| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Division of Neurophysiology, Department of Medical Physiology, University of Copenhagen, Copenhagen, Denmark
Submitted 6 August 2004; accepted in final form 11 October 2004
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
10 mV) all-or-none depolarizing events of 10-ms duration. These spikelets often occurred in bursts at high frequency (
250 Hz); they were present despite the application of synaptic and gap junction antagonists, but were abolished by TTX and intracellularly applied QX314. The spikelets were interpreted as attenuated sodium spikes initiated in different branches of the granule cells dendrites. They occurred spontaneously, but could also be evoked by excitatory postsynaptic potentials (EPSPs) to the distal dendrites. Spikelets initiated by distal excitation could function as prepotentials for full sodium spikes, in part depending on the level of proximal depolarization. Somatic depolarization by the electric field evoked full sodium spikes as well as low-threshold calcium spikes (LTSs). Calcium imaging revealed that the electrophysiologically identified LTS evoked from the soma was associated with calcium transients in the proximal and the distal dendrites. Our data suggest that the LTS in the soma/proximal dendrites plays a major role in boosting excitability, thus contributing to the initiation of sodium spiking in this compartment. The results furthermore suggest that the LTS and the sodium spikes may act independently or cooperatively to regulate dendritic calcium influx. |
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
; Shepherd et al. 2004). In granule cells, synaptic output is localized to the distal dendrites, which connect them to mitral/tufted cell dendrites through dendrodendritic reciprocal synapses (Jahr and Nicoll 1980; Shepherd et al. 2004; Yokoi et al. 1995). The dendritic output and the axonless nature (Shepherd et al. 2004) of granule cells make these cells an interesting and practical example for the study of dendritic physiology, since dendritic properties can be directly related to cellular output. As the distal dendrites form the output region of the cell, the interaction and synchronization of dendritic spikes become of interest, since this will affect the degree to which the cell may function as a loosely connected ensemble of individual, isolated input-output regions, e.g., each dendritic branch or group of branches versus a synchronized collective providing a global output. Furthermore, the interaction between the proximal and the distal dendrites is of interest (Midtgaard 1994, 1995), since the proximal dendrites do not engage in dendrodendritic synapses, but receive traditional axodendritic input from centrifugal fibers and from axonal collaterals of mitral/tufted cells (Price and Powell 1970).
Granule cells produce short bursts of sodium action potentials (Hall and Delaney 2002; Wellis and Kauer 1994) and sodium spike–activated T-type calcium channels contribute to dendritic calcium transients (Egger et al. 2003). In addition, a low-threshold spike (LTS) due to T-type calcium channels has been shown to boost sodium spike initiation and can occur in isolation subthreshold for the sodium spike (Pinato and Midtgaard 2003), but the somatodendritic localization of this LTS has not been characterized. A dendritic localization of the LTS could suggest that the LTS, independently from the sodium spikes, contributes to local synaptic processing and dendritic calcium-dependent mechanisms, e.g., transmitter release.
Small fast spikes, "spikelets," have been observed in granule cells in vivo (Luo and Katz 2001; Mori and Takagi 1978; Wellis and Scott 1990). Their nature and localization have not been analyzed, however. The small diameter of the distal dendrites (1 µm, Woolf et al. 1991) makes dendritic patch-clamp recording impracticable in granule cells. The application of a constant electric field on neurons, especially when their morphology is characterized by extended processes, can be a useful tool to activate, modulate, or silence selective parts of the cells (Chan and Nicholson 1986; Tranchina and Nicholson 1986). The combination of field stimulation and intrasomatic current injection through the recording electrode can provide insight into spatial integration along the dendrites. Granule cells in the olfactory bulb are suitable for this purpose since they are polarized cells with the somatodendritic axis aligned in the ependymal-pial direction (Shepherd et al. 2004) and therefore the application of an external electric field represents an alternative experimental approach in the study of active properties in fine dendrites, in particular the initiation and spread of action potentials.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
Adult (carapace length 20–25 cm) turtles (Trachemys scripta elegans) were anesthetized with pentobarbitone (100 mg/kg, im) and decapitated. The surgical procedures comply with Danish legislation and are approved by the controlling body under The Danish Ministry of Justice. The olfactory bulb was separated from the rest of the brain and transferred to oxygenated Ringer fluid (Mori et al. 1981). The olfactory bulb was divided into the two hemispheres, and each of them was hemisected along the horizontal plane. Some experiments were performed in one-half of a hemisphere of the olfactory bulb. Coronal slices of 300–400 µm thickness were cut on a Vibratome (DSK). Both dissection and experiments were done at room temperature (22–24°C).
Slices were transferred to the recording chamber (1 ml vol.) and perfused with Ringer fluid at a rate of 1 ml/min. Ringer fluid had the following composition (in mM): 120 NaCl, 15 NaHCO3, 5 KCl, 2 MgCl2, 3 CaCl2, and 20 glucose, pH 7.6, oxygenated with 98% O2-2% CO2.
Recording electrodes (15–30 M
resistance) were filled with a solution of the following composition (in mM): 127 KCH3SO4, 1.53 Mg-gluconate, 3.7 MgCl2, 1 glucose, 5 HEPES, 5 Na-HEPES, and 2 Na2-ATP, adjusted to pH 7.5 with KOH (all chemicals from Sigma). Whole cell recordings were made by conventional "blind" recording technique or visually guided patching (see below). The data presented here were not corrected for the liquid junction potential between the recording electrode and the bath. Potassium-Lucifer yellow-CH (5 mM) was included in the recording electrodes for morphological identification in some cases. Standard procedures were followed to fix and clear the tissue for conventional fluorescence microscopy.
TTX was obtained from Alomone Labs (Jerusalem, Israel); D-2-amino-5-phosphonopentanoic acid (AP-5), 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX), and SR-95531 were obtained from Tocris (Bristol, UK); Mibefradil was kindly provided by Roche Pharma (Basel, Switzerland). Octanol and carbenoxolone (Sigma) were bath-applied for 30 min.
The data were amplified (Axoclamp 2B, Axon Instruments), digitized at 20 kHz (Digidata, Axon Instruments), and analyzed using pClamp8 software (Axon Instruments). Statistical significance was assessed using Student's t-test.
Two Ag/AgCl electrodes were positioned in the recording chamber as close as possible to the slice to deliver an electric field along the ependymal-pial direction, i.e., orthogonal to the slice. The electric field electrodes were connected to a constant current stimulus isolation unit (Isolator 11, Axon Instruments). The actual field strength was calculated by fitting the voltage gradient measured at different locations between the electrodes. It was homogeneous in the direction of the granule cell dendrites and zero in the horizontal direction. Any offset added to the recordings in the presence of the electric field was subtracted on-line using an extracellular reference electrode placed as close as possible to the recording site.
Fluorescence imaging
Calcium imaging combined with patch recordings were performed on a Zeiss Axioskop FS equipped with water-immersion optics (40x/0.8; 63x/0.9), an excitation light source (Polychrome II, TILL Photonics), and a cooled CCD camera (Sensicam, 640 x 480 pixels, PCO). The electrodes and the microscope were positioned using an electronically controlled system (Luigs and Neumann). Cells were visualized using conventional video enhancement in 300-µm-thick slices and were identified as granule cells based on their size and the location of the soma and dendrites. Recordings from granule cell somata showed electrophysiological properties similar to the results obtained by blind patching. For calcium imaging, 300 µM Oregon BAPTA was included in the patch pipette solution. Following break-in, a filling period 45 min was necessary for the distal tips of dendrites to show an adequate fluorescence signal. For inclusion in this study, cells were required to have a stable resting membrane potential with no bias current and be able to generate full size sodium spikes after filling with calcium indicator. Twenty-one cells were included in this study. Images were most often collected using 200- or 400-ms exposure for each frame in simultaneous acquisition mode. Pixels were binned at 2 x 2 or 4 x 4. Usually a series of 20 or 40 frames was collected for each stimulation of the cell. Electrical stimulation of the cell was synchronized to frame acquisition. Background fluorescence was measured in a region of the slice away from the cell recorded from and subtracted from the signal, and the data are presented as dF = Factive – Freference or dF/F = 100 x dF/Freference, using an average of four prestimulus frames as reference. Images were analyzed using custom built routines (IDL, Research Systems). The pixel values in a region of interest were averaged for a detailed analysis of fluorescence changes. For illustrative purposes, pseudocolor images were spatially smoothed using a 3 x 3-pixel boxcar filter. To visualize the finest type of spines (cf. Price and Powell 1970), a 63x objective was ideal. However, the limited field of view made the mapping of calcium transients over the dendritic arbor impractical, given the large extent of dendrites (up to about 600 µm length) in adult turtles. Therefore a 40x objective was used. Often an intermediate lens of 0.5x was added in the light path to increase the field of view at the cost of resolving details, e.g., the finest spines, only clearly visualizing the larger types of spines (Price and Powell 1970). Even so, with large cells, the field of view was often restricted in relation to the extent of the dendrites. In such cases, a proximal region near the cell body and a distal dendritic region in the external plexiform layer (EPL), respectively, were selected for close analysis. Electronic storage of the X-Y-Z position of the microscope allowed multiple, precise movements between the two chosen regions following pharmacological manipulations. After the experiment, the patch electrode was withdrawn; the objective was changed to a 20x lens, and a series of overview pictures was acquired to show the extent of the dendrites. For illustration, a montage of the cell was produced, often by collapsing a Z-stack of pictures acquired with 2-µm separation into one plane with all dendrites in focus using Automontage (Syncroscopy).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
). For conventional current injection through the somatic recording electrode, the voltage-current relation shown in Fig. 1B was typical for granule cells (Hall and Delaney 2002; Pinato and Midtgaard 2003; Wellis and Kauer 1994). For intracellular current injection, the mean sodium spike amplitude was 33.98 ± 8.15 (SD) mV (20 cells), the duration was 3.91 ± 1.24 ms, and the threshold for activation was –25.28 ± 7.19 mV. A pulsed electric field was applied to the cell to polarize the soma versus the dendrites. AP-5 (30 µM), CNQX (50 µM), and SR95531 (10 µM) were added to the perfusate to block excitatory and inhibitory synaptic inputs. The cell was stimulated by different field intensities (Fig. 1, C and D) to evoke sub- and suprathreshold responses at opposite polarities, depolarizing the soma (Fig. 1C) or the dendrites (Fig. 1D). A field pulse of 60 V/m depolarized the soma below sodium spike threshold (Fig. 1C, left), while the same field strength with opposite polarity evoked somatic hyperpolarization (Fig. 1D, left). A field pulse of 120 V/m produced a sodium spike in the soma for somatic-depolarization polarity (Fig. 1C, middle) and somatic hyperpolarization accompanied by spikelets for dendritic-depolarization polarity (Fig. 1D, middle). Individual spikelets had an amplitude of 8.8 ± 4 mV (10 cells) and a duration of 8.6 ± 2 ms. The spikelets had a relatively slow decay phase, which contributed to a longer duration than the full sodium spikes. On average, the threshold field strength for sodium spikes was 142.5 ± 50.8 V/m for somatically evoked spikes and 151.55 ± 41.9 V/m for dendritically evoked spikelets. Spikelets were not observed during somatic field depolarization or on top of an LTS.
|
|
; see also Fig. 5). We cannot rule out that cells might differ in active and passive physiological properties in their ability to initiate dendritic spikelets, but difficulty in aligning the field across the slice and any deviation in the orientation of the dendritic branches from this axis could contribute to the inability to evoke spikelets in some cells with the field strengths available.
The occurrence and synchrony of spikelets induced by dendritic field depolarization could be controlled by fine-tuning the membrane potential through somatic current injection (Fig. 2). The occurrence of single spikelets was all-or-none as shown in the two superimposed lower traces of Fig. 2A, which were obtained during a field pulse of 45 V/m with somatic bias current injection (+23 pA). Increased somatic current injection (+30 pA, at the same field strength) led to an increase in the number of spikelets (Fig. 2A, top), occurring at a frequency (range, 150–250 Hz) that was much higher than the maximum frequency for somatic sodium spikes evoked by current injection (range, 10–30 Hz). When a burst of spikelets was evoked (Fig. 2B, bottom), further somatic depolarization reduced the delay to the first spikelet and evoked a full sodium spike, after which no spikelets were observed (Fig. 2B, top; same field strength in the 2 traces). The interaction between somatic and dendritic depolarization could be finely tuned by varying the amplitude of short current pulses delivered through the patch electrode during dendritic field depolarization (Fig. 2C; n = 5). In the control situation (Fig. 2C, left), dendritic field depolarization evoked a number of spikelets. When a small somatic depolarization was added during the field stimulation, the number of spikelets increased (Fig. 2C, middle), further increase in proximal depolarization resulted in a full sodium spike (Fig. 2C, right) triggered by the spikelets. The interaction along the dendrites did not only increase spiking: when a full spike, evoked by a somatic current pulse, preceded the field stimulation (Fig. 2D, thick line), the spikelets (control: Fig. 2D, thin line; n = 3) were abolished, suggesting that a single full spike evoked from the soma could spread up to the initiation site for dendritic spikes, preventing those from being activated by the dendritic field depolarization. This reduced excitability had a duration comparable with the interspike interval given by the maximum firing frequency (10–30 Hz), and could be due to dendritic potassium conductances elicited by sodium spike-evoked calcium influx (Egger et al. 2003; Fig. 5). In both traces of Fig. 2D, a rebound somatic spike was generated by the depolarization that followed the end of the field pulse (Fig. 2D, arrow). However, the preceding spikelets in the control experiment (Fig. 2D, thin line) did not prevent the rebound somatic full spike, even though it occurred within the time window where a preceding somatic full spike would abolish subsequent spikelet initiation. This supports the notion that spikelets were not actively invading the soma, and that there were independent sodium spike initiation sites proximally and distally in the cell.
|
) and thalamocortical neurons (Hughes et al. 2002) as gap junction–mediated potentials. Gap junctional coupling has been reported between granule cell somata in the olfactory bulb (Reyher et al. 1991; but see Kosaka and Kosaka 2003). Since somatic field depolarization readily evoked full spikes (Fig. 1), it could be expected that such spikes evoked in neighboring cells would be transmitted through somatic gap junctions as spikelets in the cell recorded from. However, spikelets were never observed with somatic field depolarization. To exclude that spikelets originated as gap junctional potentials of spikes in other cells, instead of dendritic spikes in the cell recorded from, carbenoxolone or octanol was added to the medium (100 µM, n = 3 and 1 mM, n = 4, respectively; data not shown); spikelets were still evoked by dendritic field depolarization under these conditions. Furthermore, when QX-314 (5 mM, n = 7; data not shown) was included in the electrode solution, the full spikes and subsequently the spikelets were abolished within 10 min following break-in. In addition, spikelet threshold was very sensitive to small variations of the membrane potential (Fig. 2, A–C), and a full spike could abolish subsequent spikelets (Fig. 2D). This suggests that the spikelets evoked by the dendritic field depolarization could not be postsynaptic gap junction potentials, reflecting the firing of the neighboring cells, but more likely they represented voltage-dependent active processes taking place within the cell recorded from. Furthermore, the granule cells would in any case be depolarized dendritically by the field, so that even if the spikelets were mediated by gap junctions they would likely reflect dendritic sodium spikes in other granule cells. Finally, sodium spikes are associated with a large calcium signal over the dendrites to the tips (Egger et al. 2003; Fig. 5), showing that the dendrites can support regenerative sodium spikes. In conclusion, spikelets most likely represent fast sodium spikes initiated at multiple sites in the granule cell dendrites and electrotonically transmitted to the soma (cf. Luo and Katz 2001; Mori and Takagi 1978; Wellis and Scott 1990). Spontaneous occurrence of spikelets
Some granule cells (5%) showed spontaneous spikelets (Fig. 3B), which were clearly distinct from spontaneous excitatory postsynaptic potentials (EPSPs; Fig. 3A). The rise time of the spontaneous spikelets was fitted by an exponential function (average time constant 
= 1.98 ± 1.14 ms, n = 20) and compared with the rise time of spontaneous EPSPs ( = 5.02 ± 2.98 ms, n = 20). Significant difference in the two populations was shown by paired t-test (P < 0.0005). Spikelets were also evoked by dendritic field depolarization (Fig. 3C), and their rising phase ( = 1.49 ± 1.03 ms, P > 0.4, n = 20) was not significantly different from that of the spontaneous ones. Groups of several spikelets (Fig. 3B, bottom) occurred spontaneously, as also observed during field depolarization (Fig. 2). Spontaneous full spikes occurring at rest could occasionally be preceded by fast prepotentials (Fig. 3D; see also Mori et al. 1982) whose rising phase was not significantly different from the one of the spikelets (1.60 ± 1.25 ms, P > 0.3, n = 20).
|
12 Hz) was not affected by the membrane potential, neither spikelets nor full spikes were detected at membrane potentials more negative than –50 mV. Above this value, spikelets started to occur at higher frequencies than full spikes, but their frequency suddenly decreased at membrane potentials around –45 mV at which full spikes were initiated. At about –25 mV, the cell fired only full spikes, and no more spikelets were observed. Spontaneous distal dendritic EPSPs can be difficult to detect in somatic recordings due to attenuation (Magee and Cook 2000; Williams and Stuart 2002), and such distal events could have initiated the spontaneously occurring spikelets. Initiation of spikelets by EPSPs
To test whether dendritic spikelets could be evoked systematically by stimulation of known synaptic inputs, extracellular stimulus electrodes (Fig. 4A, black triangle) were placed in the EPL for activation of excitatory synaptic input to the distal dendrites of granule cells. This was done in the presence of the GABAA antagonist SR95531 (10 µM) to reduce inhibitory synaptic input. The strength of the electrical stimulus was gradually increased showing that a spikelet would be evoked by the EPSP and eventually become a full spike by further increasing the stimulation (Fig. 4B). The stimulus intensity could be also fine-tuned to reveal a fast prepotential preceding full spike activation (Fig. 4B, right). The rise time of the prepotentials ( = 1.42 ± 0.13 ms, n = 12) was not significantly different from the field-evoked spikelets (P > 0.4). At the same stimulus intensity, but holding the cell at different prestimulus membrane potentials (Fig. 4C), a spikelet could be evoked on top of the EPSP, which would elicit a full sodium spike with a further change in bias current. The occurrence of the spikelets on top of the EPSPs was all-or-none (Fig. 4D, left), and similarly, spikelets/prepotentials could alternate with full spikes from sweep to sweep (Fig. 4D, right).
|
A LTS has been identified in olfactory granule cells (Pinato and Midtgaard 2003). The LTS has a significant influence on excitability and can boost sodium spike firing; conversely, the LTS may occur in isolation, subthreshold for the sodium spikes. Since different regions of the cell are functionally specialized, we wished to examine the contribution of different parts of the cell in the generation of the LTS. During an electrophysiologically identified LTS, fluorescence imaging at high magnification revealed LTS-associated calcium increases in the distal and proximal dendrites (Fig. 5, B and C, respectively). Thus an LTS could spread to all parts of the dendrites including the very distal tips in the EPL. This suggests that the LTS contributes to calcium-dependent processes in the dendrites even in the absence of sodium spikes. The time course of the fluorescence changes in relation to the electrophysiological events are shown at lower magnification for three different locations along the dendrites (Fig. 5, D–H). In all locations, including the distal dendrites in the EPL, a single LTS [Fig. 5F, full black line labeled LTS(TTX); Fig. 5G, LTS(F)] evoked sizable increases in intracellular calcium. The LTS was often crowned by a sodium spike (Fig. 5F: LTS + Na; Pinato and Midtgaard 2003) and a sodium spike combined with the LTS to evoke a larger calcium signal [Fig. 5G: dashed line LTS + Na(F)] than each of the spikes in isolation. The calcium signal elicited by a single sodium spike (Fig. 5E) is indicated by a dotted line [Na (E), Fig. 5G]. Depolarizations subthreshold for the LTS (Fig. 5F, control traces) did not evoke calcium transients [Fig. 5G, controls(F)]. The maximum amplitudes of the calcium transients are summarized in Table 1 for the proximal and the distal dendrites.
|
). Paired-pulse stimulation (Fig. 5H) revealed that the second, attenuated LTS was associated with a modest additional calcium signal [Fig. 5G, full gray line: LTS + LTS (H)] compared with the single LTS [Fig. 5G, full black line LTS(F)]. The contribution of the LTS to the intradendritic calcium concentration was therefore dependent on stimulation frequency, since this strongly influenced LTS amplitude.
The LTS generates a powerful depolarization, which boosts sodium spike initiation (Pinato and Midtgaard 2003). This suggests that the LTS has other effects than increasing the intradendritic calcium concentration. We wished to test whether local excitation could initiate an LTS, which would boost the local excitability. Depolarizing the cell body and simultaneously hyperpolarizing the dendrites with an electric field was a useful tool to test the relative contribution of distally versus proximally located T-channels to the somatically recorded LTS. By depolarizing the soma from –70 mV, it was possible to evoke the LTS (Fig. 6, average field strength for LTS activation: 144.7 ± 19.4 V/m, n = 4), reaching a peak amplitude similar to the LTS evoked by a current pulse. Dendritic field depolarization did not show conclusive electrophysiological evidence for a regenerative LTS (data not shown). However, we cannot exclude that the conditions used were not suitable for initiating and detecting such events, e.g., if the dendrites were not sufficiently prehyperpolarized, if the relative density of T-channels and dendritic potassium channels (Schoppa and Westbrook 1999) antagonized local LTS initiation or if an LTS, generated in the dendrites by a highly localized field depolarization (Chan and Nicholson 1986), could not be detected in somatic recordings because of electrotonic attenuation. The possibility of a regenerative LTS in the distal dendrites under these circumstances may be tested by voltage-sensitive dye imaging. The selective blockers mibefradil and NiCl2 (Lacinova et al. 2000; Pinato and Midtgaard 2003; Todorovic and Lingle 1998) had the same effect on both the field induced LTS and the LTS evoked by intracellular current injection (Fig. 6A). As previously described (Pinato and Midtgaard 2003), a refractory period followed the LTS evoked by paired pulses (Fig. 6B). The LTS evoked by somatic field depolarization was abolished when it was preceded by a current-pulse evoked LTS (Fig. 6, C and E). In contrast, a preceding field-evoked LTS (Fig. 6, D and E) left part of the current-pulse evoked LTS, but with a significantly reduced amplitude (49.3 ± 7.7%, P < 0.0001, n = 4). Since the somatic field depolarization was associated with a dendritic hyperpolarization, these results showed that the proximal part of the cell can sustain a regenerative LTS, which add to the excitability of this compartment. The interaction of the field and current pulse-evoked LTS suggests that the proximal parts of the cell contributes significantly to the amplitude and rising phase of the ordinary LTS evoked by proximal current injection. These results suggest that the LTS has a dual function and that the LTS localized proximally could have a major function in boosting excitability and initiating sodium spikes.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
Dendritic initiation of sodium spikelets
The data presented here (Figs. 1–4) are consistent with the notion that the spikelets are due to electrophysiological events in the cell recorded from e.g., dendritic sodium spikes. The dendritic field depolarization suggests that they can be initiated some distance from the cell body, since the field depolarization is strongest at the end of the neuronal process, but the exact dendritic initiation site(s) for the spikelets is (are) unknown. Spikelet amplitude at the initiation site is not known, but dendritic origin and passive decay is consistent with the prolonged decay phase of the spikelets compared with the full spikes. Patch recordings from the distal dendrites would help resolve this issue, but proved technically unfeasible during this study.
Multiple initiation sites for dendritic spikelets were inferred from their apparent firing frequency (bursts 250 Hz). This value was much higher than the maximum somatic sodium spike firing frequency (< 30 Hz) observed by injection of current pulses, and could be accounted for by a number of independent initiation sites. The possibility of multiple initiation sites may be analyzed further using voltage-sensitive dye imaging. Many dendrites originated close to the cell body and could be followed for long distances into the EPL (Figs. 1 and 5). Field depolarization of the apical dendritic tips may be particularly efficient for the induction of spikelets at multiple initiation sites (e.g., 1 spikelet per dendrite), because the accompanying proximal hyperpolarization could prevent dendritic spikes from propagating between dendrites originating below the neutral point for the field (Tranchina and Nicholson 1986). In this way, the field may be acting similar to a dynamic synaptic balance between distal excitation and proximal inhibition, which confined spiking to the distal dendrites in mitral cells (Chen et al. 1997, 2002). The results suggest that distal and proximal dendritic inhibition could be particularly efficient in controlling the spatial extent of lateral inhibition from a granule cell, by controlling initiation and propagation of sodium spikes.
Synaptic input to the distal dendrites activated sodium spikelets/prepotentials consistent with a dendritic origin of these active responses (Fig. 4; cf. Mori and Takagi 1978). In this, the results are reminiscent of the fast dendritic prepotentials recorded in mitral cell somata (cf. Chen et al. 2002; Shen et al. 1999). In contrast to their activation by distally located synapses, no spikelets/prepotentials were observed when a sodium spike was initiated by an LTS, suggesting that, under such circumstances, the sodium spike was initiated close to the somatic stimulation and recording site. During field depolarization of the distal dendrites, a short-lasting somatic depolarization would increase spikelet firing and allow the spikelets to trigger full somatic sodium spikes (Fig. 2C). Thus interactions along the somato–dendritic axis combined to produce the final spiking pattern of the cell.
Granule cells can be classified according to their dendritic branching pattern and synaptic connections with mitral and tufted cells (Mori et al. 1983). It remains to be seen whether the sodium spiking patterns may vary depending on the specific dendritic branching pattern of the cell.
Low-threshold calcium spike: contributions to calcium signaling and excitability
The calcium imaging experiments (Fig. 5) showed that calcium transients could be detected both in the proximal and the distal dendrites, when an electrophysiologically identified LTS was evoked. Thus the LTS contributed independently to calcium influx, in addition to T-type calcium channel activation during sodium spike firing (Chemin et al. 2002; Egger et al. 2003; Kozlov et al. 1999). The widespread LTS-evoked calcium signal suggests that the LTS in itself contribute to calcium dependent processes in the dendrites, e.g., transmitter release or synaptic plasticity. If the LTS was part of such functions, they would be affected by the properties of the LTS, e.g., the long refractory period (Fig. 5; Pinato and Midtgaard 2003). The amplitude of the LTS was modulated by prior local depolarization (Fig. 6), and it is likely that this, in combination with local excitatory and inhibitory synapses along the dendrites, controls the initiation, amplitude, and spatial extent of the LTS (cf. Destexhe 2000).
In adult turtles, the distribution of the LTS-evoked calcium transient in spines and stem-dendrites (Fig. 5) was similar to the distribution of calcium transients in mammalian granule cells, which were due to TTX-sensitive sodium spike-mediated calcium influx, about one-half of which was mediated by T-type channels (Egger et al. 2003). In other aspects, granule cells from different species appear rather similar (Hall and Delaney 2002; Pinato and Midtgaard 2003; Wellis and Kauer 1994). Further experiments may show whether calcium signals due to the LTS and the sodium spike have different consequences, for instance in the control of dendritic transmitter release, such as the functionally distinct but anatomically co-localized high- and low-threshold axonal spikes (Meech and Mackie 1993). However, the proximal dendrites and spines do not participate in presynaptic functions, as far as is known (Price and Powell 1970), and the role of the LTS- and sodium spike calcium increase here (Egger et al. 2003; Fig. 5) is not clear. Other sources of intracellular calcium that may interact with the LTS include high-threshold voltage-gated calcium channels (Isaacson 2001; Pinato and Midtgaard 2003), N-methyl-D-aspartate (NMDA) receptors (Chen et al. 2000), and release of calcium from internal stores; little is known about the latter in granule cells. An NMDA receptor (NMDAR) antagonist was present during the imaging experiments here, but it is conceivable that the LTS and the sodium spikes may cause different degrees of Mg2+ unblocking and provide different time windows for interactions with dendritic NMDARs (Kampa et al. 2004; Vargas-Caballero and Robinson 2003).
In addition to the role of LTS as a source of intracellular calcium, the LTS contributes to synaptic excitability and sodium spike initiation (Pinato and Midtgaard 2003). Therefore we analyzed the factors that influenced LTS excitability. The field experiments (Fig. 6) showed that local depolarization of the proximal cell membrane initiated a regenerative LTS, which influenced the amplitude of subsequent low-threshold spikes evoked by ordinary current pulses. Thus local control of T-channel activation/inactivation could influence subsequent cellular responses. In contrast to spikelets and prepotentials initiated by distal dendritic depolarization by the external field or distally located EPSPs, no such small spikes were observed during somatic field depolarization or on top of an LTS. This can be explained if the LTS or other proximal depolarization initiates sodium spiking at a proximal location, close to the recording site, from where the sodium spikes could propagate throughout the dendrites.
Functional considerations
While the LTS boosted granule cell excitability and increased intracellular calcium, the LTS did not contribute to membrane potential oscillations under the current experimental conditions (Pinato and Midtgaard 2003). Several functional possibilities exist for the LTS, which are not mutually exclusive. In other brain regions, local circuit dynamics are shaped by synapses showing depressant or facilitating behavior (Gupta et al. 2000). The properties of the LTS could allow olfactory granule cells to express both types of dynamics: at low levels of excitation, the refractory period of the LTS (Pinato and Midtgaard 2003) would likely appear as synaptic depression of the granule cell lateral inhibition, while higher levels of excitation would lead to repetitive sodium spiking, presumably resulting in a more persistent lateral inhibition. The results suggest that sodium spikelets had multiple initiation sites in granule cell dendrites. This imply that the spike-mediated functional coupling of the dendrites is controllable (cf. Baccus et al. 2000; Combes et al. 1993; Golding and Spruston 1998): depending on synaptic interactions along the dendrites, each dendrite would act independently or act in concert with other dendrites in a more uniformly activated cell. Sodium spikelets/prepotentials could be synaptically activated (Fig. 5), and spikelets are observed in vivo during olfactory stimulation (Luo and Katz 2001), showing that naturally occurring synaptic activity influence spike initiation and propagation. This fine-tuning of sodium spiking makes lateral inhibition from granule cells sensitive to the context of excitation and inhibition along their distal and proximal dendrites.
|
GRANTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
FOOTNOTES |
|---|
Address for reprint requests and other correspondence: J. Midtgaard, (E-mail: j.midtgaard{at}mfi.ku.dk)
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
Chan CY and Nicholson C. Modulation by applied electric fields of Purkinje and stellate cell activity in the isolated turtle cerebellum. J Physiol 371: 89–114, 1986.
Chemin J, Monteil A, Perez-Reyes E, Bourinet E, Nargeot J, and Lory P. Specific contribution of human T-type calcium channel isotypes (alpha(1G), alpha(1H) and alpha(1I)) to neuronal excitability. J Physiol 540: 3–14, 2002.
Chen WR, Midtgaard J, and Shepherd GM. Forward and backward propagation of dendritic impulses and their synaptic control in mitral cells. Science 278: 463–467, 1997.
Chen WR, Shen GY, Shepherd GM, Hines ML, and Midtgaard J. Multiple modes of action potential initiation and propagation in mitral cell primary dendrite. J Neurophysiol 88: 2755–2764, 2002.
Chen WR, Xiong W, and Shepherd GM. Analysis of relations between NMDA receptors and GABA release at olfactory bulb reciprocal synapses. Neuron 25: 625–633, 2000.[CrossRef][ISI][Medline]
Combes D, Simmers J, Nonnotte L, and Moulins M. Tetrodotoxin-sensitive dendritic spiking and control of axonal firing in a lobster mechanoreceptor neurone. J Physiol 460: 581–602, 1993.
Destexhe A. Modelling corticothalamic feedback and the gating of the thalamus by the cerebral cortex. J Physiol Paris 94: 391–410, 2000.[CrossRef][ISI][Medline]
Egger V, Svoboda K, and Mainen ZF. Mechanisms of lateral inhibition in the olfactory bulb: efficiency and modulation of spike-evoked calcium influx into granule cells. J Neurosci 23: 7551–7558, 2003.
Golding NL and Spruston N. Dendritic sodium spikes are variable triggers of axonal action potentials in hippocampal CA1 pyramidal neurons. Neuron 21: 1189–1200, 1998.[CrossRef][ISI][Medline]
Gupta A, Wang Y, and Markram H. Organizing principles for a diversity of GABAergic interneurons and synapses in the neocortex. Science 287: 273–278, 2000.
Hall BJ and Delaney KR. Contribution of a calcium-activated non-specific conductance to NMDA receptor-mediated synaptic potentials in granule cells of the frog olfactory bulb. J Physiol 543: 819–834, 2002.
Hughes SW, Blethyn KL, Cope DW, and Crunelli V. Properties and origin of spikelets in thalamocortical neurones in vitro. Neuroscience 110: 395–401, 2002.[CrossRef][ISI][Medline]
Isaacson JS. Mechanisms governing dendritic gamma-aminobutyric acid (GABA) release in the rat olfactory bulb. Proc Natl Acad Sci USA 98: 337–342, 2001.
Jahr CE and Nicoll RA. Dendrodendritic inhibition: demonstration with intracellular recording. Science 207: 1473–1475, 1980.
Kampa BM, Clements J, Jonas P, and Stuart GJ. Kinetics of Mg2+ unblock of NMDA receptors: implications for spike-timing dependent synaptic plasticity. J Physiol 556: 337–345, 2004.
Kosaka T and Kosaka K. Neuronal gap junctions in the rat main olfactory bulb, with special reference to intraglomerular gap junctions. Neurosci Res 45: 189–209, 2003.[CrossRef][ISI][Medline]
Kozlov AS, McKenna F, Lee JH, Cribbs LL, Perez-Reyes E, Feltz A, and Lambert RC. Distinct kinetics of cloned T-type Ca2+ channels lead to differential Ca2+ entry and frequency-dependence during mock action potentials. Eur J Neurosci 11: 4149–4158, 1999.[CrossRef][ISI][Medline]
Lacinova L, Klugbauer N, and Hofmann F. Low voltage activated calcium channels: from genes to function. Gen Physiol Biophys 19: 121–136, 2000.[ISI][Medline]
Luo M and Katz LC. Response correlation maps of neurons in the mammalian olfactory bulb. Neuron 32: 1165–1179, 2001.[CrossRef][ISI][Medline]
Magee JC and Cook EP. Somatic EPSP amplitude is independent of synapse location in hippocampal pyramidal neurons. Nat Neurosci 3: 895–903, 2000.[CrossRef][ISI][Medline]
Meech RW and Mackie GO. Ionic currents in giant motor axons of the jellyfish, Aglantha digitale. J Neurophysiol 69: 884–893, 1993.
Midtgaard J. Processing of information from different sources: spatial synaptic integration in the dendrites of vertebrate CNS neurons. Trends Neurosci 17: 166–173, 1994.[CrossRef][ISI][Medline]
Midtgaard J. Spatial synaptic integration in Purkinje cell dendrites. J Physiol Paris 89: 23–32, 1995.[CrossRef][ISI][Medline]
Mori K, Kishi K, and Ojima H. Distribution of dendrites of mitral, displaced mitral, tufted, and granule cells in the rabbit olfactory bulb. J Comp Neurol 219: 339–355, 1983.[CrossRef][ISI][Medline]
Mori K, Nowycky MC, and Shepherd GM. Electrophysiological analysis of mitral cells in the isolated turtle olfactory bulb. J Physiol 314: 281–294, 1981.
Mori K, Nowycky MC, and Shepherd GM. Impulse activity in presynaptic dendrites: analysis of mitral cells in the isolated turtle olfactory bulb. J Neurosci 2: 497–502, 1982.[Abstract]
Mori K and Takagi SF. An intracellular study of dendrodendritic inhibitory synapses on mitral cells in the rabbit olfactory bulb. J Physiol 279: 569–588, 1978.
Pinato G and Midtgaard J. Regulation of granule cell excitability by a low-threshold calcium spike in turtle olfactory bulb. J Neurophysiol 90: 3341–3351, 2003.
Price JL and Powell TP. The morphology of the granule cells of the olfactory bulb. J Cell Sci 7: 91–123, 1970.
Reyher CK, Lubke J, Larsen WJ, Hendrix GM, Shipley MT, and Baumgarten HG. Olfactory bulb granule cell aggregates: morphological evidence for interperikaryal electrotonic coupling via gap junctions. J Neurosci 11: 1485–1495, 1991.[Abstract]
Schmitz D, Schuchmann S, Fisahn A, Draguhn A, Buhl EH, Petrasch-Parwez E, Dermietzel R, Heinemann U, and Traub RD. Axo-axonal coupling. a novel mechanism for ultrafast neuronal communication. Neuron 31: 831–840, 2001.[CrossRef][ISI][Medline]
Schoppa NE and Westbrook GL. Regulation of synaptic timing in the olfactory bulb by an A-type potassium current. Nat Neurosci 2: 1106–1113, 1999.[CrossRef][ISI][Medline]
Shen GY, Chen WR, Midtgaard J, Shepherd GM, and Hines ML. Computational analysis of action potential initiation in mitral cell soma and dendrites based on dual patch recordings. J Neurophysiol 82: 3006–3020, 1999.
Shepherd GM, Chen WR, and Greer CA. Olfactory bulb. In: The Synaptic Organization of the Brain, edited by Shepherd GM. New York: Oxford, 2004, p. 165–216.
Todorovic SM and Lingle CJ. Pharmacological properties of T-type Ca2+ current in adult rat sensory neurons: effects of anticonvulsant and anesthetic agents. J Neurophysiol 79: 240–252, 1998.
Tranchina D and Nicholson C. A model for the polarization of neurons by extrinsically applied electric fields. Biophys J 50: 1139–1156, 1986.
Vargas-Caballero M and Robinson HP. A slow fraction of Mg2+ unblock of NMDA receptors limits their contribution to spike generation in cortical pyramidal neurons. J Neurophysiol 89: 2778–2783, 2003.
Wellis DP and Kauer JS. GABAergic and glutamatergic synaptic input to identified granule cells in salamander olfactory bulb. J Physiol 475: 419–430, 1994.
Wellis DP and Scott JW. Intracellular responses of identified rat olfactory bulb interneurons to electrical and odor stimulation. J Neurophysiol 64: 932–947, 1990.
Williams SR and Stuart GJ. Dependence of EPSP efficacy on synapse location in neocortical pyramidal neurons. Science: 295: 1907–1910, 2002.
Woolf TB, Shepherd GM, and Greer CA. Local information processing in dendritic trees: subsets of spines in granule cells of the mammalian olfactory bulb. J Neurosci 11: 1837–1854, 1991.[Abstract]
Yokoi M, Mori K, and Nakanishi S. Refinement of odor molecule tuning by dendrodendritic synaptic inhibition in the olfactory bulb. Proc Natl Acad Sci USA 92: 3371–3375, 1995.
This article has been cited by other articles:
|
|
 |
|
  G. Pinato, J. Rievaj, S. Pifferi, M. Dibattista, L. Masten, and A. Menini Electroolfactogram Responses from Organotypic Cultures of the Olfactory Epithelium from Postnatal Mice Chem Senses, April 1, 2008; 33(4): 397 - 404. [Abstract] [Full Text] [PDF]
|
|
