| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
The Journal of Neurophysiology Vol. 79 No. 3 March 1998, pp. 1518-1534
Copyright ©1998 by the American Physiological Society
1 Anatomisches Institut, Albert-Ludwigs-Universität Freiburg, D-79104 Freiburg; 2 Max-Planck-Institut für Medizinische Forschung, Abteilung Zellphysiologie, D-69120 Heidelberg, Germany; and 3 Department of Neurobiology and Physiology, Institute for Neuroscience, Northwestern University, Evanston, Illinois 60208-3520
 |
ABSTRACT |
|---|
 Abstract
 Introduction
Methods
Results
Discussion
References
|
|---|
Lübke, Joachim, Michael Frotscher, and Nelson Spruston. Specialized electrophysiological properties of anatomically identified neurons in the hilar region of the rat fascia dentata. J. Neurophysiol. 79: 1518-1534, 1998. Because of their strategic position between the granule cell and pyramidal cell layers, neurons of the hilar region of the hippocampal formation are likely to play an important role in the information processing between the entorhinal cortex and the hippocampus proper. Here we present an electrophysiological characterization of anatomically identified neurons in the fascia dentata as studied using patch-pipette recordings and subsequent biocytin-staining of neurons in slices. The resting potential, input resistance (RN), membrane time constant (
m), "sag" in hyperpolarizing responses, maximum firing rate during a 1-s current pulse, spike width, and fast and slow afterhyperpolarizations (AHPs) were determined for several different types of hilar neurons. Basket cells had a dense axonal plexus almost exclusively within the granule cell layer and were distinguishable by their low RN, short m, lack of sag, and rapid firing rates. Dentate granule cells also lacked sag and were identifiable by their higher RN, longer m, and lower firing rates than basket cells. Mossy cells had extensive axon collaterals within the hilus and a few long-range collaterals to the inner molecular layer and CA3c and were characterized physiologically by small fast and slow AHPs. Spiny and aspiny hilar interneurons projected primarily either to the inner or outer segment of the molecular layer and had a dense intrahilar axonal plexus, terminating onto somata within the hilus and CA3c. Physiologically, spiny hilar interneurons generally had higher RN values than mossy cells and a smaller slow AHP than aspiny interneurons. The specialized physiological properties of different classes of hilar neurons are likely to be important determinants of their functional operation within the hippocampal circuitry.
The hilar region of the hippocampus is likely to play an important role in processing information as it travels through the hippocampal formation and therefore may be important for the formation of memories that occurs in the hippocampus (see Buckmaster and Schwartzkroin 1994 The methods employed for this study are identical to those described in a related paper on stratum lucidum neurons (Spruston et al. 1997 Patch-pipette recordings and biocytin filling
Transverse hippocampal slices (300-400 µm) were prepared from 13- to 35-day-old Wistar rats. Slices were visualized using infrared differential interference contrast (IR-DIC) videomicroscopy (Stuart et al. 1993
Data acquisition and analysis
Voltage recordings were filtered at 3-10 kHz and acquired online using a VME-bus computer system. Data analysis was performed using Igor Pro (Wavemetrics) on a Power Macintosh computer using analysis procedures identical to those previously described for neurons in the stratum lucidum region (Spruston et al. 1997 Morphological analysis
During recording, cells were filled with 0.5% biocytin (15-30 min) to reveal their morphology. At the end of the recording, slices were immersion-fixed in a phosphate buffered solution containing 2% paraformaldehyde and 1% glutaraldehyde (100 mM; pH 7.4) overnight. Slices then were processed using an avidin-horseradish peroxidase (HRP), avidin-biotin-complex (ABC)-solution (1:150, Vector Laboratories) as described (Lübke et al. 1996 Morphological identification of hilar neurons
Some morphological features of hilar neurons could be distinguished in the living slice using IR-DIC videomicroscopy. Putative basket cells were identified by their relatively large somata and position adjacent to the granule cell layer. Putative mossy cells could be identified by their large ovoid or triangular somata, although in some cases, further morphological analysis after biocytin staining revealed that these cells lacked the dense thorny spines characteristic of mossy cells. We could not distinguish between other spiny and aspiny hilar neurons using IR-DIC videomicroscopy alone. Recordings therefore were made from neurons in all areas of the hilus. Successful morphological identification of cell type was achieved in >90% of biocytin-filled neurons after avidin-HRP staining and light microscopic examination of somatic shape and location, dendritic arborization, and axonal projection. The morphological and physiological properties are described here for six types of neurons: dentate granule cells, dentate basket cells, hilar mossy cells, spiny hilar interneurons, and aspiny hilar interneurons projecting either to the inner molecular layer (IML) or outer molecular layer (OML).
Dendritic arborizations and axonal projection of dentate granule cells
Although not part of the hilar region, granule cells of the fascia dentata were characterized because they are an integral part of the circuitry of the hilus (Frotscher et al. 1994 Physiology of dentate granule cells
An example of the physiological properties of a dentate granule cell is shown in Fig. 1, B-D. At the maximum firing rate, little spike frequency accommodation is apparent because the patch pipette contained 10 mM EGTA. In other neurons recorded with 0.5 mM internal EGTA, more accommodation was observed, but the maximum number of spikes during a 1-s current injection was not significantly different. Also of note is that granule cells displayed little or no sag during hyperpolarizing voltage changes (Fig. 1, B and C). The physiological properties of 16 granule cells are summarized in Table 1. The values for
Basket cells with axonal projection to the granule cell layer
A representative example of a basket cell is given in Figs. 2A and 3A. Interneurons of this type resemble those described in the early Golgi studies by Ramón y Cajal (1911)
Physiology of basket cells
An example of a recording from a basket cell is shown in Fig. 3, B-D, and a summary of the physiological properties determined from three such recordings is given in Table 1. Basket cells displayed several characteristic physiological properties. The maximum number of action potentials during a 1-s current injection was significantly higher and the action potential half-width significantly narrower than any of the other cell types studied. In addition, RN was lower and Mossy cells with local axonal collaterals in the hilus
Besides granule cells, the major spiny cell type of the fascia dentata are the mossy cells (Amaral 1978
Physiology of mossy cells
An example of the physiological properties of a mossy cell is shown in Fig. 5, B-D, and the summary of the properties measured from nine neurons is given in Table 1. Mossy cells had the lowest firing rate, on average, but this difference was only statistically significant for the comparison with basket cells. Action potentials in mossy cells also had the smallest fast AHPs. Our data indicate, however, that some of these differences are not statistically significant and that there is considerable overlap in the ranges of these parameters measured from mossy cells and other hilar neurons. The difference in mossy cell firing rate and spike width was only statistically significant for the comparison with basket cells, and the difference in fast AHP was significant for the comparison with basket cells and granule cells but not with spiny and IML aspiny hilar interneurons. Mossy cells had the slowest Spiny hilar interneurons with axonal projections to the outer molecular layer
In our sample, we identified three spiny interneurons that were distinguishable from mossy cells. These neurons lacked the characteristic excrescences of mossy cells, had somewhat smaller somata, longer and thinner spines than mossy cells, and had an axonal projection pattern that differed markedly from that of the mossy cells. An example of such a spiny interneuron is shown in Fig. 6A. The soma of this neuron is situated near the granule cell layer. Several primary dendrites emerge from the poles of the fusiform cell body, some of which bifurcate into two long secondary branches running within the hilar region or passing through the granule cell layer toward the molecular layer (Fig. 6A). The distal portions of all dendrites are covered with numerous long, thin spines.
Physiology of spiny hilar interneurons
An example of the physiological properties of a spiny hilar interneuron is shown in Fig. 6, B-D, and the properties of three such cells summarized in Table 1. As shown in Fig. 6, these cells reached maximum firing rates of ~70 action potentials during a 1-s current injection, considerably lower than the maximum rate of basket cell action potentials but comparable with the other cell types studied. Aspiny interneurons with axonal projection to the inner molecular layer
Hilar interneurons lacking spines fell into two distinct categories: those with axons projecting to the inner molecular layer and those with axons projecting to the outer molecular layer. Examples of each of these two cell types are shown in Figs. 7, 8, A and B, and 9A.
Aspiny interneurons with axonal projections to the outer molecular layer
A representative example of an aspiny hilar interneuron with an axonal projection to the outer molecular layer is given in Fig. 8B. These neurons (n = 4) usually had somata located within the deep hilus, 20-50 µm beneath the granule cell layer, and were fusiform in shape, with two- to four-thick smooth primary dendrites emerging from the two poles of the soma. After a short distance, these primary dendrites give rise to several secondary and tertiary branches, which either remain within the hilus or terminate in the granule cell layer or molecular layer, sometimes ending in a small terminal tuft (Fig. 8B). The characteristic feature of these neurons is their axonal projection terminating in the outer molecular layer in a particular vertically oriented fashion (Fig. 8B). The main axon emerges from either the soma or one of the primary dendrites (Fig. 8B). Shortly after leaving the soma, the axon gives off several very long axonal collaterals ( Physiology of aspiny hilar interneurons
The physiological properties obtained from four IML-projecting aspiny interneurons and two OML-projecting aspiny interneurons were similar (Table 1). An example of a recording from an identified IML aspiny neuron is shown in Fig. 9, B-D. This neuron is typical of both IML and OML aspiny interneurons in that it fires at moderate rates but exhibits considerable spike frequency accommodation and never approaches the rapid firing rates observed in basket cells. Aspiny hilar interneurons also displayed considerable sag in hyperpolarizing responses and hence could be distinguished from both granule cells and basket cells on the basis of this feature. The most characteristic physiological feature of aspiny interneurons was the large slow AHP after a train of action potentials (Fig. 9, B-D, Table 1). One IML and one OML aspiny interneuron, however, had relatively small slow AHP potentials, so it is clear that this parameter is not a perfect fingerprint for these neurons. RN of IML aspiny interneurons was significantly higher than for basket cells or mossy cells, but there was overlap in the range of values for all cell types except basket cells. The role of the hippocampus in information processing is determined by the physiological properties and synaptic connections of its various neuronal types. Our study of identified cell types in the hilar region focused on the axonal plexus and physiological characteristics of the neurons because previous Golgi studies did not provide sufficient staining of the axon of the cells and many physiological studies lacked the identification of cell types. Although our approach has the advantage of visually identifying neurons with infrared-DIC, some caveats should be raised.
Classification of cell types
In addition to the granule cells and basket cells, three major cell types are described: the mossy cell and the spiny and aspiny interneurons. Granule cells and mossy cells are likely to be excitatory, using glutamate as a neurotransmitter, whereas basket cells and hilar interneurons are thought to be GABAergic inhibitory interneurons (e.g., Halasy and Somogyi 1993 Physiological distinctions between cell types
The values of RN and Flow of information through the fascia dentata and hilus
As information arrives at the fascia dentata, it is segregated into inputs arriving via the associational/commissural afferents, medial perforant path, and lateral perforant path, which correspond to the inner, middle, and outer thirds of the molecular layer (e.g., Blackstad 1956
INTRODUCTION
Abstract
Introduction
Methods
Results
Discussion
References
). Furthermore, neurons in this region were shown to degenerate in temporal lobe epilepsy of humans (de Lanerolle et al. 1989) and experimentally induced epilepsy in rats (Scharfman and Schwartzkroin 1990; Sloviter 1987, 1991). Hilar neurons are also extremely vulnerable to kindling and ischemia (Benviniste and Diemer 1988; Cavazos and Sutula 1990; Crain et al. 1988; Hsu and Buzsáki 1993). Despite the importance of this region, the physiological and morphological properties of the different cell types in the hilus are not as well characterized as those of the principal cell types in the hippocampal formation (Amaral and Witter 1989, 1995; Jonas et al. 1993; Li et al. 1994; Lorente de Nó 1934; Ramón y Cajal 1911; Sik et al. 1993; Spruston and Johnston 1992).
; Frotscher et al. 1991; Lübbers and Frotscher 1987; Ribak and Seress 1983, 1988) and their immunoreactivity for neuronal transmitters and neuropeptides (e.g., Buckmaster et al. 1994; Hendry and Jones 1985; Léránth and Frotscher 1986; Léránth et al. 1984, 1988, 1990; Sloviter and Nilaver 1987; Soriano and Frotscher 1994) have revealed that the population of neurons in the hilus is very heterogeneous, which probably accounts in part for the lack of a detailed characterization of the electrophysiology and axonal projections of these neurons. In addition, the connections of these neurons are difficult to characterize because of the lack of laminated axonal efferents and afferents like those to and from the principal hippocampal cell types. Recently, intracellular labeling studies have shown that the axons of hilar interneurons preferentially terminate in different strata (Buckmaster and Schwartzkroin 1995a,b; Han et al. 1993; Mott et al. 1997; Sik et al. 1997). These studies suggest that there are distinct organizational principles that can be identified by studying the axonal projections of hilar neurons.
METHODS
Abstract
Introduction
Methods
Results
Discussion
References
). Abbreviated methods are presented here; for more details, see Spruston et al. (1997).
). An attempt was made to record only from neurons 
50 µm below the surface of the slice to reduce the probability that dendrites or axon collaterals were severed during slicing. However, neurons deeper than ~100 µm were difficult to visualize, so most recordings were made from neurons 50-100 µm below the surface of the slice. Patch-pipette recordings were made using electrodes pulled from borosilicate glass and having resistances of 3-10 M
when filled with an internal solution containing (in mM) 115 K-gluconate, 20 KCl, 10 Na-phosphocreatine, 0.5 ethylene glycol-bis(
-aminoethyl ether)-N,N,N
,N-tetraacetic acid (EGTA), 10 N-2-hydroxyethylpiperazine-N-2-ethanesulfonic acid, 4 Mg2+-ATP, and 0.3 GTP plus 50 units/ml creatine phosphokinase and 5 mg/ml biocytin, adjusted to pH 7.3 with KOH. In most cases, the pipette solution contained 0.5 mM EGTA, but in some recordings 10 mM EGTA was used instead. With the exception of the granule cell shown in Fig. 1, all recordings shown are from cells where 0.5 mM EGTA was present in the pipette. Recordings were made using an Axoclamp 2B amplifier in the bridge voltage-recording mode. Series resistance varied between 5 and 50 M in different recordings. The bath solution contained (in mM) 125 NaCl, 25 NaHCO3, 25 glucose, 2.5 KCl, 1.25 NaH2PO4, 2 CaCl2, and 1 MgCl2 and was bubbled with a 95% O2-5% CO2 gas mixture. In most experiments, 10 µM 6-cyano-7-nitroquinoxaline-2,3-dione (Tocris), 50 µM D-2-amino-5-phosphonopentanoic acid (Tocris), and 10 µM bicuculline methiodide (Sigma) were included in the extracellular solution to block spontaneous synaptic activity. Recordings were made at physiological temperature (35-37°C).
View larger version (31K):
[in a new window]
FIG. 1.
Paired morphology and physiology of a dentate granule cell. A: morphology of the biocytin-filled neuron. Dendrites (black) form a characteristic cone-shaped dendritic tree. Several axon collaterals (red) project deep into the hilus with the main axon projecting toward the pyramidal cell layer. 
, origin of the axon from the cell body; thick black line indicates the pial surface. H, hilar region; GL, granule cell layer; ML, molecular layer. Scale bar: 100 µm. B: responses to 1-s current injections of +250 and 
50 pA. Maximum firing (54 Hz) was observed with the +250-pA current injection. Sag ratio = 1.0 (50 pA response). C: voltage responses to current injections of 50 to +80 pA in 10-pA increments. Responses from 50 to +50 pA are averages of 11-22 trials; responses from +60 to +80 pA are single trials. o = 31 ms; action potential half-width = 0.67 ms. D: voltage-current relationship for the data shown in C. 
, steady state voltages; +, peak responses, most of which overlap; 
, linear regression through the steady state points between 20 and +20 pA. RN = 195 M; Vrest = 77 mV.
). The membrane time constant (m) was estimated from the slowest time constant (0) of exponential fits to the voltage transients in the linear region of the voltage-current plots with the previously described criteria for linearity (Spruston and Johnston 1992; Spruston et al. 1997). The membrane time constant could not be determined in a few cells that did not meet these criteria for linearity. "Sag" was quantified by the ratio of the steady state voltage to the peak voltage in response to current injections of 50 to 300 pA. The maximum firing rate was determined as the maximum number of action potentials that could be elicited by a 1-s depolarizing current injection (100-1,000 pA). The slow afterhyperpolarization (AHP) was measured as the most negative membrane potential (relative to the resting potential) after the maximum train of action potentials. Action potential half width and fast AHP were measured from the first action potential with a current injection just above threshold. Action potential amplitude and fast AHP were measured relative to the action potential threshold. All statistical analyses were performed using single factor analysis of variance with Tukey's multiple comparisons. Differences were regarded as statistically significant at the P < 0.05 level.
). Representative examples of well-filled hilar neurons and granule cells were examined with an Olympus BX 50 microscope, photographed, and drawn with a camera lucida at a final magnification of ×480.
RESULTS
Abstract
Introduction
Methods
Results
Discussion
References
; Lindsay and Scheibel 1981; Lorente de Nó 1934; Lübbers and Frotscher 1987; Seress and Pokorny 1981). Also, the inclusion of granule cells in this study facilitated a direct comparison of their physiology to that of the hilar neurons. Morphological analysis revealed that granule cells had their characteristic morphological features that have been well described before (Claiborne et al. 1986, 1990), namely ovoid somata with one or two primary apical dendrites, which branch and fan out to a cone-shaped dendritic field within the molecular layer (Fig. 1A). The main granule cell axon entered the hilus and gave rise to several mossy fiber collaterals (Fig. 1A). The main axon could be followed long distances running parallel to CA3c the pyramidal layer while giving off several short terminal branches. Both the main axon and axon collaterals gave rise to mossy fiber boutons spaced ~20 µm apart.
;o = 50 ms). The morphology of this neuron is reminiscent of the VIP-immunoreactive or the choline acetyltransferase-immunoreactive neurons found in the granule cell layer (Frotscher et al. 1986; Sloviter and Nilaver 1987).
o and RN are in reasonable agreement with those determined previously (Spruston and Johnston 1992; Staley et al. 1992). Granule cells typically had resting potentials that were more negative than the other cell types studied. This difference was statistically significant for all cell types except spiny hilar interneurons. Considerable overlap in the ranges of resting potentials, however, makes it impossible to identify a granule cell with certainty on the basis of resting potential. Granule cells could be distinguished, however, on the basis of a combination of resting potential, lack of sag in hyperpolarizing responses, and maximum firing rate, which was substantially lower than the only other cell type lacking sag, namely basket cells (see further).
View this table:
TABLE 1.
Physiological properties of neurons in the hilus and fascia dentata of the hippocampal formation
and Lorente de Nó (1934) and in later Golgi, Golgi/electron microscopic, and immunocytochemical studies (Amaral 1978; Lübbers and Frotscher 1987; Ribak and Seress 1983; Ribak et al. 1978; Seress and Frotscher 1991; Seress and Pokorny 1981). All basket cells (n = 6) were located 10-20 µm from the granule cell layer on the hilar side. Basket cells had large, round, ovoid or triangular somata, 20-25 µm in diameter. A thick main apical dendrite ascends through the granule cell layer where it branches off into thick secondary dendrites, which then terminate in the molecular layer, often with a terminal tuft (Fig. 2A). Two to six basal dendrites emerge from the base of the soma and extend toward the hilar region. The dendrites of all cells were aspiny, but spine-like or hair-like protrusions were found on secondary dendrites and the terminal tufts. Dendritic varicosities were found mainly on basal dendrites (Fig. 2A).
View larger version (183K):
[in a new window]
FIG. 2.
Representative example of a dentate basket cell. A: photomontage of the basket cell the camera lucida reconstruction and physiology of which is shown in Fig. 3. Note extensive axonal arborization in the granule cell layer of fascia dentata. B: high magnification of pericellular "basket" of boutons formed by a single collateral on the cell body of a granule cell. GL, granule cell layer; H, hilar region. Scale bar in A: 100 µm; B: 50 µm.
View larger version (43K):
[in a new window]
FIG. 3.
Paired morphology and physiology of a dentate basket cell. A: camera lucida reconstruction of the biocytin-filled basket cell shown in Fig. 2. Axonal projection (red) of the neuron is restricted mainly to the granule cell layer, although a few axonal collaterals enter the innermost portion of the molecular layer. Axonal collaterals extend along approximately two-thirds of the length of the granule cell layer. , origin of the axon from the cell body; thick black line indicates the pial surface. H, hilar region; GL, granule cell layer; ML, molecular layer. Scale bar: 100 µm. B: responses to 1-s current injections of +900 and 300 pA. Maximum firing (266 Hz) was observed with the +900-pA current injection. Sag ratio = 0.99 (300 pA response). C: voltage responses to current injections of 100 to +220 pA in 20-pA increments. All responses, except the action potentials, are averages of 4-15 trials. o = 11 ms; action potential half-width = 0.33 ms. D: voltage-current relationship for the data shown in C. , steady state voltages; +, peak responses, most of which overlap; , linear regression through the steady state points between 100 and +120 pA. RN = 54 M; Vrest = 52 mV.
o faster than for the other cell types in the hilar region. Most of these differences were statistically significant except for the difference in RN between basket cells and mossy cells. The combination of these differences are distinctive enough that any single basket cell recording is clearly distinguishable from other hilar neurons on the basis of the physiology alone.
; Buckmaster et al. 1993a,b, 1996; Frotscher et al. 1991; Ribak et al. 1985; Scharfman 1995; Soriano and Frotscher 1994). Mossy cells were identified by the dense "thorny excrescences" covering their dendrites (Figs. 4A, inset, and 5A). Mossy cells were encountered throughout the entire hilar region, although they were more prominent near the granule cell layer. They are large, multipolar or fusiform neurons; from the ovoid cell body up to seven thick primary dendrites originate, which usually bifurcate and give rise to several secondary and tertiary branches (Fig. 4, A and B, and 5A). Depending on the location of the neurons within the hilus, mossy cell dendrites run in different directions. Dendrites of neurons situated beneath the granule cell layer tend to run parallel to the granule cell layer (Figs. 4A and 5A). Mossy cells located more deeply in the hilus extend dendrites in all directions (Fig. 4B). Some dendritic branches pass through the granule cell layer, terminating in the inner molecular layer (not shown).
View larger version (116K):
[in a new window]
FIG. 4.
Morphological characteristics of hilar mossy cells. A and B: 2 examples of hilar mossy cells at low magnification to show the different arrangement of the dendritic tree. , main axon. Inset: high magnification of the typical excrescences on proximal mossy cell dendrites. C: local hilar collaterals of a mossy cell. D: single collateral of a mossy cell with en passant boutons traversing the granule cell layer toward the inner molecular layer. E: single axon collateral of a mossy cell entering and terminating within the pyramidal layer of CA3c. F and G: 2 examples of single, mossy cell collaterals leaving the hilus, ascending through the granule cell layer, and finally terminating in the inner molecular layer. H, hilar region; GL, granule cell layer; IML, inner molecular layer. Scale bars in A and B: 100 µm; C-G: 50 µm.
View larger version (39K):
[in a new window]
FIG. 5.
Paired morphology and physiology of a hilar mossy cell. A: morphology of the biocytin-filled neuron. Axonal projection of the neuron (red) is restricted mainly to the hilus with some collaterals projecting to CA3c. , origin of the axon from the cell body. H, hilar region; GL, granule cell layer; ML, molecular layer. Scale bar: 100 µm. B: responses to 1-s current injections of +1,000 and 300 pA. Maximum firing (44 Hz) was observed with the +1,000-pA current injection. Sag ratio = 0.75 (300 pA response). C: voltage responses to current injections of 150 to +150 pA in 50-pA increments and 40 to +40 pA in 10-pA increments. All responses, except the action potentials, are averages of 5-15 trials. o = 49 ms; action potential half-width = 0.89 ms. D: voltage-current relationship for the data shown in C. , steady state voltages; +, peak responses, most of which overlap; , linear regression through the steady state points between 50 and +50 pA. RN = 181 M; Vrest = 64 mV.
900 µm (Fig. 5A). Other axon collaterals could be followed entering and terminating within the CA3c pyramidal cell layer (Fig. 4E and 5A) or passing through the granule cell layer (Fig. 4D) and then taking a course parallel to the granule cell layer within the inner molecular layer (Fig. 4, F and G). The projection toward the inner molecular layer and CA3c seems to be variable between mossy cells, at least as far as it could be followed in the slice.
o values of the neurons characterized, yet had lower RN than all other cell types except basket cells, presumably because of the large somata and extensive dendritic arborizations in these cells. The sag in hyperpolarizing responses was prominent in mossy cells, which distinguished them from basket cells and granule cells (which lack sag) but not from spiny or aspiny hilar interneurons. In fact, the combination of physiological properties made it generally easy to distinguish mossy cells from granule cells or basket cells but overlapped enough with spiny and aspiny interneurons that it is difficult to identify mossy cells on the basis of spike firing and passive membrane properties alone.
View larger version (48K):
[in a new window]
FIG. 6.
Paired morphology and physiology of a spiny hilar interneuron. A: morphology of the biocytin-filled neuron. Axonal projection (red) to the molecular layer is restricted mainly to the outer zone. Note abundant collaterals in the hilus and in area CA3c. , origin of the axon from the cell body; thick black line indicates the pial surface. H, hilar region; GL, granule cell layer; ML, molecular layer. Scale bar: 100 µm. B: responses to a 1-s current injection of +800 pA. Maximum firing (70 Hz) was observed with the +800-pA current injection. C: voltage responses to current injections of 100 to +150 pA in 50-pA increments and 40 to +40 pA in 10-pA increments. All responses, except the 1 containing action potentials, are averages of 4-8 trials. Note that the spikes arise from a slow ramp in response to the +150-pA current injection. Presence of a larger initial depolarization in response to +100 pA suggests that a K+ current is activated by the larger current injection, which then slowly inactivates, causing the depolarizing ramp response to the +150-pA current injection. o = 35 ms; action potential half-width = 0.87 ms; sag ratio = 0.81 (100 pA response). D: voltage-current relationship for the data shown in C. , steady state voltages; +, peak responses, most of which overlap; , linear regression through the steady state points between 20 and +20 pA. RN = 286 M; Vrest = 71 mV.
View larger version (39K):
[in a new window]
FIG. 8.
Camera lucida reconstructions of 2 multipolar aspiny hilar interneurons. Arrows, origin of the axons; thick black lines indicate the pial surface. Note that the axonal projection (red) terminates mainly within the inner molecular layer closer to the granule cell layer (cell in A) and the outer molecular layer (cell in B). Both cells have a prominent axonal domain in the hilus and in CA3c. H, hilar region; GL, granule cell layer; ML, molecular layer. Scale bar: 100 µm.
o and RN were most similar to that of aspiny hilar interneurons and statistically different only from basket cells. Aside from being distinguishable from basket cells, spiny hilar interneurons could be distinguished physiologically only from granule cells, on the basis of the presence of a considerable sag in hyperpolarizing responses.
View larger version (150K):
[in a new window]
FIG. 7.
Representative example of an aspiny hilar interneuron projecting to the inner molecular layer. A: photomontage of the neuron at low power. Inset: single collateral, which forms a basket-like array of boutons ( ) around a granule cell soma. B: Hilar collaterals () of another neuron projecting mainly to the inner molecular layer. Inset: high magnification of a basket-like array of synaptic boutons () around the cell body of a hilar neuron. GL, granule cell layer; H, hilar region; IML, inner molecular layer. Scale bars, A: 100 µm, inset in A: 25 µm, B: 50 µm, inset in B: 25 µm.
View larger version (46K):
[in a new window]
FIG. 9.
Paired morphology and physiology of an aspiny hilar interneuron. A: morphology of the biocytin-filled neuron Axonal projection of the neuron (red) is confined mainly to the inner molecular layer. , origin of the axon from the cell body; thick black line indicates the pial surface. H, hilar region; GL, granule cell layer; ML, molecular layer. Scale bar: 100 µm. B: responses to 1-s current injections of +200 and 50 pA. Maximum firing (58 Hz) was observed with the +200-pA current injection. Sag ratio = 0.84 (50 pA response). C: voltage responses to current injections of 20 to +25 pA in 5-pA increments. All responses, except the action potentials, are averages of 4-5 trials. o = 40 ms; action potential half-width = 0.55 ms. D: voltage-current relationship for the data shown in C. , steady state voltages; +, peak responses, most of which overlap; , linear regression through the steady state points between 20 and +15 pA. RN = 369 M; Vrest = 59 mV.
1 mm. While passing through the granule cell layer, they establish basket-like boutons around granule cell somata (Fig. 7A, inset). A second population of axonal collaterals found within the hilus (Figs. 7B, 8A, and 9A) is variable in density and range. Some of them run parallel to the granule cell layer for a long distance (800 µm), whereas others could be followed toward CA3c. Within the hilus, these neurons also form basket-like bouton arrays around somata (Fig. 7B, inset).
900 µm). One population then takes a more or less vertical course passing through the granule cell layer, the inner molecular layer, and finally terminating within the outer molecular layer. Within the inner molecular layer, some of these collaterals give rise to tangentially oriented side branches of variable length (30-300 µm) that also terminate within the outer molecular layer. On their way through the granule cell layer, these vertically oriented axonal collaterals form basket-like bouton arrays around granule cell somata. A second population of collaterals establishes a sometimes dense axonal plexus within the hilus, with en passant boutons contacting somata of other hilar neurons. Some of these collaterals were seen to project to CA3c(Fig. 8B).
o of IML aspiny interneurons was only statistically different from basket cells, and it appeared to be comparable with that of granule cells, mossy cells, and spiny interneurons (Table 1).
DISCUSSION
Abstract
Introduction
Methods
Results
Discussion
References
), could not be examined. Second, speculations about which connections are excitatory or inhibitory remain tentative in the absence of physiological analysis of these connections with paired recordings from morphologically identified neurons. In addition, some hilar neurons are likely to perform neuromodulatory functions, as indicated by the presence of various neuroactive peptides in these cells. Third, our cells are from relatively young animals (2- to 5-wk old), and we cannot exclude that they undergo further physiological and morphological maturation. Fourth, the need to fill the neurons with biocytin precluded the use of perforated-patch recording and hence introduces possible errors in physiological properties as a result of the influence of the pipette solution on the intracellular composition. One such possible problem is the influence of the EGTA in the pipette solution on spike firing properties. Because buffering internal calcium was shown to reduce spike frequency accommodation and AHP (Madison and Nicoll 1984), we used a low concentration of internal EGTA (0.5 mM). To examine the possible effects of internal EGTA, we also compared spike train responses in granule cells with electrodes containing 0.5 and 10 mM EGTA (see also Staley et al. 1992). Although some loss of accommodation was observed with the higher EGTA concentration when using weaker stimulation, no differences between the 0.5 and 10 mM internal EGTA groups were observed when measuring maximum firing rate, fast AHP, or slow AHP using strong depolarizing current pulses. Assuming that the normal calcium-buffering capacity of granule cells is at equivalent to 0.5 mM EGTA, these data suggest that the inclusion of 0.5 mM EGTA does not affect these properties (in granule cells at least) under our experimental conditions. Furthermore, the physiological properties recorded remained generally stable over time, suggesting that any artifacts introduced by the pipette solution must occur immediately or be subtle; nevertheless, we cannot rule out the problem entirely. Finally, it should be noted that the properties examined here were measured in an attempt to identify electrophysiological signatures of the different types of hilar neurons in slices. In vivo, some of these properties are likely to differ, and the actual firing patterns of these neurons will depend on the nature of their synaptic and neuromodulatory inputs in the behaving animal.
,b; Halasy and Somogyi 1993; Han et al. 1993; Mott et al. 1997; Sik et al 1997).
; Ribak et al. 1978; Sloviter and Nilaver 1987; Soriano and Frotscher 1993, 1994). Aspiny hilar interneurons can be subdivided further into those with a predominant projection to the inner molecular layer or the outer molecular layer. Other authors have adopted similar classification schemes for hilar neurons (Buckmaster and Schwartzkroin 1995a,b; Frotscher et al. 1994; Han et al. 1993; Mott et al. 1997; Seay-Lowe and Claiborne 1992; Sik et al. 1997; Soriano and Frotscher 1993). There is evidence that the various hilar neurons projecting to subzones of the molecular layer give rise to commissural fibers terminating in the same zones of the contralateral fascia dentata (Deller et al. 1995; see also Sik et al. 1997).
; at least some of these neurons could contain cholecystokinin (Fredens et al. 1987; Kosaka et al. 1985; Léránth and Frotscher 1986; Sloviter and Nilaver 1987; Somogyi et al. 1984; Stengaard-Pedersen et al. 1983). The aspiny hilar interneurons projecting to the outer molecular layer may correspond to the HIPP cells (hilar interneurons with axons ramifying in the perforant path terminal field) described by Han et al. (1993), which are likely to contain neuropeptide Y (Deller and Léránth 1990) and/or somatostatin (Bakst et al. 1986; Léránth et al. 1990). However, as the axonal collaterals of these neurons traverse the entire molecular layer, they also may correspond to the cell with collaterals throughout the molecular layer described in Golgi preparations (Soriano and Frotscher 1993). Spiny hilar interneurons, like those stained in vivo by Buckmaster and Schwartzkroin (1995a,b), also may contain somatostatin (Bakst et al. 1986). We regard it as an important observation that all of the interneurons, except the basket cell, had additional axonal domains in the hilus and in CA3. These additional termination fields indicate that these neurons do not only act in a feedback manner on granule cell and basket cell dendrites but also have feedforward connections to other hilar and CA3 neurons. Moreover, in the absence of a detailed electron microscopic analysis, it cannot be excluded that the axons of spiny and aspiny neurons projecting to the OML give rise to numerous en passant synapses in the IML, which they traverse. The present classification of dentate hilar neurons is far from complete. In a Golgi study, Amaral (1978) described 21 neuronal cell types mainly on the basis of their dendritic configuration. Also, an important cell type, the dentate chandelier cell, was not encountered in our study. It recently has been described morphologically and physiologically in some detail (Buhl et al. 1994b; Soriano and Frotscher 1989; Soriano et al. 1990).
o determined in our study are somewhat higher than determined with microelectrodes (Buckmaster and Schwartzkroin 1995a,b; Buckmaster et al. 1993a,b; Scharfman 1992; Scharfman and Schwartzkroin 1988) probably due to a leak conductance caused by microelectrode impalement (Spruston and Johnston 1992; Staley et al. 1992). A previous patch-clamp study of hilar neurons found similar RN values to ours (Livsey and Vicini 1992). RN and o values may be effectively lower in vivo, however, because of the higher degree of spontaneous synaptic activity than in the slice (Buckmaster and Schwartzkroin 1995a,b).
,b; Livsey and Vicini 1992; Mott et al. 1997; Scharfman 1992). The basket cells were easily distinguishable from other hilar neurons on the basis of their low RN and o, lack of sag in hyperpolarizing responses, and rapid maximum firing rate. Granule cells also could be distinguished reliably because they were the only other cell type to lack sag but had much higher RN and o and much lower maximum firing rates than basket cells. Because basket cells and granule cells can be distinguished readily from all other hilar cell types studied and from one another, we then only need consider further distinctions between mossy cells, spiny, and IML and OML aspiny interneurons.
; Scharfman 1992), but again, there was considerable overlap in the ranges. The most reliable property for distinguishing mossy cells from aspiny hilar interneurons was the slow AHP after a train of action potentials, which was significantly larger for IML aspiny interneurons. Two aspiny interneurons, however, one IML and one OML, had slow AHPs as small as those of mossy cells; so even this property is not a perfect means of distinguishing mossy cells from aspiny hilar interneurons.
; Steward 1976). It is likely that these inputs convey different information to the granule cells, which may integrate these synaptic inputs in different ways (for simplification, other inputs to the granule cells, i.e., subcortical afferents, are neglected here). In this context, it is interesting that hilar interneurons project back to the molecular layer in a laminated fashion. As these neurons are presumably inhibitory, neurons projecting specifically to the outer molecular layer would be in a position to selectively shut down inputs from lateral perforant path, whereas those projecting specifically to the inner molecular layer would be in a position to shut down inputs from the contralateral hippocampus. Medial perforant path inputs could be inhibited by either type, depending on the exact location of inhibitory afferents in the middle molecular layer. As action potential initiation occurs near the soma of granule cells (Jefferys 1978), however, inhibitory interneurons projecting to the inner molecular layer could reduce the efficacy of excitatory inputs in all parts of the molecular layer via shunting inhibition. Interneurons formed basket-like arrays of boutons on granule cell somata in addition to their laminar projections to the molecular layer, however, so if the laminar specificity of the projections to the molecular layer is of functional significance, one also might expect the somatic contacts to be somehow functionally distinct from the dendritic inputs arising from the same axons.
, 1995). Such widespread inhibition of granule cells may be a general feature of inhibitory interneurons in the dentate gyrus. The axon collaterals of spiny and aspiny interneurons projecting to the molecular layer were also widespread, and in vivo, these neurons have been shown to have longitudinal projections extending 50% of the length of the septotemporal axis of the hippocampal formation and even have someaxon collaterals that extend into CA1 (Buckmaster and Schwartzkroin 1995a,b). The GABAergic dentate chandelier cell (Buhl et al. 1994b; Soriano and Frotscher 1989; Soriano et al. 1990) also contributes to the laminated inhibitory input of the granule cells by selectively contacting the axon initial segments of the granule cells.
), one possibility is that they are activated by the same inputs that activate granule cells and that they then feedback and inhibit granule cell firing with a delay, thus effectively imposing a refractory period on granule cell firing. Arguing against this idea, however, is the fact that the lower RN and o of basket cells might make them less likely to integrate inputs over tens of milliseconds, as granule cells might, but would likely require synchronous synaptic inputs to bring them to firing threshold. Hence granule cell firing might be inhibited by basket cell activation only when basket cells are activated by many synchronous inputs. The ability of basket cells to fire rapidly and the low RN and o of basket cells, as well as the rapid kinetics of 
-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptor channels in these neurons (Koh et al. 1995), imply that basket cells could reliably follow such synchronous inputs with action potential firing, even at very high frequencies, and possibly therefore provide a tonic form of granule cell inhibition (see Geiger et al. 1997).
; Soriano and Frotscher 1994), constitute one of the output neurons of the hilus, as their axons project to the contralateral hippocampus (Frotscher et al. 1991; Ribak et al. 1985). As shown in the present study, mossy cells have extensive axon collaterals within the hilus, suggesting that their role as local circuit neurons is probably as important as their function as output neurons. These local circuits could constitute an excitatory feedback network onto granule cells, pyramidal cells, basket cells, or hilar interneurons. Buckmaster and Schwartzkroin (1994) propose that just such a positive feedback loop may form an associational circuit that is important for memory formation. Experimental evidence for a positive feedback circuit from mossy cells to granule cells comes from voltage-sensitive dye and microelectrode recordings, suggesting that mossy cells can mediate feedback excitation of granule cells (Jackson and Scharfman 1996; Scharfman 1995). However, they also may play an important role in polysynaptic feedback inhibition of granule cells as mossy cells were found to innervate interneurons (Scharfman 1995)
probably including both basket cells and spiny and aspiny interneurons. This connection could play a role in the hyperexcitability of granule cells during epilepsy because mossy cells are particularly sensitive to ischemia, and their death could lead to a reduction in inhibition of granule cells (loss of excitatory drive to inhibitory interneurons, cf. Sloviter 1987, 1991).
; Soltesz and Mody 1994). The unresolved targets and inputs of hilar interneurons are likely to be addressed in the future with recordings and reconstructions of pairs of neurons in slices (see Buhl et al. 1994a, 1995; Geiger et al. 1997; Miles et al. 1996; Scharfman 1994, 1995; Scharfman et al. 1990).
| |
ACKNOWLEDGEMENTS |
|---|
The authors are grateful to B. Joch, S. Nestel, and M. Winter for technical assistance and to T. Mickus for providing some granule cell data.
This work was supported by the von Helmholtz-Programm of the Bundesministerium für Bildung und Forschung (J. Lübke), the Alexander von Humboldt Foundation (N. Spruston), and the Deutsche Forschungsgemeinschaft Grant Fr 620/1-6 and Leibniz Programm (M. Frotscher).
| |
FOOTNOTES |
|---|
Address for reprint requests: M. Frotscher, Anatomisches Institut der Universität Freiburg, PO Box 111, D-79001 Freiburg, Germany.
Received 15 July 1997; accepted in final form 27 October 1997.
| |
REFERENCES |
|---|
|
Abstract
Introduction
Methods
Results
Discussion
References
|
|---|
This article has been cited by other articles:
|
|
 |
|
  W. Zhang and P. S. Buckmaster Dysfunction of the Dentate Basket Cell Circuit in a Rat Model of Temporal Lobe Epilepsy J. Neurosci., June 17, 2009; 29(24): 7846 - 7856. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Larimer and B. W. Strowbridge Nonrandom Local Circuits in the Dentate Gyrus J. Neurosci., November 19, 2008; 28(47): 12212 - 12223. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Lee, K.-H. Lee, W.-K. Ho, and S.-H. Lee Target Cell-Specific Involvement of Presynaptic Mitochondria in Post-Tetanic Potentiation at Hippocampal Mossy Fiber Synapses J. Neurosci., December 12, 2007; 27(50): 13603 - 13613. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. A. Williams, P. Larimer, Y. Gao, and B. W. Strowbridge Semilunar Granule Cells: Glutamatergic Neurons in the Rat Dentate Gyrus with Axon Collaterals in the Inner Molecular Layer J. Neurosci., December 12, 2007; 27(50): 13756 - 13761. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. S. Richards, K. Bommert, G. Szabo, and R. Miles Differential expression of Na+/K+-ATPase {alpha}-subunits in mouse hippocampal interneurones and pyramidal cells J. Physiol., December 1, 2007; 585(2): 491 - 505. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Schmidt-Hieber, P. Jonas, and J. Bischofberger Subthreshold Dendritic Signal Processing and Coincidence Detection in Dentate Gyrus Granule Cells J. Neurosci., August 1, 2007; 27(31): 8430 - 8441. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L.-Y. Fu and A. N. van den Pol GABA excitation in mouse hilar neuropeptide Y neurons J. Physiol., March 1, 2007; 579(2): 445 - 464. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Dyhrfjeld-Johnsen, V. Santhakumar, R. J. Morgan, R. Huerta, L. Tsimring, and I. Soltesz Topological Determinants of Epileptogenesis in Large-Scale Structural and Functional Models of the Dentate Gyrus Derived From Experimental Data J Neurophysiol, February 1, 2007; 97(2): 1566 - 1587. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. M. Kerr and M. Capogna Unitary IPSPs enhance hilar mossy cell gain in the rat hippocampus J. Physiol., January 15, 2007; 578(2): 451 - 470. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Aponte, C.-C. Lien, E. Reisinger, and P. Jonas Hyperpolarization-activated cation channels in fast-spiking interneurons of rat hippocampus J. Physiol., July 1, 2006; 574(1): 229 - 243. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
V. Santhakumar, I. Aradi, and I. Soltesz Role of Mossy Fiber Sprouting and Mossy Cell Loss in Hyperexcitability: A Network Model of the Dentate Gyrus Incorporating Cell Types and Axonal Topography J Neurophysiol, January 1, 2005; 93(1): 437 - 453. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. J. Frazier, B. W. Strowbridge, and R. L. Papke Nicotinic Receptors on Local Circuit Neurons in Dentate Gyrus: A Potential Role in Regulation of Granule Cell Excitability J Neurophysiol, June 1, 2003; 89(6): 3018 - 3028. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. U. BRAUER, N. E. SAVASKAN, M. H. P. KOLE, M. PLASCHKE, L. M. MONTEGGIA, E. J. NESTLER, E. SIMBURGER, R. A. DEISZ, O. NINNEMANN, and R. NITSCH Molecular and functional analysis of hyperpolarization-activated pacemaker channels in the hippocampus after entorhinal cortex lesion FASEB J, December 1, 2001; 15(14): 2689 - 2701. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. K Ellerkmann, V. Riazanski, C. E Elger, B. W Urban, and H. Beck Slow recovery from inactivation regulates the availability of voltage-dependent Na+ channels in hippocampal granule cells, hilar neurons and basket cells J. Physiol., April 15, 2001; 532(2): 385 - 397. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Lynch, U. Sayin, G. Golarai, and T. Sutula NMDA Receptor-Dependent Plasticity of Granule Cell Spiking in the Dentate Gyrus of Normal and Epileptic Rats J Neurophysiol, December 1, 2000; 84(6): 2868 - 2879. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Acsady, I. Katona, F. J. Martinez-Guijarro, G. Buzsaki, and T. F. Freund Unusual Target Selectivity of Perisomatic Inhibitory Cells in the Hilar Region of the Rat Hippocampus J. Neurosci., September 15, 2000; 20(18): 6907 - 6919. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. E. Scharfman, J. H. Goodman, and A. L. Sollas Granule-Like Neurons at the Hilar/CA3 Border after Status Epilepticus and Their Synchrony with Area CA3 Pyramidal Cells: Functional Implications of Seizure-Induced Neurogenesis J. Neurosci., August 15, 2000; 20(16): 6144 - 6158. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
U. Kraushaar and P. Jonas Efficacy and Stability of Quantal GABA Release at a Hippocampal Interneuron-Principal Neuron Synapse J. Neurosci., August 1, 2000; 20(15): 5594 - 5607. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
V. Santhakumar, R. Bender, M. Frotscher, S. T Ross, G. S Hollrigel, Z. Toth, and I. Soltesz Granule cell hyperexcitability in the early post-traumatic rat dentate gyrus: the 'irritable mossy cell' hypothesis J. Physiol., April 1, 2000; 524(1): 117 - 134. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Lynch and T. Sutula Recurrent Excitatory Connectivity in the Dentate Gyrus of Kindled and Kainic Acid-Treated Rats J Neurophysiol, February 1, 2000; 83(2): 693 - 704. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Martina, S. Royer, and D. Pare Physiological Properties of Central Medial and Central Lateral Amygdala Neurons J Neurophysiol, October 1, 1999; 82(4): 1843 - 1854. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
