Encapsulation of Sodium Fluorescein for Dye Release Studies

Investigative Ophthalmology & Visual Science, Vol. 33, No. 7, June 1992 Copyright Association for Research in Vision and Ophthalmology Encapsulation of Sodium Fluorescein for Dye Release Studies Michael R. Niesman, Bahrain Khoobehi, and Gholam A. Peyman Recent investigations have detailed a selective dye release technique in which a pulse of laser light induces the release of a fluorescent dye from temperature-sensitive liposomes circulating in the retinal vasculature. This dye release technique has made possible a new method for measuring ocular blood flow in the retina and has spurred the development of repetitive, site-specific angiography. However, sodium fluorescein, the dye employed clinically for angiography of the retina, has not been employed in the aforementioned studies because of its rapid efflux from liposomes. This report outlines a method for stable encapsulation of sodium fluorescein in temperature-sensitive liposomes. Heat-induced leakage of the dye from liposomes in vitro was similar to that previously seen with other fluorescent dyes. Furthermore, after intravenous injection of encapsulated fluorescein in a nonhuman primate, dye released by a pulse of laser light allowed excellent visualization of the retinal architecture. These results indicate that sodium fluorescein, a dye that has proven to be the agent of choice for sensitive detection of leakage of vessels of the retina, can be released at a specific site in the retinal vasculature. Direct comparisons of the diagnostic capability of free and encapsulated sodium fluorescein are now possible. Invest Ophthalmol Vis Sci 33:2113-2119,1992 Fluorescent dyes encapsulated in liposomes have been used for many years to investigate liposome-cell interactions. 1-2 Recently, a method for the controlled release of a fluorescent dye at a predetermined location in the retinal vasculature has been described. 3 " 11 These investigations were completed by encapsulating a high concentration offluorescentdye in temperature-sensitive liposomes. The high concentration of encapsulated dye is not fluorescent, because of self quenching. 12 After the intravenous injection of the respective fluorescent dye encapsulated in liposomes, a laser pulse triggered the release of the dye at a specific location in the retinal vasculature. Release of the dye from the liposome carrier caused an increase in fluorescence, and the released dye was followed as in normal angiography. This targeted dye release method allowed investigators to measure retinal bloodflowand to perform repetitive, selective angiography of specific areas of the retina. In many previous studies of liposome-cell interaction, and in all of the earlier studies of ocular dye From the LSU Eye Center, Louisiana State University Medical Center School of Medicine, New Orleans, Louisiana. Supported in part by U.S. Public Health Service grants EY08137, EY07541, and EY02377 from the National Eye Institute, the National Institutes of Health Services, Bethesda, Maryland. Submitted for publication: May 24, 1991; accepted December 9, 1991. Reprint requests: Dr. Gholam A. Peyman, LSU Eye Center, 2020 Gravier St., Suite B, New Orleans, LA 70112-2234. release, the fluorescent dye employed was carboxyfluorescein (CF) or calcein (CAL). CF was originally suggested as preferable to sodium fluorescein for liposome-cell interaction studies because the addition of the carboxyl group greatly decreased the efflux of the dye from the small unilamellar vesicles (SUV). The half time of efflux of sodium fluorescein from SUV previously was reported to be 5 min, versus several weeks for CF. 12 CAL was suggested as a further improvement over CF because it has a greater net charge (-3 at ph 7) and does not leak out of liposomes as the ph is lowered toward 6. The efflux of CF increases as the ph is lowered. 13 CF and CAL have been successfully employed in studies of repetitive angiography and retinal blood flow. However, recent investigations have shown that the qualitative staining and leakage patterns of CF and CAL (injected as free drug) are different from sodiumfluorescein. 14 When compared to sodium fluorescein, CAL and CF exhibit less leakage in areas of the retinal vasculature with increased capillary permeability (eg, areas exposed to laser energy). Sodium fluorescein is the dye that exhibits the greatest sensitivity to small changes in capillary permeability. To detect leakage from a capillary defect when only a small amount of dye is released via the laser, it would be desirable to use encapsulated sodium fluorescein. Therefore, this investigation was undertaken to determine a method for stable encapsulation of sodium fluorescein within liposomes. 2113

2114 INVESTIGATIVE OPHTHALMOLOGY b VISUAL SCIENCE / June 1992 Vol. 33 Materials Materials and Methods Phospholipids were obtained from Avanti Polar Lipids (Alabaster, AL) and used without further purification. Sodium fluorescein powder was obtained from City Chemical Company (New York, NY). A 10% sodium fluorescein solution was purchased from Alcon (Fort Worth, TX). Glycine and boric acid were supplied by Sigma (St. Louis, MO). All other chemicals were reagent grade or better. Methods Liposome formation: Liposomes were formed as previously described. 15 The lipid composition for all formulations was a 4:1 molar ratio of dipalmitoylphosphatidylcholine (DPPC) and dipalmitoylphosphatidylglycerol (DPPG). A typical batch consisted of 100 mg of DPPC, 25 mg of DPPG, and 4 ml of the aqueous solution to be entrapped. Several buffered solutions were formulated for encapsulation. They were (1) 100 mmol/1 sodium fluorescein, 100 mmol/1 glycine, ph 7; (2) 100 mmol/1 sodium fluorescein, 100 mmol/1 glycine, ph 10; and (3) 100 mmol/1 sodium fluorescein, 95 mmol/1 borate, ph 10. The osmolality of the three buffered-dye solutions ranged between 305 and 340 mosm. After formation, the preparations were dialyzed against 4 X 250 vol excess of HEPES-buffered saline (140 mmol/1 NaCl, 6 mmol/1 KC1, 10 mmol/1 HEPES, ph 7.4, 295 osm). Calculation of liposome dye capture: After extensive dialysis, aliquots of liposomes were diluted 25 times in distilled water. Aliquots of the liposomes and aqueous buffered-dye solutions were dissolved in 3 ml methanol, which was acidified with 20 n\ 1 N HC1. Known concentrations of sodium fluorescein were used to construct a standard curve in acidified methanol, and the concentration of the dye in the liposome aliquots was determined at 440 nm, the wavelength of maximal absorbance of fluorescein in acidified MeOH. The amount of encapsulated dye was compared to the measured concentration present in the aqueous phase. The percent encapsulation was calculated as: amount of liposome associated Na-fluorescein X 100 (1) amount of Na-fluorescein aqueous phase Dye release studies: The efflux of sodium fluorescein was measured in two different assays: one to measure the efflux of dye versus temperature and one to measure the leakage over time at 37 C. The leakage of dye versus temperature was assayed in a manner similar to that previously described. 15 Briefly, this involved diluting aliquots of the liposome suspension to 10% of their original concentration in HEPES-buffered saline or fetal calf serum. The two diluted liposome suspensions then were placed in a temperature-controlled water bath (Haake model D-1; Baxter Scientific, Chicago IL). The liposome suspensions were incubated at a constant temperature for 5 min at 35 C. A 150 ^1 aliquot of diluted liposomes then was removed and the liposomes were immediately pelleted at approximately 75,000 X g in a Beckman Airfuge (Beckman Instruments, Palo Alto, CA). The temperature of the water bath was increased in a stepwise manner and the liposomes were incubated for an additional 5 min at each temperature. The leakage was measured at 35, 39, 40, 41, 43, and 45 C. After the liposomes were pelleted, 100 ix\ aliquots of the supernatant were removed from the centrifuge tubes and 50 /A aliquots were assayed to determine the amount of dye release. Aliquots (also 50 ix\) of known quantities of fluorescein in 90% serum or 90% buffer were used to construct the appropriate standard curve. Dilution of aliquots of released dye and standards were made with HEPES-buffered saline (ph 7.4), and the absorbance of fluorescein was recorded at 488 nm, the absorbance maximum of fluorescein in buffer at ph 7.4. To determine the leakage of dye versus time, the liposome suspensions were diluted in buffer and serum as described above (diluted to 20% of their original concentration). Then, 1.5 ml of the buffer- and serum-diluted sample were placed in a water bath at 37 C. Aliquots were withdrawn at 0.25, 0.5, 1, 4, and 24 hr after the sample reached 37 C. The released drug was isolated, diluted, and measured as described above for the temperature-release assay. For both dye release studies, the maximum possible release (100% release) was determined by subjecting the liposome suspension to three cycles of freezing in liquid nitrogen followed by thawing at 50 C. These 100% release samples then were pelleted and processed in an identical manner to the other leakage samples. Dye leakage after laser exposure: Preliminary experiments (data not shown) were performed in an in vitro exposure chamber that has been previously described. 3 A pulse of light from an argon laser elicited dye release from liposomes flowing in a transparent capillary tube. The results were similar to those seen previously with CF. 35 Because laser-induced release in vitro was similar to that seen previously with CF and CAL, an injection of liposome-encapsulated sodium fluorescein was given to one cynomolgus and one squirrel monkey. The dosage of fluorescein was 5.1 mg/kg for the cynomolgus monkey and 12.6 mg/ kg for the squirrel monkey. Laser-induced angiography was performed as previously described. 6 " 8 Briefly,

No. 7 ENCAPSULATION OF SODIUM FLUORESCEIN / Niesmon er ol 2115 for the exposures recorded for this work, the spot size of the laser was equal to the diameter of the optic disk of the squirrel monkey (approximately 1.2 mm) and the laser was focused on the optic disk, which allowed dye to be released simultaneously in all four quadrants of the retina." The laser exposure time was 1.5 sec at a laser power of 40 mw. The dye released into the retina could be photographed at a preset delay after the laser was turned on. The photographs presented are a composite of dye release elicited by two successive, separate laser firings in which the delay times were offset to allow the dye leakage to be followed. This was necessary because the motor drive on the camera requires photographs to be taken at intervals equal to or greater than 650 msec. The images were recorded with Kodak (Rochester, NY) Tri-X 400 film push processed in Kodak D-11 developer for an effective ISO of approximately 1600. The investigation adhered to the ARVO Resolution on the Use of Animals in Research. Results Table 1 lists the values for the encapsulation efficiency of the different liposome preparations. Liposomes made with borate buffer at ph 10 or with glycine buffer at ph 7 entrapped far less fluorescein than the vesicles formed with glycine buffer at ph 10. The liposomes made with an internal ph of 10 using glycine buffer were tested further to determine the temperature dependence of dye release. Previous work has shown that liposomes formed from DPPC and DPPG release their contents near the phase-transition temperature of 41 C.' 5 Figure 1 depicts the temperature dependence of dye release of liposomes diluted in HEPES-buffered saline or 80% serum (both at ph 7.4). Little dye release was seen at 37 C. The release of dye increased with temperature and was nearly complete at 41 C. The data shown here for sodium fluorescein are similar to that previously reported for CF 3 and CAL. 6 The leakage of sodium fluorescein from vesicles made with glycine buffer at ph 10 was compared to those made with buffer at ph 7. After 24 hr at 37 C, Table 1. Encapsulation liposome formulations Buffer Borate Glycine Glycine ph* 10 7 10 efficiency of different Encapsulated (% of original aqueous phase) 1.7 6.3 9.8 ± 0.7f * The ph of the dye-buffer mixture used to prepare the liposomes. t Mean plus or minus the SEM. N 1 1 6 vesicles made with ph 10 buffer and subsequently diluted in HEPES-buffered saline retained 95.5% of the entrappedfluorescein,while liposomes made with ph 7 glycine buffer retained only 80% (Figure 2). The difference was even more striking for the liposomes diluted in 80% serum, a dilution designed to mimic in vivo conditions. After 24 hr, the liposomes made with ph 10 buffer retained 70% of their contents, while in the ph 7 liposomes, all of the fluorescein was lost. Leakage was essentially 100%. Routinely, in vivo, repetitive angiography is completed in approximately 1 hr. After 1 hr at 37 C in serum, leakage of the ph 10 vesicles in vitro was only 5%, whereas the vesicles made with ph 7 had lost nearly 30% of their dye after 1 hr at 37 C. The release of dye in the retina of a squirrel monkey is depicted in Figures 3A-3D. The optic nervehead was exposed to laser energy for 1500 msec and the laser then was turned off. Photographs were recorded at various intervals before and after the laser was turned off (seefigurelegends). The dye released in the retinal vasculature was clearly visualized throughout the fundus without the interference of background fluorescence from the choroid. In Figures 3A and 3B, the laser is still on; scattering from it obscures the view of the optic nervehead. In Figure 3A, only the arterial circulation is visible as the dye begins to spread out into the retinal vasculature. By Figure 3D, much of the dye has left the arteries, and drainage of the dye in the venous system is seen. Discussion Early work on targeted drug delivery indicated that highly polar, hydrophilic substances were well retained in the aqueous compartment of the vesicles. Likewise, extremely hydrophobic, nonpolar substances intercalated in the lipid bilayer and remained tightly bound to the liposomes. However, compounds of intermediate polarity (eg, doxorubicin, sodium fluorescein) were difficult to entrap because of their solubility in the bilayer and surrounding aqueous media. However, recent reports have shown that it is possible to entrap lipophilic cations in liposomes. 16 " 18 The strategy employed has been to take preformed liposomes and generate a transmembrane potential (A^) 16 or ph gradient (ApH). 17-18 When the liposomes are incubated with the lipophilic cation, the drug is accumulated inside the vesicle in response to lower internal ph or A^ (inside negative). Apparently, all the agents reported in the literature that have been encapsulated in response to a ph or an electrochemical gradient have been lipophilic cations, such as Adriamycin or dibucaine. It seemed plausible that the same strategy would work with a lipophilic

2116 INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / June 1992 Vol. 33 35 37 39 41 TEMPERATURE ( C) 43 45 Fig. 1. The release of sodiumfluoresceinas a function of temperature. The release was tested for liposomes diluted in buffer (circles) and fetal calf serum (diamonds). anion such as fluorescein. It had been previously shown that when Erlich ascites tumor cells were incubated with valinomycin, an ionophore that induces a large, basic ph shift in the mitochondrial matrix, intracellular fluorescein becomes concentrated in the matrix of the mitochondria. 19 The authors of that study calculated that a ApH of 1 resulted in 95% of the total cellular fluorescein concentrated in the mitochondria. Therefore, it seemed logical that if liposomes were formed with a ph gradient across the membrane and with the inside of the vesicle basic, sodium fluorescein could be encapsulated and retained for an extended period in liposomes. In previous work in which a ph gradient was used to load lipophilic cations in liposomes, 16 " 18 the maximum internal concentration of drug was achieved when the liposomes were loaded by incubating the vesicles in drug containing solution maintained at a temperature above the phase-transition temperature (T m ) of the component phospholipids. However, the temperature-sensitive liposomes employed in this study could not be incubated at or above the phase-transition temperature of DPPC and DPPG (T m = 41 C), because at these temperatures the internal contents leak rapidly and equilibrate with the external media, 15 thereby destroying the ph gradient. To overcome this drawback, the liposomes made for this study were formed at ph 10 with a buffer containing a high concentration (100 mmol/1) of glycine. The buffer also contained a high concentration of sodium fluorescein (100 mmol/1). The pka of the amine group of glycine has been reported to be 9.6. 20 Therefore, the internal ph will be highly buffered against change. The carboxyl group (pka ~ 2.4) of glycine ensures that the buffer molecule remains charged over the entire ph range the vesicles will likely encounter, preventing glycine from leaking out of the vesicles at temperatures below T m. Thus, the glycine buffer should remain entrapped within the liposome, maintaining a high ph environment. Borate, another buffer used in this study, had a pka of 9.8. However, below the pka it is uncharged. This may account for the poor entrapment of fluorescein when borate buffer was used (Table 1). The pkas for the two titratable groups on sodium fluorescein have been reported to be pka, ~ 4.5 and pka 2 = 6.5. 17 It is possible to determine the ratio of sodium fluorescein molecules with two charged groups to those with one charge by employing the Henderson-Hasselbach equation: 21 LJJ o < LU 110 100 90 80 70 60 50 40 30 20 10 0 O Buffer ph 10 Serum ph 10 A Buffer ph 7 A Serum ph 7 Fig. 2. The leakage of sodium fluorescein from liposomes incubated at 37 C for 24 hr. One batch of liposomes was formed from an aqueous phase containing glycine buffer and sodium fluorescein at ph 7 (O, ), and a second batch was made with the glycine buffer; sodium fluorescein aqueous adjusted to ph 10 (A, A). The liposomes were diluted in HEPES-buffered saline, ph 7 (O, A), or fetal calf serum (, A). 8 12 16 20 24 TIME (hr)

No. 7 ENCAPSULATION OF SODIUM FLUORESCEIN / Niesman er al 2117 Fig. 3. The photographic record of sodium fluorescein release and drainage in the retina of a squirrel monkey. This is a combination of photographs taken after two individual laser shots to the retina. The laser was fired 400 msec after the peak of systole. The total laser exposure per shot was 1.5 seconds. (A) Taken 900 msec after laser firing, with the laser still on. (B) Dye release 1500 msec after the initiation of dye release. The laser is in the process of being turned off. (C) The retina 1650 msec after laser was fired, 150 msec after laser was turned off. (D) The retina 1 second after (C), 2.65 seconds after the lasertriggered dye release. ph = pka + log10 [unprotonated(base)/ protonated(acid)] (2) If, for calculation's sake, the vesicle interior is assumed to be ph 9.5, and using a pka2 of 6.5 for the second titratable group, the ratio of sodium fluorescein molecules with a - 2 charge versus those with a net charge o f - 1 is approximately 1000 to 1. Thus, even after the dialysis of the liposomes against several changes of ph 7.4 buffer, a high internal ph ensures that most of the encapsulated dye retains a net surface charge of - 2, which helps the dye remain encapsulated. The leakage data shown in Figure 2 indicate that the ph of the buffer used to form the liposomes is important in ensuring that fluorescein is retained in the vesicles. Liposomes made with fluorescein and glycine buffer at ph 7 retained little fluorescein after

2118 INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / June 1992 Vol. 33 24 hr in serum, while the liposomes made with ph 10 buffer retained 70% of their contents. Leakage as measured after 24 hr in buffer was approximately 4.5 times higher when the low ph buffer was used. When the liposomes made with sodium fluorescein buffered at ph 10 were injected in monkeys, the dye was not immediately visible in the circulation. This finding indicated that any leakage from the vesicles on exposure to the blood was minor or absent, in agreement with the in vitro leakage data. However, when the argon-ion laser was fired at the retina, dye release could be visualized in the retina. The pattern of dye release was similar to that seen previously with CAL and CF. The transit time of the dye through the retina seems slower when compared to the previously published work with CF. However, much more work in vivo will be needed to confirm this fact. The results shown here highlight the fact that the arterial phase and the venous phase of the dye transit can be well differentiated. As has been seen in the past with CF and CAL, the visualization of the micro vasculature is superior when the fluorescein in the choroid is essentially absent. One of the most important potential advantages of encapsulated sodium fluorescein is that it is more likely to leak from small capillary defects than are CF or CAL. 14 If only a small aliquot of dye is released into the vasculature of the retina, it would be advantageous to employ the dye that is most sensitive to capillary leakage. Thus, this formulation of encapsulated sodium fluorescein may prove superior to entrapped CF or CAL for the detection of early changes in capillary permeability. Much more work will be required to document these first impressions carefully and to determine whether sodium fluorescein released from liposomes will allow the detection of small capillary defects as effectively as free fluorescein. If subsequent work in primates proves that encapsulated fluorescein is capable of sensitive detection of early abnormalities in the retinal circulation, especially in the capillaries, liposome-encapsulated sodium fluorescein has the potential to become an important technique for diagnosing abnormalities of the vasculature of the human retina. Extensive studies that examine the systemic toxicity of this encapsulated drug would be needed before this formulation could be used in humans. Although a slightly higher level of energy was used in the present study, the recently reported method that involves exposure of the optic nerve to the pulse of laser energy has reduced the amount of laser energy required to elicit dye release to levels safe for use in humans." Because the internal volume of the liposomes injected systemically is such a small volume compared to the volume of blood, it is not likely that the high internal ph of the vesicles will cause problems with blood ph. The accumulation of vesicles in the liver and spleen, 22 which will carry a large portion of encapsulated, unreleased fluorescein to those organs, is likely to be the dose-limiting toxicity. That the drug itself currently is approved for use in humans is an advantage compared to CF or CAL. Neither of these agents is approved for use in humans and both would have to be demonstrated to be safe in free form before an encapsulated form was tested. Future testing of encapsulated fluorescein will reveal whether it can be successfully employed in humans. In the interim, the encapsulated formation will allow experiments to be completed in animals, which will make it possible to determine whether the encapsulated form is as effective or more effective at diagnosing retinal pathology compared to the free form. This must be proven if the liposome-drug combination is to be used in humans. Key words: liposomes, sodiumfluorescein,retinal blood flow References 1. Heath TD: Interaction of liposomes with cells. In Methods in Enzymology: Drug and Enzyme Targeting, Part B, Widder KJ and Green R, editors. New York, Academic Press, 1987, vol. 149, pp. 135-143. 2. Schroit AJ, Madsen J, and Nayar R: Liposome-cell interactions: In vitro discrimination of uptake mechanism and in vivo targeting strategies to mononuclear phagocytes. Chem Phys Lipids 40:373, 1986. 3. Zeimer RC, Khoobehi B, Niesman MR, and Magin RL: A potential method for local drug and dye delivery in the ocular vasculature. Invest Ophthalmol Vis Sci 29:1179, 1988. 4. Khoobehi B, Peyman GA, Niesman MR, and Oncel M: Measurement of retinal blood velocity and flow rate in primates using a liposome-dye system. Ophthalmology 96:905, 1989. 5. Zeimer RC, Khoobehi B, Peyman GA, Niesman MR, and Magin RL: Feasibility of blood flow measurement by externally controlled dye delivery. Invest Ophthalmol Vis Sci 30:660, 1989. 6. Khoobehi B, Niesman MR, Peyman GA, and Oncel M: Repetitive, selective angiography of individual vessels of the retina. Retina 9:87, 1989. 7. Khoobehi B, Aly OM, Schuele K, Stradtmann MO, and Peyman GA: Determination of retinal blood velocity with respect. to the cardiac cycle using laser-triggered release of liposome-encapsulated dye. Lasers Surg Med 10:469, 1990. 8. Khoobehi B, Schuele KM, Aly OM, and Peyman GA: Measurement of circulation time in the retinal vasculature using selective angiography. Ophthalmology 97:1061, 1990. 9. Zeimer RC, Guran T, Shahidi M, and Mori MT: Visualization of the retinal microvasculature by targeted dye delivery. Invest Ophthalmol Vis Sci 31:1459, 1990. 10. Guran T, Zeimer RC, Shahidi M, and Mori MT: Quantitative analysis of retinal hemodynamics using targeted dye delivery. Invest Ophthalmol Vis Sci 31:2300, 1990. 11. Khoobehi B, Peyman GA, and Cruz S: Repetitive total fluorescein angiogram using externally triggered liposome-encapsulatedfluoresceindye. Invest Ophthalmol Vis Sci 32(suppl):866, 1991.

No. 7 ENCAPSULATION OF SODIUM FLUORESCEIN / Niesmon er ol 2119 12. Weinstein JN, Yoshikami S, Henkart P, Blumenthal R, and Hagins WA: Liposome-cell interaction: Transfer and intracellular release of a trapped fluorescent marker. Science 195:489, 1977. 13. Allen TM: Calcein as a tool in liposome methodology. In Liposome Technology, Gregoriadis G, editor. Boca Raton, FL, CRC Press, 1983, vol. Ill, pp. 177-182. 14. Fang T, Naguib KS, Peyman GA, and Khoobehi B: Comparative study of three fluorescent dyes for angiography: Sodium fluorescein, carboxyfluorescein and calcein. Ophthalmic Surg 21:250, 1990. 15. Magin RL and Niesman MR: Temperature-dependent permeability of large unilamellar liposomes. Chem Phys Lipids 34:245, 1984. 16. Bally MB, Hope MJ, Van Echteld CJA, and Cullis PR: Uptake of safranine and other lipophilic cations into model membrane systems in response to a membrane potential. Biochim BiophysActa812:66, 1985. 17. Mayer LD, Bally MB, and Cullis PR: Uptake of adriamycin into large unilamellar vesicles in response to a ph gradient. Biochim Biophys Acta 857:123, 1986. 18. Mayer LD, Wong KF, Menon K, Chong C, Harrigan PR, and Cullis PR: Influence of ion gradients on the transbilayer distribution of dibucaine in large unilamellar vesicles. Biochemistry 27:2053, 1988. 19. Thomas JA, Kolbeck PC, and Langworthy TA: Spectrophotometric determination of cytoplasmic and mitochondrial ph transitions using trapped ph indicators. In Intracellular ph, Its Measurement, Regulation and Utilization in Cellular Function, Nuccitelli R and Deamer DW, editors. New York, AR Liss, 1982, pp. 105-123. 20. Lehninger AL: Biochemistry: The Molecular Basis of Cell Structure, 2nd ed. New York, Worth, 1975, p. 79. 21. Clark JM Jr and Switzer RL: Experimental Biochemistry, 2nd ed. San Francisco, WH Freeman, 1977, pp. 68-69. 22. Senior JH: Fate and behavior of liposomes in vivo: A review of controlling factors. Crit Rev Ther Drug Carrier Syst 3:123, 1987.