HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) All Versions of this Article: 93/3/1486    most recent 00958.2004v2 00958.2004v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (4) Citing Articles via Google Scholar Google Scholar Articles by Cohen, J. D. Articles by Castro-Alamancos, M. A. Search for Related Content PubMed PubMed Citation Articles by Cohen, J. D. Articles by Castro-Alamancos, M. A.

Skilled Motor Learning Does Not Enhance Long-Term Depression in the Motor Cortex In Vivo

Department of Neurobiology and Anatomy, Drexel University College of Medicine, Philadelphia, Pennsylvania

Submitted 15 September 2004; accepted in final form 21 October 2004

	ABSTRACT

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION GRANTS REFERENCES

 
Learning of motor skills may occur as a consequence of changes in the efficacy of synaptic connections in the primary motor cortex. We investigated if learning in a reaching task affects the excitability, short-term plasticity, and long-term plasticity of horizontal connections in layers II–III of the motor cortex. Because training in this task requires animals to be food-deprived, we compared the trained animals with similarly food-deprived untrained animals and normal controls. The results show that the excitability, short-term plasticity, and long-term plasticity of the studied horizontal connections were unaffected by motor learning. However, stress-related effects produced by food deprivation and handling significantly enhanced the expression of long-term depression in these pathways. These results are compatible with the hypothesis that the acquisition of a complex motor skill produces bi-directional changes in synaptic strength that are distributed throughout the complex neural networks of motor cortex, which remains synaptically balanced during learning. The results are incompatible with the idea that learning causes large unidirectional changes in the population response of these neural networks, which may occur instead during certain behavioral states, such as stress.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION GRANTS REFERENCES

 
Skilled motor movements rely heavily on the primary motor cortex for their accurate expression (Lawrence and Kuypers 1968; Schieber and Poliakov 1998). Lesions of the primary motor cortex of rodents dramatically impair the performance of a forelimb reaching task that requires rodents to reach and grasp food pellets (Castro 1977; Castro-Alamancos and Borrell 1993; Whishaw et al. 1991). Also, reaching in this task correlates with cell activity in the primary motor cortex contralateral to the trained forelimb (Dolbakyan et al. 1977; Kargo and Nitz 2004).

A long-held hypothesis is that skilled motor learning occurs as a consequence of changes in the efficacy of synaptic connections in the primary motor cortex (Hebb 1949; Ramon y Cajal 1904). Indeed, Ramon y Cajal noted around a century ago that pianists surely have more complex synaptic networks in motor cortex as a result of activity-dependent mechanisms. In agreement with this hypothesis, studies have shown, for example, that learning of a forelimb reaching task modifies the dendritic density in the motor cortex (Bury and Jones 2002; Greenough et al. 1985; Withers and Greenough 1989), synaptogenesis and the size of the forelimb region in primary motor cortex derived using microstimulation mapping techniques (Kleim et al. 2004), and the signal-to-noise ratios of motor cortex cell activity (Kargo and Nitz 2004). In addition, this hypothesis has propelled a parallel line of studies devoted to the characterization of the different forms of synaptic plasticity (Bear and Malenka 1994; Bliss and Collingridge 1993; Nicoll and Malenka 1995) that are expressed in the primary motor cortex. Those studies have found that both long-term potentiation (LTP) and long-term depression (LTD) can be generated in pathways of the motor cortex under the appropriate conditions, both in vivo and in vitro (Aroniadou and Keller 1995; Castro-Alamancos and Connors 1996; Castro-Alamancos et al. 1995; Hess and Donoghue 1996; Hess et al. 1996; Iriki et al. 1989; Keller et al. 1990).

This large body of evidence strongly suggests that the forelimb region within the primary motor cortex undergoes changes associated with the acquisition of a forelimb motor skill, such as the forelimb reaching task. A particularly intriguing possibility is that the extensive horizontal connections in the upper layers of the motor cortex can undergo changes in synaptic efficacy that underlie the acquisition of a skilled motor behavior (Asanuma and Pavlides 1997; Hess and Donoghue 1994; Keller 1993). Therefore, if learning occurs through changes in synaptic strength of these connections, learning of a motor skill should change the efficacy of motor cortex horizontal fibers. Indeed, recent work in the slice preparation supports this hypothesis (Rioult-Pedotti et al. 1998, 2000). These studies presented evidence that learning of a reaching task was associated with a striking global increase in the amplitude of field potential population responses evoked in layers II–III by stimulating horizontal fibers as determined using input/output (I/O) curves, and this increase occurred only in the forelimb motor cortex contralateral to the trained forelimb. Moreover, the selective increase in response amplitude in the trained hemisphere was accompanied by a reduction in the amount of LTP that could be induced by theta-burst stimulation (Rioult-Pedotti et al. 1998) as well as a corresponding increase in the expression of LTD (Rioult-Pedotti et al. 2000). These data support the idea that learning of a skilled motor behavior involving the primary motor cortex occurs by an increase in the efficacy of the population of horizontal connections in layers II–III of the motor cortex. However, the large unidirectional change in synaptic efficacy of the population response reported in these studies is difficult to interpret because it could collapse the uneasy balance of excitation in cortical networks and perhaps lead to hyperexcitation (e.g., seizures), something that is not expected to occur with motor learning. Instead, the acquisition of a complex motor skill might be expected to produce both increases and decreases in synaptic strength that are distributed throughout the complex neural networks within layers II–III. These bi-directional distributed changes would be unlikely to cause large unidirectional changes in the population response of these neural networks. Indeed, an intrinsic property of neural networks in modeling studies is that the global state (i.e., sum of total synaptic weights) of the network must remain balanced and is thereby stable during learning (Buhmann and Schulten 1986; Song et al. 2000).

In this study, conducted in vivo, we sought to determine if learning in the unilateral reaching task indeed affects the excitability, short-term plasticity, and long-term plasticity of horizontal connections in layers II–III of the motor cortex. Because training in this task requires animals to be food-deprived, we compared the trained animals with similarly food-deprived untrained animals and normal controls. The results show that the excitability, short-term plasticity, and long-term plasticity of the studied horizontal connections were unaffected by motor learning. However, food deprivation and handling significantly enhanced the expression of LTD in these pathways.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION GRANTS REFERENCES

 
Behavioral training

Fifty-five adult male Sprague-Dawley rats (225–300 g) were used in this study and cared for in accordance with National Institutes of Health guidelines for laboratory animal welfare. All experiments were approved by the Drexel University Institutional Animal Care and Use Committee.

Three groups of rats were used in this study: control, trained, and deprived. Animals in the control group were group housed and were not food deprived or handled. The trained group was food deprived and trained in a skilled reaching task. The deprived group was treated as the trained group, but these animals were not trained in the skilled reaching task. That is, although they were food deprived and placed in the training box, there were no food pellets available, and no skill was learned. Thus the deprived group serves to control for stress-related variables such as food deprivation and handling.

Food deprivation consisted of placing the animals on a food-restricted diet sufficient in maintaining their body weight at 85% free-feeding weight. Water was provided ad libitum. During training in the skilled reaching task, animals were placed in a clear Plexiglas box (30 x 15 x 12 cm) with a 1-cm slit in one of its walls. The animals learned to reach through the slit with one forelimb and retrieve a small food pellet (45 mg, Noyes Precision Food Pellets) located on a platform outside the box. Only animals that solely used one forelimb during the training process were selected for further experimentation. Two groups of trained animals were studied. One group was trained 1 h/day for 5–7 days, and another group was trained 30 min/day for 11–13 days. Performance in the reaching task was evaluated by counting misses and successful reaches for all the attempts made by the animal during the entire session. A successful reach consisted of the animal reaching with the limb through the slit of the box, grasping the food pellet, and consuming it. A miss consisted in reaching through the slit of the box and failing to grasp the food pellet or to consume it. If the animal reached and failed to grasp the pellet with the initial movement but eventually managed to grasp the food pellet by a subsequent movement (without retracting the limb), this was considered a miss. Percent success rate was the number of successes divided by the number of attempts. The day after the last training or handling session, the animals were subjected to surgical procedures.

Surgical procedures

All the animals from the control, deprived, and trained groups were henceforth subjected to the same procedures. Animals were anesthetized with urethane (1.5 g/kg, ip) and placed in a stereotaxic frame. All skin incisions and frame contacts with the skin were injected with lidocaine (2%). Body temperature was automatically maintained constant with a heating pad. The level of anesthesia was carefully monitored by continuously recording field potentials from the cortical recording electrodes and by testing limb-withdrawal reflexes. The cortical field potential activity was also displayed on-line as power spectrums derived by calculating fast-Fourier transforms (FFT), and this activity was stored for subsequent reference during data analyses. The anesthetic level was kept constant at approximately stage III/3 using supplemental doses of urethane (Friedberg et al. 1999). Small bilateral craniotomies were made over the forelimb primary motor cortex based on coordinates derived previously with microstimulation mapping (Castro-Alamancos and Borrell 1993). Although the extent of these representations can change between animals, the location of forelimb motor representations is constant and centered at around 2.5 mm lateral and 1 mm anterior to bregma. Small incisions were made in the dura to insert the electrodes, and the cortical surface was covered with small pieces of gelfoam soaked in saline solution. At the end of the experiments, animals were killed with an overdose of sodium pentobarbitone (intraperitoneally).

Electrophysiological procedures

All electrophysiological procedures were conducted with the experimental group blinded. That is, the experimenter conducting the electrophysiology was unaware of the experimental group that the animal belonged to. A total of four electrodes was used simultaneously per animal: two stimulating electrodes and two recording electrodes. In each hemisphere, a recording electrode and a stimulating electrode were inserted into layers II–III of the forelimb primary motor cortex representation, as shown in Fig. 1A. Each electrode was lowered 500 µm below the pial surface at a 60° angle. The tips of the stimulating and recording electrodes faced each other at the pial surface and began their insertion at about 1 mm apart so that they would be separated about 500 µm tip-to-tip when they reached layers II–III. The mid-point coordinate between the stimulating and recording electrodes was 2.5 mm lateral and 1 mm anterior to bregma. Minor adjustments were made in the depth of the stimulating and recording electrodes to produce the largest evoked response. The electrode arrangement allowed monitoring the efficacy of horizontal pathways in the upper layers of primary forelimb motor cortex simultaneously from each hemisphere.

View larger version (18K): [in this window] [in a new window]   FIG. 1. Schematic representation of the experimental setup and synaptic nature of field potential responses. A: location of the stimulating and recording electrodes in layers II–III of the forelimb motor cortex of 1 hemisphere. Note that in all experiments, electrodes were implanted simultaneously in both hemispheres. B: field potential responses and effect of application of a glutamate receptor antagonist (CNQX; 500 µM) with a microdialysis probe placed adjacent to the recording electrode. Application of CNQX abolished the negative field potential response that peaks between 3–5 ms poststimulus and revealed a fiber volley (i.e., antidromic and orthodromic) component in the field potential recording, which peaks between 2 and 3 ms. Abolishment of the field potential response over a 3-ms latency confirms the synaptic nature of the response measured in this study.

During baseline recordings, stimulation alternated between each hemisphere at 0.088 Hz. After an 1-h stable baseline recording period was established, I/O curves were derived by averaging 12 responses at each of seven current intensities; 0.5, 1, 2, 3, 4, 5, and 6 times the current required to elicit a field potential response of 1 mV in amplitude (maximum stimulation intensity did not exceed 300 µA). Thereafter, a minimum 30-min stable baseline was established at an intensity that elicited a half-maximal evoked response (average intensity used was 50 µA; range, 25–110 µA). LTD was induced by low-frequency stimulation (LFS) consisting of 1,800 pulses at 2 Hz at twice the baseline stimulation intensity and was applied at 30-min intervals until the intracortical pathway appeared saturated (Rioult-Pedotti et al. 2000). Saturation was defined as no significant changes in field potential amplitude after two successive LTD stimulations. This typically occurred after the LTD stimulation was delivered four to five times. Following the last LTD stimulation, responses were recorded for a minimum of 30 min. Short-term plasticity (i.e., pair-pulse stimulation) was also monitored before and after LTD stimulation using three pulses delivered at different interstimulus intervals (ISIs) of 50, 150, 250, and 350 ms. Finally, LTP induction was attempted after LTD induction using a standard theta-burst stimulation (TBS) protocol at twice the baseline stimulation intensity and consisted of 10 bursts (burst = 5 pulses at 100 Hz) delivered at 5 Hz. A total of five TBSs was delivered at 10-s intervals.

Statistical analysis

Although three groups of animals were generated experimentally (trained, deprived, and control), the statistical analyses were performed on fours groups of data, corresponding to the trained hemispheres (i.e., contralateral to the trained forelimb), untrained hemispheres (i.e., ipsilateral to the trained forelimb), deprived hemispheres (i.e., data from both hemispheres for each animal were grouped), and control hemispheres (i.e., data from both hemispheres were also grouped). Grouping of the data from both hemispheres in the deprived and control groups was performed after comparing the left versus right hemispheres for those groups, which revealed no significant differences.

The statistical analysis consisted primarily of repeated measures ANOVAs, where the within-subjects factor was either training day, current intensity for the I/O curves, response amplitude before and after LFS or TBS, or amount of facilitation before and after LTD. The between-subjects factor consisted of the different groups of hemispheres as outlined below. Simple effects were used to decompose significant main effects and interactions. Post hoc comparisons were performed with the Scheffe test. Individual comparisons between groups were performed with a one-factor ANOVA. All data are presented as mean ± SD.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION GRANTS REFERENCES

 
Motor skill learning

Animals were trained in a skilled reaching task for either 5–7 (n = 18 animals; n = 8 left-handed, n = 10 right-handed) or 11–13 days (n = 7 animals; n = 3 left-handed, n = 4 right-handed). Animals rapidly acquired the skilled reaching behavior during the extensive daily training sessions. Figure 2 shows the percent success rate for both of these groups of animals per day of training. By days 5–6 of training, the animals in both groups reached asymptotic success rates of about 65%, which were significantly higher than the first day. Thus, a repeated measures ANOVA revealed a significant effect of training on success rate for the group trained for 5–7 days [F(4,76) = 230.045; P < 0.001] and for the group trained for 11–13 days [F(10,70) = 92.412; P < 0.001]. Post hoc comparisons revealed that success rate in the task was significantly improved after the second day of training compared with the first day of training (P < 0.001). A comparison of the success rate during the final training day for the animals trained for 5–7 days and those trained for 11–13 days revealed no significant differences [F(1,26) = 1.1; P = 0.300]. Thus, since there were no apparent differences in the reaching behavior success rate between the animals trained for 5–7 days and those trained for 11–13 days, electrophysiological data for the trained animals were grouped together unless otherwise indicated.

View larger version (14K): [in this window] [in a new window]   FIG. 2. Effect of training in the reaching task on success rate. Percent success rate per day in a group of animals trained for 5–7 days and in a group of animals trained for 11–13 days in a reaching task. Success rate was calculated as the number of successful reaches divided by the number of attempts. A successful reach consisted of reaching, grasping a food pellet, and consuming it.

Twenty-four hours after the last training session, animals underwent stereotaxic surgery to compare the excitability of horizontal pathways in layers II–III of motor cortex by deriving I/O curves across a range of seven current intensities (i.e., 0.5, 1, 2, 3, 4, 5, 6 times the intensity required to evoke a 1-mV response). The I/O curve plots the amplitude of the field potential–evoked response as a function of current intensity. Figure 3A shows an example of this procedure in a trained animal for both the trained and the untrained hemispheres. Note that the amplitude of the evoked response increases with current intensity as a consequence of the recruitment of more fibers with increases in current spread. This procedure was performed for each of the four groups of hemispheres studied: trained hemispheres (n = 23), untrained hemispheres (n = 24), deprived hemispheres (n = 19), and control hemispheres (n = 36). Since the control hemispheres group and the deprived hemispheres group were formed by combining the left and right hemispheres of those animals, we first compared the left and right hemispheres of the control and deprived animals to assure that they were not significantly different. Note that in the statistical analysis, the 1-times intensity values were excluded because the field potential amplitudes are set at 1 mV for each group. Figure 3B shows I/O curves for the left and right hemispheres of the control and deprived animals. A two-factor repeated measures ANOVA of the amplitude of the evoked responses revealed that there were no significant differences between the left and right hemispheres for either the control animals [F(1,34) = 0.486; P = 0.491] or the deprived animals [F(1,17) = 0.172; P = 0.684]. Thus the left and right hemispheres of these animals could be combined to form the control hemispheres group and the deprived hemispheres group. An additional comparison was made between the trained and untrained hemispheres of the trained animals to test if motor skill training affected the excitability of the horizontal motor cortex pathways contralateral to the trained forelimb. Figure 3C shows the I/O curves for these animals. A two-factor repeated measures ANOVA revealed no significant differences between the trained and untrained hemispheres of the trained animals [F(1,45) = 0.341; P = 0.562], indicating that the tested motor cortex pathways ipsilateral and contralateral to the trained forelimb did not differ in excitability. Finally, to test if there were significant differences between the I/O curves derived from the four groups of hemispheres taken together (Fig. 3C, right), we performed an analysis to compare them. A two-factor repeated measures ANOVA revealed an obvious significant effect of intensity on the amplitude of the evoked response [within-subjects factor; F(5,490) = 2,142.1; P < 0.001], but no significant differences between the four groups of hemispheres [between-subjects factor; F(3,98) = 1.8; P = 0.14] and no interaction between these two factors [F(15,490) = 1.4; P = 0.11]. This indicates that neither motor learning nor food deprivation and handling affected the excitability of primary motor cortex pathways as assessed using I/O curves.

View larger version (34K): [in this window] [in a new window]   FIG. 3. Input/output (I/O) curves reveal no differences between the groups. A: examples of field potential responses used to derive I/O curves from trained and untrained hemispheres of a trained animal. I/O curves are derived across a range of 7 current intensities (i.e., 0.5, 1, 2, 3, 4, 5, and 6 times the intensity required to evoke a 1-mV response). Note that amplitudes of evoked responses increase with intensity similarly in the trained and untrained hemisphere. B: I/O curves for left and right hemispheres of control and deprived animals. Amplitudes of left and right hemispheres did not differ significantly in either control or deprived animals. C: I/O curves for trained and untrained hemispheres of trained animals (left). Amplitudes of trained and untrained hemispheres did not differ significantly, and these did not differ significantly from deprived and control hemispheres (right). Statistical analyses are described in RESULTS.

LTD is enhanced by food deprivation and handling but not by skilled motor learning

The next step was to investigate if there were differences between the different groups of hemispheres in their ability to undergo LTD. For example, one possibility is that motor skill learning may enhance the ability of horizontal pathways to undergo LTD. To test this hypothesis, we first established the conditions that allow to record stable baselines in urethane-anesthetized rats for extended periods of time. Thus in a group of control animals (n = 5), we found that we could maintain stable baselines for 3 h as long as we kept constant the slow spontaneous oscillatory activity that occurs in neocortex under this anesthesia (Fig. 4A). This activity was measured on-line by computing FFTs and deriving power spectrum analyses of the ongoing field potentials recorded from the electrodes implanted in each hemisphere that are used to obtain the evoked responses (Fig. 4, A and B). The FFT power spectrums were carefully monitored during the experiment and stored together with the evoked responses for further comparison off-line. Figure 4A shows an example of such a control experiment and displays both the power spectrum analysis and the amplitude of evoked responses for an extended period of time. Regardless of the group, we established that, for an experiment to be included in the dataset, the power spectrum of the slow oscillatory activity (0.1–4 Hz) should not vary >40% between the baseline period and 30 min after the last LTD stimulation (3 h later). Animals that did not meet this criterion were removed from the LTD analysis unless indicated. Figure 4B shows an example of the comparison of spontaneous field potential activity and the power spectrums derived using FFTs obtained during a period of baseline before LFS and after the induction of LTD (i.e., 30 min after the last LFS). Although significant LTD was induced, the power spectrum activity was not significantly affected, and thus the data were considered acceptable. Note that because I/O curves were generated immediately after an initial baseline period where stable stimulus-evoked responses were identified, all animals were included in the I/O curve analysis described in the previous section.

View larger version (34K): [in this window] [in a new window]   FIG. 4. Example of long-term depression (LTD) induced by low-frequency stimulation (LFS) in trained and untrained hemispheres of a trained animal. A: example of a stable baseline maintained for 3 h during simultaneous monitoring of the spontaneous oscillatory activity through the same recording electrode by computing fast-Fourier transforms (FFTs) and deriving a power spectrum of the ongoing activity. Increases in power for a particular frequency are represented as hot colors (red), and blue represents 0. B: comparison of spontaneous field potential activity and power spectrums derived using FFTs obtained during a period of baseline before LFS and after the induction of LTD (i.e., 30 min after the last LFS). Note that the power spectrums did not change significantly before and after LTD. C: example of LTD induced by the application of LFS 4 times at 30-min intervals in trained and untrained hemispheres of a trained animal. Note a reduction in amplitude of evoked responses as a consequence of LFS, which was similar in both hemispheres.

View larger version (23K): [in this window] [in a new window]   FIG. 5. Population data of LTD in horizontal pathways of layers II–III in motor cortex. A: LTD induced in control, deprived, trained, and untrained hemispheres after the last LFS stimulation compared with baseline before the 1st LFS. Note that control hemispheres produced less LTD than other hemispheres. B: LTD induced 30 min after the last LFS in the 4 groups of hemispheres. Note that control hemispheres produced significantly less LTD than other hemispheres (*P < 0.01). Data are mean ± SD.

LTD is associated with a significant change in short-term plasticity

The previous results showed that significant LTD can be evoked in horizontal pathways of the motor cortex in vivo and that control animals produce significantly less LTD than the other animals. In an effort to characterize further LTD in this pathway, we evaluated the effect of LTD on short-term plasticity. In pathways where LTP and LTD are believed to be expressed by a change in neurotransmitter release probability, there are consistent effects on short-term plasticity, such as facilitation and depression (Manabe and Nicoll 1994; Weisskopf and Nicoll 1995; Zucker 1989). By first evaluating short-term plasticity in naïve pathways before the induction of LTD, we were able to study if motor learning or food deprivation and handling had any effect on short-term plasticity in these pathways. We found that at the 50-ms ISI, horizontal pathways of the motor cortex consistently displayed facilitation (Fig. 6). Interestingly, before the induction of LTD, facilitation appeared to be stronger in control hemispheres than in the deprived, trained, and untrained hemispheres, but this difference was not statistically significant [1-factor ANOVA; F(3,57) = 1.2; P = 0.31]. Thus these results indicate that, although there is a tendency for food deprivation and handling to reduce the amount of short-term facilitation in horizontal pathways of layers II–III, this is not statistically significant.

View larger version (28K): [in this window] [in a new window]   FIG. 6. Effect of LTD on short-term plasticity. A: ratio of amplitude of the field potential (FP) response between the 3rd and 1st responses in a train delivered at 4 different interstimulus intervals (ISIs) for control, deprived, trained, and untrained hemispheres. Note that at 50-ms ISI, responses show facilitation and there were no significant differences between the groups. B: effect of LTD on facilitation measured at 50-ms ISI. Facilitation was measured as in A, before and after the induction of LTD (measured 30 min after the last LFS). Note that facilitation was significantly reduced by LTD in control hemispheres and significantly enhanced in trained and untrained hemispheres (*significant difference between before and after LTD). See RESULTS for statistical analyses. Data are mean ± SD. C: sample traces showing facilitation at 50-ms ISI (3 stimulus train) of an untrained hemisphere before (black) and after (light gray) induction of LTD with LFS. D: traces shown in C are overlaid to reveal changes in facilitation after LTD. Left: responses to the 1st stimulus scaled so that they match each other in amplitude. Right: same scaling applied to the 3rd responses, which reveals an enhancement in facilitation after LTD (light gray).

LTP is not induced

Preliminary experiments conducted in naïve pathways of control animals revealed that application of theta-burst stimulation was ineffective in inducing LTP in horizontal pathways of the motor cortex in vivo. This is in agreement with studies in slices, which have shown that LTP is not generally induced in these pathways unless inhibition is suppressed (Castro-Alamancos et al. 1995; Hess et al. 1996; Rioult-Pedotti et al. 1998). We did not attempt to study LTP during the suppression of inhibition in vivo because of the dramatic consequences on spontaneous activity that disinhibition produces in the motor cortex in vivo (Castro-Alamancos 2000). Instead, we attempted to evaluate if, after the induction of LTD, there was any significant difference in the ability of the different groups to undergo LTP. A two-factor repeated measures ANOVA of the amplitude of the evoked responses before and 30 min after LTP stimulation revealed no significant effect of the LTP stimulation [F(1,62) = 0.04; P = 0.825] and no significant group effect [F(3,62) = 1.034; P = 0.384] or interaction [F(3,62) = 1.802; P = 0.156]. The results indicate that, under these conditions (i.e., without suppressing inhibition), LTP is not induced in the horizontal pathways of layers II–III in any of the four groups of hemispheres studied.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION GRANTS REFERENCES

 
In this study, we assessed the effects of learning a skilled reaching behavior and the stress-related effects of food deprivation and handling on the excitability, short-term plasticity, and long-term plasticity of horizontal pathways in layers II–III of motor cortex. Neither skilled motor learning nor stress-related variables affected the excitability of horizontal pathways of the motor cortex as evaluated using I/O curves. Interestingly, stress-related variables, such as food deprivation and handling, strongly enhanced LTD in these pathways, whereas skilled motor learning had no effect on LTD. In addition, the ability to induce LTP, which was absent in control animals, was not affected by food deprivation and handling or by skilled motor learning. Finally, neither skilled motor learning nor stress-related variables affected the short-term plasticity of these pathways, which display slight facilitation in response to pulses delivered at 50-ms ISIs. However, the expression of LTD was accompanied by a tendency for changes in facilitation that differed between the experimental groups. While these effects were small, they are worth mentioning. In control animals, the expression of LTD is accompanied by a reduction of facilitation. Food deprivation and handing eliminated this reduction of facilitation caused by LTD expression that is observed in control animals. Moreover, this effect of food deprivation and handling was augmented by motor training since the expression of LTD is accompanied by an increase of facilitation in trained animals. Notably, in this study, we never found a difference between the trained and untrained hemispheres of trained animals, indicating that skilled motor learning did not affect the excitability, short-term plasticity, or long-term plasticity of population responses in horizontal pathways of layers II–III of the forelimb primary motor cortex. In contrast, stress-related variables, such as food deprivation and handling, did enhance the ability of these pathways to undergo LTD.

Data from this study do not support the hypothesis that skilled motor learning leads to a global increase in the synaptic efficacy of a synchronously activated neuronal population in horizontal networks of layers II–III of motor cortex. Thus the data presented here contrast with those previously reported (Rioult-Pedotti et al. 1998, 2000), which found that the excitability and LTD of horizontal pathways of motor cortex changed as a consequence of skilled motor learning. There are some differences between our study and the previous studies that may explain the discrepancies. For example, although this study and the previous slice studies used adult animals, the reported weight of the animals is higher in this study, suggesting that the slice work was carried out in younger animals, which may contribute to the differences observed. Clearly, the most obvious difference is that our study was conducted in anesthetized animals in vivo, while the previous studies were done in slices. This may well have been a contributing factor, which we cannot discard. In fact, one possibility is that the effects of learning cannot be detected because they are masked by the anesthetic used in vivo. However, if the anesthetic was having a major effect, it would likely have affected the induction of LTD, but we found that LTD was induced in urethane-anesthetized animals at similar levels than in slices. Also, in previous studies that were conducted simultaneously in vivo using urethane anesthesia, and in vitro using slices, the response characteristics of synaptic pathways and their plasticity were found to be quite similar in vivo and in slices (Castro-Alamancos 2002a, b; Castro-Alamancos and Calcagnotto 2001). Furthermore, many characteristics of the responses recorded in this study in vivo are identical to those recorded in our laboratory in slices in the same pathways. For example, the shape, amplitude, excitability, and short-term plasticity of field potential responses in layers II–III horizontal pathways in vivo and in slices are indistinguishable (Castro-Alamancos et al. 1995). Since LTD is expressed in vivo and in slices similarly, it would be expected that its modulation by learning should also be expressed similarly under both conditions. In fact, in this study, LTD induction is clearly modulated by food deprivation and handling despite the anesthesia. In any case, as indicated above, this is a major difference that cannot be discarded and may explain why the previous slice studies found an effect of learning while this study did not. Another related question is why did the slice studies not reveal an effect of food deprivation and handling on LTD? The paper by Rioult-Pedotti et al. (1998) consisted of naïve controls (n = 12) and food-deprived controls (n = 8), and these animals were mixed together to form one control group (n = 20). Moreover, the paper by Rioult-Pedotti et al. (2000) did not include naïve controls. To reveal the effect of food deprivation, it is necessary to compare naïve controls with food-deprived animals, which was never done in those studies. Finally, it is important to note that the fact that we observed an effect of food deprivation and handling on LTD induction does not explain the discrepancy between the previous studies in slices and this study on the effect of motor learning. That is, it is unclear why we found no differences between the trained and untrained hemispheres while the slice studies did. In conclusion, at least in vivo, skilled motor learning in a reaching task does not enhance the induction of LTD in motor cortex pathways, while food deprivation and handling does.

How would food deprivation change the ability of horizontal pathways to undergo LTD? An interesting finding from our study is that the food-deprived and handled animals consistently expressed more LTD than the unmanipulated control animals, and this effect was not further affected by motor learning. It is important to consider that food deprivation, and the consequent loss of 20% of body weight, is a strong stressor comparable to, for example, electric footshocks (Carlson et al. 1987). In fact, food deprivation per se strongly enhances the activity of the hypothalamus-pituitary-adrenal axis and also increases the levels of several neuromodulators such as norepinephrine and acetylcholine, among others (El Fazaa et al. 2000; Endou et al. 2001; Heiderstadt et al. 2000; Kiss et al. 1994; Savard et al. 1983). That food-deprived animals are stressed is apparent when an investigator handles these animals; deprived animals are hyperactive and more excitable (unpublished observations; Abraham and Gogate 1989; Endou et al. 2001). In addition, handling animals and placing them in novel environments can also lead to stress-related changes (Davis et al. 2004; Enrico et al. 1998; Feenstra et al. 1998; Kawahara et al. 2000), although these changes tend to decline with repeated exposures, and handled animals are not hyperactive and as excitable as food-deprived animals. In this study, it cannot be ascertained which of these factors (i.e., food deprivation, handling) was the major contributor to the observed effects. We speculate that food deprivation was the more relevant of the two factors. Thus the effects of food deprivation and handling are circumscribed within the realm of stress-related changes. In accordance with our findings, studies that have evaluated the effects of stress on synaptic plasticity in the hippocampus have consistently shown a reduced capacity for LTP and a corresponding increase in the ability to induce LTD (Kim et al. 1996; Manahan-Vaughan 2000; Pavlides et al. 1996; Xu et al. 1997). This observation is also found in other pathways, such as the hippocampal-prefrontal cortex (Rocher et al. 2004) and amygdala-prefrontal cortex (Maroun and Richter-Levin 2003) pathways. The generality of these effects suggest that neuromodulators and/or circulating neurohormones are the cause. Potential mediators of these modulations of plasticity are high levels of circulating hormones released by the hypothalamus-pituitary-adrenal axis such as corticotrophin releasing hormone, ACTH, and corticosterone, and increases in the release of neuromodulators, such as acetylcholine and norepinephrine, among others, which are all increased during stress (Carrasco and Van de Kar 2003; Degroot et al. 2004; Mark et al. 1996; Valentino et al. 1993). Indeed, activation of both type I and type II glucocorticoid receptors in the hippocampus lead to an inhibition of LTP induction and enhanced LTD (Pavlides et al. 1995). Also, both norepinephrine and acetylcholine have been shown to facilitate the induction of LTD in layers II–III visual cortex synapses (Kirkwood et al. 1999) and in other brain regions (Fujii and Sumikawa 2001; Massey et al. 2001; Scheiderer et al. 2004). Therefore, it is plausible that our findings of an increased expression of LTD in food-deprived and handled animals are mediated by a stress-induced increase in one or several of these neurochemicals. Future studies could be designed to decipher which of these factors are responsible for the increase in LTD produced in motor cortex by food deprivation and/or handling. However, this was not a goal of this study, which was instead interested in testing the effects of skilled motor learning on the excitability and synaptic plasticity of motor cortex pathways in vivo.

An additional finding in this study is that food deprivation not only enhanced LTD, but the expression of this enhanced LTD seemed to be produced via a different mechanism because facilitation (short-term plasticity) changed differently in control and food-deprived animals after the induction of LTD. Moreover, this effect was exacerbated by motor training since, in trained animals, the expression of LTD resulted in an increase in facilitation, which contrasts with the reduction of facilitation produced by LTD in the control animals. We believe that the exacerbation of this effect by training must be considered in the context that trained animals are actually more stressed than food-deprived animals, because in addition to being hungry, trained animals have to work for food. It is difficult to definitively interpret the effects observed on facilitation because these were small and also because inhibitory networks were intact in vivo, making it difficult to interpret changes in facilitation. As a reference, excitatory synapses enhance facilitation as a consequence of a reduction in neurotransmitter release probability (Manabe and Nicoll 1994; Weisskopf and Nicoll 1995; Zucker 1989). In a speculative note, it is possible that the neuromodulators or neurohormones increased by stress associated with food deprivation and handling enhanced LTD by affecting neurotransmitter release probability in the excitatory horizontal synapses.

Do these results imply that skilled motor learning does not involve changes in the efficacy of horizontal connections of the motor cortex? One interpretation of our results is that skilled motor learning may not involve changes in the efficacy of synaptic connections in the primary motor cortex. We disagree with this assertion and believe that these results, although compatible with this initial interpretation, can be interpreted in a different way. We believe that the changes in synaptic efficacy produced in networks of the primary motor cortex during skilled motor learning are highly distributed and bidirectional, and thus would not be apparent in any study that measures population responses that engage the whole distributed network, because the synaptic efficacy of the population remains stable over the course of motor learning. This interpretation seems in agreement with neural network modeling studies of learning and memory that have long been dependent on using synaptic weight change algorithms that implicitly maintain a stable global state (Abbott and Nelson 2000). This serves to avoid both extremes of the network state; that of permanent inactivity and the state of epileptic hyperactivity (Buhmann and Schulten 1986). Thus, studies using models based on spike-timing dependent plasticity rules (Bi and Poo 1998; Magee and Johnston 1997; Markram and Tsodyks 1996) also require a global balance between potentiation and depression of synaptic weights (Carpenter and Milenova 2002; Fusi et al. 2000; Matsumoto and Okada 2003; Song et al. 2000). These models suggest that, during learning, competition for synaptic strengthening occurs through the control of timing of the postsynaptic action potential. This competitive nature of the network components automatically creates a distributed balance between potentiation and depression within the population of synaptic connections (Song et al. 2000; Zhou et al. 2003). These studies support the hypothesis that learning occurs via a distribution of synaptic strengths across the network and thus, a unidirectional change would not be readily expressed when population responses engaging large portions of the network are simultaneously assessed. In contrast, it would be expected that changes caused by neuromodulators or neurohormones, which are more widespread than activity-driven changes, may be manifested more clearly in the population responses measured in these studies. Indeed, in agreement with this interpretation, we found that changes potentially caused by neuromodulators or neurohormones drive the population of synaptic connections in the neural network in a similar direction (i.e., enhancement of LTD), and thus, are clearly expressed by the population responses measured in both hemispheres of all the food-deprived animals irrespective of whether they were trained or not. In conclusion, stress-related mechanisms produced by food deprivation and handling enhance LTD in the population responses of horizontal pathways in layers II–III of the motor cortex, whereas skilled motor training per se has no significant effect on these responses.

	GRANTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION GRANTS REFERENCES

 
This work was supported by the National Institutes of Health and the EJLB Foundation.

	FOOTNOTES

 
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Address for reprint requests and other correspondence: M. Castro-Alamancos, Dept. of Neurobiology and Anatomy, Drexel Univ. College of Medicine, 2900 Queen Ln., Philadelphia, PA 19129 (E-mail: manuel.castro{at}drexel.edu)

	REFERENCES

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION GRANTS REFERENCES

 
Abbott LF and Nelson SB. Synaptic plasticity: taming the beast. Nat Neurosci 3(Suppl): 1178–1183, 2000.

Abraham ME and Gogate MG. Effect of stress on behaviour in rats. Ind J Physiol Pharmacol 33: 84–88, 1989.[Medline]

Aroniadou VA and Keller A. Mechanisms of LTP induction in rat motor cortex in vitro. Cereb Cortex 5: 353–362, 1995.[Abstract/Free Full Text]

Asanuma H and Pavlides C. Neurobiological basis of motor learning in mammals. Neuroreport 8: i–vi, 1997.[ISI][Medline]

Bear MF and Malenka RC. Synaptic plasticity: LTP and LTD. Curr Opin Neurobiol 4: 389–399, 1994.[CrossRef][Medline]

Bi GQ and Poo MM. Synaptic modifications in cultured hippocampal neurons: dependence on spike timing, synaptic strength, and postsynaptic cell type. J Neurosci 18: 10464–10472, 1998.[Abstract/Free Full Text]

Bliss TV and Collingridge GL. A synaptic model of memory: long-term potentiation in the hippocampus. Nature 361: 31–39, 1993.[CrossRef][Medline]

Buhmann J and Schulten K. Associative recognition and storage in a model network of physiological neurons. Biol Cybern 54: 319–35, 1986.[CrossRef][ISI][Medline]

Bury SD and Jones TA. Unilateral sensorimotor cortex lesions in adult rats facilitate motor skill learning with the "unaffected" forelimb and training-induced dendritic structural plasticity in the motor cortex. J Neurosci 22: 8597–8606, 2002.[Abstract/Free Full Text]

Carlson JN, Herrick KF, Baird JL, and Glick SD. Selective enhancement of dopamine utilization in the rat prefrontal cortex by food deprivation. Brain Res 400: 200–203, 1987.[CrossRef][ISI][Medline]

Carpenter GA and Milenova BL. Redistribution of synaptic efficacy supports stable pattern learning in neural networks. Neural Comput 14: 873–888, 2002.[Abstract/Free Full Text]

Carrasco GA and Van de Kar LD. Neuroendocrine pharmacology of stress. Eur J Pharmacol 463: 235–272, 2003.[CrossRef][ISI][Medline]

Castro AJ. Limb preference after lesions of the cerebral hemisphere in adult and neonatal rats. Physiol Behav 18: 605–608, 1977.[CrossRef][Medline]

Castro-Alamancos MA. Origin of synchronized oscillations induced by neocortical disinhibition In vivo. J Neurosci 20: 9195–9206, 2000.[Abstract/Free Full Text]

Castro-Alamancos MA. Different temporal processing of sensory inputs in the rat thalamus during quiescent and information processing states in vivo. J Physiol 539: 567–578, 2002a.[Abstract/Free Full Text]

Castro-Alamancos MA. Properties of primary sensory (lemniscal) synapses in the ventrobasal thalamus and the relay of high-frequency sensory inputs. J Neurophysiol 87: 946–953, 2002b.[Abstract/Free Full Text]

Castro-Alamancos MA and Borrell J. Reversal of paw preference after ablation of the preferred forelimb primary motor cortex representation of the rat depends on the size of the forelimb representation. Neuroscience 52: 637–644, 1993.[CrossRef][ISI][Medline]

Castro-Alamancos MA and Calcagnotto ME. High-pass filtering of corticothalamic activity by neuromodulators released in the thalamus during arousal: in vitro and in vivo. J Neurophysiol 85: 1489–1497, 2001.[Abstract/Free Full Text]

Castro-Alamancos MA and Connors BW. Short-term synaptic enhancement and long-term potentiation in neocortex. Proc Natl Acad Sci USA 93: 1335–1339, 1996.[Abstract/Free Full Text]

Castro-Alamancos MA, Donoghue JP, and Connors BW. Different forms of synaptic plasticity in somatosensory and motor areas of the neocortex. J Neurosci 15: 5324–5333, 1995.[Abstract]

Davis CD, Jones FL, and Derrick BE. Novel environments enhance the induction and maintenance of long-term potentiation in the dentate gyrus 1. J Neurosci 24: 6497–6506, 2004.[Abstract/Free Full Text]

Degroot A, Wade M, Salhoff C, Davis RJ, Tzavara ET, and Nomikos GG. Exposure to an elevated platform increases plasma corticosterone and hippocampal acetylcholine in the rat: reversal by chlordiazepoxide. Eur J Pharmacol 493: 103–109, 2004.[CrossRef][ISI][Medline]

Dolbakyan E, Hernandez-Mesa N, and Bures J. Skilled forelimb movements and unit activity in motor cortex and caudate nucleus in rats. Neuroscience 2: 73–80, 1977.[CrossRef][ISI][Medline]

El Fazaa S, Gharbi N, Kamoun A, and Somody L. Vasopressin and A1 noradrenaline turnover during food or water deprivation in the rat. Comp Biochem Physiol C Toxicol Pharmacol 126: 129–137, 2000.[ISI][Medline]

Endou M, Yanai K, Sakurai E, Fukudo S, Hongo M, and Watanabe T. Food-deprived activity stress decreased the activity of the histaminergic neuron system in rats. Brain Res 891: 32–41, 2001.[CrossRef][ISI][Medline]

Enrico P, Bouma M, De Vries JB, and Westerink BH. The role of afferents to the ventral tegmental area in the handling stress-induced increase in the release of dopamine in the medial prefrontal cortex: a dual-probe microdialysis study in the rat brain 3. Brain Res 779: 205–213, 1998.[CrossRef][ISI][Medline]

Feenstra MG, Botterblom MH, and van Uum JF. Local activation of metabotropic glutamate receptors inhibits the handling-induced increased release of dopamine in the nucleus accumbens but not that of dopamine or noradrenaline in the prefrontal cortex: comparison with inhibition of ionotropic receptors 2. J Neurochem 70: 1104–1113, 1998.[ISI][Medline]

Friedberg MH, Lee SM, and Ebner FF. Modulation of receptive field properties of thalamic somatosensory neurons by the depth of anesthesia. J Neurophysiol 81: 2243–2252, 1999.[Abstract/Free Full Text]

Fujii S and Sumikawa K. Nicotine accelerates reversal of long-term potentiation and enhances long-term depression in the rat hippocampal CA1 region. Brain Res 894: 340–346, 2001.[CrossRef][ISI][Medline]

Fusi S, Annunziato M, Badoni D, Salamon A, and Amit DJ. Spike-driven synaptic plasticity: theory, simulation, VLSI implementation. Neural Comput 12: 2227–2258, 2000.[Abstract/Free Full Text]

Greenough WT, Larson JR, and Withers GS. Effects of unilateral and bilateral training in a reaching task on dendritic branching of neurons in the rat motor-sensory forelimb cortex. Behav Neural Biol 44: 301–314, 1985.[CrossRef][ISI][Medline]

Hebb DO. The Organization of Behavior: A Neuropsychological Theory. New York: Wiley, 1949.

Heiderstadt KM, McLaughlin RM, Wright DC, Walker SE, and Gomez-Sanchez CE. The effect of chronic food and water restriction on open-field behaviour and serum corticosterone levels in rats. Lab Anim 34: 20–28, 2000.[Abstract/Free Full Text]

Hess G, Aizenman CD, and Donoghue JP. Conditions for the induction of long-term potentiation in layer II/III horizontal connections of the rat motor cortex. J Neurophysiol 75: 1765–1778, 1996.[Abstract/Free Full Text]

Hess G and Donoghue JP. Long-term potentiation of horizontal connections provides a mechanism to reorganize cortical motor maps. J Neurophysiol 71: 2543–2547, 1994.[Abstract/Free Full Text]

Hess G and Donoghue JP. Long-term depression of horizontal connections in rat motor cortex. Eur J Neurosci 8: 658–665, 1996.[CrossRef][ISI][Medline]

Iriki A, Pavlides C, Keller A, and Asanuma H. Long-term potentiation in the motor cortex. Science 245: 1385–1387, 1989.[Abstract/Free Full Text]

Kargo WJ and Nitz DA. Improvements in the signal-to-noise ratio of motor cortex cells distinguish early versus late phases of motor skill learning. J Neurosci 24: 5560–5569, 2004.[Abstract/Free Full Text]

Kawahara H, Kawahara Y, and Westerink BH. The role of afferents to the locus coeruleus in the handling stress-induced increase in the release of noradrenaline in the medial prefrontal cortex: a dual-probe microdialysis study in the rat brain 16. Eur J Pharmacol 387: 279–286, 2000.[CrossRef][ISI][Medline]

Keller A. Intrinsic synaptic organization of the motor cortex. Cereb Cortex 3: 430–441, 1993.[Abstract/Free Full Text]

Keller A, Iriki A, and Asanuma H. Identification of neurons producing long-term potentiation in the cat motor cortex: intracellular recordings and labeling. J Comp Neurol 300: 47–60, 1990.[CrossRef][ISI][Medline]

Kim JJ, Foy MR, and Thompson RF. Behavioral stress modifies hippocampal plasticity through N-methyl-D-aspartate receptor activ