A voltage-clamp analysis of inward (anomalous) rectification in mouse spinal sensory ganglion neurones

The action of N-methyl-D-aspartic acid on mouse spinal neurones in culture

Neurones from the ventral half of mouse embryo spinal cord were grown in dissociated culture and voltage clamped. The current-voltage relation of responses evoked by N-methyl-D-aspartic acid (NMDA), L-glutamic acid and kainic acid was recorded in media of different ionic composition. On removal of Mg2+ from the extracellular solution, responses to NMDA and L-glutamate became less voltage sensitive, such that NMDA responses were no longer associated with a region of negative slope conductance. The antagonism of NMDA responses produced by application of Mg2+ to neurones bathed in nominally Mg2+-free solutions shows voltage dependence and uncompetitive kinetics. Voltage-jump experiments showed that the voltage-dependent action of Mg2+ occurred rapidly, and with complex kinetics. Ni2+ and Cd2+, two potent blockers of calcium currents in spinal cord neurones, had significantly different potencies as NMDA antagonists, Ni2+ being of greater potency than Mg2+, and Cd2+ considerably weaker. The voltage-dependent block of NMDA responses produced by physiological concentrations of Mg2+ is sufficient to explain the apparent increase in membrane resistance produced by NMDA in current-clamp experiments, and the ability of NMDA to support repetitive firing. Substitution of choline for Na+ produced a hyperpolarizing shift in the reversal potential for responses evoked by kainic acid consistent with an increase in permeability to Na+ and K+. In choline-substituted solutions, the reversal potential of NMDA responses was more positive than that recorded for kainic acid, and in addition NMDA responses showed enhanced desensitization.

Electrical properties of rat dorsal root ganglion neurones with different peripheral nerve conduction velocities

The electrical characteristics of individual rat dorsal root ganglion neurones were studied and related to the peripheral axon conduction velocity and morphological cell type. Neurones were divided into four groups based on the conduction velocity of their peripheral axons (A alpha, 30-55 m/s; A beta, 14-30 m/s; A delta, 2.2-8 m/s and C less than 1.4 m/s). Electrophysiological parameters examined included membrane potential, action potential amplitude and duration, after-potential height and duration, input resistance and the occurrence of time-dependent rectification. The mean duration of the somatic action potentials was found to be characteristic for each of the conduction velocity groupings. However, there was considerable overlap between groups. The fast-conducting (A alpha) and slowly conducting (A delta) myelinated fibres had short-duration action potentials, within the ranges 0.49-1.35 and 0.5-1.7 ms at the base respectively. The A beta and C cells had somatic action potentials with durations in the ranges of 0.6-2.9 and 0.6-7.4 ms respectively. The longer action potential durations could be related to the presence of an inflexion on the repolarizing phase seen in a third of A beta neurones (called A beta I neurones) and in all C neurones. The action potential overshoot was larger in C neurones and A beta I neurones than in the other neurone groups. The mean duration of the after-hyperpolarization was several times greater in C neurones than in A neurones. A delta neurones displayed the shortest and greatest amplitude after-hyperpolarizations. Large, long-lasting after-hyperpolarizations were not limited to neurones displaying an inflexion. The electrophysiological properties of the soma membrane of A delta neurones closely resembled those of A alpha neurones, while in several respects those of C neurones resembled the A beta I neuronal properties. The input resistance was found to be much greater in C than in A cells, although there was no significant difference between specific membrane resistance values calculated for the different groups. A number of A cells exhibited time-dependent rectification.

Sodium-potassium-ATPase activity in the dorsal root ganglia of rats with streptozotocin-induced diabetes

Sodium-potassium-ATPase activity was measured in excised dorsal root ganglia of streptozotocin-diabetic rats, 2 months after induction of diabetes. In comparison with age-matched controls, there was a decrease in both the total and ouabain-insensitive activity, indicating an overall reduction in ouabain-sensitive activity of 46%. This decrease may explain the reduced amino-acid uptake exhibited by diabetic sensory ganglia and could be relevant to the development of diabetic neuropathy.

Acetylcholine responses of identified neurons in Helix pomatia--I. Interactions between acetylcholine-induced and potential-dependent membrane conductances

The influence of potential-dependent membrane conductances on amplitude and time course of acetylcholine (ACh) responses was studied. The investigations were performed on the identified neurons B1 and B3 of the buccal ganglion of Helix pomatia. The neurons B1 and B3 were depolarized by ACh. The depolarization was accompanied by a decrease of membrane resistance. An inward rectification occurring negative to the resting membrane potential (RMP) reduced the amplitude of the ACh depolarizations. An outward rectification occurring positive to the RMP consisted of two parts and ceiled the ACh responses. The early outward current reduced the amplitude and modified the time course of ACh responses. Local responses or axonal action potentials increased the amplitude of the ACh depolarizations.

A voltage-clamp analysis of membrane currents in solitary bipolar cells dissociated from Carassius auratus

Membrane properties of solitary bipolar cells, mechanically dissociated from the enzyme-treated goldfish retina, were studied under current- and voltage-clamp conditions with 'giga-seal' suction pipettes (pipette solution 138 mM-K). The resting potential of solitary bipolar cells was about -30 mV. They responded to depolarizing current pulses with sustained depolarization, and to hyperpolarizing current pulses with an initial hyperpolarizing transient followed by a sag to a less hyperpolarized level. The current-voltage relationship determined under voltage-clamp conditions showed strong outward and inward rectification. The membrane currents consisted of four components; Ca current (ICa), voltage- and Ca-dependent K currents (IK(V) and IK(Ca), respectively), and an inward current activated by membrane hyperpolarization (Ih). ICa was activated by membrane depolarization beyond -40 mV, was maximum at +10 mV and became smaller with further depolarization. No polarity reversal was seen. ICa was enhanced by equimolar replacement of Ca with Ba, and was blocked by 4 mM-Co. IK(Ca) was observed by membrane depolarization beyond -10 mV, was maximum at about +40 mV, and became smaller with further depolarization. This current was suppressed by 4 mM-Co, 1.6 mM-Ba, 35 mM-TEA or 30 microM-quinine. IK(V) was activated by membrane depolarization beyond -60 mV, and had slower kinetics that ICa or IK(Ca). The reversal potential of the tail current was close to the K equilibrium potential (EK), suggesting that this current is carried purely by K ions. IK(V) was inactivated slowly and nearly completely by sustained depolarization. IK(V) was blocked by 35 mM-TEA. Ih was activated by membrane hyperpolarization (less than -60 mV). The current showed a time-dependent increase. It was also dependent on the membrane potential, but not on the driving force of K ions. This current seems to be carried by a mixture of Na and K ions, since (1) in low Na solution, Ih became small in amplitude, and (2) the reversal potential of the tail current was between the Na equilibrium potential (ENa) and EK X Ih was blocked by 10 mM-Cs, but was resistant to 0.2 mM-Ba. The resting potential and voltage responses of solitary bipolar cells are discussed in reference to the characteristics of each membrane conductance isolated in the present study.

Membrane resistivity estimated for the Purkinje neuron by means of a passive computer model

A multicompartment passive electrotonic computer model is constructed for the cerebellar Purkinje cell of the guinea-pig. The model has 1089 coupled compartments to accurately represent the morphology of the Purkinje cell. In order that the calculated behavior of the model fit the published electrophysiological observations of somatic and dendritic input conductance, the neural membrane resistivity must be spatially non-uniform. The passive electrical parameter values for which the model best fits the observations of input conductances, pulse attenuation and current-clamp voltage transients are rm,dend = 45,740 omega cm2, rm,soma = 760 omega cm2, ri = 225 omega cm and cm = 1.16 microF/cm2 (the membrane and cytoplasm specific resistivities and membrane specific capacitance, respectively). The model with these parameter values is electrically compact, with electrotonic length X = 0.33 and dendritic dominance ratio p = 0.44. Analysis of the calculated voltage transient of the multicompartment model by the methods of equivalent-cylinder cable theory is shown to result in very different and unreliable conclusions. The significance for neuronal function of the estimated electrical parameter values is discussed. The possible effect of active conductances on these conclusions is assessed.

Relations among passive electrical properties of lumbar alpha-motoneurones of the cat

The relations among passive membrane properties have been examined in cat motoneurones utilizing exclusively electrophysiological techniques. A significant relation was found to exist between the input resistance and the membrane time constant. The estimated electrotonic length showed no evident tendency to vary with input resistance but did show a tendency to decrease with increasing time constant. Detailed analysis of this trend suggests, however, that a variation in dendritic geometry is likely to exist among cat motoneurones, such that the dendritic trees of motoneurones projecting to fast-twitch muscle units are relatively more expansive than those of motoneurones projecting to slow-twitch units. Utilizing an expression derived from the Rall neurone model, the total capacitance of the equivalent cylinder corresponding to a motoneurone has been estimated. With the assumption of a constant and uniform specific capacitance of 1 mu F/cm2, the resulting values have been used as estimates of cell surface area. These estimates agree well with morphologically obtained measurements from cat motoneurones reported by others. Both membrane time constant (and thus likely specific membrane resistivity) and electrotonic length showed little tendency to vary with surface area. However, after-hyperpolarization (a.h.p.) duration showed some tendency to vary such that cells with brief a.h.p. duration were, on average, larger than those with longer a.h.p. durations. Apart from motoneurones with the lowest values, axonal conduction velocity was only weakly related to variations in estimated surface area. Input resistance and membrane time constant were found to vary systematically with the a.h.p. duration. Analysis suggested that the major part of the increase in input resistance with a.h.p. duration was related to an increase in membrane resistivity and a variation in dendritic geometry rather than to differences in surface area among the motoneurones. The possible effects of imperfect electrode seals have been considered. According to an analysis of a passive membrane model, soma leaks caused by impalement injury will result in underestimates of input resistance and time constant and over-estimates of electrotonic length and total capacitance. Assuming a non-injured resting potential of -80 mV, a comparison of membrane potentials predicted by various relative leaks (leak conductance/input conductance) with those actually observed suggests that the magnitude of these errors in the present material will not unduly affect the presented results.+4

Nerve Na+-K+-ATPase, conduction, and myo-inositol in the insulin-deficient BB rat

Nerve conduction slowing in acute diabetes in animals has been associated with both a diminished axolemmal transmembrane Na+ potential and a myo-inositol-related defect in nerve Na+-K+-ATPase activity. The interaction between these two potentially related defects, their reversibility, and their possible role in the nerve conduction slowing and axonopathy of diabetes are not well defined. Therefore, the effects of rigorous insulin replacement on peripheral nerve conduction velocity, myoinositol content, and Na+-K+-ATPase activity were studied in the spontaneously diabetic BB-Wistar rat, an animal model that manifests both conduction slowing and a characteristic progressive axonopathy. Twelve weeks of sustained hyperglycemia reduced both motor conduction velocity and Na+-K+-ATPase activity in sciatic nerve. Six weeks of subsequent vigorous insulin replacement normalized the enzymatic defect but only partially corrected diminished nerve conduction velocity. Hence, nerve conduction slowing in diabetic animals may be partly attributable to reduced nerve Na+-K+-ATPase activity, but a less readily reversible component of conduction slowing probably reflects structural alterations that occur in nerve within the first 3 mo of diabetes.

Mixed-agonist action of excitatory amino acids on mouse spinal cord neurones under voltage clamp

Neurones from the ventral half of mouse embryo spinal cord were grown in tissue culture and voltage clamped with two micro-electrodes. The current-voltage relation of responses evoked by brief pressure applications of excitatory amino acids was examined over a membrane potential range of -100 to +70 mV. Three types of current-voltage relation were observed. Responses to kainic and quisqualic acids were relatively linear within +/- 20 mV of the resting potential. N-methyl-D-aspartate (NMDA) and L-aspartic acid responses had a negative slope conductance at membrane potentials more negative than -30 mV. In contrast, over the same potential range the slope conductance of responses evoked by L-glutamic and L-homocysteic acids was close to zero. The membrane potential-chord conductance relation of the ionic mechanism activated by excitatory amino acids, derived using the driving force for ionic current, showed two types of behaviour. The conductance linked to NMDA receptors was highly voltage sensitive and increased on depolarization; a much weaker voltage sensitivity was observed for responses evoked by kainic and quisqualic acids. L-glutamic and L-homocysteic acid responses behaved as though due to simultaneous activation of both NMDA and either kainate or quisqualate receptors. In the presence of the NMDA receptor antagonist (+/-)-2-aminophosphonovaleric acid (2-APV) the response to L-glutamate became less voltage sensitive and resembled responses evoked by kainate or quisqualate. Simultaneous activation of both conductance mechanisms by mixtures of kainate and NMDA produced current-voltage and membrane potential-chord conductance relations similar to those of L-glutamate. The voltage sensitivity of the L-glutamate response was inversely related to the dose; for low doses of L-glutamate the slope conductance of responses recorded near the resting potential was close to zero. However, larger doses of L-glutamate evoked responses with a voltage sensitivity similar to that of kainate. We suggest that L-glutamate acts as a mixed agonist at both NMDA and non-NMDA receptors. This can explain the results of previous experiments that failed to demonstrate a membrane resistance change during L-glutamate-induced depolarizations.

Glutamate currents in mammalian spinal neurons: resolution of a paradox

Mouse spinal cord neurons grown in tissue culture were impaled with a pair of microelectrodes containing 1 M CsCl and voltage-clamped. Membrane currents evoked by excitatory amino acids were studied over the potential range -70 to +20 mV. Glutamate currents behaved as though generated by simultaneous activation of two conductance mechanisms, one voltage-sensitive, the other conventional. Block of NMDA receptors with the competitive antagonist 2-APV removed the voltage-sensitive component of the glutamate response. These results help to explain the paradoxical lack of conductance-change previously reported for glutamate responses recorded in the mammalian CNS.

Voltage-dependent block by Mg2+ of NMDA responses in spinal cord neurones

Acidic amino acids are putative excitatory synaptic transmitters, the ionic mechanism of which is not well understood. Recent studies with selective agonists and antagonists suggest that neurones of the mammalian central nervous system possess several different receptors for acidic amino acids, which in turn are coupled to separate conductance mechanisms. N-methyl-D-aspartic acid (NMDA) is a selective agonist for one of these receptors. The excitatory action of amino acids acting at NMDA receptors is remarkably sensitive to the membrane potential and it has been suggested that the NMDA receptor is coupled to a voltage-sensitive conductance. Recently, patch-clamp experiments have shown the voltage-dependent block by Mg2+ of current flow through ion channels activated by L-glutamate. We now show using voltage-clamp experiments on mouse spinal cord neurones that the voltage-sensitivity of NMDA action is greatly reduced on the withdrawal of physiological concentrations (approximately 1 mM) of Mg2+ from the extracellular fluid. This provides further evidence that Mg2+ blocks inward current flow through ion channels linked to NMDA receptors.

Effect of changes in intra- and extracellular sodium on the inward (anomalous) rectification in salamander photoreceptors

Solitary rod inner segments were obtained by enzymic dissociation of the tiger salamander retina. Ih, an inward current activated by membrane hyperpolarization, was studied using the single-pipette voltage-clamp technique with patch pipettes. In order to investigate Ih in isolation from voltage-dependent potassium and calcium currents, it was necessary to superfuse with a solution containing TEA and cobalt. When the solution in the patch pipette contained 45 mM-KCl and 50 mM-NaCl, the characteristics of Ih were indistinguishable from those previously described with fine-tip micro-electrodes: the reversal potential was near-30 mV and Ih was blocked by extracellular caesium and enhanced by an increase in the extracellular potassium concentration. The increase in Ih observed when the extracellular potassium concentration is raised is due to an increase in conductance and in driving force. Replacement of sodium in the patch pipette with choline caused a 15 mV displacement of the reversal potential for Ih in the depolarized direction. When using sodium-free patch pipettes, replacement of extracellular sodium displaced the reversal potential for Ih to -74 mV, a value in the range of the potassium equilibrium potential in solitary inner segments. Intracellular or intra- and extracellular sodium substitution affected neither the activation range of Ih nor the maximum conductance. From points 3-6 it can be concluded that Ih is carried mainly, if not exclusively, by sodium and potassium and that the channel responsible for Ih is insensitive to modifications of the intra- or extracellular sodium concentration. The results of long-term hyperpolarization, of partial block with caesium and of total sodium substitution are consistent with sodium and potassium permeating the same type of channel.

Diabetic neuropathy

Peripheral nerve disorders are important late complications of diabetes mellitus. Polyneuropathy, which may involve varying proportions of sensory, motor, and autonomic fibers, is considered the consequence of metabolic derangements that result from chronic hyperglycemia. Symmetrical proximal motor neuropathy ("diabetic amyotrophy") also may have a metabolic basis. Mononeuropathies in diabetes may have an ischemic or compressive cause. Advances have been made in understanding the biochemical basis for diabetic polyneuropathy. The treatment of symptomatic diabetic neuropathy should be directed toward long-term normalization of blood glucose until more specific therapies become available.