Potent Inhibition of Macropinocytosis by Niclosamide in Cancer Cells: A Novel Mechanism for the Anticancer Efficacy for the Antihelminthic

Abstract

Niclosamide, a drug used to treat tapeworm infection, possesses anticancer effects by interfering with multiple signaling pathways. Niclosamide also causes intracellular acidification. We have recently discovered that the amino acid transporter SLC38A5, an amino acid-dependent Na⁺/H⁺ exchanger, activates macropinocytosis in cancer cells via amino acid-induced intracellular alkalinization. Therefore, we asked whether niclosamide will block basal and SLC38A5-mediated macropinocytosis via intracellular acidification. We monitored macropinocytosis in pancreatic and breast cancer cells using TMR-dextran and the function of SLC38A5 by measuring Li⁺-stimulated serine uptake. The peptide transporter activity was measured by the uptake of glycylsarcosine. Treatment of the cancer cells with niclosamide caused intracellular acidification. The drug blocked basal and serine-induced macropinocytosis with differential potency, with an EC₅₀ of ~5 μM for the former and ~0.4 μM for the latter. The increased potency for amino acid-mediated macropinocytosis is due to direct inhibition of SLC38A5 by niclosamide in addition to the ability of the drug to cause intracellular acidification. The drug also inhibited the activity of the H⁺-coupled peptide transporter. We conclude that niclosamide induces nutrient starvation in cancer cells by blocking macropinocytosis, SLC38A5 and the peptide transporter. These studies uncover novel, hitherto unknown, mechanisms for the anticancer efficacy of this antihelminthic.

Keywords: Na+/H+ exchanger; SLC38A5; amino acid transporter; amino acid-mediated Na+/H+ exchange; antihelminthic; intracellular acidification; macropinocytosis; niclosamide; peptide transport.

Figure 1

Intracellular acidification caused by niclosamide (three different concentrations) in TNBC cell line TXBR-100 (A) and pancreatic cancer cell line HPAF-II (B). The experiment was repeated in another TNBC cell line (MB231) and pancreatic cancer cell line (BxPC3). The magnitude of cellular acidification was quantified for each concentration of niclosamide; data (means ± S.E.M.) for all four cell lines are given (C).

Figure 2

Comparison of the effects on intracellular acidification between niclosamide and pyrvinium. The time course of cellular acidification is shown in (A) for MB231 cells. The experiment was repeated for two additional cell lines (TXBR-100 and HPAF-II), and the magnitude of acidification for all three cell lines are shown (means ± S.E.M.) (B).

Figure 3

Effect of niclosamide on basal macropinocytosis. (A) Cellular uptake of TMR-dextran was used to monitor macropinocytosis activity in MB231 cells. The perifusion buffer contained NaCl. The assay was done in the absence or presence of niclosamide (2 μM). (B) The experiment was repeated in three other cell lines (TXBR-100, HPAF-II and BxPC3), and fluorescence signals were quantified as CTCF (corrected total cell fluorescence) for all four cell lines. Data (mean ± S.E.M.) are shown. ***, p < 0.001.

Figure 4

Effect of niclosamide on SLC38A5-mediated macropinocytosis. (A) Cellular uptake of TMR-dextran was used to monitor macropinocytosis activity in MB231 cells. The perifusion buffer contained NaCl and serine (1 mM). The assay was done in the absence or presence of niclosamide (2 μM). (B) The experiment was repeated in three other cell lines (TXBR-100, HPAF-II and BxPC3), and fluorescence signals were quantified as CTCF (corrected total cell fluorescence). Data are shown as mean ± S.E.M. ***, p < 0.001.

Figure 5

Dose-response for the inhibition of basal and SLC38A5-mediated macropinocytosis (TMR-dextran uptake) by niclosamide in HPAF-II (A) and TXBR-100 cells (B). Basal macropinocytosis was monitored in a NaCl-buffer without or with varying doses of niclosamide. The assay was done again, but using the NaCl-buffer with serine (1 mM). SLC38A5-mediated macropinocytosis was calculated by subtracting TMR-dextran uptake in NaCl-buffer from that in NaCl-serine buffer. Dose-responses for niclosamide to inhibit basal and SLC38A5-mediated macropinocytosis are shown.

Figure 6

Direct effect of niclosamide on SLC38A5-mediated serine uptake. Uptake of [³H]-serine (0.4 μM) was measured in HPAF-II (A) and TXBR-100 (B) cells in two buffers: NMDGCl-buffer, pH 8.5 and LiCl-buffer, pH 8.5. Both buffers contained 5 mM tryptophan. Uptake was measured in the absence or presence of varying concentrations of niclosamide. The uptake in NMDGCl-buffer was subtracted from the corresponding LiCl-buffer to calculate the serine uptake that occurred via SLC38A5. Data (mean ± S.E.M.) are given as % control (i.e., uptake in the absence of niclosamide). *, p < 0.05; **, p < 0.01; ***, p < 0.001.

Figure 7

Potentiation of inhibition by niclosamide on SLC38A5 with preexposure of the cells to the drug. (A) Serine uptake (0.4 μM) was measured in MB231 cells with niclosamide present only during uptake measurement. SLC38A5-mediated serine uptake was calculated as described in Figure 6. (B) The experiment was repeated again, but this time the cells were preexposed to 5 μM niclosamide for 15, 30 or 60 min prior to initiation of uptake in the presence of 5 μM niclosamide. Thus, niclosamide was present during preexposure as well as during uptake. (C) Lack of inhibition of SLC38A5-mediated uptake by pyrvinium. MB231 cells were used for serine uptake in the absence or presence of varying concentrations of pyrvinium during uptake. Measurements were made in two different buffers as described in Figure 6 to allow the determination of SLC38A5-mediated uptake. *, p < 0.05; ***, p < 0.001.

Figure 8

Effect of preexposure of MB231 cells to niclosamide on SLC38A5-mediated uptake. The cells were preexposed to niclosamide (2.5 μM) for 30 min in a NaCl-buffer. DMSO was used for control. Following the preexposure, cells were washed thrice with the same NaCl-buffer, but without niclosamide or DMSO. Serine uptake (0.4 μM) was then initiated in two different buffers (NMDGCl-buffer and LiCl-buffer), both at pH 8.5 and both containing 5 mM tryptophan. Niclosamide was not present during uptake. Data (means ± S.E.M) are given as percent of control (i.e., SLC38A5-specific serine uptake in cells preexposed only to DMSO). ***, p < 0.001.

Figure 9

Molecular docking of glutamine, niclosamide and pyrvinium to SLC38A5. (A) Predicted model for SLC38A5 as it is present in the lipid bilayer. Binding of glutamine (B), niclosamide (C) and pyrvinium (D) to the binding pocket of SLC38A5. The ligands (glutamine, niclosamide and pyrvinium) are shown as stick models. The amino acid residues predicted to be involved in the binding of the three ligands are also shown.