| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
,1
* Vascular Biology Program at Children’s Hospital Boston, Department of Surgery, and
Department of Ophthalmology, Harvard Medical School, Boston, Massachusetts, USA
2Correspondence: Vascular Biology Program, Children’s Hospital Boston, Harvard Medical School, 300 Longwood Ave., Boston, MA 02115, USA. E-mail: ofer.fainaru{at}childrens.harvard.edu
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: pulmonary edema • angiogenesis • delayed-type hypersensitivity
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. It was then found to be a powerful promoter of angiogenesis, which stimulates microvessel sprouting from existing blood vessels (2)
. These actions are complementary, as tumor angiogenesis and other forms of pathological angiogenesis are preceded and/or accompanied by enhanced vascular permeability (3)
. The integrity of the endothelium lining blood vessel walls serves as a barrier against vascular hyperpermeability, and its maintenance is critical for tissue homeostasis. Enhanced vascular permeability due to disruption, loss, or disorganization of the interendothelial junctions—a hallmark of inflammation that can be triggered by various inflammatory mediators [e.g., interleukin (IL) -2, lipopolysacharide] induced by different inciting events (e.g., pathogens, trauma, etc.) (4)
—is associated with intercellular leakage of large molecular solutes and associated fluid flux across the vascular endothelium, which can lead to tissue edema and organ failure. Conditions in which exaggerated vascular permeability leads to morbidity and mortality include pulmonary edema secondary to traumatic lung injury, stroke-induced cerebral edema, and various inflammatory and hypersensitivity disorders. Yet no therapies exist to prevent or restore junctional integrity in endothelium.
Tetracycylines, potent inhibitors of the matrix metalloproteinase (MMP) proteins, have been used to reduce tissue degradation in arthritis and periodontal disease (5)
. Doxycycline, a tetracycline derivative, has been shown to inhibit angiogenesis in both humans (6)
and animal models (7)
; however, its antiangiogenic effect is MMP independent in vitro (8)
. We now report that oral doxycycline prevents VPF/VEGF-induced vascular permeability, IL-2-induced pulmonary edema, and delayed-type hypersensitivity (DTH) in mice. We further show that doxycycline prevents tumor vessel hyperpermeability and suppresses tumor growth. Finally, we show that a possible mechanism for these effects is the specific targeting of adherens junctions by doxycycline, inducing the expression of VE-cadherin and decreasing its phosphorylation on the membranes of endothelial cells.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Mice
C57Bl/6J mice (6–8 wk) were purchased from Jackson Laboratories (Bar Harbor, ME, USA). All animal procedures were performed in compliance with Boston Children’s Hospital guidelines, and protocols were approved by the Institutional Animal Care and Use Committee.
Miles vascular permeability assay
C57Bl/6J mice were treated with oral doxycycline (intragastric gavage) at the specified doses or vehicle for 3–5 days before the Miles assay was performed (9
10
11)
. Of note, as it has been previously shown (12)
, a doxycycline dose of 100 mg/kg/day achieved a mean plasma concentration similar to plasma levels of human patients taking the recommended dose of 200 mg/day. We therefore used a similar dosing range in all our in vivo experiments. For the Miles assay, Evans blue dye (100 µl of a 1% solution in 0.9% NaCl) was injected intravenously into mice. Evan’s blue dye binds to plasma proteins and leaks with them at sites of vessel permeability. After 10 min, 50 µl of human VEGF165 (1 ng/µl), platelet activating factor (PAF; 100 µM), histamine (1.2 µg/ml), or PBS was injected intradermally into the preshaved back skin. After 20 min the animals were sacrificed, and an area of skin that included the entire injection site was removed. Evans blue dye was extracted from the skin by incubation with formamide for 5 days at room temperature, and the absorbance of extracted dye was measured at 620 nm.
Tumor vascular permeability
Mice were inoculated with subcutaneous 1 x 106 Lewis lung carcinoma cells. When tumors reached a volume of 
150 mm3, mice were treated with doxycycline (80 mg/kg/day) for 9 days. Evans blue dye was then injected intravenously, and dye was extracted from the tumors and assessed as above. The results were normalized to tumor weight.
DTH reactions
DTH reactions were induced in the ears of 8-wk-old C57Bl/6J male mice (n=5) as described previously (13)
. Mice were sensitized by topical application of 2% oxazolone solution in vehicle (acetone:olive oil, 4:1 v/v), to the shaved abdomen (50 µl). Mice were treated with oral doxycyline (80 mg/kg/day) for 5 days beginning on day 3, and after 5 days the right ears were challenged by topical application of 10 µl of a 1% oxazolone solution; the left ears were treated with vehicle alone. Ear thickness was then measured daily as a measure of inflammation intensity (14)
. Some mice from each experimental group were sacrificed 24 h after oxazolone challenge. Their ears were fixed in 10% formalin and processed for H&E-stained paraffin sections.
IL-2-associated pulmonary edema
Mice were pretreated with oral doxycycline (80 mg/kg/day) or vehicle for 5 days. On the sixth day, mice received an i.p. injection with IL-2 (1.2x106 U/100 µl) or saline 3x/day for 5 days. Doxycycline or vehicle treatment was continued through the course of IL-2 injections. At termination, mice were sacrificed and lungs were dissected, weighed, fixed, and processed for H&E staining.
Permeability of endothelial monolayers
Endothelial permeability was analyzed in vitro by the diffusion of 2000 kDa fluorescein isothiocyanate (FITC) -dextran (Sigma) through the endothelial monolayer (15)
. Human microvascular endothelial cells (HMVECs) were grown on Transwell inserts (Costar, Cambridge, MA, USA) up to confluence. The cells were pretreated with doxycycline (20 µM) for 16 h. Medium containing 2.5 mg/ml 2000 kDa FITC-dextran was then loaded in the upper compartment of the Transwell. The amount of FITC-dextran diffused through the endothelial monolayer into the lower compartment was measured by a microplate reader (Vactor3, PerkinElmer, Waltham, MA, USA).
Cell culture and confocal microscopy
Primary human dermal microvascular endothelial cells were grown in 131 complete medium (Cascade Biologics Inc., Portland, OR, USA), treated at passage 4–6, and used for further studies. In some experiments, cells were grown on coverslips precoated with collagen in 131 complete medium. After reaching confluence, the cells were starved overnight (0.5% BSA in 131 medium) with or without doxycycline (0–30 µM) and then treated with 50 ng/ml VEGF for 10 min. The concentration range of doxycycline used in all the in vitro experiments (0–30 µM) corresponds to human serum levels, which were reported to reach up to 160 µM (16)
.
The cells were fixed with 4% paraformaldehyde for 10 min, followed by incubation with 1% Triton X-100 for 10 min at room temperature and blocking with 3% BSA for 30 min at room temperature. The slips were incubated with the primary antibodies at 4°C overnight. After washing, the cells were incubated with appropriate fluorochrome-conjugated secondary antibodies. Coverslips were washed and mounted with Vectashield HardSet Mounting Media with DAPI (Vector Laboratories, Burlingame, CA, USA). Optical sections were scanned using a Leica TCS SP2 AOBS confocal system fitted to a DM IRE2 inverted microscope (Leica Microsystems, Wetzlar, Germany) with an x40 objective and 488 nm argon, 543 nm HeNe, 633 nm HeNe, and 405 nm diode lasers. Images were scanned sequentially to avoid fluorescence crossover, and z stacks were produced by scanning optical sections every 366 nm.
Western blots
Cells were lysed in RIPA buffer [50 mM Tris HCl (pH 7.5), 150 mM NaCl, 1% Nonidet P-40, 0.5% sodium deoxycholate, 0.1% SDS, 0.2 mM PMSF, and 2 mM sodium orthovanadate] supplemented with protease inhibitor cocktail (Sigma). Protein was quantified, and 40 µg was run on 10% PAGE, blotted, and incubated with primary antibodies. Pierce detection reagents were used to visualize horseradish peroxidase-labeled secondary antibodies. Relative band intensity was quantified using Image J 1.37 v software (U.S. National Institutes of Health, Bethesda, MD, USA).
Data analysis
Tumor volume (mm3) was calculated using the following formula: (tumor width)2 x tumor length x 0.52. Statistical comparisons of continuous data were performed using Student’s t test. Unpaired 2-tailed values of P < 0.05 were considered statistically significant. For multiple comparisons, the differences between groups were compared using analysis of variance (ANOVA); values of P < 0.05 were considered statistically significant. Subsequent pairwise comparisons between groups were performed using Tukey-Kramer HSD test. All statistical analyses were performed using Microsoft Excel datasheets (Microsoft, Redmond, WA, USA) and JMP 7.01 software (SAS Institute, Cary, NC, USA).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, 17)
to determine the effect of doxycycline on VEGF-induced vascular permeability (Fig. 1
). Doxycycline pretreatment decreased VEGF-induced extravasation of Evans Blue dye in a dose-dependent fashion (Fig. 1A, B
). Bevacizumab, an antibody that recognizes all isoforms of VEGF-A, is known to block vascular permeability (18)
. When mice were pretreated with oral doxycycline (80 mg/kg) for 5 days or with a single injection of i.p. Bevacizumab (100 µg) before carrying out the assay, both doxycycline and bevacizumab inhibited vascular permeability (2.4- and 2.1-fold, respectively) when compared to controls (Fig. 1C
). To determine the specificity of this effect, the relative efficacy of various tetracycline analogues was compared with respect to their ability to prevent VEGF-induced vascular hyperpermeability following i.p. administration. Remarkably, when compared to minocycline, chloro-tetracycline, and tetracycline, doxycycline was the most effective in preventing Evan’s blue dye leakage (Fig. 1D
).
|
Doxycycline decreases IL-2-induced pulmonary edema
Treatment of patients with metastatic melanoma and renal cell carcinoma by i.v. administration of IL-2 is often limited by the development of a capillary leak syndrome, which, in turn, may cause life-threatening systemic edema (19)
. To explore the possibility that doxycycline may prevent this potentially lethal complication, mice were pretreated with oral doxycycline (80 mg/kg/day) or vehicle for 4 days before administering i.p. IL-2 for 5 days in the continued presence of drug. IL-2-treated mice developed severe pulmonary edema, as demonstrated by a 2.75-fold increase in wet lung weight when compared to vehicle-treated controls (Fig. 2A, B
), confirming previously published results (17)
. Histological sections of the lungs from control mice revealed severe congestion and edema with intra-alveolar fibrin deposition, as well as perivascular and peribronchial mononuclear cell infiltrates (Fig. 2C, D
). Impressively, doxycycline almost completely inhibited the IL-2-induced increase in lung weight (Fig. 2A, B
) and prevented tissue edema (Fig. 2C
), without producing any evidence of systemic toxicity or weight loss.
|
Doxycycline decreases microvascular permeability in delayed type hypersensitivity
The DTH reaction is also characterized by enhanced vascular permeability and edema formation (20)
; thus, we tested whether doxycycline can inhibit this reaction in a mouse model of contact dermatitis. Mice were sensitized by applying the hapten oxazolone to their abdominal skin and then were treated either with oral doxycycline (80 mg/kg/day) or vehicle beginning on day 3 after sensitization. Six days after sensitization, we challenged the mice by application of oxazolone or vehicle to the right and left ears, respectively. Mice treated with doxycycline exhibited significantly reduced (P<0.05) erythema and ear swelling compared with vehicle-treated control mice at 24, 36, and 48 h (Fig. 3
). Histological analysis demonstrated similar inflammatory mononuclear cell infiltration into the oxazolone challenged ear in both doxycycline-treated and untreated mice (not shown).
|
Doxycycline inhibits tumor blood vessel permeability and tumor growth
We next tested whether doxycycline would also prevent tumor vessel hyperpermeability. Mice bearing subcutaneous implants of Lewis lung carcinoma 150 mm3 in volume were treated with either oral doxycycline (80 mg/kg/day) or vehicle for 8 days, and on day 9 Evans blue was injected intravenously. Treatment with doxycycline reduced Evans Blue leakage from the tumor vasculature when compared to vehicle-treated mice (Fig. 4A
). Remarkably, doxycycline treatment for 12 days also resulted in a decrease in tumor volume by 40% (Fig. 4B
).
|
Doxycycline inhibits permeability of endothelial cell monolayers
We next tested the effect of doxycycline on endothelial permeability in vitro by quantifying the diffusion of 2000 kDa FITC-dextran through endothelial monolayers. HMVECs were grown to confluence on Transwell inserts. The cells were incubated overnight with doxycycline (20 µM). FITC-dextran was then loaded in the upper compartment of the Transwell. The amount of FITC-dextran diffused through the endothelial monolayer into the lower compartment was measured after 90 min. As depicted in Fig. 5
, doxycycline pretreatment significantly inhibited vascular leakage when compared to control endothelial monolayers.
|
Doxycycline increases membrane expression of VE-cadherin and inhibits its phosphorylation in endothelial cells
We next sought to determine the mechanism by which doxycycline prevents vascular permeability in these mouse models. It is known that VEGF/VPF enhances endothelial cell permeability by promoting endocytosis of cell surface VE-cadherin (21)
. This effect disturbs the function of the adherens junction and leads to leakiness of the endothelial monolayer. When serum-starved confluent human dermal microvascular endothelial cell monolayers were cultured overnight in the presence of doxycycline (0–30 µM) before stimulation with VEGF (10 min, 50 ng/ml), VE-cadherin staining was found to increase as the doxycycline concentration was raised when analyzed using both confocal immunofluorescence microscopy (Fig. 6A
) and Western blots (Fig. 7A
). Remarkably, doxycycline significantly inhibited the VEGF-induced decrease in VE-cadherin expression (Fig. 7B
). This increase in VE-cadherin expression appeared to be a direct effect of doxycycline on endothelial cells, as a similar increase was observed in doxycycline-treated cells that were not starved or exposed to VEGF (Fig. 6B
). In contrast, neither the expression of β-catenin, another adherens junction protein that colocalizes with VE-cadherin (22)
, nor the tight junction protein ZO-1 (23)
was influenced by doxycycline treatment (Figs. 6B
and 7A)
. As expected (24)
, VEGF treatment led to phosphorylation of VE-cadherin (Fig. 7B
). Doxycycline pretreatment, however, reduced this phosphorylation. Taken together, these results suggest that doxycycline specifically targets the adherens junctions in endothelial cells by increasing membrane expression of VE-cadherin and inhibiting its phosphorylation.
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, 26)
. This increased permeability results in edema, clotting, and deposition of a provisional matrix that favors new vessel formation with mature stroma generation (27)
. Vascular hyperpermeability also contributes to the morbidity and mortality caused by fluid accumulation in body cavities (e.g., pleura and peritoneum) and organs (e.g., lung and brain) secondary to trauma or inflammation. Given that many angiogenesis inhibitors act by interfering with VEGF function, we tested the ability of the FDA-approved oral angiogenesis inhibitor, doxycycline, on vascular permeability in various mouse models of pathological angiogenesis and edema formation (6
, 7)
.
We first demonstrated that doxycycline inhibited VPF/VEGF-induced vascular permeability in a dose-dependent manner using the in vivo vascular permeability (Miles) assay. We confirmed this effect of doxycycline on the vascular endothelium by showing its inhibition of FITC-dextran leakage through endothelial cell monolayers in vitro. Remarkably, the effect of doxycycline was comparable to that achieved by bevacizumab, a powerful inhibitor of VEGF action (18)
, in the Miles assay. When compared to tetracycline and other related compounds (i.e., minocycline, chloro-tetracycline), doxycycline was the most effective at preventing Evan’s blue dye leakage. The effect of doxycycline was not specific for VEGF/VPF, as it also prevented vascular leak induced by histamine but not that induced by PAF (data not shown). PAF was reported to modulate vascular permeability through inducing nitric oxide synthesis (28)
, which affects the integrity of the tight junction (29)
rather than the adherens junction. Taken together with our immunostaining results, our findings suggest that doxycycline treatment specifically enhances the structural integrity of the adherens junction.
Of the molecular structures comprising the endothelial cell-cell contacts, the adherens junctions, composed of cadherins and catenins, are the primary adhesions between the cells and are essential for barrier integrity (22)
. At the endothelial adherens junction, the key transmembrane protein is VE-cadherin (30
, 31)
, which clusters together in these regions and mediates cell-cell adhesion through homophilic binding to other VE-cadherins expressed on adjacent endothelial cells. Paracellular permeability induced by inflammatory mediators, such as histamine and VEGF, is accompanied by phosphorylation of VE-cadherin (24)
, disruption of the VE-cadherin/catenin complex, and loss of cadherin from the cell borders (4
, 21)
. In fact, disruption of cadherins within the adherens junction mediates the increase in permeability and lung edema that are induced by inflammatory stimuli (32
, 33)
. Our results show that doxycycline treatment increases the expression of VE-cadherin at the intercellular junctions of human dermal endothelial cells in vitro without significantly altering the expression of β catenin or ZO-1. It has been suggested (34)
that the phosphorylation of VE-cadherin leads to a decrease in cell contact integrity. We now show that doxycycline inhibits VEGF-induced phosphorylation of VE-cadherin. Taken together, these effects on VE-cadherin expression and phosphorylation may explain the potential of doxycycline to protect against VEGF-induced vascular hypermeability in vivo, as observed in the Miles assay.
In light of the ability of doxycycline to prevent vascular hyperpermeability, we tested mouse models of clinical conditions where vascular leak serves a major source of morbidity. Immunotherapy with IL-2 represents an important modality in the management of human metatstatic renal cell carcinoma and malignant melanoma. However, this effective treatment is often limited by myriad complications, mainly due to a vascular leak syndrome resulting in pulmonary edema (35
, 36)
. Although it was shown previously (17)
that treatment with i.v. administration of the experimental antiangiogenic agent TNP-470 can prevent this complication in mice, the present results show that IL-2-induced pulmonary edema may be effectively prevented by oral administration of the FDA-approved drug doxycycline. Furthermore, we show that this drug effectively inhibits edema caused by delayed type hypersensitivity reactions, and thus it may also prove to be of value in treating allergic conditions in humans. As mentioned above, all forms of pathological angiogenesis, including tumor vessel formation, are preceded by vascular leak and deposition of fibrin in the matrix (3)
. Doxycycline also prevented the permeability of these vessels in mouse tumors. It is tempting to speculate that this aspect of antiangiogenesis may be responsible for the decrease in tumor size observed in the doxycycline-treated mice, however, the contribution of the reduction in tumor edema to tumor volume was not assessed.
Taken together, our results indicate that doxycycline may prove useful as a potent oral antivascular permeability drug. The mechanism for this effect appears to be the up-regulation of total VE-cadherin at the adherens junctions while decreasing the degree of its phosphorylation, thereby enhancing intercellular adhesion and improving the barrier function of the endothelium. The potential use of this drug in the treatment of human conditions associated with increased vascular permeability warrants clinical trials.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Received for publication March 19, 2008. Accepted for publication June 12, 2008.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  S.-U. Chen, C.-H. Chou, C.-W. Lin, H. Lee, J.-C. Wu, H.-F. Lu, C.-D. Chen, and Y.-S. Yang Signal mechanisms of vascular endothelial growth factor and interleukin-8 in ovarian hyperstimulation syndrome: dopamine targets their common pathways Hum. Reprod., March 1, 2010; 25(3): 757 - 767. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Albini, G. Pennesi, F. Donatelli, R. Cammarota, S. De Flora, and D. M. Noonan Cardiotoxicity of Anticancer Drugs: The Need for Cardio-Oncology and Cardio-Oncological Prevention J Natl Cancer Inst, January 6, 2010; 102(1): 14 - 25. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. K. Lucio-Eterovic, Y. Piao, and J. F. de Groot Mediators of Glioblastoma Resistance and Invasion during Antivascular Endothelial Growth Factor Therapy Clin. Cancer Res., July 15, 2009; 15(14): 4589 - 4599. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
µm) over the original ear thickness (mean±SD). Mice treated with doxycycline showed a significantly decreased ear swelling 24, 36, and 48 h after challenge when compared to controls. Control ears challenged with vehicle alone showed no swelling. *P < 0.05. Ox, oxazolone; dox, doxycycline.
