| Nucleic Acids Research | Pages |
Bioactivity of anti-angiogenic ribozymes targeting Flt-1 and KDR mRNA
Introduction
Materials And Methods
Ribozymes
In vitro ribozyme cleavage assays
Cell culture
Ribozyme and ribozyme/LipofectAMINETM formulation
Proliferation assay
RNase protection assay
Rat corneal pocket assay of VEGF-induced angiogenesis
Results
Ribozyme-mediated reduction of VEGF-induced cell proliferation
Specific reduction of Flt-1 or KDR mRNA by ribozyme treatment
In vitro activity of anti-Flt and anti-KDR ribozymes
Ribozyme-mediated reduction of VEGF-induced angiogenesis in vivo
Discussion
Anti-Flt and KDR ribozymes in cell culture
Anti-Flt and KDR ribozymes in vivo
Anti-Flt and KDR ribozymes and VEGF receptor function
Anti-Flt-1 and KDR ribozymes as therapeutics
Summary
Acknowledgements
References
Bioactivity of anti-angiogenic ribozymes targeting Flt-1 and KDR mRNA
Received as resubmission May 12, 1999; Accepted May 14, 1999
ABSTRACT Vascular endothelial growth factor (VEGF) and its receptors Flt-1 and KDR play important roles in physiological and pathological angiogenesis. Ribozymes that target the VEGF receptor mRNAs were developed and their biological activities in cell culture and an animal model were assessed. Ribozymes targeting Flt-1 or KDR mRNA sites reduced VEGF-induced proliferation of cultured human vascular endothelial cells and specifically lowered the level of Flt-1 or KDR mRNA present in the cells. Anti-Flt-1 and KDR ribozymes also exhibited anti-angiogenic activity in a rat corneal pocket assay of VEGF-induced angiogenesis. This report illustrates the anti-angiogenic potential of these ribozymes as well as their value in studying VEGF receptor function in normal and pathophysiologic states.
INTRODUCTION
Physiologic processes such as vascular development, wound healing and the endometrial cycle depend upon angiogenesis (1-4). Angiogenesis also plays an important role in the pathogenesis of several diseases such as diabetic retinopathy, neovascular glaucoma, solid tumor growth and metastasis, endometriosis, rheumatoid arthritis and psoriasis (2,5,6). Because few effective therapies exist for these diseases, the development of anti-angiogenic agents could have a substantial impact on their treatment.
Vascular endothelial growth factor (VEGF) has been shown to be an important mediator of vascular endothelial cell proliferation and migration (4,7). As a consequence, VEGF-induced angiogenesis has been shown to support solid tumor growth (6,8) and to underlie the pathology of certain neovascular diseases of the eye (9,10). In these diseases, ischemia-induced local hypoxia is thought to enhance VEGF secretion and thereby stimulate angiogenesis (4,11-13). A causative role for VEGF has been established in diabetic retinopathy (9,10,14-16) and is implicated in age-related macular degeneration (17,18). VEGF production is also elevated in synovial explants from patients with rheumatoid arthritis (19) and in psoriatic skin (20).
Endothelial cells are the primary cell type known to express VEGF receptors (21). The two human VEGF receptor subtypes are known as Flt-1 and KDR. Although both are tyrosine kinases, the precise function of each VEGF receptor subtype in the process of normal angiogenesis is not known. In addition, their roles in pathologic angiogenesis may be obscured by changes in their relative expression as a function of disease (20,22) and by the presence of other cytokine and inflammatory mediators. The reduction of VEGF receptor signaling represents a possible anti-angiogenic strategy as well as an avenue to explore the relative role of each receptor in pathologic neovascularization.
Ribozymes are RNA-based enzymes that bind to and cleave RNA molecules in a sequence-specific manner. This results in the destruction of the message and decreased expression of the encoded protein. Because ribozymes can be designed to target practically any RNA, the potential use of ribozymes to reduce gene expression is quite broad. Therapeutic applications of ribozymes have been facilitated by the development of catalytically active ribozymes that are chemically stabilized to resist nuclease degradation (23) and maintain activity in vivo (24).
Here we describe the use of hammerhead ribozymes targeting Flt-1 andKDR mRNA as potential inhibitors of angiogenesis. Ribozymes targeting each receptor message were screened in cultured human microvascular endothelial cells (HMVEC-d) to determine their effect on VEGF-induced cell proliferation. Ribozymes that inhibited cell proliferation were also shown to decrease the abundance of targeted mRNA present in the cell. In addition, angiogenesis was significantly reduced when either an Flt-1 or aKDRtargeted ribozyme was tested in a rat corneal model of VEGF-induced angiogenesis. Our results suggest that anti-Flt-1 and/or anti-KDR ribozymes may be effective therapeutic agents as well as useful tools for identifying the roles of Flt-1 and KDR receptors in angiogenesis.
MATERIALS AND METHODS
Ribozymes
Hammerhead ribozymes and controls designed to have attenuated activity (attenuated controls) were synthesized and purified as previously described (25). The attenuated ribozyme controls maintain the binding arm sequence of the parent ribozyme and thus are still capable of binding to the mRNA target. However, they have two nucleotide changes in the core sequence that substantially reduce their ability to carry out the cleavage reaction. Ribozymes were designed to target Flt-1 or KDR mRNA sites conserved in human, mouse and rat. In general, ribozymes with binding arms of 7 nucleotides (nt) were designed and tested. If, however, only 6 nt surrounding the cleavage site were conserved in all three species, 6-nt binding arms were used. A subset of ribozyme and attenuated control sequences and modifications are listed in Table 1. Other sequences are available upon request. In studies reported here, two previously described ribozyme motifs (23) were tested. While both motifs yielded similar results in all test systems, data are presented for 2[prime]-NH2 uridine modified ribozymes in cell proliferation studies and for 2[prime]-C-allyl uridine modified ribozymes in RNase protection, in vitro cleavage and corneal studies.
Table 1. Ribozyme and attenuated control sequences and locations of modified nucleotides
[dagger]Modifications are indicated as follows: 2[prime]-O-methyl nucleotides, lowercase; ribonucleotides, uppercase G, A; inverted 3[prime]-3[prime] deoxyabasic, B. Two ribozyme motifs were tested (23). For the NH2-modified ribozymes, U indicates the two core positions where 2[prime]-NH2 uridines are present. For the C-allyl-modified ribozyme, one 2[prime]-C-allyl uridine is present at the 5[prime]-most U, while the second or 3[prime]-most U is 2[prime]-O-methyl uridine. The positions of four phosphorothioate linkages at the 5[prime] end are indicated by a subscript s. Changes in the attenuated controls are underlined. Nucleotides comprising the base-paired region of stem II are in italics. A 3 bp stem II is shown above. Ribozymes or attenuated controls referred to in the text with a 4 bp stem II have one additional base pair such that the stem II/loop sequence is ggccgaaaggcc.
In vitro ribozyme cleavage assays
In vitro RNA cleavage rates on a 15-nt synthetic RNA substratewere measured as previously described (23). Briefly, the ribozymes or attenuated controls (500 nM) were incubated with stoichiometric amounts of a short complementary RNA substrate (and trace amounts of 5[prime]-32P-end-labeled substrate) in 50 mM Tris-HCl buffer, pH 8.0, 40 mM MgCl2 at 37°C. The magnesium-dependent cleavage reactions were quenched at specified timepoints with EDTA-containing buffer. Following polyacrylamide gel electrophoresis, full-length and product bands were quantified using a Molecular Dynamics PhosphorImager.
Cell culture
Human dermal microvascular endothelial cells (HMVEC-d, Clonetics Corp.) were maintained at 37°C in flasks or plates coated with 1.5% porcine skin gelatin (300 bloom, Sigma) in Growth medium (Clonetics Corp.) supplemented with 10-20% fetal bovine serum (FBS, Hyclone). Cells were grown to confluency and used up to the seventh passage. Stimulation medium consisted of 50% Sigma 99 media and 50% RPMI 1640 with L-glutamine and additional supplementation with 10 µg/ml insulin-transferrin-selenium (Gibco BRL) and 10% FBS. Cell growth was stimulated by incubation in Stimulation medium supplemented with 20 ng/ml of either VEGF165 or bFGF. VEGF165(165 amino acids) was selected for cell culture and animal studies because it is the predominant form of the four native forms of VEGF generated by alternative mRNA splicing (4). Cell culture assays were carried out in triplicate.
Ribozyme and ribozyme/LipofectAMINETM formulation
Cell culture. Ribozymes or attenuated controls (50-200 nM) were formulated for cell culture studies and used immediately. Formulations were carried out with LipofectAMINETM (Gibco BRL) at a 3:1 lipid to phosphate charge ratio in serum-free medium (Opti-MEMTM, Gibco BRL) by mixing for 20 min at room temperature. For example, a 3:1 lipid to phosphate charge ratio was established by complexing 200 nM ribozyme with 10.8 µg/µl LipofectAMINETM (13.5 µM DOSPA).
In vivo. For corneal studies, lyophilized ribozyme or attenuated controls were resuspended in sterile water at a final stock concentration of 170 µg/µl (highest dose). Lower doses (1.7-50 µg/µl) were prepared by serial dilution in sterile water.
Proliferation assay
HMVEC-dwere seeded (5 × 103 cells per well) in 48-well plates (Costar) and incubated for 24-30 h in Growth medium at 37°C. After removal of the Growth medium, cells were treated with 50-200 nM LipofectAMINETM complexes of ribozyme or attenuated controls for 2 h in Opti-MEMTM. The ribozyme/control-containing medium was removed and the cells were washed extensively in 1× PBS. The medium was then replaced with Stimulation medium or Stimulation medium supplemented with 20 ng/ml VEGF165 or bFGF. After 48 h, the cell number was determined using a CoulterTM cell counter. Data are presented as cell number per well following 48 h of VEGF stimulation.
RNase protection assay
HMVEC-d were seeded (2 × 105 cells per well) in 6-well plates (Costar) and allowed to grow for 32-36 h in Growth medium at 37°C. Cells were treated with LipofectAMINETM complexes containing 200 nM ribozyme or attenuated control for 2 h as described in Proliferation Assay and then incubated in Growth medium containing 20 ng/ml VEGF165 for 24 h. Cells were harvested and an RNase protection assay was carried out using the Ambion Direct Protect kit and protocol with the exception that 50 mM EDTA was added to the lysis buffer to eliminate the possibility of ribozyme cleavage during sample preparation. Antisense RNA probes targeting portions of Flt-1 and KDR were prepared by transcription in the presence of [32P]UTP. Samples were analyzed on polyacrylamide gels and the level of protected RNA fragments was quantified using a Molecular Dynamics PhosphorImager. The levels of Flt-1 and KDR were normalized to the level of cyclophilin (human cyclophilin probe template, Ambion) in each sample. The coefficient of variation for cyclophilin levels was 11% [265 940 c.p.m. 29 386 (SD)] for all conditions tested here (i.e. in the presence of either active ribozymes or attenuated controls). Thus, cyclophilin is useful as an internal standard in these studies.
Rat corneal pocket assay of VEGF-induced angiogenesis
Animal guidelines and anesthesia. Animal housing and experimentation adhered to standards outlined in the 1996 Guide for the Care and Use of Laboratory Animals (National Research Council). Male Sprague Dawley rats (250-300 g) were anesthetized with ketamine (50 mg/kg), xylazine (10 mg/kg) and acepromazine (0.5 mg/kg) administered intramuscularly (im). The level of anesthesia was monitored every 2-3 min by applying hind limb paw pressure and examining for limb withdrawal. Atropine (0.4 mg/kg, im) was also administered to prevent potential corneal reflex-induced bradycardia.
Preparation of VEGF soaked disk. For corneal implantation, 0.57 mm diameter nitrocellulose disks, prepared from 0.45 µm pore diameter nitrocellulose filter membranes (Millipore Corporation), were soaked for 30 min in 1 µl of 30 µM VEGF165 in 82 mM Tris-HCl (pH 6.9) in covered Petri dishes on ice.
Corneal surgery. The rat corneal model used in this study was modified from Kochet al. (26) and Pandey et al. (27). Briefly, corneas were irrigated with 0.5% povidone iodine solution followed by normal saline and two drops of 2% lidocaine. Under a dissecting microscope (Leica MZ-6), a stromal pocket was created and a presoaked filter disk (see above) was inserted into the pocket such that its edge was 1 mm from the corneal limbus.
Intraconjunctival injection of test solutions. Immediately after disk insertion, the tip of a 40-50 µm OD injector (constructed in our laboratory) was inserted within the conjunctival tissue 1 mm away from the edge of the corneal limbus that was directly adjacent to the VEGF-soaked filter disk. 600 nl of test solution (ribozyme, attenuated control or sterile water vehicle) were dispensed at a rate of 1.2 µl/min using a syringe pump (Kd Scientific). The injector was then removed, serially rinsed in 70% ethanol and sterile water and immersed in sterile water between each injection. Once the test solution was injected, closure of the eyelid was maintained using microaneurism clips until the animal began to recover gross motor activity. Following treatment, animals were warmed on a heating pad at 37°C.
Animal treatment groups/experimental protocol. Ribozymes targeting Flt-1 site 4229 and KDR mRNA site 726 were tested in the corneal model along with their attenuated controls. Five treatment groups were assigned to examine the effects of five doses of each test substance over a dose range of 1-100 µg on VEGF-stimulated angiogenesis. Negative (30 µM VEGF soaked filter disk and intraconjunctival injection of 600 nl sterile water) and no stimulus (Tris-soaked filter disk and intraconjunctival injection of sterile water) control groups were also included. Each group consisted of five animals (10 eyes) receiving the same treatment.
Quantitation of angiogenic response. Five days after disk implantation, animals were euthanized following im administration of 0.4 mg/kg atropine and corneas were digitally imaged. The neovascular surface area (NSA, expressed in pixels) was measured post-mortem from blood-filled corneal vessels using computerized morphometry (Image Pro Plus, Media Cybernetics, v.2.0). The individual mean NSA was determined in triplicate from three regions of identical size in the area of maximal neovascularization between the filter disk and the limbus. The number of pixels corresponding to the blood-filled corneal vessels in these regions was summated to produce an index of NSA. A group mean NSA was then calculated. Data from each treatment group were normalized to VEGF/ribozyme vehicle-treated control NSA and finally expressed as percent inhibition of VEGF-induced angiogenesis. Statistics. After determining the normality of treatment group means, group mean percent inhibition of VEGF-induced angiogenesis was subjected to a one-way analysis of variance. This was followed by two post-hoc tests for significance including Dunnett's (comparison to VEGF control) and Tukey-Kramer (all other group mean comparisons) at alpha = 0.05. Statistical analyses were performed using JMP v.3.1.6 (SAS Institute).
RESULTS
Ribozyme-mediated reduction of VEGF-induced cell proliferation
Ribozyme cleavage of Flt-1 or KDR mRNA should result in a decrease in the density of cell surface VEGF receptors. This decrease should limit VEGF binding and consequently interfere with the mitogenic signaling induced by VEGF. To determine if cell proliferation was impacted by anti-Flt-1and/or anti-KDR ribozyme treatment, proliferation assays using cultured human microvascular cells were carried out. Ribozymes included in the proliferation assays were initially chosen by their ability to decrease the level of VEGF binding to treated cells (data not shown). In these initial studies, ribozymes targeting 20 sites in the coding region of each mRNA were screened. The most effective ribozymes against two sites in each target (Table 1), Flt-1 sites 1358 and 4229 and KDR sites 726 and 3950, were included in the proliferation assays reported here (Fig. 1). In addition, attenuated analogs of each ribozyme were used as controls (Table 1). These attenuated controls are still capable of binding to the mRNA target since the binding arm sequence is maintained. However, these controls have two nucleotide changes in the core sequence that substantially reduce their ability to carry out the cleavage reaction.
Figure 1. Ribozyme-mediated inhibition of cell proliferation. Cultured HMVEC-d were treated with ribozyme or attenuated controls as LipofectAMINETM complexes. After treatment, cells were stimulated with VEGF165or bFGF and allowed to grow for 48 h prior to determining the cell number. Each ribozyme was tested in triplicate at three concentrations and data are presented as mean cell number per well + SD. The data obtained following ribozyme treatment and VEGF stimulation are presented in (A) and (B) for anti-Flt-1 ribozymes and (D) and (E) for anti-KDR ribozymes. Representative data obtained following ribozyme treatment and bFGF stimulation are shown in (C) for one anti-Flt-1 ribozyme and in (F) for one anti-KDR ribozyme. In all panels, active ribozymes are represented with filled symbols; attenuated controls with open symbols. In addition to the ribozymes and attenuated controls listed in Table 1, a second set having the same sequences but with an additional basepair in the `stem II' region of the ribozyme are also shown for VEGF-induced proliferation studies. These 4 bp stem II ribozymes and attenuated controls have one additional base pair such that the stem II/loop sequence is ggccgaaaggcc. Therefore, ribozymes and controls with 3 or 4 bp stem IIs are denoted with circles and squares, respectively. The data for one irrelevant ribozyme [filled triangle, (B)] are also shown. This irrelevant ribozyme contains an active core sequence but has no binding site in either Flt-1 or KDR mRNA. Its sequence is 5[prime]-gsasasgsgaacUGAuGaggccgaaaggccGaaAgauggcT-3[prime] with modifications as in Table 1 except that T indicates a 3[prime]-3[prime] inverted deoxythymidine. For reference, the average number of cells in control wells after 48 h in the absence of VEGF or bFGF for each of the panels are as follows: (A), (B) and (C), 12 477 617; (D), (E) and (F), 17 182 1053.
The anti-proliferative effect of active ribozymes targeting two lead sites on each VEGF receptor mRNA is shown in Figure 1. The active ribozymes tested decreased the relative proliferation of HMVEC-d after VEGF stimulation, an effect that increased with ribozyme concentration. This concentration dependency was not observed following treatment with the attenuated controls designed for these sites. In fact, little or no change in cell growth was noted following treatment with the attenuated controls, even though these controls can still bind to the specific target sequences. At 200 nM, there was a distinct `window' between the anti-proliferative effects of each ribozyme and its attenuated control; a trend also observed at lower doses. This window of inhibition of proliferation (56-77% based on total cells per well) reflects the contribution of ribozyme-mediated activity. In comparison, no effect of anti-Flt-1 or anti-KDR ribozymes was noted on bFGF-stimulated cell proliferation (Fig. 1C and F). Moreover, an irrelevant, but active, ribozyme whose binding sequence is not found in either Flt-1 or KDR mRNA had no effect in this assay (Fig. 1B). These data are consistent with the basic ribozyme mechanism in which binding and cleavage are necessary components. Although the relative surface distribution of Flt-1 and KDR receptors in this cell type is not known, the anti-proliferative effects of these ribozymes indicate that, at least in cell culture, both receptors are functionally coupled to proliferation.
Specific reduction of Flt-1 or KDR mRNA by ribozyme treatment
To confirm that anti-Flt-1 and anti-KDR ribozymes reduce their respective mRNA targets, cellular levels of Flt-1 or KDR were quantified using an RNase protection assay with specific Flt-1 or KDR probes. For each target, one ribozyme/attenuated control pair was chosen for continued study. Data from a representative experiment are shown in Figure 2. Exposure of HMVEC-d to active ribozyme targeting Flt-1 site 4229 decreased Flt-1 mRNA, but not KDR mRNA. Likewise, treatment with the active ribozyme targeting KDR site 726 decreased KDR, but not Flt-1 mRNA. Both ribozymes decreased the level of their respective target RNA by >50%. The degree of reduction associated with the corresponding attenuated controls was not >13%.
Figure 2. Target specificity of anti-Flt-1 and KDR ribozymes. Cultured HMVEC-d were treated with LipofectAMINETM complexes containing 200 nM active ribozyme (A) or attenuated control (C) and analyzed by RNase protection following 24 h of VEGF-stimulated growth. Data obtained for ribozymes and attenuated controls that target Flt-1 site 4229 or KDR site 726 are shown. Data were normalized to the level of an internal mRNA control (cyclophilin) and are presented as percent decrease in Flt-1 (left) or KDR mRNA (right) relative to an untreated control. Error bars indicate the range of duplicate samples.
In vitro activity of anti-Flt and anti-KDR ribozymes
To confirm further the necessity of an active ribozyme core, in vitro cleavage activities were determined for the Flt-1 site 4229 ribozyme and the KDR site 726 ribozyme as well as their paired attenuated controls (Fig. 3). The first order rate constants calculated from the timecourse of short substrate cleavage for the anti-Flt-1 ribozyme and its attenuated control were 0.081 0.0007 min-1 and 0.001 6 × 10-5min-1, respectively. For the anti-KDR ribozyme and its paired control, the first order rate constants were 0.434 0.024 min-1 and 0.002 1 × 10-4min-1, respectively. Although the attenuated controls retain a very slight level of cleavage activity under these optimized conditions, the decrease in in vitro cleavage activity between each active ribozyme and its paired attenuated control is about two orders of magnitude. Thus, an active core is essential for cleavage activity in vitro and is also necessary for ribozyme activity in cell culture.
Figure 3. In vitro cleavage analysis. A timecourse of in vitro cleavage on a complementary RNA substrate was carried out for the ribozymes and paired attenuated controls targeting (A) Flt-1 site 4229 and (B) KDR site 726. The upper and lower bands correspond to full-length and cleaved substrate, respectively. A trace amount of 5[prime]-32P-end-labeled substrate was included for detection purposes. The first-order rate constants (k1) were determined using a double exponential curve fit and are given at the bottom of each gel panel.
Ribozyme-mediated reduction of VEGF-induced angiogenesis in vivo
To assess whether ribozymes targeting VEGF receptor mRNA could impact the complex process of angiogenesis, prototypic anti-Flt-1 and KDR ribozymes that were identified in cell culture studies were screened in a rat corneal pocket assay of VEGF-induced angiogenesis. In this assay, corneas implanted with VEGF-containing filter disks exhibited a robust neovascular response in the corneal region between the disk and the corneal limbus (from which the new vessels emerge; Fig. 4B). Disks containing a vehicle solution elicited no angiogenic response (Fig. 4A). In separate studies (data not shown), intraconjunctival injections of sterile water vehicle did not affect the magnitude of the VEGF-induced angiogenic response. In addition, ribozyme injections alone did not induce angiogenesis.
Figure 4. Representative digital corneal images detailing the effects of intraconjunctival injection of ribozymes on VEGF-induced angiogenesis. Key features are indicated on (A). In panels (B), (C) and (D) the disk was soaked in VEGF whereas in (A) the disk was soaked in vehicle only. (A) and (B) illustrate the angiogenic range of the assay between vehicle and VEGF-treated corneas in the absence of ribozyme treatment. As shown in (A), the ribozyme vehicle does not induce any angiogenic response. (C) demonstrates the impact of an active ribozyme targeting Flt-1site 4229 on VEGF-induced angiogenesis. The angiogenic response following treatment with an attenuated control targeting the same site is given in (D). Ribozymes injected into the conjunctiva elicited no angiogenic response in this model (data not shown).
Representative images of corneas treated with either 10 µg of an anti-Flt-1 ribozyme (site 4229) or its attenuated control are shown in Figure 4C and D. Figure 4C illustrates the typical ribozyme-mediated reduction in VEGF-induced neovascularization of the cornea compared with Figure 4B. The attenuated control (Fig. 4D) had little impact on neovascularization. Changes in the observed angiogenic response could be quantified reproducibly.
The dose-related effects of anti-Flt-1 or KDR ribozymes on the VEGF-induced angiogenic response were then examined. Figure 5 illustrates the quantified anti-angiogenic effect of the anti-Flt-1 (site 4229) and KDR (site 726) ribozymes and their attenuated controls over a dose range from 1 to 100 µg. For both ribozymes, the maximal anti-angiogenic response (48 and 36% for anti-Flt-1and KDRribozymes, respectively) was observed at a dose of 10 µg.
A
 |
B
 |
Figure 5. Anti-angiogenic efficacy of ribozymes in the rat corneal model of VEGF-induced angiogenesis. The percent inhibition of VEGF-induced angiogenesis for locally administered (A) anti-Flt-1 (site 4229) and (B) anti-KDR(site 726)ribozymes (filled circles) and their attenuated controls (open circles) are plotted over the dose range tested. Pixels associated with background structures including the iris (Fig. 3A) were subtracted from all treatment groups. Data are expressed as mean percent reduction in VEGF-induced angiogenesis SEM. *p < 0.05 relative to VEGF/vehicle treated controls by Dunnett's, **p < 0.05 relative to attenuated dose-matched controls by Tukey-Kramer.
The anti-Flt-1ribozyme produced a significantly greater anti-angiogenic response than its attenuated control at 3 and 10 µg (p < 0.05; Fig. 5A). Its attenuated control exhibited a small but significant anti-angiogenic response at doses above 10 µg compared to vehicle treated VEGF controls (p < 0.05; Fig. 5A). At its maximum, this response was not significantly greater than that observed with the lowest dose of active anti-Flt-1 ribozyme. The anti-KDR ribozyme significantly inhibited angiogenesis from 3 to 30 µg (p < 0.05; Fig. 5B). The anti-KDR attenuated control had no significant effect at any dose tested. Interestingly, anti-Flt-1orKDRribozyme doses >10 µg failed to yield a greater anti-angiogenic effect (Fig. 5). It is possible that ribozyme bioavailability was limited at the higher doses because ribozyme injectates at 30 and 100 µg were considerably more viscous than lower doses.
DISCUSSION
We have shown that ribozymes targeting each VEGF receptor subtype message specifically reduce Flt-1 or KDR mRNA levels, decrease proliferation of vascular endothelial cells in culture and inhibit VEGF-stimulated angiogenesis. As a result, we believe that anti-VEGF receptor ribozymes present a novel therapeutic strategy for the control of VEGF-mediated angiogenesis. This study also illustrates the potential use of ribozymes to probe the biological contribution of each VEGF receptor subtype in the process of neovascularization.
The involvement of Flt-1 and KDR receptors in the formation of blood vessels is supported by several lines of evidence. In addition to their localized expression in adult mouse endothelial cells (21), the expression of VEGF receptors has been demonstrated in vascular structures of developing mouse embryos (28) and studies indicate that both receptors are required for vasculogenesis (29,30). In fact, mutations in either Flt-1 or KDR are lethal to the embryo and result in vascular abnormalities (29,30). Inhibition of Flk-1 (the mouse KDR homolog) has also been shown to be anti-angiogenic in a chorioallantoic membrane assay (31). Finally, interfering with the expression of VEGF (32) or the binding of VEGF to its receptors (15,33,34), results in decreased new vessel formation. Although these studies do little to describe the VEGF receptor-specific mechanisms involved in pathologic angiogenesis, they demonstrate that VEGF receptors are important targets for the development of anti-angiogenic agents.
Anti-Flt and KDR ribozymes in cell culture
Numerous variables can contribute to the activity of different ribozymes in the cell. These variables include, but are not limited to, the accessibility of a ribozyme to its mRNA binding site and/or differences in cleavage activities in the intracellular milieu. Thus, the most catalytically active ribozyme in biochemical assays may not be the most effective ribozyme in cell culture or in vivo. A systematic screening strategy was used to identify effective ribozymes in cultured HMVEC-d that could then be tested in anangiogenesis model. In the cell culture assays, VEGF binding and VEGF-induced cell proliferation were used as endpoints. In initial screens, several ribozymes targeting either Flt-1 or KDRmRNA decreased VEGF binding by as much as 50-60% (data not shown). Subsequent proliferation assays with the most effective ribozymes targeting either Flt-1 or KDR resulted in a 56-77% decrease in VEGF-induced cell proliferation. Because both receptors are present on HMVEC-d (35), a ribozyme-mediated decrease of only one of the receptors is unlikely to lead to a complete inhibition of binding or proliferation. However, preliminary experiments involving combinations of anti-Flt-1and KDRribozymes showed that the reduction in VEGF-induced proliferation of HMVEC-d was potentiated when both receptor subtypes were targeted. Although KDR has been implicated as the primary mediator of VEGF-induced endothelial cell proliferation (7,36), our results reinforce the possibility that the Flt-1 receptor is also involved.
The specificity and mechanism of the anti-Flt and KDR ribozymes were addressed in several ways. In cell culture assays, anti-Flt-1 and KDRribozymes specifically reduced their targeted mRNA without impacting that of the other receptor. While ribozymes targeting Flt-1 or KDR mRNA effectively reduced VEGF-induced cell proliferation, they had no effect on bFGF stimulated cell proliferation. In addition, an irrelevant but active ribozyme lacking appropriate binding arm sequences had no effect on HMVEC-d proliferation. Anti-Fltor KDRribozymes also had no effect on the binding of two other ligands, bFGF or urokinase-like plasminogen activator, to their respective receptors (data not shown). Since attenuated controls had little or no impact on any endpoint tested, it can be concluded that the effects of the active ribozymes are due to a cleavage mechanism. Taken together, these results indicate that anti-Flt-1and KDR ribozymes reduce the level of their respective molecular targets and impact VEGF-mediated cell signaling through a sequence-specific cleavage mechanism.
Anti-Flt and KDR ribozymes in vivo
Corneal pocket assays are widely used to study the complex biology of angiogenesis and to identify potential therapeutic agents which impinge upon this process (26,37-43). These assays are reproducible, adaptable to specific pharmacologic stimulation (i.e. VEGF), exhibit low background and are amenable to quantitation. In the present study, we used a rat corneal pocket assay to examine whether anti-Flt-1and KDR ribozymes (targeting conserved sites between human, rat and mouse) impact VEGF-induced angiogenesis.
In general, ribozymes that target either VEGF receptor subtype mRNA significantly reduced angiogenesis in the rat cornea in a dose-dependent manner. Attenuated controls produced only a low level of anti-angiogenic activity at the highest doses tested (30 or 100 µg). This low level of activity was never significantly greater than that observed with the lowest dose of active ribozyme, and could be due to the slight amount of cleavage activity retained by these controls. Therefore, the greater anti-angiogenic effect of the active anti-Flt-1 and KDR ribozymes versusthe attenuated controls strongly supports a ribozyme-specific mechanism.
The possibility exists that a greater anti-angiogenic activity may be achieved by combining ribozymes targeting both receptor mRNAs. Although initial studies in cultured HMVEC-d suggest a potentiated response with combined anti-Flt-1and KDRribozyme treatment, preliminary anti-Flt-1/KDR ribozyme combination studies in the rat cornea did not enhance the anti-angiogenic activity over either ribozyme alone (data not shown). We have not yet explored ribozyme dosing combinations to assess their enhancement of anti-angiogenic efficacy.
Anti-Flt and KDR ribozymes and VEGF receptor function
Many factors complicate our understanding of the functional contributions of the two VEGF receptor subtypes to neovascularization, particularly when considering their role in disease. For instance, there is some evidence that KDR but not Flt-1 stimulation mediates VEGF-induced endothelial cell proliferation as well as enhanced motility, cell shape change and actin reorganization (7,36). Our results indicate that down-regulation of KDR or Flt-1 receptor mRNA leads to decreased endothelial cell proliferation and decreased VEGF-induced angiogenesis. It is possible that the Flt-1 receptor can either directly mediate these processes or that there may be interplay between receptors (7,44,45).
Mutations in either receptor lead to abnormalities in neovascularization through presumably different pathways (29,30). For KDR, vasculogenesis may be mediated directly by changes in endothelial cell development (30) while Flt-1 may be important in cellular interactions with the extracellular matrix (29). It appears that this effect of Flt-1 on embryologic neovascularization is mediated by the binding domain of Flt-1 rather than the kinase domain (46). In addition, there are differences in the relative abilities of isoforms of VEGF and placenta growth factor (and dimers thereof) to induce homo- and heterodimerization of KDR and Flt-1 receptors (7,44,45). Further complicating this picture is the existence of a native soluble form of the Flt-1 receptor that may serve to downregulate VEGF-mediated angiogenesis (47,48)by sequestering VEGF and/or forming inactive heterodimers with membrane-spanning VEGF receptors. Thus, changes in post-transcriptional receptor processing (47,48) as well as the effect of disease or hypoxia on receptor expression (49,50) may also impact the relative role of VEGF receptor subtypes in angiogenesis.
By virtue of their high degree of selectivity, ribozymes targeting VEGF receptor mRNA can be used to determine the relative role of each receptor subtypes in signal transduction and neovascularization. The current study demonstrates that these ribozymes are effective in both cell culture and in vivo conditions. Thus, it is possible to use anti-Flt-1 and KDR ribozymes to address receptor-specific concerns in isolated cells (vascular endothelial, pericytes, macrophages, etc.) and tissue explants as well as in animal systems.
Anti-Flt-1 and KDR ribozymes as therapeutics
In addition to their utility in differentiating the role of VEGF receptors, ribozymes that reduce VEGF receptor expression have broad therapeutic potential, especially where VEGF may mediate angiogenesis in solid tumors (22,33,34,51-55) and neovascular diseases of the eye (9,10,14-16,34,56). Several studies have shown that reducing VEGF signaling and subsequent angiogenesis through a variety of mechanisms can have therapeutic effects. Treatment with neutralizing VEGF antibodies reduces tumor growth (33,34), metastasis (34) and tumor neovascularization (33,49) in a variety of human tumor xenograft models. In a VEGF-dependent model of retinal neovascularization (56,57), soluble VEGF-receptor chimeric proteins (15) or antisense oligonucleotides targeting VEGF message (32) inhibited retinal neovascularization in response to hypoxia. Also, retinal vein occlusion-induced neovascularization of the non-human primate iris was reduced following treatment with an anti-VEGF monoclonal antibody (16).
Summary
In conclusion, we have identified ribozymes directed against Flt-1and KDR mRNA that specifically reduce VEGF binding and endothelial cell proliferation in cell culture as well as VEGF-induced angiogenesis in vivo. The therapeutic potential of these ribozymes is currently being evaluated in a variety of disease models involving VEGF-mediated angiogenesis. Results from these studies will not only support the development of a potential therapeutic agent, but will also help clarify the relative importance and functional contribution of each VEGF receptor subtype in pathophysiological angiogenesis.
ACKNOWLEDGEMENTS
We wish to thank Dr Denis Gospodarowicz (Chiron Corp.) for supplying VEGF165 and bFGF. Dr Lori Stephenson (Chiron Corp.) provided the templates for RNase protection probes.
REFERENCES
*To whom correspondence should be addressed. Tel: +1 303 449 6500; Fax: +1 303 449 6995; Email: pavco{at}rpi.com
Present address: Cynthia Cushman, Elan Corporation, 800 Gateway Boulevard, South San Francisco, CA 94080, USA
This page is run by Oxford University Press, Great Clarendon Street, Oxford OX2 6DP, as part of the OUP Journals
Comments and feedback: jnl.info{at}oup.co.uk
Last modification:
Copyright© Oxford University Press, 1999.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  J Heidemann, D G Binion, W Domschke, and T Kucharzik Antiangiogenic therapy in human gastrointestinal malignancies. Gut, October 1, 2006; 55(10): 1497 - 1511. [Full Text] [PDF]
|
|
|
|
 |
|
 D.-G. Bae, T.-D. Kim, G. Li, W.-H. Yoon, and C.-B. Chae Anti-Flt1 Peptide, a Vascular Endothelial Growth Factor Receptor 1-Specific Hexapeptide, Inhibits Tumor Growth and Metastasis Clin. Cancer Res., April 1, 2005; 11(7): 2651 - 2661. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Zhang, C. R. Dass, E. Sumithran, N. Di Girolamo, L.-Q. Sun, and L. M. Khachigian Effect of Deoxyribozymes Targeting c-Jun on Solid Tumor Growth and Angiogenesis in Rodents J Natl Cancer Inst, May 5, 2004; 96(9): 683 - 696. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Ruggeri, J. Singh, D. Gingrich, T. Angeles, M. Albom, H. Chang, C. Robinson, K. Hunter, P. Dobrzanski, S. Jones-Bolin, et al. CEP-7055: A Novel, Orally Active Pan Inhibitor of Vascular Endothelial Growth Factor Receptor Tyrosine Kinases with Potent Antiangiogenic Activity and Antitumor Efficacy in Preclinical Models Cancer Res., September 15, 2003; 63(18): 5978 - 5991. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Stopeck, M. Sheldon, M. Vahedian, G. Cropp, R. Gosalia, and A. Hannah Results of a Phase I Dose-escalating Study of the Antiangiogenic Agent, SU5416, in Patients with Advanced Malignancies Clin. Cancer Res., September 1, 2002; 8(9): 2798 - 2805. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Yang, X. Xin, C. Zlot, G. Ingle, G. Fuh, B. Li, B. Moffat, A. M. de Vos, and M. E. Gerritsen Vascular Endothelial Cell Growth Factor-Driven Endothelial Tube Formation Is Mediated by Vascular Endothelial Cell Growth Factor Receptor-2, a Kinase Insert Domain-Containing Receptor Arterioscler. Thromb. Vasc. Biol., December 1, 2001; 21(12): 1934 - 1940. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Bramlage, E. Luzi, and F. Eckstein HIV-1 LTR as a target for synthetic ribozyme-mediated inhibition of gene expression: site selection and inhibition in cell culture Nucleic Acids Res., November 1, 2000; 28(21): 4059 - 4067. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Amarzguioui, G. Brede, E. Babaie, M. Grotli, B. Sproat, and H. Prydz Secondary structure prediction and in vitro accessibility of mRNA as tools in the selection of target sites for ribozymes Nucleic Acids Res., November 1, 2000; 28(21): 4113 - 4124. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. C. Jarvis, K. S. Bouhana, M. E. Lesch, S. A. Brown, T. J. Parry, D. J. Schrier, S. W. Hunt III, P. A. Pavco, and C. M. Flory Ribozymes as Tools for Therapeutic Target Validation in Arthritis J. Immunol., July 1, 2000; 165(1): 493 - 498. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. A. Pavco, K. S. Bouhana, A. M. Gallegos, A. Agrawal, K. S. Blanchard, S. L. Grimm, K. L. Jensen, L. E. Andrews, F. E. Wincott, P. A. Pitot, et al. Antitumor and Antimetastatic Activity of Ribozymes Targeting the Messenger RNA of Vascular Endothelial Growth Factor Receptors Clin. Cancer Res., May 1, 2000; 6(5): 2094 - 2103. [Abstract] [Full Text]
|
|
|
|
 |
|
 G. McMahon VEGF Receptor Signaling in Tumor Angiogenesis Oncologist, April 1, 2000; 5(90001): 3 - 10. [Abstract] [Full Text]
|
|
|
|
 |
|
 B. Li, G. Fuh, G. Meng, X. Xin, M. E. Gerritsen, B. Cunningham, and A. M. de Vos Receptor-selective Variants of Human Vascular Endothelial Growth Factor. GENERATION AND CHARACTERIZATION J. Biol. Chem., September 15, 2000; 275(38): 29823 - 29828. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||