Quantitation of the changes in vascularity during arthritis in the knee joint of a mouse with a digital image analysis system.код для вставкиСкачать
THE ANATOMICAL RECORD 262:420 – 428 (2001) Quantitation of the Changes in Vascularity During Arthritis in the Knee Joint of a Mouse With a Digital Image Analysis System PIETER BUMA,1* MARTIN GROENENBERG,1 PAUL F.J.W. RIJKEN,2 WIM B. VAN DEN BERG,3 LEO JOOSTEN,3 AND HANS PETERS2 1 Department of Orthopaedics, Orthopaedic Research Laboratory, University Medical Center, Nijmegen, The Netherlands 2 Department of Radiotherapy, University Medical Center, Nijmegen, The Netherlands 3 Department of Rheumatology, University Medical Center, Nijmegen, The Netherlands ABSTRACT Many joint and bone diseases are caused by, or associated with vascular changes. Particularly in rheumatoid arthritis, vascular sprouting of synovial vessels plays a major role in the generation of joint pathology. To assess the effects of pharmaceuticals that are designed to inhibit neovascularization, we developed a quantitative procedure to measure vascular changes in cross-sections of the mouse knee joint during arthritic inflammation. Arthritis was induced in the knee joint of C57Black6 mice by a single subpatellar injection of methylated BSA after previous immunization. Total vascularity was visualized with a specific monoclonal rat anti-mouse antibody (9F1). Functional vessels were detected with the fluorescent perfusion marker Hoechst 33342. The localization of Hoechst and the vascular marker 9F1 were analyzed in separate images with an automated digital image processing system. By combining the two images, total vascularity and the perfusion status of the vessels during arthritis could be established. The digital image system measures synovial area (SA), number of all blood vessels (NBV) and the number of perfused blood vessels (NpBV). From these parameters the percentage of perfused vessels (perfusion fraction; PF), the vessel density (VD ⫽ NBV/SA) and the density of perfused vessels (VDp ⫽ NpBV/SA) can be calculated. The measurements showed that the area of synovial tissue had increased during arthritis. Moreover, both the number of blood vessels (NBV) and the number of perfused vessels (NpBV) in the synovial area had increased significantly on Days 4 and 7 after arthritis induction. This procedure enabled quantitation of total vascularity and of functional blood vessels in cross-sections of synovial tissue. It is expected to be a powerful tool, not only to analyze the effects of anti-angiogenic therapies in animal models of arthritis, but could also be applicable to study vascular and perfusion changes in vascular related diseases of the skeleton. Anat Rec 262:420 – 428, 2001. © 2001 Wiley-Liss, Inc. Key words: joint pathology; rheumatoid arthritis; neovascularization; digital image processing; synovial tissue Many joint diseases are associated with vascular changes. For instance, changes in vascularity may play a role in the etiology of osteoarthritis, avascular necrosis and bone marrow edema. Particularly in rheumatoid arthritis, the synovium shows a chronic inflammatory reaction, with neovascularization, infiltration of lymphocytes, plasma cells and macrophages. Cells in this tissue that may even grow out as pannus tissue over the surface of the cartilage, produce catabolic factors for cartilage (Ben et al., 1995; Colville-Nash and Scott, 1992; Fava et al., 1994; © 2001 WILEY-LISS, INC. FitzGerald et al., 1991). The expansion of this hyperplastic synovial tissue is associated with an extensive prolif- *Correspondence to: Dr. P. Buma, Dept. Orthopaedics, Orthopaedic Research Laboratory, PO Box 9101, 6500 HB Nijmegen, The Netherlands. E-mail: P. Buma@orthp.azn.nl Received 17 July 2000; Accepted 19 October 2000 Published online 28 February 2001 DIGITAL IMAGE ANALYSIS OF MOUSE KNEE JOINT eration of blood vessels (Blake et al., 1989; Buma et al., 2000; Colville-Nash and Scott, 1992). The regulation of neovascularization in the synovium of arthritic joints is complex and various growth factors and inflammatory mediators may be involved (Ben et al., 1995; Blake et al., 1989; Claffey and Robinson 1996; FitzGerald et al., 1991; Nagashima et al., 1995; Stevens et al., 1991). Inflammatory cells that may contribute to the catabolic effect on cartilage cells caused by cytokine release can only reach the joint via the blood vessels. In addition to the well-known therapies for rheumatoid arthritis on the basis of fast-acting non-steroidal anti-inflammatory agents (NSAID) and long-term suppressive drug therapy with penicillamine, azathioprine, methotrexate and gold, new drugs have now been developed that specifically inhibit angiogenesis only (Oliver et al., 1995). In theory, the influx of inflammatory cells would be prevented or diminished if vascular proliferation could be inhibited, potentially reducing the disease activity of rheumatoid arthritis. To assess the primary effects of any new therapy strategy, it is essential to obtain quantitative information on the number of vessels, but also on the perfusion status of the vascular network in the synovial tissues during the arthritic process. Therefore, the aim of this study was to adapt a methodology that is already available for the assessment of vascular perfusion of tumour tissue in joints (Bernsen et al., 1999; Bussink et al., 1998, 1999; Rijken et al., 1995). This will allow quantitative assessment of the effects of drug therapies on time-specific changes in vascularity during arthritis. As a model system we used an arthritic model in the mouse (van Beuningen et al., 1989; van de Loo et al., 1995; van den Berg et al., 1981; van Lent et al., 1987). All vessels were visualized with immunofluorescent staining with a monoclonal marker 9F1, a specific marker for endothelial cells (Westphal et al., 1997). The fluorescent dye Hoechst 33342 was used as a marker for perfused vessels (Rijken et al., 1995; Smith et al., 1988). Hoechst is a small molecule that rapidly binds to the nucleus of all cells, irrespective of their location. Images of Hoechst and 9F1 staining were analyzed quantitatively by a scoring system that was programmed to combine the two images (Rijken et al., 1995). MATERIALS AND METHODS Experimental Animals Twenty-four adult female C57Black6 mice were used in this experiment. All mice were immunized by subcutaneous injection of an emulsion of 100 g methylated bovine serum albumin (mBSA) with complete Freund’s adjuvant in the foreleg and a subcutaneous injection of 100 g of this emulsion in the back of the mice (Kruijsen et al., 1983; van den Berg et al., 1981; van Beuningen et al., 1989; van de Loo et al., 1995). Additionally, an intraperitoneal injection of 1 ml of desiccated Bordetella pertussis was given. On Day 7, the same procedure was repeated as a boost. Then, after 21 days, arthritis was induced by a single intra-articular injection of 60 g mBSA in 6 l saline through the patellar ligament of the right knee joint. The left knee joint served as control in this study. Fixation Both on Days 4 and 7 after inducing arthritis, 12 animals were killed for routine histology (five animals in each 421 time point) and for vascular quantification (seven animals in each time point). For routine histology, joints were fixed by immersion in 0.1 M phosphate-buffered (pH 7.4) 4% paraformaldehyde, embedded in polymethylmethacrylate and sectioned (7 M) in the frontal plane. For vascular analysis animals were injected intravenously in one of the lateral tail veins with a 0.2 ml solution of phosphatebuffered saline (PBS; pH ⫽ 7.4) containing 0.3 mg Hoechst 33342 (Sigma Chemical Company, St. Louis, MO). One minute after injection with Hoechst, the mice were killed by cervical dislocation and both knee joints were quickly removed (in circa 10 sec), embedded in Tissue Tek Optimal Cutting Temperature (OCT; Miles Diagnostics Division, Elkhart, IN) and snap frozen in isopentane (⫺80°C). This prevented the Hoechst dye from diffusing too far into the tissues. Histological Procedures The knee joints were sectioned (see Fig. 1 for schematic representation), without previous decalcification in the frontal plane into 7 m sections at ⫺22°C on a Bright 3050 cryostat with a tungsten edged knife and mounted on slides precoated with aminopropyltrietoxysilaan (Sigma). The sections were air dried for 1 hr and stored at ⫺80°C until required. In each knee, three sections were analyzed. The middle section was the most central section containing synovial membrane tissue. The other two sections were located at equal distance (200 m) before and after this central section. After thawing, the sections were fixed in cold acetone for 10 min and then washed in phosphate-buffered saline (PBS, pH 7.4) for 5 min. Sections were subsequently incubated at room temperature in a moistened chamber for 45 min with an undiluted monoclonal rat-anti-mouse endothelial cell antibody (9F1, a gift from the Department of Pathology, University of Nijmegen; Westphal et al., 1997). Subsequently the sections were washed in PBS for 5 min (3 changes), followed by a 30-min incubation with TRITClabeled secondary antibody (goat-anti-rat-TRITC, Jackson ImmunoResearch Laboratories, Inc., West Grove, PA; Sanbio; dilution 1:50). After again three washes in PBS, the sections were incubated for 30 min with the TRITClabeled third antibody (donkey-anti-goat-TRITC, Jackson ImmunoResearch Lab, Sanbio; 1:50). The second and third antibodies were diluted in PBS containing 0.5% bovine albumin solution. The wet sections were then shortly washed in PBS and quickly mounted under a coverslip. At each time point five mice were fixed in 4% buffered paraformaldehyde (ph 7.4) for routine histology. Serial sections were sectioned in the frontal plane and stained with hematoxylin and eosin and with thionine for assessment of the glycosaminoglycan content of the matrix (Bulstra et al., 1993). Analysis of the Fluorescent Signals Immediately after the immunocytochemical procedure, labeled structures in the joint were visualized with epifluorescence. A filter with 550-nm excitation and 570-nm emission was used to visualize the TRITC-labeled antibodies, whereas a filter with 365-nm excitation and 420-nm emission nm was used to visualize the Hoechst 33342. A custom-made program was developed to automate all the operations needed for scanning the microscopic slides and analyzing synovial areas. Each slide was 422 BUMA ET AL. Fig. 1. Summary diagram of frontal view of cross-sectioned femur and ligaments. The left side shows the control situation and the right side the situation of arthritic joint after one week. Note the synovial swelling. The dotted line indicates the boundary of measured synovial area. F, femur; M, meniscus; JC, joint capsule; JS, joint space; P, patella; S, synovial membrane. scanned twice by field movement of the motorized scanning stage (EK 32, Märzhäuzer, Wetzler, Germany) using the two different filters for scanning the vascular structures and the perfused synovial areas, respectively. The video signals were digitized, processed and analyzed on a Macintosh Quadra 650 computer, using the digital imaging application TCL-Image (TNO, Delft, The Netherlands). Before the scanning procedure, a threshold for the vascular structures was determined for both stain signals. For each section, composite images were reconstructed from the individually processed microscopic images: one with the vascular structures, the other with the perfused areas. Using the image with the vascular structures, a contour line was drawn around the synovial area from the patella to the femur on a new image (see Fig. 1 for location of synovial area). The lumen of the knee joint was excluded. This area was used as a mask for further image analysis. After scanning, the coverslips were removed and the sections were mounted in Fluoritab (Organon Teknika, Cappel-ICN), and coverslipped again. right arthritic knee of each mouse, because these results were not independent. Furthermore, these data did not meet the requirements for analysis using parametric procedures. To analyze the difference between the two points of time, i.e., Day 4 and Day 7, the Mann-Whitney U-test was used; these data were also non-parametric. P values of less than 0.05 were considered to represent significance. RESULTS Clinical Behavior and Routine Histology The digital image system measured synovium area (SA), number of all blood vessels (NBV) and the number of perfused blood vessels (NpBV). From these data the percentage of perfused vessels of all vessels (perfusion fraction; PF ⫽ NpBV/NBV), the vessel density (VD ⫽ NBV/ SA) and the density of perfused vessels (VDp ⫽ NpBV/SA) could be calculated. The gross appearance of the affected joints had not changed dramatically after induction of mBSA arthritis. The mice functioned well clinically. Macroscopic examination of the right knee joint exhibited the normal characteristic signs of inflammation, such as swelling and redness. All micrographs were viewed with the patella on top of the femur and the tibia below (see Fig. 1 and 2). In control joints a layer of connective tissue in the synovial membrane (Fig. 2A) covered the surface of the joint cavity around the femur, tibia and cruciate ligaments. Four and 7 days after induction of arthritis the joint cavity was enlarged, the synovial membrane was thickened and both were invaded by massive quantities of polymorphonuclear lymphocytes (Fig. 2B). The tissues underneath the synovial membrane were filled with massive numbers of leukocytes. HE sections exhibited an increase in the number of small blood vessels, which seemed to penetrate the tissue of the synovial membrane from the joint capsule (Fig. 2C–E). Statistical Analysis Fluorescence Signals In the statistical analysis, the Wilcoxon signed rank test was used for the data from the left control knee and the Background staining with 9F1 was rather low (Fig. 3B,F). Controls for method specificity (omission of the first Quantification of the Perfusion and Vascularity in the Synovial Area of the Knee Joint DIGITAL IMAGE ANALYSIS OF MOUSE KNEE JOINT 423 Fig. 2. Thionine- (A) and HE-stained (B–E) stained sections of control (A) and inflamed joints (B–E) 4 (B–D) and 7 (E) days after induction. A: Note the thin synovial lining (arrows) of joint space (JS). JC, joint capsule. B: Synovial tissues (ST) are swollen and loaded with leukocytes. PT, patella. C–E: Small blood vessels are found in the thick synovial lining (arrows in D). D: Enlargement of boxed area in C. E: Seven days after induction of arthritis, the number of vessels containing erythrocytes (arrows) was higher. A, B: ⫻35; C: ⫻100; D: enlargement of C, ⫻170; E: ⫻140. incubation step) were completely negative. Nuclei of endothelial cells and occasionally cells directly adjacent to the perfused vessels were stained by the Hoechst marker, and were thus clearly distinguishable from the rest of the tissues (Fig. 3A,D,G). Diffusion of the Hoechst marker outside the endothelial lining was observed only occasionally in a small very small area around the perfused vessels. In all tissues, including muscle (Fig. 3A–C) and med- 424 BUMA ET AL. ullar tissue of bone (Fig. 3D,E) there was a close match between vascular 9F1 staining and the perfusion marker Hoechst. In control knees relatively few 9F1 positive blood vessels were found in periosteum and tissue directly under the synovial membrane (Fig. 3I). In some of the areas with no Hoechst staining, small vessels were stained with 9F1, and careful comparison of Hoechst images with 9F1 images showed that the larger vessels in particular were stained by Hoechst (Fig. 3F–H,K). The strong 9F1-fluorescence signal and the low signal from the surrounding soft tissue enabled segmentation of the vessels from the background by the image analysis system (Fig. 4A,B). Quantitation of Tissue Changes Associated With mBSA Arthritis The measurements exhibited an increase in the area of the synovial membrane (SA) during arthritis from ca. 0.2 mm2 in control knees to ca. 2.3 mm2 on Day 7 in arthritis cases (Fig. 5A). On Days 4 and 7, the SA during arthritis was significantly larger than that in the control knee (Fig. 5A; P ⫽ 0.043 and P ⫽ 0.016, respectively), and the SA was significantly larger on Day 7 than on Day 4 (P ⫽ 0.025). No significant difference was observed between the control knees. The number of blood vessels in the area of the synovial membrane varied between 70 (control) and 1,250 (inflamed joint; Fig. 5B). On Days 4 and 7, the number of vessels was significantly higher in the SA than in the controls (Fig. 5B; P ⫽ 0.043 and P ⫽ 0.022, respectively). Also, significantly more vessels were present on Day 7 than on Day 4 (Fig. 5B; P ⫽ 0.025). Surprisingly, a significant increase in vessel number was also found between Days 4 and 7 in the control knee (P ⫽ 0.025). The number of perfused blood vessels (NpBV) in arthritic synovial membrane was always lower than the total number of vessels (Fig. 5B), again indicating that not all vessels were perfused, but that they were significantly larger compared with the number in control knees on the same day (P ⫽ 0.043, P ⫽ 0.043, respectively). On Day 4 the synovial membrane had significantly less perfused vessels than on Day 7 (Fig. 5B; P ⫽ 0.025). No differences were observed between controls on Day 7 and Day 4. Changes in Vessels and Vascular Perfusion of the Synovial Membrane Associated With mBSA Arthritis On Day 4, the vessel density (VD) in the arthritic knee was slightly higher than that in the control knee (Fig. 5C; P ⫽ 0.043). On Day 7, this difference was no longer present, and the VD was significantly larger than that on Day 4 (P ⫽ 0.025), but the absolute differences in vessel density were not very large. The VD of the control knee on Day 7 was significantly larger than that on Day 4 (P ⫽ 0.025). In all measurements, the smallest VD was 250 per mm2, whereas the largest VD was 675 per mm2 (Fig. 5C). There were no differences in the perfused vessel density (VDp) on Days 4 and 7 between the control knee and the arthritic knee (Fig. 5C). Another observation was that the VDp on Day 7 in the arthritic knee was significantly larger than that on Day 4 (P ⫽ 0.025). In all measurements the VDp was in the range of 45–280 per mm2 (Fig. 5C). The perfusion fraction (PF) in all measurements was in the range of 0.1– 0.72 (Fig. 5D). The PF in the control knees on Day 4 was significantly higher than that in the arthritic knees (P ⫽ 0.043). On Day 7, the difference was no longer significant. The PF in the control knees was significantly higher on Day 4 than on Day 7 (P ⫽ 0.025). DISCUSSION mBSA-Induced Arthritis Mouse Model The main purpose of this study was to develop a procedure to quantify the changes in vascular structures during arthritic inflammation in mouse models. In humans these changes are difficult to study. Human joints are very large, which makes it impossible to study all joint structures in one section. Moreover, the specimens are often incomplete and the soft tissues may be disorientated in surgically removed specimens (FitzGerald et al., 1991). In addition, the stages of inflammation in the arthritic process may be very different from one individual to another, which makes it difficult to demonstrate an effect of vascular inhibitors. Therefore, it is important to use a model with reproducible tissue changes of the synovial membrane. This requirement was met in the present murine arthritic model (van de Loo et al., 1995). It has been described previously that the antigen-induced inflammatory process is highly reproducible in C57Black6 mice (van Beuningen et al., 1989). The chronicity of this type of experimental arthritis is dependent on T-cell reactivity and on the duration of antigen retention in joint tissues (van Lent et al., 1987). It is also related to age and gender (van Beuningen et al., 1989). In this experiment the mBSA-induced arthritis model exhibited highly reproducible swelling of the tissues of the synovial membrane on Days 4 and 7. Other experiments have shown that between 2 and 8 weeks the tissues of the synovial membrane will slowly return to their pre-arthritic configuration (van Beuningen et al., 1989; Buma et al., 2000). Digital Image Analysis System The first advantage that also contributed to the success of our digital image analysis system, was the possibility to scan entire sections of the synovial membrane, which made it impossible for the investigator to subjectively select parts of the synovial membrane. The second advantage is that this procedure not only yields quantitative data on the total number of vessels, but also allows combinations of this picture with the perfusion marker in the same area of the synovial membrane. The Hoechst marker stained nuclei of endothelial cells and some cells located in very close vicinity to the perfused blood vessels in the synovial membrane. Perfused vessels could be clearly distinguished from the nonvascularized area of the synovial membrane. To avoid variation in the diffusion of the Hoechst marker in the tissues, the time interval between perfusion and killing the animals must remain constant between different animals. In this study, the animals were killed after 1 min, which had proven to be an appropriate time interval in previous studies using the marker in a mouse model (Rijken et al., 1995). The scanning procedure, which takes about 20 min, should follow the staining procedure within a few hours. If not the Hoechst marker will diffuse away into the tissues, because it is only loosely bound to the nuclei of stained cells. Figure 3. 426 BUMA ET AL. Changes in Synovial Membrane Tissue Associated With mBSA Arthritis Previous quantitative studies of the human vasculature of the synovial membrane during inflammation exhibited an increase in vessel number as well as a change in vessel morphology in synovial membranes when compared with clinically unaffected joints (FitzGerald et al., 1991). These findings conflict with a report by Stevens et al. (1991) that suggested reduced vascularity with advanced disease. It is possible that these differences are a reflection of different stages of disease progression in rheumatoid arthritis, and they illustrate that it is extremely difficult to evaluate vascular changes in clinical human material. In this study, synovial membrane tissues were examined after two different time intervals during mBSA arthritis. On Day 4 after induction of arthritis, the swollen subsynovial membrane tissues were avascular (Buma et al., 2000). An important observation was that in the first stages of the inflammation process, the area of the synovial membrane (SA) increased significantly. This caused a situation in which parts of the SA were avascular. The SA kept on swelling and on Day 7 it was approximately six times bigger than the size in control knees and one and a half times bigger than the SA on Day 4 of mBSA arthritis. In this study we demonstrated an increase in the area of the synovial membrane, an increased number of vessels and changes in vessel densities in arthritis. The relatively low vessel density on Day 4 and the increase in vessel density on Day 7 as compared with Day 4, strongly suggests that directly after the induction of arthritis, the size of the synovial area increases rapidly without vessel formation and that after Day 4, vessel formation increases. This was confirms previous preliminary findings in that blood vessels sprouted, particularly between Days 4 and 7 (Buma et al., 2000). The relative avascularity of the synovial membrane during the first few days after the induction of arthritis might induce hypoxia, which is a powerful inducer of angiogenesis (Blake et al., 1989; Fava et al., 1994; Paleolog, 1996; Stevens et al., 1991). The relatively low number of perfused vessels on Day 4 indicates that the newly formed vessels were initially non-functional. This is in agreement with our observation that sprouting vessels end blindly in the synovial tissue, making actual flow impossible (Buma et al., 2000). The perfused vessel density (VDp) on Day 7 was again comparable to that of the control knee on day seven, indicating that after a few days, newly formed vessels become functional. So it might be speculated that there is in synovial tissue a fixed perfusion rate of vessels that is independent of vessel density. Fig. 3. 9F1 staining (B,F) Hoechst staining (A,D,G) and double photographs showing co-localization of 9F1 and Hoechst (C,E,H,I,K). A–C: Same location of muscle (M) of mouse. Note the close overlap between 9F1 (B,C) and Hoechst (A,C) in endothelial cells (arrows). Hoechst staining is only found in the nuclei of the endothelial cells (⫻120). D,E: Bone marrow (BM) of tibia. Close overlap between Hoechst (D) and 9F1 staining (E) (⫻180). F–H: 9F1, Hoechst and the combination in arthritic synovial membrane four days after induction (S). Note that some of the small 9F1-positive vessels are not stained by Hoechst (arrow) (⫻200). I,K: Control joint (I) and synovial membrane seven days after induction of arthritis (K). The arrow is pointed at a vessel not stained by Hoechst. BV, blood vessel; BM, bone marrow; C, capsule of joint; JS, joint space; M, muscle; P, periosteum; S, synovial membrane (⫻60). Fig. 4. Digitized images of control joint (A) and joint seven days after induction of arthritis (B). JS, joint space; P, periosteum; PA, patella; S, synovial membrane. An interesting observation was the increase in NBV and the VD in the control knee on Day 7 compared with Day 4. We do not know exactly what adaptive mechanism is involved in the increase in NBV and VD. Firstly, circulating inflammatory mediators released from the inflamed knee joint could have a more general effect on vascularity in the whole animal. This seems very plausible in the light of the observed proliferative effect of AIA induced arthritis on vessels outside the joint in the inflamed joint (unpublished observations). A second explanation could be that the AIA in the inflamed joint induced a changed load distribution in the contra lateral joint. This might have induced an adaptive response in the vessel density. This does not seem very plausible, however, because a changed walking pattern after induction of AIO was never ob- Fig. 5. Quantitative data of synovial area (A), number of vessels and number of perfused vessels (B), the vessel density in the synovial area (C) and the perfusion fraction (D). 428 BUMA ET AL. served. Irrespective of the mechanism involved, because the number of perfused vessels and the synovial area did not change, we speculate that no further vascular capacity was needed and that angiogenesis of non-functional vessels took place. Finally, we think that the measurements will be highly reproducible. In a pilot experiment we analyzed Day 7 arthritis. All parameters in the experiment exhibited the same significant differences between arthritic and control joints, indicating that the arthritis model in the mouse is highly reproducible. In similar studies of vascularization of tumour tissue the variations of the measured parameters were relatively limited (Rijken et al., 1995). Therefore, we think that this procedure yields very accurate measurements of the relationships between tissue changes during disease and the concomitant changes in vascularity. Moreover, this procedure may also prove very useful in other fields of research in which vascularity plays an important role. Important fields in orthopaedic research are studies related to avascular necrosis (Simank et al., 1997), tissue engineering studies of different parts of the locomotion system (Germain et al., 2000) and to follow revascularization of bone grafts and other substitutes (Kirschner et al., 1999). In conclusion, the Hoechst perfusion combined with immunocytochemistry allows an efficient quantitation of total vascularity and of functional vessels during arthritis. The procedure described here should allow us not only to quantify the effects of inhibitors of vascular sprouting on the vascularity but also to assess the effects on the arthritic processes in our mouse models. Moreover, there are clear applications of this procedure in the study of other bone and joint diseases in animal models. LITERATURE CITED Ben Av P, Crofford LJ, Wilder RL, Hla T. 1995. Induction of vascular endothelial growth factor expression in synovial fibroblasts by prostaglandin E and interleukin-1: a potential mechanism for inflammatory angiogenesis. FEBS Lett 372:83– 87. Bernsen HJ, Rijken PF, Hagemeier NE, van der Kogel AJ. 1999. A quantitative analysis of vascularization and perfusion of human glioma xenografts at different implantation sites. Microvasc Res 57:244 –257. Blake DR, Merry P, Unsworth J, Kidd BL, Outhwaite JM, Ballard R, Morris CJ, Gray L, Lunec J. 1989. Hypoxic-reperfusion injury in the inflamed human joint. Lancet 1:289 –293. Bulstra SK, Drukker J, Kuijer R, Buurman WA, van-der-Linden AJ. 1993. Thionin staining of paraffin and plastic embedded sections of cartilage. Biotech Histochem 68:20 –28. Buma P, Elmans L, van den Berg WB, Schrama LH. 2000. Neurovascular plasticity in the knee joint of an arthritic mouse model. Anat Rec 260:51– 81. Bussink J, Kaanders JH, Rijken PF, Martindale CA, van-der-Kogel AJ. 1998. Multiparameter analysis of vasculature, perfusion and proliferation in human tumour xenografts. Br J Cancer 77:57– 64. Bussink J, Kaanders JH, Rijken PF, Peters JP, Hodgkiss RJ, Marres HA, van der Kogel AJ. 1999. Vascular architecture and microenvironmental parameters in human squamous cell carcinoma xenografts: effects of carbogen and nicotinamide. Radiother Oncol 50:173–184. Claffey KP, Robinson GS. 1996. Regulation of VEGF/VPF expression in tumor cells: consequences for tumor growth and metastasis. Cancer Metastasis Rev 15:165–176. Colville Nash PR, Scott DL. 1992. Angiogenesis and rheumatoid arthritis: pathogenic and therapeutic implications. Ann Rheum Dis 51:919 –925. Fava RA, Olsen NJ, Spencer Green G, Yeo KT, Yeo TK, Berse B, Jackman RW, Senger DR, Dvorak HF, Brown LF. 1994. Vascular permeability factor/endothelial growth factor (VPF/VEGF): accumulation and expression in human synovial fluids and rheumatoid synovial tissue. J Exp Med 180:341–346. FitzGerald O, Soden M, Yanni G, Robinson R, Bresnihan B. 1991. Morphometric analysis of blood vessels in synovial membranes obtained from clinically affected and unaffected knee joints of patients with rheumatoid arthritis. Ann Rheum Dis 50:792–796. Germain L, Remy-Zolghadri M, Auger F. 2000. Tissue engineering of the vascular system: from capillaries to larger blood vessels. Med Biol Engineer Computing 38:232–240. Kirschner MH, Suess L, Wanders A, Harris AG, Krombach F, Hofmann GO, Messmer K. 1999. Long-term preservation with UW solution vs. EC improves revascularization of transplanted bone. Langenbecks Archives Surg 1(Suppl):83– 87. Kruijsen MWM, van den Berg WB, van de Putte LBA. 1983. Sequential alterations of periarticular structures in antigen-induced arthritis in mice: histological observations on fibrous capsule, ligaments, bone and muscles, using whole mount sections. Br J Exp Pathol 64:298 –305. Nagashima M, Yoshino S, Ishiwata T, Asano G. 1995. Role of vascular endothelial growth factor in angiogenesis of rheumatoid arthritis. J Rheumatol 22:1624 –1630. Oliver SJ, Cheng TP, Banquerigo ML, Brahn E. 1995. Suppression of collagen-induced arthritis by an angiogenesis inhibitor, AGM-1470, in combination with cyclosporin: reduction of vascular endothelial growth factor (VEGF). Cell Immunol 166:196 –206. Paleolog EM. 1996. Angiogenesis: a critical process in the pathogenesis of RA—a role for VEGF? Br J Rheumatol 35:917–919. Rijken PF, Bernsen HJ, van der Kogel AJ. 1995. Application of an image analysis system to the quantitation of tumor perfusion and vascularity in human glioma xenografts. Microvasc Res 50:141– 153. Simank HG, Graf J, Kerber A, Wiedmaier S. 1997. Long-term effects of core decompression by drilling— demonstration of bone healing and vessel ingrowth in an animal study. Acta Anatomica 158:185– 191. Smith KA, Hill SA, Begg AC, Denekamp J. 1988. Validation of the fluorescent dye Hoechst 33342 as a vascular space marker in tumors. Br J Cancer 57:247–253. Stevens CR, Blake DR, Merry P, Revell PA, Levick JR. 1991. A comparative study by morphometry of the microvasculature in normal and rheumatoid synovium. Arthritis Rheum 34:1508 –1513. van Beuningen HM, van den Berg WB, Schalkwijk J, Arntz OJ, van de Putte LB. 1989. Age- and sex-related differences in antigen-induced arthritis in C57Bl/10 mice. Arthritis Rheum 32:789 –794. Van de Loo AAJ, Joosten LAB, van Lent PLEM, Arntz OJ, van den Berg WB. 1995. Role of interleukin-1, tumor necrosis factor alpha and interleukin-6 in cartilage proteoglycan metabolism and destruction. Effect of in situ cytokine blocking in murine antigen- and zymosan-induced arthritis. Arthritis Rheum 38:164 –172. Van den Berg WB, Kruijsen MWM, van de Putte LBA, van Beusekom HJ, van der Sluis M, Zwarts WA. 1981 Antigen-induced and zymosan-induced arthritis in mice: studies on in vivo cartilage proteoglycan synthesis and chondrocyte death. Brit J Exp Pathol 62:308 – 316 Van Lent PL, van den Berg WB, Schalkwijk J, van de Putte LB, van den Bersselaar L. 1987. Allergic arthritis induced by cationic antigens: relationship of chronicity with antigen retention and Tcell reactivity. Immunology 62:265–272. Westphal JR, van ’t Hullenaar RG, van der Laak JA, Cornelissen IM, Schalkwijk LJ, van Muijen GN, Wesseling P, de Wilde PC, Ruiter DJ, de Waal RM. 1997. Vascular density in melanoma xenografts correlates with vascular permeability factor expression but not with metastatic potential. Br J Cancer 76:561–570.