July 2010
Volume 51, Issue 7
Free
Retina  |   July 2010
Evaluation of CXCR4 Inhibition in the Prevention and Intervention Model of Laser-Induced Choroidal Neovascularization
Author Affiliations & Notes
  • Edwin Lee
    From the Department of Ocular Biology, Pfizer, Inc., San Diego, California.
  • David Rewolinski
    From the Department of Ocular Biology, Pfizer, Inc., San Diego, California.
  • Address correspondence to: Michael Shiue, PGRD-La Jolla, Pfizer, Inc., 10646 Science Center Drive, San Diego, CA 92121; michael.shiue@pfizer.com
Investigative Ophthalmology & Visual Science July 2010, Vol.51, 3666-3672. doi:10.1167/iovs.09-3802
  • Views
  • PDF
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Edwin Lee, David Rewolinski; Evaluation of CXCR4 Inhibition in the Prevention and Intervention Model of Laser-Induced Choroidal Neovascularization. Invest. Ophthalmol. Vis. Sci. 2010;51(7):3666-3672. doi: 10.1167/iovs.09-3802.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

Purpose.: Endothelial precursor cells (EPCs) derived from hematopoietic stem cells (HSCs) have been shown to contribute to choroidal neovascularization by signaling through the SDF-1/CXCR4 axis. In a prevention and treatment/intervention modality of the laser choroidal neovascularization (CNV) model, the efficacy of CXCR4 inhibition on reducing choroidal leakage and angiogenesis was evaluated.

Methods.: CNV in rats was generated by focal rupture of Bruch's membrane with an 810-nm diode laser. In the prevention mode, a CXCR4 antagonist (AMD3100) was delivered via an osmotic pump 1 day after laser induction. In the intervention mode, AMD3100 delivery commenced 14 days after laser induction. Inhibition of CXCR4 was determined through leukocyte and SDF-1 actin polymerization blood biomarker assays. Leakage was assessed by fluorescein angiography, and CNV lesion size was quantified after isolectin B4 endothelial cell staining. SU14813, an anti-VEGFR, PDGFR-β, KIT, and FLT3 inhibitor, was also assessed in an intervention study protocol.

Results.: Inhibition of CXCR4 was demonstrated by an increase in the number of blood leukocytes, and diminished SDF-1 induced actin polymerization in whole blood. CNV leakage and neovascularization were inhibited when the dose regimen was initiated 1 day after laser-induced CNV induction. AMD3100 did not show efficacy when administered 14 days after lasering. Treatment with SU14813 significantly decreased CNV leakage and lesion size in an intervention modality.

Conclusions.: Inhibition of CXCR4 may be useful in preventing neovascularization but does not appear to have an effect on already established angiogenesis. A multiple receptor tyrosine kinase (RTK) inhibitor approach shows promise for the treatment of wet age-related macular degeneration.

Wet age-related macular degeneration (AMD) affects approximately 1.6 million adults over 50 years of age within the United States and is expected to affect 3 million by 2020. 1 Globally, 25 to 30 million are affected by AMD, and it is the leading cause of blindness in developing countries. Wet AMD is characterized by aberrant choroidal neovascularization and leaky angiogenic vessels in the eye. Accumulation of fluid and blood, in addition to vascular scarring, ultimately reduces central vision by triggering photoreceptor cell death. There are currently two approved therapies for the treatment of wet AMD that focus on the inhibition of VEGF signaling: pegaptanib sodium (Macugen; Eyetech/OSI Pharmaceuticals, New York, NY) and ranibizumab (Lucentis; Genentech, South San Francisco, CA). 
Recent findings have implicated hematopoietic stem cells (HSCs) in the formation of new pathologic vessels observed in wet AMD. In mice with GFP+ bone marrow transplantation, HSCs derived from the bone marrow were shown to be incorporated into sites of retinal and choroidal neovascularization. 24 Recruitment of endothelial precursor cells (EPCs) to the site of neovascularization is mediated, in part, by the chemokine SDF-1, and its receptor, CXCR4. CXCR4 is a G-protein-coupled receptor found on lymphocytes, monocytes, hematopoietic, endothelial progenitor cells, and mature endothelial cells. 58 Sequestration of HSCs in bone marrow is also mediated through the SDF-1/CXCR4 axis in which the HSC receptor is agonized by SDF-1 released from bone marrow stromal cells. 9,10 SDF-1 also helps regulate angiogenesis by causing EPCs to home in on the vessel lumens. 1114  
In many of the studies showing HSC/EPC incorporation, the rodent laser-induced choroidal neovascularization (CNV) model has been used, in which laser rupture of Bruch's membrane leads to choroidal angiogenesis and leakage. The pathology of this model mimics the wet AMD phenotype with its characteristic subretinal fluid accumulation, vascular scarring, macrophage and leukocyte recruitment, and penetration of Bruch's membrane by choroidal capillaries. The rodent CNV model of wet AMD also has been used to test the efficacy of various small molecule and biological therapeutics. 1525 Almost all the studies reported in the literature have evaluated potential therapies in a prevention modality, in which the drug is first administered around the time of lesion induction, and often only lesion size was measured. In the present study, we administered AMD3100, a bicyclam CXCR4 functional antagonist, 26 to inhibit CXCR4 signaling, and tested the efficacy of leakage and angiogenesis in both the prevention and intervention modality of the laser-induced CNV model. Biomarker studies were performed to demonstrate in vivo/ex vivo CXCR4 target modulation. In addition, a receptor tyrosine kinase inhibitor was evaluated in an intervention model of laser-induced CNV. 
Materials and Methods
Animal Use and Welfare
Female Brown Norway Rats (200–250 g) were obtained from Charles River Laboratories. Animal experiments were performed in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research and Pfizer Institutional Animal Care and Use Committee guidelines in the Laboratory Animal Research Center of Pfizer Global Research and Development (La Jolla, CA). 
Actin Polymerization and Leukocytosis Assay
Whole blood (WB) samples collected from the animals on day 12 were incubated with WB assay buffer (nonphenol RPMI, 10 mM HEPES, and 0.5% FBS) and stimulated with 30 nM of SDF-1 (R&D Systems, Minneapolis, MN). Flow cytometry lysis buffer (FACS; BD Biosciences, Franklin Lakes, NJ), 1.6% EM grade formaldehyde, and water were added to fix and lyse the cells. The cells were centrifuged and washed two times with PBS, 2% FBS, and 0.2% sodium azide before addition of permeabilization/stain solution (1× HBSS, lysophosphatidyl choline, 4% formaldehyde, and FITC phalloidin (Phallicidin; Sigma-Aldrich, St. Louis, MO). The cells were then run on a flow cytometer (BD Biosciences). The lymphocyte population was gated, and median fluorescence was measured. For the leukocytosis assay, blood was added to a diluent (Isoton Diluent II; BD Biosciences), mixed, and lysed with 2 drops of lytic reagent (Zapoglobin II; BD Biosciences). WBC counts were measured with a cell counter (Coulter, Hialeah, FL). The mean ± SEM was calculated from results in three to four animals per group in both assays. 
Laser-Induced CNV
Animals were anesthetized with an IP injection of 50 mg/kg ketamine HCl mixed with 5 mg/kg xylazine. The eyes were dilated with 1 drop each of 2.5% phenylephrine HCl and 1% cyclopentolate HCl (Wilson, Mustang, OK). Lesions were induced with an 810-nm red diode (Oculight SLx; Iridex, Mountain View, CA) laser mounted on a slit lamp (SL 130; Carl Zeiss Meditec, Inc., Thornwood, NY). In each animal, five to six laser lesions were induced between major retinal vessels at a distance from the optic nerve head of approximately two times the diameter of the optic disc. The laser parameters were: wavelength, 810 nm; spot size, 75-μm diameter; power, 150 to 200 mW; and duration, 100 ms. Cavitation bubbles were observed, indicating the rupture of Bruch's membrane. 
In Vivo Efficacy Studies
In the prevention mode studies, animals were lasered on day 0, and on day 1, a 14-day continuous release osmotic pump (Alzet, Cupertino, CA) was implanted that contained vehicle or AMD3100 (30 mg/kg/d). Fourteen days after lasering, fluorescein angiography was conducted for leakage assessment, and the eyes were stained with isolectin B4 on day 15. For AMD3100 interventional studies, the eyes were lasered on day 0, and baseline fluorescein angiography images were taken on day 13. An osmotic pump containing vehicle or AMD3100 was implanted on day 14. Endpoint fluorescein angiography was conducted 28 days after lasering, and the eyes were stained with isolectin B4 on day 29. 
Animals were prepared for osmotic pump implantation by clipping of dorsal hair followed with a betadine scrub. While the animal was under anesthesia, bupivacaine was delivered subcutaneously to the upper dorsal region. An incision was made in the skin at the surgical prep site with surgical Mayo scissors. The osmotic pump was placed subcutaneously, and the incision was stapled shut with stainless-steel sutures. 
For SU14813 intervention mode studies, the animals were lasered on day 1, and baseline fluorescein angiography images were taken on day 13. On day 14, SU14813 was delivered via 10-mg/kg IP injections, twice daily for 8 days. Endpoint fluorescein angiography was conducted 21 days after lasering, and the eyes were stained with isolectin B4 on day 22. 
Fluorescein Angiography and Leakage Grading
Fluorescein (100 mg/kg body weight; AK-Fluor 25%; Akorn, Lake Forest, IL) was administered in intraperitoneal injections and late-stage fluorescein angiography images were taken at 7 to 9 minutes after injection. Fluorescence scoring was as follows: 3, lesions with fluorescence intensity greater than the neighboring major artery or vein; 2, lesions with fluorescence intensity equal to the neighboring major artery or vein; 1, lesions with fluorescence intensity less than the neighboring major artery or vein; and 0, lesions with no apparent leakage. 
In the intervention mode, only lesions with scores of 2 or 3 that were observed in the 14-day baseline fluorescein angiography images were measured at the endpoint and used in calculating mean leakage. Leakage scores were conducted by a blinded grader. The mean ± SEM was calculated from results in six to nine animals per group. 
Isolectin B4 Staining and Lesion Size Measurements
Enucleated eyes were fixed with 4% paraformaldehyde and washed three times with PBS. Cornea and lens were removed and eye cups were dehydrated and rehydrated in a graded series of alcohol washes. Eye cups were blocked with PBS containing 1% BSA and 0.5% Triton X-100 and stained overnight with FITC-isolectin B4 (VWR, West Chester, PA) in PBS containing 0.2% BSA and 0.1% Triton X-100. The retina was removed from the eye cups and the RPE/choroid/sclera complex was flatmounted in aqueous mounting medium. Specimens were excited at 488 nm, and images were taken with a digital camera (Microfire; Optronics, Goleta, CA) mounted on a fluorescence microscope (AX70; Olympus, Center Valley, PA). Isolectin B4-stained lesion areas were automatically measured (Image Pro 6.0; Media Cybernetics) after image size calibration with a micrometer was completed. The mean ± SEM was calculated from results in four to nine animals per group. 
Statistics
The unpaired t-test was used for statistical analysis, and P < 0.05 was considered significant. 
Results
Effect of Systemic Administration of AMD3100 on CXCR4 In Vivo
AMD3100 was administered to rats via an osmotic pump implanted subcutaneously to deliver 30 mg/kg body weight per day for 14 days. Inhibition of CXCR4 in whole blood was determined with an ex vivo flow cytometry–based actin polymerization assay. Lymphocytes from vehicle animals stimulated with exogenous SDF-1 were gated and shown to increase fluorescence intensity by ∼30% through actin polymerization over unstimulated control subjects (Fig. 1A). Lymphocytes from AMD3100-treated animals, however, showed a marked reduction in fluorescence increase after SDF-1 stimulation (Fig. 1B). AMD3100-treated animals on day 12 of the regimen showed ∼70% inhibition of SDF-1-induced actin polymerization, indicating that drug was present in the blood and CXCR4 was robustly inhibited (P < 0.05, Fig. 1C). CXCR4 inhibition was also reflected in the 1.6-fold elevation of WBC count in whole blood (P < 0.05, Fig. 1D). This leukocytosis effect has been observed in patients receiving AMD3100 in clinical trials. 27  
Figure 1.
 
In vivo and ex vivo CXCR4 target modulation of AMD3100-treated animals in the prevention laser-induced CNV mode. Blood from AMD3100-and vehicle-treated animals was collected 12 days after osmotic pump implantation and was used in actin polymerization and WBC count assays. (A) Flow cytometry showed that ex vivo SDF-1 induction increased actin polymerization fluorescence in the gated lymphocyte population from blood of vehicle-treated animals compared with non-SDF-1-treated vehicle blood. (B) Compared with vehicle, lymphocytes of AMD3100-treated animals had reduced FITC-phalloidin fluorescence after SDF-1 stimulation, as shown by the slight right shift of the graph compared with that of untreated blood. (C) AMD3100 significantly decreased SDF-1 induced actin polymerization fluorescence compared with vehicle control subjects (*P < 0.05, unpaired t-test). (D) WBC counts increased in the blood of AMD3100-treated animals compared with the control (*P < 0.05, unpaired t-test).
Figure 1.
 
In vivo and ex vivo CXCR4 target modulation of AMD3100-treated animals in the prevention laser-induced CNV mode. Blood from AMD3100-and vehicle-treated animals was collected 12 days after osmotic pump implantation and was used in actin polymerization and WBC count assays. (A) Flow cytometry showed that ex vivo SDF-1 induction increased actin polymerization fluorescence in the gated lymphocyte population from blood of vehicle-treated animals compared with non-SDF-1-treated vehicle blood. (B) Compared with vehicle, lymphocytes of AMD3100-treated animals had reduced FITC-phalloidin fluorescence after SDF-1 stimulation, as shown by the slight right shift of the graph compared with that of untreated blood. (C) AMD3100 significantly decreased SDF-1 induced actin polymerization fluorescence compared with vehicle control subjects (*P < 0.05, unpaired t-test). (D) WBC counts increased in the blood of AMD3100-treated animals compared with the control (*P < 0.05, unpaired t-test).
Effect of CXCR4 Inhibition on Choroidal Leakage and Angiogenesis in the Laser-Induced CNV Prevention Model
The efficacy of CXCR4 inhibition in reducing choroidal leakage and angiogenesis was evaluated in the laser-induced CNV prevention model. In the prevention study protocol, treatment generally occurs before, concurrent with, or shortly after laser rupture and inhibits the development of angiogenesis and subsequent leakage. 
AMD3100 30 mg/kg was administered 1 day after laser induction and the regimen continued for 14 consecutive days before endpoint imaging by fluorescein angiography and measurement of CNV lesion size. Compared with vehicle-treated animals, AMD3100-treated animals had reduced leakage of fluorescein from neoangiogenic lesions after late-phase fluorescein angiography (Figs. 2A, 2B). Overall, there was ∼40% inhibition of choroidal leakage (P < 0.0001, Fig. 2E). Isolectin B4 staining of choroidal flat mounts was performed to visualize angiogenic endothelial cells (Figs. 2C, 2D). Angiogenesis was also reduced in AMD3100-treated animals, as the mean size of stained endothelial cell lesions was diminished ∼30% compared with that in vehicle control subjects (P < 0.005, Fig. 2F). These results, coupled with target modulation data showing inhibition of CXCR4, indicate that disrupting the SDF-1/CXCR4 axis is effective in reducing angiogenesis and leakage in the prevention model. 
Figure 2.
 
Antileakage and antiangiogenesis efficacy of AMD3100 in the prevention model of laser-induced CNV. Laser rupture of Bruch's membrane was introduced to five to six sites around the optic disc of each eye on day 0. An osmotic pump delivering AMD3100 or vehicle was implanted on day 1, fluorescein angiography was conducted on day 14, and isolectin B4 staining was completed on day 15. (A, B) Fluorescein leakage from AMD3100-treated animals was reduced compared with that in animals in the vehicle group. (C, D) Isolectin B4-stained endothelial cells in choroidal flat mounts showed a smaller CNV area in AMD3100-treated animals than in the vehicle-treated control subjects. (E) Average leakage score per blister of CXCR4 inhibitor–treated animals was significantly less than that in control animals (****P < 0.0001). (F) The mean CNV lesion area was significantly reduced in the treated versus the control animals (***P < 0.005). Scale bars: 100 μm.
Figure 2.
 
Antileakage and antiangiogenesis efficacy of AMD3100 in the prevention model of laser-induced CNV. Laser rupture of Bruch's membrane was introduced to five to six sites around the optic disc of each eye on day 0. An osmotic pump delivering AMD3100 or vehicle was implanted on day 1, fluorescein angiography was conducted on day 14, and isolectin B4 staining was completed on day 15. (A, B) Fluorescein leakage from AMD3100-treated animals was reduced compared with that in animals in the vehicle group. (C, D) Isolectin B4-stained endothelial cells in choroidal flat mounts showed a smaller CNV area in AMD3100-treated animals than in the vehicle-treated control subjects. (E) Average leakage score per blister of CXCR4 inhibitor–treated animals was significantly less than that in control animals (****P < 0.0001). (F) The mean CNV lesion area was significantly reduced in the treated versus the control animals (***P < 0.005). Scale bars: 100 μm.
Efficacy of AM3100 in the Laser-Induced CNV Intervention Model of Wet AMD
Observed efficacy in the prevention model prompted us to determine whether CXCR4 inhibition would also be effective in an intervention model of laser-induced CNV. In the intervention modality, angiogenesis and choroidal leakage are established first, before administration of the therapeutic agent. In our study, the lesions were established for 14 days before a 2-week administration of AMD3100 at 30 mg/kg per day via an osmotic pump. By 13 days after laser induction, peak leakage was observed through fluorescein angiography. 
Target modulation was confirmed with the SDF-1-induced actin polymerization assay as before, and ∼60% inhibition of actin polymerization was observed in blood lymphocytes from animals treated in the intervention mode (P < 0.05, Fig. 3A). WBC counts were also elevated approximately twofold, indicating in vivo inhibition of CXCR4 (P < 0.001, Fig. 3B). 
Figure 3.
 
In vivo and ex vivo CXCR4 target modulation of AMD3100-treated animals in the intervention laser-induced CNV mode. (A) Significantly less fluorescence from decreased actin polymerization was measured in leukocytes from animals receiving AMD3100 versus control animals (*P < 0.05). (B) The number of WBCs increased significantly in AMD3100 animals over the vehicle control (**P < 0.01).
Figure 3.
 
In vivo and ex vivo CXCR4 target modulation of AMD3100-treated animals in the intervention laser-induced CNV mode. (A) Significantly less fluorescence from decreased actin polymerization was measured in leukocytes from animals receiving AMD3100 versus control animals (*P < 0.05). (B) The number of WBCs increased significantly in AMD3100 animals over the vehicle control (**P < 0.01).
Baseline images of fluorescein angiography were taken 13 days after laser induction, and endpoint images were captured 28 days after lasering, for assessment of choroidal leakage (Figs. 4A–D). Endpoint isolectin B4 staining of choroidal endothelial cells was performed the following day. AMD3100 inhibition of CXCR4 did not significantly reduce leakage in this intervention mode and failed to decrease lesion size compared with that in vehicle control subjects (Figs. 4E–H). 
Figure 4.
 
Antileakage and antiangiogenesis efficacy of AMD3100 in the intervention model of laser-induced CNV. Laser rupture of Bruch's membrane was induced in five to six sites around the optic disc of each eye on day 0. An osmotic pump delivering AMD3100 or vehicle was implanted on day 14. Baseline fluorescein angiography (FA) images were taken on day 13 and endpoint FA on day 28. Isolectin B4 staining was completed on day 29. (A, C) Baseline and endpoint leakage in vehicle animals were similar. (B, D) Fluorescein leakage in baseline and endpoint images of AMD3100-treated animals did not appear to change. (E, F) Isolectin B4 stained endothelial cells in choroidal flat mounts displayed similar sized CNV areas in the AMD3100-treated animals compared with the vehicle control. (G) Average endpoint leakage score per blister of CXCR4 inhibitor–treated animals were nearly equal to that of the baseline control. (H) The mean CNV lesion area did not differ significantly in the treated versus the control animals. Scale bar: 100 μm.
Figure 4.
 
Antileakage and antiangiogenesis efficacy of AMD3100 in the intervention model of laser-induced CNV. Laser rupture of Bruch's membrane was induced in five to six sites around the optic disc of each eye on day 0. An osmotic pump delivering AMD3100 or vehicle was implanted on day 14. Baseline fluorescein angiography (FA) images were taken on day 13 and endpoint FA on day 28. Isolectin B4 staining was completed on day 29. (A, C) Baseline and endpoint leakage in vehicle animals were similar. (B, D) Fluorescein leakage in baseline and endpoint images of AMD3100-treated animals did not appear to change. (E, F) Isolectin B4 stained endothelial cells in choroidal flat mounts displayed similar sized CNV areas in the AMD3100-treated animals compared with the vehicle control. (G) Average endpoint leakage score per blister of CXCR4 inhibitor–treated animals were nearly equal to that of the baseline control. (H) The mean CNV lesion area did not differ significantly in the treated versus the control animals. Scale bar: 100 μm.
Efficacy of SU14813 in the Laser-Induced CNV Intervention Model of Wet AMD
As the intervention model of laser-induced CNV better represents the disease state in patients needing treatment, we wanted to determine whether other angiogenic inhibitors would be efficacious in this model. SU14813, a multiple receptor tyrosine kinase inhibitor of VEGFR, PDGFR-β, KIT, and FLT3, 28 was tested in the intervention model. Lesions were established for 14 days before a 7-day regimen of vehicle or SU14813 began. At the 3-week endpoint, most SU14813-treated animals had decreased leakage, as observed through fluorescein angiography compared with the control animals (Figs. 5A–D). SU14813 10 mg/kg IP administered twice daily was effective in reducing leakage by ∼40% and decreased lesion size by ∼50% (P < 0.05, P < 0.005, Figs. 5E–H). 
Figure 5.
 
Antileakage and antiangiogenesis efficacy of SU14813 in the intervention model of laser-induced CNV. Laser rupture of Bruch's membrane was induced in five to six sites around the optic disc of each eye on day 0. SU14813 was administered via IP injection twice daily at 10 mg/kg, starting at day 14. Baseline fluorescein angiography (FA) images were taken on day 14 and endpoint FA on day 21. Isolectin B4 staining was completed on day 22. (A, C) Baseline and endpoint leakage from vehicle animals were similar. (B, D) Fluorescein leakage in endpoint images of SU14813-treated animals was decreased compared with that at baseline. (E, F) Isolectin B4-stained choroidal flat mounts showed reduced CNV area in SU14813-treated animals compared with that in the vehicle control. (G) The average endpoint leakage score per blister in animals administered a multiple RTK inhibitor was significantly reduced compared with that in baseline control animals (*P < 0.05). (H) Mean CNV lesion areas were significantly smaller in treated versus control animals (***P < 0.005). Scale bar: 100 μm.
Figure 5.
 
Antileakage and antiangiogenesis efficacy of SU14813 in the intervention model of laser-induced CNV. Laser rupture of Bruch's membrane was induced in five to six sites around the optic disc of each eye on day 0. SU14813 was administered via IP injection twice daily at 10 mg/kg, starting at day 14. Baseline fluorescein angiography (FA) images were taken on day 14 and endpoint FA on day 21. Isolectin B4 staining was completed on day 22. (A, C) Baseline and endpoint leakage from vehicle animals were similar. (B, D) Fluorescein leakage in endpoint images of SU14813-treated animals was decreased compared with that at baseline. (E, F) Isolectin B4-stained choroidal flat mounts showed reduced CNV area in SU14813-treated animals compared with that in the vehicle control. (G) The average endpoint leakage score per blister in animals administered a multiple RTK inhibitor was significantly reduced compared with that in baseline control animals (*P < 0.05). (H) Mean CNV lesion areas were significantly smaller in treated versus control animals (***P < 0.005). Scale bar: 100 μm.
Discussion
In our studies, CXCR4 inhibition was efficacious in the prevention modality of the laser-induced CNV model, but failed to reduce choroidal leakage and angiogenesis in the intervention modality, despite showing positive target modulation. This finding suggests that therapies targeting the SDF-1/CXCR4 axis may be beneficial in blocking the induction of ocular neoangiogenesis, but are unlikely to reduce already established angiogenesis. 
There is strong evidence that CXCR4 inhibition disrupts the recruitment of endothelial precursor cells (EPCs) to sites of angiogenesis, most likely the major mechanism leading to efficacy in the prevention model. 24 In this scenario, CXCR4 inhibition of circulating EPCs derived from bone marrow would have disrupted homing of the cells along an SDF-1 gradient to the vascular area of the lesion. Indeed, we observed a reduction in the area of endothelial cells composing the CNV lesions, thus corroborating evidence in other studies in which the SDF-1/CXCR4 axis was inhibited with anti-SDF-1 antibodies, CXCR4 peptide inhibitors, and CXCR4 small molecule inhibitors. 20,29 In two independent studies, 3,4 the contribution of EPCs to the CNV lesion was thought to be 20% to 45%. 
Hypothetically, systemic inhibition of CXCR4 had the potential to increase CNV size by augmenting the mobilization of bone marrow–derived cells and therefore increasing the supply of circulating cells that could be incorporated in the lesion. We did not observe such an effect in our study. Although the circulating EPCs may have increased, they were likely blocked from recruitment to the vascular angiogenic sites by CXCR4 inhibition. Local intravitreal delivery of AMD3100 could circumvent the potential negative effects of increased EPC supply due to systemic inhibition of CXCR4 and may increase local drug concentration and exposure. In a previous study, Lima e Silva et al. 20 delivered other CXCR4 inhibitors locally to the eye, reducing CNV formation in a prevention study protocol. 
In addition to suppressing CNV, CXCR4 inhibition reduced choroidal vascular leakage in the prevention modality. It is not known whether CXCR4 inhibition decreases leakage directly or as a secondary effect of the reduction of the angiogenic vessel area. Inhibition of CXCR4 may block amplification of angiogenesis by disrupting the positive feedback loop of VEGF expression. CXCR4 signaling has been shown to induce VEGF expression in CXCR4-positive endothelial cells. 30,31  
The observation that CXCR4 inhibition did not decrease choroidal leakage or angiogenic lesion size in the intervention modality suggests that after a 2-week generation of laser-induced CNV, there is limited, if any, contribution of EPC cells to the already established vessels. In addition, it suggests that disruption of the SDF-1/CXCR4 axis will not be effective as a monotherapy in treating patients with existing CNV. 
A multiple receptor tyrosine kinase inhibitor may still be an effective monotherapy, as SU14813 reduced the size of previously formed lesions. In the treatment mode of the laser-induced CNV model, both leakage and angiogenesis decreased even after the pathologic effect was given 14 days to fully establish before drug intervention. Our results suggest that blockade of the VEGF receptor is an effective alternative method of inhibiting the crucial VEGF pathway versus targeting the ligand by anti-VEGF aptamers or antibodies. 
SU14813 is a small molecule with broad target RTK selectivity, inhibiting VEGFR, PDGFR-β, KIT, and FLT3. 28 Although the primary mechanism that reduces preexisting angiogenesis and leakage is the blockade of the central VEGF pathway, the additional inhibition of PDGFR-β may augment efficacy in this model over single anti-VEGFR agents. PDGFR-β is known to regulate pericyte recruitment to developing vessels, and inhibition of PDGFR-β in combination with VEGF inhibition has been shown to cause regression of corneal angiogenesis and laser-induced CNV better than either inhibitor alone. 15  
Many potential wet AMD therapies have been tested in the prevention modality of laser-induced CNV and are capable of reducing CNV size in this study design. 1623 Although these therapies are useful for proof-of-concept studies, patients seeking treatment have an established disease, and it is unlikely that the prevention model would be translatable to the human condition in predicting efficacy. The intervention laser-induced CNV model is likely to be a better model for testing the efficacy of candidate monotherapies or combination therapies. Studies have shown that an anti-VEGF aptamer and a multikinase inhibitor cause vessel regression in a mouse intervention laser-induced CNV model. 24,25 In a primate laser-induced CNV model in an interventional study protocol, ranibizumab (Lucentis; Genentech) was shown to be effective in reducing leakage from preexisting lesions before demonstrating success in the clinic, therefore highlighting its predictive power. 32 The leakage endpoint shown by fluorescein angiography in the laser-induced CNV model is an important indicator of therapeutic response and has been used as a secondary outcome measurement in clinical trials. However, it may still be useful to develop an endpoint assessment of angiogenesis in the primate model to test single and combination treatments on CNV vessel regression. 
Footnotes
 Supported by Pfizer, Inc.
Footnotes
 Disclosure: E.S. Lee, Pfizer (F, E), D. Rewolinski, Pfizer (F, E)
The authors thank Larry Kahn and Jill Hallin for assistance with the AP biomarker protocol and flow cytometry, Vicky Holway and Philip Morton for the supply of AMD3100, and the Pfizer La Jolla Comparative Medicine Group for animal care support. 
References
Friedman DS O'Colmain BJ Muñoz B . Prevalence of age-related macular degeneration in the United States. Arch Ophthalmol. 2004;122:564–572. [CrossRef] [PubMed]
Grant MB May WS Cabellero S . Adult hematopoietic stem cells provide functional hemangioblast activity during retinal neovascularization. Nat Med. 2002;8:607–612. [CrossRef] [PubMed]
Espinosa-Heidmann DG Suner IJ Hernandez EP . Macrophage depletion diminishes lesion size and severity in experimental choroidal neovascularization. Invest Ophthalmol Vis Sci. 2003;44:3586–3592. [CrossRef] [PubMed]
Sengupta N Caballero S Mames RN . The role of adult bone marrow-derived stem cells in choroidal neovascularization. Invest Ophthalmol Vis Sci. 2003;44:4908–4913. [CrossRef] [PubMed]
Federsppiel B Melhado IG Duncan AM . Molecular cloning of the cDNA and chromosomal localization of the gene for a putative seven-transmembrane segment (7-TMS) receptor isolated from human spleen. Genomics. 1993;16:707–712. [CrossRef] [PubMed]
Nomura H Nielsen BW Matsushima K . Molecular cloning of cDNAs encoding a LD78 receptor and putative leukocyte chemotactic peptide receptors. Int Immunol. 1993;5:1239–1249. [CrossRef] [PubMed]
Loetscher M Geiser T O'Reilly T . Cloning of a human seven-transmembrane domain receptor, LESTR, that is highly expressed in leukocytes. J Biol Chem. 1994;269:232–237. [PubMed]
McGrath KE Koniski AD Maltby KM . Embryonic expression and function of the chemokine SDF-1 and its receptor, CXCR4. Dev Biol. 1999;213:442–456. [CrossRef] [PubMed]
Aiuti A Webb IJ Bleul C . The chemokine SDF-1 is a chemoattractant for human CD34+ hematopoietic progenitor cells and provides a new mechanism to explain the mobilization of CD34+ progenitors to peripheral blood. J Exp Med. 1997;185:111–120. [CrossRef] [PubMed]
Jo DY Rafii S Hamada T . Chemotaxis of primitive hematopoietic cells in response to stromal cell-derived factor-1. J Clin Invest. 2000;105:101–111. [CrossRef] [PubMed]
Ceradini DJ Kulkarni AR Callaghan MJ . Progenitor cell trafficking is regulated by hypoxic gradients through HIF-1 induction of SDF-1. Nat Med. 2004;10:858–864. [CrossRef] [PubMed]
De Falco E. Porcelli D Torella AR . SDF-1 involvement in endothelial phenotype and ischemia-induced recruitment of bone marrow progenitor cells. Blood. 2004;104:3472–3482. [CrossRef] [PubMed]
Walter DH Haendeler J Reinhold J . Impaired CXCR4 signaling contributes to the reduced neovascularization capacity of endothelial progenitor cells from patients with coronary artery disease. Circ Res. 2005;97:1142–1151. [CrossRef] [PubMed]
Butler JM Guthrie SM Koc M . SDF-1 is both necessary and sufficient to promote proliferative retinopathy. J Clin Invest. 2005;115:86–93. [CrossRef] [PubMed]
Takeda A Hata Y Shiose S . Suppression of experimental choroidal neovascularization utilizing KDR selective receptor tyrosine kinase inhibitor. Graefes Arch Clin Exp Ophthalmol. 2004;241:765–782. [CrossRef]
Kinose F Roscilli G Lamartina S . Inhibition of retinal and choroidal neovascularization by a novel KDR kinase inhibitor. Mol Vis. 2005;27:366–373.
Murnaka K Yanagi Y Tamaki Y . Suppression of laser-induced choroidal neovascularization by oral administration of SA3443 in mice. FEBS Lett. 2005;579:6084–6088. [CrossRef] [PubMed]
Takahashi H Obata R Tamaki Y . A novel vascular endothelial growth factor receptor 2 inhibitor, SU11248, suppresses choroidal neovascularization in vivo. J Ocul Pharmacol Ther. 2006;228:213–218. [CrossRef]
Fu Y Ponce ML Thill M . Angiogenesis inhibition and choroidal neovascularization suppression by sustained delivery of an integrin antagonist, EMD478761. Invest Ophthalmol Vis Sci. 2007;48:5184–90. [CrossRef] [PubMed]
Lima e Silva R Shen J Hackett SF . The SDF-1/CXCR4 ligand/receptor pair is an important contributor to several types of ocular neovascularization. FASEB J. 2007;21:3219–3230. [CrossRef] [PubMed]
Marneros AG She H Zambarakji H . Endogenous endostatin inhibits choroidal neovascularization. FASEB J. 2007;21:3809–3818. [CrossRef] [PubMed]
Nagai N Oike Y Izumi-Nagai K . Suppression of choroidal neovascularization by inhibiting angiotensin-converting enzyme: minimal role of bradykinin. Invest Ophthalmol Vis Sci. 2007;48:2321–2326. [CrossRef] [PubMed]
Yamada K Sakurai E Itaya M . Inhibition of laser-induced choroidal neovascularization by atorvastatin by downregulation of monocyte chemotactic protein-1 synthesis in mice. Invest Ophthalmol Vis Sci. 2007;48:1839–1843. [CrossRef] [PubMed]
Jo N Mailhos C Ju M . Inhibition of platelet-derived growth factor B signaling enhances the efficacy of anti-vascular endothelial growth factor therapy in multiple models of ocular neovascularization. Am J Pathol. 2006;168:2036–2053. [CrossRef] [PubMed]
Chung EJ Yoo S Lim HJ . Inhibition of choroidal neovascularization in mice by systemic administration of the multikinase inhibitor, sorafenib. Br J Opthalmol. 2008;93:958–963. [CrossRef]
De Clercq E . Potential clinical applications of the CXCR4 antagonist bicyclam AMD3100. Mini Rev Med Chem. 2005;5:805–824. [CrossRef] [PubMed]
Hendrix CW Flexner C MacFarland RT . Pharmacokinetics and safety of AMD-3100, a novel antagonist of the CXCR-4 chemokine receptor, in human volunteers. Antimicrob Agents Chemother. 2000;44:667–673.
Patyna S Laird AD Mendel DB . SU14813: a novel multiple receptor tyrosine kinase inhibitor with potent antiangiogenic and antitumor activity. Mol Cancer Ther. 2006;5:1774–1782. [CrossRef] [PubMed]
Sengupta N Caballero S Mames RN . Preventing stem cell incorporation into choroidal neovascularization by targeting homing and attachment factors. Invest Ophthalmol Vis Sci. 2005;46:343–348. [CrossRef] [PubMed]
Salcedo R Wasserman K Young HA . Vascular endothelial growth factor and basic fibroblast growth factor induce expression of CXCR4 on human endothelial cells: in vivo neovascularization induced by stromal-derived factor-1alpha. Am J Pathol. 1999;154:1125–1135. [CrossRef] [PubMed]
Mirshahi F Pourtau J Li H . SDF-1 activity on microvascular endothelial cells: consequences on angiogenesis in in vitro and in vivo models. Thromb Res. 2000;99:587–594. [CrossRef] [PubMed]
Krzystolik MG Afshari MA Adamis AP . Prevention of experimental choroidal neovascularization with intravitreal anti-vascular endothelial growth factor antibody fragment. Arch Ophthalmol. 2002;120:338–346. [CrossRef] [PubMed]
Figure 1.
 
In vivo and ex vivo CXCR4 target modulation of AMD3100-treated animals in the prevention laser-induced CNV mode. Blood from AMD3100-and vehicle-treated animals was collected 12 days after osmotic pump implantation and was used in actin polymerization and WBC count assays. (A) Flow cytometry showed that ex vivo SDF-1 induction increased actin polymerization fluorescence in the gated lymphocyte population from blood of vehicle-treated animals compared with non-SDF-1-treated vehicle blood. (B) Compared with vehicle, lymphocytes of AMD3100-treated animals had reduced FITC-phalloidin fluorescence after SDF-1 stimulation, as shown by the slight right shift of the graph compared with that of untreated blood. (C) AMD3100 significantly decreased SDF-1 induced actin polymerization fluorescence compared with vehicle control subjects (*P < 0.05, unpaired t-test). (D) WBC counts increased in the blood of AMD3100-treated animals compared with the control (*P < 0.05, unpaired t-test).
Figure 1.
 
In vivo and ex vivo CXCR4 target modulation of AMD3100-treated animals in the prevention laser-induced CNV mode. Blood from AMD3100-and vehicle-treated animals was collected 12 days after osmotic pump implantation and was used in actin polymerization and WBC count assays. (A) Flow cytometry showed that ex vivo SDF-1 induction increased actin polymerization fluorescence in the gated lymphocyte population from blood of vehicle-treated animals compared with non-SDF-1-treated vehicle blood. (B) Compared with vehicle, lymphocytes of AMD3100-treated animals had reduced FITC-phalloidin fluorescence after SDF-1 stimulation, as shown by the slight right shift of the graph compared with that of untreated blood. (C) AMD3100 significantly decreased SDF-1 induced actin polymerization fluorescence compared with vehicle control subjects (*P < 0.05, unpaired t-test). (D) WBC counts increased in the blood of AMD3100-treated animals compared with the control (*P < 0.05, unpaired t-test).
Figure 2.
 
Antileakage and antiangiogenesis efficacy of AMD3100 in the prevention model of laser-induced CNV. Laser rupture of Bruch's membrane was introduced to five to six sites around the optic disc of each eye on day 0. An osmotic pump delivering AMD3100 or vehicle was implanted on day 1, fluorescein angiography was conducted on day 14, and isolectin B4 staining was completed on day 15. (A, B) Fluorescein leakage from AMD3100-treated animals was reduced compared with that in animals in the vehicle group. (C, D) Isolectin B4-stained endothelial cells in choroidal flat mounts showed a smaller CNV area in AMD3100-treated animals than in the vehicle-treated control subjects. (E) Average leakage score per blister of CXCR4 inhibitor–treated animals was significantly less than that in control animals (****P < 0.0001). (F) The mean CNV lesion area was significantly reduced in the treated versus the control animals (***P < 0.005). Scale bars: 100 μm.
Figure 2.
 
Antileakage and antiangiogenesis efficacy of AMD3100 in the prevention model of laser-induced CNV. Laser rupture of Bruch's membrane was introduced to five to six sites around the optic disc of each eye on day 0. An osmotic pump delivering AMD3100 or vehicle was implanted on day 1, fluorescein angiography was conducted on day 14, and isolectin B4 staining was completed on day 15. (A, B) Fluorescein leakage from AMD3100-treated animals was reduced compared with that in animals in the vehicle group. (C, D) Isolectin B4-stained endothelial cells in choroidal flat mounts showed a smaller CNV area in AMD3100-treated animals than in the vehicle-treated control subjects. (E) Average leakage score per blister of CXCR4 inhibitor–treated animals was significantly less than that in control animals (****P < 0.0001). (F) The mean CNV lesion area was significantly reduced in the treated versus the control animals (***P < 0.005). Scale bars: 100 μm.
Figure 3.
 
In vivo and ex vivo CXCR4 target modulation of AMD3100-treated animals in the intervention laser-induced CNV mode. (A) Significantly less fluorescence from decreased actin polymerization was measured in leukocytes from animals receiving AMD3100 versus control animals (*P < 0.05). (B) The number of WBCs increased significantly in AMD3100 animals over the vehicle control (**P < 0.01).
Figure 3.
 
In vivo and ex vivo CXCR4 target modulation of AMD3100-treated animals in the intervention laser-induced CNV mode. (A) Significantly less fluorescence from decreased actin polymerization was measured in leukocytes from animals receiving AMD3100 versus control animals (*P < 0.05). (B) The number of WBCs increased significantly in AMD3100 animals over the vehicle control (**P < 0.01).
Figure 4.
 
Antileakage and antiangiogenesis efficacy of AMD3100 in the intervention model of laser-induced CNV. Laser rupture of Bruch's membrane was induced in five to six sites around the optic disc of each eye on day 0. An osmotic pump delivering AMD3100 or vehicle was implanted on day 14. Baseline fluorescein angiography (FA) images were taken on day 13 and endpoint FA on day 28. Isolectin B4 staining was completed on day 29. (A, C) Baseline and endpoint leakage in vehicle animals were similar. (B, D) Fluorescein leakage in baseline and endpoint images of AMD3100-treated animals did not appear to change. (E, F) Isolectin B4 stained endothelial cells in choroidal flat mounts displayed similar sized CNV areas in the AMD3100-treated animals compared with the vehicle control. (G) Average endpoint leakage score per blister of CXCR4 inhibitor–treated animals were nearly equal to that of the baseline control. (H) The mean CNV lesion area did not differ significantly in the treated versus the control animals. Scale bar: 100 μm.
Figure 4.
 
Antileakage and antiangiogenesis efficacy of AMD3100 in the intervention model of laser-induced CNV. Laser rupture of Bruch's membrane was induced in five to six sites around the optic disc of each eye on day 0. An osmotic pump delivering AMD3100 or vehicle was implanted on day 14. Baseline fluorescein angiography (FA) images were taken on day 13 and endpoint FA on day 28. Isolectin B4 staining was completed on day 29. (A, C) Baseline and endpoint leakage in vehicle animals were similar. (B, D) Fluorescein leakage in baseline and endpoint images of AMD3100-treated animals did not appear to change. (E, F) Isolectin B4 stained endothelial cells in choroidal flat mounts displayed similar sized CNV areas in the AMD3100-treated animals compared with the vehicle control. (G) Average endpoint leakage score per blister of CXCR4 inhibitor–treated animals were nearly equal to that of the baseline control. (H) The mean CNV lesion area did not differ significantly in the treated versus the control animals. Scale bar: 100 μm.
Figure 5.
 
Antileakage and antiangiogenesis efficacy of SU14813 in the intervention model of laser-induced CNV. Laser rupture of Bruch's membrane was induced in five to six sites around the optic disc of each eye on day 0. SU14813 was administered via IP injection twice daily at 10 mg/kg, starting at day 14. Baseline fluorescein angiography (FA) images were taken on day 14 and endpoint FA on day 21. Isolectin B4 staining was completed on day 22. (A, C) Baseline and endpoint leakage from vehicle animals were similar. (B, D) Fluorescein leakage in endpoint images of SU14813-treated animals was decreased compared with that at baseline. (E, F) Isolectin B4-stained choroidal flat mounts showed reduced CNV area in SU14813-treated animals compared with that in the vehicle control. (G) The average endpoint leakage score per blister in animals administered a multiple RTK inhibitor was significantly reduced compared with that in baseline control animals (*P < 0.05). (H) Mean CNV lesion areas were significantly smaller in treated versus control animals (***P < 0.005). Scale bar: 100 μm.
Figure 5.
 
Antileakage and antiangiogenesis efficacy of SU14813 in the intervention model of laser-induced CNV. Laser rupture of Bruch's membrane was induced in five to six sites around the optic disc of each eye on day 0. SU14813 was administered via IP injection twice daily at 10 mg/kg, starting at day 14. Baseline fluorescein angiography (FA) images were taken on day 14 and endpoint FA on day 21. Isolectin B4 staining was completed on day 22. (A, C) Baseline and endpoint leakage from vehicle animals were similar. (B, D) Fluorescein leakage in endpoint images of SU14813-treated animals was decreased compared with that at baseline. (E, F) Isolectin B4-stained choroidal flat mounts showed reduced CNV area in SU14813-treated animals compared with that in the vehicle control. (G) The average endpoint leakage score per blister in animals administered a multiple RTK inhibitor was significantly reduced compared with that in baseline control animals (*P < 0.05). (H) Mean CNV lesion areas were significantly smaller in treated versus control animals (***P < 0.005). Scale bar: 100 μm.
×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×