点击显示 收起
【摘要】
Basic fibroblast growth factor (bFGF, FGF-2) is a potent angiogenic factor and endothelial cell mitogen. Although bFGF levels are increased in chronically inflamed tissue, its role in inflammation is unclear. We investigated the effect of bFGF on acute dermal inflammation and the recruitment of monocytes, T cells, and neutrophils. Leukocyte recruitment to inflamed sites was quantified with radiolabeled leukocytes. Intradermal injection of bFGF in rats did not induce leukocyte recruitment or inflammation. However, the recruitment of leukocytes to inflammation induced by tumor necrosis factor-, interferon-, C5a, or a delayed hypersensitivity reaction was enhanced by bFGF by 55 to 132% (P < 0.05). Either acute or prolonged bFGF treatment of dermal sites had this effect. The potentiating effect of bFGF on leukocyte recruitment was also seen in joints. There was no associated modulation of vascular permeability, blood flow, or angiogenesis in the sites by bFGF. However, the expression of the endothelial cell adhesion molecules (CAMs) for leukocytes, P-selectin, E-selectin, and ICAM-1, was significantly up-regulated in the inflamed tissue by bFGF, as quantified by radiolabeled anti-CAM antibody binding in vivo. Thus, although not directly proinflammatory, bFGF synergistically potentiates inflammatory mediator-induced leukocyte recruitment, at least in part, by enhancing CAM up-regulation on endothelium.
--------------------------------------------------------------------------------
The interaction of leukocytes with vascular endothelial cells (ECs) and their emigration from the blood by transendothelial migration are essential components of the inflammatory reaction, defense against infections, and development of immunity. This process involves a cascade of events in which engagement between cell adhesion molecules (CAMs) on the surface of activated ECs and on the surface of leukocytes occurs. Leukocyte capture by and rolling on ECs are mediated by endothelial E- and P-selectin and 4 integrins while firm adhesion and migration of rolling leukocytes is primarily dependent on ß2 integrins on leukocytes binding to intercellular CAM-1 (ICAM-1) and ICAM-2 on ECs.1 The activation of ECs for expression of CAMs is mediated in part by inflammatory cytokines, such as tumor necrosis factor- (TNF-) and interferon- (IFN-). Chemotactic factors, such as interleukin (IL)-8, C5a, or MCP-1, are also involved in the recruitment of leukocytes to sites of inflammation.2
The inflammatory process is often accompanied by angiogenesis (formation of new blood vessels from pre-existing vasculature) and this is essential for tissue repair. Tumor growth is also dependent on angiogenesis. A variety of angiogenic factors are known to induce EC migration and proliferation. Basic fibroblast growth factor (bFGF, FGF-2) has been associated with tumor angiogenesis3,4 and tissue healing.5-7 bFGF binds to cell-surface tyrosine kinase receptors (FGFR), an interaction that is modulated by the presence of heparin-heparan sulfate8 and that increases the stability of the FGF-FGFR complex and its efficacy of signaling.9
The role of CAMs in leukocyte adhesion and emigration has been studied in normal microvessels. However, the mechanisms of leukocyte interaction with ECs in angiogenic vessels are poorly understood. Angiogenic vessels may have significantly altered interactions with blood leukocytes.10-12 For example, in tumor neovessels leukocyte interactions with the ECs are diminished.13-17 These findings have been related to diminished expression of EC-CAMs such as ICAM-1, ICAM-2, VCAM-1, and E-selectin in tumor vessels,18-20 perhaps due to inhibition of expression by bFGF and another angiogenic factor found in tumors, vascular endothelial growth factor.10,21 Recently, our laboratory demonstrated in vitro that long-term (ie, more than 24 hours) treatment of human umbilical vascular endothelial cells (HUVECs) with bFGF down-modulated the adhesion and transendothelial migration of monocytes and neutrophils (PMNs). Furthermore, endothelial expression of adhesion molecules and chemokines was also inhibited.22,23
It is well known that expression and secretion of angiogenic factors including bFGF is increased at sites of chronic inflammation. bFGF is increased in serum and affected tissue of patients with rheumatoid arthritis, inflammatory bowel disease, or asthma.24-27 However, it is not known whether the putative effect of such factors on EC and leukocyte recruitment is comparable in tumors and in inflammation. During inflammation, the presence of angiogenic factors has been associated with more intense accumulation of leukocytes and exacerbation of injury.28-30
The aim of this work was to elucidate the effects of bFGF on the recruitment of leukocytes during inflammation in normal tissue using skin as a model. Here, we provide in vivo evidence that bFGF synergistically enhances the recruitment of monocytes, T cells, and PMNs in response to a variety of inflammatory mediators. We also demonstrate that one of the mechanisms involved in the bFGF effect is increased expression of adhesion molecules on cytokine-activated ECs.
【关键词】 fibroblast potentiates leukocyte recruitment inflammation enhancing endothelial adhesion molecule expression
Materials and Methods
Animals
Inbred male Lewis rats (Charles River Canada, St. Constant, QC, Canada), weighing 250 to 300 g, were used. Some animals were used exclusively for harvesting of blood leukocytes for purification as previously described,31 whereas others were used for dermal inflammatory reactions and were recipients of labeled leukocytes. Animals in which a delayed type hypersensitivity (DTH) reaction was to be induced were immunized with 0.05 mg of killed Mycobacterium butyricum (Difco Laboratories, Inc., Detroit, MI) in 0.05 ml of mineral oil subcutaneously in two sites at the base of the tail.
Reagents
Human bFGF, rat TNF-, rat IFN-, and rat MIP-2 were purchased from Peprotech Inc. (Rocky Hill, NJ). Recombinant rat bFGF was a kind gift from E. Kardami and P. Cattini (the Department of Physiology and the Department of Anatomy, respectively, University of Manitoba, Winnipeg, MB, Canada). Heparin (molecular weight 3000 d, from porcine intestinal mucosa) was from Sigma Chemical Co. (St. Louis, MO). Zymosan-activated serum (ZAS) was generated by activating complement in normal rat serum (37??C, 60 minutes) with Zymosan A (Sigma), followed by removal of the zymosan by centrifugation as previously described.31 This ZAS was used as a source of the chemotactic factor C5ades-arg. Fetal calf serum (Medicorp, Montreal, QC, Canada) contained <0.0315 ng/ml of endotoxin. Human serum albumin (HSA), endotoxin- and pyrogen-free, was obtained from Connaught Laboratories (Downsview, ON, Canada). RPMI 1640 (Sigma) and all salts for buffers were made with analytical grade chemicals and dissolved in endotoxin- and pyrogen-free water (Abbott Laboratories, Ltd., St.-Laurent, QC, Canada). The following anti-CAM mouse IgG mAbs were used: RMP-1 and RP-2 (to rat P-selectin);32 RME-1 and 14G2 (to rat E-selectin);33 1A29, WT-3, and TA-2 (to rat ICAM-1, rat CD18, and rat VLA-4, respectively; kind gifts from M. Miyasaka, Osaka, Japan and T.B. Issekutz, Halifax, NS, Canada);34-36 and 5F10 (to VCAM-1; a kind gift from L. Burkly, Biogen Inc., Cambridge, MA).37
Rat Leukocyte Purification and Labeling
Rat blood PMNs and monocytes for migration studies were isolated by hydroxyethyl starch exchange transfusion technique as described previously.31 Briefly, via the femoral vein of an anesthetized heparinized donor rat, blood was gradually exchanged with 6% hydroxyethyl starch/saline (Hespan; American Hospital Supply Corp., McGaw Park, IL) into acid-citrate-dextrose (ACD, formula A; Fenwal-Travenol, Malton, ON, Canada) anti-coagulant, and red blood cells were sedimented at 1 x g. The leukocyte-rich plasma was harvested, and the leukocytes were pelleted (200 x g for 10 minutes) and resuspended in calcium- and magnesium-free Tyrode??s solution (TySC/C) containing 10% platelet-poor plasma (PPP). Leukocytes were layered on discontinuous Percoll gradients (63%/74%). Cells recovered at the interphase between 63% and 74% Percoll were >95% neutrophils. The PMNs were washed and labeled with 111In-labeled oxine (Amersham Corp., Oakville, ON, Canada) at 1.5 µCi/1 x 107 PMNs in 0.1 ml of TySC/C solution for 10 minutes at room temperature. Labeled PMNs were washed three times with and resuspended in TySC/C-10% PPP for intravenous injection.
Monocytes were purified as described,31 with minor modifications. Briefly, the mononuclear layer on the top of 63% Percoll was removed, washed, and resuspended in 3.6 ml of TySC/C-10% PPP, and the osmolarity of this suspension was slightly increased by the addition of 20 µl of 9% NaCl to improve the separation of monocytes from lymphocytes. After incubation for 10 minutes at 37??C, this cell suspension was layered onto another discontinuous Percoll gradient composed of 40% Percoll, on top of 55% Percoll, on top of 59% Percoll. All of the Percoll layers had 12.5 µl of 9% NaCl/2.5 ml of gradient. The gradient was centrifuged at 400 x g for 30 minutes at room temperature, and monocytes were harvested at the 40%/55% and the 55%/59% interphases. The purified monocytes (>85%) were washed and radiolabeled with 75 µCi of Na251CrO4 (Amersham) per 5 x 107 monocytes in 1 ml of TySC/C-10% PPP for 30 minutes at 37??C. Labeled monocytes were washed twice and resuspended in TySC/C-10% PPP for intravenous injection.
T lymphocytes were purified from rat spleen as previously described38 with minor modifications. Briefly, the spleen was minced in RPMI 1640 to release the cells. The cells were washed, resuspended in RPMI-10% fetal calf serum, and allowed to penetrate a sterile, pyrogen-free nylon wool column. The column was clamped and incubated 1 hour at 37??C. The purified T cells (>90%) were eluted with RPMI-fetal calf serum, centrifuged, and labeled with 111In as described for PMNs. Each rat received 5 x 106 51Cr-labeled monocytes together with 10 x 106 111In-labeled PMNs, or 20 x 106 111In-labeled T cells.
Vascular Permeability, Vascular Density, and Blood Flow Measurements
Exudation due to enhanced vascular permeability was quantitated using HSA labeled with 125I, as described.39 An aliquot of 125I-HSA (5 µCi/kg) was administered intravenously 2 hours before sacrifice. The content of 125I per site and of 125I per µl of blood plasma was quantified, and content in skin sites expressed as µl of plasma albumin equivalents/site.
Vascular density was measured by blood volume in the skin sites by determining the content of the intravascular marker 125I-IgG. This is a modification of the method using fluorescein isothiocyanate-labeled dextran (molecular weight 150 kd).40 A 5- to 10-µCi/kg aliquot of 125I-IgG was injected intravenously into each rat 15 minutes before sacrifice. The content of 125I per site was quantified and expressed as the fold increase over the control sites. Blood flow was measured as previously described41 with minor modifications. Briefly, just before sacrifice, 50 µCi of 86RbCl (Amersham) were injected intravenously. Forty-five seconds later, 1 ml of saturated KCl solution was injected intravenously to cause cardiac arrest. The content of 86Rb in the skin sites was quantified and expressed as fold increase over the control sites.
Growth Factor Pretreatments and Inflammatory Reactions
The dorsal skin of Lewis rats was shaved and dermal sites were injected intradermally in triplicate with 50 µl of 2, 10, 50, 100, or 200 ng of bFGF and 5 µg of heparin or heparin alone (control) in RPMI-HSA (1 mg/ml). Because bFGF and heparin did not induce inflammation, neither was considered an inflammatory stimulus. These injections were made at various stages: 48 hours before and at the time of injection of inflammatory stimuli (days C2 and 0), daily for 5 days (days C5 to 0), only 48 hours before (day C2), or only at the time of inflammatory stimulus (day 0). Inflammatory reactions were induced by intradermal injection (50 µl) of various inflammatory stimuli in triplicate sites in the dorsal skin in the bFGF and heparin or heparin (control)-pretreated sites. These included 3 ng of TNF-, 100 U of IFN-, 3 ng of TNF- and 100 U of IFN-, 10% ZAS as source of C5ades-Arg, 10 µg of killed M. butyricum and 20 ng of MIP-2, all in RPMI-HSA (0.1%). Inflammatory agents were injected at the time of intravenous injection of radiolabeled leukocytes. The animals were sacrificed 18 hours or 2 hours later as indicated.
In some experiments, the recruitment of leukocytes to inflamed sites was blocked by intravenous injection of anti-rat CD18 and anti-rat CD49d antibodies (WT-3, 1 mg/rat and TA-2, 0.5 mg/rat) immediately before induction of inflammation and injection of radiolabeled leukocytes. The absence of leukocyte recruitment after 2 hours was verified as reported previously.38,42,43 Where the effect of rat bFGF was compared with its human equivalent, dermal sites were injected intradermally in triplicate with 50 µl of 50 ng rat or human recombinant bFGF and 5 µg of heparin ?? 3 ng of TNF- and 100 U of IFN-. 111In-labeled PMNs were injected intravenously. The animals were sacrificed 2 hours later.
In some experiments, PMN recruitment to joints was investigated. Inflammatory reactions were induced in the ankles and wrists by injection in 20 µl of 100 ng of TNF- and 300 U of IFN- with 5 µg of heparin ?? 50 ng of human bFGF or rat bFGF in diluent (RPMI-HSA 1 mg/ml). Immediately thereafter, 111In-labeled PMNs (10 x 106) were injected intravenously. The animals were sacrificed 2 hours later.
Measurement of Leukocyte Accumulation
At the time of sacrifice, 2 ml of blood were collected in acid-citrate-dextrose anti-coagulant, and dorsal skin was removed and cleaned. The inflamed sites were punched out (12-mm diameter punch) and the joints were removed. Samples of spleen, liver, lung, and lymph nodes were taken for determination of 111In and 51Cr content. Counts of 111In and 51Cr in the tissues were expressed as cpm detected per 106 cpm injected. The amount of radioactivity (111In, 51Cr, 125I, 86Rb) was measured in a four-channel gamma spectrometer (LKB1282; Fisher Scientific Co., Ottawa, ON, Canada). Automatic corrections were made for the spill of isotope emissions into adjacent channels.
Histology in Dermal Inflammatory Sites
The dermal lesions were transected and fixed in 2.5% acetic acid, 2% formalin, and 75% ethanol (AFA fixative) for 24 hours at room temperature. Then, 1-mm slices were placed in 70% ethanol until paraffin embedding. Sections (5 µm thick) were cut, deparaffinized, and rehydrated by sequential passage in xylol and 100%, 95%, and 70% ethanol. The endogenous peroxidase was quenched with 3% H2O2-methanol 1:1 (30 minutes), and nonspecific binding of antibodies was blocked with Tris-buffered saline, 1% bovine serum albumin (Sigma), 10% normal goat serum, egg white avidin (1 µg/ml; Jackson ImmunoResearch Laboratories, Inc., West Grove, PA), 0.02% Tween 20 (Sigma) (20 minutes, 22??C). The slides were incubated with rabbit anti-mouse laminin antibody (Cedarlane Laboratories Ltd., Hornby, ON, Canada) or nonimmune rabbit IgG (Jackson ImmunoResearch Laboratories) overnight at 4??C. The secondary antibody was a biotin-SP-conjugated goat anti-rabbit IgG (Jackson ImmunoResearch Laboratories). The assay was developed with ABC-DAB kit (Vector Laboratories, Burlingame, CA) as suggested by the manufacturer. The slides were dehydrated and mounted in Permount (Fisher Scientific).
To visualize CAM expression, the dermal lesions were fixed in Carnoy??s fixative (60% methanol, 30% chloroform, 10% acetic acid) for 24 hours and then transferred to 80% ethanol until paraffin embedding. Sections (5 µm) were deparaffinized and rehydrated, and nonspecific binding of antibodies was blocked as above. The slides were incubated with primary mAb (eg, to rat E-selectin, to rat P-selectin, or IgG1 isotype control) overnight at 4??C. The secondary antibody, a biotin-conjugated goat anti-mouse IgG (Sigma), was incubated 1 hour at room temperature, and the binding was detected after washing with streptavidin-Alexa Fluor 568 (2 µg/ml, incubation 60 minutes; Molecular Probes, Inc., Eugene, OR) followed by washing, slide mounting, and examination with an immunofluorescence microscope.
For ICAM-1 visualization, skin sites were snap-frozen in liquid nitrogen-cooled isopentane and embedded in OCT (Tissue-Tek; Sakura Finetek U.S.A., Inc., Torrance, CA). Cryostat sections (8 µm) were air-dried and acetone-fixed for 5 minutes. After treatment with normal goat serum, the sections were incubated 1 hour with mouse anti-rat ICAM-1 mAb (1A29) or isotype control (10 µg/ml). The subsequent steps were performed under the same conditions as the ones used for E- and P-selectin staining.
Vessel Density Quantification
The skin sections were examined in a Nikon Eclipse E600 microscope (Nikon Canada Inc., Mississauga, ON, Canada) and analyzed with ACT-1 program in digital images. The number of laminin-stained vessels in a 0.8-mm2 area was counted, differentiating between small, medium, or large vessels. Vessels were considered small when their lumens were 1 leukocyte in diameter, medium when 2 to 3 in diameter and large when more than 3. Only vessels with defined borders and visualized in cross-section were counted.
Radiolabeling of mAb
The in vivo expression on ECs of E-selectin, P-selectin, VCAM-1, and ICAM-1 was quantified by the binding in the skin sites of intravenously administered 125I- or 131I-labeled antibodies as previously described.44 Briefly, 100 µg of mAb were incubated with 100 µCi of 125I or 131I in an Iodogen-precoated tube (Pierce Biotechnology, Inc., Rockford, IL) for 15 minutes at room temperature. The reaction was stopped by removing the sample from the tube. The free isotope was separated by dialyzing overnight against PBS and then diluting and reconcentrating the sample in a centrifugal filter device with a molecular-weight cutoff of 30,000 d (30,000 NMWL; Millipore Corp., Bedford, MA).
Quantification of Vascular Adhesion Molecule Expression with Radiolabeled mAb
Endothelial expression of E-selectin, P-selectin, VCAM-1, and ICAM-1 in inflamed dermal sites was determined by an intravenous injection of 125I-labeled RME-1 (anti-rat E-selectin), RMP-1 (anti-rat P-selectin), 1A29 (anti-rat ICAM-1), or 5F10 (anti-rat VCAM-1) mAb plus 131I-labeled control IgG (mAb 11D10 anti-human neutrophil, unpublished). Briefly, the animals were anesthetized with Ketamine-Innovar (SABEX Inc., Quebec, QC, Canada) subcutaneously, and 10 minutes later a mixture of 25 µg of 125I anti-CAM mAb plus 25 µg of control IgG was injected intravenously. Immediately thereafter, the animals received an intraperitoneal injection of 0.4 ml of xylazine. After 4 minutes the abdomen was opened, the ascending aorta cannulated, and 1 ml of 200 U heparin in saline was administered. The inferior vena cava was then opened, and the animals were perfused via the aorta with 75 ml of TyS+/+ buffer. The content of 125I and 131I in skin sites was quantified in a gamma counter along with a known fraction of the total injected dose of each labeled antibody. The expression of CAMs in test and control injected skin sites was calculated as the percentage of the 125I-injected dose (% ID) C the percentage of 131I ID of isotype control IgG C normal (noninjected) skin.
Statistical Analysis
Statistical significance was determined using paired t-test. P values exceeding 0.05 were considered not significant.
Results
Leukocyte Recruitment in Response to Inflammatory Mediators Is Up-Regulated by bFGF
Previously, we demonstrated in vitro that the pretreatment of ECs with bFGF for 1 and up to 3 days downmodulates EC activation for the expression of adhesion molecules and production of some chemokines, associated with a decrease in transendothelial migration of PMNs and monocytes.22,23 To study this action of bFGF in vivo, the migration of 51Cr-labeled blood monocytes and 111In-labeled spleen T cells or 111In-labeled PMNs to various inflammatory reactions in the dermis was investigated after pretreatment of the sites with bFGF for 48 hours. Sites on dorsal skin were pretreated by intradermal injection with 50 ng of bFGF and 5 µg of heparin or HSA and heparin as control 48 hours before and again at the time of the injection of inflammatory stimuli. The bFGF dose selected was based on dose-response experiments between 2 and 200 ng of bFGF as indicated in Materials and Methods. The maximum response was reached with 50 ng (163 ?? 12% increase in presence of TNF- and IFN-), with no further increase at doses of 100 and 200 ng of bFGF. However, even doses of 2 to 10 ng of bFGF induced increased leukocyte recruitment in response to TNF- and IFN- (data not shown). Heparin was used as a co-factor at the lowest concentration (5 µg), which avoided any effect on local hemostasis. The doses of inflammatory stimuli were selected to induce submaximal leukocyte accumulation (50% maximal for each stimulus). Figure 1 shows the recruitment of monocytes to various stimuli, normalized to 100% of the response to stimulus and HSA and heparin as the control. Pretreatment or repeated treatment of dermal sites with bFGF did not induce recruitment of monocytes (Figure 1) , T cells (Figure 2) , or PMNs (Figure 3) in the absence of inflammatory stimuli (ie, diluent alone). However, the recruitment of monocytes in response to IFN-, TNF-, TNF- and IFN-, C5a, and a DTH reaction was enhanced significantly by bFGF pretreatment, ranging from 155 to 212% of control (Figure 1) . The T-cell recruitment was also significantly increased in response to IFN- in the bFGF-pretreated sites relative to the control dermal sites (216% of control, P < 0.001). However, the T-cell migration in response to TNF- and IFN- or DTH reactions was not modified by bFGF pretreatment (Figure 2) .
Figure 1. Effect of pretreatment of dermal sites with bFGF on monocyte recruitment to inflammatory stimuli. Dorsal skin of Lewis rats was pretreated intradermally with 50 ng of bFGF and 5 µg of heparin or heparin alone. After 2 days, 51Cr-labeled monocytes were injected intravenously, and skin sites were reinjected with bFGF and heparin or heparin together with inflammatory stimuli (100 U of IFN-, 3 ng of TNF-, 10% of ZAS as source of C5a, 10 µg of M. butyricum for DTH reaction). After 18 hours the skin was removed, and 51Cr in the dermal sites was quantitated. Values are mean percentages ?? SEM of control (heparin-treated) sites, which are set as the 100% response. The mean absolute 51Cr-monocyte accumulation per lesion is given under each bar pair (n = 4 to 11). *P < 0.05, ***P < 0.005, paired t-test.
Figure 2. Effect of pretreatment of dermal sites with bFGF on T-cell recruitment to IFN-. Skin sites were pretreated for 2 days as in Figure 1 , and then 111In-labeled T lymphocytes were injected intravenously. At the same time, the sites were reinjected with bFGF and heparin or heparin alone plus inflammatory stimuli (100 U of IFN-, 3 ng of TNF-, 10 µg of M. butyricum for DTH reaction). After 18 hours 111In in the sites was quantitated. Values are mean percentage ?? SEM of control sites expressed as in Figure 1 (n = 4 to 8). ***P < 0.005, paired t-test.
Figure 3. Effect of pretreatment of dermal sites with bFGF on PMN recruitment to inflammatory stimuli. Skin sites were pretreated as in Figure 1 . After 2 days, 111In-labeled PMNs were injected intravenously, and the sites were reinjected with bFGF and heparin or heparin plus inflammatory stimuli (100 U of IFN-, 3 ng of TNF-, 10% ZAS for C5a, 20 ng of MIP-2). After 2 hours 111In in the sites was quantitated. Values are mean percentages ?? SEM of control sites expressed as in Figure 1 (n = 4). *P < 0.05, paired t-test.
We have shown that PMN recruitment is maximal within the initial 2 hours of induction of dermal inflammation31 ; therefore, a 2-hour PMN migration period was studied. A significant increase in the recruitment of PMNs to TNF-, TNF- and IFN-, and MIP-2 was observed in the sites pretreated with bFGF (232%, 195%, and 130% of control, respectively). In contrast, the PMN recruitment induced by C5a was not affected by bFGF pretreatment (Figure 3) .
Enhancement of Leukocyte Recruitment to Inflammatory Stimuli Induced by bFGF Is Rapid and Persistent
Because of the unexpected results above, we investigated whether the modulating effect of bFGF treatment was time-dependent. Therefore, 50 ng of bFGF and 5 µg of heparin was injected (intradermally) in dorsal skin sites in the following schedules: 1) daily for 5 days (days C5 to 0); 2) once, 48 hours before (day C2) the inflammatory stimulus; 3) twice, 48 hours before and at the time of inflammatory stimulus (days C2 and 0); or 4) only at the time of injection of inflammatory stimulus (day 0). Radiolabeled PMNs and monocytes were injected (day 0) and allowed to migrate for 2 hours. Figure 4, a and b , shows that daily pretreatment up to 5 days induced increased recruitment of monocytes (203 ?? 27% of control) and PMNs (171 ?? 16% of control) in response to IFN- and IFN- and TNF-, respectively. Co-injection of bFGF together with inflammatory mediator without pretreatment (day 0) induced a similar result. These levels of increase over the control were comparable to pretreatment at days C2 and 0 used in Figures 1 and 3 . However, pretreatment of sites once only 2 days before (day C2) the induction of inflammation did not affect leukocyte recruitment induced by IFN- or TNF- and IFN-. Similarly to PMN and monocyte recruitment, T-cell recruitment in response to IFN- was increased to 269 ?? 46% of control when bFGF was injected with IFN- but was not significantly different from control pretreatment when bFGF was administered at day C2 only (Figure 4c) .
Figure 4. Effect of acute versus chronic bFGF treatment of dermal sites on recruitment of leukocytes to inflammatory stimuli. Treatment of dermal sites with bFGF daily for 5 days (day C5 to 0); once, 48 hours before (day C2); twice, 48 hours before and at the time of inflammatory stimulus (days C2 and 0); or only at the time of injection of inflammatory stimulus (day 0) was compared. Inflammation with IFN- (a and c) or TNF- and IFN- (b) was induced 2 hours (a and b, for monocytes and neutrophils, respectively) or 18 hours (c, for T cells) before sacrifice. At the same time, 51Cr-monocytes and 111In-PMNs or 111In-T cells were injected intravenously. The 51Cr and 111In in the dermal sites were quantitated. Values are mean percentages ?? SEM of control sites (n = 4 to 10). *P < 0.05, **P < 0.01, ***P < 0.005, paired t-test.
Rat and Human bFGF Have Comparable Effect on Leukocyte Recruitment
Rat and human bFGF share 88% sequence identity. We investigated whether rat bFGF had the same effect as human bFGF on PMN recruitment in paired sets of experiments (n = 3). Rat bFGF enhanced the recruitment of PMNs in response to TNF- and IFN- by 132 ?? 51%, comparable to the increase induced by human bFGF (166 ?? 41%). This indicates that our observation with human bFGF is not due to species differences but rather to a biological effect on the inflammatory response.
The Synergistic Effect of bFGF on the Recruitment of Leukocytes Is Not Accompanied by Increased Vascular Permeability and Blood Flow
Increased vascular permeability and, as a consequence, extravasation of serum proteins are features of inflammation and can also enhance the rate of PMN emigration.41 To investigate the effect of bFGF on vascular permeability, 125I-HSA extravasation from plasma was measured in the bFGF-pretreated and control dermal sites. As shown in Table 1 , no difference in the extravastion of plasma albumin was observed with any of the inflammatory stimuli by bFGF or control (heparin) pretreatment.
Table 1. Parameters of Vascular Permeability, Hyperemia, and Angiogenesis in bFGF-Pretreated and Control Dermal Sites
Inflammatory hyperemia is known to enhance PMN recruitment induced by PMN recruiting mediators.45 Therefore, the blood flow was quantified with 86RbCl uptake. As shown in Table 1 , blood flow (expressed as fold increase) was increased in response to all inflammatory stimuli tested. However, there were no differences between bFGF-pretreated and control sites in normal or in inflamed tissue.
The Increased Recruitment of Leukocytes to Inflamed Dermal Sites by bFGF Is Not Associated with Angiogenesis
bFGF is a potent angiogenic factor, eg, in the corneal implantation or in the Matrigel angiogenesis assays.46,47 We therefore examined whether the increased leukocyte recruitment induced by bFGF, especially after 2 to 5 days of bFGF pretreatment, might be associated with angiogenesis. This was assessed by labeling the intravascular blood pool in the tissues with an injection (intravenous) of 125I-IgG40 and by immunohistochemistry for blood vessel density enumeration in the skin sites. As shown in Table 1 , there was no significant difference in the tissue vascular pool. Furthermore, the histological quantification of small, medium, and large vessels revealed no differences between control and bFGF-pretreated sites after either 2 or 5 consecutive days of bFGF treatment (not shown). Figure 5 shows vessel staining in representative bFGF-pretreated (Figure 5a) and control (Figure 5b) sites.
Figure 5. Histological assessment of bFGF treatment on dermal vessels. Dermal sites were treated daily for 5 days with bFGF (50 ng of bFGF and 5 µg of heparin) (a) or control (heparin alone) (b) and then paraffin-embedded, and 5-µm sections were stained for laminin to identify blood vessels. Antibody binding was detected indirectly with biotin-SP-conjugated goat anti-rabbit IgG and developed with a ABC-DAB kit. Original magnifications, x200.
The Enhanced Recruitment of PMNs to Inflammation Induced by bFGF Is Not Tissue-Dependent
We investigated whether the enhancement of PMN recruitment by bFGF was an effect observed only in dermal tissue or also in other tissues, such as joints. Inflammation was induced in ankle (talar) and wrist (carpal) joints by injection of 100 ng of TNF- and 300 U of IFN-?? 50 ng of bFGF. The recruitment of 111In-labeled PMNs was evaluated after 2 hours. As shown in Figure 6 , bFGF induced significantly increased PMN recruitment to these joints in response to inflammatory stimuli (138 to 165% of control, P < 0.05). A similar effect was observed with rat bFGF (135 to 140% of control, n = 4, P < 0.05).
Figure 6. Effect of bFGF on PMN recruitment to joints induced by TNF- and IFN-. Intra-articular injections (20 µl) of 50 ng of bFGF, 100 ng of TNF-, and 300 U of IFN- in diluent (RPMI-HSA-heparin) were administered, followed by intravenous injection of 111In-labeled PMNs. After 2 hours, the 111In in the joints was quantitated. Values are mean percentages ?? SEM of the response in the contralateral control joints injected with the above combination without bFGF (n = 24 to 32). ***P < 0.005, paired t-test.
Expression of CAMs in TNF-- and IFN--Inflamed Skin Is Enhanced by bFGF
To elucidate the mechanisms by which bFGF induces an increase in the recruitment of leukocytes in response to inflammatory mediators, we quantified the expression of adhesion molecules in the dermal sites as shown in Figure 7 . It was noted that diluent injection alone induced a slight but significant increase in P-selectin expression (P < 0.01) but not E-selectin, ICAM-1, or VCAM-1 when compared to noninjected skin (NS; not shown but values subtracted). Various lipopolysaccharide-free diluents were tested (Tyrode??s buffer, RPMI) with and without HSA without any effect on the small degree of P-selectin increase compared to noninjected skin. This P-selectin increase may reflect some degree of EC activation by, for example, mast cell degranulation secondary to the trauma of intradermal injection (even though only a 30-gauge needle was used) and/or platelet P-selectin associated with platelet deposition in a hemostatic plug. Injection of bFGF induced a significant increase in the expression of ICAM-1 and P-selectin in the absence of inflammatory cytokines (131% and 67% increase, respectively). TNF- and IFN- also significantly increased the expression of VCAM-1, P-selectin, and E-selectin relative to the diluent control. Co-injection of bFGF (and heparin) together with TNF- and IFN- enhanced ICAM-1 expression (P < 0.05) when compared to TNF- and IFN- (and heparin as control). Furthermore, bFGF also significantly increased P- and E-selectin expression in TNF- and IFN--stimulated sites. The expression of VCAM-1 showed an increasing tendency in bFGF- and TNF-- and IFN--stimulated sites, although due to variability, this did not reach significance (Figure 7) .
Figure 7. Effect of bFGF on adhesion molecule expression in inflamed dermal sites. The dorsal skin of Lewis rats was injected intradermally with 3 ng of TNF- and 100 U of IFN-?? 50 ng of bFGF and 5 µg of heparin, with diluent (RPMI-HSA-heparin) or with bFGF and heparin. After 2 hours, 25 µg of 125I-labeled anti-ICAM-1 mAb (1A29), anti-VCAM-1 mAb (5F10), anti-P-selectin mAb (RMP-1), or anti-E-selectin mAb (RME-1) and 25 µg of 131I-labeled isotype control IgG were injected intravenously and allowed to circulate 5 minutes. The animals were then sacrificed, the skin sites removed, and 125I and 131I quantitated. CAM expression is expressed as the mean ?? SEM of the specific anti-CAM mAb bound in the site . *P < 0.05, **P < 0.01, ***P < 0.001, paired t-test.
It was considered that the enhanced expression of CAMs observed in bFGF-co-treated sites during inflammation may be a secondary effect due to the increased leukocyte recruitment. To assess this potential mechanism, we blocked the emigration of all leukocyte types by intravenous injection of anti-CD18 and anti-CD49d antibodies as shown previously38,42,43 and measured P- and E-selectin expression in skin sites after TNF- and IFN- and bFGF injection. These experiments showed that the higher P- and E-selectin expression in bFGF-co-treated sites was not altered despite essentially complete blockade of PMN recruitment to the tissue, which was simultaneously measured (for the control rat group, TNF- and IFN- and bFGF sites, the specific CAM expression in net percentage x10C3 for E-selectin was 3.4 ?? 0.5, and 5.2 ?? 0.3 for P-selectin, and PMN accumulation was 2417 ?? 269 cpm/site; for the anti-CD18 and anti-CD49d mAb-treated group in TNF- and IFN- and bFGF sites, expression of E-selectin was 2.7 ?? 0.5 and of P-selectin was 4.6 ?? 0.2, and PMN accumulation was 35 ?? 4 cpm; n = 3). This indicates that bFGF-enhanced CAM expression in response to the inflammatory cytokines used is independent of the leukocyte infiltration.
Figure 8 shows representatives of 190 vessels examined by immunofluorescence staining for E-selectin, P-selectin, and ICAM-1 expression. Blinded observer assessment indicated an increase in the percentage of positive-staining postcapillary venules as well as an increase in the intensity of staining with anti-E- and anti-P-selectin antibodies in the sites treated with bFGF and TNF- and IFN- when compared with TNF- and IFN- alone, as seen also in the figure. ICAM-1 staining was strong under all conditions, and differences between bFGF-treated and nontreated inflamed sites were difficult to appreciate visually, although quantitation by radiolabeled anti-ICAM-1 mAb binding indicated a small but significant increase in ICAM-1 expression when bFGF was administered with TNF- and IFN- as indicated above.
Figure 8. Immunofluorescence detection of E-selectin, P-selectin, and ICAM-1 on dermal vessels. Inflamed dermal sites (TNF- and IFN- ?? bFGF) were paraffin-embedded, and 5-µm sections were stained for E-selectin (14G2 mAb), P-selectin (RP-2 mAb), or with an isotype control mAb. For ICAM-1 detection, 8-µm frozen sections were stained with 1A29 mAb or isotype control. A secondary goat anti-mouse IgG biotin-conjugated antibody and further development with streptavidin-Alexa Fluor 568 detected the primary antibody binding. Representative bright field and fluorescence fields in gray scale are shown. Original magnifications, x200.
Discussion
Infection or inflammatory cytokines have been reported to induce bFGF synthesis.48,49 bFGF is produced by several cell types, including ECs and fibroblasts.50-52 Released bFGF may be stored in the extracellular matrix by binding to heparan sulfate proteoglycans, and heparinases have a role in the release of bFGF from connective tissue stores.8,53 bFGF is one of the most potent stimulators of angiogenesis and is mitogenic and chemotactic for both ECs and fibroblasts. On ECs bFGF has been reported to down-regulate the expression of CAM and the interaction of leukocytes with ECs in vitro. Extending these observations in vivo, we systematically examined the effects of bFGF on the vascular responses and leukocyte recruitment during inflammation.
The first notable finding in this study is that bFGF (with or without heparin) did not induce leukocyte recruitment or any vascular response of inflammation (permeability or hyperemia) after either acute or repeated administration into the dermis. This finding is consistent with that of Yamashita and colleagues29 that transfecting bFGF in a recombinant Sendai virus vector delivered in rat ankle joints did not induce inflammation. Similarly, bFGF transgenic mice with increased endogenous levels of bFGF did not have spontaneous inflammation.28 Previous in vitro results in our laboratory demonstrated that bFGF treatment of HUVECs for 24 hours or longer downmodulated the transendothelial migration of monocytes and PMNs to a variety of stimuli including cytokines, suggesting that bFGF may have an anti-inflammatory role.22,23 This prompted us to investigate the effect of bFGF on various types of inflammation in vivo. Our findings indicated that bFGF, either acutely or on repeated application, enhanced the recruitment of monocytes in response to a variety of inflammatory mediators including cytokines, chemotactic factors, or to the DTH reaction (Figure 1) . The synergizing effect of bFGF was observed also in the recruitment of T cells to IFN- and acute PMN recruitment within the first 2 hours to TNF-, TNF- and IFN-, and to a lesser extent, MIP-2-induced inflammation (Figures 2 and 3) . The effect of bFGF observed here in dermal inflammation is not tissue-specific, because bFGF also acutely increased the recruitment of PMNs (within 2 hours) into normal joints in response to TNF- and IFN- (Figure 6) . Furthermore, rat bFGF had a similar enhancing effect to that of human bFGF on leukocyte recruitment to both skin and joint inflammation indicating that the observed effects are not unique to human bFGF in the rat. These results indicate that bFGF has a rapid and synergistic effect in heightening leukocyte responses to some inflammatory cytokines and that in vivo, even repeated or prolonged bFGF exposure of vascular endothelium and tissues does not alter this leukocyte recruitment-enhancing action. Our direct and controlled studies with respect to various stimuli, duration, and quantification of recruitment of various leukocyte types extends the in vivo findings of Yamashita and colleagues29 in rat adjuvant arthritis indicating that bFGF overexpression in arthritic joints aggravates inflammation. Similarly, Meij and colleagues28 observed exacerbation of inflammatory myocardial injury and T-cell infiltration in bFGF-overexpressing mice.
Inflammatory hyperemia and increase in vascular permeability are microvascular responses during the acute inflammatory reaction. These parameters can have a modulating effect on the rate of PMN recruitment.41,45 As shown in Table 1 , bFGF (and heparin) alone had no effect on these vascular responses. We used submaximal concentrations of inflammatory stimuli to allow detection of modulation of the responses by bFGF in either direction. As shown in Table 1 , these submaximal concentrations of inflammatory factors were sufficient to induce an increase in the blood flow, although the vascular permeability did not measurably change compared to nonstimulated sites with the exception of the most potent stimulus for PMN recruitment, namely C5ades-arg (ZAS). Our results demonstrate that the effect of bFGF on inflammatory stimulus-induced leukocyte recruitment is not due to modulation of the blood flow or vascular permeability responses.
As mentioned above, bFGF is a potent angiogenic factor that is increased in inflamed tissues, eg, in inflammatory bowel disease24 and in synovial fluid in severe rheumatoid arthritis.54 In these pathologies there is chronic inflammation and active angiogenesis. Although angiogenesis could affect the surface area for leukocyte emigration into inflamed tissue, our results show that bFGF did not induce angiogenesis in normal, uninflamed skin, even after daily treatment of dermal tissue for 5 days (Table 1 and text; Figure 5 ). Thus, the observed enhancement of leukocyte recruitment by prolonged, repeated treatment was not secondary to angiogenesis. Furthermore, the acute effect of bFGF in enhancing leukocyte recruitment within 2 hours of administration, when combined with the inflammatory stimulus, suggests that the action of bFGF is acute and independent of angiogenesis. Growth of new blood vessels in adult skin in the absence of injury or inflammation is not detectable. bFGF-induced angiogenesis in vivo has been reported in the chick chorioallantoic membrane assay, corneal implantation assay, and the Matrigel plug assay.55 The chick chorioallantoic membrane assay is an embryological and developmental system in which other angiogenic factors are also involved. In the Matrigel plug assay, an inflammatory reaction surrounding the plug does occur. In this assay we have observed that the bFGF used here induced angiogenesis in the plug in rats (unpublished observations), even though direct bFGF injection intradermally did not. In the corneal implant assay, some foreign body irritant and local trauma may generate inflammatory stimuli, which could potentiate the angiogenic action of bFGF. Thus, although bFGF is clearly angiogenic in pathological settings where other factors may contribute to this action, our findings indicate that bFGF has independent effects in enhancing inflammatory and immune cell recruitment, which chronically could lead to an angiogenic environment.
Our findings indicate that one mechanism for the acutely enhanced leukocyte recruitment induced by bFGF to inflammation is the up-regulation of CAM expression. Within 2 hours of TNF- and IFN- injection, the expression of ICAM-1, VCAM-1, E-selectin, and P-selectin increased over noninjected skin, and this was significantly enhanced by co-administration of bFGF for ICAM-1, P-selectin, and E-selectin (Figure 7) . This increased expression of E- and P-selectin induced by bFGF was confirmed by immunofluorescence (Figure 8) . This was not a secondary effect of the accumulated leukocytes stimulating enhanced CAM expression, because when the recruitment of PMNs was blocked by >98%, P- and E-selectin expression was still increased by bFGF treatment (see Results). We have observed similar degrees of inhibition of monocyte and lymphocyte recruitment to skin inflammation using anti-CD18 and anti-CD49d mAbs.38,42,43 Thus, the mechanism of bFGF potentiation of CAM expression in response to TNF- and IFN- is likely to be local at the level of the endothelium or perhaps secondary via an effect on other resident cells, which may co-stimulate the microvascular endothelium. Because P-selectin and E-selectin are critical adhesion molecules for the capture and rolling of leukocytes in the microvasculature and ICAM-1 mediates firm adhesion, the up-regulation of these CAMs associated with bFGF during inflammation may well result in more efficient leukocyte emigration into the surrounding tissues. Whether this is the only mechanism modulated by bFGF remains to be determined.
The literature reports differing effects of bFGF on EC-leukocyte interactions. In vitro and in vivo studies have shown a reduced expression of endothelial adhesion molecules such as ICAM-1, ICAM-2, and CD34.18,20,56,57 Results from our laboratory using HUVECs have shown that bFGF treatment for 18 hours transiently increased expression of ICAM-1 and VCAM-1, but not E-selectin, whereas treatment for more than 24 hours downmodulated the expression of ICAM-1, VCAM-1, and MCP-1 protein production.22,23 Griffioen and colleagues20 also demonstrated inhibition by 40 to 70% of CAM expression on HUVECs pretreated with bFGF for 3 days. These results are in concordance with observations in tumors,15,17 where the rolling and adherence of leukocytes to the angiogenic vasculature were found to be diminished. However, what role bFGF or other factors and the neovessels played in this observation was not determined. Our results in vivo with bFGF in normal, nonangiogenic tissue cannot be directly compared with these in vitro (HUVECs) and tumor-vessel studies because the ECs under these conditions are not quiescent but in growth phase. Tromp and colleagues58 observed inhibition of leukocyte adhesion to ECs in cremaster muscle vessels activated with IL-1ß for 3 hours after 3 days of exposure to slow-release pellets of bFGF. The state of the vessels analyzed, ie, normal or angiogenic, was not mentioned, but the bFGF pellets did induce angiogenesis in skinfold chambers. Furthermore, the surgical placement of alginate pellets in the scrotum may cause inflammation, which, after 3 days and under the exposure to bFGF, might affect the vascular responsiveness to IL-1ß. Thus, the tissue, stimulus and possibly the vascular state were different from that in our studies. Some of these differences could explain the contrasting results.
In summary, we demonstrate that bFGF has direct and acute effects on normal tissue to synergistically enhance recruitment of PMNs, monocytes, and T cells in response to inflammatory cytokines and DTH reactions independent of angiogenesis. Many conditions characterized by leukocyte infiltration are associated with both angiogenesis and the presence of bFGF in the tissue.24,54,59 Our results are the first to show that bFGF has acute effects in inflammation in vivo and that this is associated at least in part by potentiation of CAM up-regulation on ECs in vivo in response to inflammatory cytokines. Taken together with recent findings in arthritis and myocardial inflammation, bFGF may be an important positive regulator of leukocyte recruitment in acute and chronic inflammation.
Acknowledgements
We thank Ms. C. Jordan and Mr. D. Rowter for excellent technical assistance; Drs. M. Miyasaka, L. Burkly, and T.B. Issekutz for gifts of mAb; Drs. P. Cattini and E. Kardami for the rat bFGF; and the histology laboratory of the IWK Health Centre.
【参考文献】
Springer TA: Traffic signals on endothelium for lymphocyte recirculation and leukocyte emigration. Annu Rev Physiol 1995, 57:827-872
Johnston B, Butcher EC: Chemokines in rapid leukocyte adhesion triggering and migration. Semin Immunol 2002, 14:83-92
Fuhrmann-Benzakein E, Ma MN, Rubbia-Brandt L, Mentha G, Ruefenacht D, Sappino AP, Pepper MS: Elevated levels of angiogenic cytokines in the plasma of cancer patients. Int J Cancer 2000, 85:40-45
Meyer GE, Yu E, Siegal JA, Petteway JC, Blumenstein BA, Brawer MK: Serum basic fibroblast growth factor in men with and without prostate carcinoma. Cancer 1995, 76:2304-2311
Molloy T, Wang Y, Murrell G: The roles of growth factors in tendon and ligament healing. Sports Med 2003, 33:381-394
Tarnawski A, Szabo IL, Husain SS, Soreghan B: Regeneration of gastric mucosa during ulcer healing is triggered by growth factors and signal transduction pathways. J Physiol Paris 2001, 95:337-344
Ono I: The effects of basic fibroblast growth factor (bFGF) on the breaking strength of acute incisional wounds. J Dermatol Sci 2002, 29:104-113
Bikfalvi A, Klein S, Pintucci G, Rifkin DB: Biological roles of fibroblast growth factor-2. Endocr Rev 1997, 18:26-45
Vlodavsky I, Miao HQ, Medalion B, Danagher P, Ron D: Involvement of heparan sulfate and related molecules in sequestration and growth promoting activity of fibroblast growth factor. Cancer Metastasis Rev 1996, 15:177-186
Iruela-Arispe ML, Dvorak HF: Angiogenesis: a dynamic balance of stimulators and inhibitors. Thromb Haemost 1997, 78:672-677
Folkman J: Tumor angiogenesis and tissue factor. Nat Med 1996, 2:167-168
Folkman J: Angiogenesis in cancer, vascular, rheumatoid and other disease. Nat Med 1995, 1:27-31
Wu NZ, Klitzman B, Dodge R, Dewhirst MW: Diminished leukocyte-endothelium interaction in tumor microvessels. Cancer Res 1992, 52:4265-4268
Fukumura D, Salehi HA, Witwer B, Tuma RF, Melder RJ, Jain RK: Tumor necrosis factor alpha-induced leukocyte adhesion in normal and tumor vessels: effect of tumor type, transplantation site, and host strain. Cancer Res 1995, 55:4824-4829
Borgstrom P, Hughes GK, Hansell P, Wolitsky BA, Sriramarao P: Leukocyte adhesion in angiogenic blood vessels. Role of E-selectin, P-selectin, and beta2 integrin in lymphotoxin-mediated leukocyte recruitment in tumor microvessels. J Clin Invest 1997, 99:2246-2253
Jain RK, Koenig GC, Dellian M, Fukumura D, Munn LL, Melder RJ: Leukocyte-endothelial adhesion and angiogenesis in tumors. Cancer Metastasis Rev 1996, 15:195-204
Dirkx AE, Oude Egbrink MG, Kuijpers MJ, van der Niet ST, Heijnen VV, Bouma-ter Steege JC, Wagstaff J, Griffioen AW: Tumor angiogenesis modulates leukocyte-vessel wall interactions in vivo by reducing endothelial adhesion molecule expression. Cancer Res 2003, 63:2322-2329
Griffioen AW, Tromp SC, Hillen HF: Angiogenesis modulates the tumour immune response. Int J Exp Pathol 1998, 79:363-368
Melder RJ, Koenig GC, Witwer BP, Safabakhsh N, Munn LL, Jain RK: During angiogenesis, vascular endothelial growth factor and basic fibroblast growth factor regulate natural killer cell adhesion to tumor endothelium. Nat Med 1996, 2:992-997
Griffioen AW, Damen CA, Blijham GH, Groenewegen G: Tumor angiogenesis is accompanied by a decreased inflammatory response of tumor-associated endothelium. Blood 1996, 88:667-673
Griffioen AW, Molema G: Angiogenesis: potentials for pharmacologic intervention in the treatment of cancer, cardiovascular diseases, and chronic inflammation. Pharmacol Rev 2000, 52:237-268
Zhang H, Issekutz AC: Down-modulation of monocyte transendothelial migration and endothelial adhesion molecule expression by fibroblast growth factor: reversal by the anti-angiogenic agent SU6668. Am J Pathol 2002, 160:2219-2230
Zhang H, Issekutz AC: Growth factor regulation of neutrophil-endothelial cell interactions. J Leukoc Biol 2001, 70:225-232
Kanazawa S, Tsunoda T, Onuma E, Majima T, Kagiyama M, Kikuchi K: VEGF, basic-FGF, and TGF-beta in Crohn??s disease and ulcerative colitis: a novel mechanism of chronic intestinal inflammation. Am J Gastroenterol 2001, 96:822-828
Nagashima M, Yoshino S, Ishiwata T, Asano G: Role of vascular endothelial growth factor in angiogenesis of rheumatoid arthritis. J Rheumatol 1995, 22:1624-1630
Gudbjornsson B, Christofferson R, Larsson A: Synovial concentrations of the angiogenic peptides bFGF and VEGF do not discriminate rheumatoid arthritis from other forms of inflammatory arthritis. Scand J Clin Lab Invest 2004, 64:9-15
Redington AE, Roche WR, Madden J, Frew AJ, Djukanovic R, Holgate ST, Howarth PH: Basic fibroblast growth factor in asthma: measurement in bronchoalveolar lavage fluid basally and following allergen challenge. J Allergy Clin Immunol 2001, 107:384-387
Meij JT, Sheikh F, Jimenez SK, Nickerson PW, Kardami E, Cattini PA: Exacerbation of myocardial injury in transgenic mice overexpressing FGF-2 is T cell dependent. Am J Physiol 2002, 282:H547-H555
Yamashita A, Yonemitsu Y, Okano S, Nakagawa K, Nakashima Y, Irisa T, Iwamoto Y, Nagai Y, Hasegawa M, Sueishi K: Fibroblast growth factor-2 determines severity of joint disease in adjuvant-induced arthritis in rats. J Immunol 2002, 168:450-457
Reinders ME, Sho M, Izawa A, Wang P, Mukhopadhyay D, Koss KE, Geehan CS, Luster AD, Sayegh MH, Briscoe DM: Proinflammatory functions of vascular endothelial growth factor in alloimmunity. J Clin Invest 2003, 112:1655-1665
Issekutz AC, Issekutz TB: Quantitation and kinetics of blood monocyte migration to acute inflammatory reactions, and IL-1 alpha, tumor necrosis factor-alpha, and IFN-gamma. J Immunol 1993, 151:2105-2115
Walter UM, Ayer LM, Wolitzky BA, Wagner DD, Hynes RO, Manning AM, Issekutz AC: Characterization of a novel adhesion function blocking monoclonal antibody to rat/mouse P-selectin generated in the P-selectin-deficient mouse. Hybridoma 1997, 16:249-257
Walter UM, Ayer LM, Manning AM, Frenette PS, Wagner DD, Hynes RO, Wolitzky BA, Issekutz AC: Generation and characterization of a novel adhesion function blocking monoclonal antibody recognizing both rat and mouse E-selectin. Hybridoma 1997, 16:355-361
Tamatani T, Kotani M, Miyasaka M: Characterization of the rat leukocyte integrin, CD11/CD18, by the use of LFA-1 subunit-specific monoclonal antibodies. Eur J Immunol 1991, 21:627-633
Tamatani T, Miyasaka M: Identification of monoclonal antibodies reactive with the rat homolog of ICAM-1, and evidence for a differential involvement of ICAM-1 in the adherence of resting versus activated lymphocytes to high endothelial cells. Int Immunol 1990, 2:165-171
Issekutz TB, Wykretowicz A: Effect of a new monoclonal antibody, TA-2, that inhibits lymphocyte adherence to cytokine stimulated endothelium in the rat. J Immunol 1991, 147:109-116
May MJ, Entwistle G, Humphries MJ, Ager A: VCAM-1 is a CS1 peptide-inhibitable adhesion molecule expressed by lymph node high endothelium. J Cell Sci 1993, 106:109-119
Issekutz TB: Inhibition of lymphocyte endothelial adhesion and in vivo lymphocyte migration to cutaneous inflammation by TA-3, a new monoclonal antibody to rat LFA-1. J Immunol 1992, 149:3394-3402
Udaka K, Takeuchi Y, Movat HZ: Simple method for quantitation of enhanced vascular permeability. Proc Soc Exp Biol Med 1970, 133:1384-1387
Hoffmann J, Schirner M, Menrad A, Schneider MR: A highly sensitive model for quantification of in vivo tumor angiogenesis induced by alginate-encapsulated tumor cells. Cancer Res 1997, 57:3847-3851
Issekutz AC: Effect of vasoactive agents on polymorphonuclear leukocyte emigration in vivo. Lab Invest 1981, 45:234-240
Issekutz TB: In vivo blood monocyte migration to acute inflammatory reactions, IL-1 alpha, TNF-alpha, IFN-gamma, and C5a utilizes LFA-1, Mac-1, and VLA-4. The relative importance of each integrin. J Immunol 1995, 154:6533-6540
Issekutz TB, Miyasaka M, Issekutz AC: Rat blood neutrophils express very late antigen 4 and it mediates migration to arthritic joint and dermal inflammation. J Exp Med 1996, 183:2175-2184
Marshall D, Haskard DO: Quantifying inflammation in vivo using radiolabeled antibodies and leukocytes. Methods Mol Biol 2003, 225:273-282
Issekutz AC: Vascular responses during acute neutrophilic inflammation. Their relationship to in vivo neutrophil emigration. Lab Invest 1981, 45:435-441
Yang CF, Yasukawa T, Kimura H, Miyamoto H, Honda Y, Tabata Y, Ikada Y, Ogura Y: Experimental corneal neovascularization by basic fibroblast growth factor incorporated into gelatin hydrogel. Ophthalmic Res 2000, 32:19-24
Akhtar N, Dickerson EB, Auerbach R: The sponge/Matrigel angiogenesis assay. Angiogenesis 2002, 5:75-80
Prochnau D, Rodel J, Hartmann M, Straube E, Figulla HR: Growth factor production in human endothelial cells after Chlamydia pneumoniae infection. Int J Med Microbiol 2004, 294:53-57
Faris M, Ensoli B, Kokot N, Nel AE: Inflammatory cytokines induce the expression of basic fibroblast growth factor (bFGF) isoforms required for the growth of Kaposi??s sarcoma and endothelial cells through the activation of AP-1 response elements in the bFGF promoter. AIDS 1998, 12:19-27
Carroll LA, Koch RJ: Heparin stimulates production of bFGF and TGF-beta 1 by human normal, keloid, and fetal dermal fibroblasts. Med Sci Monit 2003, 9:BR97-BR108
Yoshida S, Yoshida A, Ishibashi T: Induction of IL-8, MCP-1, and bFGF by TNF-alpha in retinal glial cells: implications for retinal neovascularization during post-ischemic inflammation. Graefes Arch Clin Exp Ophthalmol 2004, 242:409-413
McNeil PL, Muthukrishnan L, Warder E, D??Amore PA: Growth factors are released by mechanically wounded endothelial cells. J Cell Biol 1989, 109:811-822
Chen CH, Poucher SM, Lu J, Henry PD: Fibroblast growth factor 2: from laboratory evidence to clinical application. Curr Vasc Pharmacol 2004, 2:33-43
Manabe N, Oda H, Nakamura K, Kuga Y, Uchida S, Kawaguchi H: Involvement of fibroblast growth factor-2 in joint destruction of rheumatoid arthritis patients. Rheumatology (Oxford) 1999, 38:714-720
Auerbach R, Lewis R, Shinners B, Kubai L, Akhtar N: Angiogenesis assays: a critical overview. Clin Chem 2003, 49:32-40
Griffioen AW, Damen CA, Martinotti S, Blijham GH, Groenewegen G: Endothelial intercellular adhesion molecule-1 expression is suppressed in human malignancies: the role of angiogenic factors. Cancer Res 1996, 56:1111-1117
Hellwig SM, Damen CA, van Adrichem NP, Blijham GH, Groenewegen G, Griffioen AW: Endothelial CD34 is suppressed in human malignancies: role of angiogenic factors. Cancer Lett 1997, 120:203-211
Tromp SC, oude Egbrink MG, Dings RP, van Velzen S, Slaaf DW, Hillen HF, Tangelder GJ, Reneman RS, Griffioen AW: Tumor angiogenesis factors reduce leukocyte adhesion in vivo. Int Immunol 2000, 12:671-676
Nagashima M, Asano G, Yoshino S: Imbalance in production between vascular endothelial growth factor and endostatin in patients with rheumatoid arthritis. J Rheumatol 2000, 27:2339-2342
作者单位：From the Departments of Pediatrics,* Pathology, and Microbiology-Immunology, Dalhousie University, Halifax, Nova Scotia, Canada