal photoreceptors. Promoter activity in the pineal was observed earlier than in the retina, consistent with the rapid organogenesis of the pineal in zebrafish. ²⁸  

Our analysis of the activity of several promoter deletions of the zebrafish TαC gene represents initial steps in defining regulatory cis elements and trans factors controlling high-level, cone-specific expression. Consistent with previous studies of ocular genes, the proximal elements appear to control expression specificity and the distal elements control expression levels. ²⁹ The cis-elements sufficient to direct weak, but specific expression in cone photoreceptors lie within the 0.7-kb fragment most proximal to the transcription initiation site. Within the 0.7-kb proximal promoter bioinformatics analyses identify consensus sites for elements likely to control cone transducin expression PCE-1, ret4, PCE-II, otx, NR2e3, E-opsin, and Crx. ^{10 12 30 31 32} In addition, we identified a 1.2-kb distal enhancer region that enables high-level cone-specific expression. We are currently applying genetic and biochemical approaches to identify the factors that bind to the distal enhancer region. 

Rescue from Achromatopsia

Achromatopsia, or rod monochromacy, is a congenital, autosomal recessive visual disorder characterized by total color blindness, photophobia, reduced visual acuity and nystagmus. ^{2 3 4 5} Although disease prevalence is rare (1:30,000), founder effects can lead to areas with significant populations of affected individuals. ⁵ In achromats, rod photoreceptor function is normal while cone photoreceptors appear viable but fail to generate an electrical response. Mutations in the α-subunit of cone transducin causally link with human achromatopsia. The zebrafish mutant nof represents an in vivo model of achromatopsia, with loss of cone visual function due to mutations in the α-subunit of cone transducin. ²⁴ In the current study, we showed that cone visual function can be restored in homozygous nof mutants by directing transgenic expression of wild-type TαC cDNA under control of the 3.2-kb cone-specific promoter. The rescue frequency was moderate at 14%. However, this is expected, as we evaluated the rescue frequency in transiently transfected embryos. This random integration procedure is relatively inefficient and results in larvae with mosaic expression of the transgene (Fig. 1D) . Small patches of rescued cells in G₀ larvae are not likely to restore a functional OKR response. However, even moderate rescue provides proof of principle that monogenic achromatopsia can be overcome in vivo. We speculate that human achromatopsia can be overcome by gene therapy approaches, even when performed on affected adults, as there is no retinal degeneration. 

In this study we identified a promoter fragment capable of directing cone-specific expression in vivo. The robust and specific expression of EGFP in cone photoreceptors enables facile monitoring of cone photoreceptor integrity. Incorporation of this transgenic line into forward genetic studies and expression profiling of purified populations of cone photoreceptors will uncover novel determinants of cone photoreceptor function and survival. These factors will help decipher the molecular genetics of cone photoreceptors. 

 Supported by the Howard Hughes Medical Institute (JBH); Macular Degeneration Research, a program of the American Health Assistance Foundation (BNK); and Science Foundation Ireland Investigator Programme Grant 04/IN3/B559 (BNK).

 Submitted for publication August 17, 2006; revised October 10, 2006; accepted December 14, 2006.

 Disclosure: B.N. Kennedy, None; Y. Alvarez, None; S.E. Brockerhoff, None; G.W. Stearns, None; B. Sapetto-Rebow, None; M.R. Taylor, None; J.B. Hurley, 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.

 

Figure 1.

 

Characterization of the zebrafish TαC gene and reporter construct activity. (A) Schematic of the zebrafish TαC gene, including the deduced intron-exon structure and mapped restriction sites. Restriction fragments of the BAC clones positive for TαC by Southern blot analysis with the indicated cDNA probes, were subcloned to generate plasmids 1A10Zero, H5.2 and H5.3. Exons are boxed and numbered. The 5′ limit of promoter-reporter constructs in the EGFP-1 vector is also indicated. Also shown is the deduced zebrafish TαC cDNA, with the base pair size of the exons and the location of the nonsense mutation in nof highlighted. (B) Schematic of the zebrafish TαC reporter constructs and their corresponding activity at 5 dpf after microinjection into zebrafish embryos. At least 100 larvae for each construct were scored based on the number of GFP-positive cells in the eye (+, 1–5; ++, 5–50; +++, >50 EGFP) and the constructs were scored based on the highest activity. (C) Wholemount labeling of a 5-dpf zebrafish embryo shows expression of endogenous zebrafish TαC in the eye. Magnification, ×25. (D) Representative example of the high-levels of EGFP expression driven by the 6.0TαCP-EGFP construct in injected G₀ larvae at 5 dpf. (E) Stacked confocal z-series images of a retinal section from G₀ larvae injected with 3.2TαCP-EGFP, demonstrating EGFP expression in cells with a cone photoreceptor morphology.

Figure 2.

 

Developmental expression of EGFP in TG(3.2TαCP-EGFP) transgenics. (A) Schematic of the 3.2TαCP-EGFP construct. (B–G) Analysis of the developmental pattern of expression directed by the 3.2-kb TαCP promoter fragment in live transgenic embryos. Transgenic EGFP expression is initially observed in the pineal (pin) at ∼28 hpf (B), progressing to strong pineal-specific expression at ∼58 hpf (C). At 70 hpf, in addition to strong pineal expression, the 3.2-kb promoter directs weak expression of EGFP in the ventronasal patch (vnp) of the retina where the first photoreceptors are known to differentiate (D). Between 75 and 80 hpf (E, F) transgenic EGFP expression directed by the promoter quickly spreads throughout the margins of the retina (ret). At 5 dpf (G), the 3.2-kb TαC promoter fragment specifically directs robust EGFP expression within the retina and pineal.

Figure 3.

 

Analysis of transgene expression directed by the 3.2 kb promoter in adult TG(3.2TαCP-EGFP) transgenics. (A) Schematic of the 3.2TαCP-EGFP construct. (B) A representative coronal section from serial sections through a whole 1-month transgenic eye. EGFP expression directed by the 3.2-kb promoter is specific to the photoreceptor layer. The transgene appears evenly expressed throughout the superior–inferior axis of the retina. (C) A representative merged fluorescence micrograph from serial sagittal sections through the eye of a 1-month transgenic fish. The EGFP expression directed by the 3.2-kb promoter is located in the photoreceptor layer (prl) flanked by the retinal pigment epithelium (RPE) and the outer nuclear layer (onl) staining blue with DAPI. The transgene appears evenly expressed throughout the anterior-posterior axis of the retina. (D) Confocal fluorescent micrograph of a coronal section through a TG(3.2TαCP-EGFP) transgenic fish demonstrating the EGFP directed by the 3.2 kb promoter (green) does not colocalize with rhodopsin (red), a rod photoreceptor marker. (E) Merged bright-field and fluorescence micrographs of a coronal section showing that transgenic EGFP expression colocalizes in the retina with endogenous TαC expression, a marker of cone photoreceptors. (F) A section through the pineal of an adult transgenic, highlighting the expression of EGFP (green) in the pineal colocalized with endogenous TαC (red). (G) Stacked confocal images of high-magnification optical z-sections through a transgenic retina. The image demonstrates specific expression of EGFP in all four types of cone photoreceptor, short single cones (ssc), long single cones (lsc), and double cones (dc), mostly localizing in inner segments, nuclei and synaptic terminals. sup, superior retina; inf, inferior retina, ant, anterior; pos, posterior; rpe, retinal pigment epithelium; prl, photoreceptor layer; onl, outer nuclear layer.

Figure 4.

 

Flow cytometry purification of cone photoreceptor cells. Cone photoreceptors can be purified from TG(3.2TαCP-EGFP) transgenics by fluorescent activated cell sorting. The histograms for dissociated retinal cells from wild-type control (A) and transgenic TG(3.2TαCP-EGFP) retinas (B) showing the forward scatter (FSC) plus green fluorescence (FITC) profiles. Insets: same samples plotted as cell number (count) versus green fluorescence (FITC). There was a large population of highly fluorescent green cells unique to the dissociated transgenic retina. Cells in equivalent areas were sorted using identical parameters for transgenic and wild-type populations and represented 10% of the total cells in the disassociated transgenic retinas, but <0.5% of total cells in corresponding wild-type retinas. (C) RT-PCR confirmation of cone photoreceptor purification. RNA was isolated from wild-type zebrafish retinas (RET), dissociated, nonsorted transgenic retinas (NS) and sorted EGFP-positive cells of transgenic retinas (S). Other templates included no template (BL), RNA without RT reaction (−RT), and a vector encoding EGFP (V). The amplified cDNAs were EGFP (690 bp), β-actin (690 bp), TαC (550 bp), and rhodopsin (180 bp). RT-PCR reactions demonstrated the presence of ubiquitous (actin), cone-specific (TαC) and rod-specific markers (rhodopsin) in wild-type retinas and nonsorted transgenic retinas. However, only the ubiquitous and cone-specific markers were present in the sorted cells from transgenic retinas.

Figure 5.

 

Genetic rescue of blindness in the nof mutant. Visual impairment in nof mutants can be rescued by injecting a transgene in which the 3.2 kb TαCP promoter fragment directs expression of the wild-type zebrafish TαC cDNA. (A) Schematic of the endogenous TαC gene alleles, the rescue transgene, and the primers used to distinguish them. (B) Embryos from incrosses of nof heterozygotes (25% of offspring homozygous for recessive phenotype) were injected with the rescue construct. At 5-dpf visual function was assessed using the optokinetic response. Genomic DNA of individual larvae with a normal optokinetic response was screened for presence of the rescue transgene by PCR. The exonic primers used span introns thus distinguishing the transgene from the native genomic alleles. Larval with a normal OKR and incorporation of the transgene were genotyped to identify those homozygous for the nof mutation. Genotyping was achieved by SSCP analysis with intronic primers flanking exon 2, which do not amplify the transgene and which resolve the nof polymorphism. Confirmation of the nof homozygosity was confirmed by DNA sequence analysis. Three larvae that typed as homozygous for the nof mutation but that had incorporated the rescue transgene and that had a normal OKR were identified, corresponding to a rescue frequency of ∼14%.

The authors thank Alfonso Blanco, Paulette Brunner, Daniel Oprian, Padraig O’Murchu, and Laura Swaim for technical assistance; Tom Vihtelic and David Hyde for zebrafish opsin antibodies; and the Zebrafish International Resource Center for the monoclonal antibody zpr1.