Thermal sensitisation by lonidamine of human melanoma cells grown at low extracellular pH

Abstract

Purpose: This study tested the ability of lonidamine (LND), a clinically applicable inhibitor of monocarboxylate transporters (MCT), to thermally sensitise human melanoma cells cultured at a tumour-like extracellular pH (pH_e) 6.7.

Materials and methods: Human melanoma DB-1 cells cultured at pH_e 6.7 and pH_e 7.3 were exposed to 150 µM LND for 3 h, beginning 1 h prior to heating at 42 °C (2 h). Intracellular pH (pH_i) was determined using 2′,7′-bis-(2-carboxyethyl)-5-(and-6)-carboxyfluorescein (BCECF) and whole spectrum analysis. Levels of heat shock proteins (HSPs) were determined by immunoblot analysis. Cell survival was determined by colony formation.

Results: Treatment with LND at pH_e 6.7 reduced pH_i to 6.30 ± 0.21, reduced thermal induction of HSPs, and sensitised cells growing at pH_e 6.7 to 42 °C. When LND was combined with an acute acidification from pH_e 6.7 to pH_e 6.5, pH_i was reduced to 6.09 ± 0.26, and additional sensitisation was observed. LND had negligible effects on cells cultured at pH 7.3.

Conclusions: The results show that LND can reduce pH_i in human melanoma cells cultured at a tumour-like low pH_e so that the 42 °C induction of HSPs are abrogated and the cells are sensitised to thermal therapy. Cells cultured at a normal tissue-like pH_e 7.3 were not sensitised to 42 °C by LND. These findings support the strategy that human melanoma cells growing in an acidic environment can be sensitised to thermal therapy in vivo by exposure to an MCT inhibitor such as LND.

• lonidamine
• thermal sensitization
• human melanoma cells
• intracellular pH
• heat shock proteins

Introduction

Lonidamine (LND) inhibits the plasma membrane monocarboxylate transporter 1 (MCT-1) [1] and the mitochondrial pyruvate carrier (MPC) [2,3]. Treatment of malignant cells and tumours with LND increases lactate production and inhibits lactate efflux, reduces intracellular pH (pH_i), and decreases the rate of oxygen utilisation [1,3–7]. LND has been reported to sensitise cells and tumours to hyperthermia [4,6,8] and a variety of chemotherapeutic drugs [3,9–12]. Thermal sensitisation by LND can be attributable primarily to intracellular acidification.

Acute extracellular acidification is a potent thermal sensitiser, with sensitisation greater for temperatures below 43 °C than above 43 °C [13,14]. However, it is the decrease in pH_i accompanying acute lowering of extracellular pH (pH_e) that correlates with thermal sensitisation [15–19]. Acute acidification sensitises to hyperthermia in part by abrogating thermal induction of heat shock proteins (HSP) [19,20]. Human melanoma cells growing at a tumour-like pH_e 6.7 contain higher constitutive levels of HSP than in cells growing at pH_e 7.3. This explains why human melanoma cells growing at pH_e 6.7 are resistant to hyperthermia compared to cells growing at pH_e 7.3 [19,20]. It is important, therefore, to block additional synthesis of HSPs during thermal therapy. Acute intracellular acidification sensitises cells and tumours to chemotherapy with platinum [21] and nitrogen mustards [22–25], as well as to hyperthermia [8,9,18,19].

Human melanoma cells depend primarily on MCT-1 to maintain pH homeostasis [20,26]. Since intracellular acidification sensitised DB1 human melanoma cells to mild hyperthermia (42 °C) [19], it was proposed that inhibition of MCT-1 would sensitise the cells to hyperthermia. Inhibition of MCT-1 with alpha-cyano-4-hydroxy-cinnamic acid (CHC) reduced pH_i in DB-1 cells cultured at pH_e 6.7, abrogated induction of HSPs, and sensitised the cells to 42 °C [20]. However, CHC is not translatable to the clinic. The purpose of this study was to determine whether LND, by virtue of its ability to reduce pH_i, would sensitise DB1 cells grown at a tumour-like low pH to hyperthermia.

The clinical experience on combining LND and chemotherapy and/or radiation therapy in the treatment of solid tumours has been reviewed [12]. The available data demonstrate a significant role of LND in modulating anthracycline and platinum compound activity. The encouraging results of phase II and III trials for the treatment of advanced breast, ovarian and lung cancer still must be confirmed by larger studies.

Materials and methods

Cell culture

The human melanoma cell line, DB-1, was derived from a patient biopsied at Thomas Jefferson University [19]. DB-1 cells express human melanoma antigens [27]. Cells were maintained in exponential growth at pH_e 7.3 or 6.7 in alpha-MEM/10% FBS (pH adjusted by modifying the sodium bicarbonate concentration) as described previously [19]. The osmolality was kept constant.

Inhibitor

LND (1-[(2,4-dichlorobenzyl)methyl]-1H-indazole-3-carboxylic acid) (Sigma, St Louis, MO, USA) was solubilised in DMSO. Cells were exposed to 150 µM LND for 1 h at 37 °C in complete medium prior to heating at 42 °C (2 h). DMSO controls were performed for all experiments.

Measurement of pH_i

Cells plated onto 18 mm glass coverslips (attached to the bottom of a 1 cm diameter hole in the center of plastic Petri dishes) were loaded with BCECF-AM as described previously [19,20] with the following exceptions. After allowing for hydrolysis of the dye ester, the Petri dish was placed into a temperature-controlled microscope stage chamber (Instec, Boulder, CO) for study at 37 °C under humidified air containing 5% CO₂. pH_i was calculated based on analysis of whole excitation spectra of attached cells as has been described [19,20,28–31]. Felix32 software (Photon Technology International (PTI), Lawrenceville, NJ) was used to generate whole excitation spectra using the RatioMaster Model RM-5/2003 microscope-based ratio fluorescence spectrometer system (PTI) integrated with a Nikon Eclipse TS100 (Nikon Instruments Inc., Melville, NY) inverted fluorescence microscope and a Plan Fluor 40X objective. The initial steady-state pH_i at pH_e 7.3, 7.1, 6.7 or 6.5, was measured several times on a field of 8–15 cells and then the medium was replaced with medium containing 150 µM LND. The pH_i was measured on a single field of 8–15 cells every 5 min for up to 1 h. The reported pH_i values are those determined after 1 h. The pH_e values of the medium in the microscope chamber Petri dishes were measured immediately following the pH_i measurements to confirm that the pH_e of the medium had not changed.

Clonal survival

All tissue culture flasks (25 cm²) used for colony survival assays had a total of 10⁵ cells, comprised of irradiated (30 Gy) Chinese hamster ovarian carcinoma feeder cells and unirradiated DB-1 cells. All media in tissue culture flasks were replaced with fresh, pre-warmed (37 °C), CO₂-equilibrated media with or without LND at the treatment pH_e 1 h prior to the hyperthermia treatment. The cells were heated by submersion of flasks in 42 °C water baths. Media was replaced with fresh, pre-warmed (37 °C), CO₂-equilibrated media without LND at the growth pH_e following heat treatment. The flasks were maintained in a humidified CO₂ incubator at 37 °C for colony formation.

Immunoblot analysis

Samples were prepared for SDS-PAGE and western immunoblot analysis as described in detail previously [19,20]. Cohorts of cells were solubilised for SDS-PAGE just before addition of LND, and immediately following the 2-h 42 °C treatment. Cells exposed to LND alone were processed following the 3-h exposure to drug. Immunodetection was performed by enhanced chemiluminescence using a Tropix Western-Star protein detection kit (Applied Biosystems, Foster City, CA). The relative protein content of individual bands on X-ray film was determined by scanning laser densitometry using ImageQuant software (Molecular Dynamics, Sunnyvale, CA). The absorbance (A) obtained for each HSP band was normalised to the A of the GAPDH band in that lane and expressed as a percentage of the respective normalised HSP band from unheated cells at their growth pH_e.

Results

Effect of LND on pH_i

The steady-state pH_i of DB-1 cells cultured at pH_e 7.3 or pH_e 6.7 was 7.20 ± 0.07 or 6.76 ± 0.04, respectively. A 1-h, 37 °C treatment with LND reduced the steady-state pH_i in cells cultured at pH_e 7.3 to 7.15 ± 0.16. However, a 1-h LND treatment reduced the steady-state pH_i in cells cultured at pH_e 6.7 to 6.30 ± 0.21. A 0.2-pH unit acute acidification combined with exposure to LND reduced pHi in cells grown at pH_e 7.3 or pH_e 6.7 to 6.93 ± 0.15 or 6.09 ± 0.26, respectively. The reduction in pH_i plateaued after 15 min and was maintained for the next 45 min of exposure after which measurement was discontinued. This decrease in steady-state pH_i indicates that MCT transport is constitutively active at pH_e 6.7. Table I summarises the results of the pH_i studies.

Table I. Summary of the effects of 150 μM LND on pH_i in DB-1 cells.

Treatment pHe	Control	150 μM lonidamine
7.3	7.20 ± 0.07 (6)	7.15 ± 0.16 (3)
7.3–7.1	7.05 ± 0.04 (3)	6.93 ± 0.15 (4)
6.7	6.76 ± 0.04 (5)	6.30 ± 0.21 (9)
6.7–6.5	6.52 ± 0.15 (7)	6.09 ± 0.26 (6)

The pH_i values (mean ± SD) are after 60 min at 37 °C. Extracellular pH (pH_e) was measured immediately following the 60 min of pH_i monitoring. The number of separate experiments is in parentheses.

Effect of LND on 42 °C-induction of HSP

Figure 1 summarises the effects of LND on the induction of Hsp70 and Hsp27 by hyperthermia. Hsp70 and Hsp27 levels were measured before and immediately following 2 h of heating. HSP levels increased within 10–40 min from the start of 42 °C heating of mammalian cells [32] and the HSP levels continued to increase during 6 h of heating [32,33]. Exposure to 150 µM LND minimally suppressed the 42 °C induction of Hsp70 and Hsp27 in cells grown at pH_e 7.3. However, LND treatment abrogated the heat-induced increase in levels of HSP, especially Hsp70, in cells grown at pH_e 6.7.

Figure 1. Abrogation of induction of HSP70 (A) and HSP27 (B) during heating by 150 μM LND. Cells grown at pH_e 7.3 and pH_e 6.7 were exposed to LND for 60 min prior to and during heating (42 °C, 2 h) without or with a 0.2-pH unit acute acidification. Flasks of cells were processed for immunoblot studies before treatments and immediately after the 2 h of heating. HSP levels are presented as percent of HSP levels in unheated cells at their growth pH_e. The means (bars) and standard errors represent results from two different experiments. 37 °C (white); LND, 37 °C (stippled); 42 °C (gray); LND + 42 °C (black).

Effect of LND on cell survival

Figure 2 summarises the effects of treatments on cell survival. Exposure to LND at 37 °C had little effect on survival of cells grown at either pH_e 7.3 or 6.7. Heating alone reduced survival for all conditions. While LND exposure sensitised cells grown at pH_e 6.7 to 42 °C by a factor of 10, the inhibitor did not sensitise cells grown at pH_e 7.3 to the heat treatment. Furthermore, an acute 0.2-pH unit acute acidification of cells grown at pH_e 6.7 was not required for sensitisation by LND as it was for CHC.

Figure 2. Enhancement of cell killing by exposure of DB-1 cells to 150 μM LND prior to and during heating (42 °C, 2 h). Cells grown at pH_e 7.3 and pH_e 6.7 were exposed to LND for 60 min prior to and during heating (42 °C, 2 h) without or with a 0.2-pH unit acute acidification. The means (bars) and standard errors represent results from two different experiments. 37 °C (white); LND, 37 °C (stippled); 42 °C (gray); LND + 42 °C (black).

Discussion and conclusion

This study documents in vitro that the MCT inhibitor LND reduces pH_i in melanoma cells growing in a tumour-like acidic pH_e environment such that they are sensitised to hyperthermia. The previously studied MCT inhibitor, CHC, required a 0.2-pH unit acute acidification to sensitise DB-1 cells growing at pH_e 6.7 to 42 °C [20]. However, the LND-induced reduction of pH_i in DB-1 cells, resulting in abrogation of thermally induced HSP and thermal sensitisation, did not require an acute extracellular acidification. These findings suggest that hyperglycaemia may not be necessary to drive glycolysis in order to achieve tumour acidification by LND. Indeed, it was recently reported that LND reduced the pH_i of DB-1 xenografts from 6.90 ± 0.05 to 6.33 ± 0.10 in nude mice without a prior induction of hyperglycaemia [3].

Thermal sensitisation by LND is attributed to reduction in pH_i [4,6,7] and reduction in energy status [5]. The effect of LND on intracellular acidification in cells and tumours is attributed to increased production of lactate and inhibition of lactate efflux [1,3,7], inhibition of mitochondrial respiration by inhibiting pyruvate entry into mitochondria [2,3], and reduction in energy status [3,5]. These effects can be attributed to the drug’s inhibition of MCT [3,34–36] and MPC [2,3].

In conclusion, the results with LND support our previous findings with the MCT inhibitor CHC. However, an accompanying acute acidification is not required for reduction of pH_i by LND to a value that abrogates the 42 °C-induced increase of HSP levels and sensitises to 42 °C hyperthermia. Accordingly, LND should be a good candidate for sensitisation of solid human tumours to thermal therapy [37].

Declaration of interest

This research was supported by grant no. P01 CA56690 from the National Cancer Institute, the National Institutes of Health, of the US Department of Health and Human Services. The authors alone are responsible for the content and writing of the paper.

Acknowledgements

We thank Phyllis R. Wachsberger for proofreading this manuscript.