Propranolol and Reconsolidation Theory

(1)

A new perspective on the effects of propranolol on memory reconsolidation in studies of fear memory

Literature Thesis

Master in Brain and Cognitive Sciences University of Amsterdam

Cognitive Science Center Amsterdam

Author:

Steven Raaijmakers, BSc Studentnumber 6356877 Supervisor:

Prof. dr. Merel Kindt Co-assessor:

Dr. Marieke Effting Date: July, 2012

(2)

2

During the last decade the topic reconsolidation has been revived into interest and a lot has been published on this topic. Starting around the end of the previous millennium several articles have been influential in reviving this topic into public interest. As seen in Figure 1 the amount that is being published on memory reconsolidation multiplies about 50-fold around the turn of the century. These numbers are indexed by MEDLINE so they could possibly divert from the real number of articles published on memory reconsolidation but it helps to show how much of a rise in interest there has been over the last years. A possible reason for this rise of interest in memory reconsolidation is the already seminal paper by Nader, Schafe & LeDoux in 2000. They show that injection of a protein synthesis inhibitor in the amygdala of a rat during reconsolidation prevents the fear behaviour from returning.

Figure 1: Articles published on memory reconsolidation as indexed by MEDLINE

This finding sparked a hype in academia as reconsolidation, after being neglected for 50 years, was again the hot topic it was before. In 2009, Merel Kindt, showed that giving people the beta-adrenergic blocker propranolol during reconsolidation after fear conditioning could effectively erase the fear memory. In essence; they showed that the fear-potentiated startle reflex was gone after an intervention with propranolol. What kind of implications do these findings have? How can we translate this to the real world? As it seems, a lot of people are already taking the drug propranolol either for hypertension, as a means for keeping the heart in check in case a heart failure or sometimes with stage freight or migraine.

(4)

4

This thesis assumes no grand memory loss epidemic concerning the use of propranolol but is left with the question of whether or not there is one. Because it is an intrinsic property of propranolol (and metoprolol) to lower heart rate and reduce blood pressure it is commonly thought to lower the physiological response after stress. But what if propranolol disrupts reconsolidation? Will this have an effect on this (quite large) population of people taking the drug? Or are there boundary conditions for the drug to gain this effect?

As all these questions are interesting and worthy of answering and if we would try, we should get to the core of the problem. And that core problem is the main aim of this thesis. The question is whether or not propranolol disrupts reconsolidation in a manner that is analogous to the blocking of protein synthesis during reconsolidation as seen in Nader et al. (2000). A more specific analysis will be made regarding the two possibilities that come to mind; disruption of reconsolidation through blockade of beta-adrenergic receptors or a retrieval oriented concept where reconsolidation is successful but new information is incorporated in memory. Pros and cons of both sides of the issues will be discussed and commonalities will be addressed. Logical consistency will also constitute one of the measures of evidence.

The main two hypotheses surrounding the results are the reconsolidation hypothesis and the retrieval deficit hypothesis. Surely there are more alternative explanations to these results but these two are the focus of this thesis. The main aim of this paper is to seek out the evidence or rather to seek an explanation to the promising results of several experiments conducted with the beta-adrenergic antagonist propranolol.

But before we come to the core of this thesis we first have to deal with several concepts that are involved in this exciting research with propranolol. They will be outlined and expanded on in the first chapter. The history of consolidation as well as reconsolidation will be reviewed, seminal papers described and implications will be discussed. Further, the tasks involved in the propranolol studies will be explained to better understand the results gained from these experiments. After these introductory matters, the results will be discussed and the core of this thesis will be addressed, namely the exploration of evidence pro and contra to the disruption of reconsolidation hypothesis. Is reconsolidation truly disrupted or is it possible that reconsolidation is intact and the effect of propranolol is incorporated into the fear memory? Or is there some common ground between these hypotheses? But this is already ahead of time. First we must specifically define the aim of this thesis to allow us to keep on track and not dabble into the very interesting philosophical questions these studies trigger in all of us. Questions as: What is the unit of memory?; Is memory overwritten or new information stored into memory alongside old memory? As they are very interesting during this thesis I will briefly touch upon some of these issues, but only when they pose a problem to interpretations drawn upon the results from experiments.

(5)

5

Chapter 1.1 History of consolidation

The earliest mention of a concept related to consolidation is by the famous Greek philosophers Plato and Aristotle. Both of them came up with the analogy of memory with a wax tablet. In Plato’s Theaetetus he explains (in the words of Socrates) that memory is comparable to an impression on a wax tablet:

"Imagine, then, for the sake of argument, that our minds contain a

block of wax, which in this or that individual may be larger or smaller, and composed of wax that is comparatively pure or muddy, and harder in some, softer in others, and sometimes of just the right consistency .... and say that whenever we wish to remember something we hear or conceive in our own minds, we hold this wax under the perceptions or ideas and imprint them on it as we might stamp the impression of a seal ring. "

(translated by Hamilton in Plato: The collected dialogues, 1961)

This of course is not exactly consolidation as we use the word nowadays, but it resembles the idea of memory as something that leaves its impression in a substrate. This idea is analogous with our thinking today. In the last 100 years theories on memory formation have been influenced by the idea of an existing memory trace in the brain which is reminiscent of the analogy drawn by Plato and Aristotle.

The origin of the word consolidation, as found in the Oxford Dictionary, dates back to the 16th_{century and comes from the Latin word consolidare. This in itself can be} reduced to two parts: con (meaning ‘together’) and solidare (meaning ‘make firm’). This however does not concur with a more updated view of consolidation as a process whereby memories become permanently fixed during the time after registration (Polster et al., 1991). What seems to be missing from the Oxford definition (and wax tablet analogy by Aristotle and Plato) is the time dependent fashion in which consolidation occurs (a clay tablet might have done the trick). The roman orator Quintillian is credited to being the first to mention a consolidation-like process. In his

Institutio Oratoria he notices the effect that a night of sleep can have on the strength of a

memory. He likens the strengthening of memory to the ripening or maturing of fruit (Quintillian 1C AD; 1922). He also discusses several causes for the effect of sleep on memory and draws the next conclusion:

“Whatever the cause, things that could not be recalled on the spot are easily co-ordinated the next day, and time itself, which is generally accounted one of the causes of forgetfulness, is to strengthen the memory.”

(6)

6

No evidence of any use of consolidation-like analogies is found in the literature until Ribot (1892), although Müller and Pilzecker (1900) are usually credited with the first use of the term consolidation. Ribot noticed that after an accident that involved a loss of consciousness the patient besides not remembering the accident also did not remember everything from the period before the accident. Because Ribot’s observations were merely anecdotal evidence and case studies the work by Müller and Pilzecker is regarded as being the first to establish ground for a perseveration-consolidation

hypothesis. In their “Experimentelle Beitrage zur Lehre vom Gedachtniss” they show (in

40 experiments) that memory performance is affected by activities performed in between acquisition and testing. They interpreted these findings with the notion that new memories must first be in a labile state and then consolidate (McGaugh, 2000). They suggested that perseveration of the information was important and that recently learned information interfered with perseveration, this they dubbed retroactive interference or retroactive inhibition.

After this seminal study it stayed relatively quiet due to interference theory until the now famous Donald Hebb published his book “The organization of behaviour: A neuropsychological theory” (Hebb, 1949). In this book he posed his dual-trace hypothesis of memory that states that first a short term trace is formed and

reverberation in this trace causes physiological change which results in long term

memory. This certainly has much in common with the perseveration-consolidation

hypothesis by Müller and Pilzecker, but Hebb never mentions consolidation or the

authors (McGaugh, 1999).

(7)

7

Another memorable piece in the history of consolidation are the findings by Milner and Scoville of patient HM upon the bilateral removal of the medial temporal lobe (Milner and Scoville, 1957). The patient seemed to suffer from some anterograde amnesia, but more interesting was his anterograde amnesia. The patient seemed unable to form new memories. In 1965, Milner proposed that HM could suffer from a consolidation deficit. The short term memory traces were not transformed by Hebbian neural reverberation into long term memory traces. This finding implicated the supposed important role of the medial temporal lobe in the consolidation of memory. Hebb’s work influenced many research in the field of memory research and provided a workable framework and testable hypotheses. One of these hypotheses was the existence of a time-dependent process involved in memory storage. McGaugh showed in 1966 that this process could be interfered with through electroconvulsive shock and could be enhanced through stimulant drugs.

When all these findings are to be combined the view emerges of the medial temporal lobe as a key structure in memory consolidation. But a proper consolidation model did not come to pass until Squire et al. in 1984 formulated that information is stored in the temporal lobe (presumably the hippocampus) and through a rehearsal-like process that information, or a part of that information, is stored in circuits outside of the hippocampus (presumable the neocortex). Although not a very precise definition, the consolidation model by Squire et al. could deal with a lot of the findings regarding the timing of the processes and both the psychological and physiological data that was found.

Last, but not least, is the multiple trace theory provided by Nadel and Moscovitch (1997) that in essence states that the hippocampus functions as a binding element in creating a hippocampal-neocortical ensemble (referred to as the memory trace). Reactivation of the memory trace in altered context results in a new memory trace. In time the result of the overlapping multiple traces is the extraction of the gist of the episode and this information is regarded as independent of the episode itself.

Chapter 1.2 What is memory consolidation?

Beforehand, we must acknowledge the fact that when we are talking about constructs like a memory-unit that the nature of memory presentations and the different memory systems involved are highly complex and not even thoroughly defined. In this section I will try to set out some definitions from literature that will help to frame questions and make interpretations more clear.

From a historical perspective the division between work on physiological consolidation and psychological consolidation is noticeable. It was not until Hebb that these two perspectives merged. Still, there are distinctions to draw and the first notable distinction that can be drawn is between synaptic consolidation and systems

(8)

8

consolidation (Dudai, 2012). Synaptic consolidation is explained as the “post-encoding transformation of information into a long-term form at local nodes in the neural circuit that encodes the memory”. Essentially this entails the cellular cascades that influence the synaptic efficacy after learning. There clearly is an emphasis on local level changes, inferences are made on cellular level and not beyond. The next level is brain systems level and where systems consolidation comes in. Systems consolidation is explained as the “post-encoding reorganization of long-term memory over distributed brain circuits”. Whereas synaptic consolidation works within a timeframe of hours, systems consolidation works within days and maybe years.

A clear operational definition of consolidation may be hard to come up with but commonly agreed is that it is a process that strengthens memory and exhibits a time-specific period upon which the memory is susceptible to amnestic agents. This is also the definition upon which the standard consolidation theory (SCT) is based upon (Squire, 2004). It accounts for observations made on retrograde amnesia and states that memory resides in the hippocampus first and is then transferred to the neocortex. Memory is first dependent on the hippocampus, but afterwards it is independent of the hippocampus. This theory did not make any distinction between kinds of memory, but after work by Schacter & Tulving it is regarded that consolidation refers to declarative memory (Schacter & Tulving, 1994; Squire, 2004). The successor of the SCT is the multiple trace theory (MTT), which posits that the hippocampus initially encodes all information of the memory and provides an index for neocortical neurons (Nadel & Moscovitch, 1997). After initial encoding there exists a hippocampal-neocortical ensemble which is dubbed the memory trace. Every reactivation in a new context result in hippocampal traces and because these traces exhibit extensive overlap, the gist or semantic information of the event becomes more easily extractable. The MTT is a theory on systems level and depends on interaction between hippocampus and neocortex over time. This theory thus shows a more active view of consolidation than former theories of memory consolidation.

Although after Hebb’s Organization of Behaviour the physiological and psychological perspectives on consolidation have been merging, the MTT is more a theory of systems consolidation than synaptic consolidation and the cellular view is rather neglected. Clearly there are connections to be made between these separate views of memory and they will yield much appreciated new knowledge of memory formation.

(9)

9

Chapter 1.3 History of reconsolidation

The traditional view of consolidation, the standard consolidation theory, implies that consolidation is a passive time-dependent process with as an outcome a fixed, stabile memory. In this view each item undergoes consolidation once and is immutable afterwards. During the 60’s evidence against the supposed immutability arose and the seminal account by Misanin and colleagues showed that administration of an amnestic agent upon reactivation of a consolidated memory was able to change the memory (Misanin et al., 1968). This meant that where it was previously thought that memory was stable, reactivation of the memory is able to return the memory into its previous labile state. This process of relabilization is was it known now as reconsolidation (Przybyslawski & Sara, 1997).

This might seem like a breakthrough, but the concept did not catch on until more than thirty years after discovery it was shown that an amnestic agent administered in the amygdala upon reactivation of a fear-conditioned memory could ‘block’ the long term memory (Nader et al., 2000). This renewed the interest into the idea of disruption after reactivation or recall and resulted in a 50-fold increase of publications (Dudai, 2012). This was not due to the novelty of the use of an amnestic agent after reactivated conditioned fear, but the fact that its manipulation was specifically aimed at a brain structure that underlies the circuit involved in fear conditioning. Judge and Quartermain (1982) for instance also induced retrograde amnesia following reactivation with the same amnestic agent (anisomycin, a protein inhibitor), but they injected it systemically. But is this process different from consolidation or does it constitute a reopening of the consolidation window?

Chapter 1.4 What is memory reconsolidation?

So we are left with the question whether or not reconsolidation is anything different from consolidation. But if consolidation is hard to define, than surely reconsolidation is even harder to define? The reconsolidation hypothesis is maybe best captured by its essence that is the relabilization of a formerly consolidated memory. But that is also where confusion arises, because the boundaries of systems consolidation mentioned previously are thought to range from days to years. For practical purposes we will suppose that reconsolidation occurs after a memory has already been consolidated and thus is stabilized. But then is this relabilization really a different process or just the restarting of another consolidation process? Alberini (2005) has reviewed the literature and stated that “although both processes appear to engage the same molecules and mechanisms, brain areas involved in consolidation … are not required for reconsolidation”.

(10)

10

Previous research has shown that protein synthesis after memory formation is crucial to the successful formation of the memory and this also occurs after reactivation (Judge and Quartermain, 1982; Pedreira et al., 2002). Systemic administration of protein synthesis inhibitors results in a loss of memory formation in both consolidation paradigms and reconsolidation paradigms. But there are also differences that must be attended to. For instance protein synthesis is needed in the auditory cortex during consolidation in gerbils, but it is not needed during reconsolidation (Kraus et al., 2002). But then again Nader and colleagues have shown that protein synthesis was indeed need during reconsolidation of auditory fear conditioning (Nader et al., 2000). So it seems as though there must exist some boundary conditions that separate reconsolidation from consolidation. The existence of boundary conditions could still mean that consolidation and reconsolidation might be similar processes, but first of all; what might these boundary conditions be?

A primary boundary condition could be recognized in the finding of Pedreira et al. in 2004. They found that impairments induced by a protein synthesis inhibitor after reactivation only occurred when there was a mismatch between what was expected and what happened. So it might be that there has to be a mismatch between the old stored information and the new information. This would mean that reconsolidation could function as an updating mechanism. This perspective of reconsolidation still leaves open the possibility for a similar mechanism underlying both consolidation and reconsolidation.

Chapter 1.5 Fear conditioning

To study emotion and anxiety in humans and non-human animals researchers have made much use of the fear conditioning paradigm. The basic fear conditioning paradigm consists of the pairing of a former non fearful stimulus (conditioned stimulus or CS) with a fearful stimulus (unconditioned stimulus or US). The paradigm is meant to induce a fear response to a formerly neutral stimulus. This is a useful paradigm to research the working of anxiety disorders and to test possible clinical treatments of these disorders. As we have seen that there are manipulations that can influence memory consolidation and memory reconsolidation.

(11)

11

Chapter 1.6 Memory modulation

With the findings that amnestic agents have a negative effect (decrease of memory strength) and stimulant drugs have a positive effect (increase of memory strength) during consolidation, the next step is to see if there are manipulations that enhance memory during reconsolidation (McGaugh, 1966).

It seems that there exists a memory modulatory system more involved with enhance certain kinds of memory events that actively decreasing memory event (Cahill and McGaugh, 1996). This underlying neurobiological system is thought to be the noradrenergic system. One of the ‘normal’ ways memory is modulated is by emotional arousal. The activation of ß-adrenergic receptors is required for the modulation by this emotional arousal. Note that this modulation both entails enhancement of memory through positive arousal as well as negative arousal.

Two elements are important in the modulation of memory by emotional stimulation: the amygdaloid complex and ß-adrenergic activation (Cahill and McGaugh, 1996; McGaugh et al., 1993). Cahill and colleagues already showed in 1994 that blocking the ß-adrenergic receptors with propranolol (a non-selective ß-blocker) impaired memory for an emotional story, but not for a neutral story. In other words the effects of emotional arousal mediated by the noradrenergic system were rendered ineffective. In 1999, it was shown that through this noradrenergic system emotional memory could be either upregulated, by means of yohimbine (which stimulates noradrenergic activity), or downregulated, by means of metoprolol (blocks ß1-adrenergic receptors) (O’Carroll et al., 1999a). Recently, with functional magnetic resonance imaging, it has been established that after administration of propranolol the amygdala is no longer reactive to emotional pictures (Hurlemann et al., 2010).

But as propranolol is administered orally, one might ask whether or not the effects on memory are due to a central blockade of adrenergic receptors or a peripheral blockade. With the same paradigm used by Cahill et al. (1994) researchers set out to test out this question that resembles the old debate between James-Lange and Cannon. They administered participants placebo, propranolol (works both peripherally and centrally) or nadolol (works peripherally only) and found no differences between groups (O’Carrol et al., 1999b). This means that both propranolol and nadolol were not able to block the effects of noradrenergic stimulation on memory for emotional items. This is of course inconsistent with the results found by Cahill et al. (1994) and raises questions as to why the manipulation did not work. This could be due to the fact that maybe the wrong test had been used and episodic memory is not the memory system most prone to modulation by the noradrenergic system.

(12)

12 Chapter 2 Fear memory and propranolol

2.1. Differential fear conditioning and fear-potentiated startle

The group led by dr. Merel Kindt set out to see what the clinical implications would be of administration of propranolol after retrieval of a fearful memory. They used a variant of the fear conditioning paradigm dubbed differential fear conditioning. In differential fear conditioning the CS-US connection is learned on day 1 (acquisition), on day 2 the CS-US is reactivated and day 3 consists of extinction and reinstatement. In the work by Kindt et al. (2009) on day 2 before reactivation the participants received propranolol. As an outcome measure of fear memory the group used the fear-potentiated

startle effect, which is defined as the increase of the startle reflex by the presence of a

stimulus that was previously paired with a shock. The fear-potentiated startle effect has been demonstrated to be reliably elicited by loud noise or air puffs using auditory or visually conditioned stimuli (Davis, 1986). This is measured in rats by the height of their jump after acoustic startle and in humans it is measured by electromyography of the right orbicularis oculi muscle (essentially it records the blinking reflex after the loud noise).

What is seen in humans is that in the presence of a conditioned stimulus that has before been paired with an unconditioned stimulus the startle reflex is potentiated, i.e. the EMG measured from their eye-blinks is bigger when they are shown a CS+_(paired with shock) than when they are shown a CS-_{. It must be said that this essentially is a test} of non-declarative memory and is thus separate from the knowledge that is obtained by the participant on the pairing of the CS with the US. Other measures of fear are also used in research on emotional memory, such as the skin conductance response (SCR) but it has been suggested that the SCR reflects contingency learning more than fear memory (Weike et al., 2007).

(13)

13

In a standard differential fear conditioning paradigm the participant is first habituated with the auditory stimulus that is used to measure the potentiated startle reflex. The participant will be shown pictures of fear relevant stimuli (Kind et al., 2009 used pictures of spiders); this is done because conditioning to fearful stimuli is less prone to extinction and therefore more like the fear expressed in patient populations suffering from anxiety disorders.

The reason behind is formulated in the biological preparedness theory first proposed by Seligman and evidence for it involves for instance the ability of fear relevant stimuli to elicit fear conditioning using visual masking techniques (Öhman & Soares, 1993). The preparedness theory suggests that we are biologically prepared to acquire a fear response associated with the fearful stimuli if they are paired with a fearful situation. This could be simulated in the laboratory by using an electric shock as a fearful situation that is paired with the fearful stimulus, a spider in case of Kindt et al. (2009). The biological preparedness would results in organisms that acquire associations faster with certain stimuli over other stimuli. It is useful for an organism to learn rapidly from one fearful event than have to require multiple encounters with for instance a snake that could result in harm and possibly death. We adapted biological preparedness to learn faster from singular events and not have to experience dangerous situations multiple times. But what kinds of stimuli are associated with this biological preparedness? As mentioned before dangerous animals are one category that fulfils requirements, but also angry or fearful facial expressions and even faces from other social groups are able to elicit fast learning. The first two categories (animals and fearful/angry faces) are easy to understand but the last group is not easily explained. Faces from other social groups might have acquired negative connotation through kin selection as Palaeolithic groups tended to be related. It is thought that out-group hostility and in-group altruism stems from kin selection in early ancestors (Brewer, 2007).

The differential fear conditioning paradigm is clinically exceptionally useful because it incorporates prominent features of anxiety disorders such as extinction of the fear memory and the reinstatement of the fear memory. After habituation of the participant to the noise, further background noise is used as well to habituate the participant to the presence of sound and to reduce the effect of surprise in the results.

The participant is than shown pictures of fearful stimuli (spiders or snakes usually) paired (or not) with a shock delivered to the wrist. The participant is sometimes asked to report expectancy of the shock when shown the picture. During acquisition the participant is shown the picture for 8 seconds and after 7 seconds the startle probe is triggered. In case of the CS+_{the shock is delivered 500msec after}

(14)

14

the startle probe. In the picture shown here a stimulus from the International Affective Picture System is used to induce the CS-US association. A selectively higher startle reflex to the startle probe after the conditioned stimulus shows if the association has been successfully been obtained. The term reactivation refers to the showing of the fear conditioned stimulus to the participant. The administration of propranolol (or placebo) was done before the reactivation of the fear memory. To be more specific, propranolol and placebo were given 90 minutes before reactivation to be sure the concentration of propranolol would be at peak concentration during reactivation.

On day 3 the participation went through a set of extinction trials; they were only shown the pictures and received the startle probe, not the shock. As it is the case with the acquisition of fear that the contingency of picture with shock is learned, with extinction this contingency is unlearned. This is a method congruent to general treatments of phobia, the so called exposure and response prevention. The patient is exposed to the object of the phobic anxiety in a non-fearful setting. This is done multiple times in order to eventually get rid of the phobic response. This therapy is successful some cases, but not all. A problem to this kind of therapy is the notion of reinstatement. As the therapy is usually carried out in a safe environment, it is possible for the object of the anxiety to retain its ability to induce a phobic response in other contexts.

As explained, the differential fear conditioning paradigm mimics the main characteristics postulated to be most important in anxiety disorders that are: the acquisition of fear, extinction of the fear and the following reinstatement that interferes with successful therapy outcome. The differential fear conditioning paradigm might provide potential clinical purpose searching for ways to improve the existing therapy or explore new ways of treating problems with fear. One way of treating fear in clinical populations might be through the administration of propranolol, a beta-blocker, after reactivation of the patient’s fear. Merel Kindt and colleagues explored this experimentally and this resulted in their paper in 2009: “Beyond extinction: erasing human fear responses and preventing the return of fear” (Kindt et al., 2009). Below the main findings of this influential paper will be explained.

2.2 Propranolol administration during reactivation

The first paper by the group of Merel Kindt at the University of Amsterdam was a straight forward differential fear conditioning paradigm with a reactivation with propranolol or placebo. The main outcome measure was the fear potentiated startle response, described above. Also taken into account was expectancy, a rating whether or not the participant expected to get shocked. This could be regarded as a declarative or cognitive component of the fear memory. The potentiated startle is regarded as non-declarative. Next to the placebo control, they also ran a control experiment that did not include the reactivation, to see whether or not the reactivation of the fear memory is vital to the propranolol manipulation.

(15)

15

What they found was that in all groups fear was learned to an adequate level and there were no differences between the levels of the fear-potentiated startle between any of the groups. During the reactivation all groups showed the same amount of startle regardless of propranolol or placebo (in itself an interesting finding because propranolol is regularly prescribed as an anxiolytic, but this shows that the potentiated startle response is not affected immediately upon administration of the drug). On day three extinction result show that in the placebo group with reactivation and in the propranolol group without reactivation the potentiated startle was still present, but for the group that received propranolol and a reactivation of the fear memory the potentiated startle was no longer present. After extinction, the unconditioned stimulus was delivered to the participants (reinstatement) and this reliably brought back the potentiated startle response to levels equivalent to levels before extinction for both the placebo group and the propranolol group without reactivation. The propranolol with reactivation group once again showed no trace of a potentiated startle.

The inability of the reinstatement procedure to elicit the conditioned fear response receives two general explanations by the authors: (1) the fear memory might be erased and (2) the fear memory is unavailable due to retrieval failure (Kindt et al., 2009). It must also be mentioned that the results for the declarative memory of the fear memory did not show any influence of propranolol. This is explained by the possibility that propranolol leaves the hippocampal component intact while it erases the amygdalan component. Thus declarative and non-declarative memories are differentially affected by propranolol.

Figure 5: Differential fear conditioning paradigm extended with a long term testing phase

After these promising results the group set out the find out what the long term effect was of the propranolol treatment and published their results in 2010 (Soeter & Kindt, 2010). The long term effect is quite interesting because it yield information regarding the ability of the fear memory to return over time (spontaneous recovery). This represents a recurrent phenomenon, especially in clinical settings. They also added the skin conductance response as a measure. This measure is often used in experiments on fear conditioning, but recent research shows that it may reflect contingency learning instead of proper fear learning and it therefore also correlates with declarative knowledge of the fear association (Weike et al., 2007; Hamm & Weike, 2005).

(16)

16

The results again showed that propranolol does not affect the startle response per se (startle response was intact during reactivation; although the response was significantly reduced in the placebo group) and the fear response did not return after reactivation in the propranolol and reactivation group. So effectively they replicated the findings from Kind et al. (2009). The long-term results showed that even after 28 days since extinction the fear association was reinstated in the placebo with reactivation group, but not in the propranolol with reactivation group. This was evidence that with this manipulation the fear association could not be reinstated and also did not re-emerge after a month. Because the skin conductance response was associated with contingency learning of fear associations, it was hypothesized that propranolol should have no effect on skin conductance responses. The results confirmed these expectations. The results of this study were interpreted as evidence for the suggestion that “the drug manipulation specifically affected the molecular processes mediating reconsolidation” in a manner analogous to the findings by Nader et al. (2000) (Soeter & Kindt, 2010).

Following the previous paper, Soeter & Kindt published another research article in 2011 (Soeter & Kindt, 2011a) showing that the effects of the administration of propranolol during reactivation of fear memory extend to category related information. And again, findings from the former two studies were replicated (Kind et al., 2009; Soeter & Kindt, 2010). The dissociation between skin conductance and potentiated startle response was also seen in the results of this study, showing that the drug left the declarative information intact but not the non-declarative information.

Another paper from Soeter & Kindt in 2011 (Soeter & Kindt, 2011b) showed that stimulation of the noradrenergic system during acquisition of a fear association impaired extinction learning and resulted in faster recovery of the fear association after reinstatement and re-acquisition. These results suggest that raised levels of noradrenaline could be implicated in the persistence of fear in anxiety disorders. The involvement of the noradrenergic system prompted Soeter and Kindt to evaluate the effect an increased noradrenergic level during acquisition on their propranolol paradigm (Soeter & Kindt, 2012). In this study they showed that the addition of yohimbine to the paradigm resulted in an impairment of extinction learning, but it did not abolish the effect of propranolol on the final test of the fear memory. The potentiated startle response was effectively removed from the fear memory.

(17)

17

2.3. Other recent comparable work

On obvious candidate for application of this propranolol manipulation are patients with post-traumatic stress disorder (PTSD) who suffer from recurrent bouts of fear. The application to PTSD is especially interesting because it might provide a singularly enough reactivation trigger. The post-retrieval administration of propranolol could yield a treatment for PTSD. This experiment was tried using script-driven traumatic imagery and yielded promising results (Brunet et al., 2008). An interesting experiment was carried out where instead of a pharmacological intervention the authors chose for a less invasive approach. They tried to see if extinction during the reconsolidation window would be more effective than extinction outside of the reconsolidation window (Schiller et al., 2010). The participants in the group where extinction was done during the reconsolidation window showed no return of fear after extinction, whilst the control groups did. The group that received extinction during reconsolidation was also resistant to reinstatement. The difference with the studies carried out by the group from Merel Kindt is that in this experiment the measure of fear was the skin conductance response, which is might resemble more contingency learning than fear learning.

The prime candidate of these exciting results is the explanation formulated by some authors that reconsolidation is disrupted, meaning that either the fear memory is not reconsolidated or that the fear memory is not reconsolidated properly. Another candidate could be the failure of the fear memory to be retrieved. This explanation assumes that reconsolidation is not disrupted. There may be more possible explanations at this point, but these two candidates will be discussed in the next chapter. These two hypotheses will be contrasted with each other and implications will be thought through. The main question for the next chapter thus will be: which fails, reconsolidation or retrieval?

(18)

18

Chapter 3 Personal critical opinion

The main suggestion of the possible working mechanism behind the results obtained from administering propranolol during reactivation in a differential fear conditioning paradigm is that it might be related to the results obtained by blocking protein synthesis in the (basolateral) amygdala in rats after reactivation (Nader et al., 2000). In effect this blocks reconsolidation in the amygdala as the fear memory is no longer present after this treatment. In the article the authors propose that old memories can become active and during this period they can be disrupted by the inhibition of protein synthesis. But is this really what is happening? Are there other explanations? And are they as plausible as the reconsolidation disruption hypothesis?

3.1 Arguments from observation

As propranolol is a very common drug amongst people with heart failure and heart disease is the second largest cause of death in The Netherlands, there are a lot of people taking this drug to prevent heart failure (CSB Statline, 2012). Another case in which propranolol is prescribed is anxiety attacks or stage freight, but for the most part the beta-blocker is given to people with heart failure. This population consist mostly of older adults (50+). Could it be the case that propranolol is affecting the memory of these adult by disrupting the reconsolidation of their memories when reactivated?

What is to be kept in mind when asking a question like that is that we have to deal with boundary conditions. The chronic use of propranolol can be phenomenally different from the way it is used in fear conditioning experiment. Still one could ask what the effect might be of longer term blocking of adrenergic receptors of the amygdala. Do these people no longer benefit from the memory enhancement of arousing events? Is the arousal accompanying every arousing memory they activate during peak propranolol level then removed from the memory? These are interesting questions because even if there are boundary conditions to the effect of propranolol on memory, these conditions could be met in people taking propranolol to prevent heart failure.

3.2 Arguments from methodology

The differential fear conditioning paradigm used could be inadequate to reliably mimic the fear experienced by anxiety disorder patients. For instance the fear association with the shock could not be as highly arousing as the fear in anxiety disorder patients. This might lead to less resistant fear memories that are susceptible to manipulation during fear conditioning. Also the healthy participants used in these studies might have profoundly different fear memory systems compared to the patient population. Another issue is the simplicity of the CS-US association compared with the complexity of the fear in anxiety disorder patients. One could imagine the fear in a

(19)

19

patient to be heavily intertwined and thus there may also be multiple paths by with the fear response could be triggered. On a more basic notion of comparing methodologies between the Kindt et al. (2009) study and the Nader et al. (2000) study is the agent used by which reconsolidation is disrupted. Nader et al. use anisomycin, a direct protein inhibitor, and Kindt et al. use propranolol, an indirect protein inhibitor. Also the method of administration differs between studies, as Nader et al. administer their agent locally and Kindt et al. use systemic administration (local administration obviously not possible in human participants). Are the two working mechanisms therefore comparable? Reasonably they accomplish the same thing, the deactivation of the amygdala and the abolishing of the fear response, but is this then accomplished through the disrupting of reconsolidation or is there another candidate hypothesis?

3.3 Arguments from interpretation

As Alberini mentions, the formation of a new memory could also be an explanation for studies concerned with memory reconsolidation (Alberini, 2007). There is a distinction to be made between the reconsolidation of an original memory and the making of associations between new and reactivated information. Other authors also propose the interpretation that instead of the amnesic treatment impairing the storage of information, the old reactivated information becomes associated with the new “altered internal contextual state” (Riccio et al., 2006). This could be viewed in the MSDR model by Hinderliter et al. (1975). The modified state dependent retention model states that the absence of the internal context at testing could lead to a retrieval deficit. But the observation by Kind et al. that reinstatement is unable to elicit a fear response poses problems for this model. The inability of the fear memory to be recovered is in fact a good argument for the storage failure interpretation. The retrieval failure interpretation assumes that the memory is still there, but competing new memories make it difficult for the old memory to be recovered.

In the storage failure (disruption of reconsolidation) the line of reasoning is the following:

Reactivation makes the fear memory labile

During the reconsolidation window protein synthesis is inhibited Due to the failure of reconsolidation memory traces are weakened The fear memory is lost

The retrieval failure (new memory interferes with the old) the line of reasoning is the following:

Reactivation activates the information from the old memory

During the reconsolidation window new information is incorporated with the old information into a new memory that is reconsolidated

The new memory competes with the old memory The fear memory fails to be retrieved

(20)

20

Kindt et al point to these two possibilities (storage and retrieval failure) in their discussion, but it remains a question which one is the right interpretation (or could they both be true?) (Kindt et al., 2009). Are the two reconcilable? Is there some common ground? Earlier the possibility of two types of reconsolidation was discussed: synaptic reconsolidation and systems reconsolidation. Maybe this could yield us with a framework that is able to bring these interpretations closer together.

Maybe this is a first step to reconcile both of the hypotheses:

Reactivation activates the fear memory and makes it labile (able to incorporate new information into old information)

During the reconsolidation memory the activation of the amygdala is down regulated by the local blocking of synaptic reconsolidation in the amygdala by adrenergic receptor blockade The non-blocked systems reconsolidation incorporates the new information into a newly updated representation of the memory

Activation of the new representation is less able to retrieve the fear response because that response was not stored as new information

In essence this interpretation perceives the fear memory to exist not merely of the fear component mediated by the amygdala, but also the declarative component. In this view systems reconsolidation is not disrupted, but it has incorporated a non-fearful response because the amygdala is not able to signal arousal because synaptic

reconsolidation in the amygdala is disrupted by adrenergic blockade. This yields a

reconsolidated new memory without the fearful response.

In the wording of the Multiple Trace Theory the gist of the fear memory would be profoundly altered. The associations between the gist and the fearful component within the complete memory are no longer able to activate the fear response. Interesting research would involve looking into the possibility of the fearful component to be reconstructed from the cognitive component. The reconstruction of the connection between the fear response and the gist would however yield no information to the question if the loss of the fear response was due to a storage failure or a retrieval failure. If the storage failure hypothesis (disruption of reconsolidation) would predict the total loss of connections to the fear response, the reconstruction of the fear component from the cognitive component seem more improbable than the retrieval failure hypothesis. But this interpretation (complete loss of connections) is not used much. Instead, it is thought that the connections are dampened and thus reconstruction of the fear component would be more probable. The retrieval failure hypothesis predicts that if one retrieves the memory in a context more similar to the context wherein the fearful event is experienced, the probability that the fear response is triggered increases. Although significant overlap will continue to exist, research into the reconstruction of the fear response after reconsolidation disruption remains an interesting opportunity.

(21)

21

Chapter 4 Conclusion

4.1 Recapitulation

The main purpose of this thesis was to get more insight to what is the more probable interpretation of the studies performed by the group of Merel Kindt. They used a differential fear conditioning paradigm and administered propranolol (a beta-blocker) during reactivation (analogous to Nader et al., 2000). This resulted in an inability of the fear response to reappear during the first extinction trial. The first interpretation is that propranolol disrupts reconsolidation and afterward the fear memory trace is weakened or deleted. A second possibility could be that a new memory without the accompanied fear response is reconsolidated into memory and this yields an inability of the fear response to be triggered.

4.2 Different interpretations

It is often said in the literature that: "The fact that reactivation of a consolidated memory can return it to a labile state from which it must be re-stabilized in order to prevent memory loss is well established." (Nader & Hardt, 2009). But is finding than also what explains the prevention of the return of fear in the studies performed at the lab of Merel Kindt? As set out in the previous chapter another interpretation is a viable option (as justly stated in Soeter & Kindt, 2010). Next to the storage failure during reconsolidation, there exists a possibility that a retrieval failure causes the fear response to not be exhibited. And one might even make a distinction between synaptic reconsolidation and system reconsolidation, which might reconcile these two hypotheses.

(22)

22

The picture above shows a simplified network representation of what seems to be happening when propranolol is administered during reactivation. But again; the same end situation could be attributed to either retrieval failure (gist loses the fear component) or storage failure (reducing strength of connections). They both end up with the same conclusions and might even both contribute to the phenomenon. Clear definitions of both hypotheses are needed to come up with clear predictions that are testable by science.

On a more philosophical note; thinking about memory has often been in the form of thinking about a memory as a unit. Not to negate that this method has yield us a lot of benefits to our knowledge about how we remember and forget, but a more reconstructivist or more plastic view of memory is more consistent with results as those found be the group of Merel Kindt. Any interpretation that views memory as more dynamic instead of static is beneficial to gain insight into these complex matters, especially when dealing with intrusive memories or emotional memory disorders. The dynamic view helps us understand that for instance exposure and prevention therapy in a particular context might not help once the patient is in another context. This knowledge will push us forward to come up with new and better forms of treatment.

4.3 End statement

For us researchers it is imperative that if we want to better understand the boundary conditions of this memory unit (or the diffuse distribution of the memory), the boundary conditions of (re)activation of the memory and the subsequent (re)consolidation of the memory, we need operational definitions (philosophy), standard ways of testing memory (cognitive psychology), an increased understanding of the molecular underpinnings of memory (neurobiology) and a systems perspective on the link between molecular changes and changes in memory (systems neuroscience).

This involves interdisciplinary research and cooperation between different fields within research. Traditional memory researchers within cognitive psychology could benefit by increasing their understanding of molecular processes that underlie memory formation and vice versa memory researchers within neurobiology could benefit by increasing their understanding of the different memory systems that exist in human memory. The cross-pollination of fields within memory research will expand our ideas about memory and might yield unknown amounts of new information.

(23)

23

Chapter 5 References

Alberini, C.M. 2005. Mechanisms of memory stabilization: Are consolidation and reconsolidation similar or distinct processes? Trends Neurosci. 28: 51-56.

Alberini, C.M. 2007. Reconsolidation: the samsara of memory consolidation. Debates in Neuroscience, 1: 17-24.

Brewer, M.B. 2007. Social identity and close relationships: What is the connection? Sydney Symp. Soc.

Psych.

Brunet, A., Orr, S.P., Tremblay, J., Robertson, K., Nader, K., and Pitman, R.K. 2008. Effect of post-retrieval propranolol on psychophysiologic responding during subsequent script-driven traumatic imagery in post-traumantic stress disorder. J. of Pysch. Res., 42: 503-506.

Butler, H.E. 1920. Quintilianus, Marcus Fabius. Institutio Oratoria. Trans.. Loeb Classical Library. Cambridge: Harvard University press,

Cahill, L., Prins, B., Weber, M., and McGaugh. 1994. Beta-adrenergic activation and memory for emotional events. Nature, 371: 702-704.

Cahill, L. and McGaugh, J.L. 1996 Modulation of memory storage. Curr. Op. in Neurobiol., 6: 237-242. CBS Statline (2012) Statistics on cause of death. Available at http://statline.cbs.nl (accessed 6 July 2012). Davis, H.P. and Squire, L.R. 1984. Protein synthesis and memory: a review. Psychol. Bull., 96: 518-559. Dudai, Y. 2012. The restless engram: consolidations never end. Annu. Rev. Neurosci., 35: 227-247. Hamilton, E. 1961. Plato: The collected dialogues. Bollingen Foundation, NY.

Hamm, A.O. and Weike, A.I. 2005. The neuropsychology of fear learning and fear regulation. Int. J. of

Psychoph., 57: 5-14.

Hebb, D.O. 1949. The organization of behavior: A neuropsychological theory. Wiley, New York. Hinderliter, C.F., Webster, and T., Riccio, D.C. 1975. Amnesia induced by hypothermia as a function of treatment-test interval and recooling in rats. Anim. Learn. Behav. 3: 257:263.

Hurlemann, R., Walter, H., Rehme, A.K., Kukolja, J., Santoro, S.C., Schmidt, C., Schnell, K., Musshoff, F., Keysers, C., Maier, W., Kendrick, K.M., and Onur, O.A. 2010. Human amygdala reactivity is diminished by the ß-noradrenergic antagonist propranolol. Psych. Med. 1-10.

Judge, M.E. and Quartermain, D. 1982. Characteristics of retrograde amnesia following reactivation of memory in mice. Physiol. Behav., 28: 535-590.

Kindt, M., Soeter, M., and Vervliet, B. 2009. Beyond extinction: erasing human fear responses and preventing the return of fear. Nature Neuroscience, 12: 256-258.

Kraus, M., Schicknick, H., Wetzel, W., Ohl, F., Staak, S., and Tischmeyer, W. 2002. Memory consolidation for the discrimination of frequency-modulated tones in mongolian gerbils is sensitive to protein-synthesis inhibitors applied to the auditory cortex. Learn. Mem., 9: 293-303.

McGaugh, J.L. 1966. Time dependent processes in memory storage. Science, 153: 1351-1358.

McGaugh, J.L., Introini-Collison, I.B., Cahill, L.F., Castellano, C., Dalmaz, C., Parent, M.B., and Williams, C.L. 1993. Neuromodulatory systems and memory storage: role of the amygdala. Behav. Brain Res., 58: 81-90. McGaugh, J.L. 1999. The perseveration-consolidation hypothesis: Müller and Pilzecker, 1900. Brain Res.

Bull., 50: 445-446.

McGaugh, J.L. 2000. Memory: A century of consolidation. Science, 287: 248-251.

Milner, B. 1965. Memory disturbances after bilateral hippocampal lesions. In Cognitive Processes and the

brain, Princeton, NJ.

Misanin, J.R., Miller, R.R., and Lewis, D.J. 1968. Retrograde amnesia produces by electroconvulsive shock after reactivation of a consolidated memory trace. Science, 160: 554-555.

Müller, G.E. and Pilzecker, A. 1900. Experimentelle Beitrage zur Lehre vom Gedachtniss. Zeit. Für Psych., 1: 1-288.

(24)

24

Nadel, L. and Moscovitch, M. 1997. Memory consolidation, retrograde amnesia and the hippocampal complex. Curr. Opin. Neurobiol., 7: 217-227.

Nader, K., and Hardt, O. 2009. A single standard for memory: the case for reconsolidation. Nat. Rev.

Neurosci., 10: 224-234.

Nader, K., Schafe, G.E., and LeDoux, J.E., 2000. Fear memories require protein synthesis in the amygdala for reconsolidation after retrieval. Nature, 406: 722-726.

O’Carroll, R.E., Drysdale, E., Cahill, L., Shajahan, P., and Ebmeier, K.P. 1999a. Memory for emotional material: a comparison of central versus peripheral blockade. J. Physchopharmacol., 13: 32-39. O’Carroll, R.E., Drysdale, E., Cahill, L., Shajahan, P., and Ebmeier, K.P. 1999b. Stimulation of the

noradrenergic system enhances and blockade reduces memory for emotional material in man. Psychol.

Med., 29: 1083-1088.

Oxford Dictionaries. 2010. University Press.

Pedreira, M.E., Perez-Cuesta, L.M, and Maldonado, H. 2002. Reactivation and reconsolidation of long-term memory in the crab Chasmagnathus: protein synthesis requirement and mediation by NMDA-type glutamatergic receptors. The Journal of Neuroscience, 22: 8305-8311.

Pedreira, M.E., Perez-Cuesta, L.M., and Maldonado, H. 2004. Mismatch between what is expected and what actually occurs triggers memory reconsolidation or extinction. Learn. Mem., 11: 579-585.

Polster, M.R., Nadel, L., and Schacter, D. 1991. Cognitive neuroscience analyses of memory: a historical perspective. J. of Cog. Neurosci., 3: 95-116.

Przybyslawski, J. and Sara, S.J. 1997. Reconsolidation of memory after its reactivation. Behav. Brain Res. 84: 241-246.

Ribot, T. 1892. Memory. In A dictionary of psychological medicine, Blakiston, PHY.

Riccio, D.C., Millin, P.M., and Bogart, A.R. 2006. Reconsolidation: A brief history, a retrieval view, and some recent issues. Learn. Mem., 13: 536-544.

Schacter, D. and Tulving, E. 1994 Memory Systems 1994, MIT Press, Cambridge.

Schiller, D., Monfils, M.H., Raio, C.M., Johnson, D.C., LeDoux, J.E., and Phelps, E.A. 2010. Preventing the return of fear in humans using reconsolidation update mechanisms. Nature, 463: 49-53.

Scoville, W.B. and Milner, B. 1957. Loss of recent memory after bilateral hippocampal lesions. J.Neurol.

Neurosurg. Psychiatry, 20: 11-21.

Soares, J.J.F. and Öhman, A. 1993. Backward masking and skin conductance responses after conditioning to nonfeared but fear-relevant stimuli in fearful subjects. Psychophysiology, 30: 460-466.

Soeter, M. and Kindt, M. 2010. Dissociating response systems: Erasing fear from memory. Neurobiology of

Learning and Memory, 94: 30-41.

Soeter, M. and Kindt, M. 2011a. Disrupting reconsolidation: pharmacological and behavioral manipulations. Learn. Mem., 18: 357-366.

Soeter, M. and Kindt, M. 2011b. Noradrenergic enhancement of associative fear memory in humans.

Neurobiol. Learn. Mem., 96: 263-271.

Soeter, M. and Kindt, M. 2012. Stimulation of the noradrenergic system during memory formation impairs extinction learning but not the disruption of reconsolidation. Neuropsychopharmacology, 37: 1204-1215. Squire, L.R., Cohen, N.J., and Nadel, L. 1984. The medial temporal region and memory consolidation: a new hypothesis. In Memory Consolidation, Hillsdale, NJ.

Squire, L.R. 2004. Memory systems of the brain: A brief history and current perspective. Neurobio. of

Learning and Mem., 82: 171-177.

Thomson Reuters Science Web of Knowledge (2012) Available at http://www.webofknowledge.com (accessed on 23 June 2012).

Weike, A.I., Schupp, H.T., and Hamm, A.O. 2007. Fear acquisition requires awareness in trace but not delay conditioning. Psychophysiology, 44: 170-180.

Propranolol and Reconsolidation Theory

Table of contents