Regulation by Spermine of Native Inward Rectifier K (cid:49) Channels in RBL-1 Cells*

Polyamines have been shown to participate in the rectification of cloned inwardly rectifying potassium channels, a class of potassium channel proteins that conducts inward current more readily than outward current. Here, basophil leukemia cells were used to de- termine the effects of polyamines on a native, inwardly rectifying potassium current. Rat basophil leukemia cells were cultured in the presence of two different polyamine biosynthesis inhibitors, and both the electro- physiological properties and the polyamine levels were monitored. Treatment with (cid:97) -difluoromethylornithine, a specific ornithine decarboxylase inhibitor, resulted in no significant change of electrophysiological properties. In contrast, (cid:42) inhibitor outward through inwardly rectifying channels pu- trescine markedly increased spermidine and spermine decreased. Fluctuations of intra- cellular polyamine as imposed by directly in an altered cell excitability.

Polyamines have been shown to participate in the rectification of cloned inwardly rectifying potassium channels, a class of potassium channel proteins that conducts inward current more readily than outward current. Here, basophil leukemia cells were used to determine the effects of polyamines on a native, inwardly rectifying potassium current. Rat basophil leukemia cells were cultured in the presence of two different polyamine biosynthesis inhibitors, and both the electrophysiological properties and the polyamine levels were monitored. Treatment with ␣-difluoromethylornithine, a specific ornithine decarboxylase inhibitor, resulted in no significant change of electrophysiological properties.

In contrast, treatment with 5-{[(Z)-4-amino-2-butenyl]methyl-amino}-5-deoxyadenosine (MDL73811)
, an inhibitor of S-adenosylmethionine decarboxylase, resulted in increased outward currents through inwardly rectifying potassium channels while intracellular putrescine was markedly increased and spermidine and spermine levels were decreased. Fluctuations of intracellular polyamine concentrations as imposed by MDL73811 were directly translated in an altered cell excitability. Based on these results we conclude that the rectification properties of native inwardly rectifying potassium channels are largely controlled by intracellular spermine.
Rectification of cloned inwardly rectifying potassium channels (IRKs) 1 depends on two processes: (a) a fast voltage-dependent block of the open channel pore by internal Mg 2ϩ ions (1,2), and (b) a much slower voltage-dependent block of the open channel by cytoplasmic polyamines, in particular spermidine and spermine (3)(4)(5). The block by polyamines (PAs) is important, as it controls the shape of the current-voltage rela-tionship in strong IRKs such as IRK1, HIR, or hIRK (6 -8), where outward currents reach a maximum with increasing depolarizations and then shut down completely. PA block creates a physiologically important region of negative slope conductance at potentials just above the potassium equilibrium potential, E K , which first limits and then terminates the stabilizing effects of strong IRKs on the membrane potential (9). Weakly rectifying IRKs, such as ROMK1, are more than 4 orders of magnitude less sensitive to spermidine or spermine, and their mild rectification properties are mainly controlled by Mg 2ϩ ions (10 -12).
PA block is not restricted to IRKs. Block at micromolar concentrations has been reported for a subfamily of glutamate receptors where the rectification properties of Ca 2ϩ -permeable forms of the ␣-amino-3-hydroxy-5-methyl-isoxazolepropionate receptor are controlled by intracellular spermine (13)(14)(15). In addition, intracellular polyamines were found to bind to and modulate voltage-gated and Ca 2ϩ -gated K ϩ channels (16,17).
Putrescine, spermidine, and spermine are essential for cell growth and differentiation (18). Membrane proteins that depend on polyamines for proper function are exposed to a wide range of intracellular polyamine concentrations, since biosynthesis is regulated by growth factors, mitogens, and hormones. Polyamine biosynthesis is usually enhanced during cell growth but may also be increased in brain after excessive electrical stimulation of neurons or after epileptic episodes (19,20). Pharmacological manipulations that inhibit polyamine biosynthesis result in decreased growth rates and/or cell death. The importance of polyamines in malignant cell growth has made their biosynthetic enzymes prime targets for therapeutic interventions leading to the development of potent inhibitors for key enzymes in polyamine metabolism. The most widely studied inhibitors include ␣-difluoromethylornithine (DFMO) for ornithine decarboxylase, which catalyzes the formation of putrescine from L-ornithine, and 5Ј-{[(Z)-4-amino-2-butenyl]methyl-amino}-5Ј-deoxyadenosine (MDL73811), which inhibits S-adenosylmethionine decarboxylase, the enzyme that provides aminopropyl groups for the synthesis of spermidine and spermine from putrescine (21).
In the present study, DFMO and MDL73811 were used as pharmacological tools to manipulate the internal polyamine content of rat basophil leukemia (RBL-1) cells. RBL-1 cells were selected because their major membrane current is an inwardly rectifying K ϩ conductance (22). This conductance has been shown recently to originate from a strong inward rectifier, rIRK1, with 94% homology to the cloned mouse IRK1 channel (6,23). Here, we show for the first time that the rectification properties of native IRK channels are controlled by intracellular spermine and can be changed by pharmacological manipulation of the intracellular spermine level.
Cell Culture and Treatments-RBL-1 cells were grown in RPMI 1640 medium supplemented with 10% fetal calf serum, streptomycin (10 g/ml), and penicillin (10 units/ml) and maintained at 37°C in 5% CO 2 . RBL-1 cells were replated either at low densities (5-8 ϫ 10 5 cells/ml of medium) on glass coverslips for electrophysiological recordings or at a 5 times higher density into 6-well culture plates for determination of intracellular polyamines. For treatment with polyamine biosynthesis inhibitors, the culture medium was replaced 24 h after plating by medium supplemented with appropriate inhibitor concentrations. DFMO and MDL73811 were used in aqueous stocks. During treatment the culture medium was replaced daily. Control cell cultures were handled the same way. Prior to electrophysiological recordings or HPLC measurements cells were allowed to recover for at least 8 h after the last medium change.
Determination of Intracellular Polyamine Levels-Cellular polyamine concentrations were determined as described previously (24). In brief, cells were washed twice with ice cold phosphate-buffered saline, scraped from wells, and frozen at Ϫ80°C. Thawed samples were sonicated, and proteins were precipitated with perchloric acid. After the addition of 1,7-diaminoheptane as internal standard, samples were neutralized with K 2 CO 3 , and polyamines were derivatized with 5-dimethylamino-napthalene-1-sulfonyl chloride.
Derivatized polyamines were partially purified with Sep-Pak C 18 cartridges. Samples were further fractionated by HPLC, employing a Partisil-10 ODS column, utilizing one of two elution methods. The first method employed acetonitrile in water as mobile phase. The column was equilibrated with 56% acetonitrile. One minute after sample injection a linear gradient to 78% acetonitrile over 10 min was carried out. The gradient was then increased to 86% acetonitrile over 10 min, followed by 90% acetonitrile for 15 min. The second procedure, which provides enhanced resolution of putrescine (25), utilized as mobile phases 92.5% acetonitrile, 7.5% methanol (Solvent A) and 10 mM monopotassium phosphate (pH 4.4, Solvent B). First, the column was equilibrated with 35% Solvent A and 65% Solvent B. One minute after sample injection a linear gradient to 60% Solvent A was carried out. The gradient was then increased to 90% Solvent A over 5 min and held there for an additional 20 min. Column effluent was monitored with a fluorescence detector using excitation and emission filters of 305-395 and 435-650 nm, respectively.
Authentic putrescine, spermidine, and spermine standards were car-ried through the entire procedure to establish column retention times and calibration curves for each polyamine. Concentrations were expressed as pmol of polyamine/g of DNA and given as means Ϯ S.E. DNA concentrations were measured according to the method of Burton (26). Electrophysiology-Membrane currents were measured in the whole cell version of the patch clamp technique (31) using an Axoclamp 1B amplifier (Axon Instruments, Foster City, CA). Pipettes had resistances of 2-5 M⍀ when filled with (in mM): 130 potassium aspartate, 10 NaCl, 4 CaCl 2 , 2 MgCl 2 , 10 HEPES, and 10 EGTA (pH 7.4). In some experiments spermine (0.1 or 1 mM) was added to the pipette solution. The extracellular bathing solution had the following composition (in mM): 130 NaCl, 4.4 KCl, 2 CaCl 2 , 2 MgCl 2 , 10 HEPES, and 5 D-glucose (pH 7.4). All recordings were done at room temperature. The seal resistance was usually higher than 10 G⍀. The input resistance of cells included in the analysis was 2-6 G⍀. Capacitive currents were evoked by small voltage clamp pulses of Ϫ5 mV from a holding potential of 0 mV and compensated by the analogue circuit of the amplifier. The readout of the compensation circuit was taken as an estimate of the membrane capacitance and used to normalize current amplitudes and calculate current densities. Such capacitance measurements give an estimate of cell size, since the specific membrane capacity is assumed to be rather constant at 1 F/cm 2 . Current recordings were not corrected for leak. If not otherwise stated, they were filtered at 1 kHz and sampled at 5 kHz for off-line analysis. PClamp programs were employed for data acquisition and analysis. Data were analyzed after stable whole cell recordings were obtained, generally after about 2 min. Data are given as mean values Ϯ S.E.

RESULTS
Whole cell patch clamp recordings in RBL-1 cells reveal large inwardly rectifying K ϩ currents that are blocked by external Ba 2ϩ in the micromolar concentration range (Fig. 1A). With 4.4 mM potassium in the extracellular bath solution, current amplitudes at Ϫ140 mV were on average Ϫ786.5 Ϯ 61.4 pA (n ϭ 20). Since the IRK current is by far the dominating membrane current expressed in RBL-1 cells the small outward current component flowing through these channels is easily identified. Fig. 1B illustrates a typical steady state I-V relationship measured in RBL-1 cells. Outward currents are seen in a potential range slightly more positive than the reversal potential for potassium ions (from Ϫ70 to Ϫ30 mV; E K ϭ Ϫ87 mV) with a maximum at Ϫ55 mV. At more positive potentials the outward Ϫ70 and Ϫ15 mV are plotted in 5-mV increments. Note, cell was selected for a large outward "hump" (current density at Ϫ55 mV is 0.54 pA/pF) to illustrate that 50 M Ba 2ϩ are sufficient to block outward as well as inward currents through IRK channels. current becomes smaller, producing a negative slope conductance and a characteristic "hump" in the I-V relationship. This negative slope conductance observed in the I-V relationship is characteristic for outward currents flowing through "strong" inward rectifier channels (Fig. 1B). Similar IRK currents have been reported in RBL cells (22,23).
Polyamines Regulate Outward Currents through IRK Channels-DFMO and MDL73811 were used to manipulate intracellular polyamine concentrations in RBL-1 cells. Treatment of RBL-1 cells with 500 M DFMO for 2 days resulted in a decrease of the internal spermidine concentration by 85%, from 765 to 114 M ( Fig. 2C; Table II). At the same time the internal spermine content more than doubled, from 252 to 567 M, while putrescine concentrations remained fairly constant between 5 and 10 M. In contrast, treatment of RBL-1 cells with 50 M MDL73811 resulted in a 50 -100-fold increase of the intracellular putrescine content to about 500 M. This concentration change was accompanied by a prominent decrease of both spermidine and spermine concentrations from 765 and 252 M to 121 and 31 M (n ϭ 5), respectively (Table II and Fig. 2C; see also Table I).
The effects on the rectification properties of IRK currents were analyzed by studying I-V relationships in DFMO-and MDL73811-treated RBL-1 cells (Fig. 2, A and B). Treatment of RBL-1 cells with 500 M DFMO produced no obvious alterations in the I-V relationships even though the internal spermidine concentration had decreased ( Fig. 2A). Current amplitudes measured at Ϫ55 mV were Ϫ0.074 Ϯ 0.073 pA/pF in control (n ϭ 33) and Ϫ0.032 Ϯ 0.033 pA/pF in DFMO-treated cells (n ϭ 16). However, when RBL-1 cells were treated with 50 M MDL73811 to decrease both the internal spermine and spermidine concentrations much larger outward current amplitudes were observed between Ϫ70 and Ϫ15 mV ( Fig. 2A). Current amplitudes at Ϫ55 mV were 1.54 Ϯ 0.13 pA/pF (n ϭ 26; Fig. 2B) with slightly smaller inward currents. At test potentials of Ϫ140 mV the current density was Ϫ28.52 Ϯ 1.67 pA/pF for cells treated with 50 M MDL73811 (n ϭ 26). In control cells, Ϫ35.35 Ϯ 3.05 pA/pF (n ϭ 35) were measured ( Fig. 2A). A similar reduction of inward current amplitudes could be observed for all other groups treated with either DFMO or combinations of DFMO and MDL73811 (Table I) and might reflect a decrease in protein biosynthesis imposed by MDL73811 and DFMO on rapidly growing cells.
Passive electric properties such as input resistance or resting membrane potential were not different between control and MDL73811-or DFMO-treated cells. The resting membrane potential was Ϫ66.  Table II). The reported differences in cell size do not invalidate our conclusion that outward currents through IRK channels are increased under MDL73811 treatment, since all electrophysiological data are normalized with respect to cell capacitance and therefore cell size.
To define the pharmacological conditions that could give us maximal outward currents we also examined cells treated for 48 h with 500 M MDL73811 and cells treated with a combination of 50 M MDL73811, 500 M DFMO or 500 M MDL73811, 500 M DFMO. Data analyzed with respect to polyamine content and electrophysiological properties are summarized in Table I. Treatment of RBL-1 cells with 50 or 500 M MDL73811 proved to be most effective, resulting in a large increase in outward current through inward rectifier channels in response to a pronounced decrease of internal spermine concentrations (Table I).
To estimate the turnover in the intracellular polyamine pool the time course of the developing MDL73811 effect was examined (Fig. 3A). RBL-1 cells were treated for various times with 50 M MDL73811, and changes in outward currents were monitored at Ϫ55 mV. In 50 M MDL73811 outward currents reached a plateau after about 3 days. These data are contrasted with outward currents in cells cultured in the presence of 500 M DFMO, where outward currents fluctuate around base line (Fig. 3A). Although these experiments indicate that the maximal outward current was developed after about 3 days, most of the data presented in this paper were acquired from cells exposed for only 2 days for two reasons: (a) short exposure times were used to minimize the cytotoxicity reported for MDL73811, and (b) after 2 days cell densities were most suit-able for electrophysiological experiments in this fast growing cell line.
To exclude nonspecific drug effects the dose-response relationship was analyzed by treating RBL-1 cells for 48 h with different concentrations of MDL73811. Outward currents at Ϫ55 mV were increased by 50% with 54 nM MDL73811 (Fig.  3B). This is about 10 times lower than the K i value of about 600 nM found for MDL73811 and rat S-adenosylmethionine decarboxylase in an in vitro assay (27). Similar experiments were done to determine concentration-dependent effects of MDL73811 on intracellular polyamines. After exposing RBL-1 cells for 48 h to different MDL73811 concentrations, cells were harvested, and the polyamine content was analyzed by means of HPLC. Mean values of intracellular polyamine levels were plotted against MDL concentrations and fitted by Hill equations. The estimated IC 50 values were 62.0, 45.1, and 5.9 nM for putrescine, spermidine, and spermine, respectively (Fig. 4).
Intracellular Application of Spermine Decreases Outward Currents through IRK Channels-If the increase in outward currents under MDL73811 treatment is mainly due to a decline in intracellular spermine, then addition of spermine to the cytoplasm of spermine deprived RBL-1 cells should restore the outward current pattern of untreated control cells. Fig. 5A shows steady state I-V relations recorded immediately after establishing the whole cell configuration in RBL-1 cells treated for 2 days with MDL73811 (con MDL) and 37 min after perfusion with a pipette solution containing 100 M spermine. Dur-  ing the time course of the experiment with spermine added to the cytoplasm via the patch pipette, the outward current "hump" that could be blocked completely by 50 M Ba 2ϩ (n ϭ 5; inset to Fig. 5A) decreased gradually with time. Since the diffusion rate of spermine from the pipette reservoir into the cell is limiting, it is assumed that recordings made immediately after breaking into the cell (t ϭ 0) are similar to those made without spermine in the pipette solution (as shown in Fig. 3, A  and B). The steady increase in intracellular spermine concentrations under such recording conditions is reflected in gradually decreasing outward currents. Outward currents were stable when cells were perfused with standard pipette solution, whereas a time-and concentration-dependent decline in outward currents was observed by addition of spermine to MDL73811-treated cells (Fig. 5B). The observed decrease could be characterized by fitting monoexponential equations. Time constants were 34 min for perfusion of cells with 100 M exogenous spermine (n ϭ 3) and 98 s in experiments done with 1 mM spermine in the recording pipette (n ϭ 10; three independent experiments). When inward currents were monitored at Ϫ140 mV during the time course of such experiments, they proved to be stable as expected for application of a strongly voltage-dependent blocker such as spermine (Fig. 5B, lower panel).
Outward Currents through Inward Rectifier Channels Stabilize Membrane Potentials of RBL-1 Cells-Inward rectifier currents mediate the resting K ϩ conductance in RBL-1 cells.
Increasing amounts of outward current conducted by such channels should have a highly stabilizing effect on the resting membrane potential. This was demonstrated using whole cell recordings done under current clamp conditions immediately after gaining access to RBL-1 cells treated with 50 M MDL73811. Voltage responses were limited in amplitude to Ϫ40 mV for depolarizing current injections up to ϩ5 pA. On average, cells were depolarized by 20.36 Ϯ 1.28 mV (n ϭ 3) with Current amplitudes were normalized with respect to cell capacitances. Each data point represents mean Ϯ S.E. of six to eight cells. B, normalized outward current amplitudes at Ϫ55 mV plotted against MDL73811 concentrations. RBL-1 cells were exposed to MDL73811 for 2 days. Each data point represents mean Ϯ S.E. of 4 -10 cells. Solid line shows fit to data points with the following equation: y ϭ I max /(1 ϩ X/K d ) n , where X is the MDL concentration and n the Hill coefficient. 54 nM MDL73811 increased outward currents through IRK channels by 50%. FIG. 4. Concentration-dependent effects of MDL73811 on intracellular polyamine content. Polyamine concentrations were measured as described under "Experimental Procedures" and plotted against MDL73811 concentrations. Cells were exposed to MDL73811 for two days prior to experiments. Upper panel shows changes in putrescine content (PUT, n ϭ 3); middle panel shows spermidine content (SPD, n ϭ 4); lower panel shows spermine content (SPM, n ϭ 4). Each data point represents means Ϯ S.E. Solid line shows fit to data points with the following equation: y ϭI max /(1 ϩ X/K d ) n , where X is the MDL concentration and n the Hill coefficient. Note that putrescine concentrations increase, while spermidine and spermine concentrations decrease with increasing MDL73811 concentrations. a ϩ5 pA current injection (Fig. 6B). The same experiment was repeated a few minutes later after spermine included in the recording pipette had equilibrated with the cell interior. Subsequent current injections resulted in voltage responses of up to ϩ10 mV. On average cells were now depolarized by 63.7 Ϯ 2.45 mV (n ϭ 3) with a ϩ5 pA current injection, while voltage responses evoked by hyperpolarizing current injections were not significantly altered (Fig. 6D). The diffusion of spermine from the pipette into the cell interior was continuously monitored during such experiments with current ramps recorded under voltage clamp conditions (Fig. 6, A and C). Outward currents were gradually reduced by increasing internal spermine concentrations, and correspondingly, the cells became more sensitive to depolarizing stimuli as shown with current-clamp experiments. DISCUSSION This study indicates that the negative slope conductance of native IRK channels in RBL-1 cells is largely conferred by intracellular spermine. Our experiments strengthen the concept previously developed in experiments with heterologous expressed channel and receptor proteins that polyamines are widely used cytoplasmic gating molecules (3)(4)(5)(13)(14)(15)(16)(17).
Polyamine biosynthesis blockers were used to manipulate the polyamine content of RBL-1 cells. Treatment with DFMO led to a reduction in spermidine levels while the internal spermine content actually increased. Usually, a small reduction in spermine is reported upon treatment with DFMO, since spermidine stores are no longer replenished (see Ref. 18). We suspect that this difference is due to the short exposure time to DFMO (2 days) and the rather low concentrations used. On the other hand, this experimental design proved that an increase in spermine levels could compensate completely for the dramatic reduction observed in spermidine levels. I-V relationships measured both for DFMO treated and control cells were identical, although the combined cellular content for spermine and spermidine was reduced from 1020 to 680 M. A direct interference of DFMO with IRK channels is unlikely, since it has been shown that ornithine by itself has negligible effects on recombinant IRK1 channels (5).
In contrast, MDL73811 led to a decline of both spermidine and spermine and to the accumulation of large amounts of putrescine. While the decrease in spermidine levels was of the same magnitude as that observed with DFMO, spermine levels were increased by DFMO but decreased by MDL73811. Since significant increases in outward currents were observed only following MDL73811 treatment, we can conclude that intracellular spermine is the major determinant of outward currents in RBL-1 cells. The proposed higher efficiency of spermine in blocking native IRK channels is further supported by data on cloned IRK channels. The affinity of polyamines for IRK1 channels correlates strictly with the number of charges putrescine (ϩ2), spermidine (ϩ3), and spermine (ϩ4) carry (28).
The interaction of the various polyamines and Mg 2ϩ at their common binding site(s) in the channel pore is not very well defined. This has to be taken into account for all experiments done with MDL73811 in which the internal putrescine concentration is increased by about 100 times. The voltage-dependent block exerted by putrescine on IRKs, however, is rather shallow (4). A significant interaction of high internal putrescine concentrations with IRKs should be reflected in a reduction of IRK inward currents in a voltage range between Ϫ80 and Ϫ100 mV, where the block by spermine and spermidine is weak. Inward currents in MDL73811-treated cells, however, were not different from currents in DFMO-treated cells. For this reason the interaction of putrescine with the polyamine-Mg 2ϩ binding site must be rather weak, as predicted from experiments with cloned IRK channels where the affinity for putrescine was found to be much lower than that for the higher charged polyamines (3,4).
A significant problem with the use of enzyme inhibitors such as MDL73811 is the possibility that the observed changes in outward currents are due either to cytotoxic side effects (29) or to nonspecific drug interactions. A decreased cell viability should become evident in more depolarized resting membrane potentials of MDL73811-treated cells. There were no signs of cytotoxicity as measured by analysis of resting membrane potentials. In general, MDL73811 is considered a very effective and specific inhibitor of polyamine biosynthesis (21). Three arguments are in favor of a specific interaction of MDL73811 with its enzyme target, S-adenosylmethionine decarboxylase, under our experimental conditions: (a) the MDL73811 effect developed slowly with time as expected for manipulations of metabolic turnover rates; (b) 54 nM MDL73811 were sufficient to increase outward currents by 50% (this sensitivity is even higher than the reported "in vitro" K i of MDL73811 for S-adenosylmethionine decarboxylase (600 nM)); and (c) exogenous spermine could be used to replace authentic intracellular polyamines depleted by the pharmacological regimen.
Moreover, the reported data show clearly that pharmacologically induced changes in polyamine metabolism are directly translated into physiological response schemes. In RBL-1 cells, IRKs are mainly used to clamp the membrane potential at negative values to maintain a large driving force for Ca 2ϩ influx, essential for stimulus-secretion coupling in nonexcitable RBL-1 cells (30). Low concentrations of intracellular spermine allow for large outward currents, thereby stabilizing negative membrane potentials. High internal spermine concentrations, in contrast, destabilize RBL-1 cells and lower the threshold for membrane depolarizations. Any spermine concentration change is therefore directly translated into an altered cell excitability.
FIG. 6. Outward currents through IRK channels stabilize membrane potential in RBL-1 cells. A, I-V relationship determined under voltage clamp conditions in response to voltage ramps of 150 ms (from Ϫ120 to ϩ60 mV; holding potential, Ϫ60 mV; dV ϭ 1.2 mV/ms) in a cell treated for 2 days with 50 M MDL73811. The voltage ramp shown was recorded immediately after getting access to the cell. B, membrane voltage versus current injected into cell shown in panel A under current clamp conditions (V-I relationship). V-I relationship was determined in response to current injections between Ϫ25 and ϩ5 pA starting from resting membrane potential. C, I-V relationship determined under voltage clamp conditions in response to voltage ramps of 150 ms (from Ϫ120 to ϩ60 mV; holding potential, Ϫ60 mV; dV ϭ 1.2 mV/ms) for the cell shown in panel A 4 min after the start of intracellular perfusion with 1 mM spermine in the recording pipette. Note block of outward current "hump." D, V-I relationship determined under current clamp conditions for cell in panel C. Note the large changes of membrane potential following on positive current injections.