Neurosciencc Lcm'rs', 142 (1992) 205 210 ', 1992 Elsevier Scientific Publishers Ireland Ltd. All rights reserved 0304-3940/92/S 05.00
205
NSL 08826
Potentiation of spontaneous synaptic activity in rat mossy cells Ben W . Strowbridge ~', Paul S. B u c k m a s t e r ~ a n d Philip A. S c h w a r t z k r o i n ~,.b D~Tarlmentx ol "Physioh~gy and Biophysics and/'Neurological Surger)', l_'nivcrsilv o/W.shinglon, Seattle, VVA 9~;195 ( USA (Received 16 December 1992; Revised version received 27 April 1992: Accepted 4 May 1992)
Kin, wordw
Mossy cell: Hippocampus: Dentate hilus: Excitatory postsynaptic potential
Recent studies have demonstrated the vulnerability of dentate mossy cells to seizure-induced damage. One source of potentially damaging synaptic input are spontaneously active granule cell terminals Cmossy terminals'.) We sought to test whether there were activity-dependent changes in the spontaneous excitatory input to mossy ceils, Using the in vitro slice preparation, we examined the frequency and amplitude of spontaneous excitatory postsynaptic potentials (EPSPs) after intracellular current injection designed to mimic the extreme depolarization these neurons receive during repetitive afferent stimulation. In 4 of 7 neurons, depolarization with trains of current pulses resulted in a significant and persistent increase in frequent? of spontaneous synaptic depolarizations (to an average of 178% of the initial baseline rate). In 3 of these affected neurons, an increased frequency of large amplitude, last-rising EPSPs accounted for the majority of this change. Injection of hyperpolarizing current pulses failed to alter spontaneous activity in 3 other mossy cells. These results suggest spontaneous synaptic input to mossy cells is plastic and can be potentiated by depolarization of a single postsynaptic mossy cell. The ability of mossy cells to potentiate their excitatory input may be relewmt to their vulnerabilit~ to excitotoxic injury during repetitive afferent stimulation.
Mossy cells, the major cell type in the dentate hilus, have attracted much recent attention because of their vulnerability to some types of experimentally induced seizure activiD in the rodent dentate gyrus [15, 16, 19]. These neurons and certain types of non-spiny hilar neurons also are damaged in epileptic patients with mesial temporal sclerosis [3, 5]. Mossy cells and other 'sensitive' hilar neurons share several common properties such as strong excitatory input from granule cells [11] and prominent spontaneous excitatory postsynaptic potentials (EPSPs) [12]. In addition, mossy cells show little or 11o immunostaining for neuropeptides or calcium-binding proteins such as calbindin and parvalbumin [17, 18]. Also. Scharflnan and Schwartzkroin [13] demonstrated that intracellular injection of a calcium chelator 1,2bis(2-aminophenoxy)ethane-N,N,N',N'-tetraacetic acid (BAPTA) could protect mossy cells and other 'sensitive" hilar neurons from damage during repetitive stimulation of the perforant path in vitro. These studies have focused attention on the regulation of intracellular calcium levels during excitotoxic injury in vulnerable hilar neurons. We now report evidence for another mechanism that may contribute to the vulnerability of mossy cells. We have Corre,s7~ondence: P.A. Schwartzkroin, Department of Neurological Surgery, RI-20, University of Washington, Seattle. WA 98195, USA.
observed that the rate and amplitude of mossy cell spontaneous EPSPs can be increased by intracellular depolarization of the postsynaptic mossy cell. Hippocampal slices (400 pro) were prepared from anesthetized (sodium pentobarbital, 50 mg/kg i.p.) adult female Sprague Dawley rats with a Vibroslice (Frederick Haer). Slices were maintained in an interface chamber at 35°C and bathed in artificial cerebrospinal fluid (in mM: NaCl 126, KCI 5, CaCI~ 2. MgSO4 1.2, NaH2PO 4 1.25, NaHCO 3 26 and dextrose 10: pH 7.4). Intracellular recordings were made through glass microelectrodes (60 100 ME2) filled with either 4 M potassium acetate or 2% biocytin dissolved in 2 M potassium acetate. Standard current clamp recording equipment and procedures were used [12]. Spontaneous fluctuations in the membrane potential were amplified and digitized at 1 5 kHz. To permit morphological identification of putative mossy cells, most neurons were filled iontophoretically with biocytin using either depolarizing or hyperpolarizing current pulses (1 2 nA, 300 ms duration, 50% duty cycle for 10 min.) The same current injection protocol was used to test the effect of intracellular depolarization on spontaneous EPSPs. Previous studies [11, 12, 14] demonstrated that activation of presynaptic afferents can result in massive depolarization (>20 mV) of mossy
206
¢
A
/
/
f/
/
B
t
z,// ,I /
Fig. 1. Electrophysiology and morphology of a mossy cell. A: responses of a typical mossy cell (celt A in Table 11to depolarizing and hyperpolarizing current pulses. Note relatively wide action potentials with little afterhyperpolarization following each spike. Apparent inpt,1 resistanc'e calculated from the ohmic deflection in response to the hyperpolarizing current pulse was 43 M(2: the resting membrane potential or"this neurotl ~t~s 65 m¥~ Calibration: I0 mV, 0.5 nA, 10 ms. B: camera lucida drawing of cell A after it was filled with biocytin and visualized with DAB (see te.\t for details). C: photomicrograph of the same cell showing large soma, thorny excrescences (arrows) typical for thi~ hilar cell type Bar -200 tim.
cells; by using this current injection protocol, we hoped to mimic some of the postsynaptic effects of repetitive afferent stimulation. After the current injection protocol, slices remained in the incubation c h a m b e r for 10-30 min, then r e m o v e d and fixed overnight in 4% p a r a f o r m a l d e hyde in 0.1 M p h o s p h a t e buffer. Slices were sectioned at 40 # m and biocytin-filled neurons were visualized using the A B C - p e r o x i d a s e m e t h o d [6] with the c h r o m o g e n diaminobenzidine. Quantitative electrophysiological data were recorded f r o m 10 hilar neurons which were classified as mossy cells on the basis o f their characteristic electrophysiology [11, 12, 14]. M o s s y cells generate b r o a d action potentials with little afterhyperpolarization (see Fig. 1A) which contrasts with other hilar cell types [8, 12, 14]. Eight neurons identified as m o s s y cells based on their electrophysiology (the other two were not stained) subsequently were
shown to exhibit the morphological characteristics of mossy cells as described by A m a r a l et al. [1, 10] such as thorny excrescences on the proximal dendrites, classical spines on distal dendrites, large s o m a diameter (>25/.tin), and dendrites largely confined to the hilus (Fig. 1B,C). Previous studies [11, 12, 14, 19] have reported large (5-10 mV) s p o n t a n e o u s depolarizations in rodent mossy cells. These depolarizations are likely to be EPSPs since they are blocked completely by 6-cyano-7-nitroquinoxaline2,3-dione ( C N Q X , 10 /tM; unpublished observations) and alterations o f the m e m b r a n e potential consistently failed to reveal 'hidden' IPSPs. All morphologicallyidentified mossy cells in our study (n--8) displayed these depolarizations. These neurons can be contrasted with CA3 pyramidal cells in which both s p o n t a n e o u s EPSPs and IPSPs have been observed [7] (unpublished observations).
207
A
B After Depolarization
Before Depolarization
6o
I",
\,
P
V,
?
Fig. 2. Spontaneous activity before and after treatment with depolarizing current (same neuron as in Fig. 1 ). Straight line drawn at - 7 4 mV. A: spontaneous activity before treatment (traces 1-6). B: spontaneous activity after 10 min of depolarizing current pulses ( 1.25 hA, 301) ms in duration, 50% duty cycle). Note that the lowest excursions during most sweeps, before and after treatment, reached 74 mV which was therefore set as RM R While both the tYequency of spontaneous EPSPs and the m a x i m u m amplitudes increased following treatment (a change that persisted for more than 5 rain), R M P was not altered. Little change was detected in the apparent input resistance (43 MQ before treatment, 47 Mg2 alter treatment). Also note that Ihe sweeps after treatment contain a 'new" population of large amplitude, fast-rising EPSPs. Calibration: 10 inV. 60 ms.
Spontaneous activity was assessed during a baseline period several minutes before starting the 'treatment' in which 300 ms current pulses were delivered at 1.7 Hz for 10 min. Spontaneous synaptic activity was again quantified several minutes after stimulation offset. The spontaneous EPSP frequency was estimated by counting the number of distinct depolarizations greater than 2 mV in 3 12 sweeps (either 600 or 1,000 ms in duration} recorded before and alter treatment. These estimates, made independently by two investigators who were blinded as to the 'treatment' received by the cell (see below), ranged from 5 to 26 Hz. The high frequency of overlapping EPSPs precluded an accurate estimation of resting membrane potential; therefore, a statistical description of the EPSP amplitude was not attempted. The neurons reported in this study were selected from a larger population of treated cells based on the follow-
ing criteria: (i) stability of the apparent resting membrane potential (RMP) to within 5 mV (RMP estimated by the most hyperpolarized troughs between the EPSPs in one sweep), (ii) stability of the apparent input resistance (less than a 20% change), (iii) absence of transient changes in the frequency of spontaneous depolarizations. The latter criteria was imposed because our analysis method required sampling a large set of recordings over several minutes, from which we determined properties of the spontaneous EPSPs: transient changes in PSP frequency would make it difficult to assess longer-term alterations due to treatment. Seven neurons treated with depolarizing current pulses (1 2 nA) and 3 neurons treated with hyperpolarizing pulses (1 2 nA) met these criteria. In 4 mossy cells, treatment with depolarizing current pulses significantly increased the frequency of spontane-
TABIA-I I Effect of current injection on spontaneous EPSP rate. Average rates of spontaneous depolarizations (in Hz) are listed in order of increasing baseline rate; mean ~ S.E.M. are given for each investigator (invest~ invcstO. Cells A G were treated with depolarizing current pulses (see text for protocol); cells H-J were treated with hyperpotarizmg current pulses. Cells D and 1 were recorded with electrodes which contained only the electrolyte (4 M potassium acetate): the other 8 cells were filled with biocytin dissolved in 2 M potassium acetate. The last column lists the rate of spontaneous EPSPs after treatment as a percentage of Ihe rates before treatment (averaged for the two investigator counts), as determined 1 5 rain after treatment. Cell
Depolarized A B C D E F G
Frequency before (EPSPs/s)
Frequency after (EPSPs/s)
6.1_+l.3/ 5,8_+1.1 16.1+2.0/15.8_+1.9 8.5_+1.4/ 7.8_+1.8 7.3_+1.3/8.0_+1.1 tl.8_+1.2/10.3+1.0 17.0.+1.5/16.3_+1.8 14.8_+1.1/14.6_+0.522.8+_l.1/21.5.+1.3 15.0.+0.9/t5.9_+0.9 12.8_+0.6/14.4_+0.6 18. l_+0.8/18.5_+0.6 26.3++_1.7/24.5_+1,1 19.8+0.7/18.9_+0.921.9_+1.1/21.9+1.2
Percentof baseline
268%** 94% 152%* 152%** 89% 139%** 114% Mean: 144%
Hyperpolarized H 7.5_+0.6/ 7.4_+0.6 5.6_+1.3/5.1_+0.9 73% I 14.4_+1.3/18.2+2.2 15.7_+1.2/19.7_+2.4 109"/;, J 16.8_+2.2/14.2.+1.6 13.7.+0.9/12.1.+0.8 84% Mean: 88% *Before and after frequencies are statistically different, P
205
NSL 08826
Potentiation of spontaneous synaptic activity in rat mossy cells Ben W . Strowbridge ~', Paul S. B u c k m a s t e r ~ a n d Philip A. S c h w a r t z k r o i n ~,.b D~Tarlmentx ol "Physioh~gy and Biophysics and/'Neurological Surger)', l_'nivcrsilv o/W.shinglon, Seattle, VVA 9~;195 ( USA (Received 16 December 1992; Revised version received 27 April 1992: Accepted 4 May 1992)
Kin, wordw
Mossy cell: Hippocampus: Dentate hilus: Excitatory postsynaptic potential
Recent studies have demonstrated the vulnerability of dentate mossy cells to seizure-induced damage. One source of potentially damaging synaptic input are spontaneously active granule cell terminals Cmossy terminals'.) We sought to test whether there were activity-dependent changes in the spontaneous excitatory input to mossy ceils, Using the in vitro slice preparation, we examined the frequency and amplitude of spontaneous excitatory postsynaptic potentials (EPSPs) after intracellular current injection designed to mimic the extreme depolarization these neurons receive during repetitive afferent stimulation. In 4 of 7 neurons, depolarization with trains of current pulses resulted in a significant and persistent increase in frequent? of spontaneous synaptic depolarizations (to an average of 178% of the initial baseline rate). In 3 of these affected neurons, an increased frequency of large amplitude, last-rising EPSPs accounted for the majority of this change. Injection of hyperpolarizing current pulses failed to alter spontaneous activity in 3 other mossy cells. These results suggest spontaneous synaptic input to mossy cells is plastic and can be potentiated by depolarization of a single postsynaptic mossy cell. The ability of mossy cells to potentiate their excitatory input may be relewmt to their vulnerabilit~ to excitotoxic injury during repetitive afferent stimulation.
Mossy cells, the major cell type in the dentate hilus, have attracted much recent attention because of their vulnerability to some types of experimentally induced seizure activiD in the rodent dentate gyrus [15, 16, 19]. These neurons and certain types of non-spiny hilar neurons also are damaged in epileptic patients with mesial temporal sclerosis [3, 5]. Mossy cells and other 'sensitive' hilar neurons share several common properties such as strong excitatory input from granule cells [11] and prominent spontaneous excitatory postsynaptic potentials (EPSPs) [12]. In addition, mossy cells show little or 11o immunostaining for neuropeptides or calcium-binding proteins such as calbindin and parvalbumin [17, 18]. Also. Scharflnan and Schwartzkroin [13] demonstrated that intracellular injection of a calcium chelator 1,2bis(2-aminophenoxy)ethane-N,N,N',N'-tetraacetic acid (BAPTA) could protect mossy cells and other 'sensitive" hilar neurons from damage during repetitive stimulation of the perforant path in vitro. These studies have focused attention on the regulation of intracellular calcium levels during excitotoxic injury in vulnerable hilar neurons. We now report evidence for another mechanism that may contribute to the vulnerability of mossy cells. We have Corre,s7~ondence: P.A. Schwartzkroin, Department of Neurological Surgery, RI-20, University of Washington, Seattle. WA 98195, USA.
observed that the rate and amplitude of mossy cell spontaneous EPSPs can be increased by intracellular depolarization of the postsynaptic mossy cell. Hippocampal slices (400 pro) were prepared from anesthetized (sodium pentobarbital, 50 mg/kg i.p.) adult female Sprague Dawley rats with a Vibroslice (Frederick Haer). Slices were maintained in an interface chamber at 35°C and bathed in artificial cerebrospinal fluid (in mM: NaCl 126, KCI 5, CaCI~ 2. MgSO4 1.2, NaH2PO 4 1.25, NaHCO 3 26 and dextrose 10: pH 7.4). Intracellular recordings were made through glass microelectrodes (60 100 ME2) filled with either 4 M potassium acetate or 2% biocytin dissolved in 2 M potassium acetate. Standard current clamp recording equipment and procedures were used [12]. Spontaneous fluctuations in the membrane potential were amplified and digitized at 1 5 kHz. To permit morphological identification of putative mossy cells, most neurons were filled iontophoretically with biocytin using either depolarizing or hyperpolarizing current pulses (1 2 nA, 300 ms duration, 50% duty cycle for 10 min.) The same current injection protocol was used to test the effect of intracellular depolarization on spontaneous EPSPs. Previous studies [11, 12, 14] demonstrated that activation of presynaptic afferents can result in massive depolarization (>20 mV) of mossy
206
¢
A
/
/
f/
/
B
t
z,// ,I /
Fig. 1. Electrophysiology and morphology of a mossy cell. A: responses of a typical mossy cell (celt A in Table 11to depolarizing and hyperpolarizing current pulses. Note relatively wide action potentials with little afterhyperpolarization following each spike. Apparent inpt,1 resistanc'e calculated from the ohmic deflection in response to the hyperpolarizing current pulse was 43 M(2: the resting membrane potential or"this neurotl ~t~s 65 m¥~ Calibration: I0 mV, 0.5 nA, 10 ms. B: camera lucida drawing of cell A after it was filled with biocytin and visualized with DAB (see te.\t for details). C: photomicrograph of the same cell showing large soma, thorny excrescences (arrows) typical for thi~ hilar cell type Bar -200 tim.
cells; by using this current injection protocol, we hoped to mimic some of the postsynaptic effects of repetitive afferent stimulation. After the current injection protocol, slices remained in the incubation c h a m b e r for 10-30 min, then r e m o v e d and fixed overnight in 4% p a r a f o r m a l d e hyde in 0.1 M p h o s p h a t e buffer. Slices were sectioned at 40 # m and biocytin-filled neurons were visualized using the A B C - p e r o x i d a s e m e t h o d [6] with the c h r o m o g e n diaminobenzidine. Quantitative electrophysiological data were recorded f r o m 10 hilar neurons which were classified as mossy cells on the basis o f their characteristic electrophysiology [11, 12, 14]. M o s s y cells generate b r o a d action potentials with little afterhyperpolarization (see Fig. 1A) which contrasts with other hilar cell types [8, 12, 14]. Eight neurons identified as m o s s y cells based on their electrophysiology (the other two were not stained) subsequently were
shown to exhibit the morphological characteristics of mossy cells as described by A m a r a l et al. [1, 10] such as thorny excrescences on the proximal dendrites, classical spines on distal dendrites, large s o m a diameter (>25/.tin), and dendrites largely confined to the hilus (Fig. 1B,C). Previous studies [11, 12, 14, 19] have reported large (5-10 mV) s p o n t a n e o u s depolarizations in rodent mossy cells. These depolarizations are likely to be EPSPs since they are blocked completely by 6-cyano-7-nitroquinoxaline2,3-dione ( C N Q X , 10 /tM; unpublished observations) and alterations o f the m e m b r a n e potential consistently failed to reveal 'hidden' IPSPs. All morphologicallyidentified mossy cells in our study (n--8) displayed these depolarizations. These neurons can be contrasted with CA3 pyramidal cells in which both s p o n t a n e o u s EPSPs and IPSPs have been observed [7] (unpublished observations).
207
A
B After Depolarization
Before Depolarization
6o
I",
\,
P
V,
?
Fig. 2. Spontaneous activity before and after treatment with depolarizing current (same neuron as in Fig. 1 ). Straight line drawn at - 7 4 mV. A: spontaneous activity before treatment (traces 1-6). B: spontaneous activity after 10 min of depolarizing current pulses ( 1.25 hA, 301) ms in duration, 50% duty cycle). Note that the lowest excursions during most sweeps, before and after treatment, reached 74 mV which was therefore set as RM R While both the tYequency of spontaneous EPSPs and the m a x i m u m amplitudes increased following treatment (a change that persisted for more than 5 rain), R M P was not altered. Little change was detected in the apparent input resistance (43 MQ before treatment, 47 Mg2 alter treatment). Also note that Ihe sweeps after treatment contain a 'new" population of large amplitude, fast-rising EPSPs. Calibration: 10 inV. 60 ms.
Spontaneous activity was assessed during a baseline period several minutes before starting the 'treatment' in which 300 ms current pulses were delivered at 1.7 Hz for 10 min. Spontaneous synaptic activity was again quantified several minutes after stimulation offset. The spontaneous EPSP frequency was estimated by counting the number of distinct depolarizations greater than 2 mV in 3 12 sweeps (either 600 or 1,000 ms in duration} recorded before and alter treatment. These estimates, made independently by two investigators who were blinded as to the 'treatment' received by the cell (see below), ranged from 5 to 26 Hz. The high frequency of overlapping EPSPs precluded an accurate estimation of resting membrane potential; therefore, a statistical description of the EPSP amplitude was not attempted. The neurons reported in this study were selected from a larger population of treated cells based on the follow-
ing criteria: (i) stability of the apparent resting membrane potential (RMP) to within 5 mV (RMP estimated by the most hyperpolarized troughs between the EPSPs in one sweep), (ii) stability of the apparent input resistance (less than a 20% change), (iii) absence of transient changes in the frequency of spontaneous depolarizations. The latter criteria was imposed because our analysis method required sampling a large set of recordings over several minutes, from which we determined properties of the spontaneous EPSPs: transient changes in PSP frequency would make it difficult to assess longer-term alterations due to treatment. Seven neurons treated with depolarizing current pulses (1 2 nA) and 3 neurons treated with hyperpolarizing pulses (1 2 nA) met these criteria. In 4 mossy cells, treatment with depolarizing current pulses significantly increased the frequency of spontane-
TABIA-I I Effect of current injection on spontaneous EPSP rate. Average rates of spontaneous depolarizations (in Hz) are listed in order of increasing baseline rate; mean ~ S.E.M. are given for each investigator (invest~ invcstO. Cells A G were treated with depolarizing current pulses (see text for protocol); cells H-J were treated with hyperpotarizmg current pulses. Cells D and 1 were recorded with electrodes which contained only the electrolyte (4 M potassium acetate): the other 8 cells were filled with biocytin dissolved in 2 M potassium acetate. The last column lists the rate of spontaneous EPSPs after treatment as a percentage of Ihe rates before treatment (averaged for the two investigator counts), as determined 1 5 rain after treatment. Cell
Depolarized A B C D E F G
Frequency before (EPSPs/s)
Frequency after (EPSPs/s)
6.1_+l.3/ 5,8_+1.1 16.1+2.0/15.8_+1.9 8.5_+1.4/ 7.8_+1.8 7.3_+1.3/8.0_+1.1 tl.8_+1.2/10.3+1.0 17.0.+1.5/16.3_+1.8 14.8_+1.1/14.6_+0.522.8+_l.1/21.5.+1.3 15.0.+0.9/t5.9_+0.9 12.8_+0.6/14.4_+0.6 18. l_+0.8/18.5_+0.6 26.3++_1.7/24.5_+1,1 19.8+0.7/18.9_+0.921.9_+1.1/21.9+1.2
Percentof baseline
268%** 94% 152%* 152%** 89% 139%** 114% Mean: 144%
Hyperpolarized H 7.5_+0.6/ 7.4_+0.6 5.6_+1.3/5.1_+0.9 73% I 14.4_+1.3/18.2+2.2 15.7_+1.2/19.7_+2.4 109"/;, J 16.8_+2.2/14.2.+1.6 13.7.+0.9/12.1.+0.8 84% Mean: 88% *Before and after frequencies are statistically different, P
