Two-compartment model for whole-cell data analysis and transient compensation

A novel way to go whole-cell in patch-clamp experiments

With a conventional patch-clamp electrode, an Ag/AgCl wire sits stationary inside the pipette. To move from the gigaseal cell-attached configuration to whole-cell recording, suction is applied inside the pipette. We have designed and developed a novel Pushpen patch-clamp electrode, in which a W wire insulated and wound with Ag/AgCl wire can move linearly inside the pipette. The W wire has a conical tip, which can protrude from the pipette tip like a push pen, a procedure we call the Pushpen Operation. We use the Pushpen operation to impale the cell membrane in cell-attached configuration to go whole-cell without disruption of the gigaseal. We successfully recorded whole-cell currents from chinese hamster ovarian cells expressing influenza A virus protein A/M2, after obtaining whole-cell configuration with the Pushpen operation. This novel method of achieving whole-cell configuration may have a higher success rate than is the case with the conventional patch clamp technique.

Serotoninergic control of glycinergic inhibitory postsynaptic currents in rat hypoglossal motoneurons

This report presents the results of a study of the frequency potentiation of inhibitory postsynaptic currents (IPSCs) in hypoglossal motoneurons and its modulation by serotonin. A release-site model of synaptic plasticity was used to characterize the frequency-related potentiation of evoked IPSCs. Data were obtained to determine if the frequency potentiation of IPSCs occurs as a consequence of a low baseline quantal content of evoked IPSCs using whole cell patch-clamp recordings from hypoglossal motoneurons in the neonatal rat brainstem slice preparation. In these motoneurons, EPSCs and GABAergic IPSCs were blocked by the application of CNQX, AP-5 and bicuculline. Glycinergic IPSCs were evoked by threshold stimulation of inhibitory neurons in the nucleus of Roller, which is located ventro-lateral to the hypoglossal nucleus. IPSC responses to trains of stimuli were recorded in control solutions and in solutions containing serotonin, which is known to reduce IPSPs in this preparation. The amplitude of non-potentiated IPSCs was reduced and their frequency potentiation was enhanced when serotonin was added to the bath. These data were examined using a release-site model of synaptic plasticity in which facilitation is attributed to a time-dependent increase in the probability of transmitter release; depression is attributed to a time-dependent decrease in the number of sites available for release. Using this model, the effect of serotonin on frequency potentiation was explained by a combination of a reduction in the baseline probability of transmitter release and an increase in the time constant of decay of the increase in probability of release that follows a stimulus.

Interaction between rod and cone inputs in mixed-input bipolar cells in goldfish retina

One class of goldfish bipolar cells, the mixed-input bipolar cell, contacts both rods and cones. Although the morphology of the different mixed-input bipolar cell subtypes has been described, insight into the interaction between rods and cones at the bipolar cell level is scarce. The aim of this study was to characterize this interaction in the different physiological types of mixed-input bipolar cells. We found mixed-input bipolar cells that depolarized, hyperpolarized, or showed a combination of the two types of response after center stimulation. The relative contributions of rod and cone inputs varied strongly in these cell populations. Depolarizing mixed-input bipolar cells are rod-dominated, having the highest sensitivity and the smallest dynamic range. Hyperpolarizing mixed-input bipolar cells, on the other hand, have a more balanced rod-cone input ratio. This extends their dynamic range and decreases their sensitivity. Finally, opponent mixed-input bipolar cells seem to be mostly cone-dominated, although some rod input is present. The antagonistic photoreceptor inputs form a push-pull system that makes these mixed-input bipolar cells very sensitive to changes in light intensity. Our finding that spectral tuning changes with light intensity conflicts with the idea that the separate non-opponent and opponent channels are related to coding of brightness and color, respectively. The organization of mixed-input bipolar cells into various classes with different dynamic ranges and absolute sensitivities might be a strategy to transmit information about all visual aspects most efficiently, given the sustained nature of bipolar cell responses and their limited voltage range.

Ca2+ channel characteristics in neuroendocrine tumor cell cultures analyzed by color contour plots

Neuroendocrine tumor (NET) cells express several voltage-operated Ca2+ channels (VOCCs). To address the question, if clinically distinct entities of NETs could be associated by specific patterns of VOCC expression, electrophysiological properties of NET primary cultures derived from consecutive surgical resections and permanent NET cell cultures were determined and analyzed by a color-coding contour diagram method. Using whole-cell patch-clamp technique, electrophysiological data were obtained from human pancreatic (foregut) BON and mouse intestinal (midgut) STC-1 cells as well as from human primary cultivated NET cells from fore- and midgut tumors or liver metastasis. To describe definite Ca2+ channel characteristics, we suggested a color-coding method to depict the contours of time- and voltage-dependence. In this study, we could demonstrate specific Ca2+ channel properties in NET cells from midgut tumors which were not found in NET cells from foregut location. This may be important functionally in respect of different cell biological functions of NET cells such as release of bioamines and neuropeptides. In addition, definite differences between the effect of specific and unspecific Ca2+ channel modulators on NET cells were detected. Although most of this information can be obtained by superimposing current-voltage curves at different times after the onset of the voltage step, the color-coding contour diagrams clearly provides a better visual summary, and hence might be generally quite useful.

Correction of conductance measurements in non-space-clamped structures: 1. Voltage-gated K+ channels

To understand functions of a single neuron, such as propagation and generation of synaptic or action potentials, a detailed description of the kinetics and distribution of the underlying ionic conductances is essential. In voltage-clamp experiments, incomplete space clamp distorts the recorded currents, rendering accurate analysis impossible. Here, we present a simple numerical algorithm that corrects such distortions. The method performs a stepwise approximation of the conductance density at the site of a local voltage clamp. This is achieved by estimating membrane conductances in a simulation that yields simulated clamp currents, which are then fitted to the distorted recordings from the non-space-clamped structure, relying on accurately reconstructed cell morphology and experimentally determined passive properties. The method enabled accurate retrieval of the local densities, kinetics, and density gradients of somatic and dendritic channels. Neither the addition of noise nor variation of passive parameters significantly reduced the performance of the correction algorithm. The correction method was applied to two-electrode voltage-clamp recordings of K(+) currents from the apical dendrite of layer 5 neocortical pyramidal neurons. The generality and robustness of the algorithm make it a useful tool for voltage-clamp analysis of voltage-gated currents in structures of any morphology that is amenable to the voltage-clamp technique.

Parameter-extraction of a two-compartment model for whole-cell data analysis

Neuronal modeling of patch-clamp data is based on approximations which are valid under specific assumptions regarding cell properties and morphology. Certain cells, which show a biexponential capacitance transient decay, can be modeled with a two-compartment model. However, for parameter-extraction in such a model, approximations are required regarding the relative sizes of the various model parameters. These approximations apply to certain cell types or experimental conditions and are not valid in the general case. In this paper, we present a general method for the extraction of the parameters in a two-compartment model without assumptions regarding the relative size of the parameters. All the passive electrical parameters of the two-compartment model are derived in terms of the available experimental data. The experimental data is obtained from a DC measurement (where the command potential is a hyperpolarizing DC voltage) and an AC measurement (where the command potential is a sinusoidal stimulus on a hyperpolarized DC potential) performed on the cell under test. Computer simulations are performed with a circuit simulator, XSPICE, to observe the effects of varying the two-compartment model parameters on the capacitive transients of the current response. Our general solution for the parameter-estimation of a two-compartment model may be used to model any neuron, which has a biexponential capacitive current decay. In addition, our model avoids the need for simplifying and perhaps erroneous approximations. Our equations may be easily implemented in hardware/software compensation schemes to correct the recorded currents for any series resistance or capacitive transient errors. Our general solution reduces to the results of previous researchers under their approximations.

NRSF causes cAMP-sensitive suppression of sodium current in cultured hippocampal neurons

The neuron restrictive silencer factor (NRSF/REST) has been shown to bind to the promoters of many neuron-specific genes and is able to suppress transcription of Na(+) channels in PC12 cells, although its functional effect in terminally differentiated neurons is unknown. We constructed lentiviral vectors to express NRSF as a bicistronic message with green fluorescent protein (GFP) and followed infected hippocampal neurons in culture over a period of 1-2 wk. NRSF-expressing neurons showed a time-dependent suppression of Na(+) channel function as measured by whole cell electrophysiology. Suppression was reversed or prevented by the addition of membrane-permeable cAMP analogues and enhanced by cAMP antagonists but not affected by increasing protein expression with a viral enhancer. Secondary effects, including altered sensitivity to glutamate and GABA and reduced outward K(+) currents, were duplicated by culturing GFP-infected control neurons in TTX. The striking similarity of the phenotypes makes NRSF potentially useful as a genetic "silencer" and also suggests avenues of further exploration that may elucidate the transcription factor's in vivo role in neuronal plasticity.

Compartmental models of rat cerebellar Purkinje cells based on simultaneous somatic and dendritic patch-clamp recordings

1. Simultaneous dendritic and somatic patch-clamp recordings were made from Purkinje cells in cerebellar slices from 12- to 21-day-old rats. Voltage responses to current impulses injected via either the dendritic or the somatic pipette were obtained in the presence of the selective I(h) blocker ZD 7288 and blockers of spontaneous synaptic input. Neurons were filled with biocytin for subsequent morphological reconstruction. 2. Four neurons were reconstructed and converted into detailed compartmental models. The specific membrane capacitance (C(m)), specific membrane resistance (R(m)) and intracellular resistivity (R(i)) were optimized by direct fitting of the model responses to the electrophysiological data from the same cell. Mean values were: C(m), 0.77 +/- 0.17 microF cm(-2) (mean +/- S.D.; range, 0.64-1.00 microF cm(-2)), R(m), 122 +/- 18 kOmega cm(2) (98-141 kOmega cm(2)) and R(i), 115 +/- 20 Omega cm (93-142 Omega cm). 3. The steady-state electrotonic architecture of these cells was compact under the experimental conditions used. However, somatic voltage-clamp recordings of parallel fibre and climbing fibre synaptic currents were substantially filtered and attenuated. 4. The detailed models were compared with a two-compartment model of Purkinje cells. The range of synaptic current kinetics that can be faithfully recorded using somatic voltage clamp is predicted fairly well by the two-compartment model, even though some of its underlying assumptions are violated. 5. A model of I(h) was constructed based on voltage-clamp data, and inserted into the passive compartmental models. Somatic EPSP amplitude was substantially attenuated compared to the amplitude of dendritic EPSPs at their site of generation. However, synaptic efficacy of the same quantal synaptic conductance, as measured by the somatic EPSP amplitude, was only weakly dependent on synaptic location on spiny branchlets. 6. The passive electrotonic structure of Purkinje cells is unusual in that the steady-state architecture is very compact, while voltage transients such as synaptic potentials and action potentials are heavily filtered.