Departments of Otolaryngology (P.J.F., C.C.), Pharmacology and
Toxicology (P.J.F.), and Oncology (P.J.F., M.D.V), University of
Western Ontario and London Regional Cancer Centre, London, Canada
Platinum drugs comprise one of the main classes of chemotherapy
drugs that can induce remissions in various solid tumors. Although
tumors often regress on treatment with
cis-diamminedichloroplatinum II (cisplatin) or
cis-diammine-1,1-cyclobutane dicarboxylate platinum II
(carboplatin), they usually relapse as a drug-resistant tumor. Most
mechanisms of platinum resistance could be overcome by increasing the
amount of drug that is accumulated by tumor cells. Amphotericin B (Amph
B) is efficient at increasing platinum drug uptake, but because of
nephrotoxicity associated with extended usage, and the potential for
synergistic nephrotoxicity when used with platinum drugs, Amph B has
not been used clinically for this purpose. A liposomal preparation of
Amph B (LipoAmph B), which is substantially less nephrotoxic, was
studied for its ability to enhance platinum-drug toxicity to a human
oral squamous cell carcinoma line, HN-5a, and its carboplatin-resistant
variant, 5a/carbo-15a, in which cisplatin accumulation was reduced by
approximately 40%. Amph B at 10 µg/ml enhanced cisplatin
accumulation by approximately 100% in both cell lines, enhancing
cytotoxicity of the drugs by 35 to 60%, and completely reversed
resistance to both cisplatin and carboplatin. LipoAmph B in the
presence of phospholipase A2-II (PLA2-II) was able
to enhance cisplatin and carboplatin cytotoxicity as effectively as
free Amph B in both cell lines. At optimal concentrations, LipoAmph B
plus PLA2-II enhanced drug uptake sufficiently to abolish resistance in
the platinum-resistant line. Because PLA2-II is elevated in some tumor
microenvironments and in plasma of ill patients, LipoAmph B has
potential clinical usefulness as a modulator of platinum-drug efficacy.
 |
Introduction |
Since the discovery of the anticancer activity
of cis-diamminedichloroplatinum II (cisplatin;
Rosenberg et al.,
1969
)1, this drug
has become the mainstay of cytotoxic treatment of various neoplasms.
Particularly, when used in combination protocols, cisplatin has greatly
improved the treatment of several solid tumors (Durant, 1980).
cis-Diammine-1,1-cyclobutane dicarboxylate platinum II
(carboplatin) is effective against a similar spectrum of tumors, but
causes less toxicity (Wagstaff et al., 1989). In spite of a good
response to cisplatin and carboplatin, the majority of
tumors recur, refractory to additional platinum treatment, resulting in
the death of 90% of patients within 2 years (Wagstaff et al., 1989).
To improve the success of this treatment, methods must be developed to
enhance the antitumor effectiveness of this important class of drugs.
Because clinical chemotherapy is administered at the highest tolerable
doses, a tumor need only acquire a 2- to 5-fold level of resistance to
escape the maximum effect of a drug (Jekunen et al., 1992). A host of
phenotypes has been described in association with platinum drug
resistance, the four most common being: 1) reduced intracellular
drug accumulation; 2) increased removal of platinum-DNA adducts; 3)
increased expression of glutathione and/or glutathione-metabolizing
enzymes; and 4) increased expression of metallothionein (Andrews and
Howell, 1990; Eastman, 1991; Scanlon et al., 1991; Gately and Howell,
1993; Ferguson, 1995). Because these mechanisms of resistance are time
and/or concentration dependent, each could be overcome, at least in
part, by increasing the cellular accumulation of drug. Of a variety of
agents that have been shown to reverse resistance, amphotericin B (Amph
B) is the most universal with respect to the spectrum of cell lines,
and most efficient in terms of degree of enhancement, by increasing the
cellular accumulation of platinum drugs (Morikage et al., 1991a, 1993; Kikkawa et al., 1993; Sharp et al., 1994; Beketic-Oreskovic and Osmak,
1995). However, the danger of cumulative nephrotoxicity limits the
clinically achievable plasma concentrations of Amph B to 7 to 20% of
the concentration required to effectively enhance platinum-drug
cytotoxicity to tumor cells (Sculier et al., 1988; Barriere, 1990;
Morikage et al., 1991a). Therefore, the present study examined a
liposomal preparation of Amph B, AmBisome, which may be used more
safely in the clinic because it largely avoids the kidney toxicity
caused by free Amph B (Meunier et al., 1991). It is demonstrated herein
that the liposome component must be destabilized by phospholipase to
release sufficient Amph B to enhance platinum-drug uptake into cells.
This may be accomplished by phospholipase A2-II
(PLA2-II) found in the microenvironment of various tumors (Abe et al.,
1997).
It is important that studies on cellular resistance to anticancer drugs
be conducted in a model system that reflects the clinical situation as
closely as possible. To this end, a 4-fold carboplatin-resistant variant was selected from a unique head and neck squamous cell carcinoma (HNSCC) cell line in two increments of carboplatin
concentration, to reflect the resistance levels that are most likely to
be observed clinically. This is the first HNSCC cell line selected for
resistance to carboplatin.
| |
Materials and Methods |
Drugs.
Carboplatin (CBDCA; JM-8; NSC 241240; Paraplatin) and cisplatin
(BMY-25936; Platinol-AQ) were a generous donation from Bristol-Myers Squibb (Saint-Laurent, Quebec). Amph B was purchased as Fungizone (Life
Technologies, Inc., Burlington, Ontario). Liposomal Amph B (LipoAmph B;
as AmBisome) and empty liposomes were generously provided by NeXstar,
Inc. (Boulder, CO). Melphalan (phenylalanine mustard) was a gift from
Burroughs Wellcome (Research Triangle Park, NC). Ouabain was kindly
provided by Dr. M. Karmazyn (Department of Pharmacology and Toxicology,
University of Western Ontario).
Other Supplies.
Cell culture plasticware was obtained from Life Technologies and Fisher
Scientific (Unionville, Ontario). Cell culture medium (Dulbecco's
modified Eagle's medium) and fetal bovine serum were purchased from
Life Technologies. PLA2-II (bee venom, specific activity approx. 2400 U/mg) was purchased from Boehringer Mannheim Canada (Laval, Quebec)
(one unit of PLA2-II activity is equivalent to 1 µmol of fatty acid
released per minute). All other reagents were obtained through
commercial sources.
Cell Culture.
HNSCC cell line HN-5a was established at this institution from the
gingival tumor of a patient not previously treated with chemotherapy or
radiation (Lapointe et al., 1992). HN-5a was cloned from the primary
HNSCC cell line by limiting dilution. Cells were maintained in
Dulbecco's modified Eagle's medium plus 10% fetal bovine serum and
penicillin (50 U/ml)/streptomycin (50 µg/ml) (growth medium).
Cultures were incubated in a humidified atmosphere of 5%
CO2 at 37°C. The carboplatin-resistant cell
lines were selected by culturing HN-5a cells in the presence of
carboplatin without any exposure to mutagens. Colonies, which had
propagated from single cells, were selected and expanded into separate
variant lines. In the first step of selection, cells were exposed
continuously for 3 weeks to 6 µM carboplatin
(104 cells/60-mm Petri dish) or 10 to 20 µM
carboplatin (105 cells/dish), with a weekly
change of drug-containing medium. At concentrations less than 6 µM,
cells grew to confluence. Distinct colonies were obtained from cells in
6 and 15 to 20 µM carboplatin, but those in 15 to 20 µM were unable
to sustain proliferation in this concentration of drug. The cells in 10 µM carboplatin, not having formed distinct colonies at 2 weeks, were
harvested collectively by trypsinization, and recultured at
105/dish for an additional 3 weeks in 10 µM
carboplatin. This reduced the viable cell number to approximately
10/dish, allowing propagation of colonies from single cells. Cells
expanded from one of these colonies, 5a/carbo-10e, then were reselected
in higher concentrations of carboplatin. Colonies were obtained after 3 weeks from 15 and 20 µM carboplatin, including the 5a/carbo-15a line
used in these studies. Colonies were not obtained in higher
concentrations of carboplatin. The carboplatin-resistant lines expanded
from individual colonies were exposed to drug at every alternate
passage, and were maintained drug-free for at least 3 days before
initiation of an experiment.
For the purpose of cytotoxicity assays, cultured cells were exposed to
drug as reported previously (Ferguson and Cheng, 1987). Briefly,
replicate flasks of rapidly proliferating cells were exposed to a range
of drug concentrations, including modifiers where indicated. Drug
exposure was initiated by addition of 0.2 volume of an appropriate
concentration of the agent of interest, in growth medium. At initiation
of drug exposure and after 4 days, cell numbers were determined using
an electronic particle counter (Coulter Electronics, Hialeah, FL). The
drug effect was measured by the inhibition of proliferation of the
cultured cells as a percentage of the respective controls (absence of
platinum drug). Controls for Fungizone included desoxycholate, the
solubilizing agent for Amph B. IC50 and
IC90 values were obtained by interpolation of
plotted data. For experiments involving 4-h exposures to drugs, pretreatments were added to cultures 24 h after establishment of
cultures, and after an additional 24-h incubation, cisplatin in growth
medium was added directly (without changing medium). After 4 h,
drug-containing medium was aspirated and substituted with growth
medium, and cells were incubated for another 4 days.
Determination of Cisplatin Uptake.
After 3 days of proliferation to approximately 50% confluence,
replicate 25-cm2 flasks of cells were exposed to
cisplatin by addition of drug as described above. Cells were exposed to
50 µM cisplatin for 4 h at 37°C, using as a background control
a 0°C, 1-s exposure (nonspecific, external binding). After the 4-h
exposure, samples were prepared according to Mann et al. (1990). Each
flask was washed 3× with ice-cold phosphate-buffered saline (0.15 M
NaCl + 0.67 mM KH2PO4, pH
7.4), and then removed by scraping with a plastic scraper in 0.5 ml of
Triton-HCl [0.1% (v/v) Triton X-100 in 0.35 N HCl (modified)]. Cells
were disrupted by a cycle of freezing and thawing followed by
sonication. The protein concentration was determined using Coomassie
staining (BioRad reagent) (Bradford, 1976) for normalization of
results. Duplicate samples of triplicate flasks were analyzed by
flameless (graphite furnace) atomic absorption spectroscopy (FAAS)
using a Varian Automated Graphite Furnace Atomizer attached to a Varian
Spectra 30 Atomic Absorption Spectrometer (Varian Canada, Georgetown,
Ontario). Platinum concentrations were determined by the computer from
a standard curve generated before each run.
Statistical Analysis.
Statistical significance was determined using a 2-tailed Student's
t test, except for drug accumulation studies, which were analyzed by a paired t test.
| |
Results |
Establishment of a Carboplatin-Resistant Human Tumor Line that Is
Cross-Resistant to Cisplatin.
The HN-5a/carbo cell lines were selected under minimal stringency to
establish resistant lines that would reflect phenotypes displayed by
refractory patient tumors. The two-step carboplatin selection of the
resistant variant cell line 5a/carbo-15a is detailed in Materials
and Methods. The carboplatin-selected cell line had the same
protein content as the parent line (Table
1). The proliferation rate of
5a/carbo-15a did not differ from the parent in the absence of drug, but
was significantly slower in 20 µM carboplatin (a slightly higher
concentration of carboplatin than the selection concentration).
5a/carbo-15a cells displayed a similar level of resistance to both
carboplatin and cisplatin. This line was more sensitive than the parent
HN-5a line to ouabain. 5a/carbo-15a cells were cross-resistant to
melphalan, but sensitivity to cadmium was unchanged.
View this table:
[in this window]
[in a new window]
|
TABLE 1
Characteristics of parent HN-5a cell line and carboplatin-resistant
variant 5a/carbo-15a
Numbers in parentheses, number of experiments.
|
|
LipoAmph B in the Presence of PLA2-II Enhances Platinum Drug
Cytotoxicity.
LipoAmph B was compared with Amph B (Fungizone) for its ability to
enhance platinum cytotoxicity in the parent HN-5a and
carboplatin-resistant 5a/carbo-15a lines. As a control, the effect of
the Amph B formulations on proliferation of both cell lines was tested
each time, in the absence of carboplatin (Fig.
1). Nearly all of the treatments were
somewhat inhibitory of proliferation, in an Amph B
concentration-dependent manner, thus limiting the concentration of Amph
B that could be used. The resistant line was more sensitive to
inhibition. Addition of PLA2-II increased the antiproliferative
activity of LipoAmph B. The combination of 25 µg/ml LipoAmph B and
400 U/liter PLA2-II, in the absence of any platinum drug,
reduced proliferation of 5a/carbo-15a cells to only 10% of the control
(data not shown). LipoAmph B (50 µg/ml) plus 200 U/liter PLA2-II
killed all 5a/carbo-15a cells (not shown). Desoxycholate, the
solubilizing agent for Fungizone, was used as a control for free Amph
B. None of the formulations that were later used along with platinum
drugs significantly inhibited proliferation in the absence of Amph B.
View larger version (60K):
[in this window]
[in a new window]
|
Fig. 1.
Effects of various treatments on
proliferation of HN-5a (A) and 5a/carbo-15a (B) cells.
Cells were exposed continuously over 4 days to the conditions
presented, as controls for the combined exposure to carboplatin or
cisplatin (Fig. 2). The data were obtained from the same experiments
presented in Figs. 2 and 3. The concentration of empty liposomes at 200 µg/ml is equivalent to that of the liposomes present in a preparation
of LipoAmph B with an Amph B concentration of 20 µg/ml. The
combination of Amph B and desoxycholate is constituted by the use of
Fungizone. The combination of Amph B and liposomes is constituted by
the use of LipoAmph B. Statistical analysis (t test,
compared with drug-free medium): *p < .05;
**p < .01; ***p < .001.
|
|
The enhancement of platinum-drug toxicity was measured using Amph B in
various formulations (Fungizone, or LipoAmphB ± PLA2-II). Data
for all treatment conditions on alteration of cytotoxicity are
presented comparing protocols with Amph B against the same formulations
without Amph B. In the experiments in Figs.
2 and 3,
all agents were coincubated with carboplatin or cisplatin for the 4-day
duration of the exposure. At 10 µg/ml, Amph B increased cytotoxicity
of carboplatin by approximately 50% in both the parent and
carboplatin-resistant lines (Fig. 2), whereas desoxycholate, the other
major component of Fungizone, caused no change.
View larger version (42K):
[in this window]
[in a new window]
|
Fig. 2.
Effects of various formulations of Amph B
on cytotoxicity of carboplatin against HN-5a (A) and 5a/carbo-15a (B)
cells.
Cells were exposed continuously (4 days) to the combinations of drugs
indicated. Bars indicate the change in IC50 of carboplatin
caused by the combination of components listed below the chart, at the
concentrations indicated. These values were derived by comparing the
carboplatin IC50 values between respective combinations
that contained or did not contain Amph B. The IC50 values
for carboplatin against the HN-5a and 5a/carbo-15a cell lines in the
experiments conducted for this figure are, respectively, 7.1 ± 2.1 µM (n = 19) and 20.1 ± 4.4 µM
(n = 20). The 5a/carbo-15a cell line was
2.95-fold ± 0.91 (n = 10) resistant compared
with the parent HN-5a line. The concentration of empty liposomes at 200 µg/ml is equivalent to that of the liposomes present in a preparation
of LipoAmph B with an Amph B concentration of 20 µg/ml. The
combination of Amph B and desoxycholate is constituted by the use of
Fungizone. The combination of Amph B and liposomes is constituted by
the use of LipoAmph B. For statistical purposes (t
test), the effect of each Amph B treatment was compared with its
respective control in the absence of Amph B: *p < .05; **p < .01; ***p < .001.
|
|
View larger version (44K):
[in this window]
[in a new window]
|
Fig. 3.
Effects of various formulations of Amph B
on cytotoxicity of cisplatin against HN-5a (A) and 5a/carbo-15a (B)
cells.
Cells were exposed continuously (4 days) to the combinations of drugs
indicated. Bars indicate the change in IC50 of cisplatin
caused by the combination of components listed below the chart, at the
concentrations indicated. The IC50 values for cisplatin for
this set of experiments, for HN-5a and 5a/carbo-15a, respectively, were
2.00 ± 0.64 µM (n = 6) and 4.89 ± 0.68 µM (n = 8). See legend to Fig. 2 for
additional information.
|
|
At a given concentration of Amph B, LipoAmph B was much less
antiproliferative than Fungizone (Fig. 1), but was also much less
capable of enhancing carboplatin cytotoxicity (Fig. 2). The liposomes
alone, at concentrations equivalent to those of LipoAmph B, did not
significantly alter the cytotoxicity of carboplatin against either cell
line (p > .5). At optimal concentrations of LipoAmph B and PLA2-II, this combination significantly enhanced the
cytotoxicity of carboplatin compared with treatment with PLA2-II and
empty liposomes (Fig. 2). LipoAmph B plus PLA2-II enhanced cytotoxicity
of carboplatin by up to 35% in the parent HN-5a line and up to 60%
(in individual experiments) in the resistant subline, approximating the
enhancement caused by free Amph B. Similar results were observed at the
level of the IC90 for carboplatin (data not shown) as that presented for the IC50.
The combination of LipoAmph B and PLA2-II was subsequently tested for
its ability to enhance cisplatin cytotoxicity in both cell lines, and
again the enhancement was similar to that achieved with free Amph B
(Fig. 3). In both cell lines, there was a significant enhancement of
cisplatin cytotoxicity, in the presence of LipoAmph B, by the addition
of PLA2-II. The incremental increase in PLA2-II concentration also
caused a significant increase in enhancement of cisplatin cytotoxicity
under the following conditions: 1) in HN-5a, 400 versus 100 U/liter
PLA2-II in 25 µg/ml LipoAmph B (p < .05); 2)
in 5a/carbo-15a, 200 versus 50 U/liter in the presence of 20 µg/ml
LipoAmph B (p < .01); and 3) 100 versus 50 U/liter in 25 µg/ml LipoAmph B (p < .05).
A 24-h Exposure to LipoAmph B plus PLA2-II Enhances Cytotoxicity of
a Subsequent 4-h Exposure to Cisplatin.
Cellular platinum uptake, in the absence or presence of modulators,
must be measured after 4 h to allow time for measurable amounts of
drug to accumulate, but before the cytotoxic action is manifested.
Therefore, the ability of Amph B to alter cisplatin cytotoxicity after
a 4-h exposure was first assessed. Figure
4 demonstrates equivalent enhancement of
cisplatin cytotoxicity in both cell lines by Amph B over a 4-h
exposure. Toxicity due to Amph B alone was minimal.
View larger version (30K):
[in this window]
[in a new window]
|
Fig. 4.
Effect of cotreatment with 10 µg/ml Amph
B on cytotoxicity of a 4-h exposure to cisplatin against HN-5a (open
columns) and 5a/carbo-15a (hatched columns) cell lines.
Exposures to drug were for 4 h, followed by 4 days in drug-free
medium. A, proliferation of cells after exposure to Amph B. B,
enhancement of cisplatin cytotoxicity by Amph B. The IC50
and IC90 values for cisplatin for this set of experiments
(4-h exposure, cisplatin alone) were, for HN-5a, 5.08 ± 1.20 µM
(n = 5) and 10.7 ± 2.3 µM
(n = 6), respectively, and for 5a/carbo-15a,
12.9 ± 0.3 µM (n = 3) and 25.2 ± 1.0 µM (n = 4), respectively. Statistical analysis
(t test, compared with cisplatin alone):
*p < .05; **p < .01;
***p < .001.
|
|
The combination of LipoAmph B and PLA2-II was then tested for its
effect on cytotoxicity of a 4-h cisplatin exposure. Figures 5 and 6
indicate the significant enhancement of cisplatin cytotoxicity by the
LipoAmph B/PLA2-II combination in both the sensitive and resistant cell
lines, respectively, equivalent to the effect observed during the
continuous exposure. Because Amph B must be liberated from LipoAmph B
by the digestion by PLA2-II, cells were incubated with the combination
of LipoAmph B and PLA2-II for 24 h before the 4-h exposure to
cisplatin. Because in previous experiments the empty liposomes were
shown to have little effect on cisplatin toxicity over an extended
period, in these experiments the influence of treatments on cisplatin
toxicity were compared with cisplatin in the absence of any other
treatment.
View larger version (41K):
[in this window]
[in a new window]
|
Fig. 5.
Effects of various treatments on
cytotoxicity of cisplatin against HN-5a cells (4-h exposure).
Cells were exposed for 24 h to the various treatments indicated,
followed by concomitant 4-h exposure to cisplatin, then 4 days drug
free. Control conditions in the absence of cisplatin (A) were assessed
as a control against which to normalize the data from assays that
included cisplatin (B). The IC50 value for cisplatin for
this set of experiments was 5.08 ± 1.20 µM
(n = 5). Statistical analysis (t
test, compared with cisplatin treatment alone): *p < .05; **p < .01.
|
|
View larger version (44K):
[in this window]
[in a new window]
|
Fig. 6.
Effects of various treatments on
cytotoxicity of cisplatin against 5a/carbo-15a cells (4-h
exposure).
Cells were exposed for 24 h to the various treatments indicated,
followed by concomitant 4-h exposure to cisplatin, then 4 days drug
free. Control conditions in the absence of cisplatin (A) were assessed
as a control against which to normalize the data from assays that
included cisplatin (B). The IC50 value for cisplatin for
this set of experiments was 12.9 ± 0.3 µM
(n = 3). Statistical analysis (t
test, compared with cisplatin treatment alone): *p < .05; **p < .01; ***p < .001. Data without error bars indicate results of a single experiment
(data from other, independent experiments with Amph B alone, but
without the 24-h pretreatment, are summarized in Fig. 4).
|
|
In HN-5a cells (Fig. 5), there were no statistically significant
differences in the enhancement of cytotoxicity between the different
concentrations of PLA2-II, although the trend was clear as the PLA2-II
concentration was increased. For the 5a/carbo-15a cells (Fig. 6), the
effect was not significantly different between 25 and 50 U/liter of
PLA2-II, but was significantly different (p < .05) between each increment in PLA2-II concentration. The enhancement
of cisplatin toxicity by LipoAmph B + 50 U/liter PLA2-II against HN-5a
cells was statistically significant (p < .05),
compared with a no-Amph B control, at the IC90
level of cisplatin (data not shown). A 24-h pre-exposure with free Amph
B reduced the effectiveness of the 4-h Amph B coexposure at enhancing
cisplatin cytotoxicity (compare with Fig. 4).
LipoAmph B plus PLA2-II Enhances Cellular Accumulation of
Cisplatin.
It was tested whether the enhancement of cisplatin cytotoxicity by the
various treatments could be attributed to an increase in cisplatin
accumulation. In the carbo-15a cell line, cisplatin accumulation was
61.6 ± 9.0% (n = 4) of that of HN-5a
(p < .01) (256 ± 67, n = 4, versus 415 ± 55, n = 6, nmol/mg protein).
Figure 7 demonstrates that LipoAmph B and
PLA2-II (24-h pre-exposure and 4-h coexposure) significantly enhanced
cisplatin accumulation in both cell lines, compared with cisplatin
alone, just slightly less than the increase caused by Amph B (4-h
coexposure only).
View larger version (32K):
[in this window]
[in a new window]
|
Fig. 7.
Effect of various treatments on cellular
cisplatin accumulation in HN-5a (open columns) and 5a/carbo-15a
(hatched columns) cell lines.
Cellular drug content was determined by exposing cells for 4 h to
50 µM cisplatin in growth medium, extracting cells in Triton-HCl (see
Materials and Methods), and measuring platinum on a
FAAS. The treatments indicated (10 µg/ml Amph B, 25 µg/ml
LipoAmphB, 200 U/liter PLA2-II, or LipoAmph B + PLA2-II) were for
24 h before and 4 h concomitant with the cisplatin exposure.
Drug content was calculated as a percentage of control within each
individual experiment, as a function of cellular protein content:
415 ± 55 nmol/mg (n = 6) for HN-5a; 256 ± 67 nmol/mg (n = 4) for 5a/carbo-15a
(significantly different, p < .01).
*p < .05 by paired t test, compared
with cisplatin alone.
|
|
| |
Discussion |
The data presented above describe the ability of a liposomal
preparation of Amph B, in the presence of PLA2-II, to enhance cytotoxicity of platinum drugs by increasing drug accumulation in
cultured tumor cells. Because anticancer drugs are dose limited by
patient toxicities, the degree of resistance required for a tumor cell
to survive a drug treatment can be conferred by a minor change in its
phenotype. Therefore, an increase of only 50% in tumor cell platinum
accumulation could be enough to overcome clinical levels of drug
resistance. Most mechanisms of resistance to platinum drugs could be
circumvented, to some degree, by enhancing cellular accumulation of the
drug, because the mechanisms rely on time dependence of repair or
uptake, or a limited store of intracellular drug scavengers (e.g.,
glutathione). The carboplatin-resistant cell line used in this study
appears to express the phenotypes of decreased drug uptake and
increased glutathione content, without any change in metallothionein
content. These resistance-associated phenotypes are evidenced by the
increased sensitivity to ouabain (Ohmori et al., 1993) and decreased
cisplatin uptake, cross-resistance to melphalan (Andrews et al., 1986)
(putatively glutathione-dependent), and lack of change in sensitivity
to cadmium (Singh et al., 1995) (Table 1). Increased glutathione could
contribute to enhanced cisplatin efflux (Ishikawa and Ali-Osman, 1993)
as well as increased DNA repair (Eastman, 1987; Andrews and Howell,
1990; Meijer et al., 1990). However, there is no evidence at present to
indicate whether DNA repair is altered.
Amph B usually enhances accumulation and cytotoxicity of platinum drugs
in both platinum-resistant and -sensitive cell lines (Morikage et al.,
1991a, 1993; Kikkawa et al., 1993; Sharp et al., 1994;
Beketic-Oreskovic and Osmak, 1995), but in some sensitive lines it has
no effect (Morikage et al., 1993; Sharp et al., 1994; Beketic-Oreskovic
and Osmak, 1995). Although the ionophore effect of Amph B should not be
ruled out completely as contributing to increased platinum uptake, the
ability to enhance cytotoxicity in some cell lines and not others
suggests a specific mode of action (Sharp et al., 1994). It has been
proposed that Amph B enhances uptake by acting directly on a putative,
metal-regulatory, gated pore protein, inducing it to assume the
"open" configuration (Gately and Howell, 1993). This pore protein
is suggested to be dependent on ion balance under the control of the
Na+,K+-ATPase (Gately and
Howell, 1993); a decrease in this enzyme activity yields decreased
cisplatin accumulation and increased sensitivity to ouabain. The
increase in ouabain sensitivity of the 5a/carbo-15a cell line suggests
involvement of altered
Na+,K+-ATPase activity in
resistance of this cell line.
The 4-h drug exposures allow for direct comparisons between drug
accumulation and cytotoxicity. This exposure period was used in many
other studies to distinguish the process of drug entry from the
intracellular binding, which occurs over longer exposures, and which is
dependent on various other potentially variable determinants. Amph B
enhanced cisplatin accumulation by the same percentage in both the
parent and resistant cell lines, as did the LipoAmph B/PLA2-II
combination, almost as effectively as Amph B (Fig. 7). The cellular
cisplatin content of carbo-15a cells induced by 25 µg/ml LipoAmph B
and 200 U/liter PLA2-II was equivalent to that of the untreated HN-5a
cells [Fig. 7; 395 ± 141 nmol/mg protein (n = 4)
and 415 ± 55 nmol/mg (n = 6), respectively],
mirroring the effect of this combination on the cytotoxicity of
cisplatin [compare Figs. 6 and 5: 5.32 ± 0.12 µM
(n = 2) versus 5.08 ± 1.20 µM
(n = 5), respectively].
In nude mice carrying human ovarian carcinoma, Amph B doubled the
survival time produced by carboplatin alone (Kojima et al., 1994).
However, the clinical usefulness of Amph B in enhancing platinum-drug
antitumor activity may be limited. Although the concentrations of Amph
B required to reverse cisplatin resistance are in the range of 5 to 30 µg/ml (depending on the cell line), the clinically achievable
plasma concentration is limited to 1 to 2 µg/ml (Sculier et al.,
1988; Barriere, 1990; Morikage et al., 1991a). Also, the nephrotoxicity
caused by free Amph B and by cisplatin could become synergistic if the
two are used in combination. Thus, the approach was taken to study
liposomally encapsulated Amph B, a preparation that has been shown to
substantially avoid the kidney toxicity observed with Amph B alone.
LipoAmph B is demonstrated to be safe in patients at efficacious doses,
which yield plasma concentrations of Amph B up to 20 µg/ml (Meunier et al., 1991).
It is hypothesized that a liposomal preparation of Amph B prevents
nephrotoxicity by maintaining a lower concentration of free Amph B than
efficacious doses of aqueous Amph B. The antifungal activity of
LipoAmph B depends on yeast cells' production of a variety of
phospholipases (Takahashi et al., 1991) that are thought to destabilize
the liposome, allowing free Amph B to be released in the immediate
vicinity of the organism. Optimal exploitation of LipoAmph B as a
chemomodulator of platinum-drug cytotoxicity may entail a process by
which free Amph B can be released in the local tumor milieu. PLA2-II,
which preferentially hydrolyzes zwitterionic phospholipids, such as
phosphatidylcholine (lecithin), effectively catabolizes the liposome
(Dr. G. Jensen, NeXstar Inc., personal communication). Both PLA2-I and
-II are found in the circulating plasma, but the latter is generally in
much higher amounts (Kortesuo et al., 1992), ranging from 2 U/liter in
healthy subjects up to 500 U/liter in patients with severe infections
and neoplasms (Kortesuo et al., 1992; Nevalainen et al., 1992;
Aufenanger et al., 1993; Rintala and Nevalainen, 1993). The PLA2-II
level in peritoneal and pleural effusions from patients with various
cancers generally averages between 40 and 60 ng/ml (approximately
100-150 U/liter), and up to 150 to 180 ng/ml (375-450 U/liter) in
some patients (Abe et al., 1997), similar to the in vitro
concentrations that were effective in this study. Also, PLA2-II is
frequently overexpressed in more aggressive colon and breast tumors
(Murata et al., 1993; Yamashita et al., 1994).
In this in vitro study, the action of PLA2-II appears to have released
Amph B from the liposomes. The antiproliferative action of PLA2-II plus
LipoAmph B (up to 45 and 70% against HN-5a and 5a/carbo-15a,
respectively; Fig. 1) was much greater than that of PLA2-II plus empty
liposomes (approximately 10-15% in HN-5a, 10-20% in 5a/carbo-15a,
specific data not shown), thus rejecting the possibility that the
release of free lipids caused the antiproliferative activity. It is of
interest that the carboplatin-resistant cells are more sensitive to the
antiproliferative action of Amph B than the parent line, a finding
observed by others (Beketic-Oreskovic and Osmak, 1995).
It was possible that the liposome would have an adverse effect on the
antitumor activity of a platinum drug, because, in several animal
studies, liposomally encapsulated platinum drugs had reduced or no
activity compared with free drug (Reszka et al., 1987; Steerenberg et
al., 1988; Fichtner et al., 1993). However, in the present study, free
liposomes alone did not significantly alter carboplatin cytotoxicity in
vitro (Fig. 2).
The efficacy of LipoAmph B in combination with platinum drugs has not
been well documented. LipoAmph B significantly elevated the activity of
cisplatin against cisplatin-resistant PC-14 in nude mice (Morikage et
al., 1991b). Otherwise, to our knowledge, there are no other reports on
the antitumor activity of the combination of LipoAmph B and cisplatin.
The treatment of 5a/carbo-15a cells with free Amph B or some
combinations of LipoAmph B plus PLA2-II completely abolished resistance
to cisplatin (this report). Also, resistance of 5a/carbo-15a cells to
carboplatin was reduced from 2.5- (at the IC90
level) to 1.3-fold, compared with the untreated parent line.
The use of LipoAmph B was proposed because this preparation largely
avoids the kidney toxicity of Amph B, but it was not clear whether
sufficient active Amph B would be released in the environment of the
tumor cells to enhance drug uptake. The data presented here suggest
that the phospholipase present in tumor microenvironments, as well as
in serum of some tumor-bearing patients, can potentially release
sufficient Amph B from LipoAmph B in the vicinity of the tumors to
enhance platinum-drug cytotoxicity. LipoAmph B could be particularly
effective in enhancing cisplatin toxicity in peritoneal tumors. The
potential efficacy of this combination warrants additional investigation.
We thank Dr. Paul Andrews (Georgetown University) for providing the
extraction procedure and burn program for measuring cisplatin uptake,
and Dr. George Cherian and Susanne Vesely (University of Western
Ontario) for technical assistance in operating the FAAS, as well as
making the machine available to us.
This work was supported by the Department of Otolaryngology,
University of Western Ontario; the Office of the Vice-President Research, University of Western Ontario; the Lawson Research Institute, St. Joseph's Health Center; the Ministry of Health, Province of Ontario; the Medical Oncology Research Fund, London Regional Cancer Center; and NeXstar, Inc., Boulder, CO.
![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
Top
Abstract
Introduction
