Abstract The trivalent lanthanide ion gadolinium (Gd*) has been shown to block various ion channels, particularly in mechanoreceptor systems, in which is has been used in a diagnostic manner. Gd' has also been shown to block voltage-gated channels in a number of systems. particularly L-, N-, and T-type Ca channels. The cells of the giant fiber lobe (GFL) of the stellate ganglion of squid are useful in a variety of electrophysiological studies, and their currents are well characterized, although the effects of Gd" on these currents have not been studied. We report a lack of effect of 20 uM Gd" on K, Na, and Ca currents in these cells, either in voltage- dependent properties or kinetics, although the Ca current (proposed to be due to a P-type Ca channel) warrants further investigation to unequivocally identify it as a Ca current and look for more subtle effects of Gd". This study adds to a long list of channels on which the effects of Gd have been studied, and provides some preliminary information on the molecular structure of the pore of P-type Ca channels, as they appear to be insensitive to Gd" block at first approximation Introduction Gadolinium The lanthanide element, gadolinium (Gd“) has been studied extensively in biological systems in its trivalent ionic form. Gd' is in the "middle" of the lanthanides, with atomic number 64, and an ionic radius of 0.938 A, which is very similar in size to the 0.99 A atomic radius of Ca“ (Hamill and McBride, 1996). Gd" ions are also similar to Ca ions in their bonding, coordination, and donor atom preference (Hamill and McBride, 1996). The first suggestion of the biological significance of Gd" came with a study by Miller and Pickard in 1988, which showed block of thigmotropism and geotropism in plants by 10-250 uM Gd' (Hamill and McBride, 1996). This effect was hypothesized to be due to block of Ca-permeable mechanogated (MG) ion channels, and Gd has subsequently been shown to blocks single MG channel currents in patch clamp studies in plants, bacteria, fungi, and animal cells (Hamill and McBride, 1996). Gd' is often used as a diagnostic tool to determine whether MG channels are involved in particular processes, although it does not block all MG channels, and has been found to block other types of channels, removing its status as a specific MG channel blocker (Hamill and McBride, 1996; Caldwell, et. al., 1998). The mechanism of block has not been established absolutely, but Gd' is thought to be an open-channel, permeant blocker of MG channels (Hamill and McBride, 1996). Its effects are often seen at a range of concentrations falling between 100 nM and luM, with complex, non-linear dose-response relationships in some systems (Tokimasa and North, 1996; Hamill and McBride, 1996). The list of non-MG channel types blocked by Gd" is growing, and includes studies performed in a wide variety of cell types, including smooth and cardiac muscle, neurons, endocrine and epithelial cells in myriad species. Most blocked channels, not surprisingly, are Ca-specific channels, suggesting that Gd" mimics Ca and binds to the pore of these channels, a mechanism that has been suggested in a number of studies (Lansman, 1990; Biagi and Enveart, 1990; Hamill and McBride, 1996). Transient receptor potential (TRP) channels were found to be blocked by 1 uM Gd" in epithelial cells from rabbit (Vennekens, et al, 2001), and dihydropyridine-sensitive Ca currents were blocked by 50 uM Gd" in mouse skeletal muscle (Lansman, 1990). L-, N-, and T-type voltage-gated Ca channel currents were blocked by Gd3 in à number of studies as well, with required concentrations of Gd ranging from 200 nM to 2.5 MM (Docherty, 1988; Mlinar, and Enyeart, 1993; Biagi and Enyeart, 1990; LaCampagne, et. al., 1994). Initial observation of the effects of Gd" on Ca current in squid GFL cells appears to show a significant block (Gilly, unpublished observations). These blocks were sometimes completely reversible (over a time course of minutes), reversible in a voltage-dependent manner. or completely irreversible. There are few studies of the effects of Gd on currents other than through Ca-specific channels, but the few there are suggest significant effects. Ca-activated Cl-currents are blocked by Gd, not surprisingly (Tokimasa and North, 1996), and a number of non-specific cation currents have been shown to disappear in the presence of Gd" as well (Koh, et. al, 2001; Zhang and Hancox, 2003; Zhang, et. al., 1998). Two studies, one in Xenopus myelinated axons and one in cardiac muscle show that some types of K and Na currents are also blocked by Gd" (Elinder and Arhem, 1994; Li and Baumgarten, 2001). The diversity of channels shown to be sensitive to block by Gd" suggests that further investigation of the effects of Gd" on various ion channels. especially Ca-specific channels, is valuable. Squid Stellate Ganglion Giant Fiber Lobe The giant fiber lobe (GFL) of the stellate ganglion of squid is a well-studied system, as it contains the cells that give rise to the squid giant axon, the first system in which voltage-clamp studies were possible. The cell bodies of GFL were first reported to be inexcitable in vivo by Miledi (1967), but later found to produce observable currents when cultured. Llano and Bookman (1986) described the currents of GFL cell bodies as dominated by a slow (reaching steady-state in 10-15ms), outward current with threshold of about -30 to -20mV, which inactivated partially during a long depolarizing pulse (250ms). K ions were identified as the main carrier of this current because of its K-dependence, sensitivity to TEA, insensitivity to TIX and Cd, and reversal potential, and kinetic analysis led to the conclusion that the conductance represented a single type of K channel (Llano and Bookman, 1986). These properties of this channel are identical to those of the channel responsible for K current in giant axon, and a putative molecular clone of the channel has been isolated by Rosenthal and Gilly (2003). The calcium current in GFL cell bodies, one component of the inward current observed in recordings in the absence of K, is slow, TTX-insensitive, blocked by Zn“, 10 mM CSCl, and 1-5 mM Cd“, and does not show significant inactivation (Llano and Bookman, 1986; Brismar and Gilly, 1987; Gilly and Brismar, 1989). This channel has been characterized as P-type, based on pharmacological properties, and a non-L-type candidate clone has been identified, but has not yet been directly compared to the identified P-type current (Rosenthal and Gilly, 2003). The Na currents in GFL cell bodies are complex in their own right. As stated above, GFL cell bodies are inexcitable in vivo, but upon culturing, significant Na current appears, usually after 4-6 days, and can be maintained for up to two weeks (Gilly and Brismar, 1989; Brismar and Gilly, 1987; Gilly, et. al., 1990). This current is thought to be the result of inappropriate trafficking of Na channels normally sent to axons, but now expressed in somatic membranes, and a number of studies have attempted to elucidate the processes necessary for this trafficking (Gilly and Brismar, 1989; Brismar and Gilly, 1987; Gilly, et. al., 1990). The properties of the somatic Na current are almost identical to those of axonic Na current (with only very few subtle kinetic differences, probably due to post-translational modifications)—it is fast- activating and deactivating, TTX- and STX-sensitive (at 50-200 nM concentrations), has characteristic voltage-dependence, threshold (-40 mV), reversal potential (which demonstrates high Na selectivity), and is partially blocked by elevated levels of external Ca (50 mM) (Gilly and Brismar, 1989; Brismar and Gilly, 1987; Gilly, et. al., 1990; Llano and Bookman, 1986). The majority of this current inactivates fairly quickly, although a small (and variable) non- inactivating component has been identified that is otherwise identical (Gilly and Brismar, 1989). Two molecular clone candidates have been isolated for this Na channel (Rosenthal and Gilly. 2003). The present study examines the effects of 20 uM Gd' on the currents of squid GFL cell bodies. The previously described Na, K, and Ca currents were present, with no observed deviation from published descriptions. No effects of 20 uM Gd" were seen on any of the three current types, either on voltage-dependent properties or kinetics. Further study of the Ca current is necessary to firmly establish its Gd' insensitivity, and identify it as a P-type voltage¬ dependent Ca channel. This result would contribute to the discrimination of various Ca channel types, as L, N, and T-type Ca channels have all shown sensitivity to Gd' in other studies (including preliminary studies of squid GFL Ca current conducted in this laboratory), and inferences about the structures of the pores of these channels could be made based on their low affinity for Gd' ions. Materials and Methods Cell Culture Neurons from the GFL of Loligo opalescens from Monterey Bay were removed and cultured following a previously published protocol (Gilly, et. al., 1990, see Brismar and Gilly. 1987 and Llano and Bookman, 1986), with minor changes. GFL tips were treated with 3mg/mL non-specific protease in seawater for 1 hour. Glass coverslips (l em in diameter) were coated with 5% concanavalin A (Con A) in distilled water, were placed in plastic culture dishes, and allowed to dry. After approximately 90 minutes, the coverslips were washed with 2 mL distilled water, and allowed to dry under UV light for at least one hour. After cells were plated as previously described, they were placed in a culture medium (Leibovitz's L-15; GIBCO. Inc.) with added 263 mM Nacl, 4.64 mM KCl, 4.25 mM CaCh, 49.54 mM MgCh, 2 mM HEPES (pH 7.8), 5 mM trehalose, 50 ug/mL penicillin G, and 0.5 mgmL streptomycin. Cells were kept in an incubator maintained at 12-14°C, and medium was changed every 2 days. Recordings of K current were made within 1-3 days of plating, and recordings of inward current were made 4-12 days after plating. Recordings were made at a maintained temperature of approximately 140C. Solutions Recording solutions are listed in Table 1. For experiments measuring K current, the pipette was filled with the K+ internal solution, and the experimental bath was Na-free ASW. For experiments measuring Na current, the pipette contained the Na- internal solution, and the bath was ASW, with added 200 or 1000 nM TIX where noted. Gadolinium experiments were conducted by perfusing the recording chamber with external solutions to which 20 uM GdCla had been added from a stock solution (ImM GdClz in distilled water). The solubility of Gd' has been demonstrated in a variety of other studies (Caldwell, et. al., 1998). Voltage-Clamp Recordings Whole-cell, voltage-clamp recordings were performed using standard procedures (detailed in Gilly and Brismar, 1989, see Llano and Bookman, 1986, and Brismar and Gilly, 1987) with a maintenance temperature of 10eC. Sylgard-coated patch pipettes with resistances. when filled, of 1-4MO were used. Cells were chosen based on the absence of axonal processes or stumps, and recordings were deemed acceptable primarily based on a seal of at least 100 MO Series resistance was electronically compensated (by eye). VClamp software was used for pulse generation and data acquisition and storage. Calculations of K+ conductance were based on an isochronal IV curve generated from the amplitudes of slow inward tail currents at a range of voltages after a 20 ms activating pulse, as described previously (Gilly, et. al, 1990). Conductance values were normalized to the maximum conductance values of each cell to give relative G/V curves. Results Kcurrent A large, outward current was observed in recordings with 150 mM internal K, and 10 mM external K. Based on the appearance, kinetic properties, and reversal potential (see Fig. 1). this current was assumed to be carried predominantly by K+ ions, as previously described in GFL preparations (Llano and Bookman, 1986). Addition of 20 uM Gd" to the external bathing solution showed no effect on the voltage-dependence of this current (Fig. 2). In addition, the conductance and its voltage-dependence (Fig. 3) did not change significantly with the application of 20 uM Gd“. Based on the lack of significant effects of Gd' on UV and G/V relationships as well as visual inspection of current traces, it is also assumed that Gdt has no effect on the kinetic properties of K channels. Inward Current As detailed in previous investigations (Brismar and Gilly, 1987, Gilly and Brismar, 1989 Gilly, et. al, 1990), the cell bodies of GFL neurons do not show significant inward current in vivo. However, after 4-6 days in culture, an inward current develops in these cells that is smaller than those found in giant axon, but is identical in its properties (Gilly and Brismar, 1989. Fig. 4) This current has two components, the first, identified as Na current by its reversal potential and 1IX-sensitivity, is hypothesized to arise from the inappropriate distribution of Na channels to cell bodies rather than axons (Gilly, et. al., 1990). Due to the presence of Na inside (220 mM) as well as outside (480 mM) the cell, the current observed is both inward at voltages more negative than the reversal potential, and outward at voltages more positive than the reversal potential. The characteristic fast kinetics of activation and inactivation are present. The majority of this current inactivates, although there is some non-inactivating current present, as seen by Gilly and Brismar (1989) A slower, non-inactivating component, can also be distinguished by its insensitivity to TIX (Fig. 4). This TIX-insensitive current is most likely carried by Ca through Ca channels based on standard criteria: insensitivity to TIX, slower kinetics (time constant of activation on the order of 5 ms, at first approximation, which is much slower than typical Na current), lack of inactivation, and comparison to currents observed in GFL cell bodies by Llano and Bookman (1986) and others (Gilly and Brismar, 1989; Brismar and Gilly, 1987). For purposes of experimental design, TTX was not applied in all cases, but instead, the peak and steady-state currents were measured separately as estimates of the two components. The voltage-dependence of both currents appears to be insensitive to 20 uM Gd', as the LV curves appear almost identical (Fig. 5). The conductance-voltage relationship also appears unaffected by 20 uM Gd for both types of current (Fig. 6). The lack of change in the UV and G/V relationships due to Gd, as well as visual inspection of current traces, suggests that there is no effect of Gd' on the kinetics of either current. Further investigation of the slower inward current is necessary to unequivocally identify it as carried by Ca, as it reversed at a potential similar to that of the TIX¬ sensitive Na current, suggesting it is carried at least in part by Na. The effects of Gd' on this slower inward current should also be further investigated, due to the conflicting nature of the data presented here with previous observations from this laboratory (Gilly, unpublished observations). Discussion The voltage-dependent ionic conductances in GFL neurons of squid are well characterized, and thus, the results of this study can be compared to an extensive amount of previously published data. Three main types of current were observed, and were identified as Na, K, and Ca currents based on kinetic and pharmacological properties. The K current. observed at day zero in all cells in recording solutions with 10 mM external and 150 mM internal K (see Fig. 1), is a large, slowly activating, outward current, which inactivates only partially during the length of pulses used in this experiment, and is identical to previous descriptions of K current in GFL cells (Llano and Bookman, 1986). A molecular candidate for the channel responsible for this current has been identified by Rosenthal and Gilly (2003) as a single type of K channel, identical to the one found in giant axon. This K current showed no effect of 20 uM Gd' on voltage-dependence of current or conductance (see Fig. 2 and 3), or on reversal potential. Although kinetic properties were not extensively studied, application of external Gde was not observed to produce a significant effect on the time courses of activation or deactivation. Two inward currents were also observed in GFL cell bodies starting 4-5 days after culturing, with recording solutions containing 480 uM Na and 10 uM Ca in the external solution and 220 uM Na in the internal solution. The first of these currents, which has fast kinetics. inactivates almost completely, and is TTX-sensitive, was identified based on these criteria as the inappropriately-expressed Na current described in detail by Brismar and Gilly (1987; see also Gilly and Brismar, 1989). Previous analysis has shown that this Na current is likely produced by à single type of Na channel (with inactivating and non-inactivating subtypes) also found in giant axon, and two molecular clones have been identified as candidates (Rosenthal and Gilly. 2003). This current did not appear to be affected by application of 20 uM Gd' externally, in its voltage¬ dependent properties, reversal potential, or kinetics. The second component of the inward current, which is slower, non-inactivating, and TIX-insensitive, was not definitively identified based on the data presented here. It likely has a component of Ca-carried current, based on its TTX-insensitivity. A similar current has been described (Llano and Bookman, 1986; Brismar and Gilly, 1987; and Gilly and Brismar, 1989) and identified as Ca current based on similar criteria and additional data showing block by Cde Gilly and Brismar (1989) also found that this current did not reverse, adding further support to its identification as Ca current. The current did reverse in the experiments conducted in this study. suggesting that it may not be pure Ca current, but may also contain a Na component (carried through non-inactivating Na channels), based on its reversal potential. Only one candidate Ca channel has been identified, described conclusively as a non-L-type channel, and possibly the same P-type channel found in giant axon (Rosenthal and Gilly, 2003). This current was unaffected by external application of 20 uM Gd" in both voltage-dependent properties and reversal potential. Its kinetics were not extensively studied due a lack of experiments conducted with application of TTX. Further study of the effects of Gd' on this current are necessary, as the results presented here are contradictory to prior observations of Ca channel block by Gd' in squid GFL (Gilly, unpublished observations). The results of this study add to a large pool of data on the effects of Gd' on various types of mechanosensitive, ligand- and voltage-gated channels. Concentrations as low as 100 nM Gd are enough to block some channel types, but others remain active even in the presence of 100 uM Gd (Tokimasa and North, 1996; Hamill and McBride, 1996). The effects of Gd' are also complex in many cases, showing non-linear dose-response curves, and seemingly opposite effects over broad concentration ranges (Hamill and McBride, 1996). Due to its consistent block of mechanosensitive channels, and the lack of studies showing effects on other channel types, Gd' was seen as a specific blocker of stretch-activated and inactivated channels (SA and SI channels), although further studies showed Gd" to block other channel types, as well as examples of SA channels not blocked by Gd", urging caution in the use of Gd' as diagnostic of mechanosensitive channels (Hamill and McBride, 1996). Other channel types found to be blocked by Gd include various types of Ca channels, including transient receptor potential (TRP) channels (Vennekens, et. al, 2001), dihydropyridine¬ sensitive Ca currents (Lansman, 1990), and voltage-gated channels of N-, L-, and T-type (Docherty, 1988; Mlinar, and Enyeart, 1993; Biagi and Enyeart, 1990; LaCampagne, et, al., 1994) in diverse systems including smooth and cardiac muscle, neurons, epithelial and endocrine cells. Gd" has also been shown to block Ca-activated Cl currents (Tokimasa and North. 1996). à number of non-specific cation currents (Koh, et. al, 2001; Zhang and Hancox, 2003; Zhang, et. al., 1998), and a few types of K and Na currents in myelinated axons and cardiac muscle (Elinder and Arhem, 1994; Li and Baumgarten, 2001). These studies suggest that it is valuable to investigate the effects of Gd" on diverse channel types, especially Ca channels. The present study does not show any effects of Gd' on K, Na, or Ca currents, however. The Ca channel present in GFL neurons is likely to be P-type (Rosenthal and Gilly, 2003), a subtype on which the effects of Gd" have not been published. Observations in this laboratory, however, suggest that Gd“ is an effective blocker of Ca current in GFL cell bodies (Gilly, unpublished observations). There is a similar lack of precedent for the block of Gd" on voltage-gated Na and K channels in neuron cell bodies or unmyelinated axons. The effects of Gd' on the channels present in GFL neurons should be investigated further. Various technical concerns exist in studies using Gd", including concerns of effective concentration in the presence of various chelating anions such as phosphate, carbonate, and EGTA. The current study avoids these specific compounds (with the exception of sulfate present in the ASW used as an external solution for recordings of K current), but it is worthwhile to keep in mind the possibility that the effective concentration of Gd" in experimental solutions might be well below expected. An interesting side note is that these anions, which are very physiologically important, may prevent Gd" from having significant effects in vivo, preventing harmful consequences of using Gd" in brain imaging studies. Gd' still has value as a pharmacological agent in experimentation, as it can be used to separate various current types in a variety of otherwise confusing situations. More detailed investigation of the effects of Gd' on the Ca current in GFL neurons would also be useful as previous studies have shown that Gde block may require significant time, or may occur only after several depolarizations, both of which would have caused an effect to be missed in the present study (Biagi and Enyeart, 1990; LaCampagne, 1994). Although most blocking effects of Gd' are seen at doses lower than 20 HM Gd", it is possible that these currents would be blocked by higher concentrations, another avenue of further investigation. If it is true that the Ca current found in GFL neuron cell bodies is carried through a P¬ type channel, this study may provide insight into the structure of such channels as compared to L-, N-, and T-type channels. The mechanism of Gd" block of such channels has been proposed to be open-channel block by Gd" binding to a site inside the channel (Lansman, 1990: Biagi and Enyeart, 1990; Hamill and McBride, 1996). Therefore, the lack of a blocking effect seen in P- type channels would suggest a significant difference in the structure of the pore of these channels as compared to L, N-, and T-type channels, whereas effective block would suggest a similarity in structures amongst the Ca channel subtypes. Further study is necessary to unequivocally describe the mechanism of action of Gd", however, before such structural inferences will be reliable. Conclusions The effects of Gd" ions on the currents in squid GFL neuron cell bodies were examined. The currents, identified as K, Na, and (most likely) P-type Ca currents, were unaffected by 20 AM Gd" in their voltage-dependent properties, reversal potentials, and kinetics. This study adds to à growing literature on the effects of Gd' on diverse channel types, and provides an avenue of further investigation of the effects of Gd" on P-type Ca channels. Acknowledgements I wish to thank Dr. Stuart Thompson and Dr. Bill Gilly for their support, instruction, and guidance, and Katrina Easton for all her help and patience. I would also like to thank Alex Norton for his help with the squid. Literature Cited: Biagi, B.A. and Enyeart, J.J. (1990). Gadolinium blocks low- and high-threshold calcium currents in pituitary cells. Am. J. Physiol. 259, C515-520. Brismar, T., and Gilly, W.F. (1987). Synthesis of sodium channels in the cell bodies of squid giant axons. Proc. Natl. Acad. Sci. USA 84, 1459-1463. Caldwell, R.A., Clemo, H. F., and Baumgarten, C.M. (1998). Using gadolinium to identify stretch-activated channels: technical considerations. Am. J. Physiol. 275, C619-621. Docherty, R.J. (1988). Gadolinium selectively blocks a component of calcium current in rodent neuroblastoma x glioma hybrid (NG108-15) cells. J. Physiol, 398, 33-47. Elinder, F. and Arhem, P. (1994). Effects of gadolinium on ion channels in the myelinated axon of Xenopus laevis: four sites of action. Biophys J. 67, 71-83. Gilly, W.F., and Brismar, T. (1989). Properties of appropriately and inappropriately expressed sodium channels in squid giant axon and its somata. J. Neurosci. 9, 1362-1374. Gilly, W.F., Lucero, M. T., and Horrigan, F. T. (1990). Control of the spatial distribution of sodium channels in giant fiber lobe neurons of the squid. Neuron 5, 663-674. Hamill, O. P., and McBride, D. W. (1996). The pharacology of mechanogated membrane ion channels. Pharmacological Reviews 48(2), 231-252. Lacampagne, A., et. al. (1994). The stretch-activated ion channel blocker gadolinium also blocks L-type calcium channels in isolated ventricular myocytes of the guinea-pig. Biochim. Et Biophys. Acta 1191, 205-208. Lansman, J.B. (1990). Blockade of current through single calcium channels by trivalent lanthanide cations. J. Gen. Physiol. 95, 679-696. Llano, I., and Bookman, R. J. (1986). lonic conductances of squid giant fiber lobe neurons. J. Gen. Physiol. 88, 543-569. Li. G. R., and Baumgarten, C. M. (2001). Modulation of cardiac Na+ current by gadolinium, a blocker of stretch-induced arrhythmias. Am. J. Physiol. Heart Circ. Physiol. 280, H272-H279. Miledi, R. (1967). Spontaneous synaptic potentials and quantal release of transmitter in the stellate ganglion of the squid. J. Physiol. 192, 379-406. Mlinar, B. and Enyeart, J.J. (1993). Block of current through T-type calcium channels by trivalent metal cations and nickel in neural rat and human cells. J. Physiol. 469, 639-652. Rosenthal, J.J.C., and Gilly, W.F. (2003). Identified ion channels in the squid nervous system. Neurosignals 12, 126-141. Song, J.B., Hood, J.D., and Davis, M.J. Gadolinium blocks calcium channels in coronary artery smooth muscle cells. AS15. Tokimasa, T. and North, R.A. (1996). Effects of barium, lanthanum and gadolinium on endogenous chloride and potassium currents in Xenopus oocytes. J. Physiol 496(3), 677-686. Vennekens, R., et. al. (2001). Pore properties and ionic block of the rabbit epithelial calcium channel expressed in HEK 293 cells. J. Physiol 530(2), 183-191. Zhang, Y.H. and Hancox, J.C. (2003). A novel, voltage-dependent nonselective cation current activated by insulin in guinea pig isolate ventricular myocytes. Circ. Res. 92, 765-768. Zhang, Y., McBride, D.W. Jr., and Hamill, O.P. (1998). The ion selectivity of a membrane conductance inactivated by extracellular calcium in Xenopus oocytes. J. Physiol. 508(3), 763- 776 Table 1. Recording solutions used in whole-cell voltage-clamp experiments.. Solution Compound Concentration K+Internal KCI 20 mM K glutamate 80 mM KE 50 mM Lysine (pH 7 with HEPES) 10 mM EGTA 1mM EDTA 1 mM Glycine 381 mM Sucrose 291 mM MgATP 4 mM TMAOH 5 mM AS KCI 10 mM Nacl 480 mM Cach 10 mM Mgcl, 10 mM MgSO. 10 mM HEPES 10 mM pH Nat Internal Na-glutamate 100 mM tetramethalammonium glutamate 240 mM NaF 50 mM Nacl 50 mM NazEGTA 10 mM tetraethylammonium chloride 25 mM HEPES 10 mM pH 78 K-free ASW Nacl 480 mM Cacl 10 mM MgCl 20 mM Mgsoa 20 mM HEPES 10 mM pH 80 Fig. 1. Inset: pulse pattern used to generate current families shown in a) and b). a) Current family generated by a series of positive voltage steps from a holding potential of -80 mV. The cell contained 150 mM K internal and 10 mM K external. b) Current family generated as in a) with the addition of 20 uM Gd to the external solution. Fig. 2. Voltage-dependence of currents in cells containing 150 mM internal and 10 mM external K, with and without the addition of 20 uM Gd to the external solution. There is no significant difference amongst the conditions. Fig. 3. Voltage-dependence of conductance in cells containing 150 mM internal and 10 mM external K, with and without the addition of 20 uM Gd' to the external solution. Curves were normalized by dividing all values by the maximum conductance of each cell. There is no significant difference amongst the conditions. Fig. 4. Inset: pulse pattern used to generate current families shown in a) and b). a) Current family generated by a series of positive voltage steps from a holding potential of -80 mV. The cell contained 220 mM Na and O Ca internal and 480 mM Na and 10 mM Ca external, b) Current family generated as in a) with the addition of 20 uM Gd to the external solution. Fig. 5. Voltage-dependence of currents in cells containing 220 mM Na and 0 Ca internal and 480 mM Na and 10 mM Ca external, with and without the addition of 20 uM Gd to the external solution. There is no significant difference amongst the conditions. Fig. 6. Voltage-dependence of conductance in cells containing 220 mM Na and 0 Ca internal and 480 mM Na and 10 mM Ca external, with and without the addition of 20 uM Gd to the external solution. Curves were normalized by dividing all values by the maximum conductance of each cell. There is no significant difference amongst the conditions. Fig. 1 70 mV -80 mV n p pnn 6000 fnntenten smn 4000 gp mspnntsennng k 2000- gnenn mmnmnmnen wssentmseturstutemttt o e gn 0 25 30 35ms Time (ms) 8000- 6o00- nn 4000- ntdnnen 2000- gnmgthndndmemntenn gpe snpme pansmnhand mgunnmtandmetantminspa O 10 15 20 25 30 35ms Time (ms) Fig. 2 Current (pA) P 9 8 23 Fig. 3 c 8 Relative Conductance (G/Gmax) . —+ 8- ++ ———— —1— + I Fig. 4 70 mV 1000 f e t e peteg r etemd emnvvantn -80 mV e efg pse o men en e ne nr md eee een eren 100 We 2000- o00 10 15 30 35me Time (ms) 1000 Frnrteeen M O (svapenaenemsgedn estmmten en se dpant eee -1000 smmngstre gna wnpn grn mnee -2000- 30 10 15 25 30 35ms 20 Time (ms) 1000 Pehen ene n henenenen een aet n Ge -1000- -2000 300 15 20 25 30 35ms 5 10 Time (ms) Fig. 5 500 — Before Gd With 20 microM Gd -500 Wash off Gd 40 -20 Voltage (mV) 1000- 500 Before Gd With 20 microM Gd -500 - Wash off Gd -1000 -40 Voltage (mV) Fig. 6 5 12 Before Gd C 10 With 20 microM Gd Wash off Gd 2 08- 0.6- 0.2- 50O -------------------------.............................. P -60 -40 -20 40 60 Voltage (mV) § 0.8 Before Gd with 20 microM Gd Wash off Gd 0.6- 0.4 0.2: O.O -60 -40 -20 20 40 60 Voltage (mV)