| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
The Journal of Neurophysiology Vol. 78 No. 5 November 1997, pp. 2682-2692
Copyright ©1997 by the American Physiological Society
1 Department of Physiology and Biophysics, University of Iowa, Iowa City, Iowa 52242; and 2 Cold Spring Harbor Laboratory, Cold Spring Harbor, New York 11724
 |
ABSTRACT |
|---|
 Abstract
 Introduction
Methods
Results
Discussion
References
|
|---|
Shirke, Aneil M. and Roberto Malinow. Mechanisms of potentiation by calcium-calmodulin kinase II of postsynaptic sensitivity in rat hippocampal CA1 neurons. J. Neurophysiol. 78: 2682-2692, 1997. Preactivated recombinant 
-calcium-calmodulin dependent multifunctional protein kinase II (CaMKII*) was perfused internally into CA1 hippocampal slice neurons to test the effect on synaptic transmission and responses to exogenous application of glutamate analogues. After measurement of baseline transmission, internal perfusion of CaMKII* increased synaptic strength in rat hippocampal neurons and diminished the fraction of synaptic failures. After measurement of baseline responses to applied transmitter, CaMKII* perfusion potentiated responses to kainate but not responses to N-methyl-D-aspartate. Internal perfusion of CaMKII*potentiated the maximal effect of kainate. Potentiation byCaMKII* did not change the time course of responses to kainate, whereas increasing response size by pharmacologically manipulating desensitization or deactivation rate constants significantly altered the time course of responses. Nonstationary fluctuation analysis of responses to kainate showed a decrease in the coefficient of variation after potentiation by CaMKII*. These data support the hypothesis that CaMKII increases postsynaptic responsiveness by increasing the available number of active -amino-3-hydroxy-5-methylisoxazole-4-propionic acid/kainate channels and suggests that a similar process may occur during the expression of long-term potentiation.
Understanding the mechanism of expression of long-term potentiation (LTP) has been a long-standing problem in the field of neurobiology (Malinow 1994 Hippocampal slices and recording techniques
Hippocampal slices were prepared from rats (Sprague-Dawley, 10- to 16-day old) on a Vibratome (TPI), as described previously (Otmakhov et al. 1993 Analysis
Stimulation artifacts were subtracted digitally from synaptic responses before analysis. Response amplitudes were quantified by calculating the average current during the 10-50 ms around the peak of the response and subtracting the average current during an interval of equal length immediately before the response. Response size after manipulations was measured as a percentage of baseline. The failure rate was estimated as double the percentage of traces in which the measured response was <0 pA after adjustment for stimulation artifact (Liao et al. 1995
INTRODUCTION
Abstract
Introduction
Methods
Results
Discussion
References
; Nicoll and Malenka 1995). The means by which synaptic connections are strengthened on a long-term basis has implications for learning and memory (Bliss and Collingridge 1993), as well as the development of the nervous system (Cline 1991). One proposed mechanism is a change in postsynaptic responsiveness (Bliss and Lomo 1973; Hebb 1949; Lynch and Baudry 1984; Malinow 1994; Nicoll and Malenka 1995; Nicoll et al. 1988).
; Kauer et al. 1988; Kullmann 1994; Kullmann et al. 1996; Liao et al. 1995; Muller and Lynch 1988; Muller et al. 1988; Perkel and Nicoll 1993; although see Bashir et al. 1991; Clark and Collingridge 1995; O'Connor et al. 1995; Xie et al. 1992). Quantal size, classically thought to be determined postsynaptically (DelCastillo and Katz 1954), increases during LTP (Isaac et al. 1996; Kullmann and Nicoll 1992; Larkman et al. 1992; Liao et al. 1992; Stricker et al. 1996). Other work has shown that synaptic transmission at hyperpolarized potentials undergoes LTP with a decrease in failure rate (Kullmann and Nicoll 1992; Malinow 1991; Malinow and Tsien 1990; Stevens and Wang 1994), whereas transmission at depolarized voltages largely is unaffected (Isaac et al. 1995; Liao et al. 1995; but see Kullmann et al. 1996). One interpretation of these data is that LTP can be expressed predominantly by non-NMDA receptors.
; Kennedy et al. 1983). Protocols that induce LTP increase CaMKII activity (Fukunaga et al. 1993) and result in phosphorylation of several substrates (Fukunaga et al. 1995). Pharmacological blockade of CaMKII prevents LTP (Malenka et al. 1989; Malinow et al. 1989); genetic elimination of CaMKII diminishes LTP (Silva et al. 1992). Increased postsynaptic activity of CaMKII potentiates synaptic transmission and occludes further LTP (Pettit et al. 1994). In frog optic tectum, increased postsynaptic CaMKII selectively potentiates -amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA)-mediated transmission and decreases synaptic failures at hyperpolarized but not depolarized potentials. Quantal size of AMPA responses also is increased (Wu et al. 1996). Autophosphorylation of CaMKII produces a constitutively active enzyme (Hanson and Schulman 1992) that could underlie the persistence of LTP (Lisman 1994). Postsynaptic application of thio-phosphorylated CaMKII (CaMKII*, an in vitro modification of the enzyme that renders it active and resistant to endogenous phosphatases) has been shown to increase somatic sensitivity to kainate in cultured hippocampal cells (Kolaj et al. 1994; McGlade-McCulloch et al. 1993) and to increase the strength of synaptic transmission and mutually occlude LTP in hippocampal slices (Lledo et al. 1995). Together, these studies provide strong evidence that CaMKII is necessary and sufficient to induce LTP.
METHODS
Abstract
Introduction
Methods
Results
Discussion
References
), and were submerged and superfused at 27°C. Whole cell recordings were acquired with a patch electrode (1.5-2.5 M
tip resistance) using an Axopatch 1-D (Axon Instruments) and a LM-900 interface (Dagan) operated by a custom program written in a Quickbasic environment (Microsoft) with the Axobasic software library (Axon Instruments). Visualized patch-clamping was performed with a CCD camera (Hamamatsu) attached to a Zeiss Axioskop fixed-stage microscope mounted on a translational stage. Synaptic responses were elicited by monopolar electrical stimulation of stratum radiatum via a glass microelectrode and a Grass SD9 stimulator. Agonist responses were elicited by pressure injection of receptor agonist (kainate or NMDA) using a Picospritzer (General Valve) via a patch pipette (~7 M) onto the soma of a CA1 neuron once every 10 s, except for saturation experiments when the interval was 20 s. We chose to use kainate to stimulate AMPA/kainate receptors for three reasons: first, kainate has been used previously to assay the effect of active CaMKII (McGlade-McCulloch et al. 1993). Second, because AMPA/kainate receptors do not desensitize completely to kainate, examination of steady state responses is possible (Patneau et al. 1993). Third, because kainate is a partial agonist (Patneau et al. 1993), the maximal response will be smaller and therefore more easily measured under voltage-clamp. Any neurotransmitter analogue may act differently from the endogenous transmitter, however. Agonists were delivered at concentrations (in the puffer pipette) of 0.1 mM for both kainate and NMDA, except where explicitly stated. Injection duration was 4-30 ms. Agonist was applied directly onto the soma as has been described previously (McGlade-McCulloch et al. 1993; Patneau et al. 1993). After establishment of stable baseline responses to either synaptic stimulation or agonist application, a volume of test (or control) solution approximately equal to the amount of solution initially present in the patch pipette (~4 µl) was perfused into the cell via a small canula placed within the pipette tip.
-aminoethyl ether)-N,N,N
,N-tetraacetic acid, 5 MgCl2, 2 ATP (Na2ATP), 0.3 GTP (Na2GTP), and 40 N-2-hydroxyethylpiperazine-N-2-ethanesulfonic acid (pH 7.2 with CsOH). In some experiments, Cs+ was substituted with K+. Internal perfusion solution contained (in mM) 100 Cs+ (or K+) gluconate, 5 MgAcetate, 5 reduced glutathione, 0.25 CaCl2, 0.006 calmodulin, 1 ATP, 0.25 ATP
S, and 0.055 EDTA, plus 0.5 mg/ml bovine serum albumin and 0.005 thio-phosphorylated constitutively active recombinant CaMKII[1-316] (courtesy of T. Soderling, Vollum Institute); in control solution, active kinase was replaced with kinase that had been heat inactivated (boiled for 15 min at 100°C) before reacting it with Ca2+, CaM and ATPS. For experiments with applied agonist, 5 mM 1,2-bis (2-aminophenoxy) ethane-N,N,N,N-tetraacetic acid (Molecular Probes, Eugene, OR) and 10 µM Microcystin-LR (RBI, Natick, MA), a phosphatase inhibitor, were added to both patch solution and internal perfusion solution. Cyclothiazide (courtesy of Lilly Pharmaceuticals, Indianapolis, IN) was prepared as a 10 mM stock solution in 1% dimethyl sulfoxide and was applied at concentrations of 
100 µM. All chemicals were from Sigma (St. Louis, MO) except where otherwise noted.
). Measurements of the rate of decay of responses were produced by least-squares best fit of a single exponential decay to averaged responses using a Levenberg-Marquardt fitting algorithm within the program Origin (Microcal Software). The half-width was measured as the time between half the peak response during the rising phase of the response and half the peak during the decay phase. Nonstationary fluctuation analysis (Sigworth 1980) was performed by calculating the point-by-point mean and standard deviation of a series of traces. The point-by-point coefficient of variant (CV) is calculated as the standard deviation divided by the mean. Theoretically, if N is the number of channels, P is the probability that the channel is open, and i is the single-channel current, then the mean response at any point during the time course of the response is
The coefficient of variation is
and
here the primed terms represent values after potentiation. Note that an increase in i should produce a change in the mean with no change in the CV of responses. Mean and CV were computed over 20-50 consecutive responses during the baseline period, and, after potentiation had stabilized, over the same number of responses before and after maximum potentiation. Baseline noise was subtracted before computation of CV. If the recording was lost before potentiation had stabilized, the data were excluded for this analysis.
Simulations
Simulations were done using a simple five-state model for non-NMDA receptor activation (Ambros-Ingerson and Lynch 1993) based on the acetylcholine receptor. The model has the following states: C = closed, CB = closed and ligand-bound, CD = desensitized, CBD = closed, ligand-bound and desensitized, and O = open. Rate constants are as shown in Fig. 5. Initial estimates for the rate constants were adapted from Ambros-Ingerson et al. (1993). The model was simulated using the program NEURON 3.1 for Windows (M. Hines, http://www.nnc.yale.edu/HTML/YALE/NNC/neuron/nrnsim.html).
|
| |
RESULTS |
|---|
|
Abstract
Introduction
Methods
Results
Discussion
References
|
|---|
Enhancement of synaptic transmission by internally perfused CaMKII*
Standard visualized whole cell patch recordings were obtained from hippocampal CA1 neurons and stable synaptic transmission was elicited for 10-15 min. To test the ability of CaMKII* to potentiate synaptic transmission, the recording pipette was perfused internally with a solution containing either CaMKII* or heat-inactivated CaMKII. Perfusion of active CaMKII* significantly potentiated synaptic transmission [Fig. 1, A-C; 146 ± 17% (mean ± SD) of preperfusion response amplitude; paired t-test: P < 0.05, n = 17], whereas inactive kinase did not (93 ± 9%; paired t-test: n.s., n = 17). Potentiation by active kinase was significantly different from the effect of inactive kinase (independent t-test: P < 0.05, n = 34). Some of these experiments were performed with K+ as the major cation instead of Cs+ with no obvious effect on the results.
|
Enhancement of responsiveness to glutamate analogues by CaMKII*
Changes in synaptic failure rates traditionally have been interpreted to indicate presynaptic modifications. However, more recently, postsynaptic mechanisms have been postulated to explain this finding (Isaac et al. 1995
Increase of the maximal effect of kainate
To investigate the mechanism of potentiation of kainate responses, several hypothetical mechanisms were tested. CaMKII* might increase responsiveness to kainate via a simple increase in receptor affinity for agonist; however, this hypothesis also implies that responses to a saturating dose of kainate, a measure of maximum efficacy, should not be potentiated by CaMKII* (Fig. 4A, Table 1). In contrast, an increase in the number of AMPA/kainate receptors would cause an increase of responses to all concentrations of kainate as well as of the maximal effect of kainate (Fig. 4B). Calculations from previous estimates of the AMPA/kainate receptor dose-response curve predict that 10 mM kainate will produce ~99% of a saturated response (Jonas and Sakmann 1992
Analysis of time course of kainate responses before and after CaMKII* perfusion
For a ligand-gated channel, the time course of the whole cell response depends on the time course of ligand concentration and on single-channel open time distribution (Colquhoun and Hawkes 1995
Nonstationary fluctuation analysis of kainate responses before and after CaMKII* perfusion
Nonstationary fluctuation analysis (Sigworth 1980
To summarize, perfusion of CaMKII* potentiated synaptic transmission and decreased failure rates. CaMKII* specifically potentiated kainate but not NMDA responsiveness. The maximal effect of kainate also was potentiated. Potentiation was not associated with a change in the time course of responses. Further, CaMKII* potentiation was accompanied by a decrease in the point-by-point CV.
The authors thank lab members for helpful criticism of the work, especially Dr. Z. Mainen and L. Johnson for comments on the manuscript, N. Dawkins for technical assistance, and Drs. D. Brickey and S. Mukherji in Dr. T. Soderling's laboratory for kindly providing recombinant CaMKII enzyme.
The authors thank the Mathers Foundation for its generous support. A. Shirke was supported by a National Institute of Mental Health predoctoral Fellowship 5 F31 MH-10524-02.
Address for reprint requests: R. Malinow, Cold Spring Harbor Laboratory, 1 Bungtown Rd., Cold Spring Harbor, NY 11724. Received 9 April 1997; accepted in final form 9 July 1997.
; Malinow and Tsien 1990), we wanted to test the effect of CaMKII* on synaptic failures. Stimulus strength was set so that synaptic failures (trials in which responses were not distinguishable from the background noise; Fig. 2A) were detected. As shown in Fig. 2, perfusion of active CaMKII* decreased the number of synaptic failures (Fig. 2B). In this example, response size increased ~8 min after internal perfusion of CaMKII* (Fig. 2C) with no significant change in series resistance (Fig. 2D). Quantitative failure analysis (see METHODS) showed a decrease in the failure rate, indicated by a decrease in the magnitude of the peak at 0 pA in histograms of response amplitudes (Fig. 2, E and F). The decrease in failure rate was a general finding after perfusion of active CaMKII* (Fig. 2G; before = 55 ± 6.6%, after = 41 ± 8.7%; paired t-test: P < 0.05, n = 11). These experiments show that active CaMKII is sufficient to potentiate synaptic transmission and decrease failure rates, as occurs during LTP (Bolshakov and Siegelbaum 1995; Kullmann and Nicoll 1992; Malinow 1991; Malinow and Tsien 1990; Stevens and Wang 1994).
View larger version (36K):
[in a new window]
FIG. 2.
Perfusion of constitutively active CaMKII* decreases synaptic failure rate. A and B: 40 consecutive superimposed traces 2 min before (A) and 12 min after (B) intrapipette perfusion of thio-phosphorylatedCaMKII*. Note the increase in response size and the decrease in the number of failures. C: synaptic response amplitude plotted as a function of time after perfusion of active kinase. A and B mark the center of epochs from which data in A and B were obtained. Filled portion of the bar indicates the period of active internal perfusion. D: series resistance, plotted vs. time, over the course of the experiment. E and F: smoothed and binned histograms of 125 response amplitudes before (E) and after (F) perfusion of CaMKII* show a marked decrease in the magnitude of the peak at 0 pA. A smoothed histogram of noise scaled to the 0 pA peak is superimposed. G: thio-phosphorylated CaMKII* decreases failure rate (n = 11; paired t-test: P < 0.05). Graph shows the failure rate before and after potentiation for each pathway analyzed. Large circles are the mean failure rates before and after perfusion.
; Liao et al. 1995). We wanted to examine more closely possible postsynaptic modulations produced by CaMKII*. First, we studied the specificity of CaMKII* potentiation for NMDA or non-NMDA receptor responses. Responses were monitored under voltage-clamp while glutamate analogues were pressure-applied to the soma. Figure 3 shows an experiment in which kainate and NMDA were puffed alternately. After achieving stable baseline responses to the two agonists of 
50 sweeps, CaMKII* was perfused into the patch pipette. About 10 min subsequent to perfusion, the response to kainate began to grow (Fig. 3, A-C). Overall, active CaMKII* significantly potentiated responses to kainate (Fig. 3H; 265 ± 54%; paired t-test: P < 0.05, n = 8), whereas inactive kinase had no significant effect (104 ± 21%, paired t-test: n.s., n = 5). The effect of active kinase was significantly different from inactive kinase (independent t-test: P < 0.05, n = 13). In the experiment shown in Fig. 3, a fluorescent marker (40 kDa RITC-dextran) was included in the internal perfusate. The time course of the increase in fluorescence intensity of marker in the cell body (Fig. 3G) closely matched potentiation of responses to kainate (Fig. 3C). This was a general finding (Fig. 3I; multiple regression analysis: P < 0.05, R = 0.95, n = 17).
View larger version (28K):
[in a new window]
FIG. 3.
Perfusion of constitutively activeCaMKII* specifically potentiates non-N-methyl-D-aspartate (NMDA) responses. A-G are all from the same experiment in which 100 µM kainate and 100 µM NMDA were pressure-applied alternately to the soma of CA1 neurons. A and B: average of 10 responses to kainate before (A) and after (B) perfusion of active CaMKII*. C: amplitude of responses to puffed kainate plotted vs. time after perfusion of CaMKII*. Filled portion of the bar indicates the period of active internal perfusion. A and B mark the center of epochs from which data in A and B were obtained. D and E: average of 10 responses to NMDA before (D) and after (E) perfusion of CaMKII*. F: amplitude of NMDA responses plotted against time after perfusion show a rapid-onset, persistent decrease. D and E mark the center of epochs from which data in D and E were obtained. G: fluorescence intensity of background-subtracted cell body images (insets) plotted vs. time. H: active CaMKII* potentiates responses to 100 µM kainate (265 ± 54%; paired t-test: P < 0.05, n = 8). Heat-inactivated CaMKII has no significant effect on responses to kainate (104 ± 21%, paired t-test: n.s., n = 5). Responses to NMDA decrease with perfusion of active CaMKII* (64 ± 8%; paired t-test: P < 0.05, n = 4); perfusion of inactive CaMKII has a similar effect (75 ± 4%; paired t-test: P < 0.05, n = 5). I: normalized amount of potentiation seen in four experiments is plotted against the normalized fluorescence at different times during an experiment. Each symbol represents the fluorescence and potentiation for an individual experiment; the squares are from the experiment shown in A-G. Note that with higher fluorescence, a higher degree of potentiation is observed. In all 4 experiments, 100% fluorescence corresponds with 100% potentiation.
). No further characterization of this effect was attempted.
; Patneau and Mayer 1991).
View larger version (22K):
[in a new window]
FIG. 4.
Perfusion of active CaMKII* increases the maximal effect of kainate. A: hypothetically, an increase of the binding affinity of the receptor would result in a leftward shift of the dose-response curve and could produce potentiation of subsaturated responses. However, such a change would not change the maximal effect of the agonist. B: an increase of receptor number could produce potentiation of responses to all concentrations of agonist and would increase the maximal effect of the agonist. C: responses to 10 mM kainate as the puff duration is increased are larger up to the maximal effect of kainate; however, further increases of peak response size are not seen. D: responses from 2 equidistant pipettes, 1 puffing 10 mM kainate (fine line) and 1 puffing 50 mM kainate (heavy line), are shown. E: internal perfusion of active CaMKII* potentiates the maximal effect of kainate. 
, size of average maximal responses to saturating applications of kainate; 
, average amplitude of subsaturated responses. Time 0 indicates the time at which the fluorescence of marker could first be detected in the cell. F: representative traces from an individual experiment with 10 mM kainate application. Fine line represents the baseline response; the heavy line represents the response after internal perfusion. Top: maximal responses. Bottom: subsaturated responses. G: responses to application of 50 mM kainate are not significantly larger than responses to application of 10 mM kainate onto the same cell (sat + 
conc.; 98 ± 2.3%; paired t-test: n.s., n = 4). Responses to 10 mM kainate are potentiated by perfusion of CaMKII*(sat + active; 116 ± 3%, paired t-test: P < 0.05, n = 4), as were responses to subsaturating doses of kainate during the same experiment (subsat + active; 114 ± 3%, paired t-test: P < 0.05, n = 4).
View this table:
TABLE 1.
Theoretical estimates of the effect of voltage-clamp compromise on measured potentiation
(paired t-test: P < 0.05, n = 4). On puffing agonist, the input resistance dropped as low as 30 M. Exchanging K+ for Cs+ in internal solutions did not affect this response. This low resistance, presumably due to activation of AMPA/kainate conductances, will compromise the voltage clamp. Note that potentiation of responses to 10 mM kainate is evidence that the voltage-clamp was not completely compromised before application of CaMKII*. Thus although potentiation could be detected, interpretation of responses must take this into account.
; Jonas and Spruston 1994). Therefore, time course analysis may provide a method of detecting mechanistic changes in receptor function as well as artifactual changes in the concentration of applied agonist. For example, relief of partial physical blockade of the puffer pipette or unintended movement of the puffer pipette toward the target cell could masquerade as potentiation. To test for such possibilities, the duration of pressure injection of agonist-containing solution was increased. This produced the expected increase in response size, but also resulted in a longer-lasting response (Fig. 6A). Decreasing the distance between the tip of the puffer pipette and the cell also increased the magnitude of responses and altered their time course; in addition, the onset time of the response decreased (Fig. 6B).
View larger version (25K):
[in a new window]
FIG. 6.
CaMKII* potentiation of responsiveness to non-NMDA agonists does not change the time course of responses. A: time course of responses changes when puff duration is varied between 6, 9, 12, and 20 ms. Raw averages are shown top; normalized responses (bottom) show the gross changes in time course associated with changing the time course of agonist concentration. B: moving the puffer pipette from 6 µM (thick line) to 3 µM (thin line) produces an increase in the magnitude of the response (top) as well as an alteration of the time course (bottom, normalized). Note the change in onset time, time to peak, and decay time course. C: CaMKII* potentiates responses (top) without changing the time course of responses (bottom, normalized). Heavy line, after potentiation, is nearly superimposable on the scaled fine line, before potentiation. D: cyclothiazide increases response magnitude (top) and slows decay time course (bottom, normalized). E and F: quantitative analysis of time course of responses augmented by cyclothiazide ( ) and potentiated by CaMKII* (). Time course measures (E: decay tau; F: half-width) normalized to baseline values are plotted against normalized amplitude increase.
; Patneau et al. 1993). Blockade of desensitization and/or deactivation has been shown to augment response size and to change the time course of responses to exogenously applied transmitter in isolated preparations (Tang et al. 1991). Application of cyclothiazide increased the magnitude of responses to puffed kainate and slowed their time course (Fig. 6D).
0.01, n = 6) while cyclothiazide-mediated augmentation was correlated with larger tau values (regression analysis: P < 0.05, R = 0.97, n = 4). CaMKII* potentiation also did not correlate with changes in half-width (regression analysis: n.s., R = 0.02, n = 6). Cyclothiazide potentiation correlated with a measurable change in half-width (regression analysis: P < 0.05, R = 0.97, n = 4).
) was used to test the hypothesis that CaMKII* potentiates kainate responsiveness by changing single-channel current. Because the number of receptors exposed to agonist is not likely to be constant during an individual response in this experimental paradigm, plots of variance versus mean derived from entire traces were not parabolic. However, over the time course of a response, the number of active channels at any point during the time course of a response should be constant from trial to trial. We therefore analyzed the point-by-point CV (standard deviation divided by mean) around the peak of the response. To determine if this analysis method is sensitive to a modification of single-channel current, the holding potential was changed. This is expected to alter the driving force and the single-channel current. Changing the driving force from 35 to 70 mV increased average response size, but had no effect on the CV (Fig. 7A), as expected (Hille 1992). If CaMKII* potentiates kainate responsiveness by increasing the single-channel current, no change in the nonstationary CV is predicted with potentiation. However, CaMKII*-mediated potentiation of responses to kainate decreased the CV (Fig. 7B), consistent with an increase in the number of AMPA receptors. The ratio of CV
2 before and after potentiation will be constant and equal to one for changes in singlechannel current. Potentiation of kainate responses byCaMKII* produced CV2 ratio values greater than one, and the ratio correlated with the amount of potentiation (Fig. 7C; regression analysis: P < 0.05, R = 0.89, n = 8). Mean potentiation for experiments used for CV analysis was 1.72 ± 0.19. The mean change in CV2 was 2.15 ± 0.58. The change in CV2 was significantly different from 1 and is therefore inconsistent with a hypothesis involving a change in single-channel conductance.
View larger version (11K):
[in a new window]
FIG. 7.
CaMKII* potentiation of responsiveness to non-NMDA agonists is accompanied by a decrease in the nonstationary coefficient of variation (CV). A: 50 consecutive responses to kainate at 35 mV (fine line) and 70 mV (heavy line) are shown bottom; the CV calculated over the entire time course of the responses is shown top. Driving force changes the size of kainate responses, but the point-by-point CV at each potential is similar. B: CaMKII* potentiates responses to exogenous kainate and decreases point-by-point CV. Mean response before (fine line) and after (heavy line) potentiation are shown bottom. Point-by-point CV is shown top. Note that CV decreased over the length of the response. C: ratio of CV2 after to CV2 before potentiation increases with the amount of potentiation. Individual points show the ratio value and amount of potentiation for experiment used for CV analysis. Circle with error bars is the average potentiation and change in ratio of CV2. Horizontal line is the predicted outcome for the effect of a simple change in the single-channel current. Diagonal line is the predicted outcome for a simple increase in receptor number.
DISCUSSION
Abstract
Introduction
Methods
Results
Discussion
References
; Pettit et al. 1994; Wang and Kelly 1995) and frog optic tectum (Wu et al. 1996). In contrast to those studies, the patch perfusion technique used in the present study allowed direct comparison of responses before and after a delimited increase of CaMKII activity. Interestingly, CaMKII*-mediated potentiation occurred long after the typical "washout" period for LTP (Malinow and Tsien 1990), implying that the processes underlying washout are before CaMKII activation. The delay between internal perfusion and potentiation varied from experiment to experiment; part of this delay may be attributable to diffusion of CaMKII* (33 kDa) to the synapses being stimulated.
; Tang et al. 1991; Wyllie et al. 1993); in this study, we assume that effects seen at somatic receptors will have analogous effects at synaptic receptors. We also assume that the observed changes of responses to kainate will be accompanied by similar changes of responses to endogenous transmitter. Two advantages of measuring responses to somatic application of agonist are that we have complete control over the distance between the application pipette and the response site and that diffusion of intrapipette perfusate to the cytoplasm underlying the response site is more rapid. Note that our application method is quite sensitive to distance. In fact, we were unable to evoke detectable responses with 100 µM kainate application unless the application pipette was within 10 µm of the cell membrane. Diffusion of kainate to greater distances (e.g., the dendrites) is thus unlikely to contribute significantly to the measured response. The greater magnitude of potentiation of somatic responses compared with synaptic responses might be due to the greater concentration of perfused CaMKII* in the soma compared with dendritic arbors and spines. Another possibility is that somatic AMPA/kainate receptors may be naive to CaMKII because CaMKII is found in the highest concentration at the PSD (Kelly et al. 1984; Kennedy et al. 1983) and are therefore more sensitive to its effect.
; Kullmann et al. 1996). The finding that CaMKII* specifically potentiated responses to exogenous application of a non-NMDA agonist is consistent with the idea that plasticity mechanisms are specific to receptor subtype. This result is also consistent with the finding that enhancement of kinase activity in CA1 cells by inhibiting phosphatases also selectively potentiated non-NMDA receptor-mediated responses (Figurov et al. 1993). Although we found that active CaMKII* had no specific effect on extra-synaptic NMDA responses, Kolaj et al. (1994) found that NMDA responses in dorsal horn neurons potentiated when activated CaMKII was included in patch solution. Different NMDA receptor subunit compositions in dorsal horn neurons and hippocampal CA1 neurons may explain this difference. In addition, Kolaj et al. did not use internal perfusion; as a result, they could not have easily observed the rapid decrease of NMDA responsiveness immediately after application of kinase solution. The rapid, kinase-independent decrease seen of NMDA responses preceded appearance of RITC-dextran (40 kDa). This effect may have been mediated by calmodulin (10 kDa), a constituent of both test and control perfusion solutions, which can down-regulate directly NMDA responses (Ehlers et al. 1996). Elucidation of this effect will require further studies. If the factor mediating this nonspecific effect can be found and excluded, the effect of CaMKII* on synaptic NMDA responses could be measured directly.
) and the increase in noise produced by the perfusion canula within the patch pipette. As an alternative, the time course and trial-to-trial variability of whole cell currents were studied as a method of examining single-channel properties (Hessler et al. 1993; Li et al. 1993; Lu et al. 1993; Numann et al. 1991; Tang et al. 1991). The finding that CaMKII*-mediated potentiation does not change kainate response time course parallels the finding that the time course of non-NMDA synaptic responses does not change during LTP (O'Connor et al. 1995). Analysis of trial-to-trial variability showed that the CV2 increased with potentiation; by itself, this finding is consistent with a change in channel number or open probability.
; McGlade-McCulloch et al. 1994; Tan et al. 1994), in the PSD as well as in vitro, supporting the hypothesis that CaMKII acts directly on AMPA/kainate receptors to potentiate responsiveness during LTP.
). A second hypothesis is that other rate constants of the channel change with CaMKII activity. One specific change might be a decrease in the desensitization rate. Experiments with cyclothiazide demonstrate that decreasing desensitization and/or deactivation results in a slowing of time course of the kainate-evoked responses. CaMKII activation, in contrast, potentiates responses without changing time course, implying that CaMKII does not potentiate transmission by reducing desensitization as cyclothiazide does. It is possible that a kinetic change distinct from that mediating the cyclothiazide effect results in increased open probability; kinetic changes that increase response size without affecting the macroscopic response time course are not ruled out by our results. Finally, the hypothesis that the single-channel conductance of receptors increases with CaMKII activity would predict an increase in the mean response with no change in the nonstationary CV; our finding that CV decreases is evidence against this case.
).
; Raymond et al. 1993; Swope et al. 1992). Phosphorylation studies show that GluR1 is phosphorylated by CaMKII (Hayashi et al. 1997; McGlade-McCulloch et al. 1993; Tan et al. 1994). Activation of previously silent channels may occur by changing open probability from 0 to a nonzero value or increasing conductance from 0 to a nonzero value. In addition, phosphorylation has been shown to induce acetylcholine receptor assembly (Green et al. 1991) as well as to cause active translocation of glucose transporters to the cell surface (Haruta et al. 1995; Pessin and Bell 1992). The phosphorylation state of synapsin I influences targeting of presynaptic vesicles and is regulated by CaMKII (Llinas et al. 1991). The phosphorylation state of the NR1 subunit of the NMDA receptor also has been shown to regulate its localization (Ehlers et al. 1995). CaMKII phosphorylation also may regulate targeting of AMPA/kainate receptors to the cell surface (Malinow 1995). Recent work in our laboratory demonstrates calcium-evoked exocytosis of dendritic vesicles that is CaMKII dependent (M. Maletic-Savatic, T. Koothan, and R. Malinow, unpublished observations). The complete resolution of this problem is of great importance to an understanding of synaptic physiology.
ACKNOWLEDGEMENTS
FOOTNOTES
REFERENCES
Abstract
Introduction
Methods
Results
Discussion
References
subunit.
Neuron
7: 659-666, 1991.[Medline]-calcium-calmodulin kinase II mutant mice.
Science
257: 201-206, 1992.
0022-3077/97 $5.00 Copyright ©1997 The American Physiological Society
This article has been cited by other articles:
|
|
 |
|
  M. Sanhueza, C. C. McIntyre, and J. E. Lisman Reversal of Synaptic Memory by Ca2+/Calmodulin-Dependent Protein Kinase II Inhibitor J. Neurosci., May 9, 2007; 27(19): 5190 - 5199. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Serrano, Y. Yao, and T. C. Sacktor Persistent Phosphorylation by Protein Kinase M{zeta} Maintains Late-Phase Long-Term Potentiation J. Neurosci., February 23, 2005; 25(8): 1979 - 1984. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. K. Andrasfalvy and J. C. Magee Changes in AMPA receptor currents following LTP induction on rat CA1 pyramidal neurones J. Physiol., September 1, 2004; 559(2): 543 - 554. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H.-W. Yang, X.-D. Hu, H.-M. Zhang, W.-J. Xin, M.-T. Li, T. Zhang, L.-J. Zhou, and X.-G. Liu Roles of CaMKII, PKA, and PKC in the Induction and Maintenance of LTP of C-Fiber-Evoked Field Potentials in Rat Spinal Dorsal Horn J Neurophysiol, March 1, 2004; 91(3): 1122 - 1133. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. L. Hinds, I. Goussakov, K. Nakazawa, S. Tonegawa, and V. Y. Bolshakov Essential function of alpha -calcium/calmodulin-dependent protein kinase II in neurotransmitter release at a glutamatergic central synapse PNAS, April 1, 2003; 100(7): 4275 - 4280. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. C. Poncer, J. A. Esteban, and R. Malinow Multiple Mechanisms for the Potentiation of AMPA Receptor-Mediated Transmission by alpha -Ca2+/Calmodulin-Dependent Protein Kinase II J. Neurosci., June 1, 2002; 22(11): 4406 - 4411. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H.-X. Chen, N. Otmakhov, S. Strack, R. J. Colbran, and J. E. Lisman Is Persistent Activity of Calcium/Calmodulin-Dependent Kinase Required for the Maintenance of LTP? J Neurophysiol, April 1, 2001; 85(4): 1368 - 1376. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Hayashi, S. Shi, J. A. Esteban, A. Piccini, J. Poncer, and R. Malinow Driving AMPA Receptors into Synapses by LTP and CaMKII: Requirement for GluR1 and PDZ Domain Interaction Science, March 24, 2000; 287(5461): 2262 - 2267. [Abstract] [Full Text]
|
|
|
|
|
|
C Stricker, A I Cowan, A. Field, and S J Redman Analysis of NMDA-independent long-term potentiation induced at CA3--CA1 synapses in rat hippocampus in vitro J. Physiol., October 15, 1999; 520(2): 513 - 525. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
V. Derkach, A. Barria, and T. R. Soderling Ca2+/calmodulin-kinase II enhances channel conductance of alpha -amino-3-hydroxy-5-methyl-4-isoxazolepropionate type glutamate receptors PNAS, March 16, 1999; 96(6): 3269 - 3274. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. C. Carroll, R. A. Nicoll, and R. C. Malenka Effects of PKA and PKC on Miniature Excitatory Postsynaptic Currents in CA1 Pyramidal Cells J Neurophysiol, November 1, 1998; 80(5): 2797 - 2800. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Rumpel, H. Hatt, and K. Gottmann Silent Synapses in the Developing Rat Visual Cortex: Evidence for Postsynaptic Expression of Synaptic Plasticity J. Neurosci., November 1, 1998; 18(21): 8863 - 8874. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Maletic-Savatic, T. Koothan, and R. Malinow Calcium-Evoked Dendritic Exocytosis in Cultured Hippocampal Neurons. Part II: Mediation by Calcium/Calmodulin-Dependent Protein Kinase II J. Neurosci., September 1, 1998; 18(17): 6814 - 6821. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
) or inactive (