www.elsevier.com / locate / bres
Research report
Responses of rat subicular neurons to convergent stimulation of lateral
entorhinal cortex and CA1 in vivo
a ,
*
b aJohn Gigg
, David M. Finch , Shane M. O’Mara
a
Department of Psychology, University of Dublin, Trinity College, Dublin 2, Ireland b
Brain Research Institute, University of California at Los Angeles, Los Angeles, CA 90095-1761, USA Accepted 15 August 2000
Abstract
There has been little electrophysiological examination of the afferent projection from lateral entorhinal cortex to dorsal subiculum. Here we provide evidence that synaptic inputs from lateral entorhinal cortex and CA1 converge onto single dorsal subicular neurons in vivo. Subicular responses to CA1 stimulation consisted of excitation and / or long-duration inhibition. Neurons excited by CA1 activation usually showed inhibition to entorhinal stimulation. The latter inhibition was usually of short duration, however, long duration inhibition was seen in a significant proportion of responses. Entorhinal stimulation produced excitatory responses in four bursting cells and it was these cells that also tended to show the longest inhibition. Only bursting cells could be driven antidromically by entorhinal stimulation. Biocytin-filled multipolar and pyramidal cells displayed excitation–inhibition sequences to CA1 and inhibition to entorhinal stimulation. These data strongly suggest that subicular inhibitory neurons receive excitatory input from CA1 and display mutual inhibition. The source of entorhinal-evoked inhibition is less clear. The relative sparseness of observed entorhinal-evoked responses suggests that the input to dorsal subiculum from any one part of lateral entorhinal cortex is spatially restricted. These data show that excitation–inhibition sequences can be seen in subicular pyramidal and multipolar cells and that single subicular neurons receive convergent inputs from CA1 and entorhinal cortex. We show for the first time that bursting cells can be driven both orthodromically and antidromically by direct entorhinal stimulation. These data support the existence of a reciprocal excitatory connection between lateral entorhinal cortex and dorsal subiculum and suggest further that this connection may involve only bursting subicular neurons. 2000 Elsevier Science B.V. All rights reserved.
Theme: Other systems of the CNS
Topic: Limbic system and hypothalamus
Keywords: Morphology; Immunohistochemistry; Biocytin; Pyramidal neuron; Multipolar neuron
1. Introduction [34,45,46,55]. It is the principal target of CA1 pyramidal cell axons and is the final relay in a polysynaptic loop The hippocampal formation of the mammalian brain is between the EC and hippocampus [3,13,32,45–47]. EC composed of the dentate gyrus, hippocampus proper (areas inputs to hippocampus may also bypass the dentate gyrus, CA1–CA4), the subicular complex and the entorhinal terminating directly in the subiculum, CA1 or CA3, cortex (EC) [45]. The EC and subicular complex have thereby shortening the EC-hippocampal loop by two to been described as retrohippocampal structures whose four synapses (for review see Amaral and Witter [4] and function is to process and transmit information between the Lopes da Silva et al. [28]).
neocortex and hippocampus. Of the two, the subiculum is In contrast to hippocampal areas, the electrophysiology the major output structure of the hippocampus proper of retrohippocampal structures has received very little attention to date, despite their acknowledged importance in memory formation [2,42,58] and as foci for *Corresponding author. Present address: Department of Psychology,
pathophysiological changes (e.g., [21,53]). Previous in University of Newcastle upon Tyne, Ridley Building, Newcastle upon
vivo recordings performed in the freely moving rat have Tyne Ne1 7RU, UK. Tel.:144-191-222-5790; fax:144-191-222-5622.
E-mail address: [email protected] (J. Gigg). established that subicular neuronal firing has a spatial 0006-8993 / 00 / $ – see front matter 2000 Elsevier Science B.V. All rights reserved.
36 J. Gigg et al. / Brain Research 884 (2000) 35 –50
correlate, that is, subicular neurons have ‘place fields’ tivity between subicular principal neurons [40,44] it can be [33,40]. Subicular neurons recorded in vivo can be sub- suggested that information processing within the divided into four classes based upon their firing charac- subiculum is at the level of single neurons rather than teristics [40]: (1) burster (whose spontaneous spiking neuronal ensembles. It is important to point out, however, activity largely consists of spike bursts); (2) non-burster that it remains to be seen whether there are long-range (cells that fire more tonically); (3) depolarized burster (see functional connections between cells in different parts of below); and (4) theta (presumed inhibitory interneurons). the subiculum. Nevertheless, it can be argued that, at least In vitro studies have added further weight to the division at the local level, afferent information may converge onto of subicular principal neurons into bursters and non-burs- single bursting and non-bursting cells where it is integrated ters (and, to some extent, depolarized bursters); these two and then passed on directly to extrasubicular sites. If there primary firing patterns are produced by distinct neuronal is indeed a difference in the innervation of bursting and classes, that is, non-bursting neurons cannot be made to non-bursting neurons from sites such as entorhinal cortex, burst [5,29,44,49]. These two classes can also be dis- hippocampus and thalamus, then these two cell classes will tinguished neurochemically; only non-bursting neurons exist in different neuronal circuits. These circuits may have express nicotinamide adenine dinucleotide phosphate- intrinsically different inputs and outputs. The present study diaphorase activity [19]. Bursting in subicular neurons was designed, therefore, to investigate the synaptic conver-appears to be a function of their membrane characteristics gence of inputs from CA1 and lateral EC onto single rather than simply the type of synaptic input that the cell dorsal subicular neurons. This has not been tested directly receives [29,43,44,49]. This suggests that subicular burst- before. In this study we also sought to correlate the ing cells may act to amplify the input that they receive, electrophysiological responses of subicular neurons with converting suprathreshold single pulse synaptic inputs into their cellular morphology by filling them with biocytin burst outputs. This burst output may be a particularly [38]. A preliminary report of these results has appeared effective means of transmitting neuronal information [26]. elsewhere [17].
The hippocampal output to subiculum from CA1 is excitatory [11,12,14,49] involving the postsynaptic
activa-tion of AMPA and, to perhaps a lesser degree, NMDA 2. Materials and methods receptors [30,49]. We have shown recently that this CA1
input to subiculum expresses long-term potentiation of 2.1. Animal preparation synaptic transmission in vivo [7–9]. Far less is known
regarding the EC input to subiculum. Jones [22] reported Experiments were performed on adult male Sprague– that in vitro stimulation of the medial EC produced Dawley rats. The anaesthetic, surgical and recording biphasic inhibition in rat subicular principal neurons. The procedures were carried out as described previously [14]. only major in vivo study to examine the nature of subicular Briefly, rats were anaesthetized (chloral hydrate, 400 mg / EC afferents is from the cat [50,51]. These authors found kg i.p.) and placed in a stereotaxic frame. Supplementary that connections between the EC and subiculum are injections of chloral hydrate were given as required (0.2– reciprocal. Anatomical evidence from the rat also supports 0.4 ml i.p.). Small craniotomies (about 2 mm diameter) the reciprocity of this connection [23,54]. were then performed to allow insertion of recording and
As described above, previous studies have examined the stimulating electrodes. individual inputs to the subiculum from CA1 and, to a
lesser degree, EC. There has, however, until now been no 2.2. Stimulation and recording in vivo analysis regarding whether single subicular neurons
electrode was replaced with a high-impedance pipette once results of cell filling were indistinguishable from those placement of the stimulating electrodes was complete. using more traditional intracellular injection methods using High-impedance electrodes (25–80 MV) contained either fine-tipped pipettes in the same laboratory (e.g., [13]). In (a) 1 M NaCl saturated with Fast Green, or (b) 0.5 M KCl any one animal, the locations for cells that were not containing 2% biocytin (Sigma). marked during the experiment were determined by com-Test shocks were applied during the search for cells in paring the microdrive depth readings for these cells to that order to maximize the number of cells encountered and recorded for either a Fast Green mark or a filled cell in the minimize the effects of any sampling bias. This meant that same animal. In some rats either a Fast Green mark was ‘silent’ neurons (i.e., those cells that showed no sponta- absent or there was a failure to recover filled cells. In these neous firing but did display an excitatory response) were rats it was possible that some of the recorded cells were in also included in the sample population. However, because the presubiculum rather than the subiculum. This is, these cells fired only a single spike in response to however, extremely unlikely as electrode tracks could be excitation and had no spontaneous discharge, they could seen penetrating only the subiculum in all cases where not be unambiguously categorized as either ‘bursters’ or indications for cell location were absent. The present ‘non-bursters.’ This is because most excitatory responses analyses are based on all cells lying in the subiculum. in classified bursting and non-bursting cells were
indis-tinguishable as they each consisted of only a single evoked 2.4. Statistical analyses spike.
Orthodromic excitatory responses were characterized by Results were analyzed using Students’ t, Chi square, evoked spikes showing significant latency jitter. Antid- Mann–Whitney, ANOVA and K-means clustering tests romic responses were characterized by constant-latency (using SPSS 9.0 software for PC). Differences were action potentials. Putative antidromic potentials were ex- considered significant if the value for P was less than 0.05. amined whenever possible for the presence of collision
(e.g., see Fig. 2). Peri-stimulus time histograms (PSTH)
were computed for all cells both on- and off-line (LabView 3. Results 2.0). Hand-counting spikes ensured accuracy of PSTH bins
in the vicinity of the stimulus artifact. 3.1. Electrode placements
2.3. Histology Recordings were made from 114 cells in 20 animals.
After histological verification, stimulating electrodes aimed At the end of the recording session stimulating loci were at CA1 were located in either dorsal alveus, dorsal CA1, marked by producing small lesions at the electrode tips (50 superficial CA1 adjacent to the dorsal alveus or dentate mA, 5 s of each polarity). During the recording session the gyrus (one rat only). Stimulating electrodes aimed at locations of particularly responsive cells were marked lateral EC were located in either lateral EC or ventral either with a Fast Green dye spot (5mA negative current angular bundle. Locations of subicular neurons (given by for 20 min) or by injecting the cell with biocytin following Fast Green mark or biocytin fill) that were activated by the juxtacellular method of Pinault [38]. Animals were stimulation of these sites are presented in Fig. 1.
then subjected to terminal anaesthesia, perfused
transcar-dially with 10% formalin and the brains removed. Prior to 3.2. Bursting versus non-bursting cells sectioning, brains were placed in a 30% sucrose solution.
38 J. Gigg et al. / Brain Research 884 (2000) 35 –50
Fig. 1. Location of cells recorded in dorsal subiculum and presubiculum. (A) Location of stimulating sites within CA1 (solid triangles: stimulus sites giving responses in subiculum; empty triangles: responses recorded in subiculum). (B) Location of subicular recording sites. Cell location (circle) was determined by either successful biocytin fill or Fast Green dye spot. Responses of cells are represented as either: (i) empty circle (inhibition to CA1 and EC); (ii) circle split vertically (response to CA1 on left, response to EC on right; black semi-circle excitation, clear semi-circle inhibition); or (iii) circle split horizontally (response to CA1 on top, response to EC on bottom; black semi-circle excitation, clear semi-circle ‘no excitation’ with equivocal inhibition). (C) Location of stimulation sites within the lateral EC (key to triangular symbols as per (A) above). The anterodorsal range for serial sections is given relative to Bregma. Sections correspond to the atlas of Paxinos and Watson (1986). CA1, hippocampal field; PreSub, presubiculum; Sub, subiculum; Ent, lateral entorhinal cortex; RF, rhinal fissure.
responses were obtained for both bursting and non-bursting sents 16.7% of the excitatory responses seen in bursting cells after stimulation of the alveus, indicating that both cells. The number of evoked spikes in these bursts was types of cell are projection neurons whose axons exit the either 2 (n53) or 3 (n51). The interspike interval of these hippocampal formation through the alveus. evoked bursts was 2.5660.58 ms (mean6S.E.M.). Note that this figure may not be based solely upon the responses 3.3. Response classification of pyramidal neurons as it includes a neuron with an evoked interspike interval of 1.4 ms; this value is similar to Responses were categorized as one of the following: (1) that found for a multipolar neuron (see below). Of the excitation (usually a single spike; see below) then long- non-bursting cells analyzed only one showed a response duration inhibition (usually c.300 ms); (2) antidromic that occasionally included two spikes to CA1 stimulation activation followed by long-duration inhibition; (3) inhibi- (7% of all excitatory responses in non-bursting cells). tion alone; (4) excitation or antidromic activation alone There were five cases in which cells responded to EC (these cells had a low or zero spontaneous firing rate so we stimulation with excitation. Four of these cases were had no indication of either the presence of inhibition or the bursting cells yet the evoked response in all four was cell type; see above); (5) no excitation (again where the almost always a single spike.
post-stimulus. Excitatory latency was calculated in those mean6S.E.M.). These values were significantly longer cases where the onset of the excitatory peak in the PSTH than those to CA1 or alvear stimulation (ANOVA, F5 clearly started to rise above the background firing rate. 30.38, df52,15, P,0.001; CA1 vs. EC, P,0.001; alveus Excitatory duration was calculated as the width of the vs. EC, P,0.001; CA1 vs. alveus n.s.). Mean values for excitatory peak in the PSTH. CA1 and alveus antidromic latencies were 2.8760.71 and 3.8060.34 (mean6S.E.M.) respectively. An example of 3.4. Subicular responses to stimulation of CA1, EC and antidromic activation following alvear stimulation is
alveus shown in Fig. 2.
As expected, excitatory and inhibitory latencies to CA1 Following CA1 stimulation 29 cells (39%) showed stimulation were significantly shorter when compared to excitation, 33 cells (33%) showed only inhibition, 11 cells those from EC stimulation (Fig. 3). This probably reflects (15%) were activated antidromically and 10 cells (13%) the longer conduction time for the EC-subiculum pathway. showed no obvious response. One cell showed an equivoc- Excitatory duration was significantly longer to EC stimula-al response and is not included in the anstimula-alysis. Excitatory tion compared to CA1 stimulation and inhibitory duration responses to CA1 were equally likely from bursting and was significantly shorter to EC stimulation (Fig. 3). On non-bursting cells. Responses to EC were markedly differ- first inspection these results may be related in that the ent from those evoked by CA1 stimulation. Following EC weaker inhibitory input following EC stimulation may stimulation five cells (5%) showed excitation, 41 cells have allowed excitation to last longer. However, values for (41%) showed only inhibition, three cells (3%) antidromic EC-evoked inhibitory duration appeared to fall into three activation and 51 cells (51%) showed no obvious response. distinct groupings with the respective values for CA1-Of the cells showing excitation to EC stimulation four evoked inhibition appearing far more homogeneous (Fig. were bursting neurons and one cell could not be classified 4). Subjecting these EC data to a K-means cluster analysis as it had a zero spontaneous rate. The mean excitatory revealed that they could be separated into three groups. latency to EC stimulation was 12.3861.54 ms Cluster centres for these were at 69.78 ms, 246.40 ms and (mean6S.E.M.). All three cells driven antidromically by 472.80 ms (Fig. 4). The number of cases for these clusters EC stimulation were bursting cells. These cells had a mean were 37, 5 and 5 respectively. These clusters were burst interspike interval of 4.1560.14 ms (mean6S.E.M.). significantly different (ANOVA, df52, 44; F5469.39; P, Antidromic response latencies to EC stimulation were 0.001). From this it appears that although most cells 11.40, 11.70 and 13.00 ms (12.0360.49 ms; responded with weak inhibition after EC stimulation, a
40 J. Gigg et al. / Brain Research 884 (2000) 35 –50
Fig. 3. Latency and duration of excitation and inhibition. (A) Excitatory latency is significantly longer to EC than CA1 stimulation. (B) Excitatory duration to EC stimulation is significantly longer than that to CA1 stimulation. (C) EC stimulation produces inhibition that has a longer latency than that to CA1 stimulation. (D) Inhibition following CA1 stimulation is significantly longer than inhibition produced by EC stimulation. Significance levels for all parts of figure: **P,0.01, *P,0.05. Error bars represent standard error of the mean.
significant number of cells were subject to strong inhibi- 3.5. Morphology of filled cells tion. Interestingly, three of the bursting cells that showed
excitation to EC stimulation also expressed clear inhibitory Seven subicular neurons were filled with biocytin during responses (the inhibitory duration in the other two driven recording. These could be classified by shape as either cells could not be determined) and the inhibitory duration pyramidal-like (n54; all bursting neurons) or multipolar in these cases was 120 ms, 437 ms and 510 ms. This (n53; putative inhibitory neurons; 1 bursting cell and 2 suggests that cells excited by EC stimulation also show non-bursting cells). Fig. 5 shows the responses for two strong inhibition. The longer duration seen in excitatory identified neurons from the same rat. The morphological responses to EC stimulation may simply reflect a more features of these cells are shown in Fig. 6. The pyramidal-variable conduction time in this longer pathway compared like neurons visualized in the present study all had the to the CA1 input. It seems unlikely that the level of evoked general form of hippocampal pyramidal neurons with one inhibition was lower in the former cases as responsive cells long apical dendrite that coursed through the stratum were all bursting cells yet were prevented (presumably pyramidale towards the subicular molecular layer. This inhibited) from showing evoked burst responses. This is morphology is in agreement with previous reports for similar to the above cases where CA1 stimulation evoked subicular pyramidal neurons [20,29,31,49]. The morpholo-single spike excitation in cells that fired spontaneous spike gy of multipolar neurons was less clear. In general, the cell
bursts. bodies for all three multipolar neurons were clearly located
42 J. Gigg et al. / Brain Research 884 (2000) 35 –50
Fig. 6. Morphology of subicular pyramidal and multipolar neuron from the same rat. Evoked responses and location of these cells are given in Fig. 5. (A1) Drawing tube reconstruction of subicular pyramidal-like neuron. Spines were evident in the dendritic processes of this cell. (A2) Photomicrograph showing enlarged view of cell body region for cell in (A1). (B1) Drawing tube reconstruction of multipolar cell. The majority of processes from this cell contained numerous varicosities. (B2) Photomicrograph showing enlarged view of cell body region for multipolar neuron in (B1). Note the presence of varicosities. Orientation for all parts of this figure: ventral is up, dorsal down, medial right and lateral left. Scale bar: (A1) 35mm; (B1) 30mm; (A2) and (B2) 10mm.
These processes had no obvious orientation to any plane stimulation sites for identified cells was that for inhibitory within stratum pyramidale. This cell most closely resem- duration; pyramidal-like neurons show significantly longer bles the fast spiking class of interneuron reported by inhibition to CA1 stimulation compared to EC (P,0.05, Greene and Totterdell [20]. Mann–Whitney U ). There was a similar trend for
multi-polar neurons.
3.6. Response characteristics of filled cells Examples of the responses of pyramidal-like neurons are presented in Figs. 5 and 7. In general, the response of In general, the response characteristics for visualized pyramidal-like bursting neurons to CA1 stimulation was cells were comparable to those for the overall subiculum either inhibition or excitation (single spike) followed by data. The only statistically significant difference between inhibition. The pyramidal-like bursting neuron in Fig. 5
44 J. Gigg et al. / Brain Research 884 (2000) 35 –50 showed a single evoked spike response to CA1
stimula-tion, followed by inhibition. There was an equivocal response to EC stimulation for this neuron with perhaps a short duration inhibitory phase. Fig. 7 shows the responses of a pyramidal-like neuron to CA1 and angular bundle stimulation. This neuron showed a single-spike response to CA1 stimulation followed by inhibition. The response to angular bundle stimulation showed a slight indication of excitation followed by an inhibitory phase that was more marked when compared to the cells in Fig. 5 with EC stimulation. The data sweeps in Fig. 7c show an example of an evoked spike following CA1 stimulation and a spontaneous burst of spikes.
Following EC stimulation the three recovered multipolar cells showed either inhibition (n52) or no response (n51).
Fig. 8. Convergence data for subicular neurons tested with CA1 and EC In contrast, all three of these multipolar neurons could be
stimulation. The different types of response to CA1 stimulation are driven by CA1 stimulation. One of these neurons fired
recorded along the abscissa. The left-hand bar represents those cells that spontaneous spike bursts and is shown in Fig. 5. This cell responded with excitation followed by inhibition to CA1 stimulation. also fired an evoked burst of 2–3 spikes in response to Stacked histogram bars represent the number of EC responses of each type (given in key) to each of the respective CA1 response types. CA1 stimulation. These spikes had a 1.6 ms inter-spike
Stimulation of EC produced either ‘no response’ or ‘inhibition’ in the interval and showed a decrease in amplitude with time
majority of cells tested regardless of convergent response to CA1. during the burst. This multipolar neuron displayed a
pronounced inhibitory phase after the evoked burst to CA1
stimulation. There was also an inhibitory period following convergence onto a single subicular neuron. Responses EC stimulation in this neuron, although the duration of from this neuron are shown in Fig. 9; note that this neuron inhibition in this case was markedly shorter. The second showed one of the few excitatory burst responses to CA1 multipolar cell driven by CA1 stimulation showed an stimulation. Note also that it was necessary to apply a evoked burst of constant amplitude spikes but fired tonic double-pulse stimulus through the EC electrode in order to spikes between stimuli. The third multipolar neuron fired a evoke activity in this subicular neuron. In the second of single spike to CA1 stimulation and spontaneous high these three cases it was possible to overcome EC excitation frequency tonic spikes (spontaneous rate for this cell was with (presumed) convergent inhibition produced through 62 spikes per second, the highest rate for all three simultaneous activation of the CA1 electrode (Fig. 10). In
multipolar neurons). the last of the three cases of EC-evoked excitation the
Overall, the results from morphologically characterized neuron displayed no excitation to CA1.
neurons demonstrate that CA1 afferents produce excita- Two other neurons that showed EC-evoked responses tion–inhibition sequences in pyramidal and multipolar were also tested with dentate gyrus stimulation. In the first cells. The responses of these cells to EC stimulation were of these two cases there was long-latency convergent far less impressive, showing only relatively weak inhibi- excitation; excitatory latencies were 15.9 ms for dentate tion of cell firing. It appears from the present data that stimulation and 16.8 ms for EC stimulation. In the second different multipolar neurons show different excitatory of these cases there was an unexpected antidromic re-responses to CA1 stimulation. These results also suggest sponse to dentate stimulation; antidromic latency to dentate that multipolar neurons receive inhibitory inputs (mutual stimulation was 12 ms, excitatory latency to EC stimula-inhibition). It is noteworthy that in the present sample none tion was 15 ms.
of the multipolar neurons could be driven by EC stimula-tion.
3.7. Convergent responses 4. Discussion
Fig. 9. Convergence of excitation from EC and CA1 onto single subicular complex neuron. (A) Response to CA1 stimulation. Consecutive responses of subicular complex cell (top to bottom). Stimuli evoked responses in 92% of presentations; 8% of these were evoked spike doublets (see top trace). PSTH at bottom of (A) shows excitatory responses to CA1 stimulation for 45 successive sweeps. (B) Response to double-pulse EC stimulation. There was a low probability of response to EC stimulation. Excitation was present in 17% of responses; 1.7% of these showed evoked spike doublets. PSTH at bottom of (B) shows excitatory responses to EC stimulation over 45 sweeps. Calibration: data sweep 2 mV, 2 ms; PSTHs 100 spikes / s, 10 ms (A4); 25 spikes / s, 10 ms (B3). Solid triangles show stimulus onset.
stimulation. The relative proportions of subicular bursting 4.1. Responses to CA1 and alveus stimulation and non-bursting neurons are in general agreement with
46 J. Gigg et al. / Brain Research 884 (2000) 35 –50
Fig. 10. Inhibition evoked by CA1 stimulation can overcome excitation produced by EC stimulation in subicular neuron. (Top Left) Responses of subicular neuron to CA1 stimulation at 500mA. Lower panel shows location of CA1 electrode. (Top Right) Excitatory response to single pulse EC stimulation at 500 mA. Lower panel shows location of EC electrode. (Central Panel) Excitation to EC stimulation is still present during simultaneous stimulation of CA1 at 100mA intensity (overlapping solid triangles show stimulus onset for both stimulating sites). (Bottom Panel) Inhibition from CA1 stimulation is powerful enough to overcome EC-evoked excitation when CA1 stimulus intensity is raised to 200mA. Recording site shown in bottom right panel. Calibration: 1 mV, 10 ms.
the present anaesthetized in vivo preparation is evidenced (2) The latency data are supported by the reproducible by a low proportion of evoked burst responses to CA1 finding from a number of laboratories that EC afferents are stimulation in bursting cells. incapable of producing action potential discharge in CA1 pyramidal cells in the anaesthetized rat (see below). This 4.2. Responses to EC stimulation strongly supports the suggestion that the polysynaptic loop between EC and subiculum via CA1 would be non-func-The predominant response of subicular neurons to EC tional in the anaesthetized animal, again supporting the stimulation was that of weak inhibition, regardless of the monosynaptic excitation of subicular neurons by EC response to CA1 input in the same cell. A previous report activation.
has described inhibitory potentials in subicular cells in The duration of excitatory responses to EC stimulation vitro following EC stimulation [22] so it is likely that the was significantly longer than those to CA1 stimulation. inhibition produced with EC stimulation has a direct This is perhaps due in part to the overall weaker inhibition synaptic origin (see below). Compared to that evoked by noted earlier in response to EC stimulation. However, cells CA1 stimulation, inhibitory duration following EC activa- driven by EC stimulation tend to show the longest duration tion was significantly shorter. However, a small but inhibition. This suggests that the initial phase of inhibition significant number of neurons expressed long duration (GABAa-dependent?) produced by EC stimulation in inhibition to EC stimulation. Stimulation of EC produced subicular neurons could be weaker than that evoked very few examples of excitation in subicular neurons. Five following CA1 stimulation. The case in Fig. 10 shows a neurons could, however, be driven by EC stimulation. Of cell that could be driven by EC stimulation and that these, four were bursting neurons and the fifth could not be showed inhibition to CA1. That the latter actually repre-classified. This is a relatively small number of excitatory sented active inhibition is supported by the fact that the EC responses and it is possible that a more optimal placing of excitation could be overcome by the presumed strong, recording and stimulating electrodes in further studies may short latency inhibition produced during simultaneous CA1 produce a stronger activation of subicular cells. However, stimulation. The weak inhibitory influence from EC in as no previous studies have examined the EC-subiculum most subicular recordings resembles the responses seen in pathway in vivo, it is impossible to predict the outcome of rat hippocampus after activation of EC afferents such alternate strategies. Neurons that showed excitation to [22,24,25,41] (but see Yeckel and Berger [56,57] for CA1 EC stimulation also tended to show the longest duration excitation in the rabbit following EC activation). Although inhibition. Antidromic activation after EC stimulation was the EC input to CA1 appears weak in the rat, it does only seen in three subicular neurons and these were all express long-term potentiation [24,25] and paired-pulse bursting cells. This provides physiological evidence in facilitation [24]. However, despite the prior induction of support of the hypothesis that lateral EC and dorsal long-term potentiation in this pathway, activation of the subiculum have a functional reciprocal connection. Fur- perforant path is still incapable of driving CA1 single-unit thermore, these results suggest that this reciprocal con- activity [24]. In this regard, the EC input to subiculum may nection may only include subicular bursting neurons. be ‘stronger’ than that to CA1 as EC afferents can drive Although we cannot prove that these subicular excitatory subicular neurons.
48 J. Gigg et al. / Brain Research 884 (2000) 35 –50
excitation produced in all subicular neurons by EC stimu- principal cells. It is important to note, however, that
lation. inhibition in subicular cells produced through EC
stimula-tion may not be a product of inhibitory neurons within the 4.3. Responses of identified cells subiculum itself. One candidate extra-subicular source of inhibition is the dentate gyrus [6]. Inhibitory neurons are Pyramidal-like and multipolar cells were filled with present throughout the entire dentate outer molecular layer. biocytin and visualized histochemically. Pyramidal neu- These neurons project across the hippocampal fissure to rons all fired spontaneous bursts of spikes. The two the subiculum [6]. Stimulation of EC strongly activates the pyramidal cells that responded to CA1 stimulation dis- dentate gyrus, so it is likely to activate this particular played single evoked spikes. It is assumed that the strength inhibitory projection neuron which, in turn, could produce of inhibition following CA1 stimulation curtailed the full feed-forward inhibition of subicular pyramidal neurons burst response in these pyramidal cells. The inhibitory (and possibly subicular inhibitory neurons as well). A duration in pyramidal cells to CA1 stimulation was second possible route for EC-evoked inhibition is a direct significantly longer than that to EC stimulation, again inhibitory input from EC to subiculum; although we know highlighting the predominantly weak inhibitory input from of no direct evidence in support of this claim, the EC does EC activation in the present study. contain GABAergic cells that project to the hippocampus The most fully characterized multipolar cell is presented [15,16]. Finally, a third possible mechanism for inhibition in Fig. 5.B1. This cell resembles most closely the fast following EC stimulation is antidromic activation of spiking type of interneuron described by Greene and inhibitory neurons. Subicular inhibitory neurons have been Totterdell [20], with its location in the subicular cell body shown to project to the EC [52]. Stimulation within EC layer and processes containing numerous varicosities. This may have activated these cells antidromically, thereby cell fired a burst of spikes to CA1 stimulation that activating their presumed recurrent collaterals within the decreased in amplitude. The other two subicular multipolar subiculum.
neurons described here fired (a) CA1-evoked spike bursts
in one case and (b) a single CA1-evoked spike with high 4.5. Possible significance of EC connections to frequency spontaneous tonic firing in the other. These subiculum
results suggest that there may be more than one class of
subicular inhibitory neuron, particularly in stratum The results presented here provide novel information pyramidale, and that these classes may be distinguished by regarding the convergence of EC and CA1 afferents onto response characteristic. The CA1-evoked burst of decreas- single neurons of the dorsal subiculum. CA1 afferents ing amplitude spikes seen in the present fast spiking cell strongly excite pyramidal cells; the influence of EC resembles very closely the response of four putative afferents is far less impressive, resulting in predominantly inhibitory subicular cells recorded by Finch and Babb [12]. weak inhibition. Present recordings from identified multi-Thus, inhibitory cells with this particular response charac- polar neurons show that these presumed inhibitory neurons teristic may represent a defined subicular inhibitory cell are excited by CA1 (but not EC) afferents. We suggest that
subtype. these putative inhibitory cells both provide and receive
neu-single and multiple episodes of theta patterned high frequency rons. It is important to note, however, that this suggested
stimulation on synaptic transmission from hippocampal area CA1 to role for bursting cells is based on only a small number of
the subiculum in rats, Neurosci. Lett. 270 (1999) 99–102. cells that showed excitation to EC stimulation. [8] S. Commins, J. Gigg, M. Anderson, S.M. O’Mara, Interaction
In contrast to EC stimulation, activation of CA1 acti- between paired-pulse facilitation and long-term potentiation in the vates a large number of subicular neurons. This is con- projection from hippocampal area CA1 to the subiculum,
Neuro-report 9 (1998) 4109–4113. sistent with anatomical data showing that the CA1 output
[9] S. Commins, J. Gigg, M. Anderson, S.M. O’Mara, The projection ‘fans out’ over a large dorsoventral extent in the subiculum
from hippocampal area CA1 to the subiculum sustains long-term [3]. Dorsal subicular neurons appear, therefore, to receive a potentiation, Neuroreport 9 (1998) 847–850.
substantial overlapping input from a large region of CA1 [10] A. Cowey, Projection of the retina on to striate and prestriate cortex but a restricted convergent input from perhaps only a small in the squirrel monkey Samir sciureus, J. Neurophysiol. 27 (1964)
366–393. number of EC neurons. In turn, as described above, the
[11] D.M. Finch, T.L. Babb, Inhibition in subicular and entorhinal reciprocal output from dorsal subiculum may reach only a
principal neurons in response to electrical stimulation of the fornix limited number of cells in lateral EC. In functional terms, and hippocampus, Brain Res. 196 (1980) 89–98.
the reciprocal connection between subiculum and EC may [12] D.M. Finch, T.L. Babb, Neurophysiology of the caudally directed represent the neuronal substrate for the hypothesized hippocampal efferent system in the rat: projections to the subicular
complex, Brain Res. 197 (1980) 11–26. universal spatial map described by Sharp [39]. In this
[13] D.M. Finch, T.L. Babb, Demonstration of caudally directed hip-scheme reciprocal connections between subiculum and EC
pocampal efferents in the rat by intracellular injection of horseradish establish a single spatial map for all environments. The
peroxidase, Brain Res. 214 (1981) 405–410.
‘point-to-point’ nature of the EC-subiculum pathway de- [14] D.M. Finch, A.M. Tan, M. Isokawa-Akesson, Feedforward inhibi-scribed here is consistent with this mapping hypothesis as tion of the rat entorhinal cortex and subicular complex, J. Neurosci.
8 (1988) 2213–2226. this type of connection is thought to preserve similar
[15] P. Germroth, W.K. Schwerdtfeger, E.H. Buhl, GABAergic neurons representation of information between two connected areas
in the entorhinal cortex project to the hippocampus, Brain Res. 494 (e.g., [10]). The present results further suggest that the
(1989) 187–192.
production of this putative EC-subiculum universal map [16] P. Germroth, W.K. Schwerdtfeger, E.H. Buhl, Ultrastructure and involves only bursting cells within dorsal subiculum and aspects of functional organization of pyramidal and nonpyramidal that only these bursting cells relay processed information entorhinal projection neurons contributing to the perforant path, J.
Comp. Neurol. 305 (1991) 215–231. back to lateral EC.
[17] J. Gigg, D.M. Finch, S.M. O’Mara, Convergent responses of subicular complex cells to stimulation of CA1 and entorhinal cortex in vivo, Soc. Neurosci. Abstr. 23 (1997) 189.
Acknowledgements [18] J. Gigg, A.M. Tan, D.M. Finch, Glutamatergic excitatory responses of anterior cingulate neurons to stimulation of the mediodorsal thalamus and their regulation by GABA: an in vivo iontophoretic JG was supported by European Commission TMR
study, Cereb. Cortex 2 (1992) 477–484. ERBFMBICT961286 and Forbairt International
Collabora-[19] J.R. Greene, H. Lin, A.J. Mason, L.R. Johnson, S. Totterdell, tion Grant; DMF was supported by NIH grant DA09543; Differential expression of NADPH-diaphorase between electrophy-SMOM was supported in part by Wellcome Trust and TCD siologically-defined classes of pyramidal neurons in rat ventral
Provost’s fund. subiculum, in vitro, Neuroscience 80 (1997) 95–104.
[20] J.R. Greene, S. Totterdell, Morphology and distribution of electro-physiologically defined classes of pyramidal and nonpyramidal neurons in rat ventral subiculum in vitro, J. Comp. Neurol. 380
References (1997) 395–408.
[21] B.T. Hyman, G.W. Van Hoesen, A.R. Damasio, C.L. Barnes, [1] B.E. Alger, R.A. Nicoll, GABA-mediated biphasic inhibitory re- Alzheimer’s disease: cell-specific pathology isolates the
hippocam-sponses in hippocampus, Nature 281 (1979) 315–317. pal formation, Science 225 (1984) 1168–1170.
[2] D.G. Amaral, Memory: anatomical organization of candidate brain [22] R.S.G. Jones, Frequency-dependent alterations in synaptic transmis-regions, in: Handbook of Physiology, the Nervous System V, sion in entorhinal–hippocampal pathways, Hippocampus 5 (1995) American Physiological Society, Bethesda, MD, 1987, pp. 211–294. 125–128.
[3] D.G. Amaral, C. Dolorfo, P. Alvarez-Royo, Organization of CA1 [23] C. Kohler, Intrinsic projections of the retrohippocampal region in projections to the subiculum: A PHA-L analysis in the rat, Hip- the rat brain I. The subicular complex, J. Comp. Neurol. 236 (1985)
pocampus 1 (4) (1991) 415–436. 504–522.
[4] D.G. Amaral, M.P. Witter, The three-dimensional organization of the [24] L.S. Leung, L. Roth, K.J. Canning, Entorhinal inputs to hippocam-hippocampal formation: a review of anatomical data, Neuroscience pal CA1 and dentate gyrus in the rat: A current-source-density 31 (3) (1989) 571–591. study, J. Neurophysiol. 73 (1995) 2392–2403.
[5] J. Behr, R.M. Empson, D. Schmitz, T. Gloveli, U. Heinemann, [25] W.B. Levy, C.M. Colbert, N.L. Desmond, Another network model Electrophysiological properties of rat subicular neurons in vitro, bites the dust: Entorhinal inputs are no more than weakly excitatory Neurosci. Lett. 220 (1996) 41–44. in the hippocampal CA1 region, Hippocampus 5 (1995) 137–140. [6] K. Ceranik, R. Bender, J.R.P. Geiger, H. Monyer, P. Jonas, M. [26] J.E. Lisman, Bursts as a unit of neural information: making
¨
50 J. Gigg et al. / Brain Research 884 (2000) 35 –50
[28] F.H. Lopes da Silva, M.P. Witter, P.H. Boeijinga, A.H.M. Lohman, of guinea pig subicular neurons in vitro, J. Neurophysiol. 70 (1) Anatomic organization and physiology of the limbic cortex, Physiol. (1993) 232–245.
Rev. 70 (2) (1990) 453–511. [45] L.W. Swanson, W.M. Cowan, An autoradiographic study of the [29] A. Mason, Electrophysiology and burst-firing of rat subicular organization of the efferent connections of the hippocampal
forma-pyramidal neurons in vitro: a comparison with area CA1, Brain Res. tion in the rat, J. Comp. Neurol. 172 (1977) 49–84.
600 (1993) 174–178. [46] L.W. Swanson, J.M. Wyss, W.M. Cowan, An autoradiographic study [30] D. Mattia, G.G.C. Hwa, M. Avoli, Membrane properties of rat of the organization of intrahippocampal association pathways in the
subicular neurons in vitro, J. Neurophysiol. 70 (1993) 1244–1248. rat, J. Comp. Neurol. 181 (1978) 681–716.
[31] D. Mattia, H. Kawasaki, M. Avoli, Repetitive firing and oscillatory [47] N. Tamamaki, Y. Nojyo, Disposition of slab-like modules formed by activity of pyramidal-like neurons in the rat subiculum, Exp. Brain axon branches originating from single CA1 pyramidal cell neurons Res. 114 (1997) 507–517. in the rat hippocampus, J. Comp. Neurol. 291 (1990) 509–519. [32] R.C. Meibach, A. Seigel, Efferent connections of the hippocampal [48] N. Tamamaki, Y. Nojyo, Preservation of topography in the
con-formation in the rat, Brain Res. 124 (1977) 197–224. nections between the subiculum, field CA1, and the entorhinal [33] R.U. Muller, J.L. Kubie, E. Bostock, J.S. Taube, G.J. Quirk, Spatial cortex in rats, J. Comp. Neurol. 353 (1995) 379–390.
firing correlates of neurons in the hippocampal formation of freely [49] J.S. Taube, Electrophysiological properties of neurons in the rat moving rats, in: J. Pollard (Ed.), Brain and Space, O.U.P., 1991. subiculum in vitro, Exp. Brain Res. 96 (1993) 304–318.
[34] P.A. Naber, M.P. Witter, Subicular efferents are organized mostly as [50] T. van Groen, F.H. Lopes da Silva, Organization of the reciprocal parallel projections: a double-labeling, retrograde-tracing study in connections between the subiculum and the entorhinal cortex in the the rat, J. Comp. Neurol. 393 (1998) 284–297. cat: I. A neuroanatomical tracing study, J. Comp. Neurol. 250 [35] N.R. Newberry, R.A. Nicoll, A bicuculline-resistant inhibitory post- (1986) 485–497.
synaptic potential in rat hippocampal inhibitory cell in vitro, J. [51] T. van Groen, F.H. Lopes da Silva, Organization of the reciprocal Physiol. (Lond.) 348 (1984) 239–254. connections between the subiculum and the entorhinal cortex in the [36] N.R. Newberry, R.A. Nicoll, Comparison of the action of baclofen cat: II. An electrophysiological study, J. Comp. Neurol. 251 (1986)
with gamma-aminobutyric acid in rat hippocampal cell in vitro, J. 111–120.
Physiol. (Lond.) 360 (1985) 161–185. [52] T. van Haeften, B. Jorritsma-Byham, M.P. Witter, Quantitative [37] G. Paxinos, C. Watson, The Rat Brain in Stereotaxic Coordinates, morphological analysis of subicular terminals in the rat entorhinal
Academic Press, New York, 1982. cortex, Hippocampus 5 (1995) 452–459.
[38] D. Pinault, A novel single-cell staining procedure performed in vivo [53] G.W. Van Hoesen, B.T. Hyman, A.R. Damasio, Entorhinal cortex under electrophysiological control: morpho-functional features of pathology in Alzheimer’s disease, Hippocampus 1 (1991) 1–8. juxtacellularly labeled thalamic cells and other central neurons with [54] M.P. Witter, Organization of the entorhinal–hippocampal system: a biocytin or neurobiotin, J. Neurosci. Methods 65 (1996) 113–136. review of current anatomical data, Hippocampus 3 (1993) 33–44. [39] P.E. Sharp, Complimentary roles for hippocampal versus subicular / [55] M.P. Witter, H.J. Groenewegen, The subiculum:
cytoarchitectonical-entorhinal place cells in coding place, context and events, Hip- ly a simple structure, but hodologically complex, Prog. Brain Res.
pocampus 9 (1999) 432–443. 83 (1990) 47–58.
[40] P.E. Sharp, C. Green, Spatial correlates of firing patterns of single [56] M.F. Yeckel, T.W. Berger, Feedforward excitation of the hippocam-cells in the subiculum of the freely moving rat, J. Neurosci. 14 (4) pus by afferents from the entorhinal cortex: redefinition of the role (1994) 2339–2356. of the trisynaptic pathway, Proc. Natl. Acad. Sci. USA 87 (1990)
´
[41] I. Soltesz, Brief history of cortico-hippocampal time with a special 5832–5836.
reference to the direct entorhinal input to CA1, Hippocampus 5 [57] M.F. Yeckel, T.W. Berger, Monosynaptic excitation of hippocampal (1995) 120–124. CA1 pyramidal cells by afferents from the entorhinal cortex, [42] L.R. Squire, Memory and the hippocampus: A synthesis from Hippocampus 5 (1995) 108–114.
findings with rats, monkeys, and humans, Psychol. Rev. 99 (1992) [58] S. Zola-Morgan, L.R. Squire, D.G. Amaral, W.A. Suzuki, Lesions of 195–231. perirhinal and parahippocampal cortex that spare the amygdala and [43] M. Stewart, Antidromic and orthodromic responses by subicular hippocampal formation produce severe memory impairment, J.
