| Background and ObjectiveAs a chronic inflammatory disorder, atherosclerosis usually progresses silently for decades before evident clinical manifestations, such as myocardial infarction and stroke. However, most diagnostic method can only detect atherosclerostic changes at the advanced stage, by either evaluating the stenoses of arteries, or imaging vascular changes such as coronary calcification, carotid intima-media thickening, and plaque morphology. Inflammation is closely involved in the pathophysiologic process of atherogenesis, which is characterized by early expression up-regulation of leukocyte adhesion molecules (including P-, L-, E-selectin, intercellular adhesion molecule-1(ICAM-1), vascular cell adhesion molecule-1(VCAM-1) etc.) on the vascular endothelial surface. By real-time visualization of these inflammatory molecules, targeted ultrasound molecular imaging (UMI) promises to offer early diagnosis prior to overt clinical manifestations. And the future risk for atherosclerosis progression such as plaque vulnerability to erosion and rupture could also be evaluated by direct assessment of the extent of vascular inflammation. The great potential of this technology in the detection of microvascular inflammation has already been confirmed. However, its practical application in large-and middle-sized arteries is still difficult. Two factors are probably responsible—the high shear stress of blood flow, which limits the rapid formation of adhesive bonds and increases the dislodgement of the adhered microbubble contrast agents; The similar rheologic behavior of microbubbles to that of erythrocytes, which makes the microbubbles tend to remain close to the axial center in most regions of blood vessels, whereas the biomarker targets of UMI are present on the luminal surface. However, studies have confirmed the co-localization of atherosclerosis to the furcations of arteries, where the "axial phenomenon" was disturbed. So, at these furcations, microbubbles have more "chance" to react with the molecular targets on the luminal surface that the application of the targeted UMI would probably accomplished. The concept of dual-/multi-ligands microbubbles bearing both sialyl Lewis" and antibody ligands to increase the adhesion efficiency against the vigorous blood flow in larger arterial vessels has also been developed these years. Recent studies have demonstrated that dual-/multi-ligands microbubbles exhibited adequate adhesive capacity at the high shear stress of4-8dyn/cm2in vitro. But the mechanism of this "dual-/multi-ligands" advantage was more of a theoretical hypothesis with no immediate evidence.So, in the present study, we tried to answer2questions. First, what was the mechanism of the "dual-/multi-ligands" advantage? Was it consistent with the theoretical hypothesis? Second, can UMI of the inflammatory endothelium at the furcation of arteries be improved by dual-ligands microbubbles bearing both anti-P-selectin monoclonal antibodies and sialyl lewisx? P-selectin, as one of the representative inflammatory molecules in atherosclerosis, was selected as the imaging target. The adhesive capability of the targeted microbubbles was evaluated by a parallel plate flow chamber in vitro. We also made further research on the mechanism why dual-ligands and mono-ligand microbubbles exhibited diverse adhesive capability by analysis of the microbubble adhesive behavior using an innovative methodology. Then, mice models of vascular endothelial inflammation located at the aorto-iliac bifurcation were prepared. The efficacy of the targeted microbubbles in improving UMI of the inflammatory endothelium was examined in these mice models.Methods1. Microbubble Preparation1.1. Biotinylated microbubbles preparation:Biotinylated, lipid shelled decafl-uorobutane microbubbles were prepared by shearing of a C3F8-saturated aqueous suspension of DSPE-PEG2000-Biotin、DPPC、PEG4000. Microbubbles were then washed by superpurified water three times to remove exeess free unincorporated lipid. The concentration and size distribution of microbubbles were analysed with a Coulter Multisizer counter.1.2. Targeted microbubbles preparation:Biotinylated sialyl lewisx or biotinylated anti-P-selectin antibody were attached to the biotinylated microbubbles surface via multi-step avidin biotin bridging chemistry. Microbubbles were washed three times, streptavidin was then added to the washed microbubble dispersion. Following incubated at4℃with1.5μg of biotinylated sialyl lewisx or7.5μg of biotinylated anti-P-selectin antibody per107microbubbles, Selectin-targeted (via sialyl lewisx) microbubbles (MB-S) and P-selectin-targeted (via P-selectin monoclonal antibodies) microbubbles (MB-P) were prepared. Dual-ligands (via both sialyl lewisx and P-selectin monoclonal antibodies) microbubbles (MB-D) were prepared by simultaneous incubation with0.75μg of biotinylated sialyl lewisx and3.75μg of biotinylated anti-P-selectin antibodies per107microbubbles. A control (via isotype control antibodies) microbubbles (MB-C), was also constructed by conjuncting7.5μg rabbit monoclonal IgG1per107microbubbles. Coulter counter was then used to evaluate the characters of the microbubbles.1.3. Targeted microbubbles valuation:Targeted microbubbles and control microbubbles were incubated with both fluorescein-conjugated affinipure goat anti-rat IgG and rhodamine-conjugated affinipure goat anti-mouse IgG, using fluorescence microscope to identify the linking antibodies on the microbubbles. Then the binding rate of ligands in microbubbles was measured by quantitative flow cytometry.2. In Vitro ExperimentThe flow chamber Petri dish was incubated with200μl of P-selectin.Fc at concentration in of1μg/ml overnight at4℃to simulate the expression of inflammatory molecular on the vascular endothelial surface.2.1. Attachment study:four types of microbubbles (5×106/ml) were drawn through the parallel plate flow chamber that coated with P-selectin at the shear stress of0.5dyn/cm2,2.0dyn/cm2and4.0dyn/cm2, respectively. Video recording lasted6min since the emergence of microbubble in the observed area under the microscope.2.2. Detachment study:the shear stress was set at0for5min to allow the microbubbles to interact with P-selectins adequately. Then, the shear stress was set at0.2dyn/cm2to remove free microbubbles. and doubled every30s until all microbubbles were detached.2.3. Image-pro-plus software6.0were used to analyze the transient velocities, rolling and firmly adherent numbers of microbubbles at various shear stress. Rolling was defined as velocitiy less than50%that of the free ones, and be0for less than5s. Firm adherence was defined as velocitiy at0, which lasted more than5s. Further analysis on the adhesive behavior of the micobubbles was performed using the video recorded in the attachment study combined with the set-point tracking function of the software.3. Prepartion of mice model of abdominal aortic inflammation:40male mice at12to14weeks of age were equally randomized into8groups (n=5in each group). A control group and an inflammatory group were assigned for each type of micobubbles. Mice in the inflammatory group were injected with0.3ml PBS which contained0.125ug IL-1β and0.5μg TNF-α into the retroperitoneal space around the aorto-iliac bifurcation, and the control group with0.3ml PBS only.4h following the retroperitoneal injection, each anesthetized mouse was secured supine with the caudal vein cannulated for the administration of microbubbles.4. Ultrasound Molecular Imaging in Vivo Ultrasound (Sequoia, Siemens Medical Systems, Mountain View, Calif) with a high-frequency linear-array probe (17L5transducer on the Sequoia system) fixed in place was employed. Abdominal aorta and bilateral iliac arteries were imaged with conventional ultrasound at15MHz in the longitudinal axis. The blood flow in the arteries was confirmed by pulsed-wave Doppler. Ultrasound molecular imaging was performed simultaneously the injection of a bolus of micobubbles via the caudal vein. The ultrasound signal (video intensity, VI) from microbubbles in the abdominal aorta was measured using MCE2.7(Yabko; University of Virginia, Charlottesville, Va) and presented with color coding.5. Histopathology and ImmunohistochemistryStaining was performed on frozen sections of the distal abdominal aortae (close to the bifurcations). Rat anti-mouse CD62P (BD, American) was used as a primary antibody with a secondary anti-rat antibody (GBI, American) to detect theendothelial P-selectin expression. All slides were visualized under a microscope (BX51, Olympus, Japan), and photographed with a charge-coupled device camera (C150L, Pixera).6. Statistical Analysis:All analyses were conducted using software (SPSS13.0). Results were expressed as Mean±SD. All p values were2-tailed, with statistical significance set at0.05. Comparisons of diameters, concentrations between microbubble agents and the microbubble detachment in the flow chamber were performed with one-way ANOVA followed by Bonferroni post-tests. The microbubble attachment in the flow chamber and the targeted UMI data in vivo were analyzed using two-way factorial design with Bonferroni post-tests.Results1. Characterization of microbubblesThe concentration of the targeted microbubbles was (2.21±0.13)×108/ml. The diameter was (2.76±0.87) μm. There was no difference statistically between groups in both concentration and size. MB-C were absent of fluorescence under the fluorescence microscopy. MB-S only fluoresced in red, while MB-P only in green, and MB-D expressed both red and green fluorescence. 2. In Vitro Experiment2.1. Attachment study:MB-S, MB-P and MB-D exhibited determinate targeting abilities to P-selectins at various shear stresses, while MB-C only had minimal microbubble attachment. The attachment of the microbubbles all decreased as the shear stress increased (P＜0.01). At2.0dyn/cm2and4.0dyn/cm2, the rolling numbers of MB-D and MB-S were more than that of MB-P (P＜0.01), and the firm adherence number of MB-D were more than those of MB-P and MB-S (P＜0.05).2.2. Detachment study:MB-C could not attach on the parallel plate even at a low shear stress. Both the half-maximal and the complete detachment shear stress increased in the following order:MB-C<MB-S<MB-D<MB-P (P＜0.05).2.3. The adhesive behavior of the micobubbles:at the same shear stress, MB-P rolled in a relatively high velocity, which then decreased to0sharply; The velocity of MB-S decreased from a relatively high level to a relatively low level, then to0gradually; MB-D rolled in from a relatively high velocity to a low one, which then descended to0sharply. The velocities of MB-P, MB-S, MB-D all decreased with the increase of the shear stress, while the characteristic tendencies of the velocity transition were alike among the same type of micobubbles. A considerable number of MB-S adhered in an unstable style, which denoted that the micobubbles were detached from P-selectins again after adhesion.3. Ultrasound Molecular Imaging in VivoThe background-subtracted color-coded ultrasound molecular images were presented. There were only minimal signals for MB-C, MB-S, MB-P and MB-D in the control group and MB-C in the inflammatory group. The greatest signal enhancement was observed for MB-D in the inflammatory group. MB-S and MB-P in the inflammatory group also showed significant signal enhancement. Quantitation of Ⅵ:For MB-S, MB-P and MB-D, the Ⅵ in the inflammatory group was much greater than that in the control group (P<0.01), while for MB-C, no significant difference was found between the two groups. In the inflammatory group, the value of VI decreased in the following order:MB-D> MB-S, MB-P> MB-C (P<0.01). No significant difference was found between MB-S and MB-P (Figure4). 4. Histopathology and Immunohistology4.1. On histology with H&E staining, no evidence of inflammation was found in the control group, while edema and infiltration of a large number of leukocytes in the arterial wall were observed in the inflammatory group.4.2. On immunostaining for P-selectin, there was minimal staining in the control group. However, abundant P-selectin expression was detected on the luminal endothelial surface of the abdominal aorta in the inflammatory group.Conclusions1. Dual-liglands microbubbles(MB-D) and their single-targeted counterparts targeted to P-selectin can be successfully constructed by combinding anti-P-selectin monoclonal antibodies and sialyl Lewisx to lipid microbubbles via "avidin-biotin" bridging chemistry.2. MB-D exhibited higher adhesion efficacy to P-selectins than MB-S, MB-P at the high shear stress in vitro. The adhesive capacity was adequate even at the high shear stress of4-8dyn/cm2.3. MB-S, MB-P and MB-D showed different ways of adhesive behavior and adhesive capacilities in vitro. The adhesive behavior of MB-D exhibited the feasure of both those of MB-S and MB-P. In the former half of rolling, the adhesive behaviors of MB-D and MB-S were alike. The velocities all decreased from a relatively high level gradually. The latter half of MB-D rolling was similar to that of MB-P. The velocities all descended to0sharply in a relatively medium or high level. These results implied that for MB-D, the role of sialyl Lewisx was to decrease the velocity of microbubble rolling in the setting of high shear stress. Thus, the duration of antibody-antigen interactions was prolonged and more "chances" for firm adhesion mediated by anti-P-selectin antibodies were created. These two ligands may play complementary roles in the targeted adhesion of microbubbles.4. Ultrasound molecular imaging using MB-D can evaluate the inflammatory change at the aorto-iliac bifurcation due to the higher adhesive capability of MB-D and the non-axial distribution of the microbubbles. The technology has great potential in the detection of vascular inflammation in early atherosclerosis. |
PDF Full Text Request