The Journal of Neurophysiology Vol. 79 No. 5 May 1998, pp. 2782-2793
Copyright ©1998 by the American Physiological Society

Electrophysiological Properties of Human Astrocytic Tumor Cells In Situ: Enigma of Spiking Glial Cells

Angélique Bordey and Harald Sontheimer

Department of Neurobiology, University of Alabama at Birmingham, Birmingham, Alabama 35924

	ABSTRACT

Abstract Introduction Methods Results Discussion References

Bordey, Angélique and Harald Sontheimer. Electrophysiological properties of human astrocytic tumor cells in situ: enigma of spiking glial cells. J. Neurophysiol. 79: 2782-2793, 1998. To better understand physiological changes that accompany the neoplastic transition of astrocytes to become astrocytoma cells, we studied biopsies of low-grade, pilocytic astrocytomas. This group of tumors is most prevalent in children and the tumor cells maintain most antigenic features typical of astrocytes. Astrocytoma cells were studied with the use of whole cell patch-clamp recordings in acute biopsy slices from 4-mo- to 14-yr-old pediatric patients. Recordings from 53 cells in six cases of low-grade astrocytomas were compared to either noncancerous peritumoral astrocytes or astrocytes obtained from other surgeries. Astrocytoma cells almost exclusively displayed slowly activating, sustained, tetraethylammonium (TEA)-sensitive outward potassium currents (delayed rectifying potassium currents; I_DR) and transient, tetrodotoxin (TTX)-sensitive sodium currents (I_Na). By contrast, comparison glial cells from peritumoral regions or other surgeries showed I_DR and I_Na, but in addition these cells also expressed transient "A"-type K⁺ currents and inwardly rectifying K⁺ currents (I_IR), both of which were absent in astrocytoma cells. I_IR constituted the predominant conductance in comparison astrocytes and was responsible for a high-resting K⁺ conductance in these cells. Voltage-activated Na⁺ currents were observed in 37 of 53 astrocytoma cells. Na⁺ current densities in astrocytoma cells, on average, were three- to fivefold larger than in comparison astrocytes. Astrocytoma cells expressing I_Na could be induced to generate slow action potential-like responses (spikes) by current injections. The threshold for generating such spikes was 34 mV (from a holding potential of 70 mV). The spike amplitude and time width were 52.5 mV and 12 ms, respectively. No spikes could be elicited in comparison astrocytes, although some of them expressed Na⁺ currents of similar size. Comparison of astrocytes to astrocytoma cells suggests that the apparent lack of I_IR, which leads to high-input resistance (>500 M), allows glioma cells to be sufficiently depolarized to generate Na⁺ spikes, whereas the high resting K⁺ conductance in astrocytes prevents their depolarization and thus generation of spikes. Consistent with this notion, Na⁺ spikes could be induced in spinal cord astrocytes in culture when I_IR was experimentally blocked by 10 µM Ba²⁺, suggesting that the absence of I_IR in astrocytoma cells is primarily responsible for the unusual spiking behavior seen in these glial tumor cells. It is unlikely that such glial spikes ever occur in vivo.

	INTRODUCTION

Abstract Introduction Methods Results Discussion References

Unlike most neurons, glial cells retain the ability to divide postnatally (Gensert and Goldman 1996; Korr 1986). Glial proliferation accompanies reactive gliosis, the brain's response to injury (Reier 1986) and the malignant transformation of glial cells. Glial-derived tumors are collectively termed gliomas and include astrocytic, oligodendroglial, ependymal, and glial-neuronal tumors (Kleihues et al. 1993a,b). The majority of gliomas fall into the astrocytic lineage (astrocytomas) and presumably derive from astrocytes. They still share some common features with astrocytes, including variable expression of glial fibrillary acidic protein (GFAP), S100, and growth factor receptors (Bigner et al. 1981; Collins 1995; Kleihues et al. 1993a,b). This is particularly true for low-grade astrocytomas, the most prevalent primary brain tumor in children. Factors that induce the transition from glia to glioma are not well understood. However, it is evident that this transition is accompanied by a number of genetic alterations, gene deletions, and gene amplifications (Collins 1995; Westermark et al. 1995) that are also typical of cancer cells outside the nervous system.

Attempts to understand mechanisms that underlie cell growth and differentiation have largely focused on studying cancers of the body. Several of these studies suggest that alterations in ion channel complement are associated with and possibly are essential to allow cell proliferation to occur (Sontheimer 1995). For example, in lymphocytes entry into the cell cycle is associated with the up-regulation of Ca²⁺-activated K⁺ channels (Cahalan et al. 1991; DeCoursey et al. 1984) and blockade of these channels retards cell proliferation. Pharmacological block of K⁺ channels also inhibits proliferation of brown fat cells (Pappone and Ortiz-Miranda 1993), melanoma cells (Nilius and Wohlrab 1992), human breast cancer cells (Woodfork et al. 1995), lung small cancer cells (Pancrazio et al. 1993), retinal glial cells (Puro et al. 1989), Schwann cells (Wilson and Chiu 1993), O2-A progenitor cells (Gallo et al. 1996), and astrocytes (Pappas et al. 1994). Mechanisms by which ion channel complement affects cell proliferation are largely unknown. Some recent studies have begun to shed light on interactions between tumor suppressor genes or oncogenes and ion channel function. For example, transfection of fibroblasts with ras21 or exposure of fibroblasts to epidermal growth factor or platelet-derived growth factor leads to the induction of a novel Ca²⁺-activated K⁺ channel that is essential for cell cycle progression in these cells (Huang and Rane 1994). In Drosophila, the tumor suppressor gene dlg interacts with "A-type" K⁺ channels leading to channel clustering (Tejedor et al. 1997).

Despite our expanding knowledge of ion channel expression in normal rat or mouse glial cells, little is known about properties of glial-derived tumor cells. The few studies that have investigated glial cell-derived tumors have largely focused on studying several well-established human cell lines. These transformed glial cells possess a plethora of ion channels (Brismar 1995; Matute et al. 1992). However, culturing cells and passing them on for generations may alter their phenotypes as has been demonstrated in other cells, including glial cells (Bigner et al. 1981; Kennedy et al. 1987; Noble et al. 1995). We recently reported that ion channel complement of astrocytes in situ differs from the channel complement of cultured astrocytes (Bordey and Sontheimer 1997). Our understanding of possible biophysical changes underlying malignant transformation of glial cells is further hampered by the fact that properties of "normal" human glial cells are largely unknown.

We set out to investigate properties of astrocyte-derived tumors as compared with noncancerous control astrocytes in patient biopsies. We focused on a comparison between the lowest grade astrocyte tumors, the pilocytic astrocytomas, because unlike higher grade gliomas, these astrocytoma cells maintain several biochemical and morphological features with astrocytes and may thus be most relevant in understanding changes that underlie the transition from glia to glioma. Our data suggest that astrocytoma cells lose expression of inwardly rectifying K⁺ channels, which are abundantly expressed in astrocytes. These channels are believed to be instrumental in aiding K⁺ buffering in the brain. They also endow astrocytes with a relatively negative and stable resting potential. Because of the absence of I_IR in conjunction with the prominent expression of I_Na in astrocytoma cells, these cells could be induced to fire slow, action potential-like responses not seen in normal astrocytes. The latter, however, could be induced to exhibit spikes when I_IR was pharmacologically inhibited suggesting that the lack of I_IR in astrocytoma cells is principally responsible for the unusual spiking behavior.

	METHODS

Abstract Introduction Methods Results Discussion References

Patient selection

Human brain specimens were obtained from six pediatric patients who underwent resections of primary central nervous system neoplasms. Clinically these cases did not present with seizures. Parallel recordings were performed in comparison tissue also obtained from three pediatric patients. These tissues were resected outside the tumor margins of a choroid plexus carcinoma and a hypothalamic pilocytic astrocytoma, respectively. One additional comparison biopsy was obtained from the cortex of a patient who underwent a hemispherectomy to surgically treat Rasmussen's encephalitis. Histopathological examination showed that none of these comparison tissues contained neoplastic cells. We refer to these tissues throughout as comparison tissue, acknowledging the fact that cells in these tissues may also have altered properties.

Slice preparation

Methods for preparing slices are described in detail in Ullrich et al. (1997). Human tissue was obtained during surgery and collected in the operating room where it was immediately placed in ice-cold (4°C) calcium-free artificial cerebrospinal fluid (ACSF-Ca-free) containing (in mM) 116 NaCl, 4.5 KCl, 0.8 MgCl₂, 26.2 NaHCO₃, 11.1 glucose, and 5 N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid (HEPES), oxygenated with 95% O₂-5% CO₂. Because of its fragile and mixed consistency, tissue was embedded in agar and then glued (cyanoacrylate glue) to the stage of a Vibratome. Slices of 150-300 µm were cut in cold oxygenated ACSF-Ca-free and transferred to a beaker filled with ACSF-Ca-free at room temperature. After a recovery period of 2 h in ACSF-Ca-free, slices were placed in a flow-through chamber continuously perfused with oxygenated ACSF containing 1.8 mM CaCl₂ at room temperature. The chamber was mounted on the stage of an upright microscope (Nikon Optiphot2) equipped with a ×40 (2-mm working distance) water immersion objective and Nomarski optics.

Whole cell recordings and data analysis

Whole cell patch-clamp recordings were obtained as previously described for hippocampal slices (Bordey and Sontheimer 1997; Edwards et al. 1989). Patch pipettes were pulled from thin-walled borosilicate glass (1.55 mm od; 1.2 mm id; TW150F-40, WPI) on a PP-83 puller (Narishige, Japan). Pipettes had resistances of 4-6 M when filled with the following solution (in mM): 145 KCl, 0.2 CaCl₂, 1.0 MgCl₂, 10 ethylene glycol-bis(-aminoethyl ether)-N,N,N',N'-tetraacetic acid (EGTA), and 10 HEPES (sodium salt), pH adjusted to 7.2 with tris(hydroxymethyl)aminomethane (Tris). To label cells for later morphological and antigenic identification, 0.1% Lucifer yellow (LY; dilithium salt) was added to the pipette solution. Voltage-clamp recordings were performed with an Axopatch-200A amplifier (Axon Instruments). Current signals were low-pass filtered at 5 kHz and digitized on-line at 25-100 kHz by using a Digidata 1200 digitizing board (Axon Instruments) interfaced with an IBM-compatible computer system. Data acquisition, storage, and analysis were done with the use of pClamp version 6 (Axon Instruments). For all measurements, capacitance and series resistance compensation (60-80%) were used to minimize voltage errors. Settings were determined by compensating the transients of a small (5 mV) 10-ms hyperpolarizing voltage step; the capacitance reading of the amplifier was used as value for the whole cell capacitance. After compensation, series resistances ranged from 6 to 10 M. Membrane resistance (R_m) was determined from the average response to 50 hyperpolarizing (10 mV) current pulses (20 ms). The resulting membrane current charging curves were fitted either to a single or double exponential.

Unless indicated otherwise, capacitive and leak conductances were subtracted on-line by a modified P/5 protocol. This P/5 protocol consisted of five hyperpolarized subpulses of ¹/₅ amplitude executed before each episode, summed and added to the current trace of interest. Subpulses holding potential was 80 mV. Peak currents were determined by using Clampfit (Axon Instruments) and statistical values (mean ± SE in the text, ±SD in the table; n = number of cells tested) were evaluated with a statistical graphing and curve-fitting program (Origin, MicroCal). Chemicals were purchased from Sigma unless otherwise noted.

Steady-state parameters for activation and inactivation of sodium current

To establish steady-state activation or inactivation curves, the peak current I was measured at each potential and the corresponding conductance G was calculated by using the following equation

(1)

where V is the membrane command potential and V_i is the predicted equilibrium (Nernst) potential for the ion (i) under consideration. At 20°C, calculated equilibrium potentials for potassium and sodium were 87.7 and 67.0 mV, respectively.

The measured peak amplitudes and the calculated peak conductance were then normalized and plotted as a function of the membrane potential during the test pulse. The resulting inactivation and activation curves were fit to the Boltzmann equation

(2)

where G_i(max) is the maximum ionic conductance at peak current, V_1/2 is the voltage where G_i is one-half of G_i(max), and k, the slope factor, determines the voltage-dependent relation of G_i.

	RESULTS

Abstract Introduction Methods Results Discussion References

Whole cell recordings were obtained from acute tissue slices of nine patient biopsies and included 53 glioma cells and 9 nonneoplastic astrocytes for comparison. Although comparison tissue was also derived from patients, these tissues did not contain neoplastic cells as confirmed by histopathological examinations. Hence, despite the possibility that cells in these tissues may also have altered properties, they represent the best possible comparison tissue for purposes of this study. All tissues were histopathologically characterized and classified with the use of criteria set forth by the World Health Organization (WHO) (Kleihues et al. 1993a,b). They included four cases of pilocytic astrocytomas (WHO grade I) and one desmoplastic and one giant cell astrocytoma, also considered to be WHO grade I tumors. These tumors were resected from children between 4-mo and 14-yr old (mean age 5 yr). Comparison tissue (3 cases) was cortical tissue associated with a choroid plexus carcinoma and a hypothalamic pilocytic astrocytoma and cortical tissue from a patient with Rasmussen's disease (3-yr-old patient).

Tumor tissue consisted of densely packed, round tumor cells (Fig. 1A; black arrows point out round cells) and their processes. Cell processes stained positive for GFAP (Fig. 1B, arrows), whereas most glioma cell bodies were only weakly GFAP positive (Fig. 1B) as this was previously reported (Bilzer et al. 1991; Bressler and Edwards 1997; Kharbanda et al. 1993). Tumor tissue could be readily distinguished from surrounding normal peritumoral tissue microscopically (Fig. 1C). The tumor tissue was fragile and showed thinner sections after cutting. Figure 2 shows two representative cells, one from a pilocytic astrocytoma cell from the hypothalamus and one from a nonneoplastic comparison astrocyte from peritumoral issue from a choroid plexus carcinoma. These cells were each filled with LY during whole cell recording to allow morphological and antigenic characterization. The comparison astrocyte was GFAP positive and its morphology resembled that of stellate astrocytes in rat hippocampus in situ (Bordey and Sontheimer 1997) with 2-3 major processes and many smaller processes of ~50-60 µm (Fig. 2A). The pilocytic astrocytoma cells had a round cell body and extended only 2-3 major processes (Fig. 2B). This cell, like many others, showed some LY coupling to adjacent cells. Typically, dye spread to less than five adjacent cells.

View larger version (95K): [in this window] [in a new window]   FIG. 1. Human astrocytoma biopsies. A: high power photograph (×400) of a live section of a pilocytic astrocytoma. Cells are densely packed and tend to have round cell bodies (arrows point to representative tumor cells). B: glial fibrillary acidic protein (GFAP) staining of the same slice as in A; primarily processes traversing the slice stain positive for GFAP (white arrows). Cell bodies are only weakly stained. C: low power photograph (×100) of same slice as in A shows tumor margins delineated by a dashed white line.

View larger version (40K): [in this window] [in a new window]   FIG. 2. Astrocytoma cells are electrophysiologically distinct from comparison astrocytes. Left panel: characteristic morphology of Lucifer yellow-filled cells; right panel: representative whole cell currents after stepping the cell membrane from 160 to + 80 mV from a holding potential of 80 mV of the corresponding recorded cell. (A: comparison nontumorous astrocyte; B: pilocytic astrocytoma cell; C: corresponding current-voltage relationships of the displayed traces from A and B.) Current amplitudes were determined 4 ms (, ) and 35 ms (, ) after the beginning of the recording and plotted as a function of applied potentials. I-V curve of the pilocytic astrocytoma cell (, ) was outwardly rectifying, whereas I-V curve of the comparison astrocyte (, ) was linear.

The recordings from these two cells (Fig. 2, A and B, right panels) are representative examples for each tissue group, and the traces show raw data without leak subtraction elicited by 50-ms voltage steps to potentials ranging from 160 to 60 mV (10-mV increments). These recordings show prominent expression of outward potassium currents in both glioma and comparison astrocytes. However, comparison cells always expressed large inward K⁺ currents in addition to outward K⁺ currents. Inward K⁺ currents were absent in glioma cells. This difference in rectification is more apparent in the corresponding current-voltage curves (I-V) plotted in Fig. 2C. Current amplitudes were determined 4 ms (, ) and 35 ms (, ) after the beginning of the recording and plotted as a function of applied potentials. I-V curves of astrocytoma cells (, ) were always outwardly rectifying, whereas I-V curves of comparison astrocytes (, ) were either inwardly rectifying or linear at potentials near the resting potential. We previously reported expression of outwardly rectifying glioma-associated chloride currents (GCC) that are activated by very positive voltages [e.g., >45 mV (Ullrich et al. 1997)]. To eliminate possible contamination of current traces with GCC, we restricted our analysis of peak currents to potentials <30 mV, where Cl currents were not activated. A more detailed dissection of inward and outward currents is discussed below and shown in Figs. 3-4.

View larger version (27K): [in this window] [in a new window]   FIG. 3. Astrocytoma cells express tetrodotoxin (TTX)-sensitive, voltage-dependent Na+ currents. A: whole cell inward currents in response to 8-ms step depolarizations starting at 70 mV, incremented by 10 mV from a holding potential of 110 mV. Currents before exposure to TTX (left), after 3-min exposure to TTX 100 nM (right). B: mean peak current amplitudes obtained in 14 cells from 5 pilocytic astrocytomas () and from traces shown in A () were plotted vs. corresponding membrane potential. Extrapolated reversal potential +66.0 mV. C: whole cell Na+ currents after a 200-ms inactivating prepulse potential. Prepulse potentials started at 160 mV and were incremented by 10-mV steps while the command potential was kept at 10 mV. D: activation () and inactivation curves (). Data from Fig. 2A plotted as activation curve for peak current. ···, best fit according to Eq. 2 withV1/2 = 20.1 mV, k = 8.2 mV1 per e-fold. Fraction of peak current measured from 18 cells is plotted vs. inactivation voltage.- - -, best fit to a Boltzmann distribution with values of -59.5 mV for V1/2 and 9.6 mV per e-fold for k.

View larger version (29K): [in this window] [in a new window]   FIG. 4. Characterization of delayed rectifying potassium IDR in astrocytoma cells. A: voltage command pulses from 70 to 80 mV from a holding potential of 70 mV evoked sustained outward currents. B: mean current amplitudes obtained in 8 cells from 2 pilocytic astrocytoma () and in 6 cells from comparison tissue () were plotted vs. corresponding membrane potential. I-V relationship suggested an activation threshold of ~ 50 mV in astrocytoma cells and 30 mV in comparison cells. Currents activated at 50 mV (Fig. 3B). C: bath application of 10 mM tetraethylammonium (TEA) reduced IDR by 32% (n = 4). D: tail current analysis was performed by stepping the cell membrane from 160 to 0 mV after a prepulse to 0 mV. Tail current amplitudes were measured 500 µs after stepping the membrane potential to 2nd varying step potentials; then basal current ("leakage" current) was subtracted from measured tail currents and resulting "tail" currents plotted against applied potential in E. E: tail current/voltage curve yielded a reversal potential of 86.6 mV (Nernst EK = 87.7 mV).

Data from all recorded cells are summarized in Table 1, which in addition to densities of current and conductance of the different current subtypes also gives mean values (±SD) of resting membrane potential (V_r), membrane capacitance (C_m), and resting membrane resistance (R_m, measured at 80 mV). For consistent comparison, all biophysical parameters (e.g., C_m, V_r, and R_m) were determined in the first 3 min of whole cell recordings. Conductance densities were used as measures that allow us to compare the relative expression of current components independent of cell size. Conductance of outward currents was determined at 30 mV and conductance of I_IR at 150 mV. Kinetic and pharmacological analysis of the outward and inward currents in tumor cells were performed in cells having series resistances <10 M and with the use of leak-subtraction protocols.

Astrocytoma cells had significantly more depolarized membrane potentials than comparison astrocytes whose potential was similar to that reported previously for rat astrocytes (Bordey and Sontheimer 1997) (P < 0.001). However, mean membrane capacitance values were similar, suggesting that cells were of approximately equal size. This is consistent with our morphological observations (Fig. 2).

Sodium current, I_Na

Transient, fast, inward currents were observed in 37 of 53 (70%) glioma cells. These currents were almost completely blocked by 100 nM TTX (Fig. 3A; n = 10 cells, 3 cases). TTX-block was partially reversible and did not significantly alter the resting potential of the cell. Currents were activated with voltage steps more positive than 50 mV that peaked within <3 ms and inactivated rapidly (Fig. 3A, left panel). Current amplitudes were voltage-dependent, peaked at ~0 mV, and decreased with more depolarized voltage steps. I-V curves were generated either in the presence of TEA (10 mM) to block K⁺ currents (Fig. 3B, ) or used the TTX-sensitive current component after subtraction (Fig. 3B, ). This analysis included 14 cells from five cases (every biopsy except giant cell astrocytoma) and mean values were plotted in Fig. 3B. The extrapolated current reversal potential was 64.7 ± 1.7 (SE) mV (n = 14), which is in good agreement with the predicted equilibrium potential for Na⁺ ions (E_Na = +68 mV) under the imposed ionic gradients suggesting that these fast, transient, inward currents were carried by Na⁺ ions. Steady-state activation and inactivation curves (Fig. 3D) were constructed according to Eqs. 1 and 2 (see METHODS) and again included 14 cells from the same five cases. Boltzmann fits of the activation curves gave values of 20.0 ± 2.5 mV for V_1/2 and 6.9 ± 0.7 mV¹ for k (n = 7). Sodium currents demonstrated typical steady-state inactivation (Fig. 3, C and D). The Boltzmann fits of the inactivation curves gave an e-fold change per 8.0 ± 0.4 mV¹ (n = 18) and a mean V_1/2 of 59.2 ± 1.9 mV (n = 18). Parameters for the activation and inactivation curves in comparison astrocytes were similar with values for V_1/2 of 21.6 and 57.4 mV and 8.7 and 7.5 mV¹ for k, respectively (data not shown).

To compare the relative Na⁺ channel expression in astrocytoma cells to comparison astrocytes, we determined sodium conductance density, G_Na (sodium conductance divided by the membrane capacitance), for each cell and pooled data from each patient. Astrocytoma cells in the four pilocytic and the desmoplastic tumors consistently expressed conductance densities that were approximately three- to fivefold higher than in comparison cells (Table 1).

Outward potassium currents

To study outward currents, depolarizing voltage steps were applied from a holding potential of 70 mV. This protocol activated sodium currents followed by sustained outward currents. The outward currents recorded in 38 of 53 (72%) tumor cells showed delayed and slow inactivation (Fig. 4A). The I-V relationship of outward currents derived from mean data of eight cells from two pilocytic astrocytomas suggested an activation threshold of approximately 50 mV (Fig. 4B) as compared with approximately 30 mV for two comparison astrocytes. The voltage-dependence of the sustained current showed kinetics reminiscent of delayed rectifying potassium currents (I_DR) described in numerous other preparations (Rudy 1988). Bath application of 10 mM TEA reduced I_DR by 32% (n = 4; Fig. 4C). Tail current analysis yielded a reversal potential of 85 mV closed to the predicted Nernst reversal potential for K⁺ (E_K = 88 mV; Fig. 4, D and E).

If currents were activated after applying a conditioning 200-ms prepulse to 110 mV (a protocol used to activate transient A-type K⁺ currents), no additional transient outward potassium current could be activated in 89% of all studied cells (data not shown). In the remaining six pilocytic astrocytoma cells, which were all from one tumor case (posterior fossa), transient A-currents with conductance densities of 391.6 ± 148.7 pS/pF could be identified and example recordings are shown in Fig. 5. These currents could be isolated by activation from a prepulse potential of 110 mV (Fig. 5A) and subsequent subtraction of currents activated from a prepulse potential of 70 mV (Fig. 5B) yielding a transient current component that activates at potentials greater than 60 mV (Fig. 5C). This is more readily visible if peak currents are plotted as a function of applied potential (Fig. 5D). Steady-state inactivation of the transient current was analyzed by varying the prepulse potential from 180 to 20 mV. The subtracted transient current (from protocols with and without prepulse; Fig. 5E) was plotted as a function of applied prepulse potential in Fig. 5D () for five tumor cells and showed a V_1/2 (for inactivation) of 84.4 ± 1.7 (n = 5). This value was more hyperpolarized than a V_1/2 of 60.3 mV in a comparison astrocyte (data not shown). Moreover, at the resting membrane potential of pilocytic astrocytoma cells (43.4 ± 4.7; n = 6), this transient outward potassium current displayed no overlap between steady-state activation and inactivation curves (Fig. 5D), suggesting that currents were not available for activation at the cells resting potential.

View larger version (40K): [in this window] [in a new window]   FIG. 5. Transient current recordings in pilocytic astrocytoma cells. Transient currents were isolated by sequentially activating currents from a prepulse potential of 110 mV (A) and subsequent subtraction of currents activated from a prepulse potential of 70 mV (B), yielding a transient current component that activates at potentials > 60 mV (C). Mean values of peak currents were plotted as a function of applied potential (D) for 5 cells. Steady-state inactivation of the transient current was analyzed by varying the prepulse potential from 180 to 20 mV (E). Mean values of the subtracted transient currents from 5 cells were plotted as a function of applied prepulse potential in Fig. 5D ().

Whole cell recordings were also obtained in which the membrane was stepped from a prepulse potential of 0 mV to more negative potentials ranging from 180 to 0 mV (data not shown). This protocol was applied to facilitate isolation of I_IR (Ransom and Sontheimer 1995). This type of current was found only in 2 of 11 cells of one pilocytic astrocytoma and was absent in 96% of all studied cells (see Table for conductance values). The inward current showed a marked current inactivation at potentials more negative than 130 mV. In addition, the I-V curve showed inward rectification typical of cesium- and barium-sensitive inwardly rectifying K⁺ channels.

Spikes induced by current injections in astrocytoma cells

Current injection into some tumor cells induced transient overshooting responses reminiscent of slow action potentials (Fig. 5, A and B, at different timescale). Only single spikes could be induced with each current injection; thus we consider these spikes not regenerative. In four cells held at 70 mV, an average of 10 action potentials gave a mean spike amplitude of 52.5 ± 3.5 mV. The threshold was 33.9 ± 1.0 mV. The mean rise time (beginning of the pulse to the peak spike value) was 17.1 ± 2.8 ms and the duration of the spike was 11.8 ± 0.6 ms. Because generation of action potentials depends on the ratio of inward to outward current, we calculated conductance densities (conductance obtained by Eq. 1, METHODS, divided by the membrane capacitance) mediated by potassium channels (G_K, corresponding to the total potassium density of conductance in these cells) and sodium channels (G_Na) to obtain the ratio G_K:G_Na. The ratio G_K:G_Na in astrocytoma cells was 2.7, a value that is significantly lower (P < 0.001) than a ratio of 11.8 ± 5.6 (n = 8) in nontumor astrocytes (Table 1).

Experimental spike induction in spinal cord astrocytes in vitro

Spinal cord astrocytes were shown to possess significantNa⁺ channel densities in vitro, allowing action potential-likeresponse induced by current injection (Sontheimer and Waxman 1992). This appears largely because of a lack of inhibition by neurons whose presence suppresses Na⁺ channel expression in these cells (Thio et al. 1993) and is probably an artifact of cell culture. However, these cells present a good model system to study whether our hypothesis, namely that the relative expression of I_IR and I_Na determines whether spikes can be induced, is correct. At 7-10 days in vitro, these cells express prominent inwardly rectifying potassium currents in addition to I_Na (Fig. 7A, left panel) (see also Ransom and Sontheimer 1995). The potassium current was blocked by 10 µM barium (Fig. 7A, right panel) as previously reported (Ransom and Sontheimer 1995). Current injection under current clamp could not elicit a spiking response from these cells (Fig. 7B). Blockade of I_IR by 10 µM Ba²⁺ transforms these "passive" cells into spiking cells (Fig. 7B, right panel). Figure 7B illustrated the progressive transformation due to the Ba²⁺ application in current-clamp recordings during the wash in of the drug. A current injection of 200 pA was applied every 5 s and the respective potential values (resting potential and potential in response to the current injection) were plotted against time (Fig. 7B, right panel). These data clearly suggest that expression of inwardly rectifying K⁺ channels prevents the spike generation by current injection.

View larger version (12K): [in this window] [in a new window]   FIG. 6. Action potential in astrocytoma cell. A: tumor cell in a pilocytic astrocytoma was recorded in current-clamp mode. Action potential was elicited by applying 100-pA current injection pulse. B: current-clamp traces in another cell from same tumor after shorter depolarizing current pulse. Spike amplitude was 52 mV, threshold was ~ 33 mV, rise-time (beginning of pulse to peak value) was 15.1 ms, duration of spike from threshold was 11.3 ms, and corresponding ratio for the same recorded cell was 1.1.

View larger version (28K): [in this window] [in a new window]   FIG. 7. Glial spikes can be induced in normal astrocytes after pharmacological blockade of IIR. A, top panel: under voltage-clamp configuration currents recorded in control (left) and in the presence of bath application of 10 µM Ba2+ (right). Holding potential was 80 mV. Currents were activated after depolarizing cell to 0 mV for 100 ms, before application of hyperpolarizing voltage step from 180 to 10 mV for 40 ms. Bottom panel: under current-clamp configuration, membrane potentials recorded from the same cell as in top panel in control (left) and in presence of 10 µM Ba2+ (right). Ba2+ blocked the IIR and allowed recorded cell to spike in response to a current injection (100-pA step). B: under current clamp, a 200-pA current was injected every 5 s from a membrane potential of 70 mV (resting potential of the recorded cell). Ba2+ was applied at trace number 6, depolarized the cell, and turned its square passive response into a spiking response (left). Right panel: resting potential of the cell () and peak amplitude of the current-injected responses (; measured between 10 and 20 ms) were measured every 5 s and plotted against time of recording. This graph illustrates the depolarizing and spiking-induced effects of Ba2+ onto the cell.

	DISCUSSION

Abstract Introduction Methods Results Discussion References

Our results suggest that marked differences exist between astrocytoma cells and comparison astrocytes, most strikingly the loss of I_IR and A-type outward potassium currents (I_A) in transformed glial cells that appeared to exclusively express delayed rectifier potassium channels. In addition, astrocytoma cells showed an up-regulation of sodium currents and were able to generate current-induced spikes that were absent in comparison tissue. Blocking I_IR in normal rat glial cells in culture allowed these cells to generate current-induced spikes, suggesting that expression of I_IR prevents depolarization and spike generation in normal astrocytes.

Delayed rectifier potassium channels

Astrocytoma cells and comparison astrocytes consistently expressed I_DR. Indeed, these channels were ubiquitously expressed in nonexcitable cells (Rudy 1988) and were implicated in the control of proliferation of various cell types (Gallo et al. 1996; Nilius and Wohlrab 1992; Pancrazio et al. 1993; Pappas et al. 1994; Pappone and Ortiz-Miranda 1993; Puro et al. 1989; Wilson and Chiu 1993; Woodfork et al. 1995). We recently reported an increase in the conductance of delayed rectifying K⁺ channels in gliotic proliferative astrocytes in an in vitro model of spinal cord injury (MacFarlane and Sontheimer 1997). A striking result from our study is the hyperpolarized shift in the activation threshold for I_DR to values close to 50 mV instead of 20 or 30 mV in other cell types, including rat hippocampal astrocytes, in situ (Bordey and Sontheimer 1997). It is possible that this change is due to some intracellular modulation affecting delayed rectifying K⁺ channels (Chung and Schlichter 1997; Cole et al. 1996; Koh et al. 1996; Perozo and Bezanilla 1990). I_DR were also present in comparison human cells in our study and in normal rat hippocampal astrocytes in situ.

Inwardly rectifying potassium channels

Our observation that astrocytoma cells lack I_IR is in contrast with a previous report that studied glioma cell lines (Brismar 1995). The reason for this discrepancy is unclear. It is possible that cell lines (in vitro) display features that differ from cells in situ or in vivo. Moreover, we studied low-grade gliomas, whereas glioma cell lines used by others were derived from high-grade gliomas (glioblastoma). Studies of rat C6 astrocytoma cells, believed to be more astrocyte-like and highly GFAP positive, also lack inwardly rectifying K⁺ channels. The absence of I_IR in pilocytic astrocytoma is consistent with other proliferating cell types that also lack I_IR, for example, leukemia cells (Missiaen et al. 1997), neuroblastoma cells (Arcangeli et al. 1995), early mouse embryos (Day et al. 1993), and HeLa cells (Takahashi et al. 1994). In addition, injury induced proliferation of spinal cord astrocytes in situ correlates with a down-regulation of I_IR (MacFarlane and Sontheimer 1997). In hippocampal slices I_IR channel expression is developmentally regulated and its expression appears to characterize primarily differentiated nonproliferating astrocytes (Bordey and Sontheimer 1997).

The lack of I_IR currents in astrocytoma cells is due either to a modulation of the channel protein or to the down-regulation of its expression. Our detection methods do not allow to discriminate between these possibilities and, for purposes of our study, this difference is not germane. However, the loss of functional I_IR currents may have important physiological consequences. This channel class is known to participate in the potassium homeostasis in the brain (Amedee et al. 1997; Newman and Reichenbach 1996) and one would predict that in the absence of this channel, potassium buffering would be impaired in tumors. In addition, these potassium channels, in conjunction with chloride channels, are involved in cell-volume regulation and, consequently, volume regulation may also be compromised in tumor cells. Interestingly, glioma cells had a more rounded-up morphology and appeared swollen (phase image). Inwardly rectifying K⁺ channels are thought to establish the hyperpolarized resting potential typical of most astrocytes. Its absence in glioma thus explains the depolarized resting potential of tumor cells. Depolarized membrane potentials are considered to be a characteristic of tumor cells (Cone 1970).

Transient outward potassium channels

In glioma cells, transient A-type K⁺ channels are largely absent. Like the loss of I_IR, the absence of transient outward potassium currents appears to be an early feature that accompanies the transformation of a normal astrocyte to become a tumor cell. The role of these channels in glial cells is not known. In neurons they modulate frequency of action potential discharge (Rudy 1988). Interestingly, A-currents were recently implicated in the extension of neurites (Wu and Barish 1996). Glioma cells displayed fewer processes than normal glial cells and show numerous alterations in extracellular matrix (McKeever et al. 1997) and membrane architecture (Kharbanda et al. 1993; Nagano et al. 1993). In analogy to the role of A-currents in neurite extension, a loss of A-currents in glioma cells might be a result of perturbations in extracellular matrix.

Na⁺ channels

Our data suggest a two- to fivefold up-regulation of Na⁺ channel density in astrocytoma cells without any significant biophysical or pharmacological alterations. This could have been due to a functional modulation of sodium channels or an increase in channel expression. Considering the second possibility and assuming a single-channel conductance of 20 pS and an open probability of 0.25 (Barres et al. 1989), mean sodium channel densities in pilocytic astrocytoma and desmoplastic astrocytoma had mean values of 0.8 ± 0.5 channels/µm² (ranging from 0.1 to 1.75) and 1.0 ± 0.5 channels/µm², respectively. These values are significantly higher than values of 0.16 ± 0.18 channels/µm² (0.02 to 0.4) in comparison human glial cells. This increase in I_Na in conjunction with the suppression of I_IR accounted for the unusual spiking behavior of astrocytoma cells after current injection.

Biophysical "spikes" are a consequence of altered channel complement

Our data suggest that the lack of I_IR allows inducement of spikes by current injection. Our experiments with spinal cord astrocytes in vitro corroborate this. These cells have the adequate ion channel complement, both Na⁺ channels and inwardly rectifying K⁺ channels. Blocking I_IR by an application of low concentrations of barium turned "passive" glial cells into spiking glial cells, suggesting that the activity of inwardly rectifying K⁺ channels prevented such spikes in Na⁺ channel-bearing glial cells.

Functional consequences and implications

The role of these slow glial spikes is an enigma. We believe that they probably never occur in vivo. However, others (Labrakakis et al. 1997) suggested that the genesis of tumor (glioblastoma multiform)-associated epileptic seizure may be due to spiking glioma cells. However, at the resting potential of glioma cells, Na⁺ channels are essentially fully inactivated, curtailing activation. The cases studied here did not present with seizure symptoms.

Much has been speculated about the potential role of glial Na⁺ channels (Ritchie 1992). These speculations include roles in fueling the Na⁺/K⁺-ATPase (Sontheimer et al. 1994), ephaptic coupling to neurons (Chao et al. 1994), or channel biosynthesis and transfer to neurons (Bevan et al. 1987). We do not want to expand on these suggestions. Instead, we are compelled to speculate that the most intriguing finding is the loss of I_IR amplifying the relative contribution of Na⁺ currents to the whole cell conductance. The latter, we believe, is an important aspect in cell differentiation. We speculate that inwardly rectifying K⁺ channels are expressed in differentiated, nondividing astrocytes, whereas they are down-regulated in dividing cells for unknown reasons. Indeed, this was observed in numerous cell types, including astrocytes, and our recent study showed the rapid loss of I_IR in conjunction with gliotic proliferation (MacFarlane and Sontheimer 1997).

Thus the observed changes are consistent with enhanced proliferation of transformed glial cells. Glial spikes can be experimentally induced as a consequence of these changes in channel complement. However, they are a biophysical exercise with probably little physiological significance.

	ACKNOWLEDGEMENTS

  This work was supported by National Institutes of Health Grants RO1-NS-31234, NS-36692, and P50-HD-32901.

	FOOTNOTES

  Address for reprint requests: H. Sontheimer, Dept. of Neurobiology, The University of Alabama at Birmingham, 1719 6th Ave. South, Cintan International Research Center, Room 545, Birmingham, AL 35294.

  Received 20 October 1997; accepted in final form 12 January 1998.

	REFERENCES

Abstract Introduction Methods Results Discussion References

• AMEDEE, T., ROBERT, A., COLES, J. A. Potassium homeostasis and glial energy metabolism. Glia 21: 46-55, 1997.[Medline]
• ARCANGELI, A., BIANCHI, L., BECCHETTI, A., FARAVELLI, L., CORONNELLO, M., MINI, E., OLIVOTTO, M., WANKE, E. A novel inward-rectifying K⁺ current with a cell-cycle dependence governs the resting potential of mammalian neuroblastoma cells. J. Physiol. (Lond.) 489: 455-471, 1995.[Abstract/Free Full Text]
• BARRES, B. A., CHUN, L. Y., COREY, D. P. Glial and neuronal forms of the voltage-dependent sodium channel: characteristics and cell-type distribution. Neuron 2: 1375-1388, 1989.[Medline]
• BEVAN, S., LINDSAY, R. M., PERKINS, M. N., RAFF, M. C. Voltage-gated ionic channels in rat cultured astrocytes, reactive astrocytes and an astrocyte-oligodendrocyte progenitor cell. J. Physiol. Paris 82: 327-335, 1987.[Medline]
• BIGNER, D. D., BIGNER, S. H., PEGRAM, C.N.B., MAHALEY, S., RUOSLAHTI, E., HERSCHMAN, H., ENG, L. F., WIKSTRAND, C. J. Heterogeneity of genotypic and phenotypic characteristics of fifteen permanent cell lines derived from human gliomas. J. Neuropathol. Exp. Neurol. 40: 210-229, 1981.
• BILZER, T., STAVROU, D., WECHSLER, W., WOHLER, B., KEIDITSCH, E. Antigen variation in a human glioblastoma: from the primary tumor to the second recurrence, permanent cell line and xenotransplantation tumors. Anticancer Res. 11: 547-554, 1991.[Medline]
• BORDEY, A., SONTHEIMER, H. Postnatal development of ionic currents in rat hippocampal astrocytes in situ. J. Neurophysiol. 78: 461-477, 1997.[Abstract/Free Full Text]
• BRESSLER, J. P., EDWARDS, N. A. Glial shape and cytoskeletal protein synthesis. Neurochem. Res. 17: 173-177, 1997.
• BRISMAR, T. Physiology of transformed glial cells. Glia 15: 231-243, 1995.[Medline]
• CAHALAN, M. D., CHANDY, K. G., GRISSMER, S. Potassium channels in development, activation, and disease in T lymphocytes. Curr. Top. Membr. Transp. 39: 357-394, 1991.
• CHAO, T. I., SKACHKOV, S. N., EBERHARDT, W., REICHENBACH, A. Na⁺ channels of Müller (glial) cells isolated from retinae of various mammalian species including man. Glia 10: 173-185, 1994.[Medline]
• CHUNG, I., SCHLICHTER, L. C. Native Kv1.3 channels are upregulated by protein kinase C. J. Membr. Biol. 156: 73-85, 1997.[Medline]
• COLE, W. C., CLEMENT-CHOMIENNE, O., AIELLO, E. A. Regulation of 4-aminopyridine-sensitive delayed rectifier K⁺ channels in vascular smooth muscle by phosphorylation. Biochem. Cell Biol. 74: 439-447, 1996.[Medline]
• COLLINS, V. P. Gene amplification in human gliomas. Glia 15: 289-296, 1995.[Medline]
• CONE, C. J. Variation of the transmembrane potential level as a basic mechanism of mitosis control. Oncology 24: 438-470, 1970.[Medline]
• DAY, M. L., PICKERING, S. J., JOHNSON, M. H., COOK, D. I. Cell-cycle control of a large-conductance K⁺ channel in mouse early embryos. Nature 307: 560-562, 1993.
• DECOURSEY, T. E., CHANDY, G., GUPTA, S., CAHALAN, M. D. Voltage-gated K⁺ channels in human T lymphocytes: a role in mitogenesis? Nature 307: 465-468, 1984.[Medline]
• EDWARDS, F. A., KONNERTH, A., SAKMANN, B., TAKAHASHI, T. A thin slice preparation for patch clamp recordings from neurones of the mammalian central nervous system. Pflügers Arch. 414: 600-612, 1989.[Medline]
• GALLO, V, ZHOU, J., MCBAIN, C., WRIGHT, P., KNUTSON, P., ARMSTRONG, R. Oligodendrocyte progenitor cell proliferation and lineage progression are regulated by glutamate receptor-mediated K⁺ channel block. J. Neurosci. 16: 2659-2670, 1996.[Abstract/Free Full Text]
• GENSERT, J. M., GOLDMAN, J. E. In vivo characterization of endogenous proliferating cells in adult rat subcortical white matter. Glia 17: 39-51, 1996.[Medline]
• HUANG, Y., RANE, S. G. Potassium channel induction by the Ras/Raf signal transduction cascade. J. Biol. Chem. 269: 1183-1189, 1994.[Abstract/Free Full Text]
• KENNEDY, P., WATKINS, B., THOMAS, D., NOBLE, M. Antigenic expression by cells derived from human gliomas does not correlate with morphological classification. J. Neuropathol. Appl. Neurobiol. 13: 327-347, 1987.
• KHARBANDA, K., SARKAR, C., DINDA, A. K., KARAK, A. K., MATHUR, M., ROY, S. Morphological appearance, growth kinetics and glial fibrillary acidic protein (GFAP) expression in primary in vitro explant culture of astrocytic neoplasms. Acta Oncol. 32: 301-306, 1993.[Medline]
• KLEIHUES, P. K., BURGER, P. C., SCHEITHAUER, B. W. In: Histological Typing of Tumours of the Central Nervous System. New York: Springer-Verlag, 1993a, p. 5-83.
• KLEIHUES, P. K., BURGER, P. C., SCHEITHAUER, B. W. The new WHO classification of brain tumours. Brain Pathol. 3: 255-268, 1993b.[Medline]
• KOH, S. D., SANDERS, K. M., CARL, A. Regulation of smooth muscle delayed rectifier K⁺ channels by protein kinase A. Pflügers Arch. 432: 401-412, 1996.[Medline]
• KORR, H. Proliferation and cell cycle parameters of astrocytes. In: Astrocytes, edited by S. Fedoroff, and A. Vernakakis Orlando, FL: Academic, 1986, p. 77-127.
• LABRAKAKIS, C., PATT, S., WEYDT, P., CERVOS-NAVARRO, J., MEYER, R., KETTENMANN, H. Action potential-generating cells in human glioblastomas. J. Neuropathol. Exp. Neurol. 56: 243-254, 1997.[Medline]
• MACFARLANE, S., SONTHEIMER, H. Electrophysiological changes that accompany reactive gliosis in vitro. J. Neurosci. 17: 316-329, 1997.
• MATUTE, C., ARELLANO, R. O., CONDE-GUERRI, B., MILEDI, R. mRNA coding for neurotransmitter receptors in a human astrocytoma. Proc. Natl. Acad. Sci. USA 89: 3399-3403, 1992.[Abstract/Free Full Text]
• MCKEEVER, P. E., VARANI, J., PAPADOPOULOS, S. M., WANG, M., MCCOY, J. P. Products of cells from gliomas: IX. Evidence that two fundamentally different mechanisms change extracellular matrix expression by gliomas. J. Neurooncol. 24: 267-280, 1997.
• MISSIAEN, L., DE SMEDT, H., PARYS, J., SIENAERT, J., SIPMA, H., VANLINGEN, S., CASTEELS, R. Slow kinetics of inositol 1,4,5-trisphosphate-induced Ca²⁺ release: is the release "quantal" or "non-quantal"? Biochem. J. 323: 123-130, 1997.
• NAGANO, N., SASAKI, H., MASARU, A., HIRAKAWA, K. Invasion of experimental rat brain tumor: early morphological changes following microinjection of C6 glioma cells. Acta Neuropathol. (Berl.) 86: 117-125, 1993.[Medline]
• NEWMAN, E., REICHENBACH, A. The Müller cell: a functional element of the retina. Trends Neurosci. 19: 307-312, 1996.[Medline]
• NILIUS, B., WOHLRAB, W. Potassium channels and regulation of proliferation of human melanoma cells. J. Physiol. (Lond.) 445: 537-548, 1992.[Abstract/Free Full Text]
• NOBLE, M., GUTOWSKI, N., BEVAN, K., ENGEL, U., LINSKEY, M., URENJAK, J., BHAKOO, K., WILLIAMS, S. From rodent glial precursor cell to human glial neoplasia in the oligofendrocyte-type-2 astrocyte lineage. Glia 15: 222-230, 1995.[Medline]
• PANCRAZIO, J. J., TABBARA, I. A., KIM, Y. I. Voltage-activated K⁺ conductance and cell proliferation in small-cell lung cancer. Anticancer Res. 13: 1231-1234, 1993.[Medline]
• PAPPAS, C. A., ULLRICH, N., SONTHEIMER, H. Reduction of glia proliferation by K⁺ channel blockers is mediated by changes in pH_i. Neuroreport 6: 193-196, 1994.[Medline]
• PAPPONE, P. A., ORTIZ-MIRANDA, S. I. Blockers of voltage-gated K channels inhibit proliferation of cultured brown fat cells. Am. J. Physiol. 264: C1014-C1019, 1993.[Abstract/Free Full Text]
• PEROZO, E., BEZANILLA, F. Phosphorylation affects voltage gating of the delayed rectifier K⁺ channel by electrostatic interactions. Neuron 5: 685-690, 1990.[Medline]
• PURO, D. G., ROBERGE, F., CHAN, C. C. Retinal glial cell proliferation and ion channels: a possible link. Invest. Ophthalmol. Vis. Sci. 30: 521-529, 1989.[Abstract/Free Full Text]
• RANSOM, C. B., SONTHEIMER, H. Biophysical and pharmacological characterization of inwardly rectifying K⁺ currents in rat spinal cord astrocytes. J. Neurophysiol. 73: 333-345, 1995.[Abstract/Free Full Text]
• REIER, P. J. Gliosis following CNS injury: the anatomy of astrocytic scars and their influence on axonal elongation. In: Astrocytes, edited by S. Fedoroff, and A. Vernakakis Orlando, FL: Academic, 1986, p. 263-324.
• RITCHIE, J. M. Voltage-gated ion channels in Schwann cells and glia. Trends Neurosci. 15: 345-351, 1992.[Medline]
• RUDY, B. Diversity and ubiquity of K channels. Neuroscience 25: 729-750, 1988.[Medline]
• SONTHEIMER, H. Ion channels in inexcitable cells. Neuroscientist 1: 64-67, 1995. [Abstract/Free Full Text]
• SONTHEIMER, H., FERNANDEZ-MARQUES, E., ULLRICH, N., PAPPAS, C. A., WAXMAN, S. G. Glial Na⁺ channels are required for maintenance of Na⁺/K⁺-ATPase activity. J. Neurosci. 14: 2464-2475, 1994.[Abstract]
• SONTHEIMER, H., WAXMAN, S. G. Ion channels in spinal cord astrocytes in vitro. II. Biophysical and pharmacological analysis of two Na⁺ current types. J. Neurophysiol. 68: 1001-1011, 1992.[Abstract/Free Full Text]
• TAKAHASHI, A., YAMAGUCHI, H., MIYAMOTO, H. Change in density of K⁺ current of HeLa cells during the cell cycle. Jpn. J. Physiol. 44, Suppl.: S321-S324, 1994.
• TEJEDOR, F. J., BOKHARI, A., ROGERO, O., GORCZYCA, M., ZHANG, J., KIM, E., SHENG, M., BUDNIK, V. Essential role for dlg in synaptic clustering of Shaker K⁺ channels in vivo. J. Neurosci. 17: 152-159, 1997.[Abstract/Free Full Text]
• THIO, C. L., WAXMAN, S. G., SONTHEIMER, H. Ion channels in spinal cord astrocytes in vitro. II. Modulation of channel expression by coculture with neurons and neuron-conditioned medium. J. Neurophysiol. 69: 819-831, 1993.[Abstract/Free Full Text]
• ULLRICH, N., BORDEY, A., SONTHEIMER, H. Expression of voltage-activated chloride currents in acute slices of human gliomas. Neuroscience 83: 1161-1173, 1997.
• WESTERMARK, B., HELDIN, C. H., NISTER, M. Platelet-derived growth factor in human glioma. Glia 15: 257-263, 1995.[Medline]
• WILSON, G. F., CHIU, S. Y. Mitogenic factors regulate ion channels in Schwann cells cultured from newborn rat sciatic nerve. J. Physiol. (Lond.) 470: 501-520, 1993.[Abstract/Free Full Text]
• WOODFORK, K. A., WONDERLIN, W. F., PETERSON, V. A., STROBL, J. S. Inhibition of ATP-sensitive potassium channels causes reversible cell-cycle arrest of human breast cancer cells in tissue culture. J. Cell. Physiol. 162: 163-171, 1995.[Medline]
• WU, R., BARISH, M. E. A-current development in hippocampal neurons: links to kinase activity and membrane insertion. Soc. Neurosci. Abstr. 22: 1250, 1996.

0022-3077/98 $5.00 Copyright ©1998 The American Physiological Society