Measuring the firing rate of high-resistance neurons with cell-attached recording

Abstract

Cell-attached recording is extensively used to study the firing rate of mammalian neurons, but potential limitations of the method have not been investigated in detail. Here we perform cell-attached recording of molecular layer interneurons in cerebellar slices from rats and mice, and we study how experimental conditions influence the measured firing rate. We find that this rate depends on time in cell-attached mode, on pipette potential, and on pipette ionic composition. In the first minute after sealing, action currents are variable in shape and size, presumably reflecting membrane instability. The firing rate remains approximately constant during the first 4 min after sealing and gradually increases afterward. Making the pipette potential more positive leads to an increase in the firing rate, with a steeper dependence on voltage if the pipette solution contains K(+) as the main cation than if it contains Na(+). Ca(2+) imaging experiments show that establishing a cell-attached recording can result in an increased somatic Ca(2+) concentration, reflecting an increased firing rate linked to an increase in the pipette-cell conductance. Pipette effects on cell firing are traced to a combination of passive electrical coupling, opening of voltage- and Ca(2+)-sensitive K(+) channels (BK channels) after action potentials, and random activation of voltage-insensitive, presumably mechanosensitive, cationic channels. We conclude that, unless experimental conditions are optimized, cell-attached recordings in small neurons may report erroneous firing rates.

Figure 1.

Effect of pipette potential and pipette solution on firing rate as measured in cell-attached recording. A, Representative traces obtained in cell-attached mode when using a K⁺-filled pipette at two different pipette potentials, yielding different firing rates (−40 mV, 1.5 Hz; −60 mV, 0.9 Hz). Pipette input resistance 11.2 GΩ. B, Semi-log graph of summary data showing the dependence of firing rate on pipette potential with K⁺-filled pipettes (6 experiments with K-gluconate and 7 experiments with KCl). There is a significant increase of firing rate on potential (p < 0.01), with a slope corresponding to a twofold factor in 27 mV. Each experiment comprised a recording taken at −60 mV, and the firing rates were normalized to the value observed at this potential (black dot). C, Likewise, the firing rate increases significantly with potential when using NaCl-filled pipettes (n = 8 cells; p < 0.01), but the slope is more shallow (2-fold factor in 62.7 mV).

Figure 2.

Dependence of firing rate on input current in whole-cell recording. Whole-cell recording data from MLIs using K-glutamate solution. A, Voltage-clamp data taken at −70 mV. Spontaneous synaptic currents are all inward at this potential (arrows, numerous GPSCs and a single EPSC). The holding current was −4.6 pA. B, Current-clamp recordings from the same cell with holding currents of −4 pA (top) and −2 pA (bottom). Note bursting pattern in which action potentials ride on sustained depolarizations (upstates) near −60 mV, whereas interburst periods have a mean potential approximately −70 mV. C, Membrane potential distributions for various holding potential values. Same cell as in A and B. D, Group data for 11 cells as in A--C, showing mean spike frequency as a function of holding current. The steepest change in firing rate occurred between −4 and −3 pA (2.4-fold change).

Figure 3.

Equivalent circuit of the cell-attached recording mode. A, Model overview. The pipette compartment (potential V_p) is connected to earth through the seal impedance Z_s. It is connected to the cell compartment (potential V_c) through the impedance of the patch membrane Z_p. The cell compartment is connected to earth across the cell impedance Z_c. B, Detail of the three impedance pathways outlined in A. Left, Z_s includes the seal resistance R_s as well as the capacitance of the pipette tip, C_s. Middle, Z_p is made up of three elements in parallel: the patch membrane capacitance C_m, the resistance of the patch membrane R_m with its associated electromotive force E, and an additional element R_a resulting from the mechanical stress generated by pressing the pipette against the cell. Right, Z_c is represented by the combination of the cell resistance R_c (in series with E) and of the cell capacitance C_c.

Figure 4.

Shape of action currents as a function of pipette solution and pipette potential. Action currents were aligned with respect to their rising phase. In A–C, the two first columns show superimposed action current traces at pipette potentials of −60 mV (blue) and 0 mV (black). The two last columns show superimposed averages of the action currents at the two pipette potentials, as well as superimposed integrals of these averages. A, NaCl in pipette solution. In this case, the action current waveforms are similar for the two pipette potentials. Their integrals (right) approach 0 at the end of the waveform (dashed line), indicating that each waveform is dominated by its capacitive component. B, K-gluconate in pipette solution. Here fluctuations can be discerned in individual traces (arrows), as outward deflections during the downstroke of the action current at −60 mV (inset in left indicate differences between single traces carrying deflections and the average of event-free traces), as well as inward deflections during the late part of the negative wave of the action current at 0 mV. As a result, average waveforms depend on the pipette potential, and the integral displays a negative value at 0 mV. C, K-gluconate + TEA in pipette solution. Including TEA (2.5–5 mm), a K channel blocker with high affinity for BK channels, abolishes trace-to-trace fluctuations as well as the sensitivity of average waveforms to pipette potential. D, Group data analysis of experiments as illustrated in A–C (n = 5–7 in each condition). Integral values are taken 10 ms after the upstroke of the action current (left and middle). At −60 mV, no significant difference appears between the three conditions (left), but at 0 mV, the integral measured in K-gluconate (KG) is significantly more negative than that obtained with either NaCl or K-gluconate + TEA (middle; p < 0.01, Mann–Whitney U test). Right, Log of the ratio between spiking frequencies measured at 0 and −60 mV. The results indicate that adding TEA to the K-gluconate pipette solution reduces the voltage sensitivity of the firing frequency (p < 0.05, Mann–Whitney U test). The K-gluconate + TEA results are indistinguishable from those obtained with an NaCl solution.

Figure 5.

Firing rate increases with time in cell-attached mode. A, Representative example showing recordings obtained during minutes 1, 3, and 8 after seal formation. Na⁺-filled pipette solution. Note that action currents appear mostly monophasic in the first trace, and biphasic in the second and third trace. B, Black curve, Group data (mean ± SEM, n = 13) showing a gradual increase in the firing rate reported in cell-attached mode as a function of time since seal formation. Pipette potential, −20 mV. Gray curve, Group data (mean ± SEM, n = 8) from loose cell-attached recordings with low input resistance (<40 MΩ) fail to display a similar increase with time. C, Comparison of action current waveforms during minutes 1 and 3. Left, Superimposed individual action currents reveal larger fluctuations in the action current waveforms during minute 1 (gray traces) than during minute 3 (black traces). Middle top, Corresponding averages display a smaller amplitude, more monophasic action current during minute 1 than during minute 3. Middle bottom, Associated SD is larger during minute 1 than during minute 3. Right, Group data (open symbols, individual experiments; filled symbols, mean ± SEM) showing a significant drop of the peak SD over peak amplitude from minutes 1 to 3. CV, Coefficient of variation. Significant differences from minute 1 data using the Wilcoxon's signed rank test are indicated by *p < 0.05 and **p < 0.01.

Figure 6.

Inward single-channel currents associated with patching. A, Single inward current openings obtained with an Na⁺-containing pipette solution, at a pipette potential of −20 mV (top trace) and 0 mV (bottom trace). B, Top trace, Single-channel openings obtained at a pipette potential of −10 mV, using a K⁺-containing pipette solution. Bottom graphs, Open time and closed time histograms for the same recording. C, Top trace, Inward current events recorded at a pipette potential of −20 mV give rise to action potentials. Bottom trace, These events have an irregular shape, with variable amplitudes and gradual open–closed transitions, suggesting that the corresponding channels are located in the space comprised between the patch membrane and the recording pipette.

Figure 7.

Ca²⁺ changes associated with patching. A, Left, Fluorescence image of an MLI under resting conditions in a slice submitted to fura-2 loading. Adult mouse, physiological temperature. Right, Fluorescence responses to cell silencing obtained by puff applications of 1 μm muscimol (bars above traces; note that fluorescence increase corresponds to a decrease in Ca²⁺ concentration). Responses are illustrated before patch recording (top trace) as well as during recording (bottom trace, with raster plot of action potentials). Thin white curves illustrate exponential fits to the fluorescence relaxations. The larger amplitude response found after patching indicates a higher mean Ca²⁺ concentration and therefore a higher firing rate. The pipette was filled with an Na⁺-rich solution and held at 0 mV. B, Summary results of fluorescence changes in response to muscimol applications for P13–P14 mice. On average, fluorescence responses to muscimol application are larger after patching than before, indicating an increased firing frequency; however, the effect is significant only for near-physiological temperature data (right). RT, Room temperature. C, Summary results of fluorescence changes in response to muscimol applications, for adult mice (P24–P31). Results are mixed at room temperature, without any significant trend emerging from the average data. At near-physiological temperature, however, there is a significant increase in the fluorescence response after patching.