| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
* Departments of Ophthalmology and Neuroscience, The Johns Hopkins University School of Medicine, Baltimore, Maryland, USA; and
Department of Ophthalmics Research, Merck & Co., West Point, Pennsylvania, USA
1Correspondence: Maumenee 719, Johns Hopkins University School of Medicine, 600 N. Wolfe St., Baltimore, MD 21287-9277, USA. E-mail: pcampo{at}jhmi.edu
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: age-related macular degeneration • angiogenesis • chemokines • diabetic retinopathy • inflammation
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, 2)
.
CXCR4 was initially cloned as an orphan chemokine receptor and was found to be expressed on monocytes, B-lymphocytes, and most T cells (3
4
5)
. It gained instant notoriety when it was identified as an essential cofactor for entry of T-tropic HIV into T cells (6
7
8)
. It was subsequently determined that CXCR4 is present on many different types of cells, is activated by only one ligand, stromal-derived factor 1 (SDF-1) (9
, 10)
, and mediates several different activities such as chemotaxis, adhesion, proliferation, survival, and, in some cells, apoptosis (11)
. Activation of CXCR4 on lymphocytes and monocytes stimulates chemotaxis, resulting in recruitment to sites of immune and inflammatory reactions. Hematopoietic and endothelial progenitor cells express CXCR4, and release of SDF-1 by bone marrow (BM) stromal cells mediates sequestration and homing of these progenitor cells to BM (12
, 13)
. SDF-1 has also been implicated in revascularization of ischemic hind limbs through recruitment of CXCR4+ hemangiocytes (14)
.
Mature vascular endothelial cells also express CXCR4 (15)
and its expression is up-regulated by inflammatory cytokines and angiogenic factors FGF2 and VEGF (16
17
18)
. Subcutaneous injections of SDF-1 in mice induce leukocytic infiltrates and small areas of neovascularization (NV) (18)
. Conversely, mice deficient in CXCR4 have defects in the formation of blood vessels of the gastrointestinal tract (19)
. In this study, we investigated the potential role of SDF-1 and CXCR4 in several types of ocular NV.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, 21)
. Seven-day-old (P7) mice and their mothers were placed in an airtight incubator and exposed to an atmosphere of 75 ± 3% oxygen for 5 days. They were returned to room air at P12 and euthanized at P13, P15, or P17, then retinal RNA was isolated or eyes were frozen for immunohistochemistry. A control group was kept at room air and the retinas were collected at the same time point as the mice with oxygen-induced ischemic retinopathy.
Transgenic mice with VEGF-induced NV
Transgenic mice with VEGF expression driven by the rhodopsin promoter (Rho/VEGF mice) have increased levels of VEGF in photoreceptors starting at P7 (22
, 23)
. Mice hemizygous for the transgene were euthanized at P14 or P21 and retinal RNA was isolated for real-time RT-PCR.
Double transgenic mice with inducible expression of VEGF
Adult double transgenic Tet/Opsin/VEGF mice with doxycycline-inducible expression of VEGF in photoreceptors (24)
were treated with 2 mg/ml of doxycycline in their drinking water for 5 days, then euthanized, and total retinal RNA was prepared for real-time PCR.
Real-time RT-PCR
Samples of retinal RNA were treated with DNA-free reagent (Ambion, Inc., Austin, TX, USA) to remove contaminating DNA, and cDNA was synthesized using SuperScript III Reverse Transcriptase (Invitrogen Corporation, Carlsbad, CA, USA). Real-time PCR was performed using a Roche Lightcycler and LightCycler FastStart DNA MasterPlus SYBR Green I reagents (Roche Applied Science) as described previously (24)
. The following primers were used: Cxcr4, forward primer 5'-AGC ATG ACG GAC AAG TAC C-3' and reverse primer 5'-GAT GAT ATG GAC AGC CTT ACA C-3'; Sdf1, forward primer 5'-GAG AGC CAC ATC GCC AGA G-3' and reverse primer 5'-TTT CGG GTC AAT GCA CAC TTG-3'; and Cyclophillin A, forward primer 5'-CAG ACG CCA CTG TCG CTT T-3' and reverse primer 5'-TGT CTT TGG AAC TTT GTC TGC AA-3'. Each sample was run in triplicate. The amount of Cyclophillin A mRNA was used for standardization and results are reported as the number of Cxcr4 or Sdf1 transcripts for per 105 Cyclophillin A transcripts.
Immunofluorescent staining
Eyes were frozen in Tissue Tek embedding compound (Sakura Finetechnical Co., Tokyo, Japan). Frozen sections (10 µm) were dried and fixed with 4% paraformaldehyde and blocked with 8% normal donkey serum. The antibodies used for staining were goat/anti-mouse CXCR4 (1:150, Novus Biologicals, Inc., Littleton, CO, USA), rat/anti-mouse PECAM-1, GFAP, and CD45 (each 1:100, PharMingen, Los Angeles, CA, USA); rabbit anti-mouse SDF-1 (1:50, Abcam, Cambridge, MA, USA), and rat/anti-mouse F4/80 (1:50, eBIOSCIENCE, San Diego, CA, USA). Secondary antibodies were Cy3 or FITC-conjugated donkey anti-rat, donkey anti-rabbit, or donkey anti-goat antibodies (Jackson ImmunoResearch Laboratories, Inc., West Grove, PA, USA). Sections were counterstained with Hoechst (1:1200, Sigma) for 3 min at room temperature. Slides were mounted with Aquamount solution and viewed with a fluorescence microscope (Nikon Instruments Inc., New York, NY, USA) using SPOT RT 3.4 software.
In vivo immunofluorescent staining for PECAM-1, CD45, F4/80, and CXCR4 was done in whole retina. Mice with ischemic retinopathy (P17) were given an intraocular injection of 1 µl containing a mixture of goat anti-CXCR4 and one of three other antibodies: anti-PECAM1-FITC, anti-CD45-phycoerythrin (PE), or F4/80-PE. After 12 h, the mice were euthanized and eyes were immersed in 10% phosphate-buffered formalin for 4 h. Entire retinas were dissected from eyecups and the intact retinas were washed, placed in small tubes, and incubated for 40 min at room temperature with donkey/anti-goat IgG conjugated with Cy3 or FITC. Radial cuts were made in the retinas and flat-mounted in Aquamount solution.
Whole eyecups consisting of RPE, choroid, and sclera from mice with laser-induced CNV were immunostained 14 days after rupture of Bruch’s membrane. Eyes were fixed in 10% phosphate-buffered formalin for 4 h, retinas were removed, and eyecups were incubated in a mixture of goat anti-CXCR4 and one of three other antibodies—anti-PECAM1-FITC, anti-CD45-phycoerythrin (PE), or F4/80-PE—at room temperature for 2 h. After three washes, eyecups were incubated for 40 min at room temperature with donkey/anti-goat IgG conjugated with Cy3 or FITC. Radial cuts were made in the eyecups and they were flat-mounted in Aquamount solution.
CXCR4 antagonists
AMD8664 (1-pyridin-2-yl-N-[4-(1,4,7-triazacyclotetradecan-4-ylmethyl)benzyl]methanamine), TC14012 (R-R-Nal-C-Y-(L)Cit-K-(D)Cit-P-Y-R-(L)citrulline-C-R-NH2, where Nal=L-3-(2-naphthylalanine), Cit=citruline and the peptide is cyclized with the cysteines), and compound 3 (the Merck CXCR4 inhibitor, N-(1-methyl-1-phenylethyl)-N-[((3S)-1-{2-[5-(4H-1,2,4-triazol-4-yl)-1H-indol-3-yl]ethyl}pyrrolidin-3-yl)methyl]amine) were synthesized at Merck and Co. (West Point, PA, USA). The compounds were dissolved in sterile Dulbecco’s PBS (Invitrogen, Carlsbad, CA, USA) and adjusted to pH 7 if necessary.
Assessment of CXCR4 antagonists in mice with ischemic retinopathy
Mice with ischemic retinopathy were given daily periocular injections starting at P12 of 3 µl of 1.25 mM of Merck compound 3 (n=8) in one eye and vehicle (n=8) in the fellow eye. At P17, the mice were euthanized and eyes were frozen in Tissue Tek. Frozen sections (10 µm) were air dried for 7 min and fixed for 30 min in 4.0% paraformaldehyde in PBS, then washed in PBS for 10 min. Sections were histochemically stained with 1:20 biotinylated Griffonia simplicifolia lectin B4 (GSA, Vector Laboratories, Burlingame, CA, USA), which selectively binds to vascular cells as described previously (21)
. Slides were incubated in methanol/H2O2 for 10 min at 4°C, washed with 0.05 M Tris-buffered saline, pH 7.6 (TBS), and incubated for 30 min in 10% normal porcine serum. Slides were incubated for 2 h at room temperature with biotinylated GSA and, after rinsing with 0.05M TBS, were incubated with 1:100 avidin coupled to peroxidase (Vector Laboratories) for 30 min at room temperature. After being washed for 10 min with 0.05 M TBS, slides were incubated for 2.5 min with diaminobenzidine (Invitrogen) to give a brown reaction product, dehydrated with an alcohol step gradient and xylene, and mounted with Cytoseal XYL (Richard-Allen Scientific, Kalamazoo, MI, USA).
To perform quantitative assessments, 10 µm frozen serial sections were cut from the iris root on one side of the eye to the iris root on the opposite side of the eye and every tenth section was stained with GSA. Retinas were examined with an Axioskop microscope and images were digitized using a 3 CCD color video camera and a frame grabber. Image-Pro Plus software was used to delineate GSA-stained vascular cells above the internal limiting membrane and the total area of staining was measured.
Treatment of VEGF transgenics with CXCR4 antagonists
Hemizygous transgene positive rho/VEGF mice were given a periocular injection of 3 µl of 0.3 mM TC14012 dissolved in PBS or 3 µl of PBS alone. At P21, mice were euthanized and the amount of subretinal NV was quantified as described previously (23)
. Briefly, mice were anesthetized, perfused with fluorescein-labeled dextran, and the number of neovascular lesions on the outer surface of the retina and their total area were measured on retinal flat mounts by image analysis.
Mouse model of laser-induced choroidal neovascularization (CNV)
Laser photocoagulation-induced rupture of Bruch’s membrane was used to generate CNV (25)
. Briefly, 4- to 5-wk-old female C57BL/6J mice were anesthetized with xylazine hydrochloride (10 mg/kg) and ketamine hydrochloride (50 mg/kg) and the pupils were dilated with 1% tropicamide (Alcon Labs, Inc., Forth Worth, TX, USA). Three burns of 532 nm diode laser photocoagulation (75 µm spot size, 0.1 s duration, 120 mW) were delivered to each retina using the slit lamp delivery system of an OcuLight GL Photocoagulator (Iridex, Mountain View, CA, USA) and a hand-held cover slide as a contact lens. Burns were performed in the 9, 12, and 3 o’clock positions of the posterior pole of the retina. Production of a bubble at the time of laser, which indicates rupture of Bruch’s membrane, is an important factor in obtaining CNV (25)
, so only burns in which a bubble was produced were included in the study.
Treatments were begun immediately after laser and included 1) intravitreous injections immediately after and 7 days after laser treatment of 1 µl of AMD8664 (0.9 mM), TC14012 (0.3 mM), or Merck compound 3 (1.25 mM) dissolved in PBS, or PBS alone, or 2) daily periocular injections of 5 µl AMD8664 (0.9 mM), TC14012 (0.3 mM), Merck compound 3 (1.25 mM), or 5 µl of PBS alone. Fourteen days after laser, some mice were used to measure the amount of CNV at Bruch’s membrane rupture sites and some were used to measure drug levels in ocular tissues.
Measurement of the area of CNV at Bruch’s membrane rupture sites
Two weeks after rupture of Bruch’s membrane, mice were anesthetized and perfused with fluorescein-labeled dextran (2x106 average mol wt, Sigma, St. Louis, MO, USA) and choroidal flat mounts were prepared as described previously (26)
. Briefly, the eyes were removed, fixed for 1 h in 10% phosphate-buffered formalin, and the cornea and lens were removed. The entire retina was carefully dissected from the eyecup, then radial cuts were made from the edge of the eyecup to the equator in all four quadrants and flat-mounted in Aquamount. Flat mounts were examined by fluorescence microscopy using an Axioskop microscope (Zeiss, Thornwood, NY, USA) and images were digitized using a 3 CCD color video camera (IK-TU40A, Toshiba, Tokyo, Japan) and a frame grabber. Image-Pro Plus software (Media Cybernetics, Silver Spring, MD, USA) was used to measure the area of each CNV lesion. Statistical comparisons were made using a linear mixed model (27)
. This model is analogous to analysis of variance (ANOVA), but allows for analysis of all NV area measurements from each mouse rather than average area of NV per mouse by accounting for correlation between measurements from the same mouse. The advantage of this model over ANOVA is that it accounts for differing precision in mouse-specific average measurements arising from a varying number of observations among mice. P values for comparison of treatments were adjusted for multiple comparisons using Dunnett’s method.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. Exposure of P7 mice to hyperoxia causes down-regulation of VEGF in the retina and other tissues. Reduction of VEGF causes dropout of many retinal vessels, and the poorly vascularized retinas become ischemic after the mice are returned to room air on P12. One day after the onset of ischemia, Vegf mRNA is markedly elevated in the retina (29)
. The level of Sdf1 mRNA was also significantly elevated compared with that in normoxic retinas from normal P13 mice (Fig. 1
A). There was further elevation of the Sdf1 mRNA level in the retina between 1 and 3 days of ischemia and no substantial change between 3 and 5 days. The level of Cxcr4 mRNA in the retina was also highly elevated 1 day after the onset of ischemia, then decreased over the next 2–4 days (Fig. 1B
).
|
In Rho/VEGF transgenic mice, the rhodopsin promoter drives expression of VEGF in photoreceptors (22)
. The ectopic expression of VEGF begins at P7 and is sustained. At P14, 1 wk after onset of VEGF expression, the level of Cxcr4 mRNA was markedly elevated in the retinas of transgenics compared with nontransgenic littermates (Fig. 1C
). Despite the sustained expression of VEGF, Cxcr4 mRNA in the retinas of transgenics dropped between P14 and P21 and was no longer different from that in nontransgenics. The increased expression of VEGF in the retinas of transgenics did not result in a detectable difference in Sdf1 mRNA in the retinas of transgenics compared with those of nontransgenics at P14 or P21 (not shown).
In Tet/opsin/VEGF double transgenic mice treated with 2 mg/ml of doxycycline for 5 days, the level of expression of VEGF in the retina is as least 5-fold higher than that in Rho/VEGF mice (30)
. At 5 days after the onset of this high level of VEGF expression in the retina, both Cxcr4 and Sdf1 mRNA were significantly elevated compared with controls (Fig. 1D
).
Localization of CXCR4 in ischemic retina
Immunofluorescent staining of ocular sections from nonischemic eyes of P17 mice showed faint staining in the RPE, and choroid and rare focal areas of staining in the retina for CXCR4 (Fig. 2
A). There was a similar pattern of staining for CD45 (Fig. 2B
), which is expressed exclusively on nucleated hematopoietic cells and is a marker commonly used to identify BM-derived cells (31)
. Staining with the F4/80 monoclonal antibody that specifically recognizes macrophages in mouse tissues (32)
also showed a similar pattern (Fig. 2C
). Sections from eyes with ischemic retinopathy showed prominent fluorescein labeling due to CXCR4 localization throughout the inner retina (Fig. 2D, G, J
). Many of the areas that stained strongly for CXCR4 also showed labeling when the same section was stained for CD45 (Fig. 2E
). Colocalization was illustrated by merging the images (Fig. 2F
). This suggests that BM-derived cells account for a substantial amount of CXCR4 staining. Staining for GFAP occurred through the inner half of ischemic retina and in a band along the outer retina (Fig. 2H
). This is due to staining of astrocytes and some Muller cell processes; staining for GFAP is generally minimal in Muller cells of normal retina, but is up-regulated in ischemic retina (33)
. Merging of this image with the image from the same section stained for CXCR4 showed many regions of colocalization (Fig. 2I
), indicating that ischemic retinal glial cells express CXCR4. Sections from ischemic eyes showed staining throughout the inner retina with F4/80, indicating macrophage infiltration (Fig. 2K
), and much of the staining colocalized with staining for CXCR4 (Fig. 2L
). Elimination of primary antibodies showed faint background fluorescence that looked the same for normoxic and hypoxic retinas (not shown).
|
Whole mounts of immunostained retinas provide a panoramic view that can provide useful perspective. Whole mounts of retinas from P17 mice with ischemic retinopathy showed aggregates of CD45+ cells (Fig. 3
A) and CXCR4+ cells (Fig. 3B
) throughout the retina, and merging of the images showed that many of the cells expressed both (Fig. 3C
). Ischemic retinas were also infiltrated with F4/80+ cells (Fig. 3D
), many of which coexpressed CXCR4 (Fig. 3E
, F). CXCR4+ cells were located in close association with retinal vessels but did not coexpress the vascular endothelial cell marker PECAM-1 (Fig. 3G-I
), indicating that CXCR4 is not expressed in retinal vascular cells. Thus, much of the increase in CXCR4 expression in ischemic retina is due to influx of BM-derived cells.
|
Choroidal whole mounts from mice with CNV occurring 14 days after laser-induced rupture of Bruch’s membrane showed prominent staining for CD45 throughout the area of CNV (Fig. 3J
) that colocalized with CXCR4 (Fig. 3K, L
). The CNV was also densely infiltrated with F4/80+ cells (Fig. 3M
), many of which also stained for CXCR4 (Fig. 3N
, O). The CXCR4+ cells were closely associated with the new vessels within the CNV lesion, but appeared distinct as they did not label with anti-PECAM-1 (Fig. 3P-R
).
Localization of SDF-1 in ischemic retina
In nonischemic P17 control retinas, there was faint nonspecific staining for SDF-1 along the outer border of the retina and in the choroid (Fig. 4
A), similar to that seen in sections stained without primary antibody (not shown). Nonischemic retinas also showed specific staining for GFAP along the inner border of the retina, which is known to label astrocytes, and nonspecific staining along the outer border of the retina and in the choroid (Fig. 4B
). A merged image that included DAPI staining (Fig. 4C
) looked similar to the GFAP-stained image alone. Sections from P17 mice with ischemic retinopathy showed staining for SDF-1 in the inner part of the retina, indicating increased levels of SDF-1 in ischemic retina (Fig. 4D
). Compared with nonischemic retina, ischemic retina showed an increase in staining for GFAP due to labeling of Muller cell processes in the inner retina as well as astrocytes near the retinal surface (Fig. 4E
). Merging of the images from Fig. 4D and E
showed substantial colocalization of SDF-1 and GFAP, indicating that ischemic retinal glia contain SDF-1 (Fig. 4F
). Staining for SDF-1 occurred adjacent to PECAM-1-stained new vessels; there did not appear to be staining for SDF-1 within endothelial cells (Fig. 4G-I
).
|
Blockade of CXCR4 reduces CD45+ and F4/80+ cells in ischemic retina
Mice with ischemic retinopathy were given a daily periocular injection of 3 µl of 1.25 mM Merck compound 3 in one eye and PBS in the fellow eye between P12 and P17. Compared with retinal whole mounts from PBS-treated eyes, those from eyes treated with the CXCR4 antagonist viewed at a magnification that allows visualization of the entire retina showed fewer CD45+ or F4/80+ cells (Fig. 5
A–D). This is more easily seen at higher magnification (Fig. 5E-H
). Image analysis confirmed that ischemic retinas from eyes treated with compound 3 had significantly fewer CD45+ cells (Fig. 5I
) or F4/80+ cells (Fig. 5J
) than retinas from corresponding contralateral eyes treated with PBS.
|
Blockade of CXCR4 suppresses ischemia-induced retinal NV
C57BL/6 mice were placed in 75% oxygen at P7 and returned to room air at P12. Between P12 and P17 the mice were given a daily periocular injection of 3 µl of 1.25 mM of Merck compound 3, a CXCR4 inhibitor, in one eye and PBS in the fellow eye. Eyes treated with the CXCR4 inhibitor had significantly less retinal NV on the surface of the retina (Fig. 6
).
|
Blockade of CXCR4 suppresses VEGF-induced subretinal NV in Rho/VEGF transgenic mice
Hemizygous Rho/VEGF transgenic mice were given daily periocular injections of 3 µl of 0.3 mM TC14012, another CXCR4 inhibitor, in one eye and 3 µl of PBS in the fellow eye between P7 and P21. At P21, the mice were perfused with fluorescein-labeled dextran, and retinal flat mounts, with the outer surface of the retina facing upward, were examined by fluorescence microscopy. There were fewer sprouts of NV on the outer surface of the retina from eyes that had been treated with the CXCR4 inhibitor (Fig. 7
A) compared to those that had been treated with PBS (Fig. 7B
). Measurements done by image analysis confirmed that inhibition of CXCR4 caused a significant reduction in subretinal NV (Fig. 7C
).
|
Periocular or intraocular injections of CXCR4 antagonists suppress CNV
We tested the effect of daily periocular injections of each of three CXCR4 antagonists in a model of CNV using the same doses described above. Daily 5 µl periocular injections of a 0.9 mM of AMD8664, 0.3 mM TC14012, or 1.25 mM Merck compound 3 each caused significant reductions in the area of CNV at Bruch’s membrane rupture sites (Fig. 8
A–C).
|
An intravitreous injection (1 µl) of one of each of the three CXCR4 antagonists (0.9 mM AMD8664, 0.3 mM TC14012, or 1.25 mM compound 3) was given on days 0 and 7 after rupture of Bruch’s membrane. Because the intraocular half-lives of the compounds were not known, the doses administered were intentionally high; at the time of injection, each dose was predicted to be > 2000-fold greater than the in vitro IC50 of the compounds. Each treatment resulted in significant reduction in the area of CNV at Bruch’s membrane rupture sites 14 days after laser. Compared with PBS-injected control eyes (Fig. 8E
), those given two injections of AMD8664 had a 60.44% reduction in the area of CNV, those injected with TC14012 had a 75.24% reduction, and those injected with Merck compound 3 (Fig. 8D
) had a 70.02% reduction (Fig. 8F
).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
35
36)
.
Pharmacologic blockade of CXCR4 suppressed ischemia-induced retinal NV, which occurs in diabetic retinopathy and other ischemic retinopathies. This indicates that increased signaling through CXCR4, along with signaling through VEGF receptors (21
, 37)
, contributes to ischemia-induced NV in the retina. Blockade of CXCR4 also blunted VEGF-induced retinal NV in the absence of ischemia. This suggests that CXCR4 signaling augments the angiogenic effects of VEGF in the retina. CNV is another type of ocular NV in which VEGF plays a central role (38
, 39)
, and CXCR4 antagonists were also very effective suppressors of CNV. Therefore, CXCR4 antagonists may provide a means to enhance the inhibitory effects of VEGF antagonists in treating several types of ocular NV.
It has been shown that increased expression of VEGF in liver or heart results in increased levels of SDF-1 and CXCR4 in those organs, and the increase in CXCR4 is due primarily to recruitment of BM-derived cells that express CXCR4 (40)
. The recruitment appears to be due to VEGF signaling through VEGF receptor 1 (34
35
36)
and the increase in SDF-1 that occurs in pericytes and smooth muscle cells functions to localize and retain the BM-derived cells adjacent to blood vessels, where they provide proangiogenic factors that work with VEGF produced by endogenous cells to stimulate NV. The situation in ischemic retina appears to be similar because there are increased levels of VEGF, SDF-1, and CXCR4, and a portion of the CXCR4 expression is due to the influx of BM-derived cells that localize adjacent to retinal vessels. Since retinal glia surround retinal blood vessels, the expression of SDF-1 by ischemic retinal glia may contribute to the localization of CXCR4-expressing, BM-derived cells around retinal blood vessels. The expression of CXCR4 as well as SDF-1 in retinal glia suggests that signaling through CXCR4 may occur in glial cells, but additional work is needed to determine its effects. However, it is clear that CXCR4 signaling, whether due solely to BM-derived cell recruitment or to effects in glial cells as well, contributes to ocular NV, because it is suppressed in several models by CXCR4 antagonists.
This study supports several earlier ones that have suggested that BM-derived cells contribute to ocular NV, although the exact nature of the contribution is still unclear. Studies utilizing lethally irradiated mice with reconstituted GFP-labeled BM cells have suggested that trans-differentiated pleuripotent cells from BM make up a large proportion of the endothelial cells in retinal and choroidal new vessels and a substantial percentage of other cell types within choroidal neovascular lesions (41
42
43)
. However, total body irradiation at a level that kills all hematopoietic cells has deleterious effects on other cells, particularly endothelial cells. Ocular exposure to even modest levels of stray radiotherapy meant for other tissues can cause radiation retinopathy in which retinal vessels shut down, resulting in nonperfused, ischemic retina (44)
. The endothelial cells of retinal vessels cannot divide and maintain the vessels; it would not be surprising if the majority of endothelial cells in new vessels complicating radiation retinopathy are trans-differentiated endothelial progenitor cells. However, even if this were proved experimentally, it would be hazardous to generalize and suggest that trans-differentiated cells make up a large proportion of new vessels of the retina or choroid in other disease processes. It is equally hazardous to ignore the potential effects of total body irradiation on retinal and choroidal endothelial cells, or other ocular cells types, and assume that percentages of trans-differentiated pleuripotent cells seen in that setting reflect what is present in other neovascular diseases. While it would not be surprising for some trans-differentiated endothelial cell progenitor cells to be incorporated into ocular NV, evidence other than that generated from models utilizing total body irradiation is needed before it is accepted that they constitute a large percentage of the endothelial cells in retinal or CNV. In the absence of such data, we suggest an alternative hypothesis that the major contribution of macrophages and other circulating cells is to increase the levels and alter the gradients of angiogenic factors, thereby contributing to maladaptive, disorganized vessel growth. Antagonists of VEGF signaling (through VEGF receptor 1) and CXCR4 signaling should help reduce this contribution to the pathological processes.
Ischemia-induced retinal neovascularization occurs in diabetic retinopathy and is the most prevalent cause of severe vision loss in working-age Americans (45
, 46)
. Our study supports previous studies that have implicated the SDF-1/CXCR4 pathway in diabetic retinopathy (47
, 48)
. Choroidal neovascularization occurs in age-related macular degeneration and is the major cause of severe vision loss in elderly Americans (49)
. Therefore, the demonstration that CXCR4 antagonists have beneficial effects in animal models for these two types of NV has important implications for a large segment of our population. These same models predicted that VEGF antagonists would provide benefit (21
, 37
38
39)
, and that prediction has been borne out. Monthly intraocular injections of Ranibizumab, an Fab fragment that binds all isoforms of VEGF-A, results in improvement of at least three lines of vision in about a third of patients with CNV due to AMD (50)
. This is the first treatment that has shown the ability to improve vision in patients with CNV, but there is room for further improvement. Clinical trials are needed to determine whether combined blockade of CXCR4 and VEGF provides added benefit. The demonstration that CXCR4 antagonists can be given by periocular injection, which is less invasive than intraocular injections, provides added impetus for such studies.
|
ACKNOWLEDGMENTS |
|---|
Received for publication September 25, 2006. Accepted for publication April 19, 2007.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  J. J. Hooks, C. N. Nagineni, L. C. Hooper, K. Hayashi, and B. Detrick IFN-{beta} Provides Immuno-Protection in the Retina by Inhibiting ICAM-1 and CXCL9 in Retinal Pigment Epithelial Cells J. Immunol., March 15, 2008; 180(6): 3789 - 3796. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
