Decreased Expression of the Insulin-like Growth Factor 1 Receptor by Ribozyme Cleavage

Abstract

purpose. Insulin-like growth factor (IGF)-1 and its receptor (IGF-1R) are associated with abnormal retinal neovascularization. Ribozymes were designed that selectively decreased the expression of the IGF-1R and these ribozymes were tested in angiogenesis models in vitro and in vivo.

methods. Two hammerhead ribozymes were designed that cleave the human IGF-1R mRNA. The ribozymes were cloned into recombinant adeno-associated viral vectors (rAAV). The rAAV constructs were transfected into human retinal endothelial cells (HRECs). IGF-1R mRNA and protein levels were examined and the modified Boyden chamber assay used to examine ribozyme effects on cell migration. These constructs were injected intravitreally into mice to determine the effect of the ribozymes on retinal neovascularization in a mouse model of oxygen-induced retinopathy.

results. Relative quantitative RT-PCR analysis showed that IGF-1R Rz1 reduced IGF-1R mRNA levels by 40% ± 10% (P = 0.003), and Western blot analysis showed a 41% ± 5% (P = 4.6 × 10⁻⁵) reduction of IGF-1R protein, confirming that this ribozyme reduces IGF-1R expression. IGF-1R Rz1 also reduced IGF-1–induced cell migration by 90% ± 5% (P = 2.9 × 10⁻⁹) showing that IGF-1R Rz1 reduces IGF-1R function in HRECs. IGF-1R Rz1 also reduced the amount of preretinal neovascularization by 65% ± 6% (P = 2.7 × 10⁻⁵), as measured by the average number of endothelial preretinal nuclei per section.

conclusions. These studies demonstrate that the IGF-1R ribozymes are effective at reducing the expression and function of the IGF-1R in vitro and in vivo. Therefore, the IGF-1R ribozymes are an effective method for studying the process of angiogenesis and may ultimately be effective as gene therapy tools for the reduction of pathologic retinal angiogenesis.

Aberrant retinal neovascularization is the hallmark of both proliferative diabetic retinopathy (PDR) and retinopathy of prematurity (ROP). Although vascular endothelial growth factor (VEGF) is currently viewed as the major effector for retinal neovascularization in both diseases, ^{1 2 3 4} recent studies point to a pivotal role for insulin-like growth factor (IGF)-1. Hellstrom et al. ⁵ used an IGF-1 knockout mouse model of ROP to demonstrate that the absence of IGF-1 prevents normal retinal vascular growth despite the presence of VEGF. They also examined IGF-1 serum levels in premature infants and postulated that serum levels of IGF-1 predict which infants will have ROP and that correction of serum IGF-1 levels toward normal may prevent the development of ROP in premature infants. In addition, several retinal cell types express IGF-1 in response to VEGF exposure, ⁶ and data indicate that VEGF induces IGF-1 and basic fibroblast growth factor (bFGF) production in human retinal endothelial cells (HRECs). 

IGF-1 is recognized as one of the progression factors that prompt competence factor–primed cells to proceed through the prereplicative (G₁) phase of the cell cycle. ^{7 8} Qureshi et al., ⁹ suggest that IGF-1 and epidermal growth factor (EGF) act synergistically to promote cell proliferation. EGF acts as a competence factor promoting cells to move from the G₀ to the G₁ phase, and then IGF-1 acts as a progression factor by stimulating mitosis (S phase). Similarly, platelet-derived growth factor (PDGF) is a competence factor that promotes cells to move from the G₀ to the G₁ phase and, with IGF-1, represents the main mitogenic action in serum. A number of studies suggest that altered IGF-1 serum levels may be clinically meaningful in diabetes. Patients with rapidly accelerating retinopathy have elevated levels of serum IGF-1. ¹⁰ In a large population-based study of 928 diabetic patients, Dills et al. ¹¹ found that higher levels of IGF-1 were associated with increased frequency of PDR after controlling for glycosylated hemoglobin, proteinuria, duration of disease, and age at diagnosis. In another study, patients with non-PDR in whom retinal neovascularization developed had elevated IGF-1 levels in serum at the onset of neovascularization compared with IGF-1 serum levels measured 3 months earlier. ¹² The local tissue levels of IGF-1 are probably as relevant as serum levels to the initiation of diabetic complications. A threefold increase of IGF-1 has been found in the vitreous of diabetic patients with PDR compared with nondiabetic individuals. ¹³  

The action of IGF-1 is mediated by binding to its cell surface receptor (IGF-1R). Cloning the IGF-1R has demonstrated that its overexpression could initiate mitogenesis and promote ligand-dependent neoplastic transformation in numerous cell types. ^{14 15 16 17} In addition, the overexpression of the IGF-1R enhances cell survival in response to death signals. ¹⁸ Conversely, antibodies to IGF-1R, antisense strategies against IGF-1 and IGF-1R, and dominant negative IGF-1R mutants all reduce cell survival and promote cell death. ^{16 19 20 21 22}  

We used hammerhead ribozymes to test the hypothesis that reduction of the IGF-1R mRNA will reduce the expression of the receptor and inhibit abnormal retinal neovascularization. Ribozymes are catalytic RNA molecules that cleave phosphodiester bonds between RNA nucleotides. ²³ Two types of ribozymes that are based on self-cleaving viral agents, hairpins and hammerheads, have been used as potential gene therapy agents. Hammerhead ribozymes have been used more commonly because they have a greater range of target sites. ²⁴  

Ribozymes offer the potential to block expression of specific growth factors or their receptors before protein translation. Several groups have used antisense RNA to control the expression of the IGF-1R in animals and in tissue culture. ^{20 25 26 27 28 29} These results indicate several sites in the IGF-1R mRNA that are accessible to antisense oligoribonucleotides and, hence, are susceptible to ribozyme attack. Resnicoff et al. ³⁰ have shown that treatment with antisense oligoribonucleotides can lead to apoptosis of glioblastoma cells in a rat model. Antisense oligodeoxynucleotides inhibit expression by activating RNAse H, which leads to degradation of the mRNA. ³¹ Because antisense DNA binding to target mRNA exists in an on-again, off-again dynamic equilibrium, high molar amounts of the antisense RNA nucleotides are necessary to achieve efficient inhibition of mRNA expression (IC₅₀ ranges from 30 to 100 nM). ³² Because of their catalytic ability, a lower molar amount of ribozyme molecules is needed to achieve efficient inhibition of mRNA expression. Therefore, ribozymes should be even more effective than antisense DNA in reducing the expression of the IGF-1R. ³³  

Our goal was to study the role that IGF-1 and the IGF-1R play in the pathways of angiogenesis. We designed hammerhead ribozymes that specifically cleave the human IGF-1R mRNA. In addition, these ribozymes anneal and cleave the mouse IGF-1R mRNA (Fig. 1) . We have performed extensive in vitro testing of these ribozymes, and we have determined the effect that they have on cultured HRECs and in an oxygen-induced mouse model of retinopathy. ³⁴ Our results demonstrate that these ribozymes are effective in inhibiting the function of the IGF-1R and in preventing preretinal neovascularization. 

Materials and Methods

Experimental Animals

All animal procedures used were in agreement with the NIH Guide for the Care and Use of Laboratory Animals and the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research and with institutional guidelines and were approved by the University of Florida Institutional Animal Care and Use Committee. Timed-pregnant C57BL/6J mice were purchased from Jackson Laboratories (Bar Harbor, MA). 

Synthetic RNA Targets and Ribozymes

RNA oligonucleotides for the active and inactive human IGF-1R hammerhead ribozymes and the human, rat, and mouse targets were purchased from Dharmacon (Boulder, CO) and deprotected according to the manufacturer’s protocol. RNA oligonucleotides were 5′-end labeled with [γ-³²P]-dATP (ICN, Irving, CA) using polynucleotide kinase (Promega, Madison, WI), as previously described. ^{35 36}  

Time Course Analysis of Ribozyme Cleavage

Time course analysis of cleavage was performed by using the RNA oligonucleotides as previously described. ^{35 36 37} For each reaction, 2 picomoles of ribozyme (15 nM final) in 40 mM Tris-HCl (pH 7.5) were incubated at 65°C for 2 minutes and then incubated at 25°C for 10 minutes. Dithiothreitol (DTT; 20 mM final), RNasin (4 U; Promega), and MgCl₂ (20 mM final) were added, and the mixture was incubated at 37°C for 10 minutes. Cleavage was initiated by the addition of the ³²P end-labeled RNA oligonucleotide target, and the reaction proceeded at 37°C. Variations on this protocol include incubation at 25°C at 1, 5, 10, and 20 mM MgCl₂. Aliquots were removed at various times and added to an equal volume of formamide stop buffer (90% formamide, 50 mM EDTA, 0.05% bromophenol blue, 0.05% xylene cyanol) and held on ice. The samples were then heat denatured at 95°C for 2 minutes and placed on ice, and the reaction products were separated on 10% polyacrylamide-8 M urea gels. The gels were analyzed on a phosphorescence imager (Amersham Biosciences, Sunnyvale, CA). 

Multiple-Turnover Kinetic Analysis

Multiple-turnover kinetic analysis was performed as previously described. ^{35 36} Reactions were performed in a final volume of 20 μL. Ribozyme (0.3 picomoles, 15 nM final) in 40 mM Tris-HCl (pH 7.5) was incubated at 65°C for 2 minutes and then incubated at 25°C for 10 minutes. DTT (20 mM final), MgCl₂ (20 mM final), and 4 U RNasin were added. The reactions were incubated at 37°C for 10 minutes, and cleavage was initiated by the addition of increasing concentrations of the target oligonucleotide (0–300 picomoles; 0–1500 nM final). The reactions were incubated at 37°C for a fixed interval determined in the time course analysis of cleavage. Variations on this protocol include incubation at 25°C in 1 mM MgCl₂. Reactions were terminated by the addition of 20 μL of formamide stop buffer and held on ice. The samples were then heat denatured at 95°C for 2 minutes and placed on ice, and the reaction products were separated on 10% polyacrylamide-8 M urea gels. The gels were analyzed on a phosphorescence imager (PhosphorImager; Amersham Biosciences). 

Cloning of the Hammerhead Ribozymes

We cloned a self-cleaving hairpin ribozyme into the SpeI and NsiI sites of the recombinant adeno-associated viral (rAAV) vector pTRUF-21. This version of the vector is designated p21NewHp. The target sequence for this hairpin ribozyme has been included just upstream of the hairpin ribozyme sequence. Ribozyme transcripts are directed by the chicken β-actin promoter/cytomegalovirus (CMV) enhancer and contain the actin introns. The plasmid also contains a neomycin resistance gene directed by the HSVtk promoter and a polyoma virus enhancer. The inverted terminal repeats of AAV2 flank the part of the plasmid encompassing the region from the β-actin promoter to the polyadenylation signal for the neomycin resistance gene. For each hammerhead ribozyme, two complementary DNA oligonucleotides (Invitrogen, Carlsbad, CA) were annealed to produce a double-stranded DNA fragment coding for each hammerhead ribozyme. All DNA oligonucleotides were synthesized with 5′-phosphate groups. The DNA oligonucleotides were designed to generate a cut HindIII site at the 5′ end and a cut SpeI site at the 3′ end after annealing. The DNA oligonucleotides were incubated at 65°C for 2 minutes and annealed by slow cooling to room temperature for 30 minutes. The resultant double-stranded DNA fragment was ligated into the HindIII and SpeI sites of p21NewHp. Transcription results in the hairpin ribozyme cleaving 8 bases downstream of the 3′ end of the hammerhead ribozyme. The ligated plasmids were transformed into electroporation-competent cells (SURE; Stratagene, La Jolla, CA) to maintain the integrity of the inverted terminal repeats. All clones were verified by sequencing. 

Transfection of Human Retinal Endothelial Cells

HRECs were grown to 70% confluence on 150-mm plates and transfected with the IGF-1R ribozyme plasmid constructs using diethylaminoethyl (DEAE)-dextran as a carrier as previously described. ^{38 39} Cells were washed once with PBS and 10.5 mL of medium containing 10% cell culture supplement (NuSerum; BD Biosciences, Bedford, MA) was added to the cells. A solution (324 μL total) of 10 μg of plasmid DNA in DEAE-dextran was added to the cells, followed by the immediate addition of 8.1 μL of 100 mM chloroquine (Sigma-Aldrich, St. Louis, MO). DEAE-dextran is a polymeric cation that associates with the plasmid DNA and carries the plasmid DNA into the cell. Chloroquine binds to the plasmid DNA and inhibits degradation of the plasmid by lysosomes. Cells were incubated for 4 hours at 37°C in 5% CO₂. The plates were shaken every 15 to 30 minutes at 37°C. After 4 hours, the cells were shocked for 1 minute by the addition of 10% dimethyl sulfoxide (DMSO) in PBS, washed twice with PBS and then placed in complete medium. DMSO increases membrane permeability and increases the efficiency of plasmid DNA entering the cell. The cells were harvested between 48 and 96 hours for further analysis. The death rate in the cells ranges from 20% to 50% with this protocol. Transfection efficiency was determined to be approximately 45% of surviving cells in cells transfected with a plasmid expressing green fluorescent protein (GFP; data not shown). This GFP-expressing plasmid is identical to the hammerhead-expressing plasmids, except that the hammerhead and hairpin ribozymes have been replaced by the GFP coding sequence. 

Relative Quantitative RT-PCR

Relative quantitative RT-PCR was performed on RNA isolated from HRECs transfected with plasmids expressing ribozymes (IGF-1R Rz1 and Rz2, active and inactive) and the control vector expressing no ribozyme. RNA was isolated from transfected HRECs by using two RNA extraction kits (GenElute Direct mRNA Miniprep Kit; Sigma-Aldrich, for mRNA; TRIzol Reagent; Invitrogen, for total RNA). Reverse transcription was accomplished with reverse transcriptase and a random hexamer (Superscript; Invitrogen) according to manufacturer’s protocol. 

PCR reactions to determine IGF-1R mRNA levels used gene-specific DNA oligonucleotides synthesized by Invitrogen (5′-AGGACGGCTACCTTTACCCGGCACAATTAC-3′ and 5′-ATCAACAGGACAGCGACGGGCAGAG-3′). The linear range of the amplification of the IGF-1R RT-PCR product was determined by using a PCR master mix (1 μL RT product/50 μL, 200 μM dNTPs, 1 mM MgCl₂, 0.4 μM IGF-1R oligonucleotides, 1× Taq DNA polymerase buffer [Sigma-Aldrich]), 2 U Taq DNA polymerase [REDTaq; Sigma-Aldrich]), 0.5 μCi/50 μL [α³²P]-dATP [ICN, Irvine, CA]). This master mix was separated into eight 0.2-mL tubes, and amplification was performed with an annealing temperature of 61°C. Samples were removed at even-numbered cycles starting at cycle 26. For each PCR sample, 5 μL was removed and 2 μL of formamide dye mix was added. The samples were heat denatured at 95°C for 3 minutes, cooled on ice, and applied to a 6% polyacrylamide-8 M urea gel. Dried gels were analyzed on the phosphorescence imager to determine the linear range of amplification. For this oligonucleotide pair, cycle 34 was determined to be within the linear range of amplification and was used in subsequent experiments. 

In the relative quantitative RT-PCR assays the level of IGF-1R mRNA was determined within each sample relative to an internal β-actin standard. β-actin mRNA levels were determined with a β-actin primer/competimer oligonucleotide set (QuantumRNA) from Ambion (Austin, TX). The competimer oligonucleotide pair from the β-actin primer set anneals to the same targets as the primer oligonucleotide pair, but they are blocked at their 3′ ends to prevent extension. This primer/competimer oligonucleotide set allowed us to determine the ratio of primer to competimer that yields a β-actin PCR fragment that is approximately equimolar to the IGF-1R PCR product. To determine the ratio of the primer/competimer oligonucleotide set necessary to achieve this, PCR reactions were performed as described earlier, and amplification proceeded for 34 cycles. The ratio of primer to competimer oligonucleotide was determined to be 10:1 at a final concentration of 0.4 μM for the combined primer/competimer mixture. 

PCR reactions were then performed to determine the relative amount of IGF-1R to β-actin, using the above conditions. PCR products were separated on 6% polyacrylamide-8 M urea gels and analyzed on the phosphorescence imager. 

Isolation of Protein from Transfected HRECs

Cells were grown in 150-mm tissue culture plates (Fisher Scientific, Atlanta, GA) and transfected as previously described. The cells were washed with PBS (BioWhittaker, Walkersville, MD) and scraped in ice-cold phenol-free Hanks’ balanced salt solution (HBSS; Invitrogen) containing 1 mM EDTA (Sigma-Aldrich). The cells were centrifuged in a tabletop centrifuge (Eppendorf 5810R; Fisher Scientific), using a swinging-bucket rotor at 1000 rpm for 5 minutes at 4°C. The pellet was suspended in 30 μL of lysis buffer (150 mM Tris-HCL and 150 mM NaCl [both from Fisher Scientific]), 1 mM EDTA, 1% Igepal CA-630 [Sigma-Aldrich], 1% protease inhibitor cocktail [Sigma-Aldrich]) and 1 mM DTT [Fisher Scientific]). The lysed cells were sonicated (Sonic Dismembrator, model 100; Fisher Scientific) for 2 seconds and centrifuged in a centrifuge (5415D Eppendorf; Fisher Scientific) at 13,200 rpm for 15 minutes at 4°C. The pellet was discarded and the amount of protein in the supernatant was determined using a bicinchoninic acid (BCA) protein assay kit (Pierce, Rockford, IL). 

Western Blot Analysis of IGF-1R Protein Levels in Transfected HRECs

A total of 80 μg of protein was separated on a 4% to 15% gradient polyacrylamide gel (Criterion; Bio-Rad Laboratories, Inc., Richmond, CA) at 120 V for 20 minutes and 140 V for 65 minutes and transferred (80 V for 5 hours) to a nitrocellulose membrane (Millipore Corp., Bedford, MA) using a blot cell apparatus (Bio-Rad Laboratories, Inc.) on ice at 4°C. The membranes was blocked in TBS containing 0.05% Tween (Sigma-Aldrich) and 5% milk for 1 hour at room temperature. For IGF-1R, the membrane was incubated with a 1:200 dilution of a rabbit polyclonal anti-human IGF-1R beta subunit IgG antibody (Santa Cruz Biotechnology, Santa Cruz, CA) at 4°C overnight. Blots were then washed with TBS containing 0.05% Tween and 5% milk for 5 minutes and incubated with a 1:2000 dilution of a horseradish peroxidase (HRP)–conjugated mouse anti-rabbit antibody (Santa Cruz Biotechnology) for 1 hour at room temperature. After incubation with the secondary antibody, the membranes were washed twice for 5 minutes and twice for 10 minutes with TBS (containing 0.05% Tween). After IGF-1R protein detection, the membranes were also used to detect β-actin protein levels. The levels of β-actin were determined with the same protocol used to determine IGF-1R levels. The primary antibody was a 1:5000 dilution of mouse monoclonal anti-β-actin antibody (Sigma-Aldrich) and the secondary antibody was a 1:7500 dilution of an HRP-conjugated anti-mouse IgG antibody (Sigma-Aldrich). The protein bands were visualized with an enhanced chemiluminescence (ECL) Western blot detection kit (Amersham Biosciences Ltd., Amersham, UK). Standard molecular weight markers (Bio-Rad Laboratories, Inc.) served to verify the molecular size of the IGF 1R β-subunit at 95.2 kDa and of β-actin at 42 kDa. Analysis of IGF-1R and β-actin protein levels was performed on computer (Image; Scion Corp., Frederick, MD). 

IGF-1–Induced Cell Migration Assays in Transfected HRECs

To determine the effect of the ribozymes on IGF-1R function in HRECs we used an assay to measure the levels of IGF-1R–induced cell migration on transfected cells. Transfected cells were assayed for their ability to migrate to increasing concentrations of IGF-1R using the modified Boyden chamber assay. ⁴⁰ Control cells, transfected with p21NewHp, were compared with cells transfected with the ribozyme constructs. After transfection HRECs were trypsinized (Trypsin-EDTA solution for endothelial cell culture, Sigma-Aldrich) until the cells became a single cell suspension. Trypsin was then inactivated, cells were washed three times in PBS and suspended in DMEM to a final concentration of 1000 cells/μL, and 30,000 cells (30 μL) were added per lower well in the blind-well chemotaxis chamber. Wells were then overlaid with a porous polyvinyl- and pyrrolidone-free polycarbonate membrane (12-μm pores) coated with 10% bovine collagen. The chemotaxis chamber was inverted and incubated in a humidified atmosphere of 5% CO₂ and room air at 37°C for 4 hours, to allow the cells to attach to the membrane. Chambers were then placed upright, and 50 μL of a cocktail containing VEGF (25 ng/mL), bFGF (25 ng/mL), and various concentrations of IGF (1 ng/mL, 10 ng/mL, or 100 ng/mL) were added to the upper wells. The chambers were then incubated for 12 hours as described earlier. Membranes were collected and cells on the attachment side (lower wells) were scraped off, leaving only those cells that migrated through the pores of the membrane into the upper wells. The cells on the membrane were then fixed in methanol, stained with a modified Wright-Giemsa stain (LeukoStat solution; Fisher Scientific, Springfield, NJ), and mounted onto glass slides. DMEM was used as a negative control in each experiment to determine the amount of random cell migration, and DMEM with 10% FBS served as a positive control in each experiment. Each test condition was assayed with a minimum of six replicate wells. 

Migrating cells were counted with a light microscope, and the number of migrating cells per well were calculated by averaging the number of cells counted in three separate, high-power (400×) fields. The values for the six replicate wells were then averaged, the statistical error calculated, and the results were compared for statistical significance by Student’s t-test. 

Intravitreal Injection into the Mouse Model

In the mouse model of oxygen-induced retinopathy, ³⁴ mice at postnatal day seven (P₇) are placed with their nursing dams in a 75% oxygen atmosphere for 5 days. On return to normal air (P₁₂), these mice show development of retinal neovascularization, with peak development occurring 5 days (P₁₇) after their return to normoxia. One day after birth (P₁), the mouse pups received a 0.5-μL intravitreal injection of plasmid (2 μg/μL) in HEPES-buffered saline into their right eyes. After the fifth day after return to normoxia (P₁₇), the animals were killed and the eyes removed and fixed in 4% paraformaldehyde and embedded in paraffin. Three hundred serial sections (6 μm) were cut sagittally through the cornea parallel to the optic disc. Every 30th section was placed on slides and stained with hematoxylin-eosin. This resulted in 10 sections from each eye being scored in a masked fashion using light microscopy to count endothelial nuclei extending beyond the inner limiting membrane into the vitreous, as previously described. ³⁴ The efficacy of treatment with a particular plasmid was then calculated as the average percentage of nuclei per section in the injected eye versus the uninjected eye. 

Statistical Analysis

All statistical analysis was performed on computer with Student’s t-test (Excel; Microsoft, Redmond, WA). P < 0.001 is signified by an asterisk on the graphs. 

Results

Time Course Analysis of Active IGF-1R Ribozymes

In a time course analysis of the active IGF-1R ribozymes (Fig. 2) , more than 90% of the target RNA was cleaved within the first 2 minutes of the reaction, demonstrating that both of these ribozymes have high catalytic activities. Figures 2B and 2C show the graphic representation of cleavage of the human targets for the gel in Figure 2A and of other time course reactions performed at 25°C and at 1, 5, 10, and 20 mM MgCl₂. This analysis was performed to establish, for multiple turnover kinetic analysis, an interval at which the reaction rate is linear and the ribozyme is saturated by target. Both ribozymes showed a high rate of cleavage at 37°C in 20 mM MgCl₂, and the reaction reached a plateau within 1 minute. Therefore, for multiple turnover kinetic analysis we dropped the reaction temperature to 25°C and varied the MgCl₂ concentration to find a convenient time point (Figs. 2B 2C) . No significant reduction in the rate of turnover was observed until the MgCl₂ concentration was dropped to 1 mM. Under these conditions a suitable interval (15% cleavage) was found at 3 minutes for IGF-1R Rz1 and 2 minutes for IGF-1R Rz2 (Figs. 2B 2C) . 

Time Course Analysis on Human and Rodent Targets

Because both ribozymes were designed to cleave the human IGF-1R mRNA, we also tested the ribozymes on targets homologous to both the rat and mouse IGF-1R mRNA, because these targets are not 100% identical with the human targets (Fig. 1) . This was important, because we planned to test these ribozymes in a mouse model. We performed time course analysis of the active IGF-1R ribozymes on RNA oligonucleotides of the mouse and rat target sequences and demonstrated that each ribozyme was capable of cleaving both the mouse and rat target sequence in vitro (Fig. 3) . 

Time Course Analysis of Inactive IGF-1R Ribozymes

We produced inactive versions of both of the IGF-1R ribozymes as shown in Figure 1 . These ribozymes still anneal to the targets but are catalytically inactive and should not cleave the targets. These inactive ribozymes are useful because they allow us to examine the antisense activities of the ribozymes in HRECs and in the mouse model. No cleavage was detected with either inactive ribozyme during the 2-hour reactions performed at 37°C in 20 mM MgCl₂ (Fig. 4) . 

Multiple Turnover Kinetic Analysis of IGF-1R Ribozymes

Multiple turnover kinetic analysis was performed on both the IGF-1R ribozymes, and the kinetic parameters were determined by using the best-fit straight-line equation from Lineweaver-Burke double-reciprocal plots (data not shown). Each analysis was performed a minimum of three times. Reactions were performed at 25°C in 1 mM MgCl₂, and the reactions were terminated after 3 minutes for IGF-1R Rz1 and after 2 minutes for IGF-1R Rz2. The parameters determined for IGF-1R Rz1 were V _max = 7.0 ± 0.3 nM/min, K _m = 47.1 ± 1.7 μM, and K _cat = 0.47 ± 0.01 minute⁻¹. The parameters determined for IGF-1R Rz2 were V _max = 2.8 ± 0.60 nM/min, K _m = 1.8 ± 0.1 μM, and K _cat = 0.2 ± 0.04 minute⁻¹. Based on these results, all active and inactive ribozymes were cloned into the p21NewHp vector for testing in HRECs and in the mouse model. 

Effect of IGF-1R Ribozymes on Expression of IGF-1R mRNA in HRECs

Plasmids expressing the active and inactive IGF-1R ribozymes and the control plasmid p21NewHp were transfected into HRECs to determine their effect on expression of IGF-1R mRNA. Figure 5 shows the results of relative quantitative RT-PCR used to determine the levels of IGF-1R mRNA relative to β-actin. Transfection with the catalytically inactive ribozymes resulted in no reduction of the IGF-1R mRNA signal (P = 0.5 for each) as expected from the in vitro cleavage results (Fig. 4) . Transfection with the active IGF-1R Rz1 resulted in a reduction of IGF-1R mRNA levels by 40% ± 10% (P = 0.003), and the active IGF-1R Rz2 resulted in a reduction of IGF-1R mRNA levels by 13% ± 6% (P = 0.003). 

Effect of IGF-1R Ribozymes on Expression of IGF-1R Protein in HRECs

Plasmids expressing the active and inactive IGF-1R Rz1 and Rz2 and the control plasmid p21NewHp were transfected into HRECs to determine their effect on expression of IGF-1R. Western blot analysis was used to determine the levels of IGF-1R relative to β-actin (Fig. 6) . Transfection with the catalytically inactive IGF-1R Rz1 reduced IGF-1R by 21% ± 3% (P = 5.4 × 10⁻⁵). Transfection with the active IGF-1R Rz1 reduced IGF-1R levels by 41% ± 5% (P = 4.6 × 10⁻⁵). This analysis demonstrates a significant antisense effect of the ribozyme in cell culture and suggests that approximately 50% of the decrease in IGF-1R in cells transfected by the active ribozyme may result from simply an antisense effect of the ribozyme binding to the target mRNA. 

Effect of IGF-1R Ribozymes on Migration of HRECs

To test whether ribozyme reduction of IGF-1R expression affects IGF-1–signaling, plasmids that express the active and inactive IGF-1R ribozymes and the vector p21NewHp were transfected into HRECs to determine their effect on growth factor–induced cell migration in a modified Boyden chamber (Fig. 7) . The p21NewHp vector was used as a control because it does not express a hammerhead ribozyme. Cells transfected with plasmid coding for active IGF-1R Rz1 reduced migration of cells by an average of 91% ± 5% (P = 2.9 × 10⁻⁹; Fig. 7A ), and cells transfected with plasmid coding for active IGF-1R Rz2 reduced the migration of cells by an average of 58% ± 11% (P = 4.1 × 10⁻⁹; Fig. 7B ). We clearly demonstrated an antisense effect in HRECs transfected with the inactive version of Rz1. The inactive version of Rz1 reduced cell migration by 51% ± 1% (P = 0.02) when compared with the control. This, along with the Western blot analysis (Fig. 6) suggests that a significant portion of the effect of the active Rz1 in HRECs results from the antisense binding of the ribozyme to the target mRNA. This type of antisense effect by hammerhead ribozymes has been demonstrated by others. ⁴¹  

Effect of IGF-1R Ribozymes on Neovascularization in the Mouse Model of Oxygen-Induced Retinopathy

We used the oxygen-induced model of retinopathy ³⁴ to test the effect of the ribozymes in vivo. Mice, at postnatal day 1 (P₁) were injected in the right eye with the plasmids expressing the active or inactive IGF-1R ribozymes or the vector p21NewHp. The extent of preretinal neovascularization was scored by counting the average number of preretinal nuclei per section of eye (Fig. 8) . The vector p21NewHp exhibited no significant difference from control (P = 0.17), as expected. The active IGF-1R Rz1 and the active IGF-1R Rz2 showed a reduction in the average number of preretinal nuclei per section of 65% ± 6% (P = 2.7 × 10⁻⁵) and 52% ± 5% (P = 2.3 × 10⁻⁵), respectively, when compared with control left eyes. The inactive versions of both ribozymes showed no significant decrease in neovascularization. This suggests that these ribozymes have no significant antisense effect in vivo and that the decrease in neovascularization found with the active ribozymes results almost entirely from cleavage of the IGF-1R mRNA. 

Discussion

We designed two hammerhead ribozymes that cleave the IGF-1R mRNA from both human and mouse, as determined by our in vitro analysis (Figs. 2 3) . In addition, cleavage of the mouse target by IGF-1R Rz1 was more efficient than cleavage of the mouse target by Rz2 (Fig. 3) . This suggests that IGF-1R Rz1 should be more effective in the mouse model than IGF-1R Rz2. The catalytically inactive versions of these ribozymes were incapable of cleaving the IGF-1R mRNA in vitro (Fig. 4) , and this was confirmed in cell culture by relative quantitative RT-PCR (Fig. 5) . Our kinetic analysis showed that both IGF-1R Rz1 and Rz2 have significant catalytic activity, and we therefore cloned both ribozymes and their inactive counterparts into rAAV vectors for testing in HRECs and in the mouse model. 

Relative quantitative RT-PCR analysis demonstrated that IGF-1R Rz1 and Rz2 reduced IGF-1R mRNA levels by approximately 40% and 13%, respectively (Fig 5) . However, because our transfection efficiency is approximately 45%, approximately 55% of the cells did not express the ribozymes, suggesting that the level of the IGF-1R mRNA is actually reduced by 87% in cells transfected with IGF-1R Rz1 and reduced by 28% in cells transfected with IGF-1R Rz2. As expected, we found no reduction of the IGF-1R mRNA by the inactive versions of IGF-1R Rz1 and Rz2 (Fig. 5) . Based on the levels of IGR-1R mRNA reduction, Western blot analysis was performed only on HRECs transfected by the active and inactive versions of IGF-1R Rz1 and the control vector. This analysis showed that the active IGF-1R Rz1 reduced IGF-1R protein levels by approximately 41% and the inactive version reduced IGF-1R protein levels by approximately 21% (Fig. 6) . The level of reduction by the active and inactive ribozymes may be as much as 91% and 46%, respectively, when allowing for transfection efficiency. These results demonstrate that the ribozymes were able to reduce the expression of the IGF-1R on both the transcriptional and translational levels. In addition, we found that a significant reduction of expression of the IGF-1R in cells occurred with the inactive IGF-1R Rz1. This demonstrated that the short targeting arms (6 bases each) of the hammerhead ribozyme were sufficient to produce an antisense effect that inhibited IGF-1R translation in HRECs. We also examined the ability of the IGF-1R ribozymes to inhibit the function of IGF-1R in HRECs (Fig. 7) . We showed that IGF-1R Rz1 and Rz2 inhibited the migration of transfected HRECs by approximately 91% and 58%, respectively. These results demonstrate that both ribozymes were effective at reducing the functional activity of the IGF-1R associated with IGF-1R–induced cell migration. We also showed that the inactive version of IGF-1R Rz1 reduced HREC migration by approximately 51%. This demonstrates a significant antisense affect for this ribozyme in HRECs. The inactive version of IGF-1R Rz2 was not tested. Overall, these results show that these ribozymes reduce the expression of the IGF-1R in HRECs and suggest that they should be effective in reducing abnormal retinal neovascularization in vivo. 

We demonstrated that the active versions of IGF-1R Rz1 and Rz2 reduce preretinal neovascularization by approximately 65% and 52%, respectively. This demonstrates that these ribozymes were effective at reducing preretinal neovascularization in this model. In addition, as suggested by the in vitro data, IGF-1R Rz1 was more effective in the mouse model than is IGF-1R Rz2. Based on the probabilities, there was no antisense effect of the inactive IGF-1R ribozymes in the mouse model. It is not unexpected to find a difference between the antisense affects found in HRECs in the in vitro cell migration assay and in the mouse model of oxygen-induced retinopathy. There is a threshold reduction in the expression of the IGF-1R that must be achieved to see a functional effect, either in cultured HRECs or in the endothelial cells of the mouse retina. We do not know what this level is, but it may be different for the two functional assays we have used. In addition, the number of IGF-1Rs on the surface of these cells differs substantially, because of species differences, physiological differences in cell state, and differences in cell culture versus the developing mouse retina. Rubini et al. ⁴² reported that, in mouse fibroblasts, quiescent cells maintain a level of approximately 15,000 to 20,000 IGF-1Rs per cell and that proliferating cells have a minimum of 30,000 IGF-1Rs per cell. Therefore, it is reasonable to assume, that the reduction of IGF-1R mRNA expression would have to be greater in proliferating cells than in cell culture for a blockade of IGF-1 signaling to occur. This reduction may be difficult to achieve, because the in vitro activity of Rz1 is modestly greater in the human mRNA than in that of the mouse (Fig. 3) . We are currently examining methods to increase the efficiency of in vivo transfection of the ribozyme expressing plasmids. 

We found a significant antisense inhibition of the IGF-1R function in vitro (Fig. 7A) , but this effect was reduced or absent in vivo (Fig. 8) . This demonstrates that most of the inhibition in vivo was due to the catalytic degradation of the target mRNA and points to the general utility catalytic ribozymes rather than antisense RNA strategies to inhibit gene expression. However, this comparison is not entirely fair, because antisense strategies generally use much longer antisense RNAs (up to several hundred nucleotides in length). 

In conclusion, hammerhead ribozymes specifically targeted against the IGF-1R mRNA inhibited the expression of IGF-1R mRNA in our study and result in the inhibition of the function of this receptor. Expression of these ribozymes resulted in the reduction of preretinal neovascularization in the oxygen-induced mouse model of retinopathy. These results confirm the involvement of the IGF-1R in the development of preretinal blood vessels. These ribozymes will be useful tools in studying the role of the IGF-1R in angiogenesis and in cell survival. In addition, it is also possible that the IGF-1R ribozymes will be useful in the treatment of disease states in which angiogenesis is involved. We have also demonstrated the general utility hammerhead ribozymes in the study of complex physiological pathways such as angiogenesis. Specifically targeting a single component of a pathway allows the knockdown of only that component and eliminates unwanted side effects or toxic concerns that can occur when using conventional drugs. In addition, we found that a single intravitreal injection of plasmid DNA resulted in expression of the ribozyme during the 17-day time course of the mouse model of oxygen-induced retinopathy. This single dose, coupled with the catalytic nature of hammerhead ribozymes, reduced the need for multiple injections or high concentrations of inhibiting agents and reduced potential discomfort in the experimental animals. 

 Supported by the Juvenile Diabetes Research Foundation International and by National Eye Institute Grants EY012601 and EY007739.

 Submitted for publication March 21, 2003; revised April 30, 2003; accepted May 20, 2003.

 Disclosure: L.C. Shaw, None; A. Afzal, None; A.S. Lewin, None; A.M. Timmers, None; P.E. Spoerri, None; M.B. Grant, 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.

 

(A) Secondary structure of a generic hammerhead ribozyme annealed to a 13-nucleotide target. Cleavage occurs 3′ of the X nucleotide. (B) The 13-nucleotide RNA target sequences recognized by the two IGF-1R ribozymes. Both ribozymes were designed specifically to cleave the human sequences after the underscored C. Boxed sequences show differences between the human and rodent sequences.

Figure 2.

 

(A) Autoradiograph of a 10% polyacrylamide-8 M urea gel used to separate the products of the ribozyme cleavage reactions. The target is a 13-base RNA oligonucleotide labeled with ³²P at the 5′ end. The labeled cleavage product is seven nucleotides in length. (B, C) Summary of time course of cleavage reactions on the human target RNA oligonucleotides for the IGF-1R Rz1 (B) and IGF-1R Rz2 (C). Reactions were performed at 37°C or 25°C and at the various MgCl₂ concentrations indicated. For example, 25/20 indicates a reaction performed at 25°C and 20 mM MgCl₂.

Figure 3.

 

Summary of time course of cleavage reactions on the human, mouse, and rat target RNA oligonucleotides for the IGF-1R Rz1 and IGF-1R Rz2. Ribozyme and target pairs are as indicated. For example, Rz1 Human is cleavage on the human target by IGF-1R Rz1.

Figure 4.

 

Summary of time course of cleavage reactions of the active and inactive versions of the two IGF-1R ribozymes on their respective human targets. IGF-1R Rz1i and IGF-1R Rz2i are the inactive versions of IGF-1R Rz1 and IGF-1R Rz2, respectively.

Figure 5.

 

Relative quantitative RT-PCR results from HRECs transfected with plasmid DNA. For each sample the level of the IGF-1R mRNA was determined relative to the level of the β-actin mRNA. The level of the IGF-1R mRNA in HRECs transfected with the p21NewHp cloning vector was set to 100%, and HRECs transfected with plasmids expressing ribozymes were compared with this standard.

Figure 6.

 

Western blot analysis of IGF-1R protein levels from HRECs transfected with plasmid DNA. (A) An autoradiograph from Western blot analysis of IGF-1R and β-actin levels on a gradient (4%–15%) SDS-polyacrylamide gel used to separate proteins isolated from HRECs transfected with the p21NewHp, p21IGF-1R Rz1 active, and pIGF-1R Rz1 inactive plasmids. (B) Summary of Western blot analysis (n = 5) on IGF-1R levels in HRECs transfected with the p21NewHp, p21IGF-1R Rz1 active, and pIGF-1R Rz1 inactive plasmids. IGF-1R levels were determined relative to the internal control of β-actin. The level of IGF-1R in cells transfected with the p21NewHp plasmid was set to 100%.

Figure 7.

 

Summary of the ability of transfected HRECs to migrate in a modified Boyden chamber in response to IGF-1. (A) HRECs were transfected with p21NewHp, p21IGF-1R Rz1 (active version), or p21IGF-1R Rz1i (inactive version). (B) HRECs were transfected with p21NewHp or p21IGF-1R Rz2. The DMEM and 10% FBS-DMEM are indicated on both graphs.

Figure 8.

 

Results of the injection of the IGF-1R ribozyme constructs in the oxygen-induced mouse model of retinopathy. Each histogram represents results from a minimum of eight eyes. The no-injection control was the left eye, and all plasmids were injected into right eyes. The y-axis indicates the average number of nuclei of preretinal endothelial cells per section. This number has been set to 100% for the no-injection control, and the data for the plasmid injections are directly compared with the control.