Inhibition of T-type Ca²⁺ channels by endostatin attenuates human glioblastoma cell proliferation and migration

Abstract

Background and purpose: Endostatin (ES) is a c-terminal proteolytic fragment of collagen XVIII with promising antitumour properties in several tumour models, including human glioblastoma. We hypothesized that this peptide could interact with plasma membrane ion channels and modulate their functions.

Experimental approach: Using cell proliferation and migration assays, patch clamp and Western blot analysis, we studied the effects of ES on the proliferation and migration of human glioblastoma U87 cells, mediated by T-type Ca²⁺ channels.

Key results: Extracellular application of ES reversibly inhibited T-type Ca²⁺ channel currents (T-currents) in U87 cells, whereas L-type Ca²⁺ currents were not affected. This inhibitory effect was associated with a hyperpolarizing shift in the voltage-dependence of inactivation but was independent of G-protein and protein tyrosine kinase-mediated pathways. All three α₁ subunits of T-type Ca²⁺ channels (Ca(V) 3), α(1G) (Ca(V) 3.1), α(1H) (Ca(V) 3.2) and α(1I) (Ca(V) 3.3), were endogenously expressed in U87 cells. Using transfected HEK293 or CHO cells, we showed that only Ca(V) 3.1 and Ca(V) 3.2, but not Ca(V) 3.3 or Ca(V) 1.2 (L-type), channel currents were significantly inhibited. More interestingly, ES inhibited the proliferation and migration of U87 cells in a dose-dependent manner. Pretreatment of the cells with the specific T-type Ca²⁺ channel blocker mibefradil occluded these inhibitory effects of ES.

Conclusion and implications: This study provides the first evidence that the antitumour effects of ES on glioblastoma cells is through direct inhibition of T-type Ca²⁺ channels and gives new insights into the future development of a new class of antiglioblastoma agents that target the proliferation and migration of these cells.

Figure 1

Characterization of voltage-gated Ca²⁺ channel currents in U87 human glioblastoma cells. Examples of traces and pooled data showing the effects of nifedipine (10 µM, A) or Bay K8644 (1 µM, B) on barium currents elicited by a 150-ms long depolarizing step pulse from the holding potential of −60 to 0 mV. (C) Examples of traces and pooled data showing the effects of NiCl₂ (100 µM) on T-currents. Currents with 10 mM barium as a charge carrier were elicited by a 80-ms-long depolarization step pulse from the holding potential of −90 to −30 mV. **P < 0.01 versus control, ***P < 0.001 versus control.

Figure 2

ES selectively inhibited T-currents. Examples of traces and pooled data of HVA L-type Ca²⁺ currents (A) or LVA T-currents (B) recorded under the control conditions and during exposure to 0.1 µM ES. (C) Time course of changes in amplitude of T-currents in control conditions, during exposure to 0.1 µM ES and washout. (D) Examples of traces and pooled data showing no inhibition of T-currents by intracellular application of ES. (E) Current–voltage (I–V) curve (evoked by a series of depolarizing pulses from a holding potential of −90 mV to test potentials between −80 and 0 mV, in 10-mV increments) for the inhibitory effects of 0.1 µM ES on T-currents. (F) Dose–response curve for the inhibitory effects of ES on T-currents. The line represents the best fit of the data points to the sigmoidal Hill equation (see Methods). Number of cells tested at each concentration of ES is indicated in parentheses. Time course (G) and pooled data (H) showing that pretreatment of cells with NNC 55-0396 (8 µM) completely abolished the inhibitory effect of ES on the T current; **P < 0.01 versus control.

Figure 3

ES shifted the steady-state inactivation curve in a hyperpolarizing direction. (A) The steady-state activation of T-type calcium channels was not altered by 0.1 ES. Tail currents were elicited by repolarization to −110 mV after 40 ms test pulses from −80 to 0 mV in increments of 10 mV. (B) ES shifted steady-state inactivation curve of T-type calcium channels to the hyperpolarizing direction. Steady-state inactivation curves were obtained by 40 ms test pulse to −30 mV after the 3 s conditioning pulses ranging from −110 to +10 mV with 10 mV increments. (C,D) Pooled data showing the changes in V_half and k (slope factor) indicated in (A) and (B), respectively.

Figure 4

G-protein and PTK are not involved in ES-induced T-current inhibition. Time course showing the effects ES (0.1 µM) on T-currents in the presence of GDP-β-S (1 mM, A) or lavendustin C (5 µM, B). Inset: an example of the current traces indicated, respectively, in (A) and (B). Numbers on plot indicate which points were used for sample traces. (C) Intracellular ATP-γ-S has no effect on ES-induced T-current inhibition. Time course of changes in amplitude of T-currents in U87 cells dialysed with a pipette solution containing 5 mM ATP-γ-S in the absence (a) or presence of 0.1 µM ES (b). (D) Pooled data showing the effects of ES (0.1 µM) on T-currents in the presence of GDP-β-S (1 mM), lavendustin C (5 µM) or ATP-γ-S (5 mM), respectively.

Figure 5

Expression analysis of α₁ subunits of T-type Ca²⁺ channels in U87 human glioblastoma cells. (A) RT-PCR analysis of voltage-gated Ca²⁺ channel α₁ subunit transcripts. RNA was isolated from cultured U87 cells and RT-PCR was performed using primers as described in Table 1. The PCR template from reverse transcription without RTase (–RT), or from water (H₂O) served as negative controls. HEK293 cells transfected with Ca_V3.1, Ca_V3.2 or Ca_V3.3 cDNA were used as corresponding positive controls. (B) Western blot analysis of α₁ subunits of Ca_V3 channels. Proteins extracted from cultured U87 cells were probed with goat polyclonal antibodies against the different Ca_V3 α1 subunits. Proteins extracted from HEK293 cells transfected with Ca_V3.1, Ca_V3.2 or Ca_V3.3 were used as corresponding positive controls. The molecular weight is shown in the left lane.

Figure 6

ES selectively inhibits Ca_V3.1 and Ca_V3.2, but not Ca_V3.3 T-type Ca²⁺ channels. (A–C) Left panels: representative traces showing the effect of 0.1 µM ES on cloned Ca_V3 channel currents elicited by a −30 mV test pulse. The holding potential (HP) was −80 mV. Right panels: current–voltage (I–V) profiles (evoked by a series of depolarizing pulses from a holding potential of −90 mV to test potentials between −80 and 0 mV, in 10-mV increments) obtained for cloned human α_1G (Ca_V3.1) (A, n= 11), α_1H (Ca_V3.1) (B, n= 13) and α_1I (Ca_V3.1) subunits (C, n= 9). (D) Pooled data showing the effect of 0.1 µM ES on cloned human Ca_V3.1, Ca_V3.2 and Ca_V3.3 channel currents indicated in (A), (B) and (C), respectively. ES blocked T-currents but not L-type Ca_V1.2 (α_1C) channel currents. Ca_V1.2 channel currents were elicited by a +10 mV test pulse applied from a HP of −80 mV. (E) ES at 0.1 µM produced a similar block of Ba²⁺ (10 mM) currents when α₁ subunits of Ca_V3 were expressed in CHO cells. **P < 0.01 versus control, ***P < 0.001 versus control.

Figure 7

Inhibition of T-type Ca²⁺ channels by ES attenuated U87 cell proliferation. (A) ES inhibited cell proliferation of U87 cells in a dose-dependent manner. U87 cells, 2 × 10⁶ per well, were seeded in 96-well plates and treated with different concentrations of ES for 3 days and were pulsed with 1 µCi [³H]-thymidine for the last 18 h. Cell proliferation values are expressed relative to those wells where no ES was added (100% control value). (B) Mibefradil (100 µM), the T-type Ca²⁺ channel blocker, but not nifedipine, a L-type Ca²⁺ channel blocker, inhibited U87 cell proliferation. Mibefradil at 100 µM occluded the inhibitory effects of cell proliferation induced by ES. (C) Western blot analysis of α_1G and α_1H expression in negative-control siRNA (ctrl siRNA) and α_1G/H siRNA-treated U87 cells. GAPDH was used as a positive control. (D) Pooled data showed the effect of α_1G/H siRNA on the inhibitory effects of 0.1 µM ES on cell proliferation in U87 cells. Values represent the mean ± SEM of six experiments. *P < 0.05 versus control, **P < 0.01 versus control.

Figure 8

Effects of ES on U87 cell migration. (A) ES inhibited U87 cell migration in a dose-dependent manner. U87 cells were seeded into the top Matrigel chamber with or without the addition of ES or mibefradil at various concentrations. A final concentration of 25 µg·mL⁻¹ fibronectin was added to the bottom chamber medium as a chemoattractant. (B) Mibefradil occluded the ES-induced inhibition of cell migration. Mibefradil (100 µM), but not nifedipine (10 µM), inhibited U87 cell migration (n= 5). Results represent means ± SEM of six experiments. *P < 0.05 versus control, **P < 0.01 versus control.