Cytokines in liposomes: preliminary studies with IL-1, IL-2, IL-6, GM-CSF and interferon-gamma.

A simple, generally applicable method to incorporate cytokine proteins into multilamellar liposomes is presented. A variety of human cytokines including granulocyte-macrophage colony stimulating factor (GM-CSF), interleukins 1 alpha, 2 and 6 (IL-1 alpha, IL-2, IL-6) and interferon-gamma (IFN-gamma) were incorporated into liposomes containing a single saturated synthetic lipid, dimyristoyl phosphatidyl choline (DMPC). Sterile cytokine liposomes were produced by gamma irradiation of DMPC lipid powder prior to use in cytokine liposome synthesis. A highly sensitive and reliable fluorescamine assay to detect microgram quantities of cytokine protein associated with liposomes is also described. When a high lipid:aqueous ratio [e.g. 300 mg DMPC lipid:1.0 ml aqueous cytokine solution] was utilized, aqueous cytokines (1 mg/ml) could be incorporated with efficiencies ranging from 19% (IL-1) to > 80% (IL-2). Combinations of cytokines (e.g. IL-2 + GM-CSF) were also co-incorporated into liposomes. Experiments with IL-2, IL-6, and GM-CSF demonstrated that these cytokines remain stably associated with the DMPC lipid and do not significantly leak from liposomes when stored at 4 degrees C for at least 3 months. Washing IL-6 liposome or GM-CSF liposome preparations reliably increased the proportion of cytokine protein associated with the liposome pellet compared to free cytokine in the supernatant of centrifuged specimens. For example the proportion of GM-CSF associated with the lipid carrier increased from 34.8% (SD 2.6%) in the original preparation to 98.0% (SD 0.6%) after three washes. Differences in the pharmacokinetics of subcutaneous (sc) free GM-CSF and GM-CSF liposomes (14 mcg/mouse) were studied in BALB/c mice. Both free GM-CSF and free GM-CSF mixed with saline loaded liposomes exhibited biphasic pharmacokinetics with very high peak levels 1 and 2 h after sc injection of 14 mcg the rapid decline to very low levels after 24 h. In contrast, sc GM-CSF liposomes provided sustained and stable levels of cytokine in the serum (approximately 100 pg/ml) for 24 h. Intraperitoneal injection of GM-CSF liposomes had > 10-fold more cytokine in the peritoneal wash than free GM-CSF mixed with saline loaded liposomes. In summary, the liposome synthesis procedure described is simple and utilizes a single synthetic lipid to reliably produce sterile cytokine preparations with in vivo depot effects after either sc or ip administration. Furthermore, the method is feasible for quantities of sterile cytokine liposomes sufficient for in vivo experiments.