Lecture Notes in Computer Science
Spontaneous Voltage Transients in Mammalian Retinal
Download 12.42 Mb. Pdf ko'rish
|
- Bu sahifa navigatsiya:
- Keywords
- 2.1 Cell Dissociation
- 2.2 Electrophysiology
- Fig. 1.
|
Spontaneous Voltage Transients in Mammalian Retinal Ganglion Cells Dissociated by Vibration Tamami Motomura, Yuki Hayashida, and Nobuki Murayama
Graduate school of Science and Technology, Kumamoto University, 2-39-1 Kurokami, Kumamoto 860-8555, Japan tama@brain.cs.kumamoto-u.ac.jp, {yukih,murayama}@cs.kumamoto-u.ac.jp
mammalian retinae by utilizing low-Ca 2+ tissue incubation and the vibro- dissociation technique, but without use of enzyme. The retinal ganglion cell somata dissociated by this method showed spontaneous voltage transients (sVT) with the fast rise and slower decay. In this study, we analyzed characteristics of these sVT in the cells under perforated-patch whole-cell configuration, as well as in a single compartment cell model. The sVT varied in amplitude with quantal manner, and reversed in polarity around − 80 mV in a normal physiological saline. The reversal potential of sVT shifted dependently on the K + equilibrium potential, indicating the involvement of some K + conductance. Based on the model, the conductance changes responsible for producing sVT were little dependent on the membrane potential below − 50 mV. These results could suggest the presence of isolated, inhibitory presynaptic terminals attaching on the ganglion cell somata. Keywords: Neuronal computation, dissociated cells, retina, patch-clamp, neuron model.
1 Introduction Elucidating the functional role of single neurons in neural information processing is intricate because the neuronal computation itself is highly nonlinear and adaptive, and depends on combinations of many parameters, e.g. the ionic conductances, the intracellular signaling, their subcellular distributions, and the cell morphology. Furthermore, the interactions with surrounding neurons/glias can alter those factors, and thereby hinder us from examining some of those factors separately. This could be overcome by pharmacologically or physically isolating neurons from the circuits. One would use the pharmacological agents those can block the synaptic signal transmission in situ, although it is hard to know whether or not such agents show any unintended side-effects. Alternatively, one can dissociate neural tissue into single neurons by means of enzymatic digestion and mechanical trituration. The dissociated single neurons often lost their fine neurites and the synaptic contacts with other cells during the dissociation procedure, and thus, are useful for examining the properties of ionic conductances at known membrane potentials [3]. Unfortunately, however,
sVT in Mammalian Retinal Ganglion Cells Dissociated by Vibration 65 several studies have demonstrated that proteolytic enzymes employed for the cell dissociations can distort the amplitude, kinetics, localization, and pharmacological properties of ionic currents, e.g. [2]. These observations lead to attempts to isolate neurons by enzyme-free, mechanical means. Recently, we developed a new protocol for dissociating single neurons from specific layers of mammalian retinae without use of any proteolytic enzymes [12], but with a combination of the low-Ca 2+ tissue incubation [9] and the vibrodissociation technique [15] which has been applied to the slices of brains and spinal cords [1]. The somata of ganglion cells dissociated by our method showed spontaneous voltage transients (sVT) with fast rise and slower decay in the time course [8]. To our knowledge, such sVT have never been reported in previous studies on the retinal ganglion cells dissociated with or without enzyme [9]. Therefore, in this study, we analyzed characteristics of these sVT in the cells under perforated-patch whole-cell configuration, as well as in a single compartment cell model. The present results could suggest the presence of inhibitory presynaptic terminals attaching to the ganglion cell somata we recorded from, as demonstrated in previous studies on the vibrodissociated neurons of brains and spinal cords [1]. If this is the case, the retinal neurons dissociated by our method would be advantageous to investigate the mechanisms of transmitter release in single tiny synaptic boutons even under the isolation from the axons and neurites.
All animal care and experimental procedures in this study were approved by the committee for animal researches of Kumamoto University.
The neural retinas were isolated from two freshly enucleated eyes of Wistar rats (P7-P25), cut into 2-4 pieces each, and briefly kept in chilled extracellular “bath” solution. This solution contained (in mM): 140 NaCl, 3.5 KCl, 1 MgCl 2 , 2.5 CaCl 2 , 10
D -glucose, 5 HEPES. The pH was adjusted to 7.3 with NaOH. A retinal piece was then placed with photoreceptor-side down in a culture dish, covered with 0.4 ml of chilled, low-Ca 2+ solution and incubated for 3-5 min. This low-Ca 2+ solution contained (in mM): 140 sucrose, 2.5 KCl, 70 CsOH, 20 NaOH, 1 NaH 2 PO 4 , 15 CaCl 2 ,
D -glucose, 15 HEPES. The estimated free Ca 2+ concentration was 100–200 nM. The pH was adjusted to 7.2 with HCl. After the incubation, the fire- blunted glass pipette horizontally vibrating in amplitude of 0.2-0.5 mm at 100 Hz was applied to the flattened surface of retina throughout under visual control with the microscope, so that the cells were dissociated from the ganglion cell layer, but least from the inner and outer nuclear layers. After removing the remaining retinal tissue, the culture dish was filled with the bath solution, and left on a vibration-isolation table for allowing cells to settle down for 15-40 min. The bath solution was replaced by a fresh aliquot supplemented with 1 mg/ml bovine serum albumin and the dissociated cells were maintained at room temperature (20-25 o C) for 2-18 hrs prior to the electrophysiological recordings described below. The ganglion cells were identified 66 T. Motomura, Y. Hayashida, and N. Murayama based on the size criteria [6]. Nearly all those cells we made recordings from in voltage-/current-clamp showed the large amplitude of voltage-gated Na + current and/or of action potentials (see Fig. 1A-B), verifying that they were ganglion cells [4]. 2.2 Electrophysiology Since the previous studies demonstrated that the membrane conductances of retinal ganglion cells can be modulated by the intracellular messengers, e.g. Zn 2+ [13] and cAMP [7], all recordings presented here were performed in perforated-patch whole- cell mode [9] to maintain cytoplasmic integrity. Patch electrodes were pulled from borosilicate glass capillaries to tip resistances of approximately 4-8 M Ω . The tip of the electrodes were filled with a recording “electrode” solution that contained (in mM): 110 K- D -gluconic acid, 15 KCl, 15 NaOH, 2.6 MgCl 2 , 0.34 CaCl 2 , 1 EGTA, 10 HEPES. The pH was adjusted to 7.2 with methanesulfonic acid. The shank of the electrodes were filled with this solution after the addition of amphotericin B as the perforating agent (260 μ g/ml, with 400 μ g/ml Pluronic F-127). The recordings were made after the series resistance in perforated-patch configuration reached a stable value (typically 20-40 M Ω , ranging 10-100 M Ω ). In the fast current-clamp mode of the amplifier (EPC-10, Heka), the voltage monitor output was analog-filtered by the built-in Bessel filters (3-pole 10–30 kHz followed by 4-pole 2-5 kHz) and digitally sampled (5–20 kHz). The voltage drop across the series resistance was compensated by the built-in circuitry. The recording bath was grounded via an agar bridge, and the bath solution was continuously superfused over each cell recorded from, at a constant flow rate (0.4 ml/min). The volume of solution in the recording chamber was kept at about 2 ml. To apply a high-K + solution (Fig. 2B), 8 mM NaCl in the bath solution was replaced by the equimolar KCl. An enzyme solution was made by supplementing 0.25 mg/ml papain and 2.5 mM L -cystein in the bath solution. All experiments were performed at room temperature. 3 Results Perforated-patch whole-cell recordings were made from the somata of ganglion cells dissociated by our recently developed protocol (see Methods), which offered us quantitative measurements of the intrinsic membrane properties with the least
distortion due to the proteolysis by enzymes [2]. Conversely, since these cells were never exposed to any enzyme, they were useful in examining the effects of the enzymes utilized for the cell dissociation in previous studies. In fact, spike firing of the ganglion cells in response to constant current injection via the patch electrode (30-pA step in the positive direction) was irreversibly altered when the enzyme solution was superfused over those cells (n=3): 1) The resting potential depolarized by 5-20 mV and the spike firing diminished during the enzyme application; 2) When the enzyme was washed out from the recording chamber, the resting potential gradually hyperpolarized near the original level and the spike firing returned in some way;
sVT in Mammalian Retinal Ganglion Cells Dissociated by Vibration 67
A: Microphotograph of the cell recorded from. Note the soma being lager than 15 μ m in diameter. B: Membrane potential changes in response to step-wise constant current injections. Four traces are superimposed. The injected current was 10 pA in the negative direction and 10, 20, and 30 pA in the positive directions. C: Spontaneous hyperpolarizations under the current- clamp. The recordings were made for 50 sec in three different episodes with breaks for 12 sec between first and second, and for 6 sec between second and third. A constant current (2 pA in the positive direction) was injected to hold the membrane potential at around –70 mV (dashed gray line). Inset: Examples of sVT in an expanded time scale. Five events are recognized. Three of them are similar in their amplitude and time course, and the other two have roughly half and quarter of the largest amplitude. 3) After ~20 min of the washing out of enzyme, the spike firing reached a steady state at which the interval between the first and second spike firings in response to the current step was shorter than that before the enzyme application, by 40 ±11 % (mean
±S.E.) (not shown, [12]). These results suggest that, in previous studies on isolated retinal ganglion cells, some of the ionic channels could be significantly distorted during the dissociation procedure because of the use of proteolytic enzymes. Moreover, we found the spontaneous voltage transients (sVT) with the fast rise and slower decay in the retinal ganglion cell somata dissociated by our method [8]. Fig. 1C shows an example of sVT recorded from the cell shown in Fig. 1A. As shown in the Fig., transient hyperpolarizations spontaneously appeared under a constant current injection. Most of these hyperpolarizations are similar in amplitude and time course at a certain membrane potential ( − 70 mV here) and in some, the peak amplitude of hyperpolarizations was roughly the half or quarter (or one-eighth, in other cells) of the largest one (Inset). Such sVT appeared in 10-20 % of the cells we made recordings from, and could be observed in particular cells as long as we kept the recordings (0.5-2 hrs). When the enzyme solution was superfused over one of those cells, the sVT disappeared completely, and then were not seen again even after 20 min of the washing out of enzyme.
68 T. Motomura, Y. Hayashida, and N. Murayama C A B
+
by injecting holding currents ranging between –8 and +8 pA in A and between –8 and +12 pA in B. C: Plots of the peak amplitude versus basal potential. Only the events having the largest amplitude (see Fig. 1C) are taken into account. Note that the amplitude of depolarizations were plotted as negative values, and vice versa. The filled circles and open circles represent the data for 3.5-mM K
+ and 11.5-mM K + , respectively. As shown in Fig. 1C, the sVT were all recorded as hyperpolarizations when the cell was held at approximately − 70 mV. Thus, the reversal potential of the ionic current producing sVT should be below this voltage. In Fig. 2, the reversal potential for sVT was measured by holding the basal membrane potential at different levels under the current-clamp, c.f. [5]. As expected, the polarity of sVT reversed around − 80 mV when the basal membrane potential was varied from about − 100 to
− 40 mV
(Fig. 2A). Based on the ionic compositions in the bath and electrode solutions used in this recording, the equilibrium potential of K + (E
K ) was estimated to be about − 90
mV, and close to the reversal potential for sVT. When the E K was shifted by +30 mV, i.e. from about − 90 to − 60 mV by applying the high-K + solution (see Method), the polarity of sVT reversed between − 54 and − 39 mV of the basal membrane potential (Fig. 2B). Fig. 2C plots the peak amplitude of sVT versus the basal membrane potential. The linear regressions on these plots (gray lines) crossed the abscissa (dashed line) at approximately –76 mV and –48 mV for 3.5-mM K + and 11.5-mM K + , respectively, showing the shift of reversal potential parallel to the E K shift. Similar results were obtained in other two cells. These results indicate that the ionic conductance responsible for producing sVT is permeable to, at least, K + .
In the present experiments, we made recordings from the cells without neurites or with neurites no longer than 10 μ m or so. Therefore, those cells can be modeled as a single compartment shown in Fig. 3A. In this model, the unknown conductance responsible for producing sVT and the reversal potential are represented by “gx” and “Ex”, respectively. The membrane properties intrinsic to the cell are represented by membrane capacitance Cm, nonlinear conductance gm, and the apparent reversal potential Em. Here, Cm was measured with the capacitance compensation circuitry of
sVT in Mammalian Retinal Ganglion Cells Dissociated by Vibration 69
Voltage responses to the current steps injected. The amplitude of current steps (Iinj) were varied from –10 to +10 pA in 5-pA increment and the corresponding voltage changes (Vm), from the bottom (black) to the top (light gray), were recorded. C: Plots of the membrane potential versus the amplitude of current steps. The voltage was measured at the time points indicated by the marks in B (circle, square, triangle, rhombus, and hexagon). The solid line shows the best fit to the plots with a single exponential function. D: Voltage-dependency of gm calculated from the plots in C. The derivative of current with respect to the membrane potential gave the slope conductance gm, which could be approximated by a hyperbolic function, gm = Iα / (Vα
–Vm), where Vm by constant current injections. The injected currents were +10, +5, 0, –10 pA, from the top (light gray) to the bottom (black). F: Conductance calculated from sVT shown in E, based on the model shown in A. Four traces are superimposed. G: Plots of the peak amplitude versus basal potential. H: Plots of the decay time constant versus basal potential. The decay time constant was obtained from the best fit to the decay phase of conductance changes with single exponential functions. In G and H, two marks (filled and open circles) represent the data obtained from two different cells, and error bars indicate ±1 SE for multiple events of sVT.
70 T. Motomura, Y. Hayashida, and N. Murayama the patch-clamp amplifier we used and Em was measured as the resting potential under current-clamp. As shown in Fig. 3B-D, gm (a function of membrane potential) was also estimated experimentally from the voltage shifts induced by the current steps injected to the cell. Likewise, the value of Ex was estimated as shown in Fig. 2C. Since, during sVT, the membrane potential (Vrec) was recorded while the known amplitude of current (Iinj) was injected from the electrode, the time-varying gx could be calculated from the Kirchhoff’s law, by the equation:
x m m E } E { dt d C − − − − =
V (t) V g (t) V I (t) g rec rec m rec inj x (1)
Fig. 3F shows the calculated time course of gx during the sVT shown in Fig. 3E. Although the sVT were different in amplitude, time course, and polarity when measured at different basal potentials (E), the calculated gx fairly resembled each other (F). When the peak amplitude of the gx change was plotted against the basal membrane potential, little voltage dependency was found at potentials especially below –50 mV (Fig. 3G). And also, the decay time constant of the gx change are almost constant over the potential we recorded (Fig. 3H). The mean values of peak amplitude and decay time constants were in the ranges of 0.1-0.3 nS and 0.1-0.15 sec, respectively. These results were quantitatively similar in two cells, as compared the plots with different marks in G and H, implying that similar populations of ionic channels operated to produce the sVT. 4 Discussions In previous studies, spontaneous voltage/current transients have never been reported in acutely dissociated retinal ganglion cells. Here, we made recordings from the mammalian retinal ganglion cells dissociated by the recently developed method [8], and showed the characteristics of sVT observed in those cells. The present study showed that the sVT have the peak amplitude in quantal manner (Fig. 1), and may involve K + conductance activation (Fig. 2) which has little or slight voltage- dependency over the range we examined (Fig. 3). These results are less compatible with the idea that the voltage-gated K + conductance are activated spontaneously by intracellular messengers like Ca 2+ , as in the dissociated retinal amacrine cells [11]. Previous studies have shown that spontaneous postsynaptic potentials or currents can be recorded in the single neurons vibrodissociated from brains and spinal cords [1]. Since the dissociation protocol we used here utilized the vibrodissociation technique [8], it is feasible for us to have the working hypothesis illustrated in Fig. 4, to infer the underlying mechanisms of sVT we observed in the retinal ganglion cells. If this hypothesis is the case, it enables one to investigate the functional machinery controlling the transmitter release in single synaptic terminals under the isolation from the axons and neurites [14].
sVT in Mammalian Retinal Ganglion Cells Dissociated by Vibration 71
Download 12.42 Mb. Do'stlaringiz bilan baham: 
|