Tetracycline-Inducible System for Photoreceptor-Specific Gene Expression

Abstract

purpose. To develop a system for inducible photoreceptor-specific gene expression in transgenic mice. The tetracycline regulatory system was chosen because it possesses the useful property of direct control of gene expression through use of an exogenous agent, doxycycline, a tetracycline derivative.

methods. Transgenic mice were generated that carried the reverse tetracycline-controlled transactivator under the control of the photoreceptor-specific promoters for rhodopsin and interphotoreceptor retinoid-binding protein. These animals were crossed with transgenic mice carrying the lacZ reporter gene under control of the tetracycline operator cassette, creating doubly transgenic mice. Doxycycline was administered to induce expression of the reporter gene. Reporter assays were then performed to evaluate lacZ expression.

results. Doxycycline administration led to photoreceptor-specific expression of the lacZ reporter gene in the doubly transgenic mice. X-gal staining was restricted to photoreceptor inner segments and synaptic termini. Induction could be achieved by addition of the drug to the animals’ drinking water or by intravitreal injection. Induction was noted within 24 hours of doxcycline administration. Because of variability among animals, there was an approximate correlation, but not a clean dose–response curve relating drug dose to level of reporter expression.

conclusions. A transgenic system for inducible photoreceptor-specific gene expression has been developed. This system is currently being exploited to study the effects of regulated expression of genes of biological interest.

Transgenic mouse technology has been used to investigate a number of aspects of retinal biology, including definition of retinal promoters, studies of phototransduction, apoptosis, growth factor signaling and neovascularization, and development of animal models of human retinal disease. ^{1 2 3 4 5 6} Although a powerful approach, traditional transgenic studies are often limited by the inability to regulate the timing and level of transgene expression. Development of an inducible, cell type-specific system for retinal gene expression that allows the precise control of target gene transcription in vivo would help overcome these limitations and likely lead to further advances in the study of retinal development, function, disease, and aging. 

Several inducible systems, based on eukaryotic as well as prokaryotic regulatory mechanisms, have been described. ⁷ Many, however, have proven to be disappointing because of poor inducibility, pleiotropic inducer effects, or both. One of the most promising is the tetracycline-controlled transcription activation system (Tet system) designed by Gossen and Bujard. ⁸ The Tet system is based on the tetracycline repressor (tetR) of Escherichia coli, which binds intracellular tetracycline (tet) with a high affinity, releasing its own transcriptional repression. The tetracycline-dependent transactivator, tTA, is a fusion of tetR and the activation domain of the viral transcriptional regulator VP16. The addition of the VP16 domain converts a bacterial repressor into a eukaryotic transcriptional activator. In the absence of tet, tTA binds to its DNA response element (tet operator [tetO]) and activates transcription of a nearby downstream target gene. In the presence of tet, tTA does not bind to tetO, and the target gene is not activated. Because tet turns off expression, this system is often referred to as “tet-off.” As a complement to this approach, a “tet-on” system was developed by mutating tTA to generate a reverse tTA (rtTA) that only binds to tetO, and thereby activates transcription in the presence of tet. ⁹ Both tet systems offer the advantage of using an inducer, tetracycline, with well-characterized pharmacokinetics. Because the affinity of the tetR for tetracycline is very high, low to moderate amounts (subtherapeutic) of the antibiotic or its analogs are sufficient to modulate gene expression. Levels of induction with tTA as high as 10⁵-fold over background have been observed, whereas with rtTA, induction of target gene expression up to 10³-fold has been reported. ⁹  

Although much of the work with the Tet system was first done in tissue culture, several reports have now shown that it also functions in transgenic applications. ^{10 11 12 13} Taking advantage of transgenic tissue-specific promoters to direct the expression of the transactivator makes possible the combination of spatial and temporal control of transgene expression. In the eye, availability of inducible cell-type–specific promoters would allow a wide variety of retinal biology studies that were not previously possible. 

In this article we describe the development of a photoreceptor-specific inducible system. The opsin and interphotoreceptor retinoid-binding protein (IRBP) promoters were used to accomplish restricted expression of the rtTA. These promoters have been previously characterized and shown to direct expression to photoreceptors. ^{14 15 16} Because it seemed to us more desirable to be able to turn on a transgene by administering an activator, rather than by removing a repressor, we chose to use the rtTA rather than the tTA system. A transgenic reporter line, Ro1-lacZ, carrying the lacZ gene downstream of the tetO-minimal promoter cassette, was used to assess the fidelity and kinetics of induction. ^{17 18} The data presented show that the expression of the lacZ reporter gene is rapidly, reproducibly, and specifically activated in photoreceptors upon induction with the tetracycline analog doxycycline. 

Materials and Methods

Constructs

The rtTA transactivator, a 1044-bp EcoRI-BamHI fragment from plasmid pUHD 172-1 neo, was subcloned downstream of the 1.9-kb murine IRBP promoter and downstream of the 2.2-kb bovine opsin promoter, ¹⁶ generating plasmids pIRBP-rtTA and pOpsin-rtTA, respectively. For plasmid pOpsin-rtTA, the “−2174” construct, gBR200-lacF, was linearized with BamHI, blunt-ended, dephosphorylated, and then ligated with the blunt-ended 1044-bp EcoRI-BamHI rtTA fragment. To generate pIRBP-rtTA, the −1783 to +101, 1884-bp IRBP KpnI-BamHI fragment from plasmid p5 ¹⁴ (a generous gift from Dr. John Nickerson, Emory University, Atlanta, Georgia) was first cloned into gBR200-lacF, replacing the opsin fragment. Then, pUHD172-1 was linearized with EcoRI and a BamHI adaptor ligated to the ends, destroying the EcoRI sites. Subsequent digestion with BamHI released the 1044-bp rtTA fragment. This was then ligated into complementary BamHI ends (dephosphorylated) of the IRBP plasmid. Restriction analysis and sequencing were performed to identify clones with the desired orientation. These plasmids were then used to generate transgenic mice by standard techniques. 

Transgenics

Founders were identified by Southern blot analysis of tail DNA using the rtTA fragment as probe. Genotyping of subsequent offspring was done by tail PCR using transgene-specific primers. Genomic DNA was obtained from a 1- to 2-mm portion of mouse tail after overnight digestion in 0.5 mg/ml proteinase K, 10 mM Tris-HCl, pH 7.5, 100 mM NaCl, 20 mM EDTA, 2% Triton X-100 at 55°C. The PCR products generated were a 377-bp transgenic band and a 582-bp mP1 internal control band. The sequences of the primers used were: rtTA primer: 5′-GTTTACCGATGCCCTTGGAATTGACGAGT-3′; IC40 primer: 5′-GATGTGGCGAGATGCTCTTGAAGTCTGGTA-3′; IC41 primer: 5′-CAAGCAACTCCTGATGCCAAAGCCCTGCCC-3′. PCR conditions were 94°C, 4 minutes, 1 cycle; 94°C, 60 seconds, 62°C, 30 seconds, 72°C, 30 seconds, 35 cycles; and 72°C, 10 minutes, 1 cycle. 

rtTA-positive animals were crossed with a lacZ reporter strain (Ro1). ¹⁹ These animals carry the lacZ gene downstream of the tet operator binding sites (TRE). Although the reporter animals carry a second (independent) transgene, Ro1, this is irrelevant to our studies of the induction of lacZ because there should be no interaction between the two. Unless otherwise specified, all experiments were carried out with offspring resulting from crossing the rtTA lines with the lacZ reporter line. All procedures were performed in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. 

Tetracycline Delivery

Intravitreal Injection.

To determine maximal induction, the tetracycline analog, doxycycline (dox; Sigma, St. Louis, MO), was directly injected into the vitreous. Animals were anesthetized with a 2.5% working solution (in PBS) of avertin (1 g 2,2,2-tribromoethanol; Aldrich Chemical Company, Inc., Milwaukee, WI, dissolved in 1 ml tertiary amyl alcohol), 17 μl/g body weight administered intraperitoneally. Doxycycline injections were made into the right eye only, leaving the left eye as control. Control littermates carrying only one transgene, rtTA or lacZ, were also similarly injected. After application of 0.5% proparacaine hydrochloride (Allergan, Irvine, CA) and 2.5% phenylephrine hydrochloride (Bausch and Lomb, Rochester, NY) to the eye, the conjunctiva was grasped with forceps and an intravitreal injection was made through the sclera, choroid, and retina of the temporal hemisphere. A Hamilton syringe with a beveled 32-gauge needle was used to deliver approximately 1 μl doxycycline at 4 μg/μl in 1× PBS into the vitreous. Backflow of material was sometimes observed after withdrawal of the needle, as is common with this procedure. Puralube ointment (Fougera, Melville, NY) was administered to the eyes to prevent drying while the animals recovered from the anesthesia. Injections were done on 2 consecutive days. Eyes were collected on the third day after the first doxycycline injection and assayed for lacZ expression. 

Doxycycline in Drinking Water.

A time course experiment was performed by giving mice doxycycline, at a concentration of 10 mg/ml in 5% sucrose, in the water bottles for a period of 10 hours. (Water bottles were removed 14 hours before the start of the experiment.) Eyes were then collected 12, 24, and 48 hours after the start of doxycycline. Samples were stored at 4°C until they were all collected. LacZ expression was then assayed as described below, using the right eyes for whole mounts and the left for solution assays. 

Whole-Mount β-Galactosidase Assay

Induction of lacZ expression was detected by incubating whole-mount samples in 5-bromo-4-chloro-3-indolyl-β-d-galactoside (X-gal) substrate. After cervical dislocation, light cautery was applied at the superior pole of the cornea to maintain orientation. After enucleation, the sclera and choroid were dissected from the retina around the limbus, and the eyes were briefly fixed for 3 to 5 minutes in 0.5% glutaraldehyde in 1× PBS at room temperature. The samples were then incubated at room temperature in X-gal staining solution (1 mg/ml X-gal, 2 mM MgCl₂, 10 mM Ke₃F(CN)₆, 10 mM Ke₄F(CN)₆ · 3 H₂O, in 1× PBS), usually overnight. A Leica MS5 dissecting microscope with an Olympus SC35 (Melville, NY) camera attachment was used to photograph the samples. 

Frozen Sections

Eyes were embedded in OCT immediately after enucleation and fixation. Twelve- to 18-μm-thick cryosections were cut and placed on glass slides. Thereafter, sections were additionally fixed for 3 minutes at room temperature in 0.5% glutaraldehyde/1× PBS. Slides were then incubated in X-gal staining solution at room temperature. Alternatively, selected whole mounts were embedded in OCT subsequent to X-gal staining and further examined after sectioning. 

β-Galactosidase Solution Assay

The Luminescent β-galactosidase Detection Kit II (Clontech Laboratories, Inc., Palo Alto, CA) was used for enzymatic assay of lacZ activity in retinal extracts. Retinas were dissected and homogenized in detergent lysis buffer. Assays were performed as indicated in the protocol, using a tube luminometer (Monolight 2010; Analytical Luminescence Laboratory, Ann Arbor, MI) to measure light emission as 10-second integrals. 

Histology

Transgenic animals were fixed by transcardiac perfusion with 1% formaldehyde/2% glutaraldehyde in 0.1 M sodium phosphate buffer, pH 7.2. Eyes were dissected and embedded in Araldite. Ultrathin and 0.5-μm sections were prepared for electron and light microscopy, respectively. Sections for light microscopy were stained with 1% Toluidine blue, and sections for electron microscopy were stained with uranium and lead salts. 

Results

Generation of pOpsin-rtTA and pIRBP-rtTA Transgenic Lines

The opsin and IRBP promoters were used to direct photoreceptor-specific expression of the reverse tetracycline transactivator, rtTA. The plasmids pOpsin-rtTA and pIRBP-rtTA were used to generate separate lines of transgenic mice (Fig. 1) . Founders were identified by Southern blot analysis. Five Opsin-rtTA founders (founders A and C through F) and ten IRBP-rtTA founders (founders G through P) were obtained. Comparison of rtTA expression in the retina, as determined by Northern blot analysis, showed that line D was the highest expressing of the opsin-driven strains, with line A expressing at much lower levels (Fig. 2) . Of the IRBP lines, after correcting for sample-loading differences, line K expressed rtTA highly, whereas lines G, H, and N showed moderate expression, and line J showed a low level of expression (Fig. 2) . It should be noted that although the Northern blot analysis showed only one band with the opsin-driven lines, it showed two bands with the IRBP-driven lines. Although the reason for this is unclear, the observation of two distinct transcripts with an IRBP-driven transgene has been reported previously. ²⁰  

Because there have been reports of toxicity in cell culture tet systems ascribed to nonspecific tTA/rtTA activities, ¹ we examined the transgenic animals for any signs of retinal abnormality. The expression of the rtTA transgene had no detrimental effect on retina morphology, as monitored by light (data not shown) and electron microscopy (Fig. 3) . Examination of transgenic photoreceptors showed normal outer and inner segments, even in aged mice, that were indistinguishable from those in nontransgenic littermate controls. 

Intravitreal Injection of Doxycycline for Induction of Transgene Expression

Several rtTA lines were selected and crossed with a reporter strain, Ro1-lacZ, ^{17 18} to evaluate transgene inducibility. In an effort to maximize clarity in describing animals, we adopted the following nomenclature scheme: The pOpsin-rtTA transgenics are identified as “Opsin” followed by a letter designating the line (A, C, D, E, or F) and a number designating the specific animal. The pIRBP-rtTA transgenics are identified as “IRBP” followed by a letter designating the line (G through P) and a number designating the specific animal. For double transgenics, the presence of the Ro1-lacZ transgene is designated by “/lacZ.” As an example, a specific animal that is an offspring of pOpsin-rtTA line D crossed with a Ro1-lacZ animal might be designated “Opsin-D28/lacZ.” 

Taking advantage of the eye’s unique accessibility to manipulation, doxycycline was injected directly into the vitreous of double transgenic animals to determine the maximum capacity for induction and to avoid possible problems with the blood–retina barrier. Of the seven lines examined by retinal whole-mount analysis, varying levels of lacZ induction were observed. Consistent with the Northern results (Fig. 2) , of the pOpsin-rtTA animals the Opsin-D/lacZ animals showed the highest induction, as measured by the extent of lacZ staining in the retina (Fig. 4C ). There was no staining observed in any other region of the eye (data not shown). In addition, the Opsin-D/lacZ animals demonstrated the important characteristic that they had essentially no detectable transgene expression in the absence of inducer (Fig. 4K) . As expected, lines with low level rtTA RNA expression had minimal lacZ staining (e.g., Opsin-A28/lacZ, Fig. 4A ), those without detectable rtTA RNA expression did not show detectable lacZ staining (e.g., Opsin-C14/lacZ, Fig. 4B ), and singly transgenic animals without Ro1-lacZ were also negative for staining (e.g., Opsin-D26, Fig. 4I ). 

All three IRBP-rtTA crosses that were tested (IRBP-K, -G, and -N) showed some level of rtTA induction with intraocular doxycycline injection, with the IRBP-K line showing the heaviest staining (IRBP-K35/lacZ, Fig. 4G ). However, the density was not as high as seen with Opsin-D animals (Fig. 4C) . Figure 4F shows an IRBP-G69/lacZ whole mount, with a moderate amount of punctate staining evident in the right injected eye. Interestingly, the staining appears to be localized to the temporal side, perhaps reflecting a phenomenon similar to that observed with bovine rhodopsin promoter transgenics in which a gradient of lacZ expression is observed. ⁶ The IRBP-N32/lacZ retina showed the weakest staining, with only small clusters of positive cells located near the injection site (Fig. 4E , but difficult to appreciate in photographic reproduction). The low level of staining with IRBP-N was unexpected, given that the Northern blot analysis indicated a moderate level of rtTA expression for this line (Fig. 2) . 

The data from the intravitreal experiment showed that five of the seven rtTA lines tested showed some degree of induction, with the responses varying from robust (Opsin-D and IRBP-K) to barely detectable (IRBP-N). Of the highest responses, the induction seen with IRBP-K/lacZ animals was less than that observed with Opsin-D/lacZ animals, probably reflecting the relative strengths of their endogenous promoters. 

Photoreceptor Cell Specific Expression of Induced Transgene

Although the whole mounts showed retinal lacZ staining, it was necessary to confirm that the opsin and IRBP promoters indeed directed lacZ expression specifically to photoreceptor cells. The cellular localization of this staining was determined by examining frozen sections of the whole mounts. Figures 4D and 4H show sections taken from the doxycycline-injected eyes of Opsin-D28/lacZ and IRBP-K35/lacZ, respectively. These sections were taken from approximately the central posterior region of the eye and show that lacZ expression was indeed highly and specifically induced in photoreceptors. Most heavily stained were the photoreceptor inner segments and synaptic terminals, with a significant amount of staining also seen in the outer nuclear layer. Close examination of frozen sections from the Opsin-D28/lacZ uninjected eye (Fig. 4L) showed one or two lightly stained cells in the photoreceptor layer across the entire section of the retina. This could be due to either low non-specific background or interocular transport of the injected doxycycline. The section from IRBP-K35/lacZ (Fig. 4G) showed the same localization of stain to the photoreceptors, with slightly less intensity. 

Transgene Expression Can Also Be Induced by Doxycycline Administered via Drinking Water

The above experiments showed that intravitreal injection of doxycycline was able to specifically induce high level photoreceptor expression of lacZ in transgenic animals carrying the reporter gene under the control of either the opsin or IRBP promoters, and that the highest expression was observed with line Opsin-D/lacZ. We next examined whether doxycycline administered in drinking water could induce transgene expression. 

After drinking water administration of 10 mg/ml doxycycline to mice for 10 hours, assays were performed on eyes taken from Opsin-D/lacZ animals at 12, 24, and 48 hours after the initial induction (Fig. 5) . The right eye from each animal was used for whole-mount staining, whereas the retina from the left eye was homogenized for β-gal solution assays. For the whole mounts, staining in X-gal was for 6 hours. 

Significant induction was observed as early as 12 hours after doxycycline delivery (Opsin-D34/lacZ, Fig. 5B ). At this time point, distribution of the staining was uneven across the retina, with less staining in the inferior portion. There was not a noticeable difference in the intensities between the temporal and nasal halves. Frozen section analysis of this whole mount revealed individual stained photoreceptor cells (Fig. 5E) . The 24-hour sample showed a similar pattern but with perhaps more staining on the temporal half and slightly more intense reaction product (Fig. 5C) . Opsin-D3/lacZ, after 48 hours of induction, was stained dark blue continuously across the entire retina (Fig. 5D) . Individual photoreceptor cells were more intensely stained than at the shorter time points (not shown). The high level of staining at 48 hours was similar to the highest level achieved with the intravitreal injections. The control animal, Opsin-D79, which carried only one of the transgenes, had no detectable staining (Fig. 5A) . 

Many of the eyes showed uneven staining across the retina, suggesting a gradient of expression. Figures 5F 5G 5H show frozen sections of Opsin-D29/lacZ (24 hours of induction), taken at the margins of the ciliary epithelium from the temporal portion (Figs. 5F 5G) moving inwards to a more nasal region (Fig. 5H) . The staining is darker and more intense in the more temporal sections, decreasing as the sections progress nasally. 

Solution assays of lacZ activity in the retinal extracts from the left eyes correlated well with that of the right eye whole mounts. Table 1 shows that the values at 48 hours were significantly greater than those at the 12-hour time point. However, other than to note that there is a time-dependent increase, it is not possible to draw quantitative conclusions from the solution assays because we noted wide variability in results, possibly related to variation in the amount of doxycycline-containing drinking water consumed and genetic differences between the animals. 

To get a better sense of how much doxycycline was needed to induce transgene expression, some animals were given as little as 0.5 mg/ml doxycycline in their drinking water. Although induction seemed to take longer, by 48 hours quite strong expression could be induced in Opsin-D/lacZ animals (Figs. 5I 5J) . 

Discussion

We have developed a system for inducible photoreceptor-specific gene expression in transgenic mice. The Tet system designed by Gossen and Bujard ¹ was modified such that the expression of the reverse tetracycline-inducible transactivator, rtTA, was put under the control of the retina-specific promoters for opsin and IRBP. Several rtTA transgenic lines were evaluated by using doxycycline to induce expression of a lacZ reporter gene, after crossing to a tetO-lacZ reporter line. The data indicate that the reporter gene can indeed be induced to varying levels in the photoreceptors of animals carrying both transgenes. 

In assessing this system, there were a number of factors to consider, including potential pathology of rtTA expression, method of delivery of the inducer to the photoreceptors, the specificity of expression of rtTA, and the kinetics and dose response of induction of the transgene. Our light and electron microscopy studies indicate that expression of rtTA does not cause observable changes in retina histology. This is consistent with earlier studies with human cytomegalovirus (hCMV), clara cell 10-kDa protein (CC10) and calcium–calmodulin-dependent kinase II (CaMKII) promoters that also failed to report evidence of rtTA-related pathology. ^{11 21 22} Another concern is possible nonspecific effects of doxycycline itself, especially given reports that it can have pleiotropic properties, particularly with respect to NOS expression ^{23 24} ; however, we saw no evidence of morphologic changes in animals that had received doxycycline for up to 5 months. 

Although it had been shown previously that activation of tTA/rtTA could be achieved across the blood–brain barrier, ^{12 21} relatively little was known about the pharmacokinetics of doxycycline delivery to the murine eye. Our results empirically demonstrate that orally administered doxycycline crosses the blood–retina barrier with sufficient efficiency to activate rtTA in photoreceptors and can do so as rapidly as 12 hours after administration. Previous reports using the rtTA transactivator in other tissues have reported induction as early as 4 hours after oral doxycyline administration. ¹¹ The demonstration that intravitreal injection can also induce transgene expression may be useful in some types of studies because it makes possible the use of the other (uninjected) eye as an essentially identical uninduced control. 

The choice of promoter to fuse with the rtTA cDNA would obviously influence transgene expression. We chose the opsin and IRBP promoters because both had been extensively characterized in transgenic mice and shown to direct expression to photoreceptor cells. ^{15 16 25 26 27 28 29 30} IRBP expresses in both rods and cones. ³¹ The rhodopsin promoter is primarily active in rods and also in pinealocytes, but low-level activity has also been reported in transgenic cones. ^{32 33}  

The data described here show excellent and specific induction of the reporter gene in photoreceptor cells, in at least two transgenic lines (Opsin-D/lacZ and IRBP-K/lacZ), by both intravitreal injection and systemic administration through drinking water. The localization of the lacZ staining to the inner segments and synaptic terminals of the photoreceptors is consistent with β-galactosidase being a cytosolic protein. These data provide clear evidence of the appropriate expression and subsequent activation of the transactivator in photoreceptor cells. 

In addition to cell-type specificity, promoters can confer subtle spatial information. There are elements within the 2.2-kb bovine opsin promoter region that, depending on its integration site, can promote a gradient of expression in the eye from the superior-temporal to the inferior-nasal quadrant. ¹⁶ The current work suggests that there may also be a gradient of expression directed by the 1.9-kb murine IRBP promoter, perhaps related to the mosaic expression pattern that has been reported previously. ³¹ Such irregular expression patterns, which are probably a combination of position effects and lack of important regulatory elements in the promoter fragments used, must be considered in designing and interpreting inducible, as well as noninducible, transgenic experiments. 

³  Present address: Dana-Farber Cancer Center, Boston, Massachusetts.

 Supported in part by National Institutes of Health Grants EY04318, EY09769, EY00444, EY00331, EY09300–02, and ES 05752–02; the Foundation Fighting Blindness; the Ruth and Milton Steinbach and Macular Vision Foundations; and unrestricted funds from Research to Prevent Blindness (RPB). RAD is an ACS Research Fellow. DJZ is a recipient of a Career Development Award from RPB. DB is the Dolly Green Professor of Ophthalmology at UCLA.

 Submitted for publication June 30, 2000; accepted August 17, 2000.

 Commercial relationships policy: N.

 

Figure 1.

 

DNA constructs used to generate the inducible retina-specific transgenic system. The tetracycline-inducible transactivator, rtTA, was cloned downstream of either the opsin or IRBP promoters, to direct photoreceptor-specific expression. These plasmids were designated Opsin-rtTA and IRBP-rtTA, respectively. The mouse protamine 1 (mP1) intron and flanking exon sequences were used in the constructs to increase message stability and provide a poly(A) + signal. The restriction enzyme sites are as follows: E, EcoRI; H, HindIII; B, BamHI; K, KpnI; N, NaeI; B/E, blunt ligation of BamHI and EcoRI sites; B/E′, EcoRI site modified with a BamHI adaptor.

Figure 2.

 

rtTA expression in retinas of transgenic mice. Top: Northern blot analysis of retinal RNA (10 μg) hybridized with a rtTA probe. (A, C, D, E, and F) RNA extracted from Opsin-rtTA lines; (G, H, I, J, K, and N) RNA from IRBP-rtTA lines. NT, RNA from a non-transgenic control animal. Bottom: photomicrograph of ethidium bromide–stained gel demonstrating relative amounts of RNA loaded in each lane.

Figure 3.

 

Photoreceptor expression of rtTA is apparently not toxic to the retina. Electron micrograph of photoreceptors from an Opsin-rtTA (line D) transgenic animal.

Figure 4.

 

Intravitreal administration of doxycycline induces photoreceptor cell-specific lacZ expression. Retinal whole-mounts (A through C, E through G, I, and K) and frozen sections of whole mounts (D, H, J, and L). After enucleation, the sclera and choroid were removed by dissection. Eyes were fixed and then immersed in X-gal staining solution. After photography, the X-gal–stained whole-mount samples were embedded in OCT and 12-μm frozen sections were cut. (A) Opsin-A28/lacZ; (B) opsin-C14/lacZ; (C and D) opsin-D28/lacZ; (E) IRBP-N32/lacZ; (F) IRBP-G69/lacZ; (G and H) IRBP-K35/lacZ; (I and J) opsin-D26; (K and L) opsin-D28/lacZ, left (uninjected) eye.

Figure 5.

 

Administration of doxycycline in drinking water induces photoreceptor cell–specific lacZ expression. (A through H) Animals were administered doxycycline (10 mg/ml) in their drinking water for 10 hours and then killed at 12, 24, or 48 hours after beginning of drug dose. Retinal whole-mounts (A through D) and frozen sections (E through H) were prepared as described in Figure 4 . (A) Opsin-D79/lacZ control that did not receive doxycycline; (B) opsin-D34/lacZ, 12 hours; (C) opsin-D29/lacZ, 24 hours; (D) opsin-D3/lacZ, 48 hours; (E) opsin-D34/lacZ, 12 hours; (F through G) opsin-D29/lacZ, 24 hours; (I) opsin-D17/lacZ control that did not receive doxycycline; (J) opsin-D19/lacZ, 48 hours after administration of doxycycline at 0.5 mg/ml.

Table 1.

 

Retinal β-Gal Solution Assay

Animal	Hours after Induction	1	2	Average
Opsin-D79 ctrl	36	48,987	46,355	47,671
Opsin-D34/lacZ	12	237,034	269,714	253,374
Opsin-D29/lacZ	24	266,443	269,590	268,017
Opsin-D3/lacZ	48	1,676,957	1,767,413	1,722,185

 Retinas were dissected from the left eyes of Opsin-D/lacZ animals at intervals of 12, 24, and 48 hours after being given 10 mg/ml doxycycline in their drinking water. Except for Opsin-D79, all animals carried both the rtTA and lacZ transgenes. Retinal extracts were made by homogenization in a detergent extraction buffer. Duplicate readings were then taken of each luminescent β-gal assay reaction.

The authors thank John Nickerson (Emory University, Atlanta, GA) for providing plasmid p5 containing the murine IRBP promoter.