0361-9230/91 $3.00 + 03
Brain Research Bulletin,Vol. 26. pp. 333-338. 0 Pergamon Press plc, 1991. Printed in the U.S.A.
Electrophysiological Comparison of Pyramidal and Stellate Nonpyramidal Neurons in Dissociated Cell Culture of Rat Hippocampus JEFFREY R. BUCHHALTER AND MARC A. DICHTER Division of Neurology, The Childrens Hospital of Philadelphia 34th and Civic Center Boulevard, Philadelphia, PA 19104 The Graduate Hospital and the University of Pennsylvania School of Medicine Received 27 August 1990
BUCHHALTER, J. R. AND M. A. DICHTER. Elech-ophysiological comparison of pyramidal and stellate nonpyramidal neurons in dissociated cell culture of rar hippocampus. BRAIN RES BULL 26(3) 333-338, 1991. -Differences in electrophysiological
properties between pyramidal and nonpyramidal neurons have been previously demonstrated in the hippocampal slice preparation. However, it has also been shown that nonpyramidal neurons from several hippocampal regions have different morphologies as well as different active membrane characteristics. In this study, active and passive electrophysiological characteristics of pyramidal and a single morphological type of nonpyramidal neuron in rat dissociated hippocampal culture were examined with intracellular recording at room temperature and at 33-3X. No significant differences were noted between the two groups of neurons, at either temperature, with regard to action potential amplitude, rate of rise and fall, duration at half maximal amplitude, adaptation to steady state depolarization or resting membrane potential and input resistance. We conclude that these electrophysiological properties cannot be used to distinguish these two neuron types in dissociated hippocampal cell culture. Hippocampus
Pyramidal neuron
Nonpyramidal neuron
Electrophysiology
Tissue culture
ramidal neurons could be distinguished on the basis of action potential duration, firing pattern, threshold, and inactivation (35). Subsequent studies have demonstrated that nonpyramidal neurons from different hippocampal layers do not have identical electrophysiological properties (12, 18, 19) and, in fact, have properties which overlap those of the pyramidal neurons. When embryonic hippocampal neurons are removed and grown in dissociated cell culture, they develop an impressive similarity to their in situ counterparts with regard to general morphologies (4, 5, 7) and synapse formation (6). However, differences in axonal development have been observed (5,7). This should not be surprising as these cells develop in an artificial two-dimensional environment with or without a serum supplemented medium. Furthermore, the extent of “functional differentiation” is unclear, although voltage- (14, 27-29) and ligand- (8, 21, 36) gated channels have been extensively described. In order to test the hypothesis that electrophysiological properties differ between neurons of different morphologies, we studied dissociated hippocampal neurons in culture for 3-4 weeks. This is an issue of significance given the underlying assumption of tissue culture studies, i.e., cells in culture mimic cells in vivo. A previous investigation of hippocampal neurons in dissociated cell culture reported that “no clear physiological differences” were noted between pyramidal and multipolar cells when properties such as input resistance, resting membrane potential, mem-
THE mammalian cortex contains a large number of neurons which can be subdivided into specific subtypes based on a variety of
criteria. A common criterion for neuronal identification is morphological, and over many years anatomists have developed different schemes to categorize cortical neurons. The simplest classification scheme distinguishes pyramidal cells from nonpyramidal cells (39) but other, more complicated schemes incorporating spine density have also been proposed (11). In addition, cortical neurons can be further identified by their relative positions in a cortical layer or stratum. For example, pyramidal neurons in layer 4 of somatosensory cortex are different from pyramidal neurons in layer 2, and pyramidal neurons in the CA1 region of the hippocampus are morphologically distinguishable from pyramidal neurons in the CA3 region. More recently, anatomic distinctions have also been made based on the content of a variety of synaptic transmitters or neuropeptides (24). Over the last several years, as the brain slice technique became more popular and dyes became available for intracellular marking after electrophysiological recording, attempts have been made to correlate different physiological properties with different morphological cell types in cortex. Most of these studies have focused on differences between spike generating properties and spike afterpotentials in pyramidal versus nonpyramidal neurons. A pioneering study which utilized the combined morphologicalphysiological approach reported that CA1 pyramidal and nonpy333
BIJCHHALTER
AND DICHTER
FIG. 1. Phase contrast photomicrographs of hippocampal neurons maintained in culture for 3 weeks. The neurons were stained with a modified Wright’& stain. (A) Pyramidal neuron-note the triangular soma, stout apical dendrite and smaller diameter basilar dendrites. (B) Nonpyramidal, stellate neuron. Processes and background cells slightly blurred due to the shallow plane of focus. In A and B, scale bar is 20 micrometers, x 624.
time constant and electronic length were studied (36). However, the analysis did not include the properties noted above which have been reported to distinguish these cell types in slice. The present study was performed to determine if hippocampal neurons of distinct pyramidal and a single nonpyramidal “stellate” morphology in dissociated cell culture exhibit the differences in electrophysiological properties described in slice. brane
METHOD Tissue Culture
Embryos from pregnant Sprague-Dawley rats (Charles River) were removed on embryonic days 19 or 20. The meninges were removed and hippocampi were incubated with trypsin (0.027%, Sigma) in a 3X, 95% 0,. 5% CO, incubator for 40 minutes. The trypsin solution was removed and the hippocampi were washed several times with a warm, Ca2+ and Mgzt free N2-hydroxyethylpiperazine-N-2-ethane sulfonic acid (HEPES) buffered saline solution (HBS). The hippocampi were placed in
warm growth media (DMEM with 25 n&l HEPES supplemented with 10% fetal calf serum, 10% Hams F-12, and 50 PI/ml penicillin-streptomycin) and triturated with a sterile Pasteur pipet. Viability, as assessed by Trypan Blue exclusion, was always greater than 75%. Six hundred thousand viable cells were plated into 35mm Petri dishes which contained 5, 12-mm coverslips which had been coated with poly-L-lysine (Peninsula Laboratories) and preincubated in growth media for 48 hours. Media was partially exchanged 3 times each week. Nonneuronal cell growth was inhibited with cytosine arabinoside (5 PM) at 7-10 days in vitro for 48 hours. Neurons were used for recording at 34 weeks in vitro. These culture methods were similar to those described by others (26,36). Neuron Morphologic
Criteria
Neurons selected for impalement had either a pyramidal or stellate morphology. Pyramidal neurons had a roughly triangular soma with one process larger than the others (presumably the apical dendrite) and 2 or 3 other processes (Fig. 1A) as described
335
ELECTROPHYSIOLOGY OF HIPPOCAMPAL NEURONS IN VITRO
in low density cultures of hippocampal neurons (5). Nonpyramidal neurons of stellate morphology had 4 or 5 processes of approximately equal diameter (Fig. 1B). Thus, bipolar and fusiform morphologies were excluded.
TABLE 1 ACTIVE AND PASSIVE MEMBRANE PROPERTIES OF PYRAMIDAL AND NONPYRAMIDAL NEURONS AS A FUNCTION OF TEMPERATURE
33-35°C
22-26°C Electrophysiology
Coverslips were placed in HBS (ionic composition in mM: 145 NaCl, 3 KCl, IO HEPES, 8 glucose 2 CaCl,, 1 MgCl,) on the stage of a Niion inverted microscope. In~cellul~ electrodes were filled with 3.5 M potassium acetate and had resistances of SO-130 megohms. Stimulation of neurons, storage and analysis of responses was performed via a DEC 11173 based data acquisition and analysis system (Indec, Cheshim Data). Neurons were included for analysis when stable recordings were obtained with resting membrane potentials of greater than -40 mV and action potential amplitudes of greater than 40 mV. The duration of the action potential was determined at one-half peak amplitude. The absence or presence of action potential inactivation during a short train of spikes induced by a 50-ms depolarization was noted. The input resistance, Ri,, was measured as the linear portion of the slope of the current-voltage (I-V) plot. The voltage deflection was measured during the steady state portion of the response to a 50-ms current square wave. Recordings were performed at room temperature (22-26”(Z) or with the bath warmed (33-3X) via a heating ring surrounding the bath controlled by a thermistor positioned adjacent to the coverslip in the center of the 35mm dish. The temperature gradient across the coverslip was 51°C. RESULTS
A total of 75 neurons which met the criteria stated above were analyzed with regard to passive and active membrane properties (Table 1). Recordings were performed at both room temperature (2%26°C N=44) and with the bath heated (33-35*C, N=31) as it has been well documented that membrane properties can be altered by heating or cooling neurons (38,41). Recordings were grouped by neuronal morphology in each temperature range for data analysis. The passive and active membrane ch~acte~stics of the pyramidal and nonpyramidal neurons were generally comparable to values reported in the literature for hippocampal neurons in the slice preparation (3 1, 33, 40) and for hippocampal (25,36) and neocortical (9) neurons in tissue culture. However, we could find no differences in either passive membrane properties (RMP, R,, tau) or action potential parameters between pyramidal neurons and nonpyramidal neurons recorded at the same temperatures. Action potentials in the vast majority of neurons of each cell type exhibited spike inactivation when repetitive fling was evoked by a 50-ms depolarization (stellate 20121 at 22-26”C, 11115 at 3335°C and pyramidal 20/23 at 22-26”C, 15/16 at 33-3X). Spike inactivation following the first action potential is illustrated in Fig. 2, although 2-3 spikes were commonly evoked before complete inactivation. Thus, no differences were found between these two cell types in their ability to sustain repetitive spike firing as
was reported for comparable neurons in the slice preparation (35). When neuronal properties were compared in the two different temperature ranges (with mo~hologic~ types pooled), clear differences emerged. Passive properties were unaffected, but action potentials had more rapid rates of rise and fall [maximum rate of rise, 72 VS. 120 V/s (t=5.8, p
Brain Research Bulletin,Vol. 26. pp. 333-338. 0 Pergamon Press plc, 1991. Printed in the U.S.A.
Electrophysiological Comparison of Pyramidal and Stellate Nonpyramidal Neurons in Dissociated Cell Culture of Rat Hippocampus JEFFREY R. BUCHHALTER AND MARC A. DICHTER Division of Neurology, The Childrens Hospital of Philadelphia 34th and Civic Center Boulevard, Philadelphia, PA 19104 The Graduate Hospital and the University of Pennsylvania School of Medicine Received 27 August 1990
BUCHHALTER, J. R. AND M. A. DICHTER. Elech-ophysiological comparison of pyramidal and stellate nonpyramidal neurons in dissociated cell culture of rar hippocampus. BRAIN RES BULL 26(3) 333-338, 1991. -Differences in electrophysiological
properties between pyramidal and nonpyramidal neurons have been previously demonstrated in the hippocampal slice preparation. However, it has also been shown that nonpyramidal neurons from several hippocampal regions have different morphologies as well as different active membrane characteristics. In this study, active and passive electrophysiological characteristics of pyramidal and a single morphological type of nonpyramidal neuron in rat dissociated hippocampal culture were examined with intracellular recording at room temperature and at 33-3X. No significant differences were noted between the two groups of neurons, at either temperature, with regard to action potential amplitude, rate of rise and fall, duration at half maximal amplitude, adaptation to steady state depolarization or resting membrane potential and input resistance. We conclude that these electrophysiological properties cannot be used to distinguish these two neuron types in dissociated hippocampal cell culture. Hippocampus
Pyramidal neuron
Nonpyramidal neuron
Electrophysiology
Tissue culture
ramidal neurons could be distinguished on the basis of action potential duration, firing pattern, threshold, and inactivation (35). Subsequent studies have demonstrated that nonpyramidal neurons from different hippocampal layers do not have identical electrophysiological properties (12, 18, 19) and, in fact, have properties which overlap those of the pyramidal neurons. When embryonic hippocampal neurons are removed and grown in dissociated cell culture, they develop an impressive similarity to their in situ counterparts with regard to general morphologies (4, 5, 7) and synapse formation (6). However, differences in axonal development have been observed (5,7). This should not be surprising as these cells develop in an artificial two-dimensional environment with or without a serum supplemented medium. Furthermore, the extent of “functional differentiation” is unclear, although voltage- (14, 27-29) and ligand- (8, 21, 36) gated channels have been extensively described. In order to test the hypothesis that electrophysiological properties differ between neurons of different morphologies, we studied dissociated hippocampal neurons in culture for 3-4 weeks. This is an issue of significance given the underlying assumption of tissue culture studies, i.e., cells in culture mimic cells in vivo. A previous investigation of hippocampal neurons in dissociated cell culture reported that “no clear physiological differences” were noted between pyramidal and multipolar cells when properties such as input resistance, resting membrane potential, mem-
THE mammalian cortex contains a large number of neurons which can be subdivided into specific subtypes based on a variety of
criteria. A common criterion for neuronal identification is morphological, and over many years anatomists have developed different schemes to categorize cortical neurons. The simplest classification scheme distinguishes pyramidal cells from nonpyramidal cells (39) but other, more complicated schemes incorporating spine density have also been proposed (11). In addition, cortical neurons can be further identified by their relative positions in a cortical layer or stratum. For example, pyramidal neurons in layer 4 of somatosensory cortex are different from pyramidal neurons in layer 2, and pyramidal neurons in the CA1 region of the hippocampus are morphologically distinguishable from pyramidal neurons in the CA3 region. More recently, anatomic distinctions have also been made based on the content of a variety of synaptic transmitters or neuropeptides (24). Over the last several years, as the brain slice technique became more popular and dyes became available for intracellular marking after electrophysiological recording, attempts have been made to correlate different physiological properties with different morphological cell types in cortex. Most of these studies have focused on differences between spike generating properties and spike afterpotentials in pyramidal versus nonpyramidal neurons. A pioneering study which utilized the combined morphologicalphysiological approach reported that CA1 pyramidal and nonpy333
BIJCHHALTER
AND DICHTER
FIG. 1. Phase contrast photomicrographs of hippocampal neurons maintained in culture for 3 weeks. The neurons were stained with a modified Wright’& stain. (A) Pyramidal neuron-note the triangular soma, stout apical dendrite and smaller diameter basilar dendrites. (B) Nonpyramidal, stellate neuron. Processes and background cells slightly blurred due to the shallow plane of focus. In A and B, scale bar is 20 micrometers, x 624.
time constant and electronic length were studied (36). However, the analysis did not include the properties noted above which have been reported to distinguish these cell types in slice. The present study was performed to determine if hippocampal neurons of distinct pyramidal and a single nonpyramidal “stellate” morphology in dissociated cell culture exhibit the differences in electrophysiological properties described in slice. brane
METHOD Tissue Culture
Embryos from pregnant Sprague-Dawley rats (Charles River) were removed on embryonic days 19 or 20. The meninges were removed and hippocampi were incubated with trypsin (0.027%, Sigma) in a 3X, 95% 0,. 5% CO, incubator for 40 minutes. The trypsin solution was removed and the hippocampi were washed several times with a warm, Ca2+ and Mgzt free N2-hydroxyethylpiperazine-N-2-ethane sulfonic acid (HEPES) buffered saline solution (HBS). The hippocampi were placed in
warm growth media (DMEM with 25 n&l HEPES supplemented with 10% fetal calf serum, 10% Hams F-12, and 50 PI/ml penicillin-streptomycin) and triturated with a sterile Pasteur pipet. Viability, as assessed by Trypan Blue exclusion, was always greater than 75%. Six hundred thousand viable cells were plated into 35mm Petri dishes which contained 5, 12-mm coverslips which had been coated with poly-L-lysine (Peninsula Laboratories) and preincubated in growth media for 48 hours. Media was partially exchanged 3 times each week. Nonneuronal cell growth was inhibited with cytosine arabinoside (5 PM) at 7-10 days in vitro for 48 hours. Neurons were used for recording at 34 weeks in vitro. These culture methods were similar to those described by others (26,36). Neuron Morphologic
Criteria
Neurons selected for impalement had either a pyramidal or stellate morphology. Pyramidal neurons had a roughly triangular soma with one process larger than the others (presumably the apical dendrite) and 2 or 3 other processes (Fig. 1A) as described
335
ELECTROPHYSIOLOGY OF HIPPOCAMPAL NEURONS IN VITRO
in low density cultures of hippocampal neurons (5). Nonpyramidal neurons of stellate morphology had 4 or 5 processes of approximately equal diameter (Fig. 1B). Thus, bipolar and fusiform morphologies were excluded.
TABLE 1 ACTIVE AND PASSIVE MEMBRANE PROPERTIES OF PYRAMIDAL AND NONPYRAMIDAL NEURONS AS A FUNCTION OF TEMPERATURE
33-35°C
22-26°C Electrophysiology
Coverslips were placed in HBS (ionic composition in mM: 145 NaCl, 3 KCl, IO HEPES, 8 glucose 2 CaCl,, 1 MgCl,) on the stage of a Niion inverted microscope. In~cellul~ electrodes were filled with 3.5 M potassium acetate and had resistances of SO-130 megohms. Stimulation of neurons, storage and analysis of responses was performed via a DEC 11173 based data acquisition and analysis system (Indec, Cheshim Data). Neurons were included for analysis when stable recordings were obtained with resting membrane potentials of greater than -40 mV and action potential amplitudes of greater than 40 mV. The duration of the action potential was determined at one-half peak amplitude. The absence or presence of action potential inactivation during a short train of spikes induced by a 50-ms depolarization was noted. The input resistance, Ri,, was measured as the linear portion of the slope of the current-voltage (I-V) plot. The voltage deflection was measured during the steady state portion of the response to a 50-ms current square wave. Recordings were performed at room temperature (22-26”(Z) or with the bath warmed (33-3X) via a heating ring surrounding the bath controlled by a thermistor positioned adjacent to the coverslip in the center of the 35mm dish. The temperature gradient across the coverslip was 51°C. RESULTS
A total of 75 neurons which met the criteria stated above were analyzed with regard to passive and active membrane properties (Table 1). Recordings were performed at both room temperature (2%26°C N=44) and with the bath heated (33-35*C, N=31) as it has been well documented that membrane properties can be altered by heating or cooling neurons (38,41). Recordings were grouped by neuronal morphology in each temperature range for data analysis. The passive and active membrane ch~acte~stics of the pyramidal and nonpyramidal neurons were generally comparable to values reported in the literature for hippocampal neurons in the slice preparation (3 1, 33, 40) and for hippocampal (25,36) and neocortical (9) neurons in tissue culture. However, we could find no differences in either passive membrane properties (RMP, R,, tau) or action potential parameters between pyramidal neurons and nonpyramidal neurons recorded at the same temperatures. Action potentials in the vast majority of neurons of each cell type exhibited spike inactivation when repetitive fling was evoked by a 50-ms depolarization (stellate 20121 at 22-26”C, 11115 at 3335°C and pyramidal 20/23 at 22-26”C, 15/16 at 33-3X). Spike inactivation following the first action potential is illustrated in Fig. 2, although 2-3 spikes were commonly evoked before complete inactivation. Thus, no differences were found between these two cell types in their ability to sustain repetitive spike firing as
was reported for comparable neurons in the slice preparation (35). When neuronal properties were compared in the two different temperature ranges (with mo~hologic~ types pooled), clear differences emerged. Passive properties were unaffected, but action potentials had more rapid rates of rise and fall [maximum rate of rise, 72 VS. 120 V/s (t=5.8, p
