One of the main problems impeding the widespread use of AmBisome™ is its considerable cost. A contributing factor to its high cost is the use of cholesterol that is purified from animal sources. The animal source of cholesterol raises the concern of viral or prion contamination (http://avantilipids.com/index.php?view=items&cid=3&id=4&option=com_quickf aq&Itemid=385
) and makes the cost of pyrogen-free injectable grade cholesterol very high. Recently new phospholipids termed sterol-modified lipids have been synthesized in which the sterol is covalently attached to the glycerol backbone (Huang and Szoka, 2008
; Huang et al., 2009
). Liposomes prepared from SML have greater bilayer stability and the sterol in a SML liposomes does not exchange (Huang and Szoka, 2008
; Huang et al., 2009
). In the synthesis of DSHemsPC, stigmasterol a plant sterol is used. Using stigmasterol, instead of cholesterol, might decrease the overall cost of the AmB preparation. Therefore, we used DSHemsPC liposomes to formulate AmB and evaluated if the SML liposomes provided a more stable AmB preparation with lower toxicity and better efficacy than currently available formulations ().
We systematically altered the mole ratios of DSHemsPC, phosphatidylcholine and phosphatidylglycerol with different aliphatic chain lengths (14, 16, 18 and mixed chain lengths HSPC) to prepare the liposomal formulation of AmB. HSPC provides a mixture of C16 and C18 aliphatic chain length (Matsumori et al., 2002
) and HSPC is used in the formulation of many of the available commercial liposomal formulations (e.g. AmBisome™ and Doxil™) and has a very good safety profile (Torchilin, 2005
). The starting point for the selection of molar ratio for each aliphatic chain length and HSPC was the formulation of AmBisome™ which is composed of Cholesterol/HSPC/DSPG/AmB in a molar ratio of 2.5/5/2.0/1.0 ((Adler-Moore and Proffitt, 2002
; Dupont, 2002
). The lipid molar ratios of F4, F8, F12 and F16 formulations in is the same as molar ratio of lipids in AmBisome™ (each mole of DSHemsPC provides two mole of stigmasterol, the plant sterol). In the other formulations (F1–F3, F5–F7, F9–F11, F13–F15), the molar ratio of DSHemsPC, PC and AmB was kept constant and only the molar ratio of PGs was decreased to determine the effect of different molar ratio of negative charge in the formulations. In formulations F17–20 the molar ratio of DSHemsPC was increased from 1.25 to 1.75 to evaluate the effect of higher molar ratio of stigmasterol in the formulations. Most of the formulations had colloidal properties like AmBisome™ ().
One of the mechanisms of stabilization of AmB in the liposomal bilayer is through electrostatic interaction between the positive charge of the mycosamine group in AmB and the negative charge on the PG (Jensen et al., 1999
; Ellis, 2002
; Guo et al., 1991
). Therefore, we hypothesized decreasing the pH of formulations would strengthen this electrostatic interaction and result to a better formulation of AmB. Based on their physical properties, potassium release properties and MTD results; F4, F8, F11, F12, F16 and F19 formulations were selected and prepared in DSSH buffer pH 6.5 () and pH 5.5 (). Most of these formulations had colloidal properties like AmBisome™.
The particle diameter of control empty liposome (F33) was less than F29 and most of the other DSHemsPC-AmB formulations. The reason behind this could be the absence of AmB in liposome bilayer in this formulation. AmB intercalates in the liposome bilayer and it is usually hard to decrease the size of AmB-liposomes.
There are a few methods for the determination of toxicity of AmB formulations. Animal lethality test in mice and in vitro
incubations of formulations with red blood cells are the most commonly used methods (Jensen et al., 1999
; Espada et al., 2008
; Szoka et al., 1987
). The red blood cell tests are based on the effect of AmB increases the leakage of intracellular constituents such as potassium or hemoglobin from RBC (Butler and Cotlove, 1971
). In this study the potassium release assay was used to assess the toxicity of formulations since potassium release has a good correlation with the animal lethality test (Jensen et al., 1999
Among commercially available AmB formulations, Fungizone™ is the most toxic and AmBisome™ is the least toxic (Espada et al., 2008
; Larabi et al., 2004
). The RBCPR studies herein also showed that Fungizone™ was much more toxic than AmBisome™ and the results obtained in this study for these two AmB commercial products was almost the same as previously reported (Jensen et al., 1999
). Among the DSHemsPC-AmB formulations prepared in HEPES buffer pH 7.4, F9- F12 formulations, which also had good vesicle properties (near 100 nm with a monomodal distribution and a low polydispersity index), showed higher IC50 for RBCPR compared to AmBisome™ (). The higher IC50 indicates less potassium leakage, hence lower toxicity of AmB for RBCs. The lower toxicity may indicate higher affinity of AmB for the lipid bilayer in the DSHemsPC-AmB formulations than for the lipid bilayer of AmBisome™. F5, F6 and F13 had also high IC50 for RBCPR but in these formulations the lesser toxicity is probably due to the multilamellar structures of formulations. Since AmB in the MLV structure has to pass several aqueous phases and bilayers of liposomes to reach the RBCs for its membrane disturbing activity and the drug is not readily available to interact with RBCs, the RBCPR activity of Formulation F5, F6 and F13 is low.
For F4, F8, F11, F12, F16 and F19, when the pH of formulations decreased to 6.5 and 5.5 the IC50 for RBCPR increased ( and ). The reason for higher RBCPR IC50 and lower toxicity of formulations could be the higher affinity of AmB for PGs in liposome bilayer at the lower pH.
The results of in vitro
antifungal activity of DSHemsPC-AmB formulations showed that most of formulations have comparable antifungal efficacy compared to AmBisome™ and Fungizone™ (). The leishmanicidal activities of the DSHemsPC-AmB formulations were tested against both the extracellular promastigote and the intracellular amastigote forms of the parasite. The ED50 of DSHemsPC-AmB formulations, AmBisome™ and Fungizone™ against promastigotes were greater than when tested versus the amastigotes. In general, intramacrophage amastigotes are more susceptible to AmB than promastigotes (Croft, 2001
; Yardley and Croft, 1997
). The ED50s of certain formulations (F9–F12) were comparable with the commercial liposomal AmB (AmBisome™). There were no significant (P
> 0.05) differences in the activities of the F1–F4, F9–F14, F19, F20, F22 and F23 formulations against L. major
amastigotes compared to AmBisome™. The ED50s of F5–F8, and F15–F18 were significantly (P
< 0.001) greater than AmBisome™ and Fungizone™. The ED50 of F21, F24–F30, and F32 was similar (P
> 0.05) to Fungizone™. The assays with promastigotes, amastigotes and antifungal also demonstrated that the processes used for the preparation of DSHemsPC formulation do not affect on the activity of the AmB.
In order to compare directly the in vivo toxicities of DSHemsPC-AmB formulations with AmBisome™ and Fungizone™, MTD was estimated in mice for formulations with favorable vesicle characteristics and RBCPR IC50 results (). After a single i.v. injection, the MTD observed for Fungizone™ was 2 mg/kg. Formulations F8, F15, F16 and F22 behaved in the MTD response like the micelle formulations of AmB and were as toxic as Fungizone™. Mice that received these formulations died in less than a minute. The reason for high toxicity of these formulations could either be due to the weak attachment of AmB to the liposomes bilayer with the consequence that AmB is released very fast when the formulation is in circulation. Alternatively the formulations might have aggregated when injected and resulted in hemostasis.
Interestingly the F11 formulation, which was prepared at pH 7.4 had an MTD of 10 mg/kg, when prepared in pH 6.5 (F23) the same formulation had an LD50 of 20 mg/kg and when prepared in pH 5.5 (F29), the MTD was increased to 60 mg/kg.
The decrease in toxicity of this composition prepared at lower pH was correlated with a decrease in the zeta potential of the formulations F11 = 69 mV, F23 = 42 mV and F29 = 25 mV(, , ). Of the three different commercialized lipid-based formulations, Amphotec™ and Abelcet™ have LD50 about 30–32 mg/kg and AmBisome™ has an LD50 about 160 mg/kg (Jensen et al., 1999
). Formulation F29 has less acute toxicity when compared to Amphotec™ and Abelcet™; however, its MTD is less than AmBisome™.