Modulation of glioma BK channels via erbB2

Abstract

Glioma cells show up-regulation and constitutive activation of erbB2, and its expression correlates positively with increased malignancy. A similar correlation has been demonstrated for the expression of gBK, a calcium-sensitive, large-conductance K(+) channel. We show here that glioma BK channels are a downstream target of erbB2/neuregulin signaling. Tyrphostin AG825 was able to disrupt the constituitive erbB2 activation in a dose-dependent manner, causing a 30-mV positive shift in gBK channel activation in cell-attached patches. Conversely, maximal stimulation of erbB2 with a recombinant neuregulin (NRG-1beta) caused a 12-mV shift in the opposite direction. RT-PCR studies reveal no change in the BK splice variants expressed in treated glioma cells. Furthermore, isolation of surface proteins through biotinylation did not show a change in gBK channel expression, and probing with phospho-specific antibodies showed no alteration in channel phosphorylation. However, fura-II Ca(2+) fluorescence imaging revealed a 35% decrease in the free intracellular Ca(2+) concentration after erbB2 inhibition and an increase in NRG-1beta-treated cells, suggesting that the observed changes most likely were due to alterations in [Ca(2+)](i). Consistent with this conclusion, neither tyrphostin AG825 nor NRG-1beta was able to modulate gBK channels under inside-out or whole-cell recording conditions when intracellular Ca(2+) was fixed. Thus, gBK channels are a downstream target for the abundantly expressed neuregulin-1 receptor erbB2 in glioma cells. However, unlike the case in other systems, this modulation appears to occur via changes in [Ca(2+)](i) without changes in channel expression or phosphorylation. The enhanced sensitivity of gBK channels in glioma cells to small, physiological Ca(2+) changes appears to be a prerequisite for this modulation.

Fig. 1

TyrAG825 decreases erbB2 phosphorylation. A: D54 glioma cells were treated with increasing concentrations (0–50 μM) of TyrAG825. Cell lysates were immunoprecipitated with erbB2-specific antibodies, and then blots were probed with a phosphotyrosine antibody. These blots were then stripped and reprobed with erbB2 as a loading control. Bands on this blot correspond to a molecular weight at ∼190 kD. B: Similarly, when treated cell lysates were immunoprecipitated with a phosphotyrosine antibody and blots probed with erbB2, a similar decrease in tyrosine phosphorylated erbB2 was observed.

Fig. 2

TyrAG825 and Nrg-1β shift the activation threshold of BK channels in glioma cells. A: Current responses at five different voltages from representative cell-attached patches of control, TyrAG825-treated (50 μM), and Nrg-1β-treated (80 nM) cells. B: The mean conductance from several cells under each condition as a function of voltage. Data points are the mean ± SE, and n = the number of cells under each condition.

Fig. 3

Biotinylation experiments demonstrate that total and surface expression of BK channels in glioma cells is unchanged. D54 glioma cells were treated with vehicle (serum-free media and 0.1% DMSO), Nrg-1β or TyrAG825 for 12 hr. The cells were biotinylated and lysates collected. The same amount of protein in each treatment condition was incubated with strepavidin beads (surface protein). Blots were probed with an anti-BK antibody and then stripped and reprobed with Na-K-ATPase antibody and an anti-actin antibody to demonstrate surface expression and a lack of intracellular proteins in surface lanes.

Fig. 4

Tyrosine, serine, and threonine phosphorylation status of BK channels is unchanged in Nrg-1β- or TyrAG825-treated cells. D54 cells were treated overnight with vehicle, Nrg-1β, or TyrAG825, and cells were lysed and protein harvested. A: Cell lysates were immunoprecipitated with an anti-BK antibody (Chemicon) and then probed with an anti-BK (top), antiphosphotyrosine (top, middle), antiphosphothreonine (bottom, middle), and antiphosphothreonine (bottom). Each band in this figure had an approximate molecular weight of 120 kD. B: Equal amounts of protein from whole-cell lysates and the unbound fractions were analyzed by Western blot analysis to demonstrate a complete immunoprecipitation of the BK channel from our samples. Note that both antibodies used in this study consistently recognized a band slightly lower than the band at 120 kD (BK); this band was always present in the HEK cell lanes, and we presume it to be nonspecific binding.

Fig. 5

RT-PCR demonstrates that the gBK insert is present in TyrAG825-treated cells as well as cells in the presence of endogenous neuregulin. A: D54 cells were incubated overnight in normal serum-containing medium (contains endogenous neuregulin), serum-free medium, or serum-free medium with 50 μM TyrAG825. mRNA was collected and RT-PCR performed. Primers were designed to target the specific gBK insert and the four common splicing sites located in the C-terminal tail of the protein shown in B. B: Schematic of the BK channel α-subunit detailing the relative location of each of the four splicing sites targeted with RT-PCR.

Fig. 6

Fura-2 imaging demonstrates that intracellular free Ca²⁺ is decreased in TyrAG825- and increased in Nrg-1β-treated cells. A: D54 cells treated with vehicle for 12 hr have mean resting free intra-cellular calcium concentrations of 89 ± 1.1 nM (n = 154). B: D54 cells treated overnight with TyrAG825 (50 μM) had a mean resting calcium concentration of 58 ± 0.8 nM (n = 169). C: D54 cells treated overnight with Nrg-1β (80 nM) had a mean intracellular free calcium concentration of 95 ± 0.9 nM (n = 218). D: Overlay of the bar graphs in A—C demonstrate the shift in mean intracellular calcium concentrations in the three different treatment conditions. *P < 0.05, **P < 0.001.

Fig. 7

Whole-cell recording parameters are unchanged in TyrAG825- and Nrg-β1-treated cells. A: Representative examples of whole-cell currents evoked from stepping the membrane from 0 to 100 mV from a holding potential of –40 mV under each treatment condition. B: I-V plot of peak current in each treatment condition. C: Plot of the mean current density (pA/pF) of control, Nrg-1β and TyrAG825-treated cells. D: Plot of the mean resting membrane potential (mV) of cells in each treatment condition. Note that the more hyperpolarized potentials are toward the top of the graph.

Fig. 8

The shift in BK activation threshold in treated cells is lost when [Ca²⁺]_i is controlled. A: Current responses at five different voltages from representative inside-out patch recordings of control, TyrAG825-treated (50 μM), and Nrg-1β-treated (80 nM) cells in the absence of extracellular Ca²⁺. B: Plot of the mean conductance (pS) from several cells under each condition as a function of voltage (mV). Data points are the mean ± SE, and n = the number of cells under each condition.