Nigel C. Phillips

Toxicity and immunomodulatory activity of liposomal vectors formulated with cationic lipids toward immune effector cells.

Liposomal vectors formulated with cationic lipids (cationic liposomes) and fusogenic dioleoylphosphatidylethanolamine (DOPE) have potential for modulating the immune system by delivering gene or antisense oligonucleotide inside immune cells. The toxicity and the immunoadjuvant activity of cationic liposomes containing nucleic acids toward immune effector cells has not been investigated in detail. In this report, we have evaluated the toxicity of liposomes formulated with various cationic lipids towards murine macrophages and T lymphocytes and the human monocyte-like U937 cell line. The effect of these cationic liposomes on the synthesis of two immunomodulators produced by activated macrophages, nitric oxide (NO) and tumor necrosis factor-alpha (TNF-alpha), has also been determined. We have found that liposomes formulated from DOPE and cationic lipids based on diacyltrimethylammonium propane (dioleoyl-, dimyristoyl-, dipalmitoyl-, disteroyl-: DOTAP, DMTAP, DPTAP, DSTAP) or dimethyldioctadecylammonium bromide (DDAB) are highly toxic in vitro toward phagocytic cells (macrophages and U937 cells), but not towards non-phagocytic T lymphocytes. The rank order of toxicity was DOPE/DDAB > DOPE/DOTAP > DOPE/DMTAP > DOPE/DPTAP > DOPE/DSTAP. The ED50s for macrophage toxicity were < 10 nmol/ml for DOPE/DDAB, 12 nmol/ml for DOPE/DOTAP, 50 nmol/ml for DOPE/DMTAP, 400 nmol/ml for DOPE/DPTAP and > 1000 nmol/ml for DOPE/DSTAP. The incorporation of DNA (antisense oligonucleotide or plasmid vector) into the cationic liposomes marginally reduced their toxicity towards macrophages. Although toxicity was observed with cationic lipids alone, it was clearly enhanced by the presence of DOPE. The replacement of DOPE by dipalmitoylphosphatidylcholine (DPPC) significantly reduced liposome toxicity towards macrophages, and the presence of dipalmitoylphosphatidylethanolamine-PEG2000 (DPPE-PEG2000: 10 mol%) in the liposomes completely abolished this toxicity. Cationic liposomes, irrespective of their DNA content, downregulated NO and TNF-alpha synthesis by lipopolysaccharide (LPS)/interferon-gamma (IFN-gamma)-activated macrophages. The replacement of DOPE by DPPC, or the addition of DPPE-PEG2000, restored NO and TNF-alpha synthesis by activated macrophages. Since macrophages constitute the major site of liposome localization after parenteral administration and play an important role in the control of the immune system, cationic liposomes should be used with caution to deliver gene or antisense oligonucleotide to mammalian cells. Cationic lipids show in vitro toxicity toward phagocytic cells and inhibit in vitro and in situ NO and TNF-alpha production by activated macrophages.

Major limitations in the use of cationic liposomes for DNA delivery

Liposomal vectors formulated with cationic lipids and the fusogenic phospholipid dioleyolphosphatidylethanolamine (DOPE) are usually used to target DNA inside mammalian cells. Since macrophages constitute the major site of liposome localisation after parenteral administration we felt it prudent to examine the effect of cationic liposomes on the production of several important immuno-inflammatory modulators secreted by activated macrophages. In addition, we have evaluated the toxicity of different cationic liposome formulations towards phagocytic macrophages and non-phagocytic T-lymphocytes. Our results indicate that cationic liposomes are able to down-regulate the synthesis of the protein kinase C (PKC)-dependent mediators nitric oxide (NO), tumour necrosis factor-α (TNF-α) and prostaglandin E2 (PGE2) by activated macrophages after in vitro incubation under non-toxic conditions or after in vivo treatment, while the production of PKC-independent IL-6 is not modified. We have shown that cationic lipids possess potent anti-inflammatory activity in vivo. Prolonged incubation (>3 h) of macrophages with cationic liposomes induced a high level of toxicity (ED50 1000 nmol/ml). The rank order of toxicity was DOPE/dimethyldioctacylammonium bromide (DDAB)>DOPE/dioleyoltrimethylammonium propane (DOTAP)=DOPE/dimethylaminoethanecarbamoyl cholesterol (DC-Chol)>DOPE/dimyristoyltrimethylammonium propane. The replacement of DOPE by dipalmitoylphosphatidylcholine (DPPC) or the incorporation of dipamitoylphosphatidylethanolamine-PEG2000 (DPPE-PEG2000) in DOPE/cationic lipids reduced the toxicity toward macrophages and restored the synthesis of PKC-dependent modulators. The incorporation of DNA, either as an antisense oligonucleotide (15-mers) or as the plasmid vector pBR322 (4363 bp), in cationic liposomes did not reduce these adverse effects. These results, in addition to the observation that cationic liposomes are extremely toxic following oral administration, indicate that DOPE/cationic lipid liposomes are not appropriate for DNA (or drug) delivery.

Drug retention and stability of solid lipid nanoparticles containing azidothymidine palmitate after autoclaving, storage and lyophilization

Solid lipid nanoparticles (SLNs) were prepared using trilaurine (TL) as the SLN core and phospholipid (PL) as coating. Neutral and negatively charged PLs were used to produce neutral and negatively charged SLNs. An ester prodrug of 3-azido-3-deoxythymidine (Zidovudine, AZT), AZT palmitate (AZT-P), was synthesized and incorporated in the SLNs. The stability of SLN formulations containing AZT-P was studied at different temperatures. Drug retention and mean particle diameter of SLNs were determined after autoclaving, during temperature stability testing, and after lyophilization (with or without cryoprotective sugars) and reconstitution. There were no significant changes in the mean diameter and the zeta potential (zeta) of SLNs after autoclaving (121 degrees C for 20 min). The amount of incorporated AZT-P was, however, slightly reduced due to the formation of hydrosoluble AZT. Autoclaved SLNs were stable for a period of 10 weeks at 20 degrees C but an increase in particle size and loss of AZT-P were observed at 4 and 37 degrees C. Trehalose was an effective cryoprotectant for preventing SLN aggregation during lyophilization and subsequent reconstitution. Thermal gravimetric analysis showed that lyophilized preparations contained approximately 1% water. Using appropriate trehalose to lipid ratios, AZT-P retention in the SLNs was 100% after reconstitution. Our results demonstrate that SLNs containing AZT-P can be autoclaved, lyophilized and reconstituted without significant changes in SLN diameter and zeta potential or in the quantity of incorporated prodrug.

Influence of surface properties at biodegradable microsphere surfaces : Effects on plasma protein adsorption and phagocytosis

AbstractPurpose. The objective of this work was to determine plasma protein adsorption and macrophage phagocytosis of biodegradable polyanhydride, polylactic acid and polylactic-co-glycolic acid microspheres prepared by both spray-drying and solvent evaporation techniques.nMethods. Microspheres were characterized by scanning electron microscopy (SEM), confocal laser microscopy, particle size distribution and zeta (ζ) potential determination. Plasma protein adsorption onto the microspheres was determined using a fluoroaldehyde reagent. Phagocytosis was evaluated by incubating microspheres containing the angiotensin II antagonist, L-158,809, with the macrophages in the presence or absence of the phagocytosis inhibitor cythochalasin D. The extent of phagocytosis was established by fluorescence determination of L-158,809 and by optical microscopy. The effect of amphiphilic poly(ethylene glycol) (PEG) derivatives on phagocytosis was determined using PEG-distearate incorporated into the microspheres.nResults. The average diameter of the microspheres, which depended on the polymer and the initial formulation, ranged from 0.9 to 3.2 micrometers. ζ potential studies showed strong negative values irrespective of the polymer used for the spray-dried formulations. The ζ potential was masked by the incorporation of PEG 400- or PEG 1,400-distearate in the formulation. Confocal laser microscopy showed a homogenous dispersion of PEG (measured as PEG-fluorescein) in the microspheres. Protein adsorption was not observed for any of the microsphere formulations following incubation with bovine serum. Incubation of microspheres with murine macrophages showed that PEG-distearate inhibited phagocytosis at appropriate levels (0,1% w/w). Higher levels >1% w/w of PEG-distearate) resulted in enhanced association with macrophages, despite the presence of the phagocytosis inhibitor cytochalasin D, indicating fusion between the microspheres and the plasma membrane.nConclusions. These results demonstrate that spray-dried PEG-containing microspheres can be manufactured and that an appropriate concentration of this excipient in microspheres results in decreased phagocytosis.

Anti-inflammatory activity of cationic lipids.

1 The effect of liposome phospholipid composition has been assumed to be relatively unimportant because of the presumed inert nature of phospholipids. 2 We have previously shown that cationic liposome formulations used for gene therapy inhibit, through their cationic component, the synthesis by activated macrophages of the pro‐inflammatory mediators nitric oxide (NO) and tumour necrosis factor‐α (TNF‐α). 3 In this study, we have evaluated the ability of different cationic lipids to reduce footpad inflammation induced by carrageenan and by sheep red blood cell challenge. 4 Parenteral (i.p. or s.c) or local injection of the positively charged lipids dimethyldioctadecylammomium bromide (DDAB), dioleyoltrimethylammonium propane (DOTAP), dimyristoyltrimethylammonium propane (DMTAP) or dimethylaminoethanecarbamoyl cholesterol (DC‐Chol) significantly reduced the inflammation observed in both models in a dose‐dependent manner (maximum inhibition: 70–95%). 5 Cationic lipids associated with dioleyol‐ or dipalmitoyl‐phosphatidylethanolamine retained their anti‐inflammatory activity while cationic lipids associated with dipalmitoylphosphatidylcholine (DPPC) or dimyristoylphosphatidylglycerol (DMPG) showed no anti‐inflammatory activity, indicating that the release of cationic lipids into the macrophage cytoplasm is a necessary step for anti‐inflammatory activity. The anti‐inflammatory activity of cationic lipids was abrogated by the addition of dipalmitoylphosphatidylethanolamine‐poly(ethylene)glycol‐2000 (DPPE‐PEG2000) which blocks the interaction of cationic lipids with macrophages. 6 Because of the significant role of protein kinase C (PKC) in the inflammatory process we have determined whether the cationic lipids used in this study inhibit PKC activity. The cationic lipids significantly inhibited the activity of PKC but not the activity of a non‐related protein kinase, PKA. The synthesis of interleukin‐6 (IL‐6), which is not dependent on PKC activity for its induction in macrophages, was not modified in vitro or in situ by cationic lipids. The synthesis of NO and TNF‐α in macrophages, both of which are PKC‐dependent, was downregulated by cationic lipids. 7 These results demonstrate that cationic lipids can be considered as novel anti‐inflammatory agents. The downregulation of pro‐inflammatory mediators through interaction of cationic lipids with the PKC pathway may explain this anti‐inflammatory activity. Furthermore, since cationic lipids have intrinsic anti‐inflammatory activity, cationic liposomes should be used with caution to deliver nucleic acids for gene therapy in vivo.

Solid lipid nanoparticles as drug carriers I. Incorporation and retention of the lipophilic prodrug 3'-azido-3'-deoxythymidine palmitate

Abstract Solid lipid nanoparticles (SLN) were prepared with trilaurin (TL) as the SLN solid core and dipalmitoylphosphatidylcholine (DPPC) or a mixture of DPPC and dimyristoylphosphatidylglycerol (DMPG) to produce neutral and negatively charged SLNs. The ester prodrug of 3′-azido-3′-deoxythimidine (Zidovudine®, AZT) with palmitic acid, AZT palmitate (AZT-P), was synthesized and its incorporation and retention in SLNs determined. The incorporation of hydrophilic AZT in SLNs was minimal; however the incorporation of AZT-P increased with increasing phospholipid (PL) content and was independent of the amount of TL used. The incorporation of AZT-P was greater in negatively charged SLNs than in neutral SLNs. The in vitro release of AZT-P from different SLNs formulation was studied at 37°C using a bulk-equilibrium reverse dialysis sac technique. Increased drug release was observed in SLNs formulated with PLs having a transition temperature below 37°C. The results obtained indicate that the highly packed TL core of the SLN is not compatible with lipophilic molecules such as AZT-P. The incorporation and subsequent retention of AZT-P appears to be dependent on the PL coating on the SLNs surface and is independent of the TL solid core.

Evidence for Phospholipid Bilayer Formation in Solid Lipid Nanoparticles Formulated with Phospholipid and Triglyceride

AbstractPurpose. Solid lipid nanoparticles (SLN) are comprised of a high-melting point triglyceride (TG) core with a phospholipid (PL) coating. This study has investigated the possible formation of multiple PL bilayers on the TG core of SLNs as a function of increasing the PL:TG molar ratio.nMethods. Trilaurin (TL) was used as the SLN core. Dipalmitoylphosphatidylcholine (DPPC) or a mixture of DPPC and dimyristoylphosphatidylglycerol (DMPG) were used to produce neutral and negatively charged SLNs. The volume of aqueous phase associated with the PL was determined using calcein and 6-carboxyfluorescein (6-CF) as hydrophilic markers incorporated during the preparation of the SLNs.nResults. The diameter of the SLNs decreased as the molar ratio of PL to TL was increased, until a PL:TL ratio of 0.15 was reached. After this point the diameter was not affected by further increases in the molar ratio. The experimental amount of PL required to prepare SLNs was significantly higher than the theoretical amount required to form a single monolayer on the surface. The aqueous volume associated with the PL was increased with increasing PL:TL molar ratios.nConclusions. The results obtained suggest that the formation of multiple PL bilayers is probable in SLNs prepared with a high molar ratio of PL to TL. The volume of the aqueous phase between the PL-bilayers, estimated from the amount of the hydrosoluble markers trapped in this phase, provides an indication of the relative number of bilayers at different PL:TL ratios

Solid lipid nanoparticles as drug carriers: II. Plasma stability and biodistribution of solid lipid nanoparticles containing the lipophilic prodrug 3′-azido-3′-deoxythymidine palmitate in mice

Abstract Solid lipid nanoparticles (SLNs) were prepared using trilaurin as the SLNs solid core and a mixture of neutral and negatively charged phospholipid. To produce SLNs with a poly(ethylene glycol) (PEG) coating, PEG was incorporated in SLNs using dipalmitoylphosphatidylethanolamine- N -[poly(ethylene glycol) 2000 ] (PE-PEG). 3′-azido-3′-deoxythymydine palmitate (AZT-P) with [ 3 H]-AZT-P as tracer were synthesized and incorporated in SLNs. Their subsequent retention in SLNs with and without PEG was determined after incubation in 50% bovine plasma. Biodistribution studies were performed in mice using free AZT-P, AZT-P incorporated in SLNs or AZT-P incorporated in PE-PEG coated SLNs (SLN-PE-PEG). The presence of PE-PEG significantly reduced the SLN zeta potential from −22 to −5 mV. Although AZT-P was rapidly released from SLNs during incubation in bovine plasma, the release rate was significantly slower in SLN-PE-PEG. AZT-P was rapidly removed from blood following i.v. injection in mice. The decrease in AZT-P blood level was biphasic and rapid, and the major excretory route of AZT-P was the kidney. Higher levels were observed after i.v. injection of AZT-P incorporated in SLNs. This effect was further increased using SLN-PE-PEG. Both SLN and SLN-PE-PEG incorporation of AZT-P significantly decreased the urinary excretion of AZT-P and increased the localization of AZT-P in the liver. The results obtained in this study indicate that using SLNs as a drug carrier increases the bioavailibility of incorporated AZT-P, and that the pharmacokinetic behaviour of the incorporated drug can be modified by changing the surface characteristics of SLNs by using the amphiphilic solvation enhancer PE-PEG.

IMMUNOGENICITY OF IMMUNOLIPOSOMES : REACTIVITY AGAINST SPECIES-SPECIFIC IGG AND LIPOSOMAL PHOSPHOLIPIDS

Immunoliposomes with surface-linked avidin-biotinylated mouse IgG2a were prepared from dipalmitoylphosphatidylcholine (DPPC), dimyristoylphosphatidylglycerol (DMPG) and biotinylated dipalmitoylphosphatidylethanolamine (DPPE-biotin) in the molar ratio 10:1:0.1 with or without 5 mol% poly(ethylene glycol) dipalmitate (PEG-(C18)2). The ability of IgG2a immunoliposomes to elicit anti-IgG2a antibodies in mice was compared with alum and N-acetylmuramyl-L-alanyl-D-isoglutamine (MDP). IgG2a 5 microgram/mouse) did not elicit an IgG1 antibody response after 4 s.c. injections. Alum-adsorbed IgG2a elicited 2.1 +/- microgram IgG1 antibody/ml serum, whereas MDP elicited 24.3 +/- microgram/ml serum. IgG2a immunoliposomes elicited 12.4 +/- 3.7 microgram IgG1 antibody/ml serum, while immunoliposomes containing lipophilic PEG-(C18)2 elicited 21.4 +/- 5.1 microgram IgG1 antibody/ml serum. Elicited antibodies were specific for IgG2a, with no cross-reactivity with IgG2b. Anti-DPPC or anti-DMPG IgG antibody levels did not change during immunization. Anti-DPPE IgG antibody levels were slightly but significantly elevated during immunization, and there was a significant increase in the level of anti-DPPE-biotin antibodies. These results demonstrate that immunoliposomes prepared with species-specific antibody are immunogenic and induce significant levels of isotypespecific antibody upon repeated injection.

Influence of phospholipid composition on antibody responses to liposome encapsulated protein and peptide antigens

The effect of phospholipid composition on mouse IgG antibody responses to liposomal bovine serum albumin (BSA), murine monoclonal antibody GK1.5 (anti-CD4) or a 21 amino acid peptide from the second conserved domain of HIV gp120 after s.c. administration, and on the IgA, IgE, and IgG antibody response to liposomal Shistosoma mansoni glutathione-S-transferase (Sm28GST) after oral administration, was determined. Antibody responses were compared with alum-adsorbed and N-acetylmuramyl-L-alanyl-D-isoglutamine (MDP)-antigen mixtures. For the s.c. route, dipalmitoylphosphatidylcholine (DPPC)/dimyristoylphosphatidylglycerol (DMPG) liposomes induced 54-60% IgG1 and 35-44% IgG(2a+2b). DPPC/dipalmitoylphosphatidylethanolamine (DPPE) liposomes induced 73-78% IgG1 and 15-25% IgG(2a+2b). DPPC/ phosphatidylserine (PS) liposomes induced 86-89% IgG1 and 8-12% IgG(2a+2b). Alum and MDP induced 79-91% IgG1 and 4-17% IgG(2a+2b). The rank order of adjuvanticity for induction of IgG antibody was DPPC/DMPGDPPC/PE > > alum > > MDPDPPC/PS for all three antigens. DPPC/DMPG liposomes were the only effective adjuvant for the induction of secretory IgA and circulatory IgE and IgG antibodies against Sm28GST after oral administration. The failure of liposome-antigen mixtures to elicit an antibody response showed that liposomal incorporation of the antigens was obligatory for adjuvant activity. These results demonstrate that the correlation between phospholipid composition and adjuvanticity is independent of liposome charge, antigen, or route of administration.