Intracellular alkalinization induced by amphotericin B derivatives in HL-60 leukemia cells

Biochirnie 71 (1989) 67-70 (~ Soci6t6de Chimie biologique /Elsevier, Paris

67

Intracellular alkalinization induced by amphotericin B derivatives in HL-60 leukemia cells Eric C H A R R E T I E R and Jaime W I E T Z E R B I N

Service de Biophysique, D~partement de Biologie, C E N / Saclay, 91191 Gif-sur-Yvette Cedex, France (Received 10-1-1988, acceptedafter revision 23-6-1988)

Summary - - The effects on intra- and extracellular pH of two polyenic derivatives of amphotericin B, N-fructosyl amphotericin B and N-fructosyl amphotericin B methyl-ester, were tested on HL-60 promyelocytic leukemia cells. Both derivatives raised the internal pH and reduced the external pH in weakly buffered medium. These results support the idea that both derivatives induce outward proton movement from the cell to the external solution. In this respect, the non-estefified derivative proved to be more powerful that the esterified one. Under the present, conditions, there was little or no regulation of pH in HL-60 cells, which exhibited an almost constant pH gradient between the external and internal pH (acid inside relative to outside). This deficiency in pH homeostasis might be due to the immature state of the HL-60 cells. amphotericin B / HL 60 / intrecellular pH

introduction

Materials and methods

The pharmacological use of amphotericin B (AmB) in anti-fungal therapy is based on its high activity against ergosterol-containing membranes of microorganisms. Recent research has revealed that this drug also has anti-tumoral and immunostimulatory properties, which have been confirmed against the human leukemia cell line HL-60 [1]. Furthermore, there is a great deal of evidence that the internal cell pH (pH0 acts as a regulator of cellular functions, especially those that are essential for proliferation and development [2]. Since AmB is known to increase the membrane permeability of small ions, e.g., H ÷, we wondered whether its immunomodulating properties are related to changes in pHi, The cytotoxic effect of AmB prompted us to study two less toxic derivatives: N-fructosy! AmB (FAmB) and N-fructosyl AmB methyl ester (FAmBME) [3]. This work constitutes a preliminary investigation of the action of these derivatives on the pHi of HL-60 cells.

FAmB and FAmBME were a generous gift from Dr. E. Borowski, Technical University of Gdansk, Poland. The two compounds were dissolved in dimethyl sulfoxide (DMSO), whose final concentration in the samples never exceeded 0.5% and did not affect the HL-60 cells. HL-60 cells were a kind gift from Drs. Jeanne Wietzerbin and Corrine Gaudelet, both of the Institut Curie, Paris. The cells were grown in suspension in RPMI 1640 medium (Gibco, Grand Island, NY) with 10% heat-inactivated fetal calf serum supplemented with 300 #g / ml of glutamine, 100 U / ml of penicillin, 100 #g / ml of streptomycin and 50 ftg/ ml of gentamycin. Stock cultures were maintained at 37°C and 5% CO2 in air and subcultured twice weekly. For the experiments, stock cells at the exponential growth stage were centrifuged at 100 rpm for 5 min, washed twice and resuspended in weakly buffered medium A (150 mM NaCl and 1 mM sodium phosphate, pH 7.4) and strongly buffered medium B (140 mM NaCl, 10 mM KCI and 30 mM Hepes) for pH meter and fluorescence experiments, respectively. At an FAmB or FAmBME concentration of

E. Charretier and J. Wietzerbin

68

2 x 10 -5 M, HL-60 cell viability, tested by trypan blue dye exclusion was 85%.

pH meter experiments The cells (106 /rnl) in the weakly buffered medium A were kept for 1 h at 37°C and then transferred into plastic cones, under a nitrogen flow with gentle stirring at 37°C. The variations in external pH (pHo) produced by the AmB derivatives were recorded with a pH meter (Radiometer, Copenhagen) connected to a Y vs time recorder. PHo was measured after pHo had reached stable values. When calculating proton release from the HL-60 cells, allowance was made for the buffering power of cell suspensions, determined with 0.1 M NaOH in isotonic sulfate. Total proton release was measured by disrupting the cells with 0.1% Triton X-100. Proton release is expressed as a percentage of this total value.

Fluorescence experiments IntraceUular pH was determined by measuring the fluorescent intensity of 6-carboxy fluorescein (6CF) (Molecular Probes). Cells (106 cells/ml) in medium B were loaded with 10/zmol of the ester derivative of 6CF for 30 rain at 25oC and pHo 6. The nonfluorescent lipophilic ester derivative crossed the plasma membrane and was hydrolyzed by nonspecific enzymes to the non-permeant fluorescent product 6CF. External fluorescence was removed by .pelleting the cells at 1000 rpm for 5 min and transferrmg them into medium B at the appropriate prin. All experiments were performed at 37oC in a PerkinElmer spectrofluo~meter with excitation at 490 nm ullu

~i,e t z t i o O l ~ l i

at

~Jb~J ii111 Ualll~

J

IlUi

blll,~.

L,I~IIt

~ttttgl-

ing accounted for less than 1% of the total signal. Cytoplasmic dye fluorescence as a function of pl-Il was obtained from H ÷ equilibration using the K + / H ÷ ionophore nigericin (10/zM), as described by Thomas et al. [4]. In all pH~ calculations, allowance was made for intracellular self-quenching and probe leak. Selfquenching was measured by disrupting the cells with

0.1% Triton X-100 after equalization of pH~ and pHo with 10/.~M nigericin. Self quenching, expressed as the ratio of fluorescent intensity after Triton X - 1 0 0 action to its intensity before this action, was 1.2, regardless of prin. Leak kinetics were established with at~d without FAmB and FAmBME, respectively, by centrifuging the cells and measuring the fluorescence in the supernatant, in order to ascertain, foi each experimental period, the fraction of external fluorescence contributing to the signal.

Results pH meter experiments We measured pHo in the presence of 10 -5 or 2 x 10-5 M F A m B or F A m B M E . Proton release, measured as the percentage of the total titratable proton content of the cells, was greater with F A m B than with F A m B M E . A t a concentration of 10 -5 M, F A m B M E had no effect on proton leakage (Table I).

Fluorescence e x p e r i m e n t s Fig. 1 depicts a typical experiment showing the action of nigericin, Triton X - 1 0 0 and both A M B derivatives on the fluorescence of HL-60loaded cells at pHo 6.7. Nigericin ( 1 0 / z g / m l , arrow N I G , Fig. 1) eiicited a rise in fluorescent intensity (panel A), indicating cell alkalinization, and reached a plateau after about 10 min, due to p H equilibration (pHI = pHo). This alkalinization was also observed at pHo 7.25 and 7.4 (not shown). Panel B of Fig. 1 depicts the effects of 10 -5 and

Table !. Proton leakage measured as the percentage of total proton release from HL-60 cells induced by 0.1% Triton X-100. Concentration (M)

Derivative 10-5

2x10-5

N-Fructosyl Arab

!5.0±0.2%

20.0_+0.3%

N-Fructosyl Aml~ methyl ester

-

10.0+0.2%

106cells/ ml were suspended in weakly buffered medium (150 mM NaCI, 1 mM sodium phosphate, pH 7.4 37oC). Data are means +- SD for 4 experiments.

69

Alkalinization in HL-OO cells pH

A

-" . 0-67

Buffer

~,,r._

30pl

. ~

Triton

Jm~ - 610

'56(~ " "

IO'SM

2.iO-SM

_,._~ s~o .S 510

G w tq W r.,. 0

1

560

C

= M.

520

! !

5

min.

.

'

Triton

Fig. 1. Recordings of fluorescent intensity at pHo 6.7 and 37'~2. 10~, cells/ml were loaded with 6CF suspended in strongly buffered medium (140 mM NaCI, 10 mM KC! and 30 mM Hepes). A shows the effect of buffer plus nigericin (NIG), and B and C show the effects of N-fiuctosyl Arab and N-fructosyi AmB methyl ester, at 10-5 and 2x 10-5 M, respectively. The final Triton concentration was 0.1%. a.u. = arbitrary units

2x 10 -5 M FAmB. The initial rapid decrease merely arose from dilution, which was also observed when only buffer was added (panel A). Panel C shows that the action of FAmB was stronger than that of FAmBME. If the initial artifactual decrease in the signal is disregarded, a rise in fluoremence occurred in all cases, reflecting an increase in pHI. From this type recording, we were able to calculate the pHI of HL-60 cells before adding nigericin, as well as the degree of alkalinization of these cells by the AmB derivatives, taking into account the intracellular fluorescent quenching and fluorescent leak (see Materials and methods). Table II shows that at pHo of 6.7, 7.25

T a b l e 11. Intracellular p H (pHI) as a function of the external p H (pHo).

pHo 6.70

7.25

7.40

pHt

6.56

7.13

7.20

ApH

0.14

0.12

0.20

10c'cells / ml were loaded with 6CF and suspended in strongly buffered medium (140 mM NaCi, l0 mM KCI, and 30 mM Hepes, 37oC). Values are means of 3 experiments with a maximum -+ SD of 0.02,

70

E. Charretier and J. Wietzerbin

and 7.4, pH~ was always more acidic than pHo. At these three pHo, 2×10 -5 M of both antibiotics induced a cell alkalinization which was measured by the increase in pHI (ApHI, Table III). FAmB was more effective in this respect than F A m B M E which, at pHo 6.7 and 7.25, had no effect. At 10 -5 M, neither derivative had any significant effect on pHI. In contrast, 10 -5 M FAmB induced a 15% H ÷ release from HL-60 cells (Table I). The reason for this difference was that the buffering power of the cells was 30 times greater than that of the surrounding medium, indicating that the pH meter technique is more sensitive than the fluorescent method (Table III).

Table Ill. Internal alkalinization (A pHi) induced by N-fructosylamphotericin B (FAmB) and its methyl ester derivative (FAmBME) at external pH (pHo) of 6.7, 7.25 and 7.40. FAmB and FAmBME concentrations were 2x 10-5 M. At 10-5 M, neither drug had any effect. pHo

FAmB

April

FAmBblE A pHI

6.70

7.25

7.40

0.12

0.11

0.11

NS

NS

0.05

Values are means of 3 experiments with a maximum -- SD of 0.03. NS: not significant,

Another possible explanation for the increase in pHI is the sequestration of H ÷ in various cell organelles, but this does not account for the decrease in pH observed in the pH meter experiments. However, these experiments enabled us to estimate the amount of protons released into the aqueous medium, taking into account the buffering power of the external solution. For 2×10 -5 M FAmB, the value calculated was 6.3+_0.8 x 10 -15 mol of H + per cell. The same calculation was made for 2 × 10-5 M FAmB from the pHI fluorescence data, on the basis of a cell buffering power of 30 .ttM/ ml.pH [6] and a cell volume of 520-/zm 3. This led to an H + loss of 4.9__.1.2 x 10-15 mol per cell, which is not significantly different from the pH meter value calculated above. The similarity of these values suggests that internal alkalinization was probably due to proton movement from the intracellular medium to the external solution. Table II shows that the pH gradient was not reversed for any of three pHo values tested and that changes in the pHo were followed by changes in the pHI. Although the reason for this lack of cell pH homeostasis is not yet clear, it might be related to the immature state of HL-60 cells.

Acknowledgments We gratefully acknowledge the assistance of Mathiide Dreyfus for English language editing and Patricia Belle for typing the manuscript.

Discussion The results of the fluorescence experiments indicate that at 2x 10-5 M, FAmB raised the pHx of HL-60 cells more than FAmBME. This finding is in agreement with previous results [5], which showed the dependence of the activity of these antibiotics upon the presence of the free carboxyl group. Several different mechanisms could be sugge~i.ed to explain tlhe alkalinization observed here. One of them might be the rise in pH~ due to the metabolic changes induced by differentiating agents, such as AmB derivatives. However, this is unlikely, owing to the rapid alkalinizatiot: kinetics recorded in our experiments (8 min)

References 1 Ciomei M., Pastori W. & Giuliani F.C (1986) Tumori 72, 545- 551 2 McJlenaar W.H. (1986) Trends Biochem. Sci.

11,141-143

3 Szponarski W., Wietzerbin J., Borowski E. & Gary-Bobo C. (1987) Biochim. Biophys. Acta 938, 97-106 4 Thomas J.A., Buchsbaum N., Zimniak A. & Racker E. (1979) Biochemistry 18, 2210-2218 5 Cybulska B., Ziminsky T., Borowski E. & GaryBobo C.M. (1983) Mol. Pharmacol. 24,270-276 6 Roos A. & Boron W.F. (1981) PhysioL Rev. 61, 296-434