| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
,¶
* Research Division and
Beetham Eye Institute, Joslin Diabetes Center, Harvard Medical School, Boston, Massachusetts, USA;
Departments of Preclinical Research and Cellular and
Analytical Development, XOMA (US) LLC, Berkeley, California, USA; and Departments of
|| Medicine and
¶ Ophthalmology, Harvard Medical School, Boston, Massachusetts, USA
1Correspondence: Joslin Diabetes Center, One Joslin Pl., Boston, MA 02215, USA. E-mail: george.king{at}joslin.harvard.edu
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: ERK • Akt • diabetic retinopathy • age-related macular degeneration • glypican 4
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, 2)
, is also expressed in epithelial cells in the trachea and the gut (3)
. Recently, BPI has been shown to mediate multiple functions in vascular cells, including inducing apoptosis in human umbilical vein-derived endothelial cells and inhibiting angiogenesis in chorioallantoic membrane (4)
. We previously reported that rBPI21, with bactericidal and LPS binding activity equivalent to or greater than the holoprotein (5
, 6)
, suppresses vascular endothelial growth factor (VEGF)-induced growth of bovine retinal endothelial cells (BREC) while, conversely, enhancing growth of bovine retinal pericytes (BRPC) (7)
. Similarly, a BPI-derived 1.4 kDa peptidemimetic promotes DNA synthesis in bovine retinal pigment epithelial cells (BRPE) (7)
.
Increased pericyte apoptosis and excessive neovascularization are classical pathological findings of diabetic retinopathy (8
, 9)
. Thus, BPI might have unique therapeutic potential for the treatment of diabetic retinopathy where promoting growth of retinal pericytes and simultaneously suppressing VEGF-induced growth of retinal endothelial cells would be beneficial. Likewise, in age-related macular degeneration (AMD), maintaining RPE viability while suppressing VEGF-induced angiogenesis would theoretically have synergistic benefits (10)
.
At the biochemical level, the molecular mechanism by which BPI signals cellular actions has not been reported. In this study, we have demonstrated the endogenous expression of BPI in nonepithelial and epithelial cells in the retina, identified cell type-specific signaling cascades and characterized the biological actions of BPI on retinal microvascular disorders, such as hypoxia-induced angiogenesis and VEGF- or diabetes-induced retinal permeability.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. BRPE were isolated by gentle scraping after removal of the neural retina and incubation with 0.2% collagenase (12)
. BREC were subsequently propagated with 10% horse serum (Life Technologies, Inc.), 50 mg/L heparin (Sigma, St. Louis, MO, USA), and 50 µg/ml endothelial cell growth factor (Roche Molecular Biochemicals, Indianapolis, IN, USA) and grown on collagen I-coated dishes (BD Biosciences, San Jose, CA, USA). BRPC were cultured in Dulbecco’s modified Eagle medium (DMEM) with 5.5 mM glucose (Glc) and 20% FBS (Life Technologies, Inc.) and BRPE in DMEM with 5.5 mM Glc and 10% FBS. Cells were cultured in 5% CO2 at 37°C, and media were changed every other day. BREC and BRPC were characterized for homogeneity by immunoreactivity with antifactor VIII antibody (Ab) and monoclonal antibody (mAb) 3G5, respectively (13)
. The authenticity of BRPE was verified by the presence of pigment granules on transmission electron microscopy (TEM) (12)
. BREC, from passages 2 through 7, and BRPC and BRPE, from passages 2 through 5, were used in these experiments. Cells remained morphologically unchanged under these conditions, as confirmed by light microscopy.
Reagents
rBPI21, a 21 kDa human recombinant modified, amino-terminal BPI fragment, and rBPI55, a 55 kDa human recombinant holo-BPI, were provided by XOMA (14)
. rBPI21 preparations have been tested in rabbit pyrogen assay, and no endotoxin was detected. These preparations of BPI’s have been used in clinical trials. Primary antibodies for immunoblotting included antiphospho-Akt (Cell Signaling, Danvers, MA, USA), Akt (Cell Signaling), phospho-p44/42 (Cell Signaling), and ERK (Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA) were obtained from commercial sources. Human VEGF165 and its antibodies were from R&D Systems (Minneapolis, MN, USA); streptozotocin (STZ) and thaumatin, from Sigma; sodium fluorescein (10%), from Akorn (Buffalo Grove, IL, USA); Evans-blue dye, from Sigma; sodium amobarbital (Eli Lilly, Indianapolis, IN, USA); sodium heparin (Elkins Sinn); tropicamide, (Alcon, Fort Worth, TX, USA); Affi-Gel 10 (Bio-Rad Laboratories, Hercules, CA, USA); and LY 294002 and PD 98059 (Calbio Chem, Inc.).
Reverse transcriptase-polymerase chain reaction (RT-PCR) analysis
RNA was extracted from frozen tissues and cultured cells with TRIzol, as described by the manufacturer (Invitrogen, Carlsbad, CA, USA). Approximately 1 µg RNA was used to generate cDNA by using reverse transcriptase and random hexamers (First-Strand cDNA Synthesis kit; Amersham Biosciences, Piscataway, NJ, USA) at 37°C for 60 min. Polymerase chain reaction (PCR) primers were: bovine BPI (Genbank accession no. X2563), 5'-TTCCCAGATAAGACCGTTGC-3' and 5'-GATTCAGACCCGCCAAAATA-3'; bovine 18S rRNA (GenBank accession no. AF176811), 5'-CGGCTACCACATCCAAGGAA-3' and 5'-GCTGGAATTACCGCGGCT-3'. PCR products were gel purified, subcloned using QIA quick PCR Purification kit (Qiagen, Valencia, CA, USA), and sequenced in both directions to confirm identity.
Immunohistochemistry
Paraffin-embedded sections (5 µm) were deparaffinized and rehydrated. Sections were incubated with 10% bovine serum albumin (BSA; Sigma), then with rabbit anti-human bactericidal permeability increasing protein polyclonal antibody (pAb) (1:1000; XOMA, Berkeley, CA, USA) overnight at 4°C. Control sections were incubated with rabbit immunoglobulin (Ig). Sections were incubated for 1 h at room temperature with biotinylated secondary Ab, followed by streptoavidin-horseradish peroxidase complex, developed with VIP (Vector VIP Substrate Kit For Peroxidase, Vector Laboratories, Burlingame, CA, USA) and counterstained with methyl green.
Collection of vitreous and plasma from patients in Joslin Diabetes Center
This study was approved by the Institutional Review Board, and informed consent was taken from each patient. Undiluted vitreous and plasma from 14 consecutive patients were collected and frozen at –80°C until assayed. BPI was measured using a capture ELISA with 0.055 ± 0.062 ng/ml detection limit (provided by XOMA).
Western blotting
Cells were stimulated with the compounds indicated after overnight starvation in 0.1% BSA for BRPE, BRPC, and BREC. Cells were lysed in 1x Laemmli buffer (50 mM Tris, pH 6.8; 2% SDS; and 10% glycerol) containing protease inhibitors (1 mM phenylmethylsulfonyl fluoride, 2 µg/ml aprotinin, 10 µg/ml leupeptin, 1 mM NaF, 0.5 mM Na3VO4). Samples were separated by SDS-PAGE, transferred to nitrocellulose membrane, and blocked with 5% skim milk. Antigens were detected using anti-rabbit horseradish peroxidase (HRP)-conjugated Ab for Western blotting and detected with the enhanced chemiluminescence (ECL) system (Amersham Biosciences).
DNA synthesis
The effect of rBPI21 on DNA synthesis was measured by incorporation of thymidine analog, bromodeoxyuridine (BrdU), using a commercial ELISA (Cell Proliferation ELISA BrdU calorimetric; Roche Diagnostics). Briefly, cells were plated at 104 cells/well on 96-well plates and incubated for 24 h. After serum starvation overnight, cells were incubated for 16 h in DMEM with 5.5 mM Glc and 0.1% BSA (Sigma) containing rBPI21 at different concentrations. BrdU (10 µM) was then added to the medium, and the cells incubated for an additional 2 h. The incorporation of BrdU was then terminated and quantified according to the manufacturer’s instructions.
DNA fragmentation analysis
DNA fragmentation was measured by quantitation of cytosolic oligonucleosome-bound DNA using ELISA (Roche Molecular Biochemicals), according to the manufacturer’s instructions. Briefly, cells were grown in 24-well plates at a density of 4 x 104 cells/well in 1 ml DMEM with 5.5 mM Glc and 5% FBS. After 10 h of incubation, cells were lysed directly on the plate. The cytosolic fraction was used as antigen source in a sandwich ELISA with primary antihistone Ab coated to the microtiter plate and a secondary anti-DNA Ab coupled to peroxidase. From the absorbance values, the fragmentation in comparison to controls was calculated according to the following formula: Ratio of control = 100 * (absorbance stimulated cells – absorbance blank)/(absorbance control cells – absorbance blank).
Murine retinal neovascularization model
This study adhered to the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. A reproducible model of hypoxia-induced neovascularization was used as described previously (15)
. Litters of 7-day-old [postnatal (P)7] C57BL/6J mice and their nursing mothers were exposed to 75% ± 2% oxygen for 5 d. At P12, the mice were returned to ambient air. Intraperitoneal injection of either vehicle alone (PBS) or rBPI21 (25 mg/kg or 50 mg/kg body wt) were administered every 24 h for 5 d (P12-P17), beginning immediately after their return to normoxic conditions.
Quantification of neovascularization
As described previously (16)
, mice at P17 were deeply anesthetized by 100 mg/kg pentobarbital sodium (Abbott Laboratories, Abbott Park, IL, USA) and sacrificed by cardiac perfusion of 4% paraformaldehyde in PBS. Eyes were enucleated and fixed in 4% paraformaldehyde overnight at 4°C before paraffin embedding. Over 50 serial 6 µm paraffin-embedded axial sections were obtained, starting at the optic nerve head. After staining with periodic acid-Schiff reagent and hematoxylin, 10 intact sections of equal length, each 30 µm apart, were evaluated for a span of 300 µm. All retinal vascular cell nuclei anterior to the internal limiting membrane were counted in each section by a fully masked protocol. The mean of all 10 counted sections yielded average neovascularization cell nuclei per 6 µm section per eye. No vascular cell nuclei anterior to the internal limiting membrane were observed in normal unmanipulated animals (17)
.
Rat preparation for intravitreal injections
Albino male Sprague-Dawley rats weighing between 200 and 250 g were used for all experiments. To induce diabetes, i.p. injection of STZ at 65 mg/kg body wt in 10 mM citrate buffer, pH 4.5, was performed after a 12-h fast. Diabetes was confirmed with blood Glc measurements 24 h after STZ injection. Blood Glc levels and body weights were monitored every other day. Blood Glc levels did not exceed 450 mg/dl in the diabetic animals, and all diabetic rats gained wt. Each animal, 24 h before the study, underwent surgical implantation of a polyvinyl catheter (i.d. 0.5 mm, o.d. 0.8 mm, length 20 cm) into the right jugular vein after anesthetization by i.p. injection of 0.1 mg/kg sodium pentobarbital. Immediately before the vitreous measurements, each rat was anesthetized as described above, and both eyes were dilated using 1% tropicamide. Intravitreal injections were performed through the limbus under direct visualization using a 32-gauge Hamilton needle and syringe, being careful to contact with the crystalline lens. A maximum vol representing < 10% of the vitreous vol was used per eye. Each eye was checked immediately after injection using a Rhodenstock scanning laser ophthalmoscope (Stamford, CT, USA) to be certain that no retinal or lenticular damage had occurred and the retinal perfusion was maintained.
Vitreous fluorophotometry protocol
Vitreous fluorophotometry was performed as described previously (19)
. Briefly, the fluorophotometry system was calibrated using a fluorescent glass standard, and incident laser power was adjusted to provide a fluorescent signal of 6000 photon counts per 2-s integration period (power
5 µW) before all experiments. The anesthetized and dilated rats were positioned such that the laser beam was incident on the cornea and the strong corneal reflex was directed away from the collection optics. After the baseline fluorescence measurement, 65 µl of 10% sodium fluorescein was injected to the rat via the externalized catheter port. Baseline measurements were made from each rat prior to all experiments to correct for intrinsic fluorescence and were subtracted from all subsequent vitreous fluorescence measurements.
Evans Blue permeation technique
The technique was performed as described by Qaum (19)
. Briefly, Evans blue was injected through the catheter port over 10 s at a dosage of 45 mg/kg. Subsequently, at 15-min intervals, 0.1 ml blood was drawn from the iliac artery for 2 h. to obtain the time-averaged plasma Evans blue concentration. Rats were perfused for 2 min via the left ventricle at 37°C with 0.05 M, pH 3.5, citrate-buffered paraformaldehyde (1% w/v). Immediately after perfusion, both eyes were enucleated and bisected away under an operating microscope and thoroughly dried in a Speed-Vac (Savant, St. Paul, MN, USA) for 5 h. The dry weight was used to normalize the quantitation of Evans blue leakage. Evans blue was extracted by incubating each retina in 240 µl formamide for 18 h at 70°C. The supernatant was filtered through Ultrafree-MC tubes (30,000 NMWL UFC3LTK00; Millipore, Bedford, MA, USA) at 3000 rpm for 2 h, and 60 µl of the filtrate was used for triplicate spectrophotometric measurements (Du-640; Beckman, Fullerton, CA, USA). The background-subtracted absorbance was determined by measuring each sample at 620 nm, the absorbance maximum for Evans blue in formamide, and 740 nm, the absorbance minimum. The concentration of dye in the extract was calculated from a standard curve of Evans blue in formamide. Retinal permeability was calculated by using the following equation, with results expressed in µl plasma·g retina dry wt–1·h–1.
 |
rBPI21 binding to VEGF by ELISA
Immobilon plates were coated with rBPI21 at 5, 10, and 20 µg/ml or thaumatin (20 µg/ml) in PBS at 4°C for 24 h. After 24 h, excessive rBPI21 was removed and blocked with 1% BSA in PBS-Tween 0.1% (PBS-T) at room temperature for 1 h. Plates were washed three times with excessive PBS-T and add VEGF (R&D Systems) 10 ng/ml in PBS, incubate at 37°C for 1 h. Plates washed as described previously, and anti-VEGF antibodies were added at 1:1000 dilution (0.5 µg/ml) in 1% BSA PBS-T at 37°C for 1 h. Plates were wash three times with PBS-T, and anti mouse-IgG-HRP was added (Amersham) at 1:1000 dilution in 1% BSA PBS-T and incubate at 37°C for 1 h. Then plates washed three times with PBS-T, developed with TMB substrate (Zymed Laboratories, Burlingame, CA, USA) and read at optical density (OD) 650.
Statistical analysis
Differences among groups with normal distribution and equal variance were analyzed using unpaired Student’s t tests. ANOVA with Tukey’s test was used for multiple comparisons of data with equal variance and normal distribution. P values < 0.05 were considered statistically significant.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
A). The PCR fragments derived from the retina, isolated RPE, and the cultured cells were confirmed to be BPI by direct nucleotide sequencing.
|
Localization of BPI expression was evaluated by using immunohistochemistry. BPI was expressed in the retinal vessel wall, most retinal layers, and the pigment epithelium (Fig. 1B
). In addition, staining also was observed in the choroidal vessel walls, including the endothelium and smooth muscle.
In human vitreous and plasma, BPI was detected in 3 out of 14 vitreous samples (0.374–0.759 µg/ml) and 10 out of 14 plasma samples (0.101–0.918 µg/ml). The three patients with positive BPI in the vitreous also showed positive BPI in the plasma.
Effect of rBPI21 on ERK phosphorylation in BRPE, BRPC, and BREC
rBPI21 and a 1.4 kDa peptidemimetic derived from BPI induce proliferation of BRPE and BRPC (7)
. Thus, we examined whether rBPI21 could induce ERK phosphorylation by Western blot analysis. In BRPE and BRPC, 7.5 µM of rBPI21 stimulated ERK phosphorylation within 5 min, reaching a maximum increases of 2.6 ± 0.4-fold (P<0.001) in BRPE and 4.8 ± 3.8-fold (P=0.028) in BRPC at 30 min (Fig. 2
A). The effect persisted for at least 60 min. ERK phosphorylation was induced by rBPI21 concentrations as low as 10 nM in BRPE (Fig. 2C
). rBPI55 (10–100 µM) was also able to induce similar levels of ERK phosphorylation as rBPI21. In contrast, rBPI21 did not stimulate ERK phosphorylation in BREC (Fig. 2A, B
). For controls, the effects of thaumatin (2.5 µM), which has similar molecular wt. (23 KD) and charge as rBPI21, and partially denatured BPI21 were studied in BRPE as shown in Fig. 2D
(20)
. Thaumatin was not able to induce ERK1/2 phosphorylation. Boiling rBPI21 at 2.5 µM for 90 min decreased its effects on ERK1/2 phosphorylation by 80% (Fig. 2D
).
|
Effects of rBPI21 on Akt phosphorylation
Because BPI accelerates apoptosis in human umbilical vein endothelial cells (HUVECs) but enhances growth in nonendothelial cells (4
, 7)
, we examined whether BPI could activate Akt preferentially in BRPE and BRPC as opposed to BREC. In BRPE and BRPC, 7.5 µM rBPI21 stimulated Akt phosphorylation within 5 min, reaching a maximum increase of 87 ± 31-fold (P< 0.001) at 30 min in BRPE and 37 ± 22-fold (P<0.001) in BRPC at 10 min (Fig. 3
A). This effect remained evident for at least 60 min. Akt phosphorylation was observed at rBPI21 concentrations as low as 1 µM (Fig. 3B, C
), and rBPI55 (7.5 µM) also induced similar lengths of Akt phosphorylation (data not shown). In contrast, rBPI21 did not stimulate any observable Akt phosphorylation in BREC (Fig. 3A, B
).
|
Effects of rBPI21 on DNA synthesis
Activation of ERK is associated with cellular proliferation (21)
. Because BPI activated ERK in BRPE and BRPC, we examined whether rBPI21 can increase DNA synthesis using BrdU incorporation assays. In BRPE, BrdU uptake was increased by 1.5-fold at rBPI21 concentration of 2.5 µM (P<0.001) and by 2.7-fold (P<0.001) at 7.5 µM (Fig. 4
A). Similarly, in BRPC, BrdU incorporation was increased by 1.5-fold (P<0.001) at 7.5 µM (Fig. 4A
). Thaumatin (7.5 µM) was not able to increase BrdU incorporation (Fig. 4B
) in BRPE. rBPI2’s effects on BrdU incorporation in BRPE were significantly decreased by the addition of inhibitors of P13K, LY 294002, and mitogen-activated protein kinase/ERK kinase (MEK)/mitogen-activated protein kinase (MAPK), PD98059, by more than 60% added separately (Fig. 4B
).
|
Inhibition of H2O2-induced apoptosis in BRPE by rBPI21
Since Akt activation has antiapoptotic actions (22
, 23)
, BPI may also have antiapoptotic properties in responsive cell types. Exposure of BPRE to 850 nM or 1000 nM hydrogen peroxide for 10 h increased apoptosis by more than 20-fold. In the presence of 2 µM rBPI21, apoptosis induced by 850 and 1000 nM of hydrogen peroxide was reduced by 75% (P=0.016) and 78% (P=0.007), respectively, as compared with cells incubated with hydrogen peroxide alone (Fig. 4C
).
Suppression of ischemia-induced retinal neovascularization in vitro and in vivo by rBPI21
VEGF, unlike BPI, stimulates ERK phosphorylation and growth in BREC (14)
. The effect of rBPI21 on VEGF-stimulated ERK phosphorylation in BREC was examined. rBPI21 alone (2.5 µM for 10 min) did not affect ERK phosphorylation, and VEGF alone (5 ng/ml for 10 min) induced ERK phosphorylation by 8-fold (Fig. 5
A, B). When VEGF and rBPI21 were preincubated together at 37°C for 15 min before addition to the cells, ERK phosphorylation in BREC was reduced by 50% as compared with VEGF stimulation alone (Fig. 5A, B
). However, when BREC were stimulated with VEGF, which had been incubated with 2.5 µM rBPI21 for 30 min, ERK phosphorylation was not altered compared with the phosphorylation induced by VEGF alone (data not shown). These data suggest that rBPI21 may partially suppress VEGF-induced ERK phosphorylation by sequestering VEGF rather than by competing with VEGF for binding to its receptors. To determine whether rBPI21 can bind directly to VEGF, we bound rBPI21 in a dose response manner at 5, 10, and 20 µg/ml to plates and added VEGF (10 ng/ml). As shown in Fig. 5C
, increasing amounts of VEGF were bound to rBPI21, but thaumatin (20 µg/ml) did not have any effects.
|
Since VEGF is a major mediator of retinal neovascularization (24
25
26)
, we examined whether rBPI21 could suppress retinal neovascularization in the murine model of ischemia-induced retinopathy (15)
. In this model, newborn mice were exposed to 75% oxygen on postnatal days 7–12 and then returned to room air. Retinal neovascularization occurs in 100% of animals by postnatal day 17 (15)
but does not occur in control animals. In preliminary studies, doses of rBPI21 in excess of 100 mg/kg/day for 5 d were well tolerated in newborn mice (XOMA, unpublished data, 2000).
RBPI21 was injected i.p. at 25 or 50 mg/kg every 24 h beginning when animals were returned to room air on day 12, the time when hypoxia of the retina initially occurs. Retinal neovascularization was evaluated at day 17 by histological examination of tissue cross sections (Fig. 6
A). Retinal neovascularization in mice treated with rBPI21 was reduced by 39% (P<0.001) and 42% (P<0.001) at BPI doses of 25 mg/kg and 50 mg/kg, respectively. (Fig. 6A, B
).
|
Suppression of VEGF-mediated and diabetes-induced retinal vascular permeability in vivo by rBPI21
Since our data suggest that rBPI21 may suppress VEGF-induced ERK phosphorylation by sequestering VEGF before binding to its receptors, we examined whether rBPI21 could suppress VEGF-mediated retinal vascular permeability (18)
. Equal vol injections of rBPI21 (0.047 and 3.6 µM final) were injected into the vitreous of rats 30 min prior to VEGF injections (0.5 nM final). Vehicle alone was administered to the contralateral eye, and vitreous fluorescein leakage was measured. Control rats demonstrated a 2.2-fold increase in vitreous fluorescein leakage after VEGF stimulation. rBPI21 at 0.047 and 3.6 µM reduced VEGF’s ability to stimulate vitreous fluorescein leakage by 48 ± 47% (P=0.036) and 93 ± 16% (P<0.001), respectively (Fig. 7
A).
|
Since intravitreal VEGF concentration is elevated in patients with diabetic retinopathy (25)
and our data suggest that rBPI21 may suppress VEGF-induced retinal vascular permeability, we evaluated whether rBPI21 could suppress diabetes-induced retinal vascular permeability. Retinal vascular permeability was evaluated following intravitreal or i.p. administration of rBPI21 in STZ-induced diabetic rats. Intravitreal injections of rBPI21 (0.047 µM, 0.95 µM, and 3.6 µM final) were performed on days 11 and 13 after onset of diabetes, while vehicle alone was administered to the contralateral eye. Retinal Evans-blue dye leakage was measured on day 14. Diabetic rats demonstrated a 2-fold increase in Evans-blue dye leakage compared with control rats. rBPI21 did not affect basal vitreous leakage; however, it reduced diabetes-induced leakage by 45 ± 55% (P=0.044) and 64 ± 47% (P=0.002) at 0.95 and 3.6 µM, respectively (Fig. 7B
).
To assess rBPI21 ability to reverse more established diabetes-induced abnormalities, diabetic rats with 2-week duration of disease were treated with i.p. injections of rBPI21 at 25 mg/kg every 24 h for 2 additional weeks, after which vitreous Evans-blue dye leakage was measured. Diabetic rats demonstrated a 1.8-fold increase in Evans-blue dye leakage compared with control rats. rBPI21 did not affect basal vitreous leakage; however, it reduced diabetes-induced leakage by 88 ± 53% (P=0.001) (Fig. 7C
).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, has recently been reported in other cell types, such as mucosal epithelial cells of diverse origin (oral, pulmonary, and gastrointestinal mucosa) and skin follicles (3
, 27)
. Our results demonstrate that mRNA of BPI is also expressed in several cell types in the retina, including RPE, pericytes, and REC. More detailed studies on the protein expression of BPI in the retina are needed to determine the specific cellular location of BPI in the retina. With the expression of BPI observed in cells that are not directly exposed to the external environment, it is likely that BPI may have biological functions in addition to its known bactericidal activities (1
, 2)
. In addition, we detected BPI in human vitreous and plasma at similar concentrations, supporting the idea that ocular tissues can express BPI. However, these results do not rule out the possibility that a nonocular source of BPI could be contributing to BPI detected in the vitreous. Future studies will be needed to clarify the correlation of BPI with diabetic retinopathy and other ocular pathologies.
In this study, we have characterized some of the mechanisms by which BPI mediates its diverse and paradoxical cellular actions in retinal endothelial cells, RPE, and pericytes. Previously, we reported that a 21 kDa modified amino-terminal fragment of BPI, rBPI21, suppressed VEGF-induced growth of endothelial cells, but in contrast, it enhanced DNA synthesis in retinal pericytes (7)
. Recently, another study showed that BPI could act as an angiogenic inhibitor where the addition of heparin neutralized the induction of apoptosis by BPI (4)
. Van der Schaft et al. speculated that heparin and/or heparan sulfate containing molecules could be the receptor for BPI, and the binding of heparin by BPI might compete for basic fibroblast growth factor (bFGF) and VEGF, which are stabilized by heparin. In the current study, we have shown that BPI can inhibit diabetes- and VEGF-induced retinal capillary permeability and hypoxia-induced angiogenesis. Yet BPI did not have any direct effect on retinal endothelial cells as measured by the activation of ERK or Akt. Interestingly, BPI was able to inhibit VEGF-induced MAPK activation in retinal endothelial cells only if it is preincubated with VEGF before adding to the cells. Direct and specific association of rBPI21 with VEGF was also demonstrated by the binding of these proteins to each other. The lack of binding of VEGF to thaumatin suggests the importance of BPI structure for its binding to VEGF, which may not be solely due to charge effects. This finding makes it likely that BPI is associating with VEGF through its heparin-binding domain in the amino-terminal region (14)
, which is the binding region of VEGF to its cellular receptors. It is also possible that the antipermeability effect of BPI in diabetic retinopathy is partly due to its ability to inhibit LPS-induced production of inflammatory cytokines secreted by circulatory cells, which also can increase vascular permeability systematically (5
, 28)
. This possibility is supported by the results from i.p. injection of rBPI21, which also inhibited retinal capillary permeability.
The ability of BPI to induce proliferation of RPE and pericytes appears to be mediated by mechanisms distinct from its antiangiogenic actions. This is the first demonstration that BPI can activate cellular signaling pathways directly. The mitogenic actions of BPI mimic those of cytokines or hormones, inducing the activation of Akt and ERK pathways. Both of these pathways are involved in cellular maintenance, DNA synthesis, and proliferation (22
, 29)
. In addition, the effect of BPI in the reduction of hydrogen peroxide-induced apoptosis in RPE is likely due to the activation of Akt cascades, a known pathway that inhibits apoptosis (22
, 23)
. The dose-response of BPI-stimulated ERK phosphorylation showed a significant effect at 10 nM in RPE, a dose much lower than that needed to stimulate Akt phosphorylation (0.5 µM). This difference in dose-response between ERK and Akt activation suggests that the effects of BPI in retinal cells may be more specific to mitogenic than metabolic activity (30)
. Interestingly, inhibitors of P13K and MAP Kinase pathways significantly reduced rBPI21’s actions on DNA synthesis in RPE, indicating that both pathways are involved.
It is unlikely that LPS contaminates in the BPI preparations are inducing Akt and ERK since we have added LPS at nM ranges that did not induce activation of Akt/MAPK or DNA synthesis in RPE. Further, assays for LPS in the rBPI21 preparations were not able to detect any endotoxins. The concentrations of BPI required to achieve maximum activity on RPE and pericytes are in the micromolar range, which suggests that the receptors or binding proteins for BPI may not be classical receptors for hormones or cytokines, but rather are more like low-affinity receptors or binding proteins for nutrients or cytoskeletal/matrix proteins (31
, 32)
.
Pathologically increased apoptosis in pericytes and excessive neovascularization are classical pathological findings in early diabetic retinal disease (8
, 9)
. Thus, these data suggest that BPI may have unique therapeutic potential for the treatment of diabetic retinopathy where promoting growth of retinal pericytes and simultaneously suppressing VEGF-induced growth of retinal endothelial cells would be dual benefits. Likewise, in macular degeneration, maintaining RPE viability while suppressing VEGF-induced angiogenesis would theoretically have synergistic benefits, because loss of RPE and choroidal neovascularization, often associated with increased vascular permeability, are pathological findings (10)
.
In this study, we have characterized numerous novel aspects of endogenous expression of BPI in nonepithelial and epithelial cells in the retina, cell type-specific signaling cascades, and biological actions on retinal microvascular disorders such as VEGF- or diabetes-induced retinal permeability and hypoxia-induced angiogenesis. The resulting dual trophic and antiangiogenic properties of BPI suggest that it may serve as a therapeutic model of diverse synergistic actions exerted by a single compound involving multiple types of cells.
|
ACKNOWLEDGMENTS |
|---|
Received for publication January 4, 2006. Accepted for publication May 25, 2006.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  A. Abedin, I. Mohammed, A. Hopkinson, and H. S. Dua A Novel Antimicrobial Peptide on the Ocular Surface Shows Decreased Expression in Inflammation and Infection Invest. Ophthalmol. Vis. Sci., January 1, 2008; 49(1): 28 - 33. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Geraldes, M. Yamagata, S. L. Rook, Y. Sassa, R. C. Ma, A. Clermont, B. Gao, L. P. Aiello, E. P. Feener, and G. L. King Glypican 4, a Membrane Binding Protein for Bactericidal/Permeability-Increasing Protein Signaling Pathways in Retinal Pigment Epithelial Cells Invest. Ophthalmol. Vis. Sci., December 1, 2007; 48(12): 5750 - 5755. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
