and activity-dependent modulation of spike back-propagation in layer ...
J Neurophysiol 105: 1372–1379, 2011. First published January 5, 2011; doi:10.1152/jn.00014.2010.
Distance- and activity-dependent modulation of spike back-propagation in layer V pyramidal neurons of the medial entorhinal cortex Sonia Gasparini Neuroscience Center, Louisiana State University Health Science Center, New Orleans, Louisiana, and Marine Biological Laboratory, Woods Hole, Massachusetts Submitted 7 January 2010; accepted in final form 2 January 2011
Gasparini S. Distance- and activity-dependent modulation of spike back-propagation in layer V pyramidal neurons of the medial entorhinal cortex. J Neurophysiol 105: 1372–1379, 2011. First published January 5, 2011; doi:10.1152/jn.00014.2010.—Layer V principal neurons of the medial entorhinal cortex receive the main hippocampal output and relay processed information to the neocortex. Despite the fundamental role hypothesized for these neurons in memory replay and consolidation, their dendritic features are largely unknown. Highspeed confocal and two-photon Ca2⫹ imaging coupled with somatic whole cell patch-clamp recordings were used to investigate spike back-propagation in these neurons. The Ca2⫹ transient associated with a single back-propagating action potential was considerably smaller at distal dendritic locations (⬎200 m from the soma) compared with proximal ones. Perfusion of Ba2⫹ (150 M) or 4-aminopyridine (2 mM) to block A-type K⫹ currents significantly increased the amplitude of the distal, but not proximal, Ca2⫹ transients, which is strong evidence for an increased density of these channels at distal dendritic locations. In addition, the Ca2⫹ transients decreased with each subsequent spike in a 20-Hz train; this activitydependent decrease was also more prominent at more distal locations and was attenuated by the perfusion of the protein kinase C activator phorbol-di-acetate. These data are consistent with a phosphorylationdependent control of back-propagation during trains of action potentials, attributable mainly to an increase in the time constant of recovery from voltage-dependent inactivation of dendritic Na⫹ channels. In summary, dendritic Na⫹ and A-type K⫹ channels control spike back-propagation in layer V entorhinal neurons. Because the activity of these channels is highly modulated, the extent of the dendritic Ca2⫹ influx is as well, with important functional implications for dendritic integration and associative synaptic plasticity. dendrite; Na⫹ channels; A-type K⫹ channels
(EC) is a key relay structure for the flow of information between the hippocampus and the neocortex (Amaral and Witter 1995). Not only does it act as a primary interface, it also plays a critical role in the computation of multisensory and cognitive modalities. The latter function is clearly supported by the direct involvement of the EC in neurodegenerative and psychiatric disorders such as Alzheimer’s disease, epilepsy, and schizophrenia (Jellinger et al. 1991; Arnold 2000; de Curtis and Paré 2004). Various studies have contributed to our knowledge of the features of EC principal neurons (Hamam et al. 2000; Spruston and McBain 2007) and their connections (Witter and Amaral 2004). Layer V neurons are the main target of the hippocampal output and send their axons to cortical regions; they are thus likely to play an important role in memory replay and consol-
THE ENTORHINAL CORTEX
Address for reprint requests and other correspondence: S. Gasparini, Neuroscience Ctr., Louisiana State Univ. Health Science Ctr., 2020 Gravier St., New Orleans, LA 70112 (e-mail: [email protected]). 1372
idation, which involve hippocampal-neocortical communications (Buzsaki 1989). To understand the function of EC layer V neurons, it is essential to understand how they integrate the inputs they receive to generate the output that is transferred to the neocortex; this process depends on the neuronal morphology and the properties of dendritic voltage-dependent channels, in addition to the input temporal and spatial patterns (Migliore and Shepherd 2005; Gasparini and Magee 2006). Dendritic channels are also responsible for the back-propagation of somatic spikes, with features that highly differ from neuron to neuron (for a review see Spruston et al. 2008). A detailed knowledge of the mechanisms of back-propagation in neurons of the central nervous system is essential because back-propagating action potentials (bAPs) are involved in the modulation of dendritic excitability and associative synaptic plasticity (Magee and Johnston 1997; Magee and Johnston 2005). Despite the fundamental role of EC layer V neurons in memory consolidation, much remains to be known about their dendritic excitability. The EC consists of two functionally distinct regions; spatial information processing is mainly restricted to the medial EC and nonspatial processing to the lateral EC (Fyhn et al. 2004; Hargreaves et al. 2005). Previous studies focused on dopaminergic modulation of hyperpolarization-activated cation and Na⫹ channels in the lateral EC (Rosenkranz and Johnston 2006, 2007). Here imaging techniques were used to measure Ca2⫹ signals associated with bAPs in regions of the apical dendrites of medial entorhinal layer V neurons that cannot be visualized under differential interference-contrast (DIC) microscopy. The amplitude of the Ca2⫹ signals associated with bAPs was found to be both distance and activity dependent and is controlled by the activity of dendritic Na⫹ and K⫹ channels. In particular, smaller distal Ca2⫹ signals could be boosted by blocking A-type K⫹ channels, suggesting that these channels control the amplitude of bAPs along the apical dendrites of medial EC layer V neurons, similar to what has been reported for CA1 pyramidal neurons (Hoffman et al. 1997). METHODS
EC slices were prepared from 7- to 9-wk-old Sprague Dawley rats, according to methods approved by the Louisiana State University Health Science Center Institutional Animal Care and Use Committee. Briefly, rats were anesthetized with ketamine and xylazine, and the brain was perfused with an oxygenated, ice-cold solution through the ascending aorta. After decapitation, the brain was rapidly removed, and 400-m-thick horizontal slices including the EC and the rest of the hippocampal formation were cut using a vibratome. After a
0022-3077/11 Copyright © 2011 the American Physiological Society
www.jn.org
BACK-PROPAGATION IN ENTORHINAL LAYER V NEURONS
recovery period of at least 1 h, an individual slice was transferred to the recording chamber; the external solution used for recordings contained (in mM) 125 NaCl, 2.5 KCl, 25 NaHCO3, 1.25 Na2HPO4, 2 CaCl2, 1 MgCl2, and 25 dextrose and was saturated with 95% O2-5% CO2 at 34 –36°C (pH 7.4). Whole cell patch-clamp somatic recordings from layer V pyramidal neurons in the ventromedial EC were performed using a Dagan BVC-700 amplifier in the active “bridge” mode. At the end of each recording, a low-magnification objective that allows a view of the whole slice was used to verify that the recording electrode was actually placed in the medial portion of layer V, just below the lamina disseccans. Patch pipettes had a resistance of 2–3 M⍀ when filled with a solution containing (in mM) 130 K-methylsulphonate, 10 HEPES, 4 NaCl, 4 Mg2ATP, 0.3 Tris2GTP, and 14 phosphocreatine (pH 7.3). Oregon Green 1,2-bis(2-aminophenoxy)ethane-N,N,N=,N=tetraacetate-1 [Oregon Green BAPTA-1 (OGB-1), 100 M; Invitrogen, Carlsbad, CA] was added to the internal solution to monitor changes in [Ca2⫹]i. Somatic action potentials were elicited by brief current steps (2 nA for 2 ms). EC layer V pyramidal neurons were visualized using a Nikon Eclipse FN-1 microscope, equipped with DIC optics under infrared illumination (Nikon, Melville, NY) coupled with a swept-field confocal system in slit mode (Prairie Technologies, Middleton, WI). A neuro-charge-coupled device (CCD) camera (RedShirt Imaging, Decatur, GA) with an 80 ⫻ 80 pixel array was used to acquire the optical signal in response to excitation at 488 nm (sequential frame rate 0.5 kHz). In other sets of experiments, ultrafast, pulsed, laser light at 920 nm (Mira 900F; Coherent, Santa Clara, CA) was used to excite OGB-1 and image local Ca2⫹ in the entorhinal apical dendrites using a multiphoton scanner (Ultima, Prairie Technologies) mounted over an Olympus BX61WI microscope (Center Valley, PA). Changes in [Ca2⫹]i associated with bAPs were quantified by calculating ⌬F/F, where F is the fluorescence intensity before stimulation, after subtracting autofluorescence, and ⌬F is the change in fluorescence during neuronal activity (Lasser-Ross et al. 1991). The autofluorescence of the tissue was measured in a region of equal size adjacent to the dye-filled dendrite. The ⌬F/F measurements were repeated three to eight times and averaged. The high frame rate of the optical imaging (0.3– 0.5 kHz) allowed the resolution of the contribution of each bAP to the cumulative Ca2⫹ signal (Figs. 4 – 6). To estimate the amplitude of the Ca2⫹ transients associated with spikes after the first one in a train, the residual fluorescence was calculated by fitting the decline of the previous peak with a single exponential (Helmchen et al. 1996) and subtracting the projected value in the absence of a subsequent spike from the actual maximum value observed in response to a given spike. In a set of experiments, local application of Ba2⫹ (375 M) was achieved by using a pressure ejection system (Picospritzer; Parker Hannifin, Fairfield, NJ) connected to a patch pipette (2–5 m diameter, pressures of 2–5 psi, and durations of 0.5–2.0 min). With these settings, an ⬃200-m distal region of the cell (⬃250 – 450 m from the soma) was perfused when the pipette was placed within 20 m of the dendritic trunk at 350 m from the soma. The area of the cell effectively reached by the local perfusion was tested using a fluorescent dye (Alexa Fluor 488). Phorbol-di-acetate (PDA) and bisindolylmaleimide (GF109203x) were purchased from Sigma-Aldrich (St. Louis, MO) and dissolved in DMSO to obtain stock solutions of 10 and 2 mM, respectively; PDA was then added to the external solution and bisindolylmaleimide to the patch pipette. The concentration of DMSO in the final solution was 0.1%. Control experiments were performed to determine that the addition of DMSO to the pipette solution did not by itself alter the response to PDA (n ⫽ 3). Data are reported as means ⫾ SE. Statistical comparisons were performed by using a one-way analysis of variance and appropriate post hoc tests. Means were considered to be significantly different when P ⬍ 0.05. J Neurophysiol • VOL
1373
RESULTS
The apical dendrites of layer V pyramidal neurons of the medial EC are too small and convoluted to be visualized under DIC microscopy; therefore, dendritic excitability cannot be examined directly using electrophysiological techniques. Instead, the Ca2⫹ transients associated with back-propagating spikes initiated by a brief current injection at the soma were analyzed. Previous studies have shown that Ca2⫹ transients are very good indicators of AP amplitude in the dendrites (Magee
Fig. 1. The amplitude of back-propagating action potential (bAP)-associated Ca2⫹ signals decreases with the distance from the soma along the apical dendrites of entorhinal cortex (EC) layer V neurons. A: 2-dimensional projection of 2-photon 3D image stacks for an EC layer V pyramidal neuron filled with Oregon Green BAPTA-1 (OGB-1) (100 M). The white boxes outline regions of the apical dendrite that are expanded on the sides, showing the presence of spines all over the apical dendrite. B: Ca2⫹ transients (expressed as ⌬F/F, where F is fluorescence) generated in response to a somatic action potential (C) measured with a line scan at the locations marked by gray lines. D: plot of the mean amplitude of the Ca2⫹ peaks associated with bAPs as a function of the distance from the soma (n ⫽ 18). The experimental points are fitted with a sigmoidal function.
105 • MARCH 2011 •
www.jn.org
1374
BACK-PROPAGATION IN ENTORHINAL LAYER V NEURONS
and Johnston 1997; Gasparini et al. 2007) and that changes in Ca2⫹ transients are correlated primarily with changes in bAP amplitude rather than in their width (see Figs. 2 and 6 in Gasparini et al. 2007). EC layer V neurons were filled with the Ca2⫹ indicator OGB-1 through a somatic patch pipette (Fig. 1). The two-dimensional projection of a series of two-photon 3D image stacks from an EC layer V neuron in Fig. 1A shows the prominent apical dendrite, which branches in a tuft in the superficial layers. The insets on the sides show that spines are distributed all over the length of the apical dendrite and tuft. Most medial EC layer V pyramidal neurons appear to lack the extensive oblique dendritic branches that characterize layer V neocortical and CA1 pyramidal neurons (Larkman 1991; Bannister and Larkman 1995). A first set of experiments used a multi-photon line scan to analyze the amplitude of the dendritic Ca2⫹ signal (⌬F/F) associated with a somatic AP at different locations (⬃60 m apart) and revealed a clear dependence of signal amplitude on the distance from the soma. As shown in Fig. 1B, the amplitude of dendritic Ca2⫹ transients was constant for the first ⬃150 m of the apical dendrite (with an average ⌬F/F of 68 ⫾ 5% at 60 m from the soma, n ⫽ 18, Fig. 1D) and started decreasing at ⬃200 m. The signal became
undetectable from baseline for distances ⬎350 m (⌬F/F of 8 ⫾ 2% at ⬃360 m, n ⫽ 18, Fig. 1D), implying that the bAPassociated depolarization was unable to activate a significant number of voltage-dependent Ca2⫹ channels at this level. To address the mechanism underlying the distance-dependent decrease in bAP-amplitude in medial EC layer V neurons, the sensitivity of distal bAP-associated Ca2⫹ transients to the blockade of A-type K⫹ channels by Ba2⫹ (150 M) and 4-aminopyridine (4-AP, 2 mM) was tested. The first set of experiments was performed using a high-speed CCD camera coupled with confocal microscopy; the orange box in Fig. 2A outlines the region of interest for the optical traces in Fig. 2B. A 20-Hz train of three APs was evoked at the soma (Fig. 2C). At 150 M, Ba2⫹ has been shown to block A-type K⫹ currents in the distal apical dendrites of hippocampal CA1 pyramidal neurons and to thereby boost the smaller bAPs at these locations (Gasparini et al. 2007). However, because Ba2⫹ also blocks inward rectifier K⫹ currents at lower concentrations, it was necessary to determine that the block of these channels did not affect back-propagation. For this reason, I first perfused Ba2⫹ at 20 M, a concentration that selectively blocks inward rectifier K⫹ channels (Chatelain et al. 2005; Gasparini et al.
Fig. 2. Ba2⫹ (150 M but not 20 M) and 4-aminopyridine (4-AP) enhance spike back-propagation into the distal apical dendrites of EC layer V pyramidal neurons. A: charge-coupled device (CCD) image of the distal (⬎300 m) portion of an EC layer V pyramidal neuron filled with OGB-1 (100 M). The orange box outlines the region of interest where the changes in fluorescence (⌬F/F, B) were monitored in response to a 20-Hz train of 3 APs (C) in control conditions (black trace), during the perfusion of Ba2⫹ (20 and 150 M, green and red trace, respectively) and upon wash out (w-o) (blue trace). The expanded time scale (C, inset) better illustrates the effect of Ba⫹ on the first action potential. D: mean data of the amplitude of the first Ca2⫹ transient measured in distal apical dendrites at ⬃350 m from the soma (n ⫽ 10) show that the block of inward rectifier K⫹ channels by Ba2⫹ 20 M did not change the Ca2⫹ transients, but Ba2⫹ 150 M significantly increased them. E: 2-photon image showing a portion of the apical dendrite at ⬃330 m from the soma and the location of the line scan. F: Ca2⫹ transients evoked by the somatic action potential in G in control conditions, during the perfusion of 4-AP (2 mM) and upon wash out. H: average data of the amplitude of the Ca2⫹ transients associated with bAPs at ⬃350 m from the soma for n ⫽ 8 neurons. *P ⬍ 0.01, **P ⬍ 0.005. J Neurophysiol • VOL
105 • MARCH 2011 •
www.jn.org
BACK-PROPAGATION IN ENTORHINAL LAYER V NEURONS
2007). As shown in Fig. 2B, the perfusion of lower Ba2⫹ concentrations did not affect the Ca2⫹ transients recorded at ⬃350 m; the application of Ba2⫹ (150 M), however, greatly increased the Ca2⫹ peaks observed in the distal dendrites (Fig. 2B), indicating that the increase in bAP amplitude at distal dendritic locations is mediated by the effect of Ba2⫹ on A-type K⫹ channels. On average, the first Ca2⫹ transient increased from 11 ⫾ 1% under control conditions to 62 ⫾ 10% in the presence of Ba2⫹ 150 M at 350 m from the soma (Fig. 2D; n ⫽ 10, P ⬍ 0.005, Wilcoxon matched-pairs signed-ranks test). The dependence of bAP amplitude on A-type K⫹ channels in the distal apical dendrites was confirmed by examining the effect of 4-AP on the distal Ca2⫹ transients, recorded through a two-photon line scan (Fig. 2, E–G). At ⬃350 m from the soma, 4-AP (2 mM) increased the Ca2⫹ transient evoked by a single somatic AP from 10 ⫾ 2% to 65 ⫾ 9% (Fig. 2H; n ⫽ 8, P ⬍ 0.01, Wilcoxon matched-pairs signed-ranks test). Although Ba2⫹ and 4-AP block other K⫹ channels, the spectra of the channels blocked do not overlap except for A-type K⫹ channels (Storm 1990); therefore, the similarity of the effects of 4-AP and Ba2⫹ is strong evidence that these effects are mediated primarily by these K⫹ channels. A putative reduction of back-propagating spike amplitude by a gradient of A-type K⫹ channels along the apical dendrites was further tested by local application of Ba2⫹ at various distances from the soma. Local perfusion of Ba2⫹ (Fig. 3) produced only a slight increase (⬃10%) of the Ca2⫹ transient in the proximal apical dendrite compared with a fourfold increase of the smaller Ca2⫹ transient at ⬃350 m from the soma. The average data (n ⫽ 15) show that the effect of local perfusion of Ba2⫹ increased with the distance from the soma, such that for locations ⱖ200 m it effectively removed the distance-dependent decrease in the Ca2⫹ transients recorded during bAP in control conditions (Fig. 3, B and C). These data suggest that the decrease in bAP amplitude along the apical dendrites of entorhinal layer V neurons can be attributed to a
1375
contribution of A-type K⫹ channels that increases with distance from the soma, similar to what has been shown for CA1 pyramidal neurons (Hoffman et al. 1997). In addition to the dependence of bAP amplitude on the spatial location, back-propagation can also depend on the firing history of a neuron. Specifically, bAP amplitude has been reported to decrease in an activity-dependent manner in CA1 and neocortical layer V pyramidal neurons attributable to a slow cumulative inactivation of dendritic Na⫹ channels (Callaway and Ross 1995; Spruston et al. 1995; Colbert et al. 1997; Stuart et al. 1997). To investigate whether this was also the case in EC layer V neurons, the Ca2⫹ transients associated with a 20-Hz train of five somatically initiated spikes were analyzed (Fig. 4). The ratio of the first Ca2⫹ peak to the peak of its proximal counterpart started decreasing at ⬃200 m from the soma (consistent with the data in Fig. 1). In contrast, the ratios of the subsequent Ca2⫹ transients decreased at more proximal locations (Fig. 4B), and the decline became progressively more pronounced with each successive spike in the train, which is also evident from the average plot (Fig. 4D). Whereas the first Ca2⫹ peak decreased only to 0.89 ⫾ 0.05 of the initial value at 200 m, the third and the fifth Ca2⫹ transients dropped to 0.73 ⫾ 0.05 and 0.51 ⫾ 0.10, respectively, at the same distance (n ⫽ 10, Fig. 4D). These data suggest that dendritic Na⫹ channels in EC layer V neurons may undergo slow cumulative inactivation. The effect of the perfusion of PKC activator PDA (10 M) was examined at various distances from the soma to determine whether phosphorylation of these channels could accelerate their recovery from inactivation, thereby reversing the activity-dependent decrease in the bAP, as has been shown in CA1 pyramidal neurons (Colbert and Johnston 1998). Figure 5 shows the line scans obtained at 100 m and 200 m from the soma for two different neurons, with the typical activitydependent decrease of the later spikes in the train at more distal locations (compare the black traces in Fig. 5, B and E). The perfusion of PDA did not affect the Ca2⫹ transients at 100 m
Fig. 3. Local Ba2⫹ perfusion counters the distance-dependent decrease in bAP amplitude. A: Ca2⫹ transients obtained under control conditions (black trace) and during local perfusion of Ba2⫹ (gray trace). The diagram on the left shows the experimental configuration, with a whole cell somatic electrode and the line scan locations for proximal and distal regions, as well as the regions affected by Ba2⫹ pressure ejection at ⬃100 and ⬃350 m from the soma, respectively. B: average data of the amplitude of the Ca2⫹ transients associated with bAPs as a function of the distance from the soma in control conditions and during local perfusion of Ba2⫹ (n ⫽ 15). The effect of Ba2⫹ is more pronounced for distal dendritic locations, as shown clearly by the ratio plot in C.
J Neurophysiol • VOL
105 • MARCH 2011 •
www.jn.org
1376
BACK-PROPAGATION IN ENTORHINAL LAYER V NEURONS
lation of the activity-dependent decrease in the amplitude of the Ca2⫹ associated with bAPs. To determine the relative contributions of the Na⫹ and K⫹ currents, A-type K⫹ currents were blocked with Ba2⫹ (150 M), and the activity dependence of the Ca2⫹ transients within a train and its modulation by phorbol esters were examined at 200 m from the soma, where the activity-dependent decline in the Ca2⫹ transients is pronounced (see Fig. 5). Ba2⫹ increased the amplitude of every
Fig. 4. bAP-associated Ca2⫹ transients show an activity-dependent decrease at more distal (⬎150 m) dendritic locations. A: reconstructed resting fluorescence image of an EC layer V neuron filled with OGB-1 and imaged with a high-speed CCD camera in combination with a swept-field confocal scanner. B: optical recordings of the Ca2⫹ influx (expressed as ⌬F/F) generated in response to a 20-Hz train of 5 APs (C) in the regions of interest marked by the white boxes in A. D: plots of the average for the first (), third (Œ), and fifth (Œ) Ca2⫹ transients (normalized to the most proximal value) associated with a train of bAPs for 10 neurons show a more pronounced decrease for the bAPs later in the train.
from the soma (Fig. 5B) but strongly affected the activity dependence of back-propagation within a train at 200 m from the soma (Fig. 5E). The average data show that PDA has a negligible effect on the first Ca2⫹ transient up to 200 m (the ratio between the amplitude in the presence of PDA and that in control conditions was 1.1 ⫾ 0.1 for the first Ca2⫹ peak) but exerts progressively larger effects on each subsequent Ca2⫹ transients at more distal locations (the ratio was 1.4 ⫾ 0.1 and 2.1 ⫾ 0.3, for the third and the fifth peak at 200 m, respectively, n ⫽ 8, Fig. 5M). To verify that PDA acted through a PKC-dependent mechanism, a series of control experiments were performed to test the ability of a PKC inhibitor, bisindolylmaleimide (GF109203x), to block the effects of PDA on reversing the activity-dependent decrease (Fig. 5). GF109203x was added to the patch pipette solution and by itself had no effect on the typical activity-dependent decline of the Ca2⫹ transients associated with later bAPs in the train for distances greater than 100 m from the soma (compare Fig. 5, K and H). However, with the PKC inhibitor in the pipette, perfusion of PDA now failed to reverse the activity-dependent decrease in the Ca2⫹ transients associated with bAPs at any distance (Fig. 5, H and K). The ratio between the amplitude in the presence of PDA and that in control conditions at 200 m from the soma was 0.9 ⫾ 0.1, 1.0 ⫾ 0.1, and 1.1 ⫾ 0.1 for the first, third, and fifth Ca2⫹ transients, respectively (n ⫽ 7, Fig. 5M). These data indicate that the activity-dependent decrease in bAP amplitude can be effectively prevented by PKC activation. Whereas the slow cumulative inactivation of dendritic Na⫹ channels is a main determinant of the activity-dependent decline of Ca2⫹ transients in a train, the higher density of A-type K⫹ channels in the dendrites of layer V EC neurons could also contribute to this phenomenon. Moreover, PKC activation could affect K⫹ currents, therefore contributing to the moduJ Neurophysiol • VOL
Fig. 5. PKC activation by phorbol esters reduces the activity-dependent decrease of bAP amplitude. A: 2-photon image showing a portion of the apical dendrite and the location of the line scan at 100 m from the soma. B: Ca2⫹ transients generated in response to a 20-Hz train of 5 APs (C). D–F: as A– C at 200 m from the soma. G: 2-photon image showing the location of the line scan (at ⬃100 m from the soma) for the Ca2⫹ transients (H) obtained in response to the train of 5 APs in I with the PKC inhibitor bisindolylmaleimide (GF109203x) added to the electrode solution. J–L: as G–I at 200 m from the soma. M: average plot of the effect of phorbol-di-acetate (PDA) (expressed as the ratio between the amplitude of the Ca2⫹ peak in the presence of PDA and the amplitude in control) for the first (), third (), and fifth (Œ) Ca2⫹ transients as a function of the distance from the soma in control conditions (red symbols, n ⫽ 8) and when GF109203x was added to the pipette (green symbols, n ⫽ 7).
105 • MARCH 2011 •
www.jn.org
BACK-PROPAGATION IN ENTORHINAL LAYER V NEURONS
Ca2⫹ peak in the train, measured as described in METHODS, by the same fraction (1.4 ⫾ 0.1 of the control values for the first, third, and fifth Ca2⫹ transient in the train, n ⫽ 7, Fig. 6). In addition, Ba2⫹ appeared to slow down the declining phase of every Ca2⫹ transient. The time course of the first Ca2⫹ transient under control conditions and in the presence of Ba2⫹ have been scaled to their respective peak amplitudes in the inset in Fig. 6B to emphasize the difference in the rate of decline. This effect resulted in a greater temporal summation of the Ca2⫹ transients in the presence of Ba2⫹. These results lead to the conclusion that A-type K⫹ currents contribute to the modulation of the amplitude of the Ca2⫹ transient associated with each spike in the train. Moreover, when applied in the presence of Ba2⫹, the effect of PDA (10 M) was still progressively larger for Ca2⫹ transients associated with later bAPs in the train (1.1 ⫾ 0.1, 1.5 ⫾ 0.1, and 2.3 ⫾ 0.2 of the values obtained in the presence of Ba2⫹ for the first, third, and fifth bAP, respectively, n ⫽ 7) and therefore activity dependent. The differential effect of Ba2⫹ and PDA on a train can be appreciated in Fig. 6E, where the ratios of the amplitude of the first, third, and fifth Ca2⫹ transients to the amplitude of the first transient are plotted for each of the three conditions. The ratios recorded in the presence of Ba2⫹ do not differ from those in control (0.46 ⫾ 0.04 vs. 0.44 ⫾ 0.02 and 0.28 ⫾ 0.04 vs. 0.28 ⫾ 0.02 for the third and the fifth peak, respectively; n ⫽ 7, P ⬎ 0.2, Wilcoxon matched-pairs signed-ranks test). On the other hand, when PDA was perfused in the presence of Ba2⫹, there was a significant increase in the ratio for subsequent spikes in the train (from 0.46 ⫾ 0.04 to 0.79 ⫾ 0.05, P ⬍ 0.02 for the third and from 0.28 ⫾ 0.04 to 0.46 ⫾ 0.07, P ⬍ 0.02 for the fifth; Wilcoxon matched-pairs signed-ranks test). Taken together, these data lead to the conclusion that both dendritic Na⫹ and A-type K⫹ channels contribute to the activity-dependent decrease of spike back-propagation in a train. Whereas we cannot exclude the contribution of other K⫹ channel to this phenomenon, the slow recovery from inactivation of Na⫹ channels,
1377
which leads to an accumulation of inactivation during a train, appears to be mainly responsible for the modulation by PKC. DISCUSSION
In this work, spike back-propagation in the apical dendrites of layer V pyramidal neurons of the medial EC was analyzed. The main findings are as follows: 1) the amplitude of the Ca2⫹ transients associated with bAPs decreases with increasing distance from the soma, and this decrease is countered by the blockade of A-type K⫹ channels; 2) the amplitude of the Ca2⫹ transients is highly activity dependent, as it decreases progressively during repetitive firing; this decline is more prominent at distal locations and is removed by PKC activation. Taken together, these results show that bAP amplitude in medial EC layer V neurons is controlled by the activity of dendritic Na⫹ and K⫹ channels and that A-type K⫹ channels play a fundamental role in the distance-dependent decrease of bAP amplitude. Under control conditions, these channels reduce the extent of depolarization and Ca2⫹ influx in the distal portions of the apical dendritic tree compared with the more proximal portions during a back-propagating spike. EC layer V neurons have an essential function in the hippocampal-neocortical circuitry that is responsible for the replay and consolidation of memory (Lavenex and Amaral 2000). Whereas the hippocampal output is aimed mainly at the proximal portion of the apical dendrites (Sørensen and Shipley 1979), the distal dendrites of most layer V neurons extend to the superficial layers, where they branch in complex tufts that receive inputs from the presubiculum (Wouterlood et al. 2004) and could be targeted by fibers from different cortical regions (as hypothesized by Canto et al. 2008). The presence of excitatory synapses throughout the dendritic tree of EC layer V neurons is confirmed by two-photon images that show a relatively high density of spines at distal locations (Fig. 1). The Ca2⫹ signals associated with bAPs could underlie a form of communication between the two relatively independent prox-
Fig. 6. PKC activators remove the activity-dependent decrease in bAP amplitude mainly by modulation of dendritic Na⫹ channels. A: 2-photon image of a dendritic portion imaged at ⬃200 m from the soma, showing the location of the line scan. B: Ca2⫹ transients obtained from the line scan in A in control conditions (black trace), in the presence of Ba2⫹ 150 M (red trace), and in the presence of Ba2⫹ ⫹ PDA 10 M (green trace) in response to the train of APs shown in C. Inset: time course of the first Ca2⫹ transients for the control and Ba2⫹ traces scaled to match their peak amplitude (scale bars 50 ms and 20% ⌬F/F). The difference in the rate during the declining phase is evident. D: average data showing the effect of Ba2⫹ and PDA activators in the presence of Ba2⫹ on the amplitude of the first, third, and fifth peak in a 20-Hz train (n ⫽ 7). E: plot of the amplitude of ratio of the first, third, and fifth peak to the first peak shows the difference between the effect of Ba2⫹ (red symbols) and PDA in the presence of Ba2⫹ (green symbols) on the activity-dependent decrease of the amplitude of the Ca2⫹ transients in a 20-Hz train measured at 200 m from the soma. J Neurophysiol • VOL
105 • MARCH 2011 •
www.jn.org
1378
BACK-PROPAGATION IN ENTORHINAL LAYER V NEURONS
imal and distal dendritic regions of these cells. The dependence of bAP amplitude on the activity of A-type K⫹ channels, which are highly regulated by neurotransmitters and plasticity mechanisms (Hoffman and Johnston 1999; Frick et al. 2004), indicates that the extent of propagation could depend on the release of specific neuromodulators and therefore on the behavioral state of the animal. In particular, the coupling between distal and proximal regions would be more efficient when the activity of A-type K⫹ channels is decreased in the presence of neurotransmitters such as norepinephrine and acetylcholine (Hoffman and Johnston 1999). Spike back-propagation could also be extended through boosting by appropriately timed synaptic inputs, as shown in other pyramidal neurons (Magee and Johnston 1997; Stuart and Häusser 2001). This type of mechanism would be in effect mostly for the small bAPs at distal locations or for later spikes in a high-frequency train throughout the apical dendrite because EC layer V neurons show a significant activity-dependent decrease in the amplitude of the Ca2⫹ transients associated with bAPs already at proximal locations (⬎100 m, Fig. 4). A similar activity-dependent decrease has been shown in neocortical layer V and hippocampal CA1 pyramidal neurons (Callaway and Ross 1995; Spruston et al. 1995; Stuart et al. 1997), where it is due to a slow cumulative inactivation of dendritic Na⫹ channels (Colbert et al. 1997; Jung et al. 1997), but it is absent in hippocampal oriens-alveus interneurons and basket cells (Martina et al. 2000; Hu et al. 2010) and neocortical layer II/III neurons (Larkum et al. 2007). This observation confirms that different neurons in the central nervous system have developed distinct strategies, presumably to adjust their dendritic features to optimally implement their specific functions. This study shows that the activity dependence of dendritic calcium influx associated with each bAP in a train in medial EC layer V neurons depends upon both Na⫹ and K⫹ channels because the peak amplitude and the temporal summation of the Ca2⫹ transients were affected by Ba2⫹ (Fig. 6B) and because the decrease in the Ca2⫹ peaks could be partially removed by phosphorylation by PKC (Figs. 5 and 6), which has been shown to speed the recovery from inactivation of dendritic Na⫹ channels in CA1 neurons (Colbert and Johnston 1998). In particular, Ba2⫹ increased the peak of each Ca2⫹ transient in the train by a constant percentage of ⬃40%, whereas the PKC activator PDA produced a larger increase for the Ca2⫹ transients associated with each subsequent bAPs in a train. These results do not imply that the A-type K⫹ current does not contribute to the activity-dependent decrease in the amplitude of the peak of each Ca2⫹ transient; on the contrary, the density of A-type K⫹ channels can be very important in determining the degree of activity-dependence. Cumulative inactivation of dendritic Na⫹ channels attributable to prior activity reduces their availability for spike back-propagation during a train of action potentials; however, at low levels of A-type K⫹ current, a decrease in sodium channel availability attributable to prior activity might have little influence on spike amplitude. At the higher levels of A-type K⫹ channels expressed at more distal dendritic locations, the inactivation of Na⫹ channels becomes more significant because more sodium-channel availability is required to counter the larger A-type K⫹ current to allow back-propagation, producing a strong activity-dependent decrease in the amplitude of the Ca2⫹ transients. Although PKC activation could in principle decrease the activity of A-type K⫹ J Neurophysiol • VOL
channels as well (Hoffman and Johnston 1999), PKC activators were very effective in removing the activity-dependent decrease in the later Ca2⫹ transients in a 20-Hz train, even when they were applied in the presence of A-type K⫹ channels blockers, and therefore were probably acting mainly on dendritic Na⫹ channels. These data support the conclusion that PKC activation modulates the activity-dependent decrease in the amplitude of Ca2⫹ transients associated with bAPs in a train mainly by removing the slow cumulative inactivation of dendritic Na⫹ channels; we cannot exclude, however, the possibility that other dendritic K⫹ channels contribute to the decrease in the amplitude of the Ca2⫹ transients associated with later bAPs in a train, nor can we exclude the possibility that other targets are also modulated by PKC. As discussed above, the decrease in amplitude Ca2⫹ transients associated with later bAPs within a train, attributable to the slow cumulative inactivation of dendritic Na⫹ channels, is an attribute that EC layer V neurons share with neocortical layer V pyramidal neurons and CA1 pyramidal neurons. However, the distance-dependent increase in A-type K⫹ channel density that mediates the attenuation of bAP amplitude along the apical dendrite is an attribute shared with CA1 neurons but not neocortical pyramidal neurons, in which the density of these channels appears to be constant in the first 500 m from the soma (Bekkers 2000). This evidence somewhat counterintuitively suggests that the dendritic functionality of EC layer V neurons is more closely aligned with hippocampal compared with neocortical neurons. Further investigations are needed to assess further differences and similarities in the intrinsic dendritic excitability and integration properties among these cell types. In conclusion, this study shows that Na⫹ and transient K⫹ channels control spike back-propagation in the apical dendrites of EC layer V pyramidal neurons. Because the activity of Na⫹ and K⫹ channels can be regulated by membrane potential and second messenger systems, bAP amplitude and the associated Ca2⫹ influx can be highly modulated as well, with important functional implications for dendritic integration and associative synaptic plasticity. ACKNOWLEDGMENTS The author thanks the Dart Foundation for support through the Dart Neuroscience Scholars Program in Learning and Memory Award at MBL. The author is grateful to Prairie Technologies and Nikon for the loan of imaging equipment and to Carmen Canavier for helpful suggestions on the manuscript.
GRANTS This work was supported by the National Institutes of Health (NS035865 and NS069714).
DISCLOSURES No conflicts of interest, financial or otherwise, are declared by the author.
REFERENCES Amaral DG, Witter MP. The Hippocampal region. In: The Rat Nervous System, edited by Paxinos G. San Diego, CA: Elsevier Academic, 1995, p. 443– 493. Arnold SE. Cellular, and molecular neuropathology of the parahippocampal region in schizophrenia. Ann NY Acad Sci 911: 275–292, 2000.
105 • MARCH 2011 •
www.jn.org
BACK-PROPAGATION IN ENTORHINAL LAYER V NEURONS Bannister NJ, Larkman AU. Dendritic morphology of CA1 pyramidal neurons from the rat hippocampus: I. Branching patterns. J Comp Neurol 360: 150 –160, 1995. Bekkers JM. Distribution and activation of voltage-gated potassium channels incell-attached and outside-out patches from large layer 5 cortical pyramidal neurons of the rat. J Physiol 525: 611– 620, 2000. Buzsaki G. Two-stage model of memory trace formation: a role for “noisy” brain states. Neuroscience 31: 551–570, 1989. Callaway JC, Ross WN. Frequency-dependent propagation of sodium action potentials in dendrites of hippocampal CA1 pyramidal neurons. J Neurophysiol 74: 1395–1403, 1995. Canto CB, Wouterlood FG, Witter MP. What does the anatomical organization of the entorhinal cortex tell us? Neural Plast 2008: 38124, 2008. Chatelain FC, Alagem N, Xu Q, Pancaroglu R, Reuveny E, Minor DL Jr. The pore helix dipole has a minor role in inward rectifier channel function. Neuron 47: 833– 843, 2005. Colbert CM, Johnston D. Protein kinase C activation decreases activitydependent attenuation of dendritic Na⫹ current in hippocampal CA1 pyramidal neurons. J Neurophysiol 79: 491– 495, 1998. Colbert CM, Magee JC, Hoffman DA, Johnston D. Slow recovery from inactivation of Na⫹ channels underlies the activity-dependent attenuation of dendritic action potentials in hippocampal CA1 pyramidal neurons. J Neurosci 17: 6512– 6521, 1997. de Curtis M, Paré D. The rhinal cortices: a wall of inhibition between the neocortex, and the hippocampus. Prog Neurobiol 74: 101–110, 2004. Frick A, Magee J, Johnston D. LTP is accompanied by an enhanced local excitability of pyramidal neuron dendrites. Nat Neurosci 7: 126 –135, 2004. Fyhn M, Molden S, Witter MP, Moser EI, Moser MB. Spatial representation in the entorhinal cortex. Science 305: 1258 –1264, 2004. Gasparini S, Losonczy A, Chen X, Johnston D, Magee JC. Associative pairing enhances action potential back-propagation in radial oblique branches of CA1 pyramidal neurons. J Physiol 580: 787– 800, 2007. Gasparini S, Magee JC. State-dependent dendritic computation in hippocampal CA1 pyramidal neurons. J Neurosci 26: 2088 –2100, 2006. Hamam BN, Kennedy TE, Alonso A, Amaral DG. Morphological and electrophysiological characteristics of layer V neurons of the rat medial entorhinal cortex. J Comp Neurol 418: 457– 472, 2000. Hargreaves EL, Rao G, Lee I, Knierim JJ. Major dissociation between medial and lateral entorhinal input to dorsal hippocampus. Science 308: 1792–1794, 2005. Helmchen F, Imoto K, Sakmann B. Ca2⫹ buffering and action potentialevoked Ca2⫹ signaling in dendrites of pyramidal neurons. Biophys J 70: 1069 –1081, 1996. Hoffman DA, Magee JC, Colbert CM, Johnston D. K⫹ channel regulation of signal propagation in dendrites of hippocampal pyramidal neurons. Nature 387: 869 – 875, 1997. Hoffman DA, Johnston D. Neuromodulation of dendritic action potentials. J Neurophysiol 81: 408 – 411, 1999. Hu H, Martina M, Jonas P. Dendritic mechanisms underlying rapid synaptic activation of fast-spiking hippocampal interneurons. Science 327: 52–58, 2010. Jellinger K, Braak H, Braak E, Fischer P. Alzheimer lesions in the entorhinal region and isocortex in Parkinson’s and Alzheimer’s diseases. Ann NY Acad Sci 640: 203–209, 1991.
J Neurophysiol • VOL
1379
Jung HY, Mickus T, Spruston N. Prolonged sodium channel inactivation contributes to dendritic action potential attenuation in hippocampal pyramidal neurons. J Neurosci 17: 6639 – 6646, 1997. Larkman AU. Dendritic morphology of pyramidal neurons of the visual cortex of the rat: I. Branching patterns. J Comp Neurol 306: 307–319, 1991. Larkum ME, Waters J, Sakmann B, Helmchen F. Dendritic spikes in apical dendrites of neocortical layer 2/3 pyramidal neurons. J Neurosci 27: 8999 – 9008, 2007. Lasser-Ross N, Miyakawa H, Lev-Ram V, Young SR, Ross WN. High time resolution fluorescence imaging with a CCD camera. J Neurosci Meth 36: 253–261, 1991. Lavenex P, Amaral DG. Hippocampal-neocortical interaction: a hierarchy of associativity. Hippocampus 10: 420 – 430, 2000. Magee JC, Johnston D. A synaptically controlled, associative signal for Hebbian plasticity in hippocampal neurons. Science 275: 209 –213, 1997. Magee JC, Johnston D. Plasticity of dendritic function. Curr Opin Neurobiol 15: 334 –342, 2005. Martina M, Vida I, Jonas P. Distal initiation and active propagation of action potentials in interneuron dendrites. Science 287: 295–300, 2000. Migliore M, Shepherd GM. Opinion: an integrated approach to classifying neuronal phenotypes. Nat Rev Neurosci 6: 810 – 818, 2005. Rosenkranz JA, Johnston D. Dopaminergic regulation of neuronal excitability through modulation of Ih in layer V entorhinal cortex. J Neurosci 26: 3229 –3244, 2006. Rosenkranz JA, Johnston D. State-dependent modulation of amygdala inputs by dopamine-induced enhancement of sodium currents in layer V entorhinal cortex. J Neurosci 27: 7054 –7069, 2007. Sørensen KE, Shipley MT. Projections from the subiculum to the deep layers of the ipsilateral presubicular and entorhinal cortices in the guinea pig. J Comp Neurol 188: 313–333, 1979. Spruston N, McBain C. Structural and functional properties of hippocampal neurons. In: The Hippocampus Book, edited by Andersen P, Morris R, Amaral D, Bliss T, and O’Keefe J. Oxford, UK: Oxford University Press, 2007, p. 133–201. Spruston N, Schiller Y, Stuart G, Sakmann B. Activity-dependent action potential invasion and calcium influx into hippocampal CA1 dendrites. Science 268: 297–300, 1995. Spruston N, Stuart G, Häusser M. Dendritic integration. In: Dendrites, edited by Stuart G, Spruston N, Häusser M. New York: Oxford University Press, 2008, p. 225–250. Storm JF. Potassium currents in hippocampal pyramidal cells. Prog Brain Res 83: 161–187, 1990. Stuart GJ, Häusser M. Dendritic coincidence detection of EPSPs and action potentials. Nat Neurosci 4: 63–71, 2001. Stuart G, Schiller J, Sakmann B. Action potential initiation and propagation in rat neocortical pyramidal neurons. J Physiol 505: 617– 632, 1997. Witter MP, Amaral DG. The hippocampal region. In: The Rat Brain, edited by Paxinos G. San Diego, CA: Elsevier, 2004, p. 635–704. Wouterlood FG, Van Haeften T, Eijkhoudt M, Baks-Te-Bulte L, Goede PH, Witter MP. Input from the presubiculum to dendrites of layer-V neurons of the medial entorhinal cortex of the rat. Brain Res 1013: 1–12, 2004.
105 • MARCH 2011 •
www.jn.org
Distance- and activity-dependent modulation of spike back-propagation in layer V pyramidal neurons of the medial entorhinal cortex Sonia Gasparini Neuroscience Center, Louisiana State University Health Science Center, New Orleans, Louisiana, and Marine Biological Laboratory, Woods Hole, Massachusetts Submitted 7 January 2010; accepted in final form 2 January 2011
Gasparini S. Distance- and activity-dependent modulation of spike back-propagation in layer V pyramidal neurons of the medial entorhinal cortex. J Neurophysiol 105: 1372–1379, 2011. First published January 5, 2011; doi:10.1152/jn.00014.2010.—Layer V principal neurons of the medial entorhinal cortex receive the main hippocampal output and relay processed information to the neocortex. Despite the fundamental role hypothesized for these neurons in memory replay and consolidation, their dendritic features are largely unknown. Highspeed confocal and two-photon Ca2⫹ imaging coupled with somatic whole cell patch-clamp recordings were used to investigate spike back-propagation in these neurons. The Ca2⫹ transient associated with a single back-propagating action potential was considerably smaller at distal dendritic locations (⬎200 m from the soma) compared with proximal ones. Perfusion of Ba2⫹ (150 M) or 4-aminopyridine (2 mM) to block A-type K⫹ currents significantly increased the amplitude of the distal, but not proximal, Ca2⫹ transients, which is strong evidence for an increased density of these channels at distal dendritic locations. In addition, the Ca2⫹ transients decreased with each subsequent spike in a 20-Hz train; this activitydependent decrease was also more prominent at more distal locations and was attenuated by the perfusion of the protein kinase C activator phorbol-di-acetate. These data are consistent with a phosphorylationdependent control of back-propagation during trains of action potentials, attributable mainly to an increase in the time constant of recovery from voltage-dependent inactivation of dendritic Na⫹ channels. In summary, dendritic Na⫹ and A-type K⫹ channels control spike back-propagation in layer V entorhinal neurons. Because the activity of these channels is highly modulated, the extent of the dendritic Ca2⫹ influx is as well, with important functional implications for dendritic integration and associative synaptic plasticity. dendrite; Na⫹ channels; A-type K⫹ channels
(EC) is a key relay structure for the flow of information between the hippocampus and the neocortex (Amaral and Witter 1995). Not only does it act as a primary interface, it also plays a critical role in the computation of multisensory and cognitive modalities. The latter function is clearly supported by the direct involvement of the EC in neurodegenerative and psychiatric disorders such as Alzheimer’s disease, epilepsy, and schizophrenia (Jellinger et al. 1991; Arnold 2000; de Curtis and Paré 2004). Various studies have contributed to our knowledge of the features of EC principal neurons (Hamam et al. 2000; Spruston and McBain 2007) and their connections (Witter and Amaral 2004). Layer V neurons are the main target of the hippocampal output and send their axons to cortical regions; they are thus likely to play an important role in memory replay and consol-
THE ENTORHINAL CORTEX
Address for reprint requests and other correspondence: S. Gasparini, Neuroscience Ctr., Louisiana State Univ. Health Science Ctr., 2020 Gravier St., New Orleans, LA 70112 (e-mail: [email protected]). 1372
idation, which involve hippocampal-neocortical communications (Buzsaki 1989). To understand the function of EC layer V neurons, it is essential to understand how they integrate the inputs they receive to generate the output that is transferred to the neocortex; this process depends on the neuronal morphology and the properties of dendritic voltage-dependent channels, in addition to the input temporal and spatial patterns (Migliore and Shepherd 2005; Gasparini and Magee 2006). Dendritic channels are also responsible for the back-propagation of somatic spikes, with features that highly differ from neuron to neuron (for a review see Spruston et al. 2008). A detailed knowledge of the mechanisms of back-propagation in neurons of the central nervous system is essential because back-propagating action potentials (bAPs) are involved in the modulation of dendritic excitability and associative synaptic plasticity (Magee and Johnston 1997; Magee and Johnston 2005). Despite the fundamental role of EC layer V neurons in memory consolidation, much remains to be known about their dendritic excitability. The EC consists of two functionally distinct regions; spatial information processing is mainly restricted to the medial EC and nonspatial processing to the lateral EC (Fyhn et al. 2004; Hargreaves et al. 2005). Previous studies focused on dopaminergic modulation of hyperpolarization-activated cation and Na⫹ channels in the lateral EC (Rosenkranz and Johnston 2006, 2007). Here imaging techniques were used to measure Ca2⫹ signals associated with bAPs in regions of the apical dendrites of medial entorhinal layer V neurons that cannot be visualized under differential interference-contrast (DIC) microscopy. The amplitude of the Ca2⫹ signals associated with bAPs was found to be both distance and activity dependent and is controlled by the activity of dendritic Na⫹ and K⫹ channels. In particular, smaller distal Ca2⫹ signals could be boosted by blocking A-type K⫹ channels, suggesting that these channels control the amplitude of bAPs along the apical dendrites of medial EC layer V neurons, similar to what has been reported for CA1 pyramidal neurons (Hoffman et al. 1997). METHODS
EC slices were prepared from 7- to 9-wk-old Sprague Dawley rats, according to methods approved by the Louisiana State University Health Science Center Institutional Animal Care and Use Committee. Briefly, rats were anesthetized with ketamine and xylazine, and the brain was perfused with an oxygenated, ice-cold solution through the ascending aorta. After decapitation, the brain was rapidly removed, and 400-m-thick horizontal slices including the EC and the rest of the hippocampal formation were cut using a vibratome. After a
0022-3077/11 Copyright © 2011 the American Physiological Society
www.jn.org
BACK-PROPAGATION IN ENTORHINAL LAYER V NEURONS
recovery period of at least 1 h, an individual slice was transferred to the recording chamber; the external solution used for recordings contained (in mM) 125 NaCl, 2.5 KCl, 25 NaHCO3, 1.25 Na2HPO4, 2 CaCl2, 1 MgCl2, and 25 dextrose and was saturated with 95% O2-5% CO2 at 34 –36°C (pH 7.4). Whole cell patch-clamp somatic recordings from layer V pyramidal neurons in the ventromedial EC were performed using a Dagan BVC-700 amplifier in the active “bridge” mode. At the end of each recording, a low-magnification objective that allows a view of the whole slice was used to verify that the recording electrode was actually placed in the medial portion of layer V, just below the lamina disseccans. Patch pipettes had a resistance of 2–3 M⍀ when filled with a solution containing (in mM) 130 K-methylsulphonate, 10 HEPES, 4 NaCl, 4 Mg2ATP, 0.3 Tris2GTP, and 14 phosphocreatine (pH 7.3). Oregon Green 1,2-bis(2-aminophenoxy)ethane-N,N,N=,N=tetraacetate-1 [Oregon Green BAPTA-1 (OGB-1), 100 M; Invitrogen, Carlsbad, CA] was added to the internal solution to monitor changes in [Ca2⫹]i. Somatic action potentials were elicited by brief current steps (2 nA for 2 ms). EC layer V pyramidal neurons were visualized using a Nikon Eclipse FN-1 microscope, equipped with DIC optics under infrared illumination (Nikon, Melville, NY) coupled with a swept-field confocal system in slit mode (Prairie Technologies, Middleton, WI). A neuro-charge-coupled device (CCD) camera (RedShirt Imaging, Decatur, GA) with an 80 ⫻ 80 pixel array was used to acquire the optical signal in response to excitation at 488 nm (sequential frame rate 0.5 kHz). In other sets of experiments, ultrafast, pulsed, laser light at 920 nm (Mira 900F; Coherent, Santa Clara, CA) was used to excite OGB-1 and image local Ca2⫹ in the entorhinal apical dendrites using a multiphoton scanner (Ultima, Prairie Technologies) mounted over an Olympus BX61WI microscope (Center Valley, PA). Changes in [Ca2⫹]i associated with bAPs were quantified by calculating ⌬F/F, where F is the fluorescence intensity before stimulation, after subtracting autofluorescence, and ⌬F is the change in fluorescence during neuronal activity (Lasser-Ross et al. 1991). The autofluorescence of the tissue was measured in a region of equal size adjacent to the dye-filled dendrite. The ⌬F/F measurements were repeated three to eight times and averaged. The high frame rate of the optical imaging (0.3– 0.5 kHz) allowed the resolution of the contribution of each bAP to the cumulative Ca2⫹ signal (Figs. 4 – 6). To estimate the amplitude of the Ca2⫹ transients associated with spikes after the first one in a train, the residual fluorescence was calculated by fitting the decline of the previous peak with a single exponential (Helmchen et al. 1996) and subtracting the projected value in the absence of a subsequent spike from the actual maximum value observed in response to a given spike. In a set of experiments, local application of Ba2⫹ (375 M) was achieved by using a pressure ejection system (Picospritzer; Parker Hannifin, Fairfield, NJ) connected to a patch pipette (2–5 m diameter, pressures of 2–5 psi, and durations of 0.5–2.0 min). With these settings, an ⬃200-m distal region of the cell (⬃250 – 450 m from the soma) was perfused when the pipette was placed within 20 m of the dendritic trunk at 350 m from the soma. The area of the cell effectively reached by the local perfusion was tested using a fluorescent dye (Alexa Fluor 488). Phorbol-di-acetate (PDA) and bisindolylmaleimide (GF109203x) were purchased from Sigma-Aldrich (St. Louis, MO) and dissolved in DMSO to obtain stock solutions of 10 and 2 mM, respectively; PDA was then added to the external solution and bisindolylmaleimide to the patch pipette. The concentration of DMSO in the final solution was 0.1%. Control experiments were performed to determine that the addition of DMSO to the pipette solution did not by itself alter the response to PDA (n ⫽ 3). Data are reported as means ⫾ SE. Statistical comparisons were performed by using a one-way analysis of variance and appropriate post hoc tests. Means were considered to be significantly different when P ⬍ 0.05. J Neurophysiol • VOL
1373
RESULTS
The apical dendrites of layer V pyramidal neurons of the medial EC are too small and convoluted to be visualized under DIC microscopy; therefore, dendritic excitability cannot be examined directly using electrophysiological techniques. Instead, the Ca2⫹ transients associated with back-propagating spikes initiated by a brief current injection at the soma were analyzed. Previous studies have shown that Ca2⫹ transients are very good indicators of AP amplitude in the dendrites (Magee
Fig. 1. The amplitude of back-propagating action potential (bAP)-associated Ca2⫹ signals decreases with the distance from the soma along the apical dendrites of entorhinal cortex (EC) layer V neurons. A: 2-dimensional projection of 2-photon 3D image stacks for an EC layer V pyramidal neuron filled with Oregon Green BAPTA-1 (OGB-1) (100 M). The white boxes outline regions of the apical dendrite that are expanded on the sides, showing the presence of spines all over the apical dendrite. B: Ca2⫹ transients (expressed as ⌬F/F, where F is fluorescence) generated in response to a somatic action potential (C) measured with a line scan at the locations marked by gray lines. D: plot of the mean amplitude of the Ca2⫹ peaks associated with bAPs as a function of the distance from the soma (n ⫽ 18). The experimental points are fitted with a sigmoidal function.
105 • MARCH 2011 •
www.jn.org
1374
BACK-PROPAGATION IN ENTORHINAL LAYER V NEURONS
and Johnston 1997; Gasparini et al. 2007) and that changes in Ca2⫹ transients are correlated primarily with changes in bAP amplitude rather than in their width (see Figs. 2 and 6 in Gasparini et al. 2007). EC layer V neurons were filled with the Ca2⫹ indicator OGB-1 through a somatic patch pipette (Fig. 1). The two-dimensional projection of a series of two-photon 3D image stacks from an EC layer V neuron in Fig. 1A shows the prominent apical dendrite, which branches in a tuft in the superficial layers. The insets on the sides show that spines are distributed all over the length of the apical dendrite and tuft. Most medial EC layer V pyramidal neurons appear to lack the extensive oblique dendritic branches that characterize layer V neocortical and CA1 pyramidal neurons (Larkman 1991; Bannister and Larkman 1995). A first set of experiments used a multi-photon line scan to analyze the amplitude of the dendritic Ca2⫹ signal (⌬F/F) associated with a somatic AP at different locations (⬃60 m apart) and revealed a clear dependence of signal amplitude on the distance from the soma. As shown in Fig. 1B, the amplitude of dendritic Ca2⫹ transients was constant for the first ⬃150 m of the apical dendrite (with an average ⌬F/F of 68 ⫾ 5% at 60 m from the soma, n ⫽ 18, Fig. 1D) and started decreasing at ⬃200 m. The signal became
undetectable from baseline for distances ⬎350 m (⌬F/F of 8 ⫾ 2% at ⬃360 m, n ⫽ 18, Fig. 1D), implying that the bAPassociated depolarization was unable to activate a significant number of voltage-dependent Ca2⫹ channels at this level. To address the mechanism underlying the distance-dependent decrease in bAP-amplitude in medial EC layer V neurons, the sensitivity of distal bAP-associated Ca2⫹ transients to the blockade of A-type K⫹ channels by Ba2⫹ (150 M) and 4-aminopyridine (4-AP, 2 mM) was tested. The first set of experiments was performed using a high-speed CCD camera coupled with confocal microscopy; the orange box in Fig. 2A outlines the region of interest for the optical traces in Fig. 2B. A 20-Hz train of three APs was evoked at the soma (Fig. 2C). At 150 M, Ba2⫹ has been shown to block A-type K⫹ currents in the distal apical dendrites of hippocampal CA1 pyramidal neurons and to thereby boost the smaller bAPs at these locations (Gasparini et al. 2007). However, because Ba2⫹ also blocks inward rectifier K⫹ currents at lower concentrations, it was necessary to determine that the block of these channels did not affect back-propagation. For this reason, I first perfused Ba2⫹ at 20 M, a concentration that selectively blocks inward rectifier K⫹ channels (Chatelain et al. 2005; Gasparini et al.
Fig. 2. Ba2⫹ (150 M but not 20 M) and 4-aminopyridine (4-AP) enhance spike back-propagation into the distal apical dendrites of EC layer V pyramidal neurons. A: charge-coupled device (CCD) image of the distal (⬎300 m) portion of an EC layer V pyramidal neuron filled with OGB-1 (100 M). The orange box outlines the region of interest where the changes in fluorescence (⌬F/F, B) were monitored in response to a 20-Hz train of 3 APs (C) in control conditions (black trace), during the perfusion of Ba2⫹ (20 and 150 M, green and red trace, respectively) and upon wash out (w-o) (blue trace). The expanded time scale (C, inset) better illustrates the effect of Ba⫹ on the first action potential. D: mean data of the amplitude of the first Ca2⫹ transient measured in distal apical dendrites at ⬃350 m from the soma (n ⫽ 10) show that the block of inward rectifier K⫹ channels by Ba2⫹ 20 M did not change the Ca2⫹ transients, but Ba2⫹ 150 M significantly increased them. E: 2-photon image showing a portion of the apical dendrite at ⬃330 m from the soma and the location of the line scan. F: Ca2⫹ transients evoked by the somatic action potential in G in control conditions, during the perfusion of 4-AP (2 mM) and upon wash out. H: average data of the amplitude of the Ca2⫹ transients associated with bAPs at ⬃350 m from the soma for n ⫽ 8 neurons. *P ⬍ 0.01, **P ⬍ 0.005. J Neurophysiol • VOL
105 • MARCH 2011 •
www.jn.org
BACK-PROPAGATION IN ENTORHINAL LAYER V NEURONS
2007). As shown in Fig. 2B, the perfusion of lower Ba2⫹ concentrations did not affect the Ca2⫹ transients recorded at ⬃350 m; the application of Ba2⫹ (150 M), however, greatly increased the Ca2⫹ peaks observed in the distal dendrites (Fig. 2B), indicating that the increase in bAP amplitude at distal dendritic locations is mediated by the effect of Ba2⫹ on A-type K⫹ channels. On average, the first Ca2⫹ transient increased from 11 ⫾ 1% under control conditions to 62 ⫾ 10% in the presence of Ba2⫹ 150 M at 350 m from the soma (Fig. 2D; n ⫽ 10, P ⬍ 0.005, Wilcoxon matched-pairs signed-ranks test). The dependence of bAP amplitude on A-type K⫹ channels in the distal apical dendrites was confirmed by examining the effect of 4-AP on the distal Ca2⫹ transients, recorded through a two-photon line scan (Fig. 2, E–G). At ⬃350 m from the soma, 4-AP (2 mM) increased the Ca2⫹ transient evoked by a single somatic AP from 10 ⫾ 2% to 65 ⫾ 9% (Fig. 2H; n ⫽ 8, P ⬍ 0.01, Wilcoxon matched-pairs signed-ranks test). Although Ba2⫹ and 4-AP block other K⫹ channels, the spectra of the channels blocked do not overlap except for A-type K⫹ channels (Storm 1990); therefore, the similarity of the effects of 4-AP and Ba2⫹ is strong evidence that these effects are mediated primarily by these K⫹ channels. A putative reduction of back-propagating spike amplitude by a gradient of A-type K⫹ channels along the apical dendrites was further tested by local application of Ba2⫹ at various distances from the soma. Local perfusion of Ba2⫹ (Fig. 3) produced only a slight increase (⬃10%) of the Ca2⫹ transient in the proximal apical dendrite compared with a fourfold increase of the smaller Ca2⫹ transient at ⬃350 m from the soma. The average data (n ⫽ 15) show that the effect of local perfusion of Ba2⫹ increased with the distance from the soma, such that for locations ⱖ200 m it effectively removed the distance-dependent decrease in the Ca2⫹ transients recorded during bAP in control conditions (Fig. 3, B and C). These data suggest that the decrease in bAP amplitude along the apical dendrites of entorhinal layer V neurons can be attributed to a
1375
contribution of A-type K⫹ channels that increases with distance from the soma, similar to what has been shown for CA1 pyramidal neurons (Hoffman et al. 1997). In addition to the dependence of bAP amplitude on the spatial location, back-propagation can also depend on the firing history of a neuron. Specifically, bAP amplitude has been reported to decrease in an activity-dependent manner in CA1 and neocortical layer V pyramidal neurons attributable to a slow cumulative inactivation of dendritic Na⫹ channels (Callaway and Ross 1995; Spruston et al. 1995; Colbert et al. 1997; Stuart et al. 1997). To investigate whether this was also the case in EC layer V neurons, the Ca2⫹ transients associated with a 20-Hz train of five somatically initiated spikes were analyzed (Fig. 4). The ratio of the first Ca2⫹ peak to the peak of its proximal counterpart started decreasing at ⬃200 m from the soma (consistent with the data in Fig. 1). In contrast, the ratios of the subsequent Ca2⫹ transients decreased at more proximal locations (Fig. 4B), and the decline became progressively more pronounced with each successive spike in the train, which is also evident from the average plot (Fig. 4D). Whereas the first Ca2⫹ peak decreased only to 0.89 ⫾ 0.05 of the initial value at 200 m, the third and the fifth Ca2⫹ transients dropped to 0.73 ⫾ 0.05 and 0.51 ⫾ 0.10, respectively, at the same distance (n ⫽ 10, Fig. 4D). These data suggest that dendritic Na⫹ channels in EC layer V neurons may undergo slow cumulative inactivation. The effect of the perfusion of PKC activator PDA (10 M) was examined at various distances from the soma to determine whether phosphorylation of these channels could accelerate their recovery from inactivation, thereby reversing the activity-dependent decrease in the bAP, as has been shown in CA1 pyramidal neurons (Colbert and Johnston 1998). Figure 5 shows the line scans obtained at 100 m and 200 m from the soma for two different neurons, with the typical activitydependent decrease of the later spikes in the train at more distal locations (compare the black traces in Fig. 5, B and E). The perfusion of PDA did not affect the Ca2⫹ transients at 100 m
Fig. 3. Local Ba2⫹ perfusion counters the distance-dependent decrease in bAP amplitude. A: Ca2⫹ transients obtained under control conditions (black trace) and during local perfusion of Ba2⫹ (gray trace). The diagram on the left shows the experimental configuration, with a whole cell somatic electrode and the line scan locations for proximal and distal regions, as well as the regions affected by Ba2⫹ pressure ejection at ⬃100 and ⬃350 m from the soma, respectively. B: average data of the amplitude of the Ca2⫹ transients associated with bAPs as a function of the distance from the soma in control conditions and during local perfusion of Ba2⫹ (n ⫽ 15). The effect of Ba2⫹ is more pronounced for distal dendritic locations, as shown clearly by the ratio plot in C.
J Neurophysiol • VOL
105 • MARCH 2011 •
www.jn.org
1376
BACK-PROPAGATION IN ENTORHINAL LAYER V NEURONS
lation of the activity-dependent decrease in the amplitude of the Ca2⫹ associated with bAPs. To determine the relative contributions of the Na⫹ and K⫹ currents, A-type K⫹ currents were blocked with Ba2⫹ (150 M), and the activity dependence of the Ca2⫹ transients within a train and its modulation by phorbol esters were examined at 200 m from the soma, where the activity-dependent decline in the Ca2⫹ transients is pronounced (see Fig. 5). Ba2⫹ increased the amplitude of every
Fig. 4. bAP-associated Ca2⫹ transients show an activity-dependent decrease at more distal (⬎150 m) dendritic locations. A: reconstructed resting fluorescence image of an EC layer V neuron filled with OGB-1 and imaged with a high-speed CCD camera in combination with a swept-field confocal scanner. B: optical recordings of the Ca2⫹ influx (expressed as ⌬F/F) generated in response to a 20-Hz train of 5 APs (C) in the regions of interest marked by the white boxes in A. D: plots of the average for the first (), third (Œ), and fifth (Œ) Ca2⫹ transients (normalized to the most proximal value) associated with a train of bAPs for 10 neurons show a more pronounced decrease for the bAPs later in the train.
from the soma (Fig. 5B) but strongly affected the activity dependence of back-propagation within a train at 200 m from the soma (Fig. 5E). The average data show that PDA has a negligible effect on the first Ca2⫹ transient up to 200 m (the ratio between the amplitude in the presence of PDA and that in control conditions was 1.1 ⫾ 0.1 for the first Ca2⫹ peak) but exerts progressively larger effects on each subsequent Ca2⫹ transients at more distal locations (the ratio was 1.4 ⫾ 0.1 and 2.1 ⫾ 0.3, for the third and the fifth peak at 200 m, respectively, n ⫽ 8, Fig. 5M). To verify that PDA acted through a PKC-dependent mechanism, a series of control experiments were performed to test the ability of a PKC inhibitor, bisindolylmaleimide (GF109203x), to block the effects of PDA on reversing the activity-dependent decrease (Fig. 5). GF109203x was added to the patch pipette solution and by itself had no effect on the typical activity-dependent decline of the Ca2⫹ transients associated with later bAPs in the train for distances greater than 100 m from the soma (compare Fig. 5, K and H). However, with the PKC inhibitor in the pipette, perfusion of PDA now failed to reverse the activity-dependent decrease in the Ca2⫹ transients associated with bAPs at any distance (Fig. 5, H and K). The ratio between the amplitude in the presence of PDA and that in control conditions at 200 m from the soma was 0.9 ⫾ 0.1, 1.0 ⫾ 0.1, and 1.1 ⫾ 0.1 for the first, third, and fifth Ca2⫹ transients, respectively (n ⫽ 7, Fig. 5M). These data indicate that the activity-dependent decrease in bAP amplitude can be effectively prevented by PKC activation. Whereas the slow cumulative inactivation of dendritic Na⫹ channels is a main determinant of the activity-dependent decline of Ca2⫹ transients in a train, the higher density of A-type K⫹ channels in the dendrites of layer V EC neurons could also contribute to this phenomenon. Moreover, PKC activation could affect K⫹ currents, therefore contributing to the moduJ Neurophysiol • VOL
Fig. 5. PKC activation by phorbol esters reduces the activity-dependent decrease of bAP amplitude. A: 2-photon image showing a portion of the apical dendrite and the location of the line scan at 100 m from the soma. B: Ca2⫹ transients generated in response to a 20-Hz train of 5 APs (C). D–F: as A– C at 200 m from the soma. G: 2-photon image showing the location of the line scan (at ⬃100 m from the soma) for the Ca2⫹ transients (H) obtained in response to the train of 5 APs in I with the PKC inhibitor bisindolylmaleimide (GF109203x) added to the electrode solution. J–L: as G–I at 200 m from the soma. M: average plot of the effect of phorbol-di-acetate (PDA) (expressed as the ratio between the amplitude of the Ca2⫹ peak in the presence of PDA and the amplitude in control) for the first (), third (), and fifth (Œ) Ca2⫹ transients as a function of the distance from the soma in control conditions (red symbols, n ⫽ 8) and when GF109203x was added to the pipette (green symbols, n ⫽ 7).
105 • MARCH 2011 •
www.jn.org
BACK-PROPAGATION IN ENTORHINAL LAYER V NEURONS
Ca2⫹ peak in the train, measured as described in METHODS, by the same fraction (1.4 ⫾ 0.1 of the control values for the first, third, and fifth Ca2⫹ transient in the train, n ⫽ 7, Fig. 6). In addition, Ba2⫹ appeared to slow down the declining phase of every Ca2⫹ transient. The time course of the first Ca2⫹ transient under control conditions and in the presence of Ba2⫹ have been scaled to their respective peak amplitudes in the inset in Fig. 6B to emphasize the difference in the rate of decline. This effect resulted in a greater temporal summation of the Ca2⫹ transients in the presence of Ba2⫹. These results lead to the conclusion that A-type K⫹ currents contribute to the modulation of the amplitude of the Ca2⫹ transient associated with each spike in the train. Moreover, when applied in the presence of Ba2⫹, the effect of PDA (10 M) was still progressively larger for Ca2⫹ transients associated with later bAPs in the train (1.1 ⫾ 0.1, 1.5 ⫾ 0.1, and 2.3 ⫾ 0.2 of the values obtained in the presence of Ba2⫹ for the first, third, and fifth bAP, respectively, n ⫽ 7) and therefore activity dependent. The differential effect of Ba2⫹ and PDA on a train can be appreciated in Fig. 6E, where the ratios of the amplitude of the first, third, and fifth Ca2⫹ transients to the amplitude of the first transient are plotted for each of the three conditions. The ratios recorded in the presence of Ba2⫹ do not differ from those in control (0.46 ⫾ 0.04 vs. 0.44 ⫾ 0.02 and 0.28 ⫾ 0.04 vs. 0.28 ⫾ 0.02 for the third and the fifth peak, respectively; n ⫽ 7, P ⬎ 0.2, Wilcoxon matched-pairs signed-ranks test). On the other hand, when PDA was perfused in the presence of Ba2⫹, there was a significant increase in the ratio for subsequent spikes in the train (from 0.46 ⫾ 0.04 to 0.79 ⫾ 0.05, P ⬍ 0.02 for the third and from 0.28 ⫾ 0.04 to 0.46 ⫾ 0.07, P ⬍ 0.02 for the fifth; Wilcoxon matched-pairs signed-ranks test). Taken together, these data lead to the conclusion that both dendritic Na⫹ and A-type K⫹ channels contribute to the activity-dependent decrease of spike back-propagation in a train. Whereas we cannot exclude the contribution of other K⫹ channel to this phenomenon, the slow recovery from inactivation of Na⫹ channels,
1377
which leads to an accumulation of inactivation during a train, appears to be mainly responsible for the modulation by PKC. DISCUSSION
In this work, spike back-propagation in the apical dendrites of layer V pyramidal neurons of the medial EC was analyzed. The main findings are as follows: 1) the amplitude of the Ca2⫹ transients associated with bAPs decreases with increasing distance from the soma, and this decrease is countered by the blockade of A-type K⫹ channels; 2) the amplitude of the Ca2⫹ transients is highly activity dependent, as it decreases progressively during repetitive firing; this decline is more prominent at distal locations and is removed by PKC activation. Taken together, these results show that bAP amplitude in medial EC layer V neurons is controlled by the activity of dendritic Na⫹ and K⫹ channels and that A-type K⫹ channels play a fundamental role in the distance-dependent decrease of bAP amplitude. Under control conditions, these channels reduce the extent of depolarization and Ca2⫹ influx in the distal portions of the apical dendritic tree compared with the more proximal portions during a back-propagating spike. EC layer V neurons have an essential function in the hippocampal-neocortical circuitry that is responsible for the replay and consolidation of memory (Lavenex and Amaral 2000). Whereas the hippocampal output is aimed mainly at the proximal portion of the apical dendrites (Sørensen and Shipley 1979), the distal dendrites of most layer V neurons extend to the superficial layers, where they branch in complex tufts that receive inputs from the presubiculum (Wouterlood et al. 2004) and could be targeted by fibers from different cortical regions (as hypothesized by Canto et al. 2008). The presence of excitatory synapses throughout the dendritic tree of EC layer V neurons is confirmed by two-photon images that show a relatively high density of spines at distal locations (Fig. 1). The Ca2⫹ signals associated with bAPs could underlie a form of communication between the two relatively independent prox-
Fig. 6. PKC activators remove the activity-dependent decrease in bAP amplitude mainly by modulation of dendritic Na⫹ channels. A: 2-photon image of a dendritic portion imaged at ⬃200 m from the soma, showing the location of the line scan. B: Ca2⫹ transients obtained from the line scan in A in control conditions (black trace), in the presence of Ba2⫹ 150 M (red trace), and in the presence of Ba2⫹ ⫹ PDA 10 M (green trace) in response to the train of APs shown in C. Inset: time course of the first Ca2⫹ transients for the control and Ba2⫹ traces scaled to match their peak amplitude (scale bars 50 ms and 20% ⌬F/F). The difference in the rate during the declining phase is evident. D: average data showing the effect of Ba2⫹ and PDA activators in the presence of Ba2⫹ on the amplitude of the first, third, and fifth peak in a 20-Hz train (n ⫽ 7). E: plot of the amplitude of ratio of the first, third, and fifth peak to the first peak shows the difference between the effect of Ba2⫹ (red symbols) and PDA in the presence of Ba2⫹ (green symbols) on the activity-dependent decrease of the amplitude of the Ca2⫹ transients in a 20-Hz train measured at 200 m from the soma. J Neurophysiol • VOL
105 • MARCH 2011 •
www.jn.org
1378
BACK-PROPAGATION IN ENTORHINAL LAYER V NEURONS
imal and distal dendritic regions of these cells. The dependence of bAP amplitude on the activity of A-type K⫹ channels, which are highly regulated by neurotransmitters and plasticity mechanisms (Hoffman and Johnston 1999; Frick et al. 2004), indicates that the extent of propagation could depend on the release of specific neuromodulators and therefore on the behavioral state of the animal. In particular, the coupling between distal and proximal regions would be more efficient when the activity of A-type K⫹ channels is decreased in the presence of neurotransmitters such as norepinephrine and acetylcholine (Hoffman and Johnston 1999). Spike back-propagation could also be extended through boosting by appropriately timed synaptic inputs, as shown in other pyramidal neurons (Magee and Johnston 1997; Stuart and Häusser 2001). This type of mechanism would be in effect mostly for the small bAPs at distal locations or for later spikes in a high-frequency train throughout the apical dendrite because EC layer V neurons show a significant activity-dependent decrease in the amplitude of the Ca2⫹ transients associated with bAPs already at proximal locations (⬎100 m, Fig. 4). A similar activity-dependent decrease has been shown in neocortical layer V and hippocampal CA1 pyramidal neurons (Callaway and Ross 1995; Spruston et al. 1995; Stuart et al. 1997), where it is due to a slow cumulative inactivation of dendritic Na⫹ channels (Colbert et al. 1997; Jung et al. 1997), but it is absent in hippocampal oriens-alveus interneurons and basket cells (Martina et al. 2000; Hu et al. 2010) and neocortical layer II/III neurons (Larkum et al. 2007). This observation confirms that different neurons in the central nervous system have developed distinct strategies, presumably to adjust their dendritic features to optimally implement their specific functions. This study shows that the activity dependence of dendritic calcium influx associated with each bAP in a train in medial EC layer V neurons depends upon both Na⫹ and K⫹ channels because the peak amplitude and the temporal summation of the Ca2⫹ transients were affected by Ba2⫹ (Fig. 6B) and because the decrease in the Ca2⫹ peaks could be partially removed by phosphorylation by PKC (Figs. 5 and 6), which has been shown to speed the recovery from inactivation of dendritic Na⫹ channels in CA1 neurons (Colbert and Johnston 1998). In particular, Ba2⫹ increased the peak of each Ca2⫹ transient in the train by a constant percentage of ⬃40%, whereas the PKC activator PDA produced a larger increase for the Ca2⫹ transients associated with each subsequent bAPs in a train. These results do not imply that the A-type K⫹ current does not contribute to the activity-dependent decrease in the amplitude of the peak of each Ca2⫹ transient; on the contrary, the density of A-type K⫹ channels can be very important in determining the degree of activity-dependence. Cumulative inactivation of dendritic Na⫹ channels attributable to prior activity reduces their availability for spike back-propagation during a train of action potentials; however, at low levels of A-type K⫹ current, a decrease in sodium channel availability attributable to prior activity might have little influence on spike amplitude. At the higher levels of A-type K⫹ channels expressed at more distal dendritic locations, the inactivation of Na⫹ channels becomes more significant because more sodium-channel availability is required to counter the larger A-type K⫹ current to allow back-propagation, producing a strong activity-dependent decrease in the amplitude of the Ca2⫹ transients. Although PKC activation could in principle decrease the activity of A-type K⫹ J Neurophysiol • VOL
channels as well (Hoffman and Johnston 1999), PKC activators were very effective in removing the activity-dependent decrease in the later Ca2⫹ transients in a 20-Hz train, even when they were applied in the presence of A-type K⫹ channels blockers, and therefore were probably acting mainly on dendritic Na⫹ channels. These data support the conclusion that PKC activation modulates the activity-dependent decrease in the amplitude of Ca2⫹ transients associated with bAPs in a train mainly by removing the slow cumulative inactivation of dendritic Na⫹ channels; we cannot exclude, however, the possibility that other dendritic K⫹ channels contribute to the decrease in the amplitude of the Ca2⫹ transients associated with later bAPs in a train, nor can we exclude the possibility that other targets are also modulated by PKC. As discussed above, the decrease in amplitude Ca2⫹ transients associated with later bAPs within a train, attributable to the slow cumulative inactivation of dendritic Na⫹ channels, is an attribute that EC layer V neurons share with neocortical layer V pyramidal neurons and CA1 pyramidal neurons. However, the distance-dependent increase in A-type K⫹ channel density that mediates the attenuation of bAP amplitude along the apical dendrite is an attribute shared with CA1 neurons but not neocortical pyramidal neurons, in which the density of these channels appears to be constant in the first 500 m from the soma (Bekkers 2000). This evidence somewhat counterintuitively suggests that the dendritic functionality of EC layer V neurons is more closely aligned with hippocampal compared with neocortical neurons. Further investigations are needed to assess further differences and similarities in the intrinsic dendritic excitability and integration properties among these cell types. In conclusion, this study shows that Na⫹ and transient K⫹ channels control spike back-propagation in the apical dendrites of EC layer V pyramidal neurons. Because the activity of Na⫹ and K⫹ channels can be regulated by membrane potential and second messenger systems, bAP amplitude and the associated Ca2⫹ influx can be highly modulated as well, with important functional implications for dendritic integration and associative synaptic plasticity. ACKNOWLEDGMENTS The author thanks the Dart Foundation for support through the Dart Neuroscience Scholars Program in Learning and Memory Award at MBL. The author is grateful to Prairie Technologies and Nikon for the loan of imaging equipment and to Carmen Canavier for helpful suggestions on the manuscript.
GRANTS This work was supported by the National Institutes of Health (NS035865 and NS069714).
DISCLOSURES No conflicts of interest, financial or otherwise, are declared by the author.
REFERENCES Amaral DG, Witter MP. The Hippocampal region. In: The Rat Nervous System, edited by Paxinos G. San Diego, CA: Elsevier Academic, 1995, p. 443– 493. Arnold SE. Cellular, and molecular neuropathology of the parahippocampal region in schizophrenia. Ann NY Acad Sci 911: 275–292, 2000.
105 • MARCH 2011 •
www.jn.org
BACK-PROPAGATION IN ENTORHINAL LAYER V NEURONS Bannister NJ, Larkman AU. Dendritic morphology of CA1 pyramidal neurons from the rat hippocampus: I. Branching patterns. J Comp Neurol 360: 150 –160, 1995. Bekkers JM. Distribution and activation of voltage-gated potassium channels incell-attached and outside-out patches from large layer 5 cortical pyramidal neurons of the rat. J Physiol 525: 611– 620, 2000. Buzsaki G. Two-stage model of memory trace formation: a role for “noisy” brain states. Neuroscience 31: 551–570, 1989. Callaway JC, Ross WN. Frequency-dependent propagation of sodium action potentials in dendrites of hippocampal CA1 pyramidal neurons. J Neurophysiol 74: 1395–1403, 1995. Canto CB, Wouterlood FG, Witter MP. What does the anatomical organization of the entorhinal cortex tell us? Neural Plast 2008: 38124, 2008. Chatelain FC, Alagem N, Xu Q, Pancaroglu R, Reuveny E, Minor DL Jr. The pore helix dipole has a minor role in inward rectifier channel function. Neuron 47: 833– 843, 2005. Colbert CM, Johnston D. Protein kinase C activation decreases activitydependent attenuation of dendritic Na⫹ current in hippocampal CA1 pyramidal neurons. J Neurophysiol 79: 491– 495, 1998. Colbert CM, Magee JC, Hoffman DA, Johnston D. Slow recovery from inactivation of Na⫹ channels underlies the activity-dependent attenuation of dendritic action potentials in hippocampal CA1 pyramidal neurons. J Neurosci 17: 6512– 6521, 1997. de Curtis M, Paré D. The rhinal cortices: a wall of inhibition between the neocortex, and the hippocampus. Prog Neurobiol 74: 101–110, 2004. Frick A, Magee J, Johnston D. LTP is accompanied by an enhanced local excitability of pyramidal neuron dendrites. Nat Neurosci 7: 126 –135, 2004. Fyhn M, Molden S, Witter MP, Moser EI, Moser MB. Spatial representation in the entorhinal cortex. Science 305: 1258 –1264, 2004. Gasparini S, Losonczy A, Chen X, Johnston D, Magee JC. Associative pairing enhances action potential back-propagation in radial oblique branches of CA1 pyramidal neurons. J Physiol 580: 787– 800, 2007. Gasparini S, Magee JC. State-dependent dendritic computation in hippocampal CA1 pyramidal neurons. J Neurosci 26: 2088 –2100, 2006. Hamam BN, Kennedy TE, Alonso A, Amaral DG. Morphological and electrophysiological characteristics of layer V neurons of the rat medial entorhinal cortex. J Comp Neurol 418: 457– 472, 2000. Hargreaves EL, Rao G, Lee I, Knierim JJ. Major dissociation between medial and lateral entorhinal input to dorsal hippocampus. Science 308: 1792–1794, 2005. Helmchen F, Imoto K, Sakmann B. Ca2⫹ buffering and action potentialevoked Ca2⫹ signaling in dendrites of pyramidal neurons. Biophys J 70: 1069 –1081, 1996. Hoffman DA, Magee JC, Colbert CM, Johnston D. K⫹ channel regulation of signal propagation in dendrites of hippocampal pyramidal neurons. Nature 387: 869 – 875, 1997. Hoffman DA, Johnston D. Neuromodulation of dendritic action potentials. J Neurophysiol 81: 408 – 411, 1999. Hu H, Martina M, Jonas P. Dendritic mechanisms underlying rapid synaptic activation of fast-spiking hippocampal interneurons. Science 327: 52–58, 2010. Jellinger K, Braak H, Braak E, Fischer P. Alzheimer lesions in the entorhinal region and isocortex in Parkinson’s and Alzheimer’s diseases. Ann NY Acad Sci 640: 203–209, 1991.
J Neurophysiol • VOL
1379
Jung HY, Mickus T, Spruston N. Prolonged sodium channel inactivation contributes to dendritic action potential attenuation in hippocampal pyramidal neurons. J Neurosci 17: 6639 – 6646, 1997. Larkman AU. Dendritic morphology of pyramidal neurons of the visual cortex of the rat: I. Branching patterns. J Comp Neurol 306: 307–319, 1991. Larkum ME, Waters J, Sakmann B, Helmchen F. Dendritic spikes in apical dendrites of neocortical layer 2/3 pyramidal neurons. J Neurosci 27: 8999 – 9008, 2007. Lasser-Ross N, Miyakawa H, Lev-Ram V, Young SR, Ross WN. High time resolution fluorescence imaging with a CCD camera. J Neurosci Meth 36: 253–261, 1991. Lavenex P, Amaral DG. Hippocampal-neocortical interaction: a hierarchy of associativity. Hippocampus 10: 420 – 430, 2000. Magee JC, Johnston D. A synaptically controlled, associative signal for Hebbian plasticity in hippocampal neurons. Science 275: 209 –213, 1997. Magee JC, Johnston D. Plasticity of dendritic function. Curr Opin Neurobiol 15: 334 –342, 2005. Martina M, Vida I, Jonas P. Distal initiation and active propagation of action potentials in interneuron dendrites. Science 287: 295–300, 2000. Migliore M, Shepherd GM. Opinion: an integrated approach to classifying neuronal phenotypes. Nat Rev Neurosci 6: 810 – 818, 2005. Rosenkranz JA, Johnston D. Dopaminergic regulation of neuronal excitability through modulation of Ih in layer V entorhinal cortex. J Neurosci 26: 3229 –3244, 2006. Rosenkranz JA, Johnston D. State-dependent modulation of amygdala inputs by dopamine-induced enhancement of sodium currents in layer V entorhinal cortex. J Neurosci 27: 7054 –7069, 2007. Sørensen KE, Shipley MT. Projections from the subiculum to the deep layers of the ipsilateral presubicular and entorhinal cortices in the guinea pig. J Comp Neurol 188: 313–333, 1979. Spruston N, McBain C. Structural and functional properties of hippocampal neurons. In: The Hippocampus Book, edited by Andersen P, Morris R, Amaral D, Bliss T, and O’Keefe J. Oxford, UK: Oxford University Press, 2007, p. 133–201. Spruston N, Schiller Y, Stuart G, Sakmann B. Activity-dependent action potential invasion and calcium influx into hippocampal CA1 dendrites. Science 268: 297–300, 1995. Spruston N, Stuart G, Häusser M. Dendritic integration. In: Dendrites, edited by Stuart G, Spruston N, Häusser M. New York: Oxford University Press, 2008, p. 225–250. Storm JF. Potassium currents in hippocampal pyramidal cells. Prog Brain Res 83: 161–187, 1990. Stuart GJ, Häusser M. Dendritic coincidence detection of EPSPs and action potentials. Nat Neurosci 4: 63–71, 2001. Stuart G, Schiller J, Sakmann B. Action potential initiation and propagation in rat neocortical pyramidal neurons. J Physiol 505: 617– 632, 1997. Witter MP, Amaral DG. The hippocampal region. In: The Rat Brain, edited by Paxinos G. San Diego, CA: Elsevier, 2004, p. 635–704. Wouterlood FG, Van Haeften T, Eijkhoudt M, Baks-Te-Bulte L, Goede PH, Witter MP. Input from the presubiculum to dendrites of layer-V neurons of the medial entorhinal cortex of the rat. Brain Res 1013: 1–12, 2004.
105 • MARCH 2011 •
www.jn.org
