Diplasmenylcholine-folate liposomes: An efficient vehicle for intracellular drug delivery

Article abstract of DOI:10.1021/ja9742949

Most pharmaceutical and gene therapy applications of targeted liposomes presently suffer from inefficient contents delivery to the cytoplasm of target cells. We report a plasma-stable liposome, composed of synthetic, naturally occurring diplasmenylcholine (1,2-di-O-(Z-1′-hexadecenyl)-sn-glycero-3-phosphocholine; DPPlsC), that rapidly and efficiently releases its contents at endosomal pHs. Acid-catalyzed hydrolysis of these liposomes produces glycerophosphocholine and fatty aldehydes, leading to greatly enhanced liposome permeability (t_{50% release} ? 1-4 h between pH 4.5-5.5) when > 20% of the vinyl ether lipid has been hydrolyzed. Plasma stability of nonhydrolyzed 9:1 DPPlsC/dihydrocholesterol liposomes exceeds 48 h at 37°C, pH 7.4 in 50% serum; pure DPPlsC liposomes remain stable in 10% serum under the same conditions. Fluorescence assays of KB cells treated with 99.5:0.5 DPPlsC/DSPE-PEG3350-folate liposomes containing encapsulated propidium iodide (PI) indicate that 83% of the PI escapes the endosomal compartment within 8 h to produce intensely stained nucleii. The IC₅₀ value of 1-β-arabinofuranosylcytosine (Ara-C) encapsulated in DPPlsC/DSPE-PEG3350-folate liposomes is 0.49 μM in KB cell cultures, a approx. 6000-fold enhancement in cytotoxicity compared with free drug (2.8 mM). Empty DPPlsC/DSPE-PEG3350-folate liposomes had no effect on DNA synthesis, indicating that DPPlsC and its degradation products are benign to cell function at these lipid concentrations. Our results suggest that concurrent application of selective targeting and membrane translocation mechanisms in drug carriers can significantly increase their efficacy. Most pharmaceutical and gene therapy applications of targeted liposomes presently suffer from inefficient contents delivery to the cytoplasm of target cells. We report a plasma-stable liposome, composed of synthetic, naturally occurring diplasmenylcholine (1,2-di-O-(Z-1'-hexadecenyl)-sn- glycero-3-phosphocholine; DPPIsC), that rapidly and efficiently releases its contents at endosomal pHs. Acid-catalyzed hydrolysis of these liposomes produces glycerophosphocholine and fatty aldehydes, leading to greatly enhanced liposome permeability (t(50% release) ? 1-4 h between pH 4.5-5.5) when >20% of the vinyl ether lipid has been hydrolyzed. Plasma stability of nonhydrolyzed 9:1 DPPlsC/dihydrocholesterol liposomes exceeds 48 h at 37°C, pH 7.4 in 50% serum; pure DPPlsC liposomes remain stable in 10% serum under the same conditions. Fluorescence assays of KB cells treated with 99.5:0.5 DPPlsC/DSPE-PEG3350-folate liposomes containing encapsulated propidium iodide (PI) indicate that 83% of the PI escapes the endosomal compartment within 8 h to produce intensely stained nucleii. The IC₅₀ value of 1-β- arabinofuranosylcytosine (Ara-C) encapsulated in DPPlsC/DSPE-PEG3350-folate liposomes is 0.49 μM in KB cell cultures, a ~6000-fold enhancement in cytotoxicity compared with free drug (2.8 mM). Empty DPPlsC/DSPE-PEG3350- folate liposomes had no effect on DNA synthesis, indicating that DPPlsC and its degradation products are benign to cell function at these lipid concentrations. Our results suggest that concurrent application of selective targeting and membrane translocation mechanisms in drug carriers can significantly increase their efficacy.

Full text of DOI:10.1021/ja9742949

VOLUME 120, NUMBER 44
NOVEMBER 11, 1998
© Copyright 1998 by the
American Chemical Society
Diplasmenylcholine-Folate Liposomes: An Efficient Vehicle for
Intracellular Drug Delivery†
Yuanjin Rui, Susan Wang, Philip S. Low, and David H. Thompson*
Contribution from the Department of Chemistry, Purdue UniVersity, West Lafayette, Indiana 47907-1393
ReceiVed December 17, 1997. ReVised Manuscript ReceiVed June 26, 1998
Abstract: Most pharmaceutical and gene therapy applications of targeted liposomes presently suffer from
inefficient contents delivery to the cytoplasm of target cells. We report a plasma-stable liposome, composed
of synthetic, naturally occurring diplasmenylcholine (1,2-di-O-(Z-1′-hexadecenyl)-sn-glycero-3-phosphocholine;
DPPlsC), that rapidly and efficiently releases its contents at endosomal pHs. Acid-catalyzed hydrolysis of
these liposomes produces glycerophosphocholine and fatty aldehydes, leading to greatly enhanced liposome
permeability (t50% release = 1-4 h between pH 4.5-5.5) when >20% of the vinyl ether lipid has been hydrolyzed.
Plasma stability of nonhydrolyzed 9:1 DPPlsC/dihydrocholesterol liposomes exceeds 48 h at 37 °C, pH 7.4 in
50% serum; pure DPPlsC liposomes remain stable in 10% serum under the same conditions. Fluorescence
assays of KB cells treated with 99.5:0.5 DPPlsC/DSPE-PEG3350-folate liposomes containing encapsulated
propidium iodide (PI) indicate that 83% of the PI escapes the endosomal compartment within 8 h to produce
intensely stained nucleii. The IC50 value of 1-â-arabinofuranosylcytosine (Ara-C) encapsulated in DPPlsC/
DSPE-PEG3350-folate liposomes is 0.49 µM in KB cell cultures, a ∼6000-fold enhancement in cytotoxicity
compared with free drug (2.8 mM). Empty DPPlsC/DSPE-PEG3350-folate liposomes had no effect on
DNA synthesis, indicating that DPPlsC and its degradation products are benign to cell function at these lipid
concentrations. Our results suggest that concurrent application of selective targeting and membrane translocation
mechanisms in drug carriers can significantly increase their efficacy.
Introduction
These applications utilize actively-3-11 and passively-12-17
targeted liposomes as the drug carrier because of their site-
Liposome-encapsulated drugs, presently used for the clinical
treatment of cancers and fungal infections, are undergoing
accelerated development for applications in gene therapy.1,2
(3) Hughes, B. J.; Kennel, S.; Lee, R.; Huang, L. Cancer Res. 1989, 49,
6214-6220.
(4) Loughrey, H. C.; Wong, K. F.; Choi, L. S.; Cullis, P. R.; Bally, M.
B. Biochim. Biophys. Acta 1990, 1028, 73-81.
(5) Morgan, J.; MacRobert, A. J.; Gray, A. G.; Huehns, E. R. Br. J.
Cancer 1992, 65, 58-64.
(6) Lee, K.-D.; Hong, K.; Papahadjopoulos, D. Biochim. Biophys. Acta
1992, 1103, 185-197.
(7) Leserman, L.; Suzuki, H.; Machy, P. Liposome Technology. 2nd ed.;
Gregoriadis, G., Ed.; CRC Press: Boca Raton, FL, 1993; Vol. III, p 139-
152.
(8) Mori, A.; Huang, L. Liposome Technology. 2nd ed.; Gregoriadis, G.,
Ed.; CRC Press: Boca Raton, FL, 1993; Vol. III, p 153-162.
(9) Lundberg, B.; Hong, K.; Papahadjopoulos, D. Biochim. Biophys. Acta
1993, 1149, 305-312.
(10) Ahmad, I.; Longenecker, M.; Samuel, J.; Allen, T. M. Cancer Res.
1993, 53, 1484-1488.
* Corresponding author: Tel: 765-494-0386. FAX: 765-496-2592.
E-mail: davethom@chem.purdue.edu.
† Abbreviations: AraC, 1-â-arabinofuranosylcytosine; DHC, 5a-choles-
tane-3b-ol (dihydrocholesterol); DPPlsC, 1,2-di-O-1′-Z-hexadecenyl-sn-
glycero-3-phosphocholine; EPC, egg phosphatidylcholine; FDMEM, folate-
deficient modified Eagle’s medium; HIFCS, heat-inactivated fetal calf
serum; IC50, 50% growth inhibition concentration; PBS, phosphate buffered
saline (150 mM NaCl, 2.0 mM Na2HPO4, pH 7.4); PI, propidium iodide.
(1) Lee, R. J.; Huang, L. In Artificial Self-Assembling Systems for Gene
DeliVery; ACS Conference Proceedings Series, American Chemical Soci-
ety: Washington, D.C., 1996; pp 169-176.
(2) Liposomes in Gene DeliVery; Lasic, D. D., Ed., Marcel Dekker: New
York, 1997.
10.1021/ja9742949 CCC: $15.00 © 1998 American Chemical Society
Published on Web 10/27/1998

11214 J. Am. Chem. Soc., Vol. 120, No. 44, 1998
Rui et al.
Figure 1. Acid-catalyzed hydrolysis of 1,2-di-O-(Z-1′-hexadecenyl)-sn-glycero-3-phosphocholine (DPPlsC) to give the degradation products
glycerophosphocholine, hexadecanal, and hexadecanoic acid.
selectivity, biocompatibility, high drug/lipid ratios, and blood
clearance characteristics. Unfortunately, existing methods for
transporting the liposomal contents to the target cell cytoplasm
are either inefficient or lack sufficient plasma stability for in
vivo applications. This obstacle is especially problematic for
the cytoplasmic delivery of peptides, antisense oligonucleotides,
and gene constructs since (i) their size and hydrophilicity
prevents efficient cytoplasmic delivery to the desired target
unless specific membrane fusion and/or cytoplasmic release
mechanisms18,19 have been incorporated, (ii) the liposomal cargo
not rapidly discharged into the cytoplasm is typically delivered
to lysosomes where it is degraded, and (iii) the cationic lipids
and polymers that are used for nucleic acid transfer are often
cytotoxic and/or have limited plasma stability. Notable excep-
tions are the first-pass use of adenovirus vectors20 or fusogenic
“virosomes” derived from influenza virus envelopes;21 however,
questions regarding the safety and processability of these carriers
remain unanswered.
We report a hybrid liposome system, incorporating both folate
receptor-mediated cell targeting (via a DSPE-PEG3350-folate
conjugate22,23) and a pH-sensitive trigger using the naturally
occurring vinyl ether-linked phospholipid diplasmenylcholine
(1,2-di-O-(1′-Z-hexadecenyl)-sn-glycero-3-phosphocholine; DP-
PlsC24,25), that was designed to obviate these problems and
efficiently deliver its contents to the cytoplasm of KB cells, a
human nasopharyngeal epidermal carcinoma cell line, upon
exposure to the acidic endosomal compartment. Acid-catalyzed
hydrolytic triggering in this system is analogous to the photo-
induced triggering previously reported by this laboratory for
plasmenylcholine18,26 and diplasmenylcholine27,28 liposomes;
both systems utilize chemical activation mechanisms to produce
lipid phase transitions by degrading lamellar phase-forming
phospholipids to a pair of single-chain surfactant products
(Figure 1).
Results and Discussion
Diplasmenylcholine Liposome Hydrolysis and Its Influ-
ence on Calcein Release. Acid-catalyzed release of calcein
from DPPlsC liposomes suspended in buffer was studied as a
model system for endosomal release. No detectable calcein
release occurs from DPPlsC liposomes maintained at pH 7.4,
37 °C for 48 h, in contrast to their leakage properties at pH 4.5
where the half-time for release (t50% release) is 76 min. Calcein
leakage rates increase with decreasing pH (Figure 2a and Table
1) and with the extent of DPPlsC hydrolysis at pH 4.5 (Figure
2b,c), however, they decrease with increasing mole fraction of
the saturated cholesterol derivative, 5a-cholestane-3b-ol (dihy-
drocholesterol, DHC29) (Figure 2d). Hydrolysis rates of DP-
PlsC, monitored by HPLC-ELS analysis, were independent of
DHC content, suggesting that a critical extent of diplasmenyl-
choline degradation is required before the onset of rapid calcein
leakage occurs (20 to ∼80% hydrolysis, depending on DHC
content; Figure 2c). DPPlsC hydrolysis kinetics at pH 4.5, a
pH regime that can occur within the endosomes of KB cells,30
are pseudo-first-order (kobs ) 6.3 × 10-5 s-1 at pH 4.5). Calcein
release rates, however, are nonlinear, with dramatic increases
in leakage rate occurring between 15 and 30% of lipid hydrolysis
(Figure 2b,c). These results suggest that membrane destabiliza-
tion occurs only after a critical concentration of diplasmenyl-
choline degradation products have accumulated within the
bilayer. The critical behavior of this coupled hydrolysis-leakage
process and the independence of the DPPlsC hydrolysis rate
on the mol % DHC in the bilayer are features shared by
semisynthetic plasmenylcholine liposomes undergoing acid-
catalyzed hydrolysis.31 1H NMR and mass spectra of the
DPPlsC hydrolysis products, isolated by Bligh-Dyer extraction
of the acidic liposome solutions, indicate that the predominant
lipophilic degradation products are hexadecanal and hexade-
canoic acid. These results indicate that palmitic acid and
glycerophosphocholine are the terminal acid-catalyzed hydroly-
sis products of DPPlsC as illustrated in Figure 1. Partially
hydrolyzed lysolipid intermediates (e.g., 1-hydroxy-2-(1′-Z-
hexadecenyl)-sn-glycero-3-phosphocholine, 1-(1′-Z-hexadece-
nyl)-2-hydroxy-sn-glycero-3-phosphocholine, or the cyclized
isomer, 1,2-hexadecylacetal-sn-glycero-3-phosphocholine), in
principle, may also participate in bilayer destabilization during
the course of the hydrolysis reaction; however, since these
(11) Wang, S.; Lee, R. J.; Cauchon, G.; Gorenstein, D. G.; Low, P. S.
Proc. Natl. Acad. Sci. U.S.A. 1995, 92, 3318-3322.
(12) Stealth Liposomes; Lasic, D. D., Martin, F. J., Eds.; CRC Press:
Boca Raton, FL, 1995.
(13) Mayer, L. D.; Tai, L. C. L.; Ko, D. S. C.; Masin, D.; Ginsberg, R.
S.; Cullis, P. R.; Bally, M. B. Cancer Res. 1989, 49, 5922-5930.
(14) Mayhew, E. G.; Lasic, D.; Babbar, S.; Martin, F. J. Int. J. Cancer
1992, 51, 302-309.
(15) Vaage, J.; Mayhew, E.; Lasic, D.; Martin, F. Int. J. Cancer 1992,
51, 942-948.
(16) Huang, S. K.; Lee, K.-D.; Hong, K.; Friend, D. S.; Papahadjopoulos,
D. Cancer Res. 1992, 52, 5135-5143.
(17) Yuan, F.; Leunig, M.; Huang, S. K.; Berk, D. A.; Papahadjopoulos,
D.; Jain, R. K. Cancer Res. 1994, 54, 3352-3356.
(18) Gerasimov, O. V.; Rui, Y.; Thompson, D. H. Vesicles; Rosoff, M.,
Ed.; Marcel Dekker: New York, 1996; pp 679-746.
(19) Gerasimov, O. V.; Boomer, J. A.; Qualls, M. M.; Thompson, D. H.
AdV. Drug DeliVery ReV. 1998, in press.
(20) Kovesdi, I.; Brough, D. E.; Bruder, J. T.; Wickham, T. J. Curr.
Opin. Biotechnol. 1997, 8, 583-589.
(21) Dijkstra, J.; Bron, R.; Wilschut, J.; deHaan, A.; Ryan, J. L. J.
Immunology 1996, 157, 1028-1036.
(22) Lee, R. J.; Low, P. S. J. Biol. Chem. 1994, 269, 3198-3204.
(23) Smith, M. D., Ph.D. Thesis, Purdue University, 1997.
(24) Rui, Y.; Thompson, D. H. Chem. Eur. J. 1996, 2, 1505-1508.
(25) Rui, Y.; Thompson, D. H. J. Org. Chem. 1994, 59, 5758-5762.
(26) Thompson, D. H.; Gerasimov, O. V.; Wheeler, J. J.; Rui, Y.;
Anderson, V. C. Biochim. Biophys. Acta 1996, 1279, 25-34.
(27) Gerasimov, O. V.; Wymer, N.; Miller, D.; Rui, Y.; Thompson, D.
H. ACS Symp. Ser., in press.
(28) Wymer, N.; Gerasimov, O. V.; Thompson, D. H. Bioconjugate
Chem. 1998, 9, 305-308.
(29) Similar release kinetics were observed when cholesterol was used
instead of DHC.
(30) Lee, R. J., Ph.D. Thesis, Purdue University, 1994.
(31) Gerasimov, O. V.; Schwan, A.; Thompson, D. H. Biochim. Biophys.
Acta 1997, 1324, 200-214.

Diplasmenylcholine-Folate Liposomes
J. Am. Chem. Soc., Vol. 120, No. 44, 1998 11215
Figure 2. (a) Calcein release rates at 37 °C from DPPlsC liposomes in different buffers.26,31 b, pH 2.3; 9, pH 3.2; 2, pH 4.5; [, pH 5.3; O, pH
6.3. (b) Calcein release rate (b) and DPPlsC hydrolysis31 rate (9) from DPPlsC liposomes at pH 4.5, 37 °C. (c) Correlation between % calcein
released and the extent of DPPlsC hydrolysis in liposomal dispersions at pH 4.5, 37 °C. b, pure DPPlsC; 9, 8:2 DPPlsC/DHC; 2, 7:3 DPPlsC/
DHC; [, 6:4 DPPlsC/DHC. Note that as the DHC content increases, more extensive DPPlsC hydrolysis is required to effect 50% calcein release.
(d) Calcein release rates at 37 °C from DPPlsC/DHC liposomes in citrate buffer (pH 4.54) containing different molar ratios of DPPlsC and DHC.
b, pure DPPlsC; 9, 9:1 DPPlsC/DHC; 2, 8:2 DPPlsC/DHC; [, 7:3 DPPlsC/DHC; O, 6:4 DPPlsC/DHC.
Table 1. pH Dependence of 50% Release Time
Table 2. Liposome Stability at pH 7.4, 37 °C
pH
t50% release (min)
50% HIFCS
10% HIFCS
48 h
liposome type
24 h
48 h
2.3
3.2
4.5
5.3
6.3
1.5
3.6
76
230
1740
DPPlsC
27%
33%
0
0
0
0
0
9:1 DPPlsC/DHC
8:2 DPPlsC/DHC
7:3 DPPlsC/DHC
6:4 DPPlsC/DHC
0
0
0
0
0
0
0
0
species were not detected by HPLC analysis,32 we believe that
they are less important in promoting membrane permeability
than the terminal hydrolytic product, hexadecanoic acid.
Plasma Stability of Diplasmenylcholine Liposomes. Unlike
many pH-sensitive liposome formulations,18,19 the plasma stabil-
ity of DPPlsC liposomes at 37 °C is quite good. Pure DPPlsC
liposomes do not leak calcein upon exposure to 10% heat-
inactivated fetal calf serum (HIFCS) for up to 48 h. The
addition of g10% DHC to the DPPlsC membrane further
stabilizes these liposomes in 50% HIFCS over the same time
period (Table 2). Incorporation of 0.5 mol % DSPE-PEG3350-
folate within the DPPlsC membrane also did not affect the
plasma stability of the liposomes. These results suggest that
a % release values are (5%.
DPPlsC liposomes are sufficiently plasma-stable for in vitro drug
delivery applications.
Cytoplasmic Delivery of Propidium Iodide. The ability
of folate-targeted DPPlsC/DHC liposomes to promote endoso-
mal delivery of hydrophilic compounds to the cytoplasm of KB
cells was evaluated by fluorometric assay using propidium
iodide (PI) as a fluorescent probe.30,33 PI fluorescence (λex
)
540 nm, λem ) 615 nm) increases approximately 50-fold upon
binding to RNA or DNA. This property makes it useful for
endosomal release assays, since cellular fluorescence at 615 nm
from internalized PI effectively arises only after it has escaped
the endosome and binds to nucleic acids. PI release kinetics
revealed that 83% of the encapsulated PI escaped both the
liposomal and endosomal compartments within 8 h when e10
(32) Lysolipids are readily detected by our HPLC-ELS methodology.31
Since they were not observed by HPLC-ELS during the course of the
1
hydrolytic reaction and were also not detected by H NMR in the isolated
organic soluble fraction from Bligh-Dyer extraction, these partially
hydrolyzed species must be cleaved more rapidly than their diplasmenyl-
choline precursors.
(33) Vogel, K.; Wang, S.; Lee, R. J.; Chmielewski, J.; Low, P. S. J.
Am. Chem. Soc. 1996, 118, 1581-1586.

11216 J. Am. Chem. Soc., Vol. 120, No. 44, 1998
Rui et al.
synthesis was not observed in KB cells treated with empty
(control) DPPlsC/DSPE-PEG3350-folate liposomes, indicating
that neither the lipid nor its degradation products have a
significant effect on this cellular function at the lipid concentra-
tions used.
Conclusion
Allen and co-workers have demonstrated the efficacy of
passively targeted, Ara-C-containing sterically stabilized lipo-
somes in a murine model and attributed their increased
therapeutic effect, relative to free Ara-C, to a sustained
extracellular release mechanism from these formulations.39 Our
results demonstrate that endosomal triggering of DPPlsC
liposomes provides a fast, efficient, and practical method for
intracellular delivery of Ara-C and other water-soluble materi-
als,40 which significantly improves their biological activity.
Synergistic action of receptor-mediated cell targeting and acid-
catalyzed endosomal escape may offer greatly enhanced drug
efficacy for many hydrophilic, nonmembrane permeable com-
pounds. Experiments designed to both elucidate the endosomal
escape mechanism and explore the applications of spontaneous
liposomes41 composed of DPPlsC:PEG formulations for drug
delivery and gene transfer are in progress.
Figure 3. Propidium iodide release kinetics in KB cells using
liposomes targeted with 0.5% DSPE-PEG3350-folate lipid.22,23 [,
egg phosphatidylcholine (EPC) liposomes (pH-insensitive control); 9,
pure DPPlsC liposomes; 2, 9:1 DPPlsC/DHC liposomes; b, 8:2
DPPlsC/DHC liposomes; O, 9:1 DPPlsC/DHC liposomes in the
presence of 25 µM monensin.
mol % DHC was present in the DPPlsC membrane; 36% release
occurred within 8 h (53% after 24 h) when the DHC content
was increased to 20 mol % (Figure 3). Endosomal unloading
of PI was also confirmed by fluorescence microscopy (Figure
4), where intense staining of nucleii and nucleoli indicates that
the released PI has also penetrated into the nucleus. Both the
extent and rate of PI release were greater for DPPlsC liposomes
than for folate-targeted EPC vesicles containing the pH-sensitive
peptide EALA either covalently attached (9% release in 8 h;
20% in 24 h) or added to the external medium (4% release in
8 h; 13% in 24 h).33 Control experiments, using KB cells treated
with DPPlsC/DSPE-PEG3350-folate/PIin liposomes in the
presence of the endosomal acidification inhibitors monensin
(25mM) and chloroquine (50mM), indicated that <5% PI
escaped into the cytoplasm, when monitored for up to 24 h after
liposomal treatment (Figure 3). These results strongly suggest
that an acidic endosomal compartment is necessary to trigger
cytoplasmic contents delivery from DPPlsC liposomes, presum-
ably via the involvement of fatty aldehydes and acids derived
from diplasmenylcholine hydrolysis.
Experimental Methods
Liposome Preparation. Calcein-containing Liposomes without
DHC. Pure DPPlsC liposomes were prepared by hydrating 3.4 mg of
DPPlsC powder in 1 mL of calcein solution (prepared by dissolving
93 mg of calcein in 3.0 mL of 0.3 M NaOH to give a final concentration
of 50 mM, pH 12.5) using five LN2 freeze-thaw-vortex cycles. The
lipid suspension was then extruded at 50 °C through two 100 nm track-
etch polycarbonate membrane filters.42 Extraliposomal calcein was
removed by using a single pass through a 40-cm Sephadex G-50 gel
column equilibrated with 150 mM NaCl. The fraction eluting at the
void volume was collected, stored at 8 °C, and used within 24 h. This
general procedure was used for all liposome preparations, except as
noted below.
Calcein-containing DPPlsC/DHC Liposomes. DPPlsC and DHC,
dissolved in chloroform that had been pre-filtered through a 1’’ plug
of anhydrous sodium carbonate to remove traces of acid and water,
were evaporated with a stream of dry N2 gas, followed by evacuation
at <200 µm for at least 4 h to give thin lipid films. Liposomes were
prepared as described above by hydrating these films in the presence
of a 50 mM calcein solution, followed by extrusion and removal of
the extravesicular calcein by gel filtration.
Cytoplasmic Delivery of 1-â-Arabinofuranosylcytosine
(Ara-C). The ability of folate-targeted 9:1 DPPlsC/DHC
liposomes to enhance cytoplasmic delivery of the cytotoxic
antimetabolite drug, Ara-C,34 was monitored by [3H]thymidine
incorporation assay (Figure 5). These liposomes exhibit a
∼6000-fold enhancement of DNA synthesis inhibition relative
to free Ara-C in KB cell cultures (IC50 ) 490 nM and 2.8 mM
for DPPlsC/folate liposomal and free Ara-C, respectively).
DPPlsC/DHC/DSPE-PEG3350-folate (9:1:0.05) liposomes
containing Ara-C represent an improvement over transferrin-
conjugated, Ara-C containing pH-sensitive phosphatidyletha-
nolamine liposomes35 by a factor of 70, pH-sensitive immuno-
liposomes36,37 by a factor exceeding 1000, and liposomal Ara-C
prodrugs by a factor of approximately 400.38 Inhibition of DNA
DPPlsC Liposomes Containing PI or Ara-C. DPPlsC (0.020
mmol), DSPE-PEG3350-folate (0.10 mmol), and DHC (0, 0.002, and
0.004 mmol) in prefiltered CHCl3 (see above) were mixed and
evaporated to a thin film. The dry lipid film was hydrated in the
presence of 10 mg/mL PI or Ara-C in phosphate-buffered saline (150
mM NaCl, 2.0 mM Na2HPO4, pH 7.4, PBS) and extruded as described
above. Unencapsulated solutes were removed using a single pass
through a 40-cm Sephadex G-50 column equilibrated and eluted with
PBS.
Calcein Release Assay. Calcein-containing liposomes (1.0 mL)
were mixed at 37 °C with 1.0 mL of 20 mM phosphate (pH 6.3 and
5.3), citrate (pH 4.5), or oxalate (pH 3.2) buffer in 150 mM NaCl;
HCl in 150 mM NaCl was used for pH 2.3 hydrolyses. Calcein
(37) Ho, R. J. Y.; Rouse, B. T.; Huang, L. J. Biol. Chem. 1987, 262,
13973-13978.
(34) The efficacy of Ara-C, the most commonly used drug for the
treatment of human acute myeloid leukemia, is limited by low lipophilicity
and a host of other biologic factors. Several prodrug approaches have been
investigated to enhance its activity: Wipf, P.; Li, W. Drugs Future 1994,
19, 49-54.
(35) Brown, P. M.; Silvius, J. R. Biochim. Biophys. Acta 1990, 1023,
341-351.
(36) Conner, J.; Huang, L. Cancer Res. 1986, 46, 3431-3435.
(38) Schwendener, R. A.; Schott, H. Int. J. Cancer 1992, 51, 466-469.
(39) Allen, T. M.; Mehra, T.; Hansen, C.; Chin, Y. C. Cancer Res. 1992,
52, 2431-2439.
(40) Qualls, M. M.; Thompson, D. H., unpublished results.
(41) Szleifer, I.; Gerasimov, O. V.; Thompson, D. H. Proc. Natl. Acad.
Sci. U.S.A. 1998, 95, 1032-1037.
(42) Mayer, L. D.; Hope, M. J.; Cullis, P. R. Biochim. Biophys. Acta
1986, 858, 161-168.

Diplasmenylcholine-Folate Liposomes
J. Am. Chem. Soc., Vol. 120, No. 44, 1998 11217
Figure 4. Micrographs of KB cells treated for 2 h with DPPlsC/folate/PI liposomes as in Figure 3, washed with phosphate-buffered saline, pH 7.4
(PBS) to remove folate-targeted liposomes, and visualized at various times. Phase contrast micrographs taken 4 h (A), 8 h (C), and 24 h (E) after
PBS wash. Fluorescence micrographs taken at 4 h (B), 8 h (D), and 24 h (F) after addition of DPPlsC/DSPE-PEG3350-folate/PIin liposomes.
was then compared with the 100% release value (determined by adding
Triton X-100 to the diluted liposome sample) using the ratio method.26
Analysis of DPPlsC Liposome Hydrolysates. At various time
points, an 80-µL aliquot of the liposome hydrolysis solution was
withdrawn and added to a mixture of 200 µL of methanol and 100 µL
of chloroform. After the solution was mixed thoroughly by vortexing,
an additional 80 µL of water and 100 µL of chloroform were added.
The sample was again vortexed and then centrifuged for 3 min at 10 000
rpm. A 25-µL sample of the chloroform phase was analyzed with a
Waters model 510 HPLC equipped with a Sedex 55 evaporative light
scattering detector. DPPlsC hydrolysis rates were determined using a
normal phase column (3.9 × 300 mm µ-Porasil silica column) and
4:5:1 (v/v) hexanes/methanol/water as eluent (2.2 mL/min flow rate)31
by monitoring the disappearance of DPPlsC as a function of hydrolysis
time. No peaks due to the accumulation of lysoplasmenylcholine
intermediates (i.e., either the sn-1 or sn-2 monohydrolysis products)
or the 2-pentadecyl-1,3-dioxolane sideproduct (i.e., the isomeric,
internally trapped cyclic acetal monohydrolysis product) were observed
Figure 5. Cytotoxicity of Ara-C in KB cell cultures. Free Ara-C ([);
Ara-C encapsulated in EPC/DSPE-PEG3350-folate liposomes (9);
9:1 DPPlsC/DHC/DSPE-PEG3350-folate liposomes (2).
in any of the hydrolysis samples analyzed by this technique.
Fully hydrolyzed DPPlsC liposome samples (as determined by the
disappearance of the DPPlsC peak in HPLC) were extracted with
chloroform (3×), evaporated, and redissolved in sodium carbonate-
1
filtered CDCl3 for H NMR analysis. NMR analysis of the liposome
samples immediately at the end of the hydrolysis reaction revealed that
the CHCl3/MeOH extracted material contained no phosphocholine
resonances and two new resonances at 2.3 ppm (2H) and 8.4 ppm (1H).
If the samples were lyophilized >10 h after completion of the hydrolysis
fluorescence dequenching was monitored as a function of time by
diluting 50 µL aliquots of the hydrolysis mixture into 2 mL of 150
mM NaCl/20 mM HEPES, pH 7.4 prior to measurement of the calcein
fluorescence intensity (λex ) 490 nm; λem ) 518 nm) using a Perkin-
Elmer MPF-66 spectrofluorimeter. The extent of liposomal leakage

11218 J. Am. Chem. Soc., Vol. 120, No. 44, 1998
Rui et al.
reaction, the 8.4 ppm resonance was replaced by a broad signal centered
at 10.5 ppm (1H). These products were identified by FAB-MS (DTT
matrix) as hexadecanal (m/z 241, M + 1) and hexadecanoic acid (m/z
257, M + 1), respectively. HPLC analysis of authentic hexadecanoic
acid samples revealed no detectable peaks under the conditions
described above, presumably due to elution at the column void volume.
Evaluation of DPPlsC Liposome Plasma Stability. Calcein-
containing DPPlsC liposomes dispersed in 20 mM HEPES/150 mM
NaCl, pH 7.4 were mixed with HIFCS in volume/volume ratios of 9:1
(10% serum) and 1:1 (50% serum) at 37 °C. Aliquots (0.1 mL) were
withdrawn as a function of time, diluted in 2.0 mL of HEPES buffer,
and the extent of calcein leakage was measured as described above.
Propidium Iodide Delivery to KB Cells. KB cells were cultured
for at least 5 weeks in folate-deficient modified Eagle’s medium
(FDMEM) to establish folate-deficiency before beginning the experi-
ments. These cells were then incubated (33 mm culture dishes, ∼2.5
× 106 cells/dish, ∼85% confluence) in 1.0 mL of FDMEM with 0.20
mL of the DPPlsC/DSPE-PEG3350-folate/5 mM PIin liposomes for
2 h at 37 °C. After the cells were washed thoroughly with PBS to
remove any unbound liposomes, 1.0 mL of fresh FDMEM was added,
and the cells were re-incubated for up to 24 h. The cells were washed
with PBS 0, 2, 4, 8, 12, and 24 h after treatment with the PI-containing
liposomes, suspended in nonenzymatic cell dissociation solution, and
the PI fluorescence measured (λex ) 540 nm, λem ) 615 nm). PI escape
from the endosomes was quantitated using the expression, % release
) [(Ft - F0)/(Fmax - F0)] × 100, where Ft ) PI fluorescence emission
intensity at time t, F0 ) background fluorescence emission intensity at
PI fluorescence as the method described above.) Phase contrast and
fluorescence micrographs were taken at 4, 8, and 24 h to visualize the
extent of PI release from the endosomes. Endosomal acidification
inhibition control experiments were performed in the same manner,
except that final medium concentrations of 25 µM monensin (obtained
by adding 20 µL of 1.25 mM monensin stock in PBS) or 50 µM
chloroquine (obtained by adding 20 µL of 2.5 mM chloroquine stock
in PBS) were maintained during the 2 h liposome-containing and 24 h
liposome-free incubations.
Ara-C Cytotoxicity in KB Cell Cultures. KB cells were plated
to 50% confluence in 24-well culture plates before treatment with free
Ara-C, Ara-C encapsulated in 6:4 EPC/cholesterol/DSPE-PEG3350-
folate liposomes, or 9:1 DPPlsC/DHC/DSPE-PEG3350-folate lipo-
somes. Liposomes were prepared as described above, except that the
lipids were hydrated in an Ara-C solution (PBS, pH 7.4); drug
concentration after gel filtration ) 500 µM, yielding a [drug]/[lipid]
ratio of 1:65. After 4 h exposure to the Ara-C suspensions at various
concentrations, the cells were washed extensively with PBS and
incubated in fresh FDMEM in the presence of [3H]thymidine (2 µCi/
well). After 24 h, the DNA was precipitated with trichloroacetic acid
and dissolved in 2 M NaOH, and [3H]thymidine incorporation was
measured by scintillation counting.
Acknowledgment. This paper is dedicated to the memories
of Robert R. Squires (deceased, 9/30/98) and Demetrios
Papahadjopoulos (deceased, 9/21/98). The authors would like
to thank Dr. Oleg Gerasimov for his assistance with the HPLC
assays. This work was supported by grants from the Indiana
Elks, Purdue Research, and Whitaker Foundations to D.H.T.
and from the Purdue Research Foundation to P.S.L.
time ) 0 h of cells not treated with PI-containing liposomes, Fmax
)
PI fluorescence emission intensity observed after sonicating the cells
for 15 min, and F′0 ) background fluorescence of untreated cells
disrupted with a 15-min sonication. (Control experiments showed that
longer sonication periods, freeze-thawing + 15 min sonication, or 0.1
M Triton ×100 treatment gave essentially the same extent of cellular
JA9742949