| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
* Angiogenesis/Laser Laboratory, Department of Ophthalmology, Massachusetts Eye and Ear Infirmary, Harvard Medical School, Boston, Massachusetts, USA; and
Department of Medical Oncology, Dana-Farber Cancer Institute, Harvard Medical School, Boston, Massachusetts, USA
1Correspondence: Angiogenesis/Laser Laboratory, Department of Ophthalmology, Massachusetts Eye and Ear Infirmary, Harvard Medical School, 243 Charles St., Boston, MA 02114, USA. E-mail: jwmiller{at}meei.harvard.edu
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, IL-1ß, and CD14 protein levels; NF-
B and HIF-1 activity; hsp90 and 70 levels and expression and phosphorylation of the tight junction proteins ZO-1 and occludin. 17-AAG treatment significantly suppressed the LPS-induced increase in retinal leukocyte adhesion; vascular leakage; NF-B, HIF-1, p38, and PI-3K activity; and VEGF, TNF-, and IL-1ß levels. 17-AAG also suppressed phosphorylation of ZO-1 and occludin by inhibiting their association with p38 and PI-3K. Although 17-AAG treatment did not reduce the LPS-induced increase in total CD14 levels in leukocytes, it significantly decreased membrane CD14 levels. These data suggest that Hsp90 inhibition suppresses several cardinal manifestations of endotoxin-induced uveitis in the rat. 17-AAG has demonstrated a favorable safety profile in clinical trials in cancer patients and represents a promising therapeutic agent for the treatment of inflammatory eye diseases.—Poulaki, V., Iliaki, E., Mitsiades, N., Mitsiades, C. S., Paulus, Y. M., Bula, D. V., Gragoudas, E. S., Miller, J. W. Inhibition of Hsp90 attenuates inflammation in endotoxin-induced uveitis.
Key Words: lipopolysaccharide • EIU • inflammatory eye disease
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. Endotoxin-induced uveitis (EIU) is an acute intraocular inflammatory condition that mimics human disease and is induced in an animal model by the systemic injection of sublethal doses of lipopolysaccharide (LPS). EIU has been used as a model for some types of uveitis in humans, including uveitis associated with seronegative arthritis, where Gram-negative bacteria may play a pathogenetic role (2
, 3)
. The inflammatory response of EIU peaks at 24 h and is characterized by leukocyte infiltration, breakdown of the blood-ocular barrier, and retinal cell death. Exposure to LPS stimulates cellular inflammatory responses and the release of cytokines and chemokines such as VEGF (4
, 5)
and IL-6 (6)
.
The exact molecular mechanisms underlying the pathogenesis of EIU have yet to be determined, although cytokines such as TNF-, IL-1ß, and IL-6 seem to play a major role (7
8
9)
. Recently, significant portions of the signal transduction pathway of LPS were elucidated. LPS, an endotoxin found in the outer membrane of Gram-negative bacteria, stimulates mononuclear cells and neutrophils to produce immunoregulatory and proinflammatory cytokines (10)
. Studies have identified three classes of receptor molecules involved in the recognition of LPS. Two classes of receptors—the CD18 antigens and scavenger receptor—mediate the disposal of LPS without initiating synthesis of proinflammatory cytokines (11)
. The myeloid differentiation antigen CD14, a 55 kDa membrane-bound glycoprotein (mCD14) expressed predominantly on the surface of monocytes, neutrophils, and macrophages, plays a major role in LPS-induced signal transduction (12)
. CD14 also exists as a soluble plasma protein (sCD14), and both forms participate in recognizing LPS and initiating cellular responses (12)
.
The first step in the molecular activation process is the binding of LPS and LPS binding protein to mCD14 on the cell surface of macrophages (13)
. Vascular endothelial cells do not express mCD14, but are stimulated by LPS via sCD14. Functional interaction of this ternary complex with Toll-like receptor 4 then leads to recruitment of the adapter protein MyD88, phosphorylation of the interleukin-1R (IL-1R) -associated kinase, and its association with tumor necrosis factor receptor (TNFR) -activated factor 6 (TRAF-6). Oligomerization of TRAF-6 activates a member of MAP3K family and IB kinase sequentially, which ultimately leads to the activation of nuclear factor-B and the induction of a wide variety of inflammatory and immune response genes that mediate the biological response of macrophages to LPS (13)
. Along with the CD14 pathway, other CD14-independent pathways have been described in the LPS signal transduction cascade (14)
.
The heat shock protein (Hsp) 90 is an intracellular chaperone that helps maintain, at the expense of ATP, the structural integrity of several membrane-, cytoplasmic-, and endoplasmic reticulum-associated proteins (15
16
17)
. Geldanamycin (GA), a benzoquinone ansamycin, and its analogs, such as 17-allylamino-17-demethoxy-geldanamycin (17-AAG), bind to and inactivate Hsp90, leading to increased degradation of its client proteins, suppressing their downstream signaling pathways, resulting in potent antiproliferative and anti-inflammatory activity in vivo (18
19
20
21
22)
. 17-AAG has been tested in preclinical and early clinical studies in a variety of solid tumors and exhibits favorable pharmacokinetic profiles with few side effects (23)
. Although early preclinical studies from our group investigated the antiangiogenic activity of the parent compound geldanamycin (24)
, in the present study we used 17-AAG, as this is the actual compound undergoing clinical development (23)
. We investigated the mode of action and the potential therapeutic effects of 17-AAG administration in a rat model of endotoxin-induced uveitis. Our study provides the rationale for therapeutic targeting of Hsp90 with 17-AAG to improve the clinical outcome of patients with uveitis.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Induction of EIU
Animals received a footpad injection of 200 µg/kg body weight LPS from Salmonella typhimurium (Sigma, St. Louis, MO, USA) in phosphate-buffered saline (PBS) or PBS alone. Two animals suffered systemic effects of the LPS injection such as hyperventilation or weakness; one died and had to be excluded from the study. None of the animals showed such an inflammatory reaction in the anterior chamber that would prevent visualization and analysis of the retina.
Treatment with 17-AAG
Thirty rats received an i.p. injection of either 2.5 mg/kg 17-AAG or the vehicle (DMSO) simultaneously with the footpad injection of LPS or PBS. This allowed a treatment group as well as three control groups (30 rats each): footpad PBS injection and DMSO i.p. injection; footpad PBS injection and 17-AAG i.p. injection; footpad LPS injection and DMSO i.p. injection. All harvesting of blood and tissues was performed 24 h after these injections.
Ex vivo quantification of retinal leukostasis 24 h after LPS
To measure leukostasis, we used a total of 24 rats (48 retinas): six rats each for the treatment and the three control groups. After the induction of deep anesthesia in the rat, the chest cavity was opened and a 14-gauge perfusion canula was introduced into the left ventricle. The right atrium was opened with a 12-gauge needle to achieve outflow. With the heart providing the motive force, 250 ml/kg PBS was administered from the perfusion canula to remove erythrocytes and nonadherent leukocytes. Fixation was then achieved by perfusion with 1% paraformaldehyde at a pressure of 100 mmHg. At this point, the heart stopped. A systemic blood pressure of 100 mmHg was maintained by perfusing a total volume of 200 ml/kg over 3 min, followed by perfusion with FITC-coupled concanavalin A lectin (20 µg/ml in PBS [pH 7.4], total concentration, 5 mg/kg body weight; Vector Laboratories, Burlingame, CA, USA). The latter stained adherent leukocytes and the vascular endothelium. Lectin staining was followed by PBS perfusion alone to remove excess concanavalin A. The retinas were flat mounted in a water-based fluorescence antifading medium (Fluoromount; Southern Biotechnology, Birmingham, AL, USA) and imaged by fluorescence microscopy (Axioplan, FITC filter, 40x; Carl Zeiss, Oberkochen, Germany). Leukocyte location was scored as being either arteriolar, venular, or capillary. The total number of adherent leukocytes per retina was counted. All experiments were performed in a masked fashion.
Measurement of retinal vascular leakage
For measurement of retinal vascular leakage, we used a total of 24 rats (48 retinas): six rats each for the treatment and control groups. Quantification of retinal vascular permeability was measured 24 h after the LPS injection, according to our previously published method (24
, 25)
. After the animals were deeply anesthetized, Evans blue dye (Sigma) dissolved in normal saline (30 mg/ml) was injected through the tail vein over 10 s at a dosage of 45 mg/kg. Just before perfusion at 2 h, blood samples were obtained from the left ventricle to obtain the Evans blue plasma concentration. These blood samples were centrifuged at 12,000 rpm for 15 min and the supernatant was diluted 1/10,000 in formamide (Sigma). The absorbance was measured with a spectrophotometer at 620 nm. The concentration of dye in the plasma was calculated from a standard curve of Evans blue in formamide. After the dye had circulated for 2 h, the chest cavity was opened and the rats were perfused through the left ventricle with citrate buffer (0.05 M, pH 3.5) for 2 min at a physiological pressure of 120 mmHg. The retinas were then carefully dissected under an operating microscope. After measuring of the retinal dry weight, Evans blue was extracted by incubating each retina in 0.3 ml of formamide for 18 h at 70°C. The extract was centrifuged through 10.000MW cut-off filters for 45 min at room temperature to remove the Evans blue bound to albumin. Sixty microliters of the supernatant was used for spectrophotometric measurement at 620 nm. Each measurement occurred over a 5 s interval, and all sets of measurements were preceded by an evaluation of known standards. 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 extracts was calculated from a standard curve of Evans blue in formamide. Blood-retinal barrier breakdown was calculated using the following equation, with results expressed in microliters of plasma per gram of retina (dry weight) x hours.
Evans blue (µg)/retina dry weight (g)/time-averaged Evans blue concentration (µg)/plasma (µl) circulation time (h)
Results were expressed as a percentage of the value in control animals.
Antibodies
The rabbit anti-rat ZO-1, the rabbit anti-rat occludin, and the anti-rat CD14 (Santa Cruz Biotechnology, Santa Cruz, CA, USA) were used for immunoblotting. The rabbit anti-PI3 kinase p85 antibody, the rabbit antiphospho-p38 MAP kinase (Thr180/Tyr 182), the mouse antiphospho-tyrosine, and rabbit anti-Hsp90 and anti-Hsp70 were purchased from Cell Signaling Inc. (Danvers, MA, USA).
Immunoblotting
For immunoblot detection of hsp70, hsp90, and phospho-p38, we used a total of 16 animals (32 retinas), four for each treatment and control group. Whole retinas were harvested 24 h after LPS administration and lysed for 30 min on ice in lysis buffer (50 mM Tris-HCl [pH 8], with 120 mM NaCl and 1% Nonidet P-40), supplemented with a mixture of proteinase inhibitors (Complete Mini; Roche Diagnostics, Basel, Switzerland). The samples were cleared by centrifugation (14,000 rpm for 30 min at 4°C). Protein concentration was assessed with the bicinchoninic acid protein assay (BCA, Pierce, Rockford, IL, USA). Thirty micrograms of protein per sample was electrophoresed in a 4–20% gradient sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) (Invitrogen Corporation, Carlsbad, CA, USA) and the proteins were electroblotted onto nylon membranes. After 1 h incubation in blocking solution (20% IgG-free normal horse serum in PBS/Tween [0.5% Tween 20 (Sigma) in PBS]), membranes were exposed to the primary antibody (1/500 dilution) overnight at 4°C. After washing in PBS supplemented with 1% Triton (Sigma), the peroxidase-labeled secondary antibodies (Amersham Pharmacia Biotech, Piscataway, NJ, USA) were added at a 1/10,000 dilution. The proteins were visualized with the enhanced chemiluminescence technique (Amersham Pharmacia Biotech).
For preparation of the cytoskeletal fractions (measurement of ZO-1 and occludin), we used a total of 16 rats (32 retinas), four for each of the treatment and control groups. In detail, the retinas were lysed for 30 min on ice in urea lysis buffer (6 mol/L urea, 0.1% Triton X-100, 10 mol/L Tris pH 8.0, 1 mmol/L dithiotreitol, 5 mmol/L MgCl2, 5 mmol/L EGTA, 150 mmol/L NaCl, 0.2 mmol/L PMSF). Samples were placed in a rotating platform for 15 min at 48°C to solubilize proteins, and insoluble material was precipitated by 10 min centrifugation at 14,000 rpm. Western blot was subsequently performed as described above.
For the immunoprecipitation experiments we used 24 rats (48 retinas), six rats from each category described above. The retinas were lysed as described in the immunoblotting experiments, and equal amounts of total protein lysate were incubated with an antibody against either p85, or ZO-1, or occludin or with the control immunoglobulin (IgG) at a dilution of 1/200 overnight at 4°C. The immunocomplexes were then incubated with 20 µl of packed protein G-agarose beads for 3 h at 4°C in a rocking platform, washed extensively and protein bound to Sepharose was eluted in loading buffer (60 mM Tris HCl pH 6.8, 2% SDS, 0.5% glycerol, 0.5% DTT, 0.002% bromphenol blue) by boiling for 2 min. Western blot was subsequently performed using the phospho-tyrosine antibody (measurement of activated PI-3 kinase, phosphorylated ZO-1 and occludin), the anti-PI3 kinase p85, or the antiphospho-p38 antibody as described above.
Detection of cell surface CD14 expression
Twenty-four hours after footpad injections of either LPS or PBS and i.p. injections of either DMSO or 17-AAG, the animals were anesthetized with 50 mg/kg pentobarbital and peripheral blood was obtained via heart puncture with a 16-gauge EDTA-flashed needle. In total, we used peripheral blood pooled from 16 rats, four for each treatment and control group. Leukocytes were isolated from whole blood by density gradient centrifugation with Histopaque 1083 (Sigma) according to the manufacturer’s instructions. The red blood cells were lysed with Red Blood Cell Lysing Buffer (Sigma). The preparations contained > 85% monocytes as determined by eosin and methylene blue staining (Leukostat Staining System, Fisher Scientific, Pittsburgh, PA, USA). Cell surface proteins were biotinylated by incubating in 0.5 mg/ml sulfo-NHS-LC-biotin (Pierce) in PBS for 30 min at room temperature. Because of its negative charge, sulfo-NHS-LC-Biotin does not cross the cell membrane, assuring that intracellular proteins are not biotinylated. Then the cells were washed three times in cold PBS, centrifuged briefly, and lysed for 30 min on ice in a lysis buffer (50 mM Tris-HCl, pH 8, containing 120 mM NaCl, 1% Igepal), supplemented with the Complete-TM mixture of proteinase inhibitors (Boehringer Mannheim, Mannheim, Germany). The samples were cleared by centrifugation (14000 rpm, 30 min, 4°C) and assessed for protein concentration. Biotinylated proteins, representing the cell surface proteins, were precipitated with streptavidin-agarose for 2 h at 4°C, electrophoresed in an SDS-PAGE, and CD14 levels were detected by Western blot as described.
Enzyme-linked immunosorbent assay for VEGF, TNF-, and IL-1ß
To measure VEGF, TNF-, and IL-1ß, we used 24 rats (48 retinas), six for each treatment and control group. Each retina was homogenized in 100 µl of solution consisting of 20 mM imidazole hydrochloride, 100 mM KCl, 1 mM MgCl2, 1 mM EGTA, 1% Triton, 10 mM NaF, 1 mM sodium molybdynate, and 1 mM EDTA supplemented with a cocktail of protease inhibitors (Complete, Roche) before use. Samples were cleared by centrifugation for 10 min at 13,000 rpm and assessed for protein concentration with the BCA assay (Mini BCA kit, Pierce). VEGF, TNF-, and IL-ß protein levels were estimated with the respective enzyme-linked immunosorbent assay (ELISA) kits (R&D Systems, Minneapolis, MN, USA) according to the manufacturer’s instructions. The reaction was stopped and absorption was measured in an ELISA reader at 450 nm. All measurements were performed in duplicate. The tissue sample concentration was calculated from a standard curve and corrected for protein concentration.
Preparation of nuclear extracts
For the measurement of NF-B and HIF-1a activity, we used 16 rats (32 retinas) in total, four rats in each of the same categories described above. Pooled retinas were homogenized with a mechanical homogenizer in five pellet volumes of buffer A (20 mM Tris pH 7.6, 10 mM KCL, 0.2 mM EDTA, 20% [by volume] glycerol, 1.5 mM MgCl2, 2 mM dithiothreitol [DTT]), 1 mM Na3VO4, and protease inhibitors (Complete, Boehringer Mannheim). Nuclei were pelleted (2500 g, 10 min) and resuspended in two pellet volumes of buffer B (identical to buffer A except that KCl was increased to 0.42 M). Nuclei debris were removed by centrifugation (15,000 g, 20 min) and the supernatant was dialyzed against one change of buffer Z (20 mM Tris-HCL [pH 7.8], 0.1 M KCL, 0.2 mM EDTA, 20% glycerol) for at least 3 h at 4°C with Dialyze Z cassettes (Pierce). Protein concentration was measured with the BCA assay.
Quantification of NF-B and HIF-1 activation
NF-B and HIF-1 activation was analyzed using respective Trans-AM Transcription Factor Assay Kits (Active Motif North America, Carlsbad, CA, USA) according to the manufacturer’s instructions. Briefly, 2 µg of the retinal nuclear extracts were incubated with an oligonucleotide containing the consensus binding element for each of the transcription factors analyzed bound to a 96-well plate. After extensive washes, the transcription factor complexes bound to the oligonucleotide were incubated with an antibody directed against the NF-B p65 subunit or HIF-1, respectively, at a dilution of 1:1000. After washes, the plates were subsequently incubated with a secondary antibody conjugated to horseradish peroxidase (1:1000) and the peroxidase reaction was quantified at 450 nm with a reference wavelength of 655 nm.
Statistical analysis
To evaluate the differences across various experimental conditions, one-way analysis of variance was performed and post hoc tests (Duncan and Dunnett’s T3 tests) served to evaluate differences between individual pairs of experimental conditions. In all analyses, P < 0.05 was considered statistically significant.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
A). We then investigated the effects of 17-AAG treatment on the levels of membranous CD14 with a modified technique of membrane labeling with streptavidin (26)
. Upon LPS challenge, the levels of membranous CD14 increased significantly. Treatment with 17-AAG significantly suppressed the LPS-induced increase in membranous CD14 levels (Fig. 1B
).
|
Hsp90 inhibition suppresses the LPS-induced increase in retinal NF-B and HIF-1 activity in vivo
The next step in the LPS-induced signal transduction pathway is activation of the transcription factors NF-B and HIF-1. We investigated the effect of 17-AAG treatment on the LPS-induced activation of NF-B and HIF-1 in vivo. Upon LPS administration, retinal NF-B and HIF-1 DNA binding activity increased compared with control (2.63±0.6-fold increase in NF-B and 1.7±0.48-fold increase in HIF-1 activity, P<0.0001 in both cases) (Fig. 2
).
|
17-AAG treatment significantly suppressed the up-regulation of both NF-B and HIF-1 levels in LPS-treated animals. Specifically, the NF-B activity increased only 1.56 ± 0.69-fold in animals treated with both 17-AAG and LPS (a 66.35% suppression of increase compared with animals treated with LPS only, P<0.01) (Fig. 2A
). Similarly, HIF-1 activity increased only 1.02 ± 0.3-fold in animals treated with both 17-AAG and LPS (a 65% suppression of increase compared to animals treated with LPS only, P<0.01) (Fig. 2B
).
Hsp90 inhibition suppresses the LPS-induced up-regulation of retinal VEGF, TNF-, and IL-1ß expression in vivo
Inflammatory cytokines that play a cardinal role in the pathogenesis of EIU and are dependent on the NF-B and HIF-1 activation include VEGF, TNF-, and IL-1ß. We investigated the effect of 17-AAG on the LPS-induced up-regulation of these three proteins. Upon LPS administration, VEGF expression increased 3.35 ± 1.73-fold, TNF- expression increased 5.3 ± 1.15-fold, and IL-1ß expression increased 2.7 ± 0.41-fold compared with controls (P<0.01 in all cases, Fig. 3
). Administration of 17-AAG reduced the expression of all three cytokines in the presence of LPS. VEGF levels rose only 1.58 ± 0.62-fold (a 77% suppression of increase), TNF- levels rose only 3.21 ± 1.97-fold (a 58% suppression of increase), and IL-1ß rose only 1.57 ± 0.405-fold (a 68% suppression of increase) (all P<0.01, Fig. 3
).
|
Hsp90 inhibition suppresses the LPS-induced increase in leukocyte adhesion in vivo
One of the main downstream actions of the inflammatory cytokines is the increase in leukocyte adhesion in the retina vasculature, which is also one of the main pathological manifestations of EIU. We investigated the effect of 17-AAG on leukocyte adhesion in the retinal vasculature in vivo. LPS administration significantly increased leukocyte adhesion compared with controls (Fig. 4
). Adhesion in veins increased 8.17 ± 1.77-fold, adhesion in arteries increased 2.64 ± 0.82-fold, and cumulative increase in arteries and veins increased 6.15 ± 1.37-fold compared with controls (all P<0.01, Fig. 4A-C
). Administration of 17-AAG significantly suppressed the LPS-induced increase in leukocyte adhesion. Adhesion rose only 3.6 ± 1.42-fold in veins (a 61% suppression of increase), 0.84 ± 0.4-fold in arteries (a 68.18% suppression of increase), and 2.7 ± 1.09-fold in arteries and veins combined (a 66.3% suppression of increase) (all P<0.01).
|
Hsp90 inhibition suppresses the LPS-induced increase in blood-retinal barrier breakdown
As we have shown before, leukocyte adhesion results in a blood-retinal barrier breakdown (27)
. Our finding that 17-AAG treatment decreases leukocyte adhesion in the retina in EIU prompted us to investigate this treatment and its effect on the blood-retinal barrier breakdown in vivo. LPS administration significantly increased blood-retinal barrier breakdown (3.86±1.03-fold increase compared with control animals, P<0.01) (Fig. 5
). Administration of 17-AAG significantly suppressed blood-retinal barrier breakdown in the presence of LPS: blood-retinal barrier breakdown increased only 2.5 ± 0.9-fold compared with the vehicle-treated animals (a 48% suppression of increase) (P<0.01) (Fig. 5)
.
|
Hsp90 inhibition attenuates the effect of LPS on retinal ZO-1 and occludin in vivo
Occludin and ZO-1 are tight junction proteins that play a major role in maintaining the blood-retinal barrier in vivo (28
, 29)
. We investigated the effect of LPS and 17-AAG on occludin and ZO-1 levels in vivo. LPS administration resulted in a significant decrease in retinal levels of (total) ZO-1 and occludin, whereas 17-AAG treatment attenuated this decrease (Fig. 6
A). Because tyrosine phosphorylation of ZO-1 and occludin plays an important role in their ubiquitination and subsequent degradation (30
31
32)
, we investigated the effect of 17-AAG on their phosphorylation status. LPS administration significantly increased the tyrosine phosphorylation of ZO-1 and occludin, whereas 17-AAG suppressed this phosphorylation (Fig. 6B
).
|
Hsp90 inhibition suppresses the association of p38/MAPK and PI3-kinase with ZO-1 and occludin in vivo
Occludin and ZO-1 are phosphorylated in vivo by tyrosine kinases, targeting them for proteasomal degradation. Because 17-AAG decreases the phosphorylation of occludin and ZO-1, we investigated the impact of 17-AAG on the interaction of p38/MAPK and PI-3 kinase, two tyrosine kinases, with ZO-1 and occludin in vivo. LPS administration stimulated the interaction of p38/MAPK with ZO-1 and occludin, an effect that was suppressed by 17-AAG (Fig. 7
A). Moreover, LPS administration stimulated the interaction of PI3-kinase with ZO-1 and occludin, an effect that was also suppressed by 17-AAG (Fig. 7B
).
|
Hsp90 inhibition suppresses the LPS-induced up-regulation of retinal PI-3 kinase and p38/MAPK phosphorylation in vivo
To determine whether the decreased association of occludin and ZO-1 with p38 MAPK and PI-3 kinase on the 17-AAG treatment is caused by a disruption of their association or by a decreased activation of these kinases, we measured the activation of p38MAPK and PI-3 kinase in vivo The PI-3 and p38 kinases have been found to participate in the LPS signal transduction pathway. Upon LPS administration, the phosphorylation of p38 MAPK and PI-3 kinase (p85) increased significantly with respect to non-LPS-treated controls (Fig. 8
). 17-AAG treatment attenuated this effect, but had no effect in the non-LPS-treated animals (Fig. 8)
.
|
Hsp90 inhibition induces up-regulation of retinal levels of Hsp90 and Hsp70 in vivo
The heat shock response is a powerful cellular defense mechanism against a variety of insults. Geldanamycin and its analogues have been shown to induce a heat shock response in a variety of models by increasing hsp70 and hsp90. We investigated the effect of 17-AAG on the retinal levels of the heat shock proteins 90 and 70 in vivo. 17-AAG treatment alone (without LPS challenge) increased Hsp90 and Hsp70 protein levels. LPS administration also increased both Hsp protein levels (Fig. 9
).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
and interleukins, which characterize EIU, and of proangiogenic factors, such as VEGF. We recently reported that GA, a benzoquinone ansamycin that inhibits Hsp90 activity, suppresses the expression of inflammatory transcription factors and cytokines in a rat model of diabetic retinopathy (24)
. We now demonstrate that the GA analog 17-AAG attenuates the cardinal manifestations of EIU by reducing the amount of leukocyte adhesion in the retinal vasculature and the blood-retinal barrier breakdown. Dissecting the molecular mechanisms of 17-AAG action, we found that it reduces the membranous expression of the LPS receptor CD14 in macrophages without affecting its total levels. 17-AAG also suppresses the activation of NF-B and HIF-1, and subsequently the expression of VEGF, TNF-, and IL-1ß. We also found that 17-AAG attenuates the LPS-induced activation of PI-3 kinase and p38 MAPK in the retina and the phosphorylation and degradation of critical proteins responsible for maintaining the blood-retinal barrier, such as ZO-1 and occludin.
The glycosylphosphatidilinositol (GPI) -anchored membrane receptor CD14 is a key molecule in innate bacterial recognition and the LPS signal transduction pathway. The CD14-LPS ligation takes place in membrane microdomains called lipid rafts, which concentrate the LPS "transducers" and their signaling machinery for a focused signaling effect (33)
. We found that 17-AAG reduces LPS-induced up-regulation of the CD14 membrane expression in vivo in the leukocytes in EIU without affecting its total levels. Our in vivo results are in accordance with recent in vitro findings (34)
. Vega et al. found that treatment of murine macrophages with GA (the parent compound of 17-AAG) reduced the surface levels of CD14 without affecting the mRNA levels or the total cellular pool of CD14. This effect involves accelerated internalization of CD14 from the plasma membrane and accumulation within the endoplasmic reticulum (ER) (34)
. GA also inhibits ER resident chaperone Grp94 and, as a result, disrupts the proper folding of proteins within the ER. It is possible that the effect observed in mCD14 expression is a result of the retention of CD14 in the ER as part of the protein "quality control" system that is present in this subcellular compartment (35)
. Another attractive possibility is that Hsp90 inhibition interferes with the GPI anchoring of CD14 that normally takes place inside the lipid raft while the molecule target is presented by the resident chaperone to the respective ER enzymes. It was found recently that Hsp90 is involved in the transport and fusion of vesicles with the plasma membrane (36)
. The inhibition of Hsp90 function by GA could interfere with the integration of CD14 to the plasma membrane and could lead to decreased cell surface mCD14 levels. The CD14 dependence of LPS-induced signaling responses provides a rational for blocking CD14 function to reduce the consequences of Gram-negative sepsis. Animal and human studies have shown that blocking CD14 inhibits almost completely the inflammatory response triggered by Gram-negative bacteria (37)
.
17-AAG specifically targets Hsp90 (and its endoplasmic reticulum homologue Grp94). However, as hsp90 interacts with a diverse group of client proteins, the effects of even specific Hsp90 inhibition are quite pleiotropic. As a result, 17-AAG has multiple effects on crucial elements of the LPS-induced signal transduction pathway and attenuates the activation of transcription factors such as NF-B and HIF-1. The transcription factor NF-B is a key signaling intermediate for LPS-induced responses of macrophages (38)
, and its activation is mediated through CD14-dependent and independent pathways. 17-AAG may reduce NF-B activation through both a direct effect and indirectly through its effect on CD14 (39
40
41
42)
. The role of HIF-1 in inflammation has recently begun to be identified, although the signal transduction pathway of the LPS-induced HIF-1 activation is not yet fully elucidated, with increased protein levels of HIF-1a found in rheumatoid synovia (43)
and HIF-1a-dependent leukocyte infiltration demonstrated in mouse models of rheumatoid arthritis (44)
. To our knowledge, this is the first report of HIF-1 activation in endotoxin-induced uveitis. Our data show that GA reduces LPS-induced HIF-1 activation; this effect can be explained by our earlier finding that GA reduces Akt and p42/44 MAPK activation, factors known to regulate HIF-1 expression (24)
.
The effect of Hsp90 inhibition on constituents of the LPS-initiated signal transduction pathway can explain its effect on the downstream expression of cytokines that are up-regulated after LPS challenge, including VEGF, TNF-, and IL-1ß. 17-AAG may affect transcription of the inflammatory cytokine genes by decreasing the activation of NF-B and HIF-1 response elements in their promoters (18)
. In addition, GA was found to have the ability to destabilize the inflammatory cytokine transcripts through inhibition of the p38 MAPK-dependent recruitment of trans-acting factors to adenine/uridine-rich elements in their 3'untranslated regions that prevent their degradation (45
, 46)
.
The 17-AAG-induced reduction of the expression levels of the inflammatory cytokines in our rat model of EIU can account for the decreased leukocyte adhesion and blood-retinal barrier breakdown we observed. As we have shown, leukocytes adhere to the retinal vasculature inducing endothelial cell dysfunction and apoptosis as well as disruption of the tight junctions and the blood-retinal barrier (27
, 47
, 48)
. In our model, LPS induced a dramatic decrease in the expression of the cytoskeleton-associated proteins occludin and ZO-1, which constitute parts of the protein complexes of the endothelial tight junctions that serve as guardians of the blood-retinal barrier. Occludin and ZO-1 are phosphorylated at tyrosine residues and targeted for degradation, as shown in various models (49
, 50)
. 17-AAG reduced the LPS-induced tyrosine phosphorylation and restored the above-mentioned tight junction proteins to almost normal levels, which helps explain its ability to stabilize the blood-retinal barrier and suppress vascular leakage. We have also demonstrated that LPS induces an association of PI-3 kinase with occludin and ZO-1, which is known in other models to target them for degradation, leading to loss of tight junction proteins; this association was disrupted by 17-AAG in our experiments.
LPS induced a heat shock response in our rat model of endotoxin-induced uveitis by increasing Hsp70 and Hsp90. We observed a similar increase in Hsp70 and Hsp90 with the administration of 17-AAG in our model. We (51)
and others (20
, 52
53
54)
have reported that Hsp90 inhibitors stimulate expression of several Hsps, including Hsp90 and Hsp70, in several models, possibly by promoting the stress-induced activation of HSF1. This probably represents a pleiotropic stress response, and has actually been shown to protect cells from apoptosis (53
, 54)
.
In conclusion, we have demonstrated that Hsp90 inhibition suppresses leukocyte adhesion and blood-retinal barrier breakdown in an animal model of endotoxin-induced uveitis through inhibitory effects on transcription factors, such as NF-B and HIF-1, and by reducing the expression of inflammatory cytokines such as VEGF, IL-1ß, and TNF-, while it stabilizes tight junction proteins such as occludin and ZO-1. Hsp90 inhibition also reduced the membranous expression of the LPS receptor CD14 in leukocytes. 17-AAG and other Hsp90 inhibitors are undergoing clinical evaluation as anticancer agents, exhibiting a favorable pharmacokinetic profile with few and manageable side effects (55
, 56)
. Therefore, Hsp90 inhibitors hold promise for the treatment of vision-threatening inflammatory conditions such as uveitis.
|
ACKNOWLEDGMENTS |
|---|
Received for publication January 25, 2007. Accepted for publication February 1, 2007.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
| |||||||||||||||||||||||
