Heteromeric channel formation and Ca(2+)-free media reduce the toxic effect of the weaver Kir 3.2 allele.

Weaver mice have a severe hypoplasia of the cerebellum with an almost complete loss of the midline granule cells. Recent genetic studies of weaver mice have identified a mutation resulting in an amino acid substitution (G156S) in the pore of the inwardly rectifying potassium channel subunit Kir 3.2. When expressed in Xenopus oocytes the weaver mutation alters channel selectivity from a potassium-selective to a nonspecific cation-selective pore. In this study we confirm by cell-attached patch-clamp recording that the mutation produces a non-selective cation channel. We also demonstrate that the cell death induced by weaver expression may be prevented by elimination of calcium from the extracellular solution as well as by coexpression with the wild-type Kir 3.2 allele, or other members of the Kir 3.0 subfamily. These results suggest that the weaver defect in Kir 3.2 may cause cerebellar cell death by cell swelling and calcium overload. Cells which express the weaver subunit, but which normally survive, may do so because of heteromeric subunit assembly with wild-type subunits of the Kir 3.0 subfamily.

Anomalous mole-fraction effects in recombinant and native cyclic nucleotide-gated channels in rat olfactory receptor neurons.

Anomalous mole-fraction effects (AMFE) were studied, using the inside-out configuration of the patchclamp technique, in both recombinant wild-type alpha-homomeric rat olfactory adenosine 3',5'-cyclic monophosphate (cAMP)-gated channels (rOCNC1) expressed in human embryonic kidney cells (HEK 293) and native cyclic nucleotide-gated (CNG) channels in acutely isolated rat olfactory receptor neurons. Single-channel and macroscopic currents were activated by 200 microM and 500 microM cAMP, respectively. Macroscopic currents, measured with mixtures of Na(+)-NH(4)(+) or Cs(+)-Li(+) in the cytoplasmic bathing solution, displayed AMFE in the rOCNC1 channels at both positive and negative membrane potentials. The rOCNC1 single-channel conductance showed a distinct minimum (or maximum) in an 80% Na(+)-20% NH(4)(+) mixture (or a 60% Cs(+)-40% Li(+) mixture), but only at positive membrane potentials. Macroscopic measurements in native olfactory CNG channels with mixtures of Na(+)-NH(4)(+) indicated similar AMFE. These results suggest that both native CNG channels and recombinant alpha-homomeric channels allow several ions to be present simultaneously within the channel pore. They also further validate the dominant role of the alpha-subunit in permeation through these channels, provide the first evidence to suggest that rOCNC1 channels have multi-ion properties and further justify the use of the rOCNC1 channel as an effective model for structure-function studies of ion permeation and selectivity in olfactory CNG channels.

Measurement of the limiting equivalent conductivities and mobilities of the most prevalent ionic species of EGTA (EGTA2- and EGTA3-) for use in electrophysiological experiments.

In many experimental biological situations, chelating agents like EGTA (ethylene glycol-bis-(beta-amino-ethyl ether) N,N,N',N'-tetra-acetic acid) are commonly used to control or suppress the concentration of divalent ions like Ca2+. The evaluation of liquid junction potentials in electrophysiological measurements, and particularly in patch-clamp situations, requires information about the ions within the solution. Where there is a significant concentration of EGTA present, it is necessary to know the values of the relative mobility of at least the most predominant ionic species of EGTA in order to complete these calculations. EGTA, with four negative charges with different pKas, can therefore exist as four differently charged ions in solution (EGTA-, EGTA2-, EGTA3- and EGTA4-) or as uncharged, although between pH 5.5 and 8 it is almost exclusively EGTA2-. We have measured limiting equivalent conductivities of the most common ionic forms of EGTA (EGTA2- and EGTA3-) encountered at physiological pHs. These were 35.9 +/- 0.7 and 56 +/- 2.5 S cm2 equiv(-1) respectively. Their mobilities relative to K+ were 0.24 +/- 0.01 for EGTA2- and 0.25 +/- 0.01 for EGTA3-. Thus for typical electrophysiological solutions, the contribution of EGTA to the liquid junction potential should be small (e.g. approximately 0.4 mV).

Glycine receptors: what gets in and why?

1. The glycine receptor channel (GlyR), a member of the ligand-gated ion channel superfamily, shares many similar permeation properties with the GABAA receptor channel. 2. The GlyR is anion permeable, with PK/PCl < 0.05, has a 5-6 A minimum pore diameter and a permeation selectivity sequence dominated by hydration energies. 3. The channels, which display multiple subconductance states, can be multiply occupied. 4. Two positive arginine rings at the ends of the pore region may contribute to the anion selectivity of the GlyR. 5. Mutation of the extracellular charged arginine ring can impair channel function by decreasing the sensitivity of glycine activation, reducing channel conductance, shifting the normal multi-subconductance states to lower values and by decoupling the link between ligand binding and channel gating. 6. These and other site-directed mutagenesis studies of recombinant GlyR, together with studies of native GlyR, are providing further insights into what controls gating and ion permeation and selectivity through this channel.

Very negative potential for half-inactivation of, and effects of anions on, voltage-dependent sodium currents in acutely isolated rat olfactory receptor neurons.

Previous measurements with CsF pipette solutions using whole-cell patch-clamp techniques in dissociated rat olfactory receptor neurons (ORNs) indicated that the sodium currents had very negative inactivation characteristics with the implication that the cell resting potential must also normally have a very negative value. This study supports the conclusions that such an effect was real and not dependent on either the nature of the pipette anions or the recording situation previously used. For all pipette solutions, sodium currents showed a threshold activation approximately -80 mV and half-maximal activation voltages approximately -55 with half-inactivation potential < or =-100 mV, without being significantly affected by the replacement of F(-) by other pipette anions (H(2)PO(-)(4) and acetate(-)) or the addition of nucleotides and glutathione (which did cause a very slight positive shift). F(-), followed by H(2)PO(-)(4) and to a much lesser extent by acetate(-), was the most favorable pipette anion for obtaining good seals and whole-cell sodium currents in these extremely small ORNs. These results implied that resting potentials, for viable responsive cells, should be more negative than about -90 mV, as supported by the observation that action potentials could only be evoked from holding potentials more negative than -90 mV.

M2 pore mutations convert the glycine receptor channel from being anion- to cation-selective.

Three mutations in the M2 transmembrane domains of the chloride-conducting alpha1 homomeric glycine receptor (P250Delta, A251E, and T265V), which normally mediate fast inhibitory neurotransmission, produced a cation-selective channel with P(Cl)/P(Na), = 0.27 (wild-type P(Cl)/P(Na) = 25), a permeability sequence P(Cs) > P(K) > P(Na) > P(Li), an impermeability to Ca(2+), and a reduced glycine sensitivity. Outside-out patch measurements indicated reversed and accentuated rectification with extremely low mean single channel conductances of 3 pS (inward current) and 11 pS (outward current). The three inverse mutations, to those analyzed in this study, have previously been shown to make the alpha7 acetylcholine receptor channel anion-selective, indicating a common location for determinants of charge selectivity of inhibitory and excitatory ligand-gated ion channels.

The startle disease mutation Q266H, in the second transmembrane domain of the human glycine receptor, impairs channel gating.

Hyperekplexia (startle disease) results from mutations in the glycine receptor chloride channel that disrupt inhibitory synaptic transmission. The Q266H missense mutation is the only hyperekplexia mutation located in the transmembrane domains of the receptor. Using recombinant expression and patch-clamping techniques, we have investigated the functional properties of this mutation. The ability of glycine and taurine to open the channel was reduced in the mutated channel, as shown by a 6-fold shift in the concentration-response curve for both agonists. This was not accompanied by similar changes in agonist displacement of strychnine binding, suggesting that the mutation affects functions subsequent to ligand binding. Taurine was also converted to a weak partial agonist and antagonized the actions of glycine, consistent with changes in its channel gating efficacy. Because the Q266H mutation is within the pore-forming second transmembrane domain, we tested for a direct interaction with permeating ions. No change in either the cation/anion selectivity ratio or in single channel conductance levels was observed. No differential effects of Zn++, pH, and diethylpyrocarbonate were observed, implying that the histidine side chain is not exposed to the channel lumen. Single-channel recordings revealed a significant reduction in open times in the mutant receptors, at both high and low agonist concentrations, consistent with the open state of the channel being less stable. This study demonstrates that residues within the second transmembrane domain of ligand-gated ion channel receptors, even those whose side chains do not directly interact with permeating ions, can affect the kinetics of channel gating.

IP3-gated channels and their occurrence relative to CNG channels in the soma and dendritic knob of rat olfactory receptor neurons.

Olfactory receptor neurons respond to odorants with G protein-mediated increases in the concentrations of cyclic adenosine 3',5'-monophosphate (cAMP) and/or inositol-1,4,5-trisphosphate (IP3). This study provides evidence that both second messengers can directly activate distinct ion channels in excised inside-out patches from the dendritic knob and soma membrane of rat olfactory receptor neurons (ORNs). The IP3-gated channels in the dendritic knob and soma membranes could be classified into two types, with conductances of 40 +/- 7 pS (n = 5) and 14 +/- 3 pS (n = 4), with the former having longer open dwell times. Estimated values of the densities of both channels from the same inside-out membrane patches were very much smaller for IP3-gated than for CNG channels. For example, in the dendritic knob membrane there were about 1000 CNG channels x microm(-2) compared to about 85 IP3-gated channels x microm(-2). Furthermore, only about 36% of the dendritic knob patches responded to IP3, whereas 83% of the same patches responded to cAMP. In the soma, both channel densities were lower, with the CNG channel density again being larger ( approximately 57 channels x microm(-2)) than that of the IP3-gated channels ( approximately 13 channels x microm(-2)), with again a much smaller fraction of patches responding to IP3 than to cAMP. These results were consistent with other evidence suggesting that the cAMP-pathway dominates the IP3 pathway in mammalian olfactory transduction.

Ion permeation and selectivity of wild-type recombinant rat CNG (rOCNC1) channels expressed in HEK293 cells.

The permeation properties of adenosine 3', 5'-cyclic monophosphate (cAMP)-activated recombinant rat olfactory cyclic nucleotide-gated channels (rOCNC1) in human embryonic kidney (HEK 293) cells were investigated using inside-out excised membrane patches. The relative permeability of these rOCNC1 channels to monovalent alkali cations and organic cations was determined from measurements of the changes in reversal potential upon replacing sodium in the bathing solution with different test cations. The permeability ratio of Cl(-) relative to Na(+) (P(Cl)/P(Na)) was about 0.14, confirming that these channels are mainly permeable to cations. The sequence of relative permeabilities of monovalent alkali metal ions in these channels was P(Na) > or = P(K) > P(Li) > P(Cs) > or = P(Rb), which closely corresponds to a high-strength field sequence as previously determined for native rat olfactory receptor neurons (ORNs). The permeability sequence for organic cations relative to sodium was P(NH3OH) > P(NH4) > P(Na) > P(Tris) > P(Choline) > P(TEA), again in good agreement with previous permeability ratios obtained in native rat ORNs. Single-channel conductance sequences agreed surprisingly well with permeability sequences. These conductance measurements also indicated that, even in asymmetric bi-ionic cation solutions, the conductance was somewhat independent of current direction and dependent on the composition of both solutions. These results indicate that the permeability properties of rOCNC1 channels are similar to those of native rat CNG channels, and provide a suitable reference point for exploring the molecular basis of ion selectivity in recombinant rOCNC1 channels using site-directed mutagenesis.