Articles in PresS. Am J Physiol Cell Physiol (July 14, 2004). 10.1152/ajpcell.00077.2004 Na+ currents and PcTX1

1

Cation Selectivity and Inhibition of Malignant Glioma Na+ channels by Psalmotoxin1

James K. Bubien1 , Hong-Long Ji1 , G. Yancey Gillespie 2 , Catherine M. Fuller1 , James M. Markert 2 , Timothy B. Mapstone 3 , and Dale J. Benos 1 1

Department of Physiology and Biophysics, University of Alabama at Birmingham, Birmingham, AL 35294 2 Department of Surgery, Division of Neurosurgery, University of Alabama at Birmingham, Birmingham, AL 35294 3 Department of Neurosurgery, Emory University, Atlanta, Georgia 30322

Corresponding Author: James K. Bubien Ph.D. Department of Physiology and Biophysics University of Alabama at Birmingham 1918 University Blvd., MCLM 726 Birmingham, AL 35294 Tel. (205) 934-6214 Fax. (205) 934-1445 Email: [email protected]

Running Title: Na+ currents and PcTX1

Copyright © 2004 by the American Physiological Society.

Na+ currents and PcTX1

2

ABSTRACT Psalmotoxin 1 (a component of the venom of a West Indies tarantula) is a 40 amino acid peptide that inhibits cation currents mediated by acid-sensing ion channels (ASIC). In this study we performed electrophysiological experiments to test the hypothesis that Psalmotoxin 1 (PcTX1) inhibits Na+ currents in high- grade human astrocytoma cells (glioblastoma multiforme or GBM). In whole-cell patch-clamped cultured GBM cells, the peptide toxin quickly and reversibly inhibited both inward and outward current with an IC50 of 36 ± 2 pM. The same inhibition was observed in freshly resected GBM cells. However, when the same experiment was performed on normal human astrocytes, the toxin failed to inhibit the whole-cell current. We also determined a cationic selectivity sequence for inward currents in three cultured GBM cell lines (SK-MG-1, U87-MG, and U215-MG).

The selectivity sequence yielded a unique biophysical

fingerprint with inward potassium conductance approximately 4- fold greater than sodium, lithium, and calcium. These observations suggest that PcTX1 may prove useful in determining whether or not GBM cells express a specific ASIC-containing ion channel type that can serve as a target for both diagnostic and therapeutic treatments of aggressive malignant gliomas.

Keywords: patch-clamp, malignant glioma, amiloride, ion channels, sodium, ASIC

Na+ currents and PcTX1

3

INTRODUCTION The ENaC/DEG (ENaC = epithelial sodium channels; DEG = degenerin) ion channel superfamily contains over 60 proteins having a similar topology: short intracellularly located N- and C- termini, two transmembrane spanning domains, and a large extracellular loop (2). All family members are cation selective ion channels and can be inhibited by amiloride (1, 17). One branch of this superfamily are the brain sodium channels (BNaC) also known as ASIC channels (acid sensing ion channel) (12, 21). The six mammalian members of this family cloned so far are primarily expressed in the brain and in sensory organs. Individual members of this family coassemble to form heteromeric channels with different properties, and are postulated to be involved in a variety of cellular responses including nociception and mechanosensation (5, 20). Data from our laboratory have revealed that all high- grade glioma cells, either derived from freshly resected tumors or from established cell lines, express a constitutively active, amiloride-sensitive inward sodium current displaying characteristics consistent with at least some of the properties of ASIC (3, 4, 25). In contrast, an amiloride-sensitive sodium current cannot be detected in astrocytes obtained from normal human brain tissue or from glioma cells derived from low- grade or benign tumors. Peptide toxins derived from the venom of poisonous animals such as snakes, spiders, coelenterates, fish, and other species have been important tools in developing strategies to inhibit ion conductance pathways (6, 9, 10, 16, 18). Escoubas et al. (7, 8) described the only compound to date that can inhibit ASIC channels with high affinity, namely, Psalmotoxin1 (PcTX1), a peptide derived from the venom of a West Indies tarantula. In experiments reported here we tested the hypothesis that PcTX1 can inhibit currents in high- grade glioma cells. We found that these currents are indeed inhibited by PcTX1. Thus, epitopes of ASIC may serve as a useful targets for therapeutic intervention as has been so successfully done for other major diseases (14, 24). One desirable prerequisite for a GBM-specific cellular "target" is that the protein expresses epitopes on the extra-cellular surface of the tumor cells. Ion channels (by definition) must have a portion of the protein

Na+ currents and PcTX1

4

outside the cell surrounding the conductive pore. Also, ion channels can be identified and characterized by a variety of biophysical properties such as conductance, kinetic activity, response to agonists, inhibition by pharmacological agents, and ionic selectivity. The development of the patch clamp technique for the investigation of ion channel biophysics and cellular regulation of ionic permeability has provided a means whereby specific ionic conductors can be identified and examined extensively in individual cells in culture or in cells isolated directly from resected malignant glioma tissue. Because ion channels present extracellular epitopes, their ionic currents can be examined with relative efficiency using patch clamp technology. We have found that GBM cells express a characteristic cation channel, and have investigated the possibility that GBM cells may express a unique cation channel that is not expressed by normal human brain astrocytes. Therapies designed to target malignant glioma cells must be specific for only the tumor cells. Therefore, we chose to develop multiple criteria for comparative analysis (malignant /non-malignant) in an attempt to isolate a specific target. In the present study, we used PcTX1 current- inhibition and cationic selectivity to show that a unique cation channel is expressed by GBM cells and that these channels are absent in normal human astrocytes. Thus, at least by these criteria, it appears as though a PcTX1-sensitive cation channel may provide an exclusive therapeutic target.

MATERIALS AND METHODS Cell Culture Primary cultures of normal human astrocytes and freshly resected glioblastoma multiforme, obtained from neurosurgical operations for epilepsy and tumor resections, respectively, along with continuous tumor cell lines (SK-MG-1, U251-MG, and U87-MG), were maintained in a 37°C humidified atmosphere containing 5% CO2 with Dulbecco’s Modified Eagle’s Medium (DMEM). This medium was supplemented with 10% fetal bovine serum (FBS) and 1% penicillin - streptomycin. All protocols for obtaining and

Na+ currents and PcTX1

5

culturing human malignant glioma tissue samples were reviewed and approved by UAB's Institutional Review Board (IRB Protocols X030403011, X011120003, and X020919005).

Electrophysiological methods: Micropipettes were constructed using a Narashigi pp-83 two stage micropipette puller. The tips of these pipettes had an inside diameter of approximately 0.3-0.5 µm, and an outside diameter of 0.7-0.9 µm. When filled with an electrolyte solution containing (in mM) K-gluconate, 100; KCl, 30; NaCl, 10; HEPES, 20; EGTA, 0.5; free Ca2+, <10 nM; ATP, 4; at a pH of 7.2, the electrical resistance of the tip was 1 to 3 MΩ. The initial bath solution was serum- free RPMI-1640 cell culture medium. These solutions approximate the usual ionic gradients across the cell membrane. The pipette was mounted in a holder connected to the headstage of an Axon 200A patch clamp amplifier affixed to a three dimensional micromanipulator system attached to the microscope.

The pipette tip was abutted to the cell and slight suction applied.

Seal

resistance was continuously monitored (Nicolet model 300 oscilloscope) using 0.1 mV electrical pulses from electrical pulse generator. After formation of a seal with a resistance in excess of 1 GΩ , another suction pulse was applied to form the whole-cell configuration by rupturing the membrane within the seal but leaving the seal intact. Successful completion of this procedure was indicated by a sudden increase in capacitance with no change in seal resistance.

The magnitude of the capacitance increase is a direct

function of the membrane available to be voltage clamped (i.e., cell size). Typically, this capacitance was between 10 and 20 picofarads (pF) both normal astrocytes and GBM astrocytes.

The cells were then be

held at a membrane potential of -60 mV, and clamped sequentially for 800 ms each to membrane potentials of ranging between -160 mV to 40 mV in 20 mV increments returning to the holding potential of -60 mV for 1 s between each test voltage. This procedure caused inward sodium (at more hyperpolarized potentials), and outward potassium (at more depolarized potentials) currents to flow across the membrane. These

Na+ currents and PcTX1

6

currents were digitally recorded and filed in real time. The entire procedure was controlled by a DOS Pentium computer modified for A/D signals with pCLAMP 6 software (Axon Instrs. Sunnyvale, CA). Once the whole cell configuration was established, the capacitance was balanced and initial currents were measured as control. In order to characterize the ion selectivity of the whole-cell currents and calculate the relative permeability for various cations, the bath solution was subsequently changed to (by perfusing the entire chamber) solutions A, B, C, D, and E, which contained Na+, Li+, Ca2+, K+ and NMDG+ respectively, as the major cation. Bath solution A contained (in mM): 140 NaCl, 5 KCl, 1 CaCl2 , 1 MgCl2 , 10 HEPES, and 5 Dextrose. Bath solution B contained (in mM): 140 LiCl2 , 1 MgCl2 , 10 HEPES, and 5 Dextrose. Bath solution C contained (in mM): 75 CaCl2 , 1 MgCl2 , 10 HEPES, 5 Dextrose and 75 Manitol. Bath solution D contained (in mM): 140 KCl, 1 CaCl2 , 1 MgCl2 , 10 HEPES, and 5 Dextrose. Bath solution E contained (in mM): 140 NMDG-Cl, 1 MgCl2 , 10 HEPES, and 5 Dextrose. The pH of the solutions were adjusted with NaOH, LiOH, Ca(OH)2 , KOH or Mg(OH)2, respectively, at 7.5. The absolute permeability coefficients for monovalent and divalent cations were retrieved by fitting the whole-cell current-voltage curves with the Goldman-Hodgkin-Katz current equation (13) using Origin version 7: I = Px ⋅ z ⋅ E test ⋅ F 2

[ X + ]out − [ K + ]in ⋅ e F ⋅Etest / R⋅T ⋅ (1 − e F⋅ Etest / R⋅T ) R ⋅T

where R, T, and F have their usual meanings, z is the valence of the cation, I represents the amiloridesensitive current carried by the cation X+, Px is the absolute permeability coefficients for the cation X+, and [X+]out and [K+]in stand for bath and pipette cation concentrations, respectively (13). As soon as the micropipettes were placed in contact with the cells, slight suction was applied. Seal resistance was continuously monitored using pCLAMP 8. After the formation of seals with resistances in excess of 1 GΩ, another suction pulse was applied to form the whole-cell configuration. Successful completion of this procedure produced a sudden increase in capacitance with no change in seal resistance,

Na+ currents and PcTX1

7

indicative of the whole-cell configuration. The cells were then voltage clamped at a membrane potential of 60 mV and clamped sequentially for 800 ms each to membrane potentials of -160 mV to +40 mV in 20 mV increments, returning to the holding potential of -60 mV for 1 s between each test voltage. The currents were recorded digitally and filed in real time. The entire procedure was controlled by pCLAMP 8 software through an A/D interface controlled by pCLAMP (Axon Instruments, Sunnyvale, CA). Data were analyzed off line at a later time. A minimum of three different cells were utilized for each experimental manipulation. Preparation of Normal Astrocytes: Brain tissue samples (non-malignant), obtained from patients undergoing surgery for epilepsy, were minced, washed in RPMI cell culture medium, and transferred to a 15 ml culture tube. The cells were dissociated using a solution of 1 ml (200X) collagenase, plus 4 ml trypsin (10X) diluted in 195 ml phosphate buffered saline. Minced tissue was suspended in this digestive solution at 37o C and agitated using a magnetic stirring bar.

The cell suspension was removed after twenty minutes of digestion, and

centrifuged to remove debris. The pelleted viable cells were then re-suspended and cultured. Because the tissue samples were obtained from epileptic patient resections, there was no possibility that the astrocytes could be contaminated with GBM cells. Also, the only viable cells in culture from these resections are astrocytes as determined in independent experiments using astrocytic or neuronal -specific markers. GBM cells were obtained from high grade freshly resected gliomas maintained in a similar manner. Also, three continuous GBM cell cultures GBM cells (SK-MG-1, U251-MG, and U87-MG) were used.

RESULTS PcTX1 inhibits currents in GBM cells. The active component of the venom of a West Indies tarantula P. cambridgei is a forty amino acid peptide containing six cystines. Escoubas et al. (7, 8) originally showed

Na+ currents and PcTX1

8

that PcTX1 was specific for rat ASIC1a. We have shown previously that glioma cells express ASIC channels (4).

PcTX1 inhibits plasma membrane ionic current in human glioma cells: We tested the hypothesis directly tha t PcTX1 can inhibit plasma membrane ionic current in human cells. In these experiments, we used a continuous glioma cell line originally derived from a human glioblastoma multiforme (GBM), namely, SK-MG-1 cells. The results of these initial experiments are shown in Figure 1. In these experiments, an individual SK-MG-1 cell was whole-cell patch-clamped, and transmembrane currents were recorded at 11 different clamp potentials to evaluate the current-voltage relationship. Also, we recorded currents in real time by continuously applying positive and negative pulses, and returning to the holding potential between each test potential. Typical results of these voltage clamp protocols are shown in Figure 1. In the top portion of Figure 1, the holding potential was -60 mV and the cell was pulsed to -120 mV, then to +60mV, and back to the holding potential at 1.25 s intervals. After stable currents were recorded, 1nM PcTX1 was superfused over the cell. A very rapid inhibition was seen such that within 30 s virtually all of the current was inhibited. The bottom records show typical whole-cell clamp currents before and after application of PcTX1. Figure 2 displays a dose-response curve for PcTX1 induced inhibition of inward sodium currents (measured at -120 mV) in SK-MG-1 cells (log scale). From these data, an apparent equilibrium inhibition constant (K i) of 36±2 pM was obtained. Because GBMs are, by nature, a heterogeneous group of tumors, we tested the effects of PcTX1 on currents obtained in a number of different tumor cells maintained in continuous culture. The findings were consistent in that PcTX1 inhibited endogenous currents in each GBM cell line (one example is shown in the inset to Figure 2). However, it is possible that transformed cells in culture may express ion channels that are

Na+ currents and PcTX1

9

not expressed by GBMs in situ. Therefore we obtained freshly resected GBM tissue, isolated the tumor cells, and performed the same electrophysiological and PcTX1 analyses. Figure 3 shows that 10 nM PcTX1 can completely block inward sodium currents in a whole-cell clamped freshly resected GBM cell with approximately the same affinity as was found in the cultured GBM cells. The control or scrambled peptide was without effect. When normal human astrocytes were whole-cell patch-clamped, ionic currents could also be recorded, however, 10 nM PcTX1 did not affect these whole-cell currents (N=8; not shown).

Astrocytes derived from high-grade gliomas have a unique plasma membrane ionic selectivity: To characterize more completely the PcTX1-sensitive current, we assessed ionic selectivity in whole-cell clamped astrocytes derived from three high grade glioma cell lines (SK-MG-1, U-87-MG, and U-251-MG). All of the cells examined had substantial inward and outward currents when normal ionic gradients for sodium, potassium and chloride were used in the pipette and bath solutions. Subsequently the bath solution was changed by replacing the normal solution with solutions that had lithium, calcium, or potassium substituted for sodium. The three solution changes were repeated no less than 3 times on each of the three cell lines examined. The findings were indistinguishable in each of the cell lines. The currents were similar when sodium, lithium, and calcium were in the bath solution. However, when sodium was substituted with potassium in the bath solution, in every case large inward currents were measured at hyperpolarizing voltage clamp potentials. A typical experiment on a U87-MG cell is shown in Figure 4. The summarized current magnitudes for each of the cell lines studied are shown in the current-voltage relations depicted in Figure 5. Also, the ability of amiloride (100 µM) to inhibit the potassium conducted inward current is shown in the lower right panel of Figure 5. These observations suggest that GBM cells conduct potassium in the inward direction with a far greater efficacy (2.5 to 6.8 times depending on the cell line) than they conduct

Na+ currents and PcTX1

10

sodium. This fact is not physiologically relevant; however, this characteristic is useful in our effort to determine whether this specific conduction pathway is, or is not, present in the plasma membrane of normal human astrocytes. Permeability coefficients derived from the Goldman-Hodgkin-Katz current equation for each of the cations on each of the cell lines are given in Table 1. Because the inward potassium currents in GBM cells were amiloride-sensitive, we hypothesized that these K+ currents would also be inhibited by PcTX1. Therefore, GBM cells were superfused with PcTX1 in the continued presence of potassium (Figure 6, upper panels). This toxin rapidly inhibited the inward currents carried by potassium as shown in the far right panel. The same experiment was also performed on normal (non- malignant) human brain astrocytes. The findings (shown in the lower panels of Figure 6) were that the plasma membrane conductance did not increase when potassium was substituted in the bath solution. Because no increase in current was observed, there was no current to inhibit with PcTX1. Thus, the evidence suggests that the ionic conductance present in tumor cells is absent from normal brain astrocytes.

Normal human astrocytes do not express whole-cell current similar to those found in GBM cells: We previously reported that amiloride-sensitive currents could be induced in normal astrocytes by disrupting simultaneously two inhibitory mechanisms that are tonically active in normal astrocytes (4). Inhibition of syntaxin 1A with Munc-18, combined with inhibition of protein kinase Cβ, induces a plasma membrane sodium permeability of approximately the same magnitude as that found constitutively in tumor cells. Because we do not precisely know the subunit composition of these induced amiloride-sensitive channels in normal astrocytes or in GBM cells, we assessed the PcTX1 sensitivity and this ionic selectivity of the induced current in normal human astrocytes, and compared the properties to those found in GBM cells.

Na+ currents and PcTX1

11

Normal human astrocytes do not express whole-cell current similar to those found in GBM cells: Normal human astrocytes were whole-cell clamped using normal ionic gradients described in the methods. Under these conditions the cells had minimal inward currents (Figure 7, upper left panel) that could be either minor leak currents across the seal, or minimal currents carried by sodium through potassium channels because there can be some small but finite ability of sodium to permeate the plasma membrane through other channel types under the relatively strong voltage clamps that were used for these experiments. However, no significant currents that could be attributed to any type of amiloride-sensitive channels were present in any of the normal astrocytes that were examined for this study under basal conditions. When these normal astrocytes were treated with agents that inhibit endogenous proteins that can themselves inhibit ENaC/DEG channels (i.e., LY379196, 100 nM) to block protein kinase C−β, and Munc18 (8 pM in the pipette solution) to inhibit syntaxin-1A, currents with electrical characteristics similar to those seen in GBM cells under basal conditions were activated. Figure 7 (upper right panel) shows the effect of these treatments on a normal human astrocyte. In the presence of both inhibitors, inward currents were activated, and the current- voltage relation was shifted to the right. These currents were subsequently inhibited with 100 µM amiloride. These experiments demonstrate that normal human astrocytes contain functional amiloride-sensitive channels, and extend our previous results that showed the presence of ENaC/DEG mRNA in these cells (4). The question we addressed using PcTX1 and selectivity measurements was: is this induced conductive pathway in normal astrocytes identical to that expressed by GBM cells? Potassium does not increase inward conductance in normal human astrocytes: We first assessed the selectivity of the basal currents of normal human astrocytes to test our hypothesis that the induced conductance in normal astrocytes was not identical to that endogenously

Na+ currents and PcTX1

12

expressed by high grade gliomas. Figure 6 shows that when the bath solution sodium was replaced with potassium, the whole-cell currents across the plasma membrane of a normal human astrocyte were not altered, and no larger inward current appeared. Thus, by selectivity criteria, we were unable to demonstrate the presence of the functional ENaC/DEG- mediated currents that we observed in malignant human astrocytes. However, further characterization was required to determine whether or not the induced current was mediated by the same channel species that is present in GBM cells, or a similar channel possibly composed of different ENaC/DEG orthologs. We therefore induced amiloride-sensitive currents in normal human astrocytes by treatment with Munc18 and a specific PKCβ inhibitor. In contrast to GBM cells, PcTX1 had no effect. Subsequently the bath solution Na+ was changed to K+. The change in the external cation inhibited the induced currents. This finding was exactly opposite of what was observed in GBM cells. Thus by two criteria, refractivity to PcTX1 and different Na+/K + ionic selectivity ratios, we conclude that the induced ENaC/DEG currents in normal astrocytes are different from those basally expressed by GBM cells. We deduce, therefore, that glioblastoma multiforme cells express a unique ionic conductor.

DISCUSSION Toxins have been instrumental for understanding structure/function relationships of ion channels as well as being very useful for therapeutic intervention in a number of diseases (11, 15, 19, 22, 23). Examples of such toxins include tetrodotoxin and saxitoxin (voltage-sensitive Na+ channels), charybdotoxin, iberiotoxin, and apamin (Ca2+ activated K+ channels), chlorotoxin (Cl- channels), and ϖ-canotoxin and ϖagatoxin (voltage-dependent Ca2+ channels). Previous studies on the crude venom and purified peptide component from the tarantula Psalmopoeus cambridgei (the Trinidad Chevron) demonstrated high specificity of PcTX1 towards ASIC1a, a member of the non- voltage gated ENaC/DEG ion channel

Na+ currents and PcTX1

13

superfamily. PcTX1 has limited homology with other known spider toxins, but it does share a conserved cysteine distribution with that found in other spider and cone snail peptides. In this study, we show that PcTX1 effectively inhibits basally active cation currents in malignant astroglioma cells, while leaving ionic currents in normal (i.e., non- malignant) human astrocytes unaffected. A constitutive amiloride-sensitive inward Na+ current exists in high- grade glioma cells (4). These inward Na+ currents persist in primary cultures of freshly resected high-grade gliomas as well as in established cell lines derived from high- grade astrocytomas. These inward Na+ currents are not present in normal astrocytes or in low- grade astrocytomas (e.g., pilocytic astrocytomas). However, the composition of the channels responsible for this inward Na+ conductance is unknown. We hypothesize, based on both functional and biological evidence, that this channel is heteromeric and is comprised of ASIC1a and some ENaC components, most likely δ- and/or γ-ENaC (4). Thus, we hypothesize that, at a minimum, ASIC1/2 plus γ- (or δ-) ENaC exist in the plasma membrane of normal astrocytes, whereas only ASIC1 plus γ- (or δ-) ENaC is present in the plasma membrane of malignant glioma cells (4). In the absence of ASIC2 in the plasma membrane, syntaxin1A, which is present in all cells of astrocytic origin, cannot inhibit channels comprised primarily of ASIC1; hence, the channel would be active. Moreover, we have previously shown that PKCβ, which is present in normal cells but not GBM cells, inhibits both ASIC1 and the currents constitutively active in GBM cells (3). The fact that PcTX1 can inhibit inward currents in GBMs supports the hypothesis that channels mediating this current are comprised, at least in part, of ASIC1 and do not contain any ASIC2. Even when inward currents are activated in normal astrocytes by the combination of PKCβ inhibitors and Munc18 (3, 4), PcTX1 is without effect. This observation is consistent with the hypothesis that amiloride-sensitive channels in normal astrocytes are different than those in high- grade gliomas.

Na+ currents and PcTX1

14

PcTX1 is a peptide toxin. Because it is a rather complex peptide, some uncertainty remains as to whether or not this ligand will be directly useful as a therapy for the treatment of malignant gliomas. Our findings thus far, have indicated the PcTX1 binds to an epitope on the surface of malignant human astrocytes. The binding is inferred because the toxin inhibits ionic conduction through a specific channel protein complex. We have not yet performed any histological studies to confirm the presence of the toxin on the cell surface. We have observed that upon washout of the toxin the currents return, therefore, the toxin's interaction with the channel is not permanent. This makes histological studies problematic. The reversibility of the effect is also of concern when postulating the use of the toxin as a therapy. All amiloride-sensitive ion channels characterized to date, including cloned members of the ENaC/DEG superfamily, have a Na+ permeability equal to or greater that that of K+, never less (see Table 2). This appears not to be the case for the amiloride- inhibitable ion conductance pathway in high- grade glioma cells. Moreover, this pathway is exquisitely sensitive to inhibition by the spider toxin PcTX1, a molecule heretofore thought to be specific for ASIC1a (8). It is unlikely that the current pathway in GBM cells is comprised exclusively of ASIC1a for several reasons. First, the GBM current is constitutively active, in contrast to ASIC1a, which must be activated either by a sudden reduction in extracellular pH (8) or by or by a reduction in intracellular Ca2+ (4, 8). Second, ASIC1a inactivates, whereas the GBM currents are always active (3). Third, the PNa+/PK+ of ASIC1a is >6, in contrast to that of the GBM pathway, namely, 0.15-0.36. Thus, we propose that the amiloride-sensitive current measured in GBM cells represents a unique ion channel undoubtedly comprised of

different ENaC/DEG subunits, one of which (ASIC1a) is an

essential component. Another unresolved question is the cell physiological function of the ion channel complex with which PcTX1 interacts. It is possible that the malignant cells require the channel complex for a specific function such as migration. At this time it is equally likely that the malignant cells can survive and migrate

Na+ currents and PcTX1

15

without utilizing the PcTX1-sensitive ionic conductance, and that transiently blocking this channel would have no long- lasting detrimental effect on the malignant astrocytes. However, these possibilities do not rule out the use of this toxin as a therapeutic agent. There are both advantages and disadvantages of using a complex peptide for therapy. For example, the amino acid sequence could be modified to produce a related peptide that interacts more strongly, or even irreversibly with the ion channel complex. Another possibility is that the peptide could be bound to a cytotoxic agent, and be used to deliver such an agent to the malignant cell specifically. This toxin may eventually be used in surgery to help guide resection, i.e., a fluorescent molecule may be attached to identify a group of abnormal cells or show upon neuroimaging that could help differentiate between necrosis and active tumor. One of the negative possibilities that must be investigated is the potential antigenicity of the toxin. Use of a complex peptide could induce a strong immune response that could result in inflammation or limit any repetitive use of the toxin in afflicted individuals. One of the most important criteria for any possibility of success for an eventual treatment strategy is the requirement for specificity for malignant cells only. The ASIC-containing conductive pathway may meet these criteria. However there are some inconsistencies that must be resolved. One inconsistency is that mRNA for ASIC1a can be identified in normal human astrocytes. For this reason we sought to increase the number of independent functional criteria that can be used to assess the question of specificity. The unique selectivity profile described in these experiments compliments the PcTX1- mediated current inhibition because PcTX1 inhibits the inward potassium in GBM cells, and we could not find a potassium conductance in normal cells. Thus by two independent criteria we have demonstrated the specificity (for malignant astrocytes) of a set of ion channels that must express an extracellular epitope that could be a target for therapeutic intervention. Our data show that PcTX1 is a specific inhibitor for GBM cells. Thus, it is possible that PcTX1 is a potential carrier for cytotoxic agents that could be used to kill the residual malignant cells at the time of surgery.

Na+ currents and PcTX1

16

In data that was not shown, we have observed that PcTX1 inhibition was readily reversible when the toxin was washed from the bath solution. This observation means that any potential treatment using this specific ligand would be transient because in situ the ligand would eventually unbind from the target and be washed away. We have not yet determined whether or not PcTX1 would be a suitable ligand. More experiments must be performed before this determination can be made. Nonetheless, because we can demonstrate efficacy by inhibition of current, PcTX1 remains a viable candidate as a carrier ligand, and offers some hope that a suitable therapy can be devised that can improve the currently poor prognosis for the treatment of glioblastoma multiforme.

ACKNOWLEDGEMENTS The authors express their thanks to Cathy Langford and Melissa McCarthy for their excellent work in preparing and culturing of GBM and normal astrocytes. This research was supported by NIH grants CA10195 and CA97247, and by funds from the Brain Tumor Foundation for Children.

REFERENCES

1.

Benos DJ. Amiloride: a molecular probe of sodium transport in tissues and cells. Am J

Physiol 242: C131-145, 1982. 2.

Benos DJ and Stanton BA. Functional domains within the degenerin/epithelial sodium

channel (Deg/ENaC) superfamily of ion channels. J Physiol 520 Pt 3: 631-644, 1999. 3.

Berdiev BK, Xia J, Jovov B, Markert JM, Mapstone TB, Gillespie GY, Fuller CM,

Bubien JK, and Benos DJ. Protein kinase C isoform antagonism controls BNaC2 (ASIC1) function. J Biol Chem 277: 45734-45740, 2002.

Na+ currents and PcTX1

4.

17

Berdiev BK, Xia J, McLean LA, Markert JM, Gillespie GY, Mapstone TB, Naren

AP, Jovov B, Bubien JK, Ji HL, Fuller CM, Kirk KL, and Benos DJ. Acid-sensing ion channels in malignant gliomas. J Biol Chem 278: 15023-15034, 2003. 5.

Biagini G, Babinski K, Avoli M, Marcinkiewicz M, and Seguela P. Regional and

subunit-specific down regulation of acid-sensing ion channels in the pilocarpine model of epilepsy. Neurobiological Disease 8: 45-58., 2001. 6.

Church JE, Moldrich RX, Beart PM, and Hodgson WC. Modulation of intracellular

Ca2+ levels by Scorpaenidae venoms. Toxicon 41: 679-689, 2003. 7.

Escoubas P, Bernard C, Lambeau G, Lazdunski M, and Darbon H. Recombinant

production and solution structure of PcTx1, the specific peptide inhibitor of ASIC1a protongated cation channe ls. Protein Sci 12: 1332-1343, 2003. 8.

Escoubas P, De Weille JR, Lecoq A, Diochot S, Waldmann R, Champigny G,

Moinier D, Menez A, and Lazdunski M. Isolation of a tarantula toxin specific for a class of proton-gated Na+ channels. Journal of Biological Chemistry 275: 25116-25121., 2000. 9.

Fan CX, Chen XK, Zhang C, Wang LX, Duan KL, He LL, Cao Y, Liu SY, Zhong

MN, Ulens C, Tytgat J, Chen JS, Chi CW, and Zhou Z. A novel conotoxin from Conus betulinus, kappa-BtX, unique in cysteine pattern and in function as a specific BK channel modulator. J Biol Chem 278: 12624-12633, 2003. 10.

Gao YD and Garcia ML. Interaction of agitoxin2, charybdotoxin, and iberiotoxin with

potassium channels: selectivity between voltage-gated and Maxi-K channels. Proteins 52: 146154, 2003. 11.

Garcia ML, Gao Y, McManus OB, and Kaczorowski GJ. Potassium channels: from

scorpion venoms to high- resolution structure. Toxicon 39: 739-748, 2001.

Na+ currents and PcTX1

12.

18

Garcia-Anoveros J, Derfler B, Neville-Golden J, Hyman BT, and Corey DP. BNaC1

and BNaC2 constitute a new family of human neuronal sodium channels related to degenerins and epithelial sodium channels. Proceedings of the National Academy of Sciences U S A 94: 1459-1464., 1997. 13.

Ji HL, Parker S, Langloh AL, Fuller CM, and Benos DJ. Point mutations in the post-

M2 region of human alpha-ENaC regulate cation selectivity. Am J Physiol Cell Physiol 281: C64-74, 2001. 14.

Jilani I, O'Brien S, Manshuri T, Thomas DA, Thomazy VA, Imam M, Naeem S,

Verstovsek S, Kantarjian H, Giles F, Keating M, and Albitar M. Transient down- modulation of CD20 by rituximab in patients with chronic lymphocytic leukemia. Blood 102: 3514-3520, 2003. 15.

Jones RM and Bulaj G. Conotoxins - new vistas for peptide therapeutics. Curr Pharm

Des 6: 1249-1285, 2000. 16.

Jouirou B, Mosbah A, Visan V, Grissmer S, M'Barek S, Fajloun Z, Van Rietschoten

J, Devaux C, Rochat H, Lippens G, El Ayeb M, De Waard M, Mabrouk K, and Sabatier J. Cobatoxin 1 from centruroides noxius scorpion venom: chemical synthesis, D structure in solution, pharmacology and docking on K+ channels. Biochem J, 2003. 17.

Kellenberger S and Schild L. Epithelial sodium channel/degenerin family of ion

channels: a variety of functions for a shared structure. Physiol Rev 82: 735-767, 2002. 18.

Lopez-Gonzalez I, Olamendi-Portugal T, De la Vega-Beltran JL, Van der Walt J,

Dyason K, Possani LD, Felix R, and Darszon A. Scorpion toxins that block T-type Ca2+ channels in spermatogenic cells inhibit the sperm acrosome reaction. Biochem Biophys Res Commun 300: 408-414, 2003.

Na+ currents and PcTX1

19.

19

Lyons SA, O'Neal J, and Sontheimer H. Chlorotoxin, a scorpion-derived peptide,

specifically binds to gliomas and tumors of neuroectodermal origin. Glia 39: 162-173, 2002. 20.

Price MP, Lewin GR, McIlwrath SL, Cheng C, Xie J, Heppenstall PA, Stucky CL,

Mannsfeldt AG, Brennan TJ, Drummond HA, Qiao J, Benson CJ, Tarr DE, Hrstka RF, Yang B, Williamson RA, and Welsh MJ. The mammalian sodium channel BNC1 is required for normal touch sensation. Nature 407: 1007-1011., 2000. 21.

Price MP, Snyder PM, and Welsh MJ. Cloning and expression of a novel human brain

Na+ channel. Journal of Biological Chemistry 271: 7879-7882., 1996. 22.

Sher E, Giovannini F, Boot J, and Lang B. Peptide neurotoxins, small-cell lung

carcinoma and neurological paraneoplastic syndromes. Biochimie 82: 927-936, 2000. 23.

Soroceanu L, Gillespie GY, Khazaeli MB, and Sontheimer H. Use of chlorotoxin for

targeting of primary brain tumors. Cancer Res 58: 4871-4879, 1998. 24.

Tobinai K. [New trends in the diagnosis and treatment of malignant lymphoma]. Nippon

Igaku Hoshasen Gakkai Zasshi 63: 81-88, 2003. 25.

Xia J, Zhou ZH, Bubien JK, Fuller CM, Markert JM, Mapstone TB, Yancey

Gillespie G, and Benos DJ. Molecular cloning and characterization of human acid sensing ion channel (ASIC)2 gene promoter. Gene 313: 91-101, 2003.

Na+ currents and PcTX1

20

Figure Legends

Figure 1. Effect of PcTX1 on Whole-Cell Currents in SK-MG-1 Cells. Upper panel: Realtime inhibition of the currents after addition of PcTX1. Lower panel: Steady-state whole-cell currents over a wide voltage-clamp range before and after treatment with PcTX1.

Figure 2. Dose-Response Curve Showing the Inhibitory Effects of Increasing Concentrations of PcTX1 on Inward Na+ Currents in SK-MG-1 Cells. Insert: Whole-cell current record sets showing the efficacy of PcTX1, and the inability of the control peptide to inhibit the currents.

Figure 3. Upper Panels: Whole-cell clamp of a cell from a freshly resected GBM. The middle panel shows that 10 nM PcTX1- ihibited the inward current. The current that was inhibited by PcTX1 is shown in the right panel. Lower Panels: The same experiment was performed on a different GBM astrocyte. This cell was superfused with a control peptide composed of the same amino acids in the same proportions but in a different sequence. This control peptide had no effect on the currents (middle panel) and, thus, there was no difference in the currents (right panel).

Figure 4. A representative selectivity sequence performed on a single whole-cell clamped GBM astrocyte showing the large inward current increase observed in every GBM cell studied when potassium was substituted for sodium in the bath solution.

Na+ currents and PcTX1

21

Figure 5. Summary data in the form of current-voltage relations for the selectivity sequences performed on 3 different cell lines derived from gliomas. The lower right panel shows the inhibition of the inward current when potassium was the charge carrier, indicating that the inward K+ current was being conducted by ASIC channels.

Figure 6. Upper Panels: Whole-cell current records from a GBM astrocyte showing the increased inward current generated when potassium was substituted for sodium in the bath solution, and the subsequent inhibition of the current by PcTX1 while the conductive species remained potassium. Lower Panels: These whole-cell current records show that in a non- malignant human astrocyte, potassium failed to alter the plasma membrane ionic permeability. The difference current record (right panel) shows the lack of any increased permeability whether the main external cation was sodium or potassium.

Na+ currents and PcTX1

22

Figure 7. The upper left panel shows a typical current set obtained from a whole-cell clamped non-malignant human astrocyte. The cells were pretreated with LY 379196 to inhibit protein kinase Cβ. This record illustrates the observation that inhibition of PKCβ was insufficient to activate ASIC currents. Records showing this current phenotype were obtained immediately (< 5s) after formation of the whole-cell configuration. The upper right panel shows the currents from the same cell approximately 2 m after the basal records were obtained. The increased currents were activated as the munc18 contained in the pipette solution diffused into the cell via the membrane breach that created the whole-cell configuration. The lower left panel shows that the activated currents were substantially inhibited by 100 mM amiloride, consistent with the properties of ASIC channels. The lower right panel shows the average current voltage relations obtained using non- malignant astrocytes and the "disinhibitory" treatment combination.

Figure 8. These whole-cell current records show a normal human astrocyte that has very little plasma membrane ionic conductance (upper left). However, a conductance can be activated in these cells by inhibition of protein kinase C and syntaxin (upper middle). The ASIC1a-specific ligand (PcTX1) has no inhibitory effect on the activated currents (upper right). When sodium is substituted for with potassium (lower left), the conductance is no longer present. Reintroduction of sodium into the bath solution restores the activated whole-cell conductance. The selectivity of the conductance and the inability of PcTX1 to inhibit it are two independent criteria that are consistent with the hypothesis that non- malignant human astrocytes do not express and ASIC1a containing ion channels.

Na+ currents and PcTX1

Figure 1

23

Na+ currents and PcTX1

Figure 2

24

Na+ currents and PcTX1

Figure 3

Figure 4

25

Na+ currents and PcTX1

Figure 4

26

Na+ currents and PcTX1

Figure 5

27

Na+ currents and PcTX1

Figure 6

28

Na+ currents and PcTX1

Figure 7

29

Na+ currents and PcTX1

Figure 8

30

Na+ currents and PcTX1

31

Table 1 Cell Line

Potassium

Sodium

Lithium

Calcium

U87-MG

1.0 X 10 -5cm/s

4.0 X 10 -6cm/s

4.0 X 10 -6cm/s

2.0 X 10 -6cm/s

U251-MG

2.0 X 10 -5cm/s

4.5 X 10 -6cm/s

5.0 X 10 -6cm/s

5.0 X 10 -6cm/s

SK-MG-1

2.0 X 10 -5cm/s

2.9 X 10 -6cm/s

3.4 X 10 -6cm/s

2.0 X 10 -6cm/s

Table 1: Permeability coefficients for each cation in the three glioma cell lines tested. All permeability coefficients have a standard error of + 1 x 10-6 cm/s.

Na+ currents and PcTX1

32

Table 2 Name

Pseudonym

PNa+/PK+

Ki AmilµM

γNa,pS

Distribution

ASIC1a

ASIC;BNaC2;

6-13

10

14.3

Sensory neurons of DR trigeminal ganglia

BNC2;ACCN2 ASIC1b

ASIC1b;BNaC1

3

21

---------

Sensory neurons

ASIC2a

BNaC; MDEG;

10

28

10.4

Widespread in nervous system

BNC1;ACCN1 ASIC3 (a,b,c)

TNAC1; SLNAC1; 5-12

60-100

13-15

Testis, lung

ACCN3; DRASIC ASIC4

SPASK

Sensory ganglia, brain

----------

-----------

----------

Pituitary, brain, Spinal cord, inner ear

ASIC1a/2a

----------------------

7

20

10

---------------------------

ASIC2a/3

----------------------

10

-----------

10

----------------------------

αβγENaC

----------------------

>50

0.1

4-5

Kidney, lung, colon, Lymphocytes, brain

δβγENaC

----------------------

>50

0.26

12

Brain, pancreas, testis, ovary

GBM Channel

LCED

0.15-0.36

10-100

----------

Malignant gliomas

Modified from Kellenberger, S. and Schild, L. Physiological Reviews 82: 735-767, 2002.