NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Bermúdez-Rattoni F, editor. Neural Plasticity and Memory: From Genes to Brain Imaging. Boca Raton (FL): CRC Press; 2007.

Cover of Neural Plasticity and Memory

Neural Plasticity and Memory: From Genes to Brain Imaging.

Show details

Chapter 8Reversible Inactivation of Brain Circuits in Learning and Memory Research



The field of learning and memory has benefited from reversible brain interventions since they were introduced more than four decades ago.1,2 Progress in developing a variety of reversible inactivation procedures has led to a wide array of potent tools that now allow us to investigate different levels of learning and organization of brain memory processes. As noted in the literature,3 the complexity and variety of processes involved in learning and memory require a complex approach to advance the understanding of the relevant neurobiological mechanisms. Although much progress has been made at the molecular and cellular levels, a complete picture will be only attained by advancing knowledge of the functional organization of learning and memory brain circuits at the system level, which it is also recognized as the most difficult approach.4

Limitations of the available techniques for undergoing a lesion approach in the field may contribute to the difficulties. This chapter will address the advantages and limitations of reversible inactivation techniques for research on the neural substrate of learning and memory at the system level. Thus, results of reversible manipulations aimed to explore the molecular mechanisms of learning and memory will not be considered because extensive reviews5–9 are available on specific topics.

Since the seminal work of Bures and Buresova10 inactivating wide brain regions to dissociate sensory and associative processes in taste aversion learning and also investigating interhemispheric transfer of memory traces, a range of procedures for temporary inactivation of discrete brain sites has been developed and applied to learning and memory research. Several reviews centering upon different learning procedures have been published in the past decade and show the value of the reversible inactivation approach for advancing knowledge in the field.11,13,14,37 This chapter does not intend to be a comprehensive review of the knowledge gained using reversible inactivation approaches, but will show representative issues that have benefited from this approach, thus leading to a better knowledge of the brain organization of learning and memory processes at the system level. The need for careful interpretation of the behavioral results based on the specific brain changes induced by the currently available reversible inactivation techniques will be stressed.


The brain creates representations of the world based on the information received through the sensory systems. Learning depends on the plastic properties of brain circuits that are able to undergo functional reorganization to adjust the representation of the world in response to ongoing changes of relevant incoming information. Memory consists of the processes involved in the maintenance of the effects induced in the brain by the learning process. Learning systems are able to detect and modify accordingly, not only to changes in the various features of a specific stimulus, but also to changes involving the relationships among stimuli and between responses and consequences. Thus, from a conceptual view, different types of learning can be considered. Based on the variety of learning procedures and various types of memory,15 dissociable brain circuits are responsible of different forms of memory. Therefore, as noted, memory is defined by the properties of brain circuits subserving the neural changes induced by learning experiences.16,17

The first step toward understanding the brain organization of learning and memory is to identify the particular brain circuit that contains the site or sites forming and retaining the memory traces induced by a specific learning experience. As plasticity is a widespread property of neurons, these sites may be located at different levels of the brain organization. In addition to the plastic brain sites, the brain circuits necessary for the basic aspects of learning include the input and output circuits that are also modified by the learning experience. This “essential memory trace circuit”17 may be dissociated from other brain areas that may also show learning-related plastic changes but are not necessary for basic learning.

Locating the essential brain circuits subserving different forms of learning has benefited from the lesion approach. Brain lesions allow us to identify the brain sites necessary for learning and memory. Other approaches are required for unraveling the specific roles of particular brain sites in the circuit and for identifying brain areas that may be involved but are not required in a learning situation. A problem for the lesion approach is that learning and memory-related changes can only be assessed by behavioral changes. Thus, a given learning situation involves not only learning and memory processes but also sensory, performance, arousal, attentional, motivational, and emotional processes. A brain lesion disrupting any of these processes can lead to similar impaired performance in a retrieval test. Thereafter, for attributing a crucial role in memory to a specific brain site using the lesion approach, it is necessary to discard its potential involvement in other required processes. Additionally, the lack of a given lesion effect may be attributed to a recovery of function due to damage-induced brain reorganization.18

Nevertheless, research using permanent lesions has contributed to draw a rather complex picture of the anatomical organization of the learning and memory systems of the brain. It has been shown that different types of learning depend on dissociable distributed brain circuits formed by areas located at different brain levels. Moreover, independent learning circuits may share components as well as modulatory actions by higher-order brain circuits. Interactions among dissociable brain circuits add complexity to this general picture. Also, memory may involve additional brain sites beyond those initially modified by the learning experience.

However, this static picture does not allow us to fully understand the functional architecture of learning and memory systems. Learning and memory consist of time-dependent dynamic processes. Various successive stages of learning such as acquisition, consolidation, retention, retrieval, reactivation, and extinction may be dissociated. Even these conceptually defined stages are not unitary and include independent processes. As these processes take place successively, permanent lesions may not be able to dissociate them, even if applied at different times in a learning situation. The disruption of different processes that takes place after a lesion could equally explain memory impairment induced by permanent damage of a particular brain site. Moreover, the same brain site may have independent roles in different learning stages, which cannot be evidenced by permanent lesions.

Reversible inactivation techniques solve some of these problems, because they allow a particular brain area to be inactivated during a particular stage, being fully functional during later stages. On the one hand, depending on the behavioral procedure and type of learning studied, a reversible lesion may be able to dissociate learning from sensory and performance processes. On the other hand, the specific contribution of a particular brain site to one stage, but not to another, may be unveiled by selective disruption at a specific time window.10 Moreover, if the same area is required to be functional at different stages, several independent roles may be envisaged. A related issue that may benefit from reversible inactivation techniques is dissociating brain circuits subserving different but temporally overlapping learning processes.11 This is the case of retrieval and extinction, because each retrieval test represents also an extinction session. Reversible inactivation allows us to explore the independent role of a given brain site on extinction by later testing on a functionally intact brain. Similar or different outcomes of the functional inactivation in retrieval and extinction would lead to valuable hypotheses concerning the brain areas involved.

In all, reversible inactivation techniques represent valuable tools for exploring not only the temporal organization of the processes involved in learning and memory, but also the structural organization of independent but overlapping learning circuits contributing to the observed outcomes.


There are a number of procedures for reversible brain inactivation since it can block the neural function by interfering with it at different levels. The spatial and temporal parameters of inactivation depend on the technique applied, thus affecting the application to system studies of learning and memory. A variety of enzyme inhibitors have been used to dissect the molecular mechanisms of learning and memory, but reviewing these techniques is beyond the scope of this chapter.5–9

8.3.1. Transcranial Magnetic Stimulation

First used as a technique to measure the conduction times of motor pathways, transcranial magnetic stimulation (TMS) is a noninvasive technique that may have various applications in humans.19–21 In addition to its application as an essay similar to the Wada test or electrical cortical stimulation in patients candidates for brain surgery, TMS may be used as a tool for creating reversible lesions. Although the action mechanisms of TMS on neuronal activity are not well understood, it may interfere with the ongoing pattern of activity with high precision timing. In single-pulse studies (duration of magnetic stimulus <1 msec) it induces neurophysiological effects lasting up to 100 msec. In repetitive TMS trains with frequencies of 1 to 25 Hz and a duration from hundreds of milliseconds to seconds, the interfering effects last throughout the stimulation and may persist at least 1 hour following magnetic stimulation in memory studies.22

Longer-lasting effects that may have applications in the treatment of depression are being explored.19,20,23 It has been calculated that TMS penetrates no deeper than 2 cm from the surface of the scalp and therefore induces focal changes of brain activity only in the area of the cortex directly underneath the coil. However, studies combining TMS and neuroimaging have shown that the effect spread to adjacent areas within 5 to 10 msec and to homologous regions in the opposite hemisphere within 20 msec.22 Cortical and subcortical distal areas that are connected with the affected area also showed changes in blood flow.21,24

8.3.2. Cryogenic Inactivation Techniques

Cooling a brain site at around 20ºC induces a neuronal block surrounded by a concentric hyperexcitable area at 30ºC at the periphery of the inactivated lesion.25 Depending on the temperature reduction, either synaptic transmission affecting the cell bodies or axonal conduction may be blocked. Local neuronal activity and synaptic transmission, but not axonal conduction, seem to be blocked if tissue temperature is kept between 8.5 and 20ºC.60 Cryogenic techniques allow a high precision control over the inactivation onset, duration, and recovery. A complete block can be induced within a few minutes, be maintained for minutes to hours, and readily terminated in minutes. Cryogenic techniques also permit repeating the inactivation of a constant extent area several times in a 24-hour period without affecting reversibility.

Minor deterioration of the tissue has been identified only when the tissue is cooled and rewarmed more than 40 times in a 12-hour period.27,28 The extent of the inactivation depends on the specific technique used. Inactivating deep brain structures require implanting cryotips — small devices formed by two stainless steel tubes.26,29,30 The inner tube delivers the coolant at the tip of the probe, which is insulated or even warmed to avoid unspecific cooling of the overlying brain tissue. While the tip of a cryoprobe was kept at 4.0 to 5.0ºC, the brain temperature recorded at a distance of 0.5 mm was 7.0 to 10.0ºC. It raised to 21.0 to 22.0ºC at a distance of 1 mm and reached 32ºC at a distance of 1.5 mm.26 The effect did not extend beyond 2.0 mm; temperature recorded at this distance was normal. The main drawbacks of cryogenic techniques are the complex technical requirements and the fact that the cooling probe is thicker than the injection cannula used for pharmacological techniques, thus inducing damage to a greater extent in the overlying brain tissue.

8.3.3. Pharmacological Techniques

Inactivating a brain site by injecting pharmacological agents can be accomplished via two main methods: (1) sodium channel blockers such as tetrodotoxin (TTX) and local anesthetics that prevent initiation and transmission of action potentials both in cell bodies and axons, and (2) agonists and antagonists of neurotransmitters that interfere with neuronal activity at the synaptic level. Both require delivering the pharmacological agent through injection cannulae connected to microsyringes driven by injection pumps.25

The injection cannula may be inserted in chronically implanted guiding cannulae in most behavioral studies. For deep brain areas, the microinjection procedure allows more deactivation of smaller regions than the previous techniques and damage to the overlying tissue is minimized due to the smaller diameter of the injection cannula. The possibility of permanent damage due to tissue displacement is minimized by controlling the drug volume (typically a maximum of 1 μl) and infusion rate (1 μl/60 sec has been widely used but lower rates around 1 μl/90 sec minimum are recommended). Although it is possible to apply repeated injections, the spread of the agent may be variable and mechanical damage to the overlying tissue may be enhanced due to repeated insertions of the injection cannula.25 Moreover, in certain brain regions, the repeated administration of some agents may cause permanent damage.28 The tissue elements inactivated and temporal parameters of the inactivation vary, depending on the agent injected. Sodium Channel Blockers

Sodium channel blockers prevent neuronal transmission both in cell bodies and fibers passing throughout the area, inducing both local and distant effects. They may be applied for inactivating any brain region because sodium channels are present in all parts of the nervous system. Both TTX and local anesthetics have been widely used.

The spatiotemporal extent of TTX-induced inactivation has been calculated for different dosages and brain sites. Zhuravin and Bures31 reported that 1 μl of TTX (10 ng) injected into the Edinger/Westphal nucleus blocked a spherical volume of tissue about 3 mm in diameter, with a maximum effect lasting 2 hours and decaying during the subsequent 20 hours. These results were in accordance with those obtained by Harlan et al.32 who monitored lordosis and other reflexes following a similar dosage of TTX in female rats. Intrahypothalamic TTX injections suppressed multi-unit activity completely within 6 min, and this suppression lasted at least several hours. Reduced lordotic responsiveness was evident 40 min after the injection and peaked 2 to 4 hours later; complete recovery required 12 to 24 hours.

In a later study using the same response, Rothfeld et al.33 reported that TTX injections in the dorsal midbrain reduced responsiveness within 2 min following TTX injection and lasted up to 8 hours. Klement et al.34 found that 5 ng/μl TTX injected in the dorsal hippocampus reduced spontaneous local field potentials within 3 min and abolished synaptic and population responses evoked by stimulation of the perforant pathway at sites 2 mm away from the TTX infusion point. The diffusion of TTX, estimated by injecting 5% India ink solution did not exceed a radius of 1.4 mm, which is in agreement with previous findings.

Local anesthetics include both amides such as lidocaine, and esters such as procaine. They have shorter induction and inactivation times than TTX due to rapid enzymatic breakdown. For lidocaine, the blockade duration may last from 15 min to 1 hour, depending on the dosage and much longer (up to 90 minutes) in fiber tracts. The usual range is 20 to 40 ng in 1 μl for single injection studies. Autoradiographic assessment in cortical tissue has shown that the maximum spread takes place within 10 to 15 min, achieving a radius of 1.5 to 2 mm.25 Glucose uptake and metabolism monitoring has indicated hypometabolism in an extensive area of 3 mm which doubles the area where the drug level declines. Thus, reduced synaptic activity of neurons efferent from the inactivated site may explain it.

Boehnke and Rasmusson35 compared the extent and duration of the neuronal inactivation induced by TTX (10 μM) and lidocaine (10%) delivered via microdialysis in the somatosensory cortex. Electrophysiological recordings of sensory-evoked potentials indicated that TTX induced a more complete neural blockade (60%) over a wider radius than lidocaine (2 and 1 mm, respectively). Responses recovered within 40 min at 0.5 mm after lidocaine infusion while they remained at least 2 hours after TTX infusion. Differences in metabolism, removal mechanisms, and relative binding strengths may explain these differences.35 Agonists and Antagonists of Neurotransmitter Receptors

The use of agonists and antagonists of neurotransmitter receptors provides useful techniques for inactivating specific brain systems and regions. The infusion of neurotransmitter receptor agonists and antagonists in a particular brain area has the advantage of temporary inactivation of the local neurons, avoiding fibers of passage.

A comparison of permanent and reversible lesions that equate neurotoxic lesions and neurotransmitter agonists and antagonists versus electrolytic lesions and lidocaine or TTX in terms of tissue affected has been established.14 For specific brain regions such as the hippocampus, antagonists of the main excitatory transmitter receptors such as glutamate may block neuronal activity in the area. However GABA-A receptor agonists that hyperpolarize neurons, preventing action potential generation, are good choices for most brain sites because GABA-A receptors have a wide distribution in the central nervous system.

The glutamate AMPA/kainate receptor antagonist LY326225 has been applied for reversible inactivation of the hippocampus in a study that examined the extent and time course of functional inactivation. Dentate gyrus field potentials in response to perforant pathway stimulation or CA1 potentials in response to stimulation of the homotopic contralateral CA1 region were monitored. Acute infusions (1 μl; 1.5 mM) reduced 90% extracellular field potentials in 4 to 6 hours. Chronic infusions (0.375 mM) through osmotic minipumps for 7 days abruptly decreased the fast synaptic transmission that returned within 1 day of pump exhaustion.14 The extent of the inactivation was measured with 2-DG autoradiography during and after drug infusion. The results showed that dorsal hippocampus (CA1-CA3 and dentate gyrus), but not ventral hippocampus, showed reduced glucose utilization. The reversibility of the inactivation was demonstrated by the fact that normal levels of glucose utilization were recorded after a 7-day inactivation.

The GABA-A agonist muscimol is the most common agonist in reversible inactivation studies. Unlike the short-lasting blockade induced by GABA with a duration similar to those of local anesthetics, the effects of muscimol persist for 12 to 24 hours.25 This can be attributed to the multiple mechanisms for clearing unbound GABA from synapses and the fact that muscimol has a higher affinity and bonds more tightly to GABA-A receptors. Thus muscimol is either immediately bound to local GABA receptors or is taken up locally by glia.25

Drug spread measured by autoradiography and spatial extent of inactivation monitored by local glucose uptake and metabolism was similar for muscimol and lidocaine.25 Muscimol produced a local block of neuronal firing by hyperpolarization. However, this does not preclude the induction of distal effects on remote brain areas connected with the inactivated site. Both these effects should be taken into account for an accurate interpretation of effects of muscimol injections on behavior. In order to dissociate both local and distant effects, agents such as muscimol that act at the synaptic level and those such as TTX that disrupt axonal transmission are usually applied.17,36

8.3.4. Protein Synthesis Blockers

Disregarding specific enzyme inhibitors that inactivate particular molecular cascades, a variety of protein synthesis blockers administered to specific brain sites have been applied to induce transient metabolic lesions. Anisomycin is the most used inhibitor of protein synthesis used in learning and memory studies, although other inhibitors such as puromycin and cycloheximide are also used. Anisomycin is considered a relatively specific inhibitor that blocks the peptidyl transferase reaction in ribosomes.27 It has been reported that local anisomycin injections in the gustatory cortex, but not intraventricular injections, reduced protein synthesis more than 90% as assessed by injection of [35S] methionine. The inhibition increased rapidly in the first 20 min and lasted for 90 min, slowly decaying in the next 240 min. The effect was localized, affecting tissue with a radius smaller than 2 mm.37

8.3.5. Genetic Inactivation Techniques

A similar reversible lesion approach at the system level by inactivating a particular gene in learning and memory has become possible nowadays through the development of mouse genetic manipulation techniques. The use of region-specific inducible knockouts solves some of the problems posed by conventional knockout mice and affecting all tissues and life stages.38 By temporarily switching off a particular gene in a specific brain region of the adult mouse, ontogenesis is left undisturbed and brain circuits involved in sensory and motor processes may be spared, thus, facilitating the interpretation of the behavioral outcome.

However, this technique presented difficulties related to the mutant mice generation method, requiring at least four independent mutant mouse strains in order to generate the final knockout mice. Moreover, the temporal resolution of the inactivation may not adjust to the time scale of some learning and memory processes. Most knockouts of gene expression take hours to days after beginning the treatment with the inducer and several days of inducer withdrawal may be required for restoring the gene expression.38 Cui et al.39 reported that the knockout of NMDAR function in inducible, reversible, forebrain-specific NMDAR1 knockout (iFB-KO) mice occurred 5 days after feeding the animals with food containing the inducer doxycycline. This long delay is due to the time required for the previously made NMDAR to be degraded, because the inducer readily switches off the transgene expression. Thus, a new approach has been applied to generate mutant transgenic mice expressing inhibitors that compete with the natural form by inactivating a target molecule at the protein level.38

Josselyn et al.40 reported a delay of 6 hours for CREB disruption after tamoxifen administration in transgenic mice that reversibly expressed a dominant negative form of CREB in the forebrain. In the absence of the inducer, the CREB repressor is inactive, but when active it competes with the endogenous CREB disrupting CREB-mediated transcription. Wang et al.41 were able to inactivate and activate CAMKII with a high temporal resolution in a range of minutes by generating a transgenic mouse strain with a modified protein containing a silent structural mutation that creates an artificial site so that a synthesized ATP inhibitor can bind to it. According to the pharmacokinetics, the inhibitor enters the brain 3 to 5 min after i.p. injection and reaches peak brain levels at 20 min, decreasing to basal level in 45 min. A single i.p. injection of the inhibitor completely suppressed kinase activity for the following 8 to 35 min. Chronic oral intake in drinking water induced partial inhibition by 6 hours and complete suppression after 24 hours. This inhibition can be maintained without observable side effects and it is easily reversed within 2 days after withdrawal of the inhibitor.

8.3.6. Advantages And Limitations Of Reversible Inactivation Techniques

Certain issues should be taken into account for the selection of the technique to be applied and the interpretation of the behavioral outcomes in learning and memory studies. First, although the particular neuronal mechanisms interfered with may be different, all the techniques mentioned share the abrupt disruption of a specific site, which induces a completely different brain state affecting also distal areas. This represents an advantage because circuit reorganization does not seem to play a relevant role in reversible lesions because no permanent damage occurs.34 However, the altered brain state induced by the abrupt local functional inactivation may affect learning and memory. It is well-known that both retrieval and extinction are context-dependent processes, being either external or internal.

The learned response may be under the control of contextual cues present during acquisition or extinction. In state-dependent learning, the relevant contextual cue is usually a centrally acting drug such as ethanol or morphine,42 but altered brain activity induced by epilepsy has also been proposed to play a role.43 The absence of the drug during the test day impairs memory retrieval following pre-training administration of drugs. This impairment may be reversed by pre-test administration of a dose similar to that applied during training.

Drug-induced state-dependent learning is well documented in different species from C. elegans44 to humans.42 Thus, the absence of the CR during retrieval with an intact brain following inactivation of a given brain site during acquisition could be attributed to a state-dependent learning impairment rather than a disruption of the acquisition processes. A similar explanation may account for retrieval deficits following pretesting inactivation of a brain site that was intact during acquisition. Control groups demonstrating that a similar inactivation in different areas or different learning tasks does not induce a similar impairment are required for excluding such an effect. Additionally, the altered brain state may have aversive properties, thus being able to act as an unconditioned stimulus. This is an important issue as aversive learning protocols are widely used for studying brain memory circuits. Demonstrating no aversive properties of the brain intervention is required if the results show intact conditioned responses.45

The second issue is that most of the techniques mentioned require stereotaxic surgery and infusion procedures. These procedures have the great advantage of allowing access to deep brain sites but show limitations regarding both the precision of the brain intervention and behavioral consequences. In addition to the issues shared with permanent lesions, stereotaxic and infusion procedures raise specific concerns in reversible inactivation studies. On the one hand, the current development of stereotaxic procedures permits limited spatial precision in target locations that vary among subjects. In permanent lesion studies, subsequent histological analysis of the damage area overcomes this limitation. However, only the location of the injection device track can be assessed in reversible inactivation studies, unless additional staining procedures are added. In fact, most of the reversible inactivation studies aimed to monitor the extent of the deactivated area have been performed in independent groups of animals. Variability of the infusion point location is increased in pharmacological studies if repeated injections in the same animal are required, because only the guiding cannula, but not the injection cannula, remains in place during the interval between infusions.

On the other hand, microinfusion procedures are also great sources of variability. Unless the injected volume and infusion rate are carefully controlled, variability among subjects can be expected due to unnoticed occlusion of the injection cannula as it is advanced through the brain because the tubing connecting the microsyringe and the infusion device can expand slightly.25 Additionally, unnoticed changes in the rate and the volume of the injection may induce irreversible tissue trauma that can be detected by subsequent histological analysis.

With respect to the behavioral effects of infusion procedure, the possibility that the procedure may provide contextual cues should be considered. A role for the i.p. injection cues has been proposed in other learning phenomena.46,47 Whenever appropriate, and depending on the technique used, previous habituation to the injection procedure or applying control vehicle injections can be an appropriate control procedure.48 A further issue is the time that infusion takes, that may exclude detecting short-lasting changes involved in learning and memory.8 Infusion usually takes about 3 min and the drug remains in the target area for several minutes before diffusion.

Finally, a third issue concerns the reversibility of the brain inactivation that represents the advantage of the mentioned techniques. On the one hand, the choice of the inactivation technique will depend on the estimated duration of the process under study. However, the temporal parameters of the transient inactivation cannot be established precisely with most techniques. Individual variability should be also taken into account. Thus, as the temporal parameters of hidden learning and memory processes are not always well known, careful interpretation of negative results is needed. Conversely, the advantage of reversible techniques for applying within-subject designs28 may be limited by restrictions imposed in repeating the intervention. Depending on the brain area and the particular technique applied, the risk of a decreased effect or permanent damage after repeated inactivation should be considered.

In all, a good knowledge of the brain and behavioral effects of the reversible inactivation techniques available is required for the advantages that they bring to the study of learning and memory brain systems to be most useful.


Considering that the functional brain architecture underlying memory is complex because it is formed by a variety of dissociable brain circuits that may interact, a choice strategy for applying the reversible approach reducing alternative interpretation may be to use simple learning procedures inducing robust learned responses that are not easily disrupted by nonspecific effects of transient brain inactivation. Previous knowledge of the behavioral parameters and well-defined sensory and performance pathways involved in the particular learning type will facilitate the experimental design and the interpretation of the reversible inactivation effects.

This is the case of eyeblink classical conditioning, one of the best defined procedures of learning at the behavioral and neurobiological levels, which benefits from the reversible inactivation approach. Eyeblink conditioning has the advantage of depending on brain plastic sites located outside the main sensory and motor pathways. The anatomical dissociation of these processes eliminates some of the interpretation problems using the lesion approach. However, permanent lesions cannot dissociate the specific role of a given brain site in the learning process. Instead, learning and performance deficits would lead to similar outcomes.

The use of one-trial learning procedures provides two main advantages when applying reversible inactivation techniques. First, the acquisition process is localized more precisely in time, making easier to dissociate acquisition and retrieval. Second, the use of one-trial learning protocols reduces the number of brain inactivation periods required, thus avoiding the problems related to repeated blocking in most of the inactivation techniques. Extensive research applying the reversible inactivation approach to inhibitory avoidance learning,12,13 fear conditioning,49–51 and taste aversion learning10,11 has achieved fruitful results. In fact, inactivation of a critical learning brain site during the acquisition stage using one-trial learning should lead to impaired CR during a subsequent retrieval test after inactivation removal. However, no effect should appear if the inactivated site is only involved in performance.

An additional advantage is the possibility of temporally separating sensory and learning processes during acquisition such as in taste aversion learning which permits long delays in ranges of minutes or even hours between the conditioned stimulus (CS) and the unconditioned stimulus (US). Thus, inactivation of a given brain area may be applied after taste processing, leaving the brain intact again during later conditioning and retrieval test.

Taking advantage of this peculiar feature, reversible inactivation techniques have facilitated identifying an associative role of brain sites that are also relay areas in the main gustatory system such as the parabrachial area.53–55 Nevertheless, the problem of dissociating sensory, motor, or performance factors from learning processes does not apply to reversible post-trial interventions because the brain may be intact during both acquisition and testing. In fact, studying the brain mechanism of the consolidation of the memory trace has been one of the fields that has benefited most from the use of reversible interventions.6,9,56

Finally, complex learning phenomena protocols may be useful behavioral tools for dissociating sensory and learning processes that overlap during acquisition or retrieval sessions. It is well-known that previous experience either with a CS or US interferes with subsequent learning. Thus, latent inhibition induced by the previous exposure to the conditioned CS and the effect of the US pre-exposure may be used as powerful tools if reversible brain inactivation during the acquisition phase results in memory impairments. As both phenomena require a previous temporally separated pre-exposure phase, disruption by reversible inactivation of the relevant brain site during this phase may disclose its sensory role.

Instead, a taste memory role for the gustatory insular cortex has been supported by Berman and Dudai37,45 using a latent inhibition procedure. Also, negative results of inactivation during the pre-exposure phase (the inactivation conditions identical to those inducing disruption during the acquisition phase) exclude an interpretation based exclusively in sensory impairment and suggest an associative role for the area. As an example, Ballesteros et al.53 showed that transient PBN blockade by TTX injections 30 min after LiCl exposure did not interfere with the US pre-exposure effect, showing that the pre-exposed group had reduced taste aversions. These results demonstrated that the disrupting effect of PBN inactivation during taste aversion acquisition using an identical procedure could not be attributed to impaired processing of US-aversive properties.

However, the use of reversible inactivation technique in complex learning behavioural protocols has the main limitation of requiring a higher number of animals, thus increasing cost and time requirements. The need of several control groups to demonstrate complex phenomena represents a serious pitfall of neurobiological studies57 and may be the reason why reversible inactivation techniques are rarely applied to studies of more complex learning protocols.


The reversible inactivation approach has proven especially valuable for localization of associative loci in the previously identified essential brain circuit of a given type of learning. Research on taste aversion learning and eyeblink classical conditioning may exemplify two different strategies aimed to dissociate sensory, motor and associative roles of a brain area. Additionally, research aimed to dissociate the neural circuits subserving overlapping learning processes can benefit from reversible inactivation studies. Research on the neural mechanisms of extinction may be representative of this issue.

8.5.1. Dissociating Sensory And Associative Processes During Acquisition

Due to the fact that taste aversion permits introducing a long delay between the taste CS and the LiCl injection usually applied for inducing malaise (US), reversible lesions are especially appropriate to leave intact taste processing both during acquisition and testing. This temporal dissociation has been essential for identifying an associative role of the parabrachial nuclei (PBN) in taste aversion learning because the area is the second brainstem relay in the main gustatory and visceral pathways.52

While total permanent lesions of the PBN interfere with autonomic functions essential for survival, partial permanent lesions may disrupt either taste or visceral processing, thus preventing a clear interpretation in terms of associative deficits.10,11 Pioneering studies using TTX for temporary inactivation of the PBN after taste processing54 or following the acquisition trial55 support an associative role of the area. These interventions prevented the acquisition of conditioned taste aversions, suggesting an associative role of the area in addition to its sensory role. Moreover, a potential explanation of the results in terms of visceral processing impairment induced by pre- or post-trial PBN inactivation has been discarded. A similar dosage TTX injection after LiCl administration does not prevent visceral processing as demonstrated by efficient US pre-exposure effects on subsequent conditioning.53

8.5.2. Dissociating Motor And Associative Processes During Acquisition

In eyeblink classical conditioning, a variety of reversible lesion techniques such as muscimol, TTX, reversible cooling, and protein synthesis inhibitors have been combined to dissociate the specific roles of previously identified brain areas forming the basic learning circuit in acquisition and motor performance.17 Inactivation of each brain site forming the basic circuit during acquisition would prevent the conditioned response (CR). However, inactivation during acquisition of those areas relevant for motor output should not prevent learning unless CR performance would be required.

Consistently, animals trained during inactivation of the cerebellar anterior interpositus nucleus and its afferent sites did not exhibit the CR in a later test after removal of the inactivation. No savings were seen in later training. In contrast, inactivation of the areas receiving efferent projections from the interpositus nucleus, such as the superior peduncle, red nucleus, and motor nuclei, prevented CR during training but this was evident in a later test leaving the brain intact. In fact, a complete learning reaching the asymptote was reported. These results dissociate the role of the explored brain sites in acquisition and performance and support an associative role for the interpositus nucleus in eyeblink conditioning (see alternative interpretations based on electrophysiological data58,59).

8.5.3. Dissociating Overlapping Learning Processes

The use of reversible inactivation during extinction tests is providing a new tool for exploring the neural substrates underlying extinction in different types of learning. The results demonstrating the possibility of dissociating anatomical and functional circuits involved in acquisition, expression, and extinction are in accordance with the current view of extinction as a form of new learning that interferes with the previous learned CR. Local injections of anisomycin into different amygdalar nuclei yielded opposite results on acquisition and extinction of conditioned taste aversion.

Injections applied 20 min before the extinction test into the basolateral (BLA) but not the central (CeA) nucleus of the amygdala impaired extinction. In contrast, the same inactivation procedure in CeA but not BLA impaired learning if applied 20 min before the conditioning session.60 Immediate post-trial anisomycin injections in the same amygdala nuclei yielded similar results. Moreover, the same basolateral amygdala inactivation induced no effect in a behavioral protocol including intensive two-trial learning which depressed extinction.61 A similar interference of extinction by anisomycin injections into the insular gustatory cortex before the retrieval test has been reported; an aversive effect of the transient inactivation “may be” discarded as anisomycin in the insular cortex did not induce aversions.45 These results suggest independent neural circuits for conditioned taste aversion extinction and other acquisition, retrieval, or consolidation processes.

Consistent results have been obtained with other learning procedures such as eyeblink classical conditioning.62 Reversible inactivation of the motor nuclei relevant for the CR by muscimol injections completely prevented extinction, which was resumed on subsequent extinction tests with an intact brain.63 However, the inactivation of the same nuclei did not prevent acquisition of the CR. An unspecific effect of preventing RC expression on extinction can be discarded because muscimol inactivation of the red nucleus that interferes with the CR does not prevent extinction.62


The current development of the lesion approach takes advantage of reversible brain inactivation for the study of the neural substrates of learning and memory at the system level. The advantages and limitations of the available techniques have been reviewed. Careful consideration should be given to behavioral and technical pitfalls of reversible brain inactivation in order to avoid inaccurate interpretation of the results. Some issues such as dissociating both independent processes contributing to a given learning situation and overlapping learning processes may benefit from the reversible approach.


The research of the author and her collaborators was supported in part by grants BSO2002-01215 (MICYT, Spain) and SEJ2005-01344 (MEC, Spain).


Bures J, Buresova O. The use of Leao’s spreading depression in the study of interhemispheric transfer of memory traces. J. Comp. Physiol. Psychol. 1960;53:558. [PubMed: 13688972]
McGaugh JL, et al. Facilitation of maze learning with post-trial injections of 5-7-dyphenyldiazadamantan-6-ol (1757 I.S.) J. Comp. Physiol. Psychol. 1962;55:710.
Cahill L, McGaugh JL, Weinberger NM. The neurobiology of learning and memory: some reminders to remember. Trends Neurosci. 2001;24:578. [PubMed: 11576671]
Kandel ER, Pittenger Ch. The past, the future and the biology of memory storage. Phil. Trans. R. Soc. Lond. B. 1999;2:354, 2027. [PMC free article: PMC1692699] [PubMed: 10670023]
Cammarota M, et al. Retrieval and the extinction of memory. Cell. Mol. Neurobiol. 2005;25:465. [PubMed: 16075375]
Dudai Y. Molecular bases of long-term memories: a question of persistence. Curr. Opin. Neurobiol. 2002;12:211. [PubMed: 12015239]
Izquierdo I, Medina JH. Memory formation: the sequence of biochemical events in the hippocampus and its connection to activity in other brain structures. Neurobiol. Learn. Mem. 1997;68:285. [PubMed: 9398590]
Izquierdo LA, et al. Molecular and pharmacological dissection of short- and long-term memory. Cell. Mol. Neurobiol. 2002;22:269. [PubMed: 12469870]
McGaugh JL, Izquierdo I. The contribution of pharmacology to research on the mechanisms of memory formation. Trends Pharmacol. Sci. 2000;21:208. [PubMed: 10838606]
Bures J, Buresova O. Reversible lesions allow reinterpretation of system level studies of brain mechanisms. Concepts Neurosci. 1990;1:69.
Bures J. Reversible lesions reveal hidden stages of learning. In: McGaugh JL, et al., editors. Plasticity in the Central Nervous System: Learning and Memory. Lawrence Erlbaum; Mahway, NJ: 1995; chap. 7.
Lorenzini CGA, et al. Analysis of mnemonic processing by means of totally reversible neural inactivation. Brain Res. Protocol. 1997;1:391. [PubMed: 9384821]
Lorenzini CGA, et al. Neural topography and chronology of memory consolidation: a review of functional inactivation findings. Neurobiol. Learn. Mem. 1999;71:1. [PubMed: 9889069]
Riedel G, et al. Reversible neural inactivation reveals hippocampal participation in several memory processes. Nat. Neurosci. 1999;2:898. [PubMed: 10491611]
Squire L. Memory systems in the brain: a brief history and current perspective. Neurobiol. Learn. Mem. 2004;82:171. [PubMed: 15464402]
Lamprecht R, Dudai Y. The amygdala in conditioned taste aversion: it’s there, but where. In: Aggleton JP, editor. The Amygdala: A Functional Aanalysis. Oxford University Press; Oxford; 2000. chap. 9.
Thompson RF. In search of memory traces. Annu. Rev. Psychol. 2005;56:1. [PubMed: 15709927]
Cimadevilla JM, et al. Inactivating one hippocampus impairs avoidance of a stable room-defined place during dissociation o0f arena cues from room cues by rotation of the arena. Proc. Natl. Acad. Sci. USA. 2001;98:3531. [PMC free article: PMC30687] [PubMed: 11248112]
Paus T. Imaging the brain before, during, and after transcranial magnetic stimulation. Neuropsychologia. 1999;37:219. [PubMed: 10080379]
Paus T, Barret J. Transcranial magnetic stimulation (TMS) of the human frontal cortex: implications for repetitive TMS treatment of depression. J. Psychiatr. Neurosci. 2004;29:3268. [PMC free article: PMC446221] [PubMed: 15309043]
Paus T, et al. Transcranial magnetic stimulation during positron emission tomography: a new method for studying connectivity of the human cerebral cortex. J. Neurosci. 1997;17:3178. [PubMed: 9096152]
Grafman J, Wassermann E. Transcranial magnetic stimulation can measure and modulate learning and memory. Neuropsychology. 1999;37:159. [PubMed: 10080373]
Barret J, et al. Mechanisms of action underlying the effect of repetitive transcranial magnetic stimulation on mood: behavioral and brain imaging studies. Neuropsychopharmacology. 2004;29:1172. [PubMed: 15029151]
Paus T, Wolforth M. Transcranial magnetic stimulation during PET: reaching and verifying the target site. Hum. Brain Mapping. 1998;6:399. [PubMed: 9788080]
Martin JH, Ghez C. Pharmacological inactivation in the analysis of the central control of movement. J. Neurosci. Meth. 1999;86:145. [PubMed: 10065983]
Wang Y, Lavond DG, Chambers KC. The effects of cooling the area postrema of male rats on conditioned taste aversions induced by LiCl and apomorphine. Behav. Brain Res. 1997;82:149. [PubMed: 9030396]
Dudai Y, Eisenberg M. Rites of passage of the engram: reconsolidation and the lingering consolidation hypothesis. Neuron. 2004;44:93. [PubMed: 15450162]
Lomber SG. The advantages and limitations of permanent or reversible deactivation techniques in the assessment of neural function. J. Neurosci. Meth. 1999;86:109. [PubMed: 10065980]
Clark RE, Zhang AA, Lavond DG. The importance of cerebellar cortex and facial nucleus in acquisition and retention of eyeblink/NM conditioning: evidence for critical unilateral regulation of the conditioned response. Neurobiol. Learn. Mem. 1997;67:96. [PubMed: 9075238]
Wikgren J, et al. Cooling of the cerebellar interpositus nucleus abolishes somatosensory cortical learning-related activity in eyeblink conditioned rabbits. Behav. Brain Res. 2006;170:94. [PubMed: 16580077]
Zhuravin IA, Bures J. Extent of the tetrodotoxin induced blockade examined by papillary paralysis elicited by intracerebral injection of the drug. Exp. Brain Res. 1991;83:687. [PubMed: 2026211]
Harlan RE, et al. Estrogenic maintenance of lordotic responsiveness: requirement for hypothalamic action potentials. Brain Res. 1983;268:67. [PubMed: 6860967]
Rothfeld JM, et al. Reversible disruption of lordosis via midbrain infusions of procaine and tetrodotoxin. Pharmacol. Biochem. Behav. 1986;25:857. [PubMed: 3786344]
Klement D, Pastalkova E, Fenton A. Tetrodotoxin infusions in the dorsal hippocampus block non-locomotor place recognition. Hippocampus. 2005;15:460. [PubMed: 15744735]
Boehnke SE, Rasmusson DD. Time course and effective spread of lidocaine and tetrodotoxin delivered via microdialysis: an electrophysiological study in cerebral cortex. J Neurosci. Meth. 2001;105:133. [PubMed: 11275270]
Nilaweera WU, Zenitsky GD, Braha V. Inactivation of the brachium conjunctivum prevents extinction of classically conditioned eyeblinks. Brain Res. 2005;1045:175. [PubMed: 15910776]
Rosenblum K, Meiri N, Dudai Y. Taste memory: the role of protein synthesis in gustatory cortex. Behav. Neural Biol. 1993;59:49. [PubMed: 8442732]
Nakajima A, Ya-Ping Tang. Genetic approaches to the molecular/neuronal mechanisms underlying learning and memory in the mouse. J. Pharmacol. Sci. 2005;99:1. [PubMed: 16141641]
Cui Z, et al. Requirement of NMDA receptor reactivation for consolidation and storage of nondeclarative taste memory revealed by inducible NR1 knockout. Eur. J. Neurosci. 2005;22:755. [PubMed: 16101757]
Josselyn SA, Kida S, Medina JH. Inducible repression of CREB function disrupts amygdale-dependent memory. Neurobiol. Learn. Mem. 2004;82:159. [PubMed: 15341801]
Wang H, et al. Inducible protein knockout reveals temporal requirement of CaMKII reactivation for memory consolidation in the brain. Proc. Nat. Acad. Sci. USA. 2003;100:4287. [PMC free article: PMC153085] [PubMed: 12646704]
Overton DA. Historical context of state dependent learning and discriminative drug effects. Behav. Pharmacol. 1991;2:253. [PubMed: 11224069]
Aldecamp AP, et al. Acute cognitive effects of nonconvulsive difficult-to-detect epileptic seizures and epileptiform electroencephalographic discharges. J. Child Neurol. 2001;16:119. [PubMed: 11292217]
Bettinger JC, McIntire SL. State-dependency in C. elegans. Genes Brain Behav. 2004;3:266. [PubMed: 15344920]
Berman DE, Dudai Y. Memory extinction, learning anew, and learning the new: dissociation in the molecular machinery of learning in cortex. Science. 2001;291:2417. [PubMed: 11264539]
DeBrugada I, Gonzalez F, Candido A. The role of the injection cues in the associative control of the US pre-exposure effect in flavour aversion learning. Q. J. Exp. Psychol B. 2003;56:241. [PubMed: 12881160]
DeBrugada I, Hall G, Symonds M. The US-preexposure effect in lithium-induced flavor-aversion conditioning is a consequence of blocking by injection cues. J. Exp. Psychol. Anim. Behav Process. 2004;30:58. [PubMed: 14709115]
Maviel T, et al. Sites of neocortical reorganization critical for remote spatial memory. Science. 2004;305:96. [PubMed: 15232109]
Phelps EA, LeDoux JE. Contribution of the amygdale to emotion processing: from animal models to human behavior. Neuron. 2005;48:175. [PubMed: 16242399]
Sanders MJ, Wiltgen BJ, Fanselow MS. The place of the hippocampus in fear conditioning. Eur. J. Pharmacol. 2003;463:217. [PubMed: 12600712]
Schafe GE, Doyere V, LeDoux JE. Tracking the fear engram: the lateral amygdale is an essential locus of fear memory storage. J. Neurosci. 2005;25:10010. [PubMed: 16251449]
Bures J, Bermúdez-Rattoni F, Yamamoto T. Conditioned Taste Aversion: Memory of a Special Kind. Oxford University Press; Oxford; 1998.
Ballesteros MA, et al. Dissociation of the associative and visceral sensory components of taste aversion learning by tetrodotoxin inactivation of the parabrachial nucleus in rats. Neurosci. Lett. 2002;322:169. [PubMed: 11897165]
Ivanova SF, Bures J. Acquisition of conditioned taste aversion in rats is prevented by tetrodotoxin blockade of a small midbrain region centered around the parabrachial nuclei. Physiol. Behav. 1990;48:543. [PubMed: 1963689]
Ivanova SF, Bures J. Conditioned taste aversion is disrupted by prolonged retrograde effects on intracerebral inyections of tetrodotoxin in rats. Behav. Neurosci. 1990;104:948. [PubMed: 2178348]
Dudai Y. The neurobiology of consolidations, or how stable is the engram? Annu. Rev. Psychol. 2004;55:51. [PubMed: 14744210]
Robbins TW, et al. Methods for assessing attention and stimulus control in the rat. In: Sahgal A, editor. Behavioural Neuroscience: A Practical Approach. Oxford University Press; Oxford; 1993. chap. 3.
Delgado-Garcia JM, Gruart A. Building new motor responses: eyelid conditioning revisited. Trends Neurosci. 2006;29:330. [PubMed: 16713636]
Jimenez-Diaz L, et al. Role of the cerebellar interpositus nucleus in the genesis and control of reflex and conditioned eyelid responses. J. Neurosci. 2004;24:9138. [PubMed: 15483132]
Bahar A, et al. The amygdalar circuit that acquires taste aversion memory differs from the circuit that extinguishes it. Eur. J. Neurosci. 2003;17:1527. [PubMed: 12713656]
Bahar A, Dorfman N, Dudai Y. Amygdalar circuits required for either consolidation or extinction of taste aversion memory are not required for reconsolidation. Eur. J. Neurosci. 2004;19:1115. [PubMed: 15009160]
Robleto K, Poulos AM, Thompson RF. Brain mechanisms of extinction of the classically conditioned eyeblink response. Learn. Mem. 2004;11:517. [PubMed: 15466302]
Kruppa DJ, Thompson RF. Inhibiting the expression of a classically conditioned behavior prevents its extinction. J. Neurosci. 2003;23:10577. [PubMed: 14627642]
Copyright © 2007, Taylor & Francis Group, LLC.
Bookshelf ID: NBK3922PMID: 21204437
PubReader format: click here to try


  • PubReader
  • Print View
  • Cite this Page

Other titles in this collection

Related information

  • PMC
    PubMed Central citations
  • PubMed
    Links to pubmed

Related citations in PubMed

See reviews...See all...

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...