| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
The Journal of Neurophysiology Vol. 80 No. 4 October 1998, pp. 2089-2101
Copyright ©1998 by the American Physiological Society
Kinsmen Laboratory of Neurological Research, Departments of Psychiatry and Physiology, University of British Columbia, Vancouver V6T 1Z3, Canada
 |
ABSTRACT |
|---|
 Abstract
 Introduction
Methods
Results
Discussion
References
|
|---|
Mackenzie, Paul J. and Timothy H. Murphy. High safety factor for action potential conduction along axons but not dendrites of cultured hippocampal and cortical neurons. J. Neurophysiol. 80: 2089-2101, 1998. By using a combination of Ca2+ imaging and current-clamp recording, we previously reported that action potential (AP) conduction is reliably observed from the soma to axonal terminals in cultured cortical neurons. To extend these studies, we evaluated Ca2+ influx evoked by Na+ APs as a marker of AP conduction under conditions that are expected to lower the conduction safety factor to explore mechanisms of axonal and dendritic excitability. As expected, reducing the extracellular Na+ concentration from 150 to ~60 mM decreased the amplitude of APs recorded in the soma but surprisingly did not influence axonal conduction, as monitored by measuring Ca2+ transients. Furthermore, reliable axonal conduction was observed in dilute (20 nM) tetrodotoxin (TTX), despite a similar reduction in AP amplitude. In contrast, the Ca2+ transient measured along dendrites was markedly reduced in low Na+, although still mediated by TTX-sensitive Na+ channels. Dendritic action-potential evoked Ca2+ transients were also markedly reduced in 20 nM TTX. These data provide further evidence that strongly excitable axons are functionally compartmentalized from weakly excitable dendrites. We conclude that modulation of Na+ currents or membrane potential by neurotransmitters or repetitive firing is more likely to influence neuronal firing before AP generation than the propagation of signals to axonal terminals. In contrast, the relatively low safety factor for back-propagating APs in dendrites would suggest a stronger effect of Na+ current modulation.
A number of recent studies contributed to the emerging view that neurons are comprised of compartments of differing electrical excitability, both in the generation and propagation of active conductances. Dual-electrode and imaging studies suggest that action potential (AP) generation does not normally occur in dendrites (Häusser et al. 1995 Hippocampal and cortical neurons and glia were dissociated from 17- to 18-day gestation rat fetuses (separate cultures were prepared from the 2 regions) placed in culture and allowed to mature for 17-26 days in vitro as previously described (Murphy and Baraban 1990 Perfusion of low-sodium solution
Computer-triggered solenoid valves (Neptune Research, Northboro, MA) were used to control a gravity-fed double-barreled
Imaging and analysis
A fiberoptically coupled intensified charged-coupled device camera with a Gen III intensifier tube was used for all experiments (Stanford Photonics, Palo Alto, CA) in combination with a Epix 4 M12-64 MB frame grabber board (Northbrook, IL). Images and electrophysiological data were analyzed together on a Pentium computer with custom routines written in IDL (Research Systems, Boulder, CO). Change in fluo-3 fluorescence (excitation 480/40 nm) was used to measure axonal and dendritic Ca2+ dynamics in response to Na+ APs elicited at the soma (Jaffe et al. 1992
Reliable axonal conduction in 150 and 60 mm Naext
Conduction reliability along axonal and dendritic processes was monitored by measuring Ca2+ transients in these compartments after AP generation at the soma (Fig. 1A). Figure 1B illustrates 38 overplotted AP waveforms in 60 and 150 mM extracellular Na+ (alternating trials). Delivery of extracellular solution containing 60 Na+ resulted in a reversible 60% decrease in the amplitude of the AP. Figure 1C illustrates a typical experimental sequence, with alternating trials of one and two APs (1 and 2 AP) in 60 and 150 Na+. For two AP the interval between the steps was 50 ms. In 60 Na+, input resistance to somatic current steps was not significantly different from control solution (Fig. 1, D and E), and basal fluo-3 fluorescence was not appreciably altered in 60 mM Na+ because of the rapid and reversible perfusion (Fig. 1F, and see METHODS).
Axonal sites have sufficient signal to noise ratio to detect intermittent failure
To determine whether the variance analysis would have been sensitive to occasional conduction failure, we created simulated failure data sets from 150 Na+ data. For each axonal site, the mean Ca2+ response was subtracted from a single trial to create data with one (artificial) failure (Fig. 4C). As predicted, adding one simulated failure to the control (150 Na+) data set only minimally perturbed the mean response (not reaching statistical significance by t-test). In contrast to the analysis of the mean, creation of a single failure resulted in a significant increase in variance at 38 of 105 axonal sites from 8 neurons (4 hippocampal and 4 cortical), indicating that it was possible to detect a single failure at these sites had one occurred. This analysis indicated that the 38 sites had signal to noise properties that were sufficient to detect a single failure. When 60 Na+ medium was compared with 150 Na+ medium (simulated failure removed), only 3 of the 38 axonal sites showed a significant increase in variance (P < 0.05). Thus, at most sites where we were confident that we would have reliably measured increased variance, no increase in variance was seen in low Na+, suggesting no increase in the number of conduction failures. Analysis of conduction reliability of two APs was carried out similarly, except that single failure simulated data sets were created by subtracting one-half of the mean response from one trial (equivalent to 1 AP). For 2 APs, 15 sites from 5 neurons were found to have sufficient signal to noise to permit detection of a failure of 1 AP. No significant difference in variance was observed when APs evoked in 60 and 150 Na+ medium were compared at these sites. Therefore, at most axonal sites that were determined to be capable of detecting conduction failure, no indicative increase in variance was observed.
Dendritic calcium transients evoked by APs are reduced by lowering Naext
Because axonal conduction (as assayed by Ca2+ imaging) was reliable under conditions of lower external Na+, we next investigated whether propagation of single or pairs of APs along dendrites was altered by similar manipulations. Figure 5A, left panel, shows a basal fluorescence image (furaptra excitation wavelength) of a representative pyramidal neuron indicating the main dendrite where Ca2+ measurements were made. Ca2+ transients evoked by single and paired APs were measured along the length of the dendrite (with the 480-nm excitation wavelength of fluo-3 for dynamic Ca2+ measurements). Figure 5B illustrates traces of average dendritic Ca2+ transient (evoked by 2 AP) versus time in 150 and 60 Na+. In 150 Na+, dendritic Ca2+ transients decreased with distance from the soma with a length constant of 150 µm (Fig. 5C). However, the decrement with distance of the Ca2+ response along the dendrite was greater in 60 Na+ than in 150 Na+ (length constant 53 µm, Fig. 5D) after either one or two AP.
Ca2+ transients in distal dendrites are mediated by TTX-sensitive Na+ channels
Because low Na+ medium selectively decreased the Ca2+ responses along the dendrite, we sought to further characterize the residual dendritic Ca2+ response. To explore whether the remaining Ca2+ response in 60 Na+ depended on regenerative potentials mediated by TTX-sensitive Na+ channels, dendritic Ca2+ transients to subthreshold and suprathreshold stimulation were measured in 1 µM TTX. Figure 7A illustrates the Ca2+ response averaged over a region of dendrite 30-80 µm from the soma (more distal regions would produce a weak signal). The threshold for AP generation in this cell, determined before addition of TTX, and again after washout, was ~600 pA (5 ms of current injection). Figure 7A illustrates the average dendritic Ca2+ response, in TTX, of one or two 5-ms current steps of 300, 700, or 1,000 pA. Ca2+ responses to 300-pA current injection were not significantly different from baseline (P > 0.30, t-test). After 700- and 1,000-pA stimulation, Ca2+ responses were significantly above baseline variation (P < 0.05, t-tests). After washout of TTX, suprathreshold stimulation that now generated APs resulted in Ca2+ responses that were 15- to 20-fold greater than responses in TTX, as illustrated in Fig. 7A, (note 6-fold change in scale on y-axis). Again, no measurable Ca2+ responses were observed after one or two 300-pA current steps (apparently below threshold). Figure 7B illustrates the spatial profile of Ca2+ responses along the same region of dendrite after two current steps of 1,000 pA in TTX, 700 pA after washout (suprathreshold: 2 AP), and 300 pA after washout (subthreshold). Figure 7B illustrates that Ca2+ responses to somatic current injection in TTX were markedly reduced along the length of the dendrite (Fig. 7B, inset). Thus the spatially decreasing responses observed under control and low Na+ conditions were dependent on TTX-sensitive Na+ channels. Similar results were observed in two other cells. A Ca2+ spike did not contribute an appreciable component to the residual Ca2+ response nor was there evidence of a Ca2+ spike in the somatic voltage records (data not shown).
Conduction safety factor in axonal and dendritic compartments in dilute TTX
To address potential concerns that the low Na+ manipulation could also affect other Na+-dependent processes such as pumps and exchangers (Bouron and Reuter 1996 Reliable axonal AP propagation in neurons from hyperpolarized potentials
A-type K+ channels have been shown to regulate excitability in dendrites of CA1 pyramidal neurons by limiting AP generation and limiting AP back-propagation in dendrites (Hoffman et al. 1997
We explored the propagation of APs along axons and dendrites of hippocampal and cortical neurons by measuring Ca2+ transients along distal regions of axon and dendrite following somatic current steps. We previously reported that single somatic APs reliably result in Ca2+ influx along all branches of axons. This Ca2+ influx occurred predominantly through high-voltage activated Ca2+ channels and was enhanced at distal functional release sites along secondary and tertiary axonal branches (Mackenzie et al. 1996 Reliable axonal conduction of truncated APs in low Na+ or dilute TTX
All axonal sites continued to show reliable Ca2+ transients after somatic APs in reduced extracellular Na+; this was observed on distal portions of axon as far as we could resolve (>400 µm from the soma) on both main and collateral branches (Figs. 2 and 3). Despite a substantial decrease in AP amplitude in 60 Na+, Ca2+ transients were still reliably observed at axonal terminals (Fig. 1, B and C). Two analyses were used to test for the occasional failure of the axonal Ca2+ response to a single AP. First, we analyzed 38 sites with sufficient signal to noise to detect intermittent failure (Fig. 4C); at these sites variance analysis indicated that intermittent AP failure does not occur, which is consistent with Dityatev and Clamann (1996) Decreased dendritic AP propagation in low external sodium
We measured Ca2+ transients caused by back-propagating APs along the main dendrite of cortical and hippocampal neurons in 150 and 60 Na+. In proximal dendritic regions (~30-50 µm from the soma), the Ca2+ transients were similar in magnitude to Ca2+ responses in the axon (Figs. 2 and 5), consistent with reports by Calleweart et al. (1996) and Llano et al. (1997) Electrical compartmentalization of the neuron
Previous studies contributed to the emerging view that the neuron is composed of compartments of differing electrical excitability (for review see Stuart et al. 1997
INTRODUCTION
Abstract
Introduction
Methods
Results
Discussion
References
; Stuart and Sakmann 1994); in contrast, the axon at or near the hillock is the site of generation, suggesting greater excitability in the axon. APs that are initiated in the axon can invade dendrites and back-propagate along them in a manner that is sensitive to branching and decrements over distance (Stuart et al. 1997). In contrast, single APs conduct reliably along axons to terminals, indicating that the distal axon is also strongly excitable (Allen and Stevens 1994; Lüscher et al. 1996; Mackenzie et al. 1996; Stevens and Wang 1995).
; Wall 1995). Second, Na+ channels, which play a dominant role in AP generation and propagation, can be modulated by second messengers, including protein kinases and G-proteins (Catterall 1993; Li et al. 1993; Ma et al. 1994). Third, several recent reports suggest the possibility that blockade of AP conduction may contribute to fluctuations of synaptic responses. For example, intense activity in principal cells of the cerebellum and hippocampus results in decreased inhibition onto those cells from presynaptic inhibitory cells (Llano et al. 1991; Pitler and Alger 1994). In hippocampus, for example, intense postsynaptic activity can lead to failure of evoked but not spontaneous inhibitory postsynaptic currents through postsynaptic Ca2+ influx, a retrograde messenger, and possibly involving presynaptic axonal conduction block (Alger et al. 1996). In cerebellar Purkinje cells, fluctuations of inhibitory currents can occur as a result of a presynaptic mechanism downstream of AP generation, possibly including fluctuations in presynaptic depolarization (Vincent and Marty 1996).
; Magee and Johnston 1995; Schwindt and Crill 1995; Stuart and Sakmann 1994). By using Ca2+ imaging, we assessed the excitability of the CNS axon under limiting conditions expected to reduce the safety factor (lowering of the external Na+ concentration). These fine distal axonal branches of mammalian CNS neurons are not directly accessible to patch-clamp or intracellular electrodes. Although brain slice preparations would be preferable for this study, Frenguelli and Malinow (1996) found that high concentrations of extracellular Ca2+ and intracellular fura-2 were required to reliably view Ca2+ dynamics in single boutons in response to a single AP in a slice preparation. Studies performed in cerebellar slices (Callewaert et al. 1996) with more physiological conditions were restricted to initial segments of axons and required averaging of trials and were unable to resolve single boutons.
), we observe that despite a 60% reduction in somatic spike amplitude reliable conduction of single and pairs of APs was maintained along main and collateral branches of axons. In contrast, along dendrites, although active propagation persisted, the Ca2+ transient, an indicator of the degree of depolarization, was greatly reduced. These results suggest that along the axon there is a large degree of conduction safety leading to reliable terminal depolarization and Ca2+ influx. Consequently, agents that modify Na+ channel function, such as modulating neurotransmitters, are unlikely to exert their effects by altering axonal conduction reliability. In the dendrite, a graded relationship exists between the weakly propagating AP and Ca2+ influx.
METHODS
Abstract
Introduction
Methods
Results
Discussion
References
), with the exception that the plating medium L-cystine concentration was supplemented to 300 µM. The whole cell recording configuration in current-clamp mode (Hamill et al. 1981) was used to fill neurons with fluo-3/furaptra (Minta et al. 1989; Ragu et al. 1989) and to generate APs by current injection. The patch pipette solution contained (in mM) 0.75 fluo-3 K+ salt, 1 furaptra (used to view basal fluorescence), 122 K+MeSO4, 20 NaCl, 5 Mg-ATP, 0.3 GTP, and 10 N-2-hydroxyethylpiperazine-N'-2-ethanesulfinic acid (HEPES), pH 7.2. Imaging was performed with a 100 × 1.3 NA Zeiss objective on a Zeiss Axiovert microscope with a custom-made stage that permits movement of the preparation during patch-clamp recording. The following extracellular solution was used to isolate the effects of AP stimulation from spontaneous synaptic activity (in mM): 137 NaCl, 5.0 KCl, 2.5 CaCl2, 1 MgCl2, 0.34 Na2HPO4(7H2O), 10 NaHEPES, 22 glucose, 1 NaHCO3, 20 µM picrotoxin, 100 µM D,L-2-amino-5-phosphonovaleric acid, and 3 µM 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX), pH 7.4.
55 to 65 mV; a calculated liquid junction potential of 12 mV was not corrected for (Neher 1995a). Membrane potential was adjusted to 60 mV by injection of a small negative or positive current, and overshooting APs were produced by injection of positive current (5- to 10-ms duration). In all trials, records of membrane potential were inspected carefully to determine whether the AP changed shape over time or whether stimulation failed to elicit APs. Experiments were terminated if a progressive change in the AP shape or the membrane potential occurred. After establishment of whole cell configuration, data were not collected for 10-15 min to allow diffusion of the Ca2+ indicators into fine processes. Pyramidal-type neurons were used for imaging experiments and were identified by their somatic size, shape, and prominent dendritic process. The axon of the neuron was identified according to the criteria outlined by Mackenzie et al. (1996) and positioned optimally for imaging.
tube (900-µm diam; flow rate 0.75 ml/min) containing normal and low Na+ solutions. By using a ×2.5 objective and a phenol red containing test solution, the tube was placed~4 mm from the center of the chamber so that its contents rapidly exchanged the bath solution of an area (~4 mm2) surrounding the cell of interest. As shown in Fig. 1A (not to scale), a compartment of the tube contained normal Na+ (150 Na+) Hanks balanced salt solution (HBSS); the other side contained low Na+ HBSS (40 mM; ionic substitution with N-methyl-D-glucamine). Figure 1 illustrates that the local perfusion system is able to alter the ionic composition of the extracellular environment surrounding the neuron; this was rapid and reversible. Examination of somatic records of membrane potential revealed that the perfusion system established a low Na+ environment around the soma of the cell (Fig. 1B, note truncated AP). However, it was also important that the solution was exchanged on distal axonal and dendritic processes. The following points argue that distal axonal and dendritic processes experienced rapid and reversible solution exchange. First, we used a large-diameter tube that perfused the solution and hence quickly exchanged it, over several millimeters, clearly encompassing the areas where most distal measurements were made (<500 µm from the soma). Second, after placement of the cell of interest into the field of view but before establishing a whole cell recording, we used a phenol red-containing HBSS in one barrel of the perfusion system to confirm both rapid delivery from one barrel and efficient clearance from the other. Third, the dendrite served as a positive control because dendritic regions that were as distal as axonal regions imaged were also affected by low Na+ perfusion. Fourth, a zero Ca2+ solution used in the perfusion system blocked Ca2+ transients evoked by APs in distal branches without altering AP generation in six of six neurons (data not shown). Trials of 150 Na+ and 60 Na+ were alternated with a 15-s intertrial interval; solution exchange for the following trial occurred at the beginning of this 15-s period. (By using the phenol red-containing medium, we calculated that the [Na+]ext. after perfusion was 80 mM within 330 ms and reached a steady state of 60 mM within 2.5 s.) Between trials of data acquisition, cells were bathed with buffer containing 150 Na+. By using the double-barreled perfusion system, we observed a small (2-5%) increase in basal fluo-3 fluorescence under low Na+ conditions (presumably caused by blockade of Na+/Ca2+ exchanger). This increase in basal fluorescence was ~10% of the increase observed in response to 1 AP and therefore did not appreciably saturate the Ca2+ indicator. This is illustrated for one axonal bouton in Fig. 1 (F), which shows changes in raw fluo-3 fluorescence in response to two somatic APs for an axonal site 174 µm from the soma. In contrast, in preliminary experiments in which continuous (bath) application of low Na+ medium was used, we observed a robust rise in basal Ca2+ concentration, thus requiring the tube application method.
View larger version (36K):
[in a new window]
FIG. 1.
Local perfusion of 60 mM extracellular Na+ reversibly reduces action potential (AP) amplitude. A: cartoon of reversible perfusion system (not to scale). A large-bore double-barreled tube permitted rapid exchange of the extracellular solution surrounding the neuron of interest and its processes. Calcium indicators were loaded with intracellular dialysis, and records of membrane potential were made in the current-clamp recording mode. B: low Na+ solution reversibly reduced the AP amplitude. Traces of membrane potential vs. time overplotted for 38 alternating trials of 150 and 60 mM extracellular Na+. C: records of membrane potential recorded at the soma during a typical experimental sequence illustrating the reversible delivery of the low Na+ solution. Trials of 1 or 2 AP were alternated between 150 and 60 mM external Na+. D: input resistance was not significantly different between 60 and 150 mM extracellular Na+. Records of somatic membrane potential after delivery of 500-ms current steps of 400, 200, 0, 200, and 400 pA in 150 and 60 mM Na+ and wash (150 mM Na+). E: voltage-current relationship showing steady-state membrane potential as a function of current step (incrementing by 100 pA). Input resistance in 60 Na+ was 98.8 ± 2.7% of input resistance in 150 Na+ (n = 3 cells, range 144-218 M
). F: traces illustrating raw fluo-3 fluorescence vs. time in the axon 174 µm from the soma after 2 somatic APs 50 ms apart. The thick trace is the mean response, and thin traces are 12 overplotted trials.
; Markram et al. 1995). These intracellular Ca2+ transients reflect AP conduction through an imaged axonal segment because of the following observations: complete block of axonal Ca2+ transients in 1 µM TTX (Fig. 8B), the Ca2+ signal was proportional to the number of APs delivered (Fig. 2C) (Mackenzie et al. 1996), the Ca2+ transients were blocked by removing extracellular Ca2+ (Mackenzie et al. 1996), and a lack of Ca2+ transients after stimulation that was subthreshold for AP generation (Figs. 2C and 6A) (Mackenzie et al. 1996). Furthermore, it is likely that, given the high indicator concentrations used, the Ca2+ signal is dominated by Ca2+ influx rather than handling (Neher 1995b). To view processes under baseline conditions and to correct for process volume and cell loading, the fluorescent Ca2+ indicator furaptra (excitation 380 nm) was coinjected with fluo-3 into cells. Furaptra was chosen for its high fluorescence signal at basal Ca2+, its low affinity for Ca2+ that produces little buffering, and its well-distinguished spectra that did not contaminate fluo-3 signals (Ragu et al. 1989). Plots of fluo-3 fluorescence versus time were produced off-line from single video frames corresponding to 33-ms intervals. After analysis, data were discarded if basal Ca2+ levels were found to rise significantly during the course of an experiment or if progressive changes in the amplitude of the Ca2+ transient were observed because of factors such as spike broadening. The integral of the Ca2+ transient was termed the Ca2+ response. Ca2+ responses were calculated for individual trials for 1.4 × 1.4 µm2 regions along dendrites and axons (including individual terminals). Percentage change in fluo-3 fluorescence (
F/F) × 100 was calculated by dividing the change in fluo-3 fluorescence (mean F over 333 ms after stimulation baseline F) by the mean fluo-3 fluorescence measured during baseline. This F/F measurement was used for analyses in which fluctuations were measured within individual sites across trials. For analyses of the spatial profile of axonal and dendritic Ca2+ transients, responses were corrected for differences in compartment volume and potential irregularities in illumination or detection by dividing the change in fluo-3 fluorescence by background-subtracted basal furaptra fluorescence and expressed as a ratio of change in fluo-3 divided by furaptra fluorescence (F480/F380). A single exponential fit was used to determine the decrement of the Ca2+ transient along a region of dendrite.
View larger version (28K):
[in a new window]
FIG. 8.
Selective impairment of dendritic but not axonal AP evoked Ca2+ response in 20 nM TTX. A, top: records of average change in Ca2+ response fluo-3/furaptra; F480/F380) vs. time in response to 1 AP for 3 dendritic sites at varying lengths from the soma. Baseline: dashed lines; TTX 20 nM: thick solid lines; Washout: thin solid lines. Below: average change in fluo-3 fluorescence versus time in response to 1 AP at an axonal segment 270 µm from the soma. Line symbols as in dendritic traces. Axonal response to current step in 1 µM TTX indicated by arrow. B: group data (mean ± SE) for axon (left, 5 cells) and dendrite (right, 5 cells).
View larger version (51K):
[in a new window]
FIG. 2.
Reliable AP-evoked Ca2+ transients along the axon in low extracellular Na+. A: basal fluorescence image of a region of axon 80-220 µm from the soma of a cortical pyramidal neuron. B: records of Ca2+ response ( fluo-3/basal furaptra; F480/F380) vs. time for 3 axonal sites (1.4 × 1.4 µm), overplotted for 7-9 trials of 1 AP, 150 vs. 60 mM Na+ (left panel); 2 AP, 150 vs. 60 mM Na+ (right panel). C: Ca2+ response integral ( fluo-3/basal furaptra; F480/F380; y-axis) plotted over repeated trials (x-axis) for axonal site 3. Ca2+ transients were evoked by 1 and 2 APs (top traces) in 150 mM Na+ (
) and 60 mM Na+ (
). Dashed line represents 2 SDs above baseline variation.
View larger version (19K):
[in a new window]
FIG. 6.
Group data illustrating the relative change in the action-potential evoked Ca2+ transient in 60 vs. 150 mM Na+ along axons and dendrites. A: axon. Average axonal Ca2+ response in 60 mM Na+ as a function of distance from the soma, expressed as a ratio of response in 60 to response in 150 mM Na+. 
, 2 APs; 
, 1 AP (8 cells). B: dendrite. Average dendritic Ca2+ response ratio (60 to 150 mM Na+), as a function of distance from the soma (6 cells). , 2 APs; , 1 AP.
RESULTS
Abstract
Introduction
Methods
Results
Discussion
References
). In 60 mM Na+, Ca2+ transients persisted despite a truncation of the AP, suggesting reliable axonal conduction in 60 Na+. Figure 2C plots the integral of the Ca2+ transient across trials for one and two APs in 60 and 150 Na+ for an axonal site. Ca2+ responses in 150 and 60 Na+ were >2 SDs above baseline variation (Fig. 2C). To investigate whether axonal Ca2+ responses in 150 Na+ varied with distance, we performed a regression analysis (Ca2+ response vs. distance) on 27 sites from 4 cells that were within 90-175 µm of the soma. The slope of the regression line was 0.08 ± 0.1%/µm (i.e., 8% increase in F480/F380 over 100 µm), indicating a small but nonsignificant tendency toward increased Ca2+ responses at more distal sites in 150 mM Na+ (regression coefficient R = 0.12, P = 0.55).
). Figure 3 plots the mean Ca2+ transient at three presumed boutons after one and two AP in both 150 and 60 mM extracellular Na+. Consistent with the proximal axon, reliable conduction was observed on synaptic bouton-rich tertiary axonal branches, both in 150 and 60 Na+. A slightly reduced (~10%) Ca2+ transient in 60 Na+ was observed at both proximal and distal components (>160 µm) of the axon; the degree of reduction did not differ between compartments (P > 0.05, unpaired t-test, distal 59 sites vs. 46 proximal sites from 8 neurons).
View larger version (58K):
[in a new window]
FIG. 3.
Reliable axonal conduction to distal release sites in 60 mM Na+. Left: basal fluorescence (furaptra; F380) image illustrating varicose boutons on a segment of tertiary axon >250 µm from the soma (cortical). Right: records of Ca2+ response ( fluo-3/basal furaptra; F480/F380) vs. time for 3 axonal sites, averaged over 8 trials of 1 AP, 150 vs. 60 mM Na+; 2 AP, 150 vs.60 mM Na+.
2), indicating a general tendency toward smaller responses in 60 Na+. The analysis by t-test thus indicated that the mean AP-evoked Ca2+ transients in 60 Na+ were only minimally affected. However, it was unclear whether this analysis would detect a failure of conduction within a small number of trials because the paired t-test is somewhat insensitive to single outlying observations. In contrast, a comparison of variance in 150 and 60 Na+ would be sensitive to intermittent conduction failure (see METHODS). Axonal sites with high signal to noise ratio were identified, and an F-test was applied to compare the variance in the amplitude of Ca2+ transients across trials in 150 and 60 Na+ at these high signal to noise sites. This variance analysis revealed increased variance at only 3 of 105 axonal sites; hence intermittent axonal conduction failure was not observed under the low Na+ condition expected to promote failure.
View larger version (29K):
[in a new window]
FIG. 4.
Group data showing reliable axonal conduction in 60 mM Na+. A: mean Ca2+ response ( F480/F480 × 100%) in 60 mM Na+ (y-axis) vs. mean response in 150 mM Na+ (x-axis). Pooled data from 105 axonal sites from 8 neurons for 1 AP. Diagonal line represents unity (response in 60 mM Na+ = response in 150 mM Na+). B: coefficient of variation (CV = SD/mean) in 60 mM Na+ vs. CV in 150 mM Na+. Diagonal line represents unity (equal CVs in the 2 conditions). C: Ca2+ response integrals evoked by 2 APs at an axonal site plotted across 7 trials in the following conditions: 
, 150 Na+ (real data set); , 60 Na+ (real data set); 
, simulated failure data (constructed from 150 Na+ data by subtracting the mean Ca2+ response, in 150 mM Na+ trials, from trial 6). A significant increase in variance between 150 Na+ (real) and the "failure" data set was observed (F-test, P < 0.05) as indicated by the asterisk. Without the simulated failure, no significant difference in variance was observed between 150 and 60 Na+ at this axonal site, indicating reliable axonal conduction.
2). Therefore, at all axonal sites, there existed no general tendency toward a greater Ca2+ response variance in 60 Na+ than in 150 Na+. We did not observe statistically significant differences between cortical and hippocampal pyramidal neurons in the magnitude of the axonal Ca2+ response in high or low Na+. Analysis of the variance across trials thus suggested no increase in conduction failure in 60 or 150 mM Na+ in both hippocampal and cortical axons (data not shown). Therefore in both cell types we observed Ca2+ influx along the axon attributed to reliable and nonattenuating AP conduction in both 150 and 60 mM Na+.
View larger version (51K):
[in a new window]
FIG. 5.
Spatial profile of action-potential evoked Ca2+ transients along dendrites in 150 and 60 mM extracellular Na+. A: basal fluorescence image (F380) of a cortical spiny pyramidal neuron illustrating the area where Ca2+ transients were imaged. B: Ca2+ transients ( fluo-3/basal furaptra F480/F380) evoked by 2 APs (50-ms interval) for the indicated dendritic sites in 150 mM Na+ (thin traces) and 60 mM Na+ (thick traces). Average of 9 trials in each condition. C: Ca2+ response (F480/F380) to 2 APs (y-axis) vs. distance from soma along the dendrite (x-axis) in 150 mM Na+ () and 60 mM Na+ (). D: Ca2+ response along the dendrite in 60 Na+ expressed as a ratio of response in 150 Na+.
0.78, P < 10
6; decrementing 43% over 100 µm). This differs from the spatial profile of the group axonal Ca2+ responses, which did not change significantly with distance. Figure 6 presents group data summarizing the Ca2+ responses in 60 mM Na+ (as a ratio of response in 150 mM Na+) in both the dendrite and axon. Along the axon (Fig. 6A), Ca2+ influx in response to both one and two APs was unchanged in 60 mM Na+. In contrast, along the dendrite (Fig. 6B), a reduction in the 60 Na+/150 Na+ response ratio was observed over a distance of 40-160 µm for both one and two AP.
View larger version (29K):
[in a new window]
FIG. 7.
Dendritic AP-evoked Ca2+ transients depend on tetrodotoxin (TTX)-sensitive Na+ channels. A: change in dendritic Ca2+ transients ( fluo-3/furaptra; F480/F380) averaged along a segment of dendrite 30-80 µm from the soma. Left panel: Ca2+ responses in the presence of 1 µM TTX to 5 ms somatic current steps of 300, 700, and 1,000 pA (1 current step: light bars; 2 current steps 50 ms apart: dark bars). Right panel: Ca2+ responses F480/F380) to current steps of 300 pA (subthreshold) and 700 pA (suprathreshold) after washout of TTX. Light bars are responses to 1 current step, and dark bars are responses to 2 current steps; note 6-fold change in y-axis scale. B: spatial profile along the dendrite of TTX-sensitive and -insensitive Ca2+ transients evoked by current steps of 1,000 pA in the presence of TTX (), 300 pA after washout of TTX (, subthreshold), and 700-pA current steps after washout (, 2 AP). Inset: expanded y-axis illustrating the spatial profile of the TTX-insensitive component (, compared with subthreshold responses after washout ().
), we used a low dose of TTX (20 nM) to attenuate the Na+ spike to a similar degree as 60 mM external Na+ (range 45-57% amplitude reduction in 20 nM TTX). Figure 8A illustrates examples of Ca2+ transients along the dendrite and axon in response to one AP. In the dendrite, the Ca2+ response at distances >100 µm from the soma was significantly reduced in dilute TTX (compared with control) and decreased with further distance. In contrast, the Ca2+ response in the axon at distances >100 µm from the soma persisted in 20 nM TTX. In the axon, the Ca2+ response was reversibly blocked by 1 µM TTX. After TTX washout, Ca2+ transients were fully restored in both the axon and dendritic compartments. Figure 8B presents group data of the normalized Ca2+ response from 23 sites from 5 cells (axon) and 39 sites from 5 cells (dendrite) that were >100 µm from the soma.
). A recent report also suggested that A-type K+ channel could regulate AP conduction in the axon (Debanne et al. 1997). We therefore examined axonal AP conduction from hyperpolarized potentials expected to remove inactivation of the A-type K+ channels and thereby increase their effect (Segal and Barker 1984; Wu and Barish 1992), possibly leading to conduction failure. Single or pairs of APs were elicited by somatic current steps from 60 or 100 mV. Figure 9A illustrates a basal fluorescence image of three representative secondary and tertiary axonal sites. Figure 9B plots traces of fluo-3 fluorescence versus time in response to pairs of APs from resting and hyperpolarized potentials. No difference in the Ca2+ transient was observed at hyperpolarized potentials. Figure 9C illustrates traces of membrane potential recorded at the soma after APs evoked from 60 and 100 mV. Figure 9D plots the average Ca2+ transient elicited by 1 AP from 100 mV versus average Ca2+ transient from 60 mV for 10 axonal sites from the neuron illustrated in Fig. 9A. The average Ca2+ response from 100 mV was not significantly different from the 60 mV response (P > 0.05, t-test). Similar results were obtained in five of five neurons.
View larger version (43K):
[in a new window]
FIG. 9.
Reliable axonal AP conduction from hyperpolarized potentials. A: basal fluorescence image illustrating secondary and tertiary axonal branches and associated boutons >250 µm from the soma (hippocampal). B: traces of change in fluo-3/furaptra (F480/F380) fluorescence vs. time after 2 APs elicited from 60 mV (thin trace) and 100 mV (thick trace). Traces are average of 8 (60 mV) and 11 trials (100 mV) for 3 axonal sites. C: traces of somatic membrane potential illustrating 2 APs 50 ms apart elicited from 60 mV (left) or from 100 mV (right). D: average Ca2+ transient (100 mV) vs. average Ca2+ response (60 mV). Diagonal line represents unity where the 60 mV response equals the 100 mV response.
DISCUSSION
Abstract
Introduction
Methods
Results
Discussion
References
). Here we compared the reliability of axonal and dendritic propagation under normal and limiting conditions expected to reduce conduction safety factor. Axons and dendrites act as separate compartments of differing excitability in normal external Na+; furthermore AP conduction is differentially sensitive to a reduction in the flux of Na+ ions along axonal and dendritic compartments.
. Second, for all 105 sites from 8 cells examined, we measured the CV in 150 Na+ versus the CV in 60 Na+ and found no consistent tendency toward increased variance in low Na+ext (Fig. 4B), suggesting reliable AP conduction in 60 mM Na+ext.
), it cannot be accounted for by intermittent failure.
). To investigate the role of A-type K+ channels, we compared AP-evoked Ca2+ transients from resting and hyperpolarized potentials. Hyperpolarized potentials would be expected to remove steady-state inactivation of A-type K+ channels and increase their effect (Segal and Barker 1984; Wu and Barish 1992). We observed no change in the magnitude of Ca2+ transients (Fig. 9), suggesting that an (inactivated) A-type K+ conductance played little role in the modulation of axonal AP conduction. This contrasts with a recent report that an A-type K+ conductance regulates axonal AP conduction in hippocampal CA3-CA1 cell pairs (Debanne et al. 1997). However, because our method is limited to imaging only a portion of the axon at a time, conduction failure could have been occurring on other possibly more distal branches that we were not imaging but that would have been detected in the study of Debanne et al. Furthermore different patterns of axonal branching could have contributed to the differences observed. As for dendritic AP conduction, recent findings of Hoffman et al. (1997) demonstrated the regulation of dendritic excitability by an A-type K+ channel. We acknowledge that a potential role for Ca2+ ions could have been lessened by the high concentration of fluo-3 that was required to a achieve a sufficient signal to noise ratio in recordings. Therefore it is possible that exogenous Ca2+ buffering may have reduced the activation of Ca2+-dependent K+ channels, thereby increasing the reliability of AP conduction. To determine whether conduction reliability was enhanced by Ca2+ buffering, we increased the contribution of K+ channels during AP repolarization by using a low (1 mM) extracellular K+ medium (to increase the K+ driving force). We did not observe AP failure in the axon in three of three neurons in preliminary experiments. Additional experiments in high extracellular Ca2+ medium (5 mM) were performed to directly increase Ca2+ influx (and potential activation of K+ channels) with no apparent effect on conduction failure (n = 4 neurons; data not shown).
. In the dendrite, the Ca2+ response decreased with distance from the soma with a length constant of 150-220 µm (Figs. 5-7). The degree of attenuation of the Ca2+ transient evoked by a single AP along the apical dendrite appears to vary according to the cell type, age, and recording conditions (Schiller et al. 1995, 1997; Spruston et al. 1995; Svoboda et al. 1997; Yuste et al. 1994), presumably reflecting the differential distribution and activation of ion channels and shunting conductances (Jaffe et al. 1992; Migliore 1996; Tsubokawa and Ross 1996; Westenbroek et al. 1992). Rather than focus on factors controlling dendritic back-propagation, we chose to compare dendritic and axonal sensitivity under conditions of low safety factor (low extracellular Na+). In proximal dendritic regions (~30 µm from the soma), AP-evoked Ca2+ responses in 60 Na+ were reduced by ~10% when compared with responses in 150 Na+. This slight attenuation observed at the proximal dendrite is likely due to the effects of Na+ substitution and not conduction, as discussed previously. However, when more distal regions were examined, the amplitude of the Ca2+ response decreased with distance in 150 Na+ and to a greater extent in 60 mM Na+ (Figs. 5 and 6).
). Cortical and hippocampal dendrites are capable of local electrogenesis, however this usually stops short of an all-or-none AP. In contrast AP generation results in faithful propagation along the axon (Allen and Stevens 1994; Lüscher et al. 1996; Mackenzie et al. 1996; Stevens and Wang 1995) and back-propagation along dendrites where the amplitude decreases with distance (Spruston et al. 1995; Stuart and Sakmann 1994; Stuart et al. 1997). We report here a marked difference in conduction reliability between axons and dendrites of hippocampal and cortical neurons. The propagation of APs at low frequency occurs in the axon even under conditions of lower external Na+. This resistance to external 60 Na+ was unique to the axon and supports the suggestion that the axon functions as a more excitable electrical compartment (Mainen et al. 1995; Rapp et al. 1996). This considerable degree of safety in the mechanism of AP propagation suggests that in the axon the AP, once generated, is indeed all or none and that Na+ channel modulation is not likely to play a major role in AP conduction. However, as larger current steps were required to generate an AP in 60 Na+, we acknowledge a possible role of Na+ channel modulation at the sites of AP initiation such as the axon hillock. Taken together, this suggests that modulation of Na+ channels or leakage channels that affect membrane resistance may have a greater impact on the AP generation or on the local depolarization of terminals than on conduction safety. In dendrites, in low Na+, APs resulted in Ca2+ transients over a more limited distance along the dendrite. These results in the dendrite are consistent with previous results suggesting a graded, modifiable dendritic AP (Jaffe et al. 1992; Spruston et al. 1995; Tsubokawa and Ross 1996). Our results support the hypothesis that regulation of dendritic Na+ channels can control the spatial extent of dendritic electrogenesis, presumably for both forward synaptic integration and for retrograde AP propagation (Colbert et al. 1997; Jung et al. 1997; Mainen et al. 1995; Rapp et al. 1996).
| |
ACKNOWLEDGEMENTS |
|---|
We thank Drs. J. Seamans, S. Wang, and J. Winslow for helpful comments on the manuscript.
This work was supported by grants from the Medical Research Council of Canada and the EJLB foundation to T. H. Murphy, an EJLB and Medical Research Council scholar. P. J. Mackenzie is supported by a Medical Research Council studentship.
| |
FOOTNOTES |
|---|
Address reprint requests to T. H. Murphy.
Received 3 March 1998; accepted in final form 9 July 1998.
| |
REFERENCES |
|---|
|
Abstract
Introduction
Methods
Results
Discussion
References
|
|---|
This article has been cited by other articles:
|
|
 |
|
  D. Okada, F. Ozawa, and K. Inokuchi Input-Specific Spine Entry of Soma-Derived Vesl-1S Protein Conforms to Synaptic Tagging Science, May 15, 2009; 324(5929): 904 - 909. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. T. Roberts, K. J. Bender, and L. O. Trussell Fidelity of Complex Spike-Mediated Synaptic Transmission between Inhibitory Interneurons J. Neurosci., September 17, 2008; 28(38): 9440 - 9450. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Vervaeke, N. Gu, C. Agdestein, H. Hu, and J. F. Storm Kv7/KCNQ/M-channels in rat glutamatergic hippocampal axons and their role in regulation of excitability and transmitter release J. Physiol., October 1, 2006; 576(1): 235 - 256. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Monsivais, B. A. Clark, A. Roth, and M. Hausser Determinants of Action Potential Propagation in Cerebellar Purkinje Cell Axons J. Neurosci., January 12, 2005; 25(2): 464 - 472. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. P. Meeks and S. Mennerick Selective Effects of Potassium Elevations on Glutamate Signaling and Action Potential Conduction in Hippocampus J. Neurosci., January 7, 2004; 24(1): 197 - 206. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Raastad and G. M G Shepherd Single-axon action potentials in the rat hippocampal cortex J. Physiol., May 1, 2003; 548(3): 745 - 752. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. B Levy and R. A. Baxter Energy-Efficient Neuronal Computation via Quantal Synaptic Failures J. Neurosci., June 1, 2002; 22(11): 4746 - 4755. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. He, C. F. Zorumski, and S. Mennerick Contribution of Presynaptic Na+ Channel Inactivation to Paired-Pulse Synaptic Depression in Cultured Hippocampal Neurons J Neurophysiol, February 1, 2002; 87(2): 925 - 936. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. Graf, R. Spencer, H. Baker, and R. Baker Vestibuloocular Reflex of the Adult Flatfish. III. A Species-Specific Reciprocal Pattern of Excitation and Inhibition J Neurophysiol, September 1, 2001; 86(3): 1376 - 1388. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. J. Hagler Jr. and Y. Goda Properties of Synchronous and Asynchronous Release During Pulse Train Depression in Cultured Hippocampal Neurons J Neurophysiol, June 1, 2001; 85(6): 2324 - 2334. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. L. Cox, W. Denk, D. W. Tank, and K. Svoboda Action potentials reliably invade axonal arbors of rat neocortical neurons PNAS, August 6, 2000; (2000) 170278697. [Abstract] [Full Text]
|
|
|
|
|
|
P. Varona, J. M. Ibarz, L. Lopez-Aguado, and O. Herreras Macroscopic and Subcellular Factors Shaping Population Spikes J Neurophysiol, April 1, 2000; 83(4): 2192 - 2208. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. L. Brody and D. T. Yue Release-Independent Short-Term Synaptic Depression in Cultured Hippocampal Neurons J. Neurosci., April 1, 2000; 20(7): 2480 - 2494. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Lopez-Aguado, J. M. Ibarz, and O. Herreras Modulation of Dendritic Action Currents Decreases the Reliability of Population Spikes J Neurophysiol, February 1, 2000; 83(2): 1108 - 1114. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. R Williams and G. J Stuart Mechanisms and consequences of action potential burst firing in rat neocortical pyramidal neurons J. Physiol., December 1, 1999; 521(2): 467 - 482. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. C. Magee and M. Carruth Dendritic Voltage-Gated Ion Channels Regulate the Action Potential Firing Mode of Hippocampal CA1 Pyramidal Neurons J Neurophysiol, October 1, 1999; 82(4): 1895 - 1901. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Jensen, M. S. Jensen, and J. D C Lambert Post-tetanic potentiation of GABAergic IPSCs in cultured rat hippocampal neurones J. Physiol., August 15, 1999; 519(1): 71 - 84. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. L. Cox, W. Denk, D. W. Tank, and K. Svoboda From the Cover: Action potentials reliably invade axonal arbors of rat neocortical neurons PNAS, August 15, 2000; 97(17): 9724 - 9728. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
