Author + information
- Received April 22, 1998
- Revision received September 25, 1998
- Accepted November 5, 1998
- Published online March 1, 1999.
- Sasha M. Demos, PhD∗,* (, )
- Hayat Alkan-Onyuksel, PhD∗,†,
- Bonnie J. Kane, BS‡,
- Kishin Ramani, MD‡,
- Ashwin Nagaraj, PhD‡,
- Rodney Greene, BS‡,
- Melvin Klegerman, PhD† and
- David D. McPherson, MD‡
- ↵*Reprint requests and correspondence: Dr. Sasha M. Demos, c/o Dr. David D. McPherson, Northwestern Memorial Hospital, Cardiology/Medicine, 250 East Superior Street—Wesley 520, Chicago, Illinois 60611
The purpose of this study was to target acoustically reflective liposomes to atherosclerotic plaques in vivo for ultrasound image enhancement.
We have previously demonstrated the development of acoustically reflective liposomes that can be conjugated for site-specific acoustic enhancement. This study evaluates the ability of liposomes coupled to antibodies specific for different components of atherosclerotic plaques and thrombi to target and enhance ultrasonic images in vivo.
Liposomes were prepared with phospholipids and cholesterol using a dehydration/rehydration method. Antibodies were thiolated for liposome conjugation with N-succinimidyl 3-(2-pyridyldithio) propionate resulting in a thioether linkage between the protein and the phospholipid. Liposomes were conjugated to antifibrinogen or anti–intercellular adhesion molecule-1 (anti–ICAM-1). In a Yucatan miniswine model, atherosclerosis was developed by crush injury of one carotid and one femoral artery and ingestion of a hypercholesterolemic diet. After full plaque development the arteries were imaged (20-MHz intravascular ultrasound catheter and 7.5-MHz transvascular linear probe) after injection of saline, unconjugated liposomes and antibody conjugated liposomes.
Conjugated liposomes retained their acoustically reflective properties and provided ultrasonic image enhancement of their targeted structures. Liposomes conjugated to antifibrinogen attached to thrombi and fibrous portions of the atheroma, whereas liposomes conjugated to anti–ICAM-1 attached to early atheroma.
Our data demonstrate that this novel acoustic agent can provide varying targeting with different antibodies with retention of intravascular and transvascular acoustic properties.
Atherosclerosis (ATH) has been shown to be extensive, of variable morphology and widely distributed in vascular beds due to variable effects of remodeling, luminal and neural influences (1–5). High resolution ultrasound, both external ultrasound for the carotid, femoral and superficial arteries, and intravascular ultrasound for the coronaries, iliacs and aortas, has been helpful in identifying atheroma location and severity (6–10). However, there continues to be difficulty in identifying surface characteristics of plaque and plaque morphology. Investigators have been able to utilize ultrasound tissue characterization to distinguish calcified lesions, but have had lesser degrees of success with fibrous, fibrofatty, fatty and thrombotic lesions (11–15). An agent that has the potential to highlight surface features of the arterial wall and atheroma as well as highlight atheroma components would have far-reaching implications for more accurate morphologic identification, quantitation of and direction of therapy toward plaque components in different stages of development.
We have previously demonstrated the generation of acoustically reflective liposomes, solely by controlling the lipid composition and method of production (16). These acoustic liposomes consist of a 60:8:2:30 molar mixture of phosphatidylcholine:phosphatidylethanolamine: phosphatidylglycerol:cholesterol and are prepared by a dehydration/rehydration method. Freeze fracture electron microscopy has illustrated that highly acoustic liposomes are multilamellar, with well separated bilayers and internal vesicles, whereas nonacoustic liposomes consist of one thick outer membrane. With no gas entrapment and a mean size of less than 1 μm, the liposomes will be stable in circulation and easily pass through pulmonary capillaries. The liposomes retain their acoustic properties at 37°C, in the presence of blood and after conjugation to antibodies. Labeling the liposomes with antibodies results in the formation of a site-specific contrast agent that can enhance ultrasound images of chosen target tissues and molecules. When conjugated to antifibrinogen antibodies, the acoustic liposomes can target to fibrin-coated surfaces and thrombi in vitro (17). Scanning electron microscopy and ultrasound imaging revealed that antifibrinogen liposomes remain attached to their target sites after saline wash, whereas unconjugated liposomes and liposomes conjugated to a nonspecific antibody did not.
Atherosclerotic plaque formation involves a complex interplay between the vascular wall and blood components. Early in plaque formation different molecules are involved in the recruitment of a variety of cells to the lesion. Intercellular adhesion molecule-1 (ICAM-1) is expressed on the endothelium and enhances the recruitment of monocytes to the plaque (18,19). Staining of ICAM-1 has shown dramatic enhancement of expression on endothelial cells in fatty streaks and fibrofatty lesions and little expression in fibrous lesions and none in normal arteries (18). Liposomes conjugated to anti–ICAM-1 were used to target early stage atherosclerotic plaques.
This study describes acoustically reflective liposomes, conjugated to antifibrinogen and anti–ICAM-1, that can target atherosclerotic plaques and thrombi in a miniswine model as well as enhance both transvascular and intravascular ultrasonic images.
Liposomes were made from L-alpha phosphatidylcholine Type XIII-E (PC), 3-(2-pyridyldithiolpropionic acid N-hydroxysuccinimide ester (SPDP) and cholesterol (Sigma Chemical Co., St. Louis, Missouri); maleimido-4 (p-phenylbutyrate)-phosphatidylethanolamine (MPB-PE) (Avanti Polar Lipids, Alabaster, AL); 1,2-dipalmitoyl-sn-glycero-3-phosphoglycerol (DPPG) (Fine Chemicals, Liestal, Switzerland); dithiothreitol (DTT) (Fisher Scientific, Itasca, IL); rabbit antihuman fibrinogen (American Diagnostica, Greenwich, CT), and intercellular adhesion molecule antibodies.
Preparation of antibody conjugated liposomes
Preparation of vesicles
Maleimido-4 (p-phenylbutyrate)-phosphatidylethanolamine, a maleimide-containing phospholipid, was included in the composition of the liposomes for conjugation to the antibodies. The maleimide group of MPB-PE reacts with thiol groups, forming a stable thioether bridge (20). A 60:8:2:30 molar mixture of PC, MPB-PE, DPPG and cholesterol was dissolved in chloroform. The solvent was evaporated by a rotary evaporator (Labconco, Kansas City, MO), rotated at 120 rpm and immersed in a thermostatted water bath, with the temperature set at 50°C under argon. The resulting lipid film was then placed in a dessicator under vacuum for 2 days for complete drying. The dry lipid film was rehydrated with deionized water to give a concentration of 10 mg lipid/1 ml water. After removing the film from the walls of the flask, the dispersion was sonicated in a water bath sonicator (Fisher Scientific, Itasca, IL) until the mean size of the liposomes was approximately 500 nm. Liposome size was determined by quasielastic light scatter as described below. D-mannitol (0.2 mol/L) was added to the liposome suspension in a 1:1 vol/vol ratio, and this suspension was frozen overnight at −70°C. The frozen samples were placed in a lyophilizer (Labconco, Kansas City, MO) for 2 days. The dried lipids were then resuspended with 0.1 mol/L phosphate buffer, pH 7.4, to give a concentration of 10 mg lipid/1 ml buffer. After hand shaking and brief vortexing, the size of the resulting liposomes was again determined. The mean size of the liposomes remained under 1 μm. In most cases the size remained below 800 nm.
An antibody solution was prepared by dissolving the protein in 0.1 mol/L phosphate buffer (pH 7.4) at a concentration of 5 mg ml−1. A 20-μmol/ml−1solution of SPDP in ethanol was prepared. The SPDP solution was added to the stirred protein solution to give a molar ratio of SPDP to protein of 15:1. The mixture reacted for 30 min at room temperature. After reaction, the protein was separated from the reactants by gel chromatography. The column was packed with Sephadex G-50 and equilibrated with 0.05 mol/L sodium citrate, 0.05 mol/L sodium phosphate and 0.05 mol/L sodium chloride at pH 7.0. Detection of protein in column elutions was determined by absorbance at 280 nm.
Reduction of derivatized protein
The pyridyl dithio-protein solution was titrated in citrate–phosphate buffer to pH 5.5 by addition of 1 mol/L HCl. A solution of 2.5 mol/L DTT in 0.2 mol/L acetate buffer, pH 5.5, was made. Ten microliters of DTT solution was added for every milliliter of protein solution. This mixture was allowed to stand for 30 min. The protein was then separated from the DTT by chromatography on a Sephadex G-50 column equilibrated with citrate–phosphate buffer at pH 7.0. Detection of protein in the elution was determined by absorbance at 280 nm. Argon was bubbled through all buffers to remove oxygen.
Conjugation of protein to liposomes
The protein was mixed with the MPB-PE–containing liposomes and allowed to react overnight under argon. Unbound protein was separated from the liposomes on a Sepharose 4B column. Detection of protein was determined by use of a bicinchoninic acid (BCA) protein assay (Pierce Labs, Rockford, Illinois).
The mean size of the liposomes was determined by quasielastic light scattering measurements using a Nicomp model 270 submicron particle sizer (Pacific Scientific, Menlo Park, CA) equipped with a 5-mW helium–neon laser at an exciting wavelength of 632.8 nm and with a 64-channel autocorrelation function, a temperature-controlled scattering cell holder and an ADM 11 video display terminal computer (Lear Siegler Inc., Anaheim, California) for analyzing the fluctuations in scattered light intensity generated by the diffusion of particles in solution. The mean hydrodynamic particle diameter, dh, was obtained from the Stokes–Einstein relation using the measured diffusion coefficient obtained from the fit.
Liposomes were imaged with a 20-MHz high frequency intravascular imaging catheter (Boston Scientific, Sunnyvale, California) and a 7.5-MHz transvascular linear probe (Acuson Inc., Mountain View, California). Instrument settings for gain, zoom, compression and rejection levels were optimized at the initiation of the experiment. To correct for any minor variation in gain that may have occurred throughout the image, the image data were scaled to a common reference feature (see following section). Images were recorded on to ½-in. VHS videotape in real time for subsequent playback and image analysis.
Videodensitometric analysis of liposome “brightness”
Relative echogenicity (apparent brightness) of all liposome formulations was objectively assessed via computer-assisted videodensitometry. This process involved image acquisition, preprocessing, liposome identification and gray scale quantification. All image processing and analysis was performed with Image Pro Plus Software (version 1.0, Media Cybernetics, Silver Spring, Maryland) running on a dedicated computer (486 CPU, 66 MHz). Images were digitized to 640 × 480 pixels spatial resolution (approximately 0.045 mm/pixel) and 8-bit (256 levels) amplitude resolution. The distribution of gray scale values within the image was adjusted to cover the entire range of possible gray levels using a linear transformation algorithm (i.e., dynamic range was maximized). Image brightness was scaled to a reference feature, the annotation text, common to each image, retaining a constant gray scale value over all images. The reference text was imprinted on each frame before imaging. Therefore as image gain and contrast were altered, the text image and gain were altered to the same extent. An automated liposome detection routine was run to identify liposomes within a region of interest that encompassed the entire portion of the artery visualized (Fig. 1). The automated liposome detection routine identified and outlined all “bright” objects within the analysis region having a gray scale level >29, a roundness ratio (ratio of maximum diameter:minimum diameter) <2.5 and a size >4 pixels to avoid artifact. A brightness area product (BAP) was calculated for each complex. Brightness area products were generally in the order of 500 to 5,000 units. Larger BAPs, compatible with image artifact, primarily from the intravascular ultrasound catheter, were highlighted and removed from image analysis. This leaves some of the larger liposomal complexes uncounted but insures that artifacts will not be part of the liposomal analysis algorithm. Thus, objects identified were considered to be “liposomes.” Areas of liposomes were then numbered by the computer program, and the average gray scale and size of liposome groups were quantified. Liposomes/liposomal complexes are much smaller than the imaging pixels, so detected liposomes will appear larger than they actually are. When conjugated liposomes attach to the target site they further aggregate, resulting in groups of liposomes contributing to each detected bright complex.
A standard Yucatan miniswine model of induced ATH was developed to evaluate in vivo enhancement of ultrasonic imaging of atheromas with the antibody-conjugated liposomes. Atherosclerosis was developed by endothelial denudement of one carotid and one iliofemoral artery under chronic sterile conditions and ingestion of a markedly hypercholesterolemic diet (21,22). After two weeks of feeding, each animal was anesthetized and underwent percutaneous angioplasty to denude the endothelium of one carotid and one iliofemoral artery. This allows active plaque development consisting of fibrous and fatty tissue, and active endothelial components. Four to six weeks later, after transvascular ultrasound demonstrated active plaque, the animal was anesthetized and artificially ventilated. Sheaths were inserted into the femoral and carotid arteries for vascular access and intravascular imaging catheter insertion. Under stable conditions either intravascular or transvascular imaging was performed. For intravascular imaging the catheters were inserted close to the region of interest. For transvascular imaging the transducer was placed over the artery to be examined. For the antifibrinogen-conjugated liposomal experiments, the animals were not heparinized for 2 h before imaging, allowing thrombus to form in the peripheral arteries. The miniswine in the anti–ICAM-1 studies were kept on heparin throughout the experiment to prevent thrombus formation. A total of 20 diseased arteries in 10 pigs were prepared and used for these experiments; six pigs for antifibrinogen experiments and four pigs for anti–ICAM-1 experiments.
Baseline imaging of the artery and luminal blood was performed, and images were recorded onto videotape. The arteries were then imaged after one local injection each of 4 cc of agitated saline, 4 cc unconjugated liposomes and 4 cc antibody-conjugated liposomes. Local injections were made into both the normal and diseased carotid and femoral arteries upstream of the intended target. The liposomes were not allowed to incubate, and blood was flowing through the arteries throughout the entire experiment. Arteries were imaged for up to five min after injection of saline and unconjugated liposomes, and up to 15 min after injection of conjugated liposomes. The images were analyzed to determine acoustic enhancement of both unconjugated and conjugated liposomes as compared to blood pool intensity. The mean gray scale of the liposomes was determined by averaging the values obtained from three image frames for each arterial segment.
Data were compared by analysis of variance and determined to be statistically different at p < 0.05. All pairwise comparisons were based on Tukey’s method for multiple comparisons. All data are reported as mean ± standard deviation.
Imaging demonstrated visual enhancement of thrombus and early stage atheroma with both sets of liposomes. The mean size of both the conjugated and unconjugated liposomes ranged from 0.5 μm to 1.0 μm consistently with an average size of 0.8 μm. We have previously demonstrated that conjugation does not affect the size of the liposomes (17). Figure 2illustrates imaging of thrombus with intravascular ultrasound. Agitated saline injection produced transient enhancement of the lumen. Similarly unconjugated liposomes could be seen transiently during injection, but there was no attachment to the thrombi or vessel wall. Figure 2,Bprovides a negative control demonstrating that unconjugated liposomes did not attach to the thrombi after injection. The antifibrinogen liposomes, however, attached to the thrombi after injection and enhanced the image at the target site (Fig. 2,C). In real time these highly acoustic liposomes attached to the plaque with marked acoustic enhancement and shadowing of intraluminal blood. Figure 3illustrates transvascular ultrasound images of the carotid artery of a miniswine before and after injection of the antifibrinogen liposomes. Here intraluminal thrombus was not present, but the liposomes attached to the atherosclerotic arterial wall—presumably at sites of exposed fibrin. Transvascular imaging for 15 min demonstrated ongoing acoustic enhancement.
Arteries were imaged by intravascular and transvascular ultrasound after injection of saline, unconjugated liposomes and the anti–ICAM-1 liposomes. Figure 4illustrates the femoral artery of a miniswine imaged with an intravascular ultrasound catheter. After injection of unconjugated liposomes, some liposomes can be seen in the lumen of the artery, but none adhere to the arterial wall (Fig. 4,B). The anti–ICAM-1 liposomes are attached directly to the endothelial surface of the vessel wall (Fig. 4,C). The transvascular ultrasound images from the carotid artery of a miniswine after injection of saline, unconjugated liposomes and anti–ICAM-1 liposomes are illustrated in Figure 5. The anti–ICAM-1 conjugated liposomes can be seen attached to the lesion (Fig. 5,C), whereas the liposomes without antibody are seen only in the lumen (Fig. 5,B).
All images were analyzed to measure acoustic reflectivity of liposomes, agitated saline and blood components. Gray scale values were determined for both the intravascular and the transvascular images. The mean gray scale values of the attached liposomes did not differ over time. Figure 6demonstrates gray scale values for the intravascular images and shows that there is no difference between the unconjugated (112 ± 9.8) and conjugated (109 ± 1.3) (gray scale; mean ± SD) liposomes, but both are brighter than the agitated saline (73 ± 4.6) and blood (37 ± 7.2) (n = 3 arterial segments per data set; three image frames per arterial segment) (see Tables 1 to 3⇓⇓for data to Figs. 6 and 7). ⇓Similar results were seen for the transvascular imaging as illustrated in Figure 7(n = 5 arterial segments per data set; three image frames per arterial segment). The unconjugated (83 ± 2.3) and conjugated (81 ± 3.6) (gray scale; mean ± SD) liposomes were brighter than the agitated saline (59 ± 4.0) and blood (29 ± 2.2).
These results highlight the effectiveness of our formulation as an acoustic contrast agent able to successfully target pathologic tissues after in vivo injection.
Liposomes as acoustic contrast agents
To improve image resolution, ultrasonic contrast agents have been employed. Many contrast agents are being developed to enhance the circulating blood pool and provide contrast between the blood and the surrounding tissues. There are several difficulties associated with the present contrast agents. One common problem with gas-containing agents is the control of the size and stability of the gas bubbles after injection. Those that are too large will have difficulty passing through the pulmonary capillary system. Encapsulated systems (microspheres) improve the stability of the systems, but not to the desired extent. The half-life of the microspheres when injected into the blood has been shown to be very short (less than 1 min), due to uptake by the reticuloendothelial system (23). Although many of these contrast agents may enhance the circulating blood pool, thereby improving image resolution, they do not provide site-specific enhancement of structural pathology.
Acoustically reflective liposomes have many advantages over other contrast agents. Liposomes, composed of phospholipid bilayers, have been shown to be safe and efficacious for use (24). Not all liposomes are reflective. Our liposomes are acoustic solely by their lipid composition and method of production (16). These changes in methodology of liposomal formulation are important. The production method and composition results in liposomes that are composed of many well separated bilayers. We have previously shown that the lipid composition greatly affects the reflectivity of the liposomes. We have also shown that liposomes of this formulation and resultant structure are highly reflective, but liposomes composed of different formulations with resultant single bilayer, unilamellar vesicles, are not (16). Mechanisms of acoustic enhancement of our formulation include a combination of membrane separation, number of membranes, thickness and possibly air entrapment between bilayers during the lyophilization process.
Additional advantages of our formulation include their reproducibly small size (less than 1 μm), allowing them to pass easily through the pulmonary circulation. The bilayer phospholipids can be utilized for coupling molecules to the external portion of the vesicle. We have shown that by coupling antibodies to the surface, these liposomes can target specific molecules (17). If conjugated to site-specific antibodies, then both conjugated and unconjugated liposomes would be applicable for systemic injection, delivery and targeting. Our data demonstrate that the mean gray scale of antifibrinogen liposomes attached to fibrous atheroma initially was the same as the mean gray scale of the same liposomes after 15 min in vivo.
We have demonstrated the ability of these liposomes to target thrombi and atherosclerotic plaques in an animal model of atherosclerosis. By changing the antibody on the surface, the liposomes can target different structures and may enhance the diagnostic differentiation of plaque morphologic characteristics. This may aid in tailoring therapy to atheroma at various stages of plaque development. Identifying topographical and surface features can help decide if drug therapy or mechanical therapy would be more appropriate. Incorporation of thrombolytic drugs into these liposomes for delivery to the diseased tissue is a potential future use. Although this system may have to be injected near the site of interest, targeted liposomes could be injected intravenously if they were conjugated to a ligand that is present only on atherosclerotic plaques.
There are currently no targeted ultrasound contrast agents commercially available. A few other groups, however, have begun investigation into the area. A biotinylated, lipid-coated, perfluorocarbon emulsion that can target thrombi is in development (25). This system, however, involves a three-step targeting approach in which a biotinylated ligand, avidin, and the biotinylated perfluorocarbon emulsion particle must each reach the target to be effective. Preliminary studies have shown enhancement of the target only when the area of interest has first been incubated with the first two steps in situ. Our system allows for injection through the imaging catheter for immediate enhancement without incubation or additional injections. Targeting of perfluorobutane-gas–filled lipid microspheres to thrombi has been demonstrated in preliminary in vitro studies (26). Our study demonstrates the ability of antibody-conjugated liposomes to target thrombi and atheroma in an animal model under constant blood flow after local injection.
Factors influencing our results
Our liposomes visually remained reflective for up to 15 min with transvascular imaging. We have previously demonstrated and reported in vivo retention of acoustic properties in solution for greater than 15 days and as a freeze-dried powder with subsequent reconstitution into solution for greater than 9 weeks. We have not yet performed a systematic study to determine the duration of acoustic properties in vivo. The exact concentration and methods of injection for maximum acoustic efficacy remain to be determined. Methods to make these liposomes for sterile injection have yet to be developed.
Implications of our data
The implications of these preliminary data are far reaching. These liposomes have the potential to be conjugated to a variety of proteins that can target various molecules and aid in the diagnosis and characterization of many pathologies. This system also has the potential for drug delivery to the target site. Although these liposomes have not been loaded with drugs, the use of liposomes as drug carriers has been widely described in literature (27). As this liposomal formulation is inherently acoustic and contain no gas, loading them with pharmaceutical agents is feasible. Either hydrophilic or lipophilic drugs could be encapsulated into the liposomes, and targeted to specific tissues through the coupling of antibodies, or other suitable molecules, to the outside of the liposomes. Our lab is currently performing DNA transfer with antibody-conjugated liposomes (28). These acoustic liposomes may therefore serve in the future for both detection and treatment of atheroma.
The authors wish to thank Cynthia Shane for her expert preparation of the manuscript.
☆ Supported in part by the National Institutes of Health HL-46550, the American Heart Association Chicago Affiliate and the Feinberg Cardiovascular Research Institute.
Presented in part at the American Heart Association National Meetings, New Orleans, Louisiana, November 1996.
- brightness area product
- intercellular adhesion molecule-1
- maleimido-4 (p-phenylbutyrate)-phosphatidylethanolamine
- 3-(2-pyridyldithiolpropionic acid N-hydroxysuccinimide ester
- Received April 22, 1998.
- Revision received September 25, 1998.
- Accepted November 5, 1998.
- American College of Cardiology
- Vlodaver Z.,
- French R.,
- Van Tassel R.A.,
- et al.
- McPherson D.D.,
- Sirna S.J.,
- Hiratzka L.F.,
- et al.
- Pasterkamp G.,
- Schoneveld A.H.,
- van Wolferen W.,
- et al.
- Miskolczi L.,
- Guterman L.R.,
- Flaherty J.D.,
- et al.
- Ng K.H.,
- Evans J.L.,
- Vonesh M.J.,
- et al.
- Benenati J.F.
- Strandness D.E.
- McPherson D.D.,
- Sirna S.J.,
- Haugen J.A.,
- et al.
- Joakimsen O.,
- Bonaa K.H.,
- Jones J.P.,
- Chandraratna P.A.N.,
- Tak T.
- Blann A.D.,
- Seigneur M.,
- Steiner M.,
- Miller J.P.,
- McCollum C.N.
- Martin F.J.,
- Papahadjopoulos D.
- Storm G.,
- Oussoren C.,
- Peeters P.A.M.,
- Barenholz Y.
- Lanza G.M.,
- Wallace K.D.,
- Scott M.J.,
- et al.
- Crommelin D.J.A.,
- Schreier H.
- Tiukinhoy S.,
- Dagar S.,
- Shively V.,
- et al.