The Contribution of the Amygdala to Aversive and Appetitive Pavlovian Processes

Abstract

Pavlovian cues predict the occurrence of motivationally salient outcomes, thus serving as an important trigger of approach and avoidance behavior. The amygdala is a key substrate of Pavlovian conditioning, and the nature of its contribution varies by the motivational valence of unconditioned stimuli. The literature on aversive Pavlovian learning supports a serial-processing model of amygdalar function, while appetitive studies suggest that Pavlovian associations are processed through parallel circuits in the amygdala. It is proposed that serial and parallel forms of information processing can be attributed to differential recruitment of amygdalar nuclei, with emphasis placed on the lateral amygdala.

Keywords

amygdala appetitive aversive motivation Pavlovian conditioning

Motivational states direct and energize animal behavior (Bindra, 1969; Hull, 1943), and are generally considered along a bivalent continuum, spanning from appetitive to aversive (Bindra, 1969). Approach and avoidance behaviors, designed to attain appetitive goal states or minimize aversive outcomes, are often guided by environmental cues (Berridge, 1999; Bindra, 1969), which can acquire their motivational valence through a process of Pavlovian conditioning. This form of learning occurs when a previously neutral conditioned stimulus (CS), such as a light or a tone, comes to predict an unconditioned stimulus (US) with an inherent motivational value, such as a food morsel or a foot shock. While Pavlov (1927) argued that conditioning results from a temporal coincidence between CS and US, subsequent scholarship indicates that the CS must carry information (i.e., reduce uncertainty) about the timing and likelihood of US delivery (Gallistel, 2003; Rescorla, 1988).

Thus, Pavlovian learning allows animals to develop expectancies about motivationally relevant outcomes, and then to determine the appropriate preparatory response. This review focuses on the acquisition of Pavlovian associations and the amygdala circuitry that underpins conditioning. Interestingly, the neural substrates of Pavlovian learning vary by the motivational valence of the US (appetitive or aversive). The acquisition of aversive associations relies on the serial progression of information through the amygdala, while evidence from appetitive studies seems to support a parallel-processing model. The following sections characterize the behavioral function of freezing, a common readout of aversive Pavlovian learning, and then contrast the neural circuitry of aversive and appetitive conditioning.

Pavlovian Processes and Innate Defensive Behavior

The Ledoux lab has long studied a model of auditory Pavlovian threat conditioning (for review, see Johansen, Cain, Ostroff, & LeDoux, 2011), in which a tone CS predicts a foot shock US and thus elicits robust freezing (immobility and the adoption of a hunched posture). It is important to note that freezing is an innate behavior (Blanchard & Blanchard, 1969; Blanchard, Blanchard, & Griebel, 2005); Pavlovian processes teach the animal when to deploy this response, but not its value as a defensive measure. Freezing is thought to be a component of the rodent’s unlearned repertoire of defenses against predation. Contact with a predator will elicit strong freezing from rats (Blanchard, Blanchard, Rodgers, & Weiss, 1990), as will the mere presentation of a novel predator odor (Endres & Fendt, 2009; Wallace & Rosen, 2000). In a natural setting, freezing occurs in the interval prior to a predatory attack, when the prey animal has noticed a predator, but has gone unnoticed by its potential attacker (Elliam, 2005; Fanselow & Lester, 1988). Freezing makes the animal less conspicuous and more likely to avoid detection, and is replaced by fleeing if and when the predator strikes (Elliam, 2005).

Freezing and other species-typical defensive reactions are considered a default response to a wide array of aversive stimulation (Bolles, 1970; Fanselow & Lester, 1988), which is why they can be elicited by artificial stimuli in a laboratory setting. Indeed, it seems that presentation of an aversive CS activates the behavior that occurs prior to an imminent predatory strike: immobility or freezing. The increase in muscle tension also prepares the rat to flee when the predicted US is presented, much as animals flee when predatory attack actually occurs. It has been proposed that Pavlovian learning allows rodents to instate oncoming threats in a predatory imminence continuum (Fanselow & Lester, 1988). Thus, rapidly acquired associations provide information about the proximity of an aversive event or outcome, allowing the animal to deploy the appropriate innate defensive strategy, such as CS-evoked freezing in the interval prior to US delivery (Bolles, 1970; Fanselow & Lester, 1988).

This complex and vital form of processing is underpinned by an elaborate neural architecture in the amygdala. The next section describes how a constellation of amygdalar nuclei acquire aversive Pavlovian information, and how these associations are used to coordinate important defensive reactions, such as conditioned freezing.

Aversive Conditioning and the Amygdala

The lateral and central nuclei of the amygdala (Figure 1) play a crucial role in the acquisition phase of Pavlovian threat conditioning. The lateral nucleus (LA) contains multimodal neurons that respond to auditory and somatosensory stimulation (Romanski, Clugnet, Bordi, & LeDoux, 1993), and are thus well situated to acquire an association between an auditory CS and a foot shock US. Aversive Pavlovian learning has strong correlates in LA activity, and conditioned freezing to an auditory CS develops in parallel with conditioned increases in LA cell firing (Repa et al., 2001). Threat conditioning also causes spiking in LA neurons to synchronize within a specific frequency range (Pare & Collins, 2000; Quirk, Repa, & LeDoux, 1995), suggesting that the predictive valence of the CS is encoded by both the rate and rhythm of action potentials in LA (Maren & Quirk, 2004). Extinction training attenuates CS-evoked behavior and returns LA activity to baseline levels, with the exception of a subpopulation in ventral LA that continues to show conditioned responses to the CS (Repa et al., 2001). These cells are thought to carry an indelible element of the memory trace that allows for the rapid return of conditioned reactions after extinction.

Figure 1.

Anatomy of the amygdala; arrows represent connections between and within amygdalar nuclei. BA, basal nucleus; CeA, central nucleus, with lateral (L) and medial (M) subnuclei; ICM, intercalated cell masses; LA lateral nucleus.

LA projects to a variety of amygdalar nuclei, and the central nucleus (CeA) receives a good deal of nonreciprocal, convergent input from within the amygdala (Pitkanen, 2000), suggesting that conditioned information originates in LA and flows forward to CeA. In support of this model, pharmacological inactivation of CeA does not prevent LA neurons from developing conditioned responses, indicating that LA does not require CeA to form aversive CS–US associations (Goosens, Hobin, & Maren, 2003). Complementary studies demonstrate that inactivation of LA (Wilensky, Schafe, Kristense, & LeDoux, 2006), selective lesions of LA (Goosens & Maren, 2001; Nader, Majidishad, Amorapanth, & LeDoux, 2001), or lesions of the entire basolateral complex (BLA; Maren, 1999) block the normal formation of an aversive association with a CS. It is important to note that pretraining lesions of the basal nucleus do not impact the initial acquisition of aversive Pavlovian associations (Anglada-Figueroa & Quirk, 2005; Nader et al., 2001), even though posttraining lesions and inactivations of the basal nucleus do have a selective effect on the expression of previously acquired associations (Amano, Duvarci, Popa, & Pare, 2011; Anglada-Figueroa & Quirk, 2005). As such, it seems acquisition deficits caused by BLA lesions depend critically on damage to LA. In a direct test of the functional anatomy, asymmetric lesion experiments demonstrate that threat conditioning does not proceed if the connection between BLA and CeA is severed (Jimenez & Maren, 2009). Thus, aversive Pavlovian learning relies on a serial form of information processing in the amygdala, with LA playing a key role in the acquisition of CS–US associations.

In spite of this serial-processing perspective, it is important to emphasize that CeA also participates in the acquisition phase of aversive Pavlovian learning (Wilensky et al., 2006), though LA and CeA make dissociable contributions to this process. Like LA, CeA neurons develop conditioned responses over the course of Pavlovian threat conditioning, displaying a varied profile of unit activity (Duvarci, Popa, & Pare, 2011). The lateral subnucleus of CeA contains distinct populations described as CS-on and CS-off cells (Ciocchi et al., 2010). Excitations among CS-on cells are thought to inhibit the CS-off population, which causes the disinhibition of the brainstem-projecting neurons in medial CeA that are critical for CS-evoked freezing (Ciocchi et al., 2010; Duvarci et al., 2011). Thus, freezing to the CS is driven by a complex disinhibitory process in CeA.

In contrast, CeA also mediates CS-evoked exploration and risk assessment, which occur at an earlier phase of the predatory imminence continuum, when a predicted threat is quite distal (Fanselow & Lester, 1988). Projection cells in lateral CeA mediate this behavior by disinhibiting basal forebrain cholinergic neurons (Gozzi et al., 2010). This lateral CeA population also inhibits medial CeA (Gozzi et al., 2010), and may conform to the CS-off cells that are inactivated under conditions that promote freezing. Exploration and immobility, then, may be driven by opposite patterns of CeA activity, suggesting that the exact nature of the reaction to the CS is determined by the function of CeA microcircuits.

Together, these data suggest that LA represents the relationship between a CS and an aversive US, while CeA evaluates conditioned information and determines the appropriate behavioral response to imminent danger. A reasonable proposal, then, is that LA acquires and stores memory for the predictive valence of the CS, while CeA acquires and stores memory linking a CS to its appropriate profile of defensive reactions. This putative arrangement preserves the elements of the serial-processing model, insofar as it requires the flow of information from LA, while ascribing both LA and CeA with active roles in the process of aversive Pavlovian learning.

Killcross, Robbins, and Everitt (1997) provide a compelling counterpoint to this serial-processing perspective. Using a complex instrumental procedure, they demonstrate a double dissociation between BLA lesions, which prevent animals from avoiding a lever punished intermittently with aversive CS–US pairings, and CeA lesions, which prevent the conditioned suppression of lever-press during the presentation of an aversive CS. From this the authors conclude that BLA and CeA function independently, thus de-emphasizing the serial flow of information through the amygdala. This divergent result may stem from the training protocol employed, which involved an average of 120 CS–US pairings—far more trials than any other aforementioned report. Studies of overtraining in rats with BLA lesions have yielded mixed results. One report demonstrates that animals with BLA lesions can acquire an aversive association with a context, but not an auditory CS, after 75 CS–US pairings (Maren, 1999). However, a study using a comparable training protocol demonstrates that animals with BLA lesions can acquire an association with both a CS and a context, but only after more pairings than intact animals require (Zimmerman, Rabinak, McLachlan, & Maren, 2007). Intriguingly, animals with BLA lesions not only acquired contextual associations more slowly, but also forgot them more quickly (Poulos et al., 2009).

It is possible that multiple systems can encode aversive Pavlovian information, and that the circuitry of serial processing simply happens to be the most efficient at producing a durable memory (Fanselow, 2010). This is a key point, because the expectancies developed by these systems are designed to counter mortal threat in the wild, and slow, incremental learning and short-lived memory fail to provide the same adaptive advantage. As such, the literature on overtraining highlights how the expedient acquisition of lasting aversive associations requires the serial flow of information from LA.

Appetitive Conditioning and the Amygdala

While the data support a serial-progression model in the case of aversion, appetitive Pavlovian processes seem to rely on a qualitatively different form of information processing in the amygdala (Figure 2). The literature on appetitive conditioning suggests that BLA and CeA function in parallel, independently encoding their own distinct representations of a given CS–US association (Balleine & Killcross, 2006; Everitt, Cardinal, Parkinson, & Robbins, 2003). BLA is thought to process associations between a CS and the specific elements of an appetitive US, as evidenced by BLA manipulations that block US devaluation (Hatfield, Han, Conley, Gallagher, & Holland, 1996), or hinder US-specific Pavlovian-to-instrumental transfer (Blundell, Hall, & Killcross, 2001; Corbit & Balleine, 2005). In contrast, CeA is thought to process associations between a CS and generalized behavioral reactions that occur regardless of the specific US, as evidenced by CeA manipulations that attenuate the conditioned orienting response (Gallagher, Graham, & Holland, 1990), or block general forms of Pavlovian-to-instrumental transfer (Corbit & Balleine, 2005).

Figure 2.

Serial versus parallel-processing models of amygdala function. Left: Serial model, in which CS–US associations are acquired and stored in LA, and then processed through CeA to produce defensive reactions. Right: Parallel model, in which BLA stores the association between the CS and the specific properties of the US, and the CeA encodes the relationship between the CS and general conditioned reactions, such as orienting.

The appetitive conditioning phenomena that support a parallel-processing model are generally measured at an expression time point, after the association has already been learned. Interestingly, the amygdala plays an unknown role in the initial acquisition of appetitive Pavlovian associations. One study found that CeA lesions attenuate the acquisition of conditioned approach (Parkinson, Robbins, & Everitt, 2000), though several others demonstrate that comparable lesions have no effect on approach learning (Corbit & Balleine, 2005; Gallagher et al., 1990; Han, McMahan, Holland, & Gallagher, 1997). However, CeA lesions do block the development of a conditioned orienting response, while sparing approach (El-Amamy & Holland, 2007; Gallagher et al., 1990; Han et al., 1997). Conditioned orienting is thought to reflect event processing (Holland & Gallagher, 1999) or attention paid to the CS when an error in US prediction occurs in a previous trial (Pearce & Hall, 1980). Commensurate with this, CeA neurons show firing increases when a US is smaller than anticipated, and this activity precedes a potentiated orienting response to subsequent CS presentations (Calu, Roesch, Haney, Holland, & Schoenbaum, 2010). Thus, CeA contributes to CS associability with a mechanism described in learning theory, potentially allowing new information to update previously acquired appetitive associations. However, Pearce-Hall requires a past trial in order for learning to occur (Pearce & Hall, 1980), making it unlikely that this mechanism drives the initial process of acquisition.

Like CeA, lesions of BLA have no effect on the acquisition of conditioned approach behavior (Corbit & Balleine, 2005; Hatfield et al., 1996; Parkinson et al., 2000). In contrast, BLA lesions do prevent appetitive second-order conditioning, in which a previously acquired CS is used to reinforce a novel CS (Hatfield et al., 1996; Setlow, Gallagher, & Holland, 2002). While these lesion data suggest that BLA is selectively involved in higher order appetitive learning processes, a divergent result comes from a recent optogenetic study. Using a light-activated ion channel to rapidly inhibit neuronal activity, Stuber et al. (2011) showed that the BLA projection to the nucleus accumbens is required during CS presentation in order for subjects to acquire a conditioned licking behavior. These disparate results may be attributable to functional accommodation that occurs after tissue damage, but not fast-acting optical stimulation, suggesting that other structures can take over for BLA in the case of lesion. Together, the data suggest that BLA might participate in first-order appetitive Pavlovian learning, but can also be accommodated for if destroyed.

At first glance, these data seem to align with the literature on aversive Pavlovian learning, in which BLA is required for first- and second-order conditioning (Maren, 1999; Parkes & Westbrook, 2010). However, excitotoxic BLA lesions strongly hamper the animal’s ability to acquire an aversive association with a CS, while first-order appetitive associations proceed unimpeded by permanent damage to that structure. Key encoding differences must exist if the brain can easily accommodate the destruction of BLA in one circumstance and not the other.

One possibility is that the two forms of Pavlovian learning recruit a different pattern of nuclei within BLA. Threat conditioning depends on LA, and in particular on its dorsal subregion, which contains cells that acquire fast conditioned responses to an aversive CS (Repa et al., 2001). LA-dependent, aversive Pavlovian conditioning occurs quickly, usually in 1–5 CS–US pairing (Johansen et al., 2011), and is strongly attenuated with damage to the LA or BLA. Conversely, the appetitive Pavlovian processes described in the articles cited here occur more slowly, and proceed unimpeded by permanent BLA damage. Thus, it seems that appetitive Pavlovian learning lacks key features of dorsal LA-dependent conditioning, suggesting that the differential recruitment of this region may explain the information-processing differences observed between appetitive and aversive domains.

This begs the question of why dorsal LA is selectively recruited in the aversive circumstance. It is possible that the difference is due to a motivational asymmetry between the aversive and appetitive stimuli employed in laboratory studies of animal behavior. Food and sucrose solutions used to reinforce appetitive CSs have considerable motivational pull, and gain further value when animals are food restricted to the moderate degree permitted in the lab. However, aversive Pavlovian learning taps into a system designed to counter the existential threat associated with predation (Bolles, 1970; Fanselow & Lester, 1988), and thus engages a profile of behaviors that function to prevent potentially lethal outcomes in the wild. While truly life-threatening forms of hunger and thirst may take motivational priority over cues that signal an imminent threat, it has long been established that aversive CSs can powerfully suppress appetitive behavior, even when animals have been food restricted (Estes & Skinner, 1941). This suggests that common aversive stimuli have a motivational edge over the appetitive stimuli used in neuroscientific experiments of behavioral learning.

Thus, dorsal LA circuitry may function as a rapid acquisition system, which is recruited in circumstances of heightened motivational salience that demand a fast behavioral response. The trade-off for quick learning and durable memory seems to be dependence on LA and an inability to accommodate its loss. Conversely, Pavlovian learning that does not recruit dorsal LA seems to trigger multiple amygdalar representations of the CS–US relationship, distributed across CeA and more ventral regions of BLA. Dorsal LA-independent learning seems more amenable to accommodation after permanent damage to a given structure, as the loss of one circuit can be accounted for by another that functions in parallel. This kind of learning seems to occur more slowly, but is also relatively resilient to amygdala damage.

Conclusions

Pavlovian learning allows organisms to determine the predictive relationship between environmental stimuli and motivationally relevant outcomes. The contribution of amygdalar circuits to this process may be determined by the motivational salience of the stimuli being conditioned. In particular, Pavlovian learning that occurs in highly motivating scenarios may recruit dorsal LA, which rapidly encodes the CS–US relationship and coordinates with CeA to guide conditioned reactions. In this circumstance, fast behavioral learning relies on the serial progression of conditioned information through the amygdala. In contrast, learning that does not recruit dorsal LA may yield multiple representations of CS–US associations that are distributed across parallel amygdalar circuits. While dorsal LA-independent conditioning generally occurs more slowly, a more diffuse representation of Pavlovian information allows for a more ready accommodation of function after damage to the amygdala.

References

Amano

Duvarci

Popa

Pare

(2011). The fear circuit revisited: Contributions of the basal amygdala to conditioned fear. Journal of Neuroscience, 31, 15481–15489.

Anglada-Figueroa

Quirk

G. J.

(2005). Lesions of the basal amygdala block expression of conditioned fear but not extinction. Journal of Neuroscience, 25, 9680–9685.

Balleine

B. W.

Killcross

(2006). Parallel incentive processing: An integrated view of amygdala function. Trends in Neuroscience, 29, 272–279.

Berridge

K. C.

(1999). Pleasure, pain, desire, and dread: Hidden core processes of emotion. In Kahenman

Diener

Schwarz

(Eds.), Well-being (pp. 525–557). New York, NY: Russell Sage Foundation.

Bindra

(1969). A unified interpretation of emotion and motivation. Annals of the New York Academy of Sciences, 159, 1071–1083.

Blanchard

D. C.

Blanchard

R. J.

(1969). Crouching as an index of fear. Journal of Comparative and Physiological Psychology, 67, 370–375.

Blanchard

D. C.

Blanchard

R. J.

Griebel

(2005). Defensive responses to predator threat in the rat and mouse. Current Protocols in Neuroscience, 8, 8.191. doi: 10.1002/0471142301.ns0819s30

Blanchard

R. J.

Blanchard

D. C.

Rodgers

Weiss

S. M.

(1990). The characterization and modeling of antipredator defensive behavior. Neuroscience and Biobehavioral Reviews, 14, 463–472.

Blundell

Hall

Killcross

(2001). Lesions of basolateral amygdala disrupt selective aspects of reinforcer representation in rats. Journal of Neuroscience, 21, 9018–9026.

10.

Bolles

R. C.

(1970). Species-specific defense reactions and avoidance learning. Psychological Review, 77, 32–48.

11.

Calu

D. J.

Roesch

M. R.

Haney

R. Z.

Holland

P. C.

Schoenbaum

(2010). Neural correlates of variations in event processing during learning in central nucleus of amygdala. Neuron, 68, 991–1001.

12.

Ciocchi

Herry

Grenier

Wolff

S. B.

Letzkus

J. J.

Vlachos

Luthi

(2010). Encoding of conditioned fear in central amygdala inhibitory circuits. Nature, 468, 277–282.

13.

Corbit

L. H.

Balleine

B. W.

(2005). Double dissociation of basolateral and central amygdala lesion on the general and outcome-specific forms of Pavlovian-instrumental transfer. Journal of Neuroscience, 25, 962–970.

14.

Duvarci

Popa

Pare

(2011). Central amygdala activity during fear conditioning. Journal of Neuroscience, 31, 289–294.

15.

El-Amamy

Holland

P. C.

(2007). Dissociable effects of disconnecting amygdala central nucleus from ventral tegmental area or substantia nigra on learned orienting and incentive motivation. European Journal of Neuroscience, 25, 1557–1567.

16.

Elliam

(2005). Die hard: A blend of freezing and fleeing as a dynamic defense—Implications for the control of defensive behavior. Neuroscience and Biobehavioral Reviews, 29, 1181–1191.

17.

Endres

Fendt

(2009). Aversion- vs. fear-inducing properties of 2,4,5-trimethyl-3-thiazoline, a component of fox odor, in comparison with those of butyric acid. Journal of Experimental Biology, 212, 2324–2327.

18.

Estes

W. K.

Skinner

B. F.

(1941). Some quantitative effects of anxiety. Journal of Experimental Psychology, 29, 390–400.

19.

Everitt

B. J.

Cardinal

R. N.

Parkinson

J. A.

Robbins

T. W.

(2003). Impact of amygdala-dependent mechanisms of emotional learning. Annals of the New York Academy of Sciences, 985, 233–250.

20.

Fanselow

M. S.

(2010). From contextual fear to a dynamic view of memory systems. Trends in Cognitive Sciences, 14, 7–15.

21.

Fanselow

M. S.

Lester

L. S.

(1988). A functional behavioristic approach to aversively motivated behavior: Predatory imminence as a determinant of the topography of defensive behavior. In Bolles

R. C.

Beecher

M. D.

(Eds.), Evolution and learning (pp. 185–212). Hillsdale, NJ: Lawrence Erlbaum Associates.

22.

Gallagher

Graham

P. W.

Holland

P. C.

(1990). The amygdala central nucleus and appetitive Pavlovian conditioning: Lesions impair one class of conditioned behavior. Journal of Neuroscience, 10, 1906–1911.

23.

Gallistel

C. R.

(2003). Conditioning from an information processing perspective. Behavioural Processes, 62, 89–101.

24.

Goosens

K. A.

Hobin

J. A.

Maren

(2003). Auditory-evoked spike firing in the lateral amygdala and Pavlovian fear conditioning: Mnemonic code of fear bias? Neuron, 40, 1013–1022.

25.

Goosens

K. A.

Maren

(2001). Contextual and auditory fear conditioning are mediated by the lateral, basal and central amygdaloid nuclei in rats. Learning & Memory, 8, 148–155.

26.

Gozzi

Jain

Giovanelli

Bertollini

Crestan

Schwarz

A. J.

Bifone

(2010). A neural switch for active to passive fear. Neuron, 67, 656–666.

27.

Han

J.-S.

McMahan

R. W.

Holland

Gallagher

(1997). The role of an amygdalo-nigrostriatal pathway in associative learning. Journal of Neuroscience, 17, 3913–3919.

28.

Hatfield

Han

J.-S.

Conley

Gallagher

Holland

(1996). Neurotoxic lesions of basolateral, but not central, amygdala interfere with Pavlovian second-order conditioning and reinforce devaluation effects. Journal of Neuroscience, 16, 5256–5265.

29.

Holland

P. C.

Gallagher

(1999). Amygdala circuitry in attentional and representational processes. Trends in Cognitive Sciences, 3, 65–73.

30.

Hull

C. L.

(1943). Principles of behavior. New York, NY: Appleton- Century-Croft.

31.

Jimenez

S. A.

Maren

(2009). Nuclear disconnection within the amygdala reveals a direct pathway to fear. Learning & Memory, 16, 766–768.

32.

Johansen

J. P.

Cain

C. K.

Ostroff

L. E.

LeDoux

J. E.

(2011). Molecular mechanisms of fear learning and memory. Cell, 147, 509–524.

33.

Killcross

Robbins

T. W.

Everitt

B. J.

(1997). Different types of fear-conditioned behavior mediated by separate nuclei within amygdala. Nature, 388, 377–380.

34.

Maren

(1999). Neurotoxic basolateral amygdala lesions impair learning and memory but not the performance of conditional fear in rats. Journal of Neuroscience, 19, 8696–8703.

35.

Maren

Quirk

G. J.

(2004). Neuronal signaling of fear memory. Nature Reviews Neuroscience, 5, 844–852.

36.

Nader

Majidishad

Amorapanth

LeDoux

J. E.

(2001). Damage to the lateral and central, but not other amygdaloid nuclei prevents the acquisition of auditory fear conditioning. Learning & Memory, 8, 156–163.

37.

Pare

Collins

D. R.

(2000). Neuronal correlates of fear in the lateral amygdala: Multiple extracellular recordings in conscious cats. Journal of Neuroscience, 20, 2701–2710.

38.

Parkes

S. L.

Westbrook

R. F.

(2010). The basolateral amygdala is critical for the acquisition and extinction of associations between a neutral stimulus and a learned danger signal but not between two neutral stimuli. The Journal of Neuroscience, 30, 12608–12618.

39.

Parkinson

J. A.

Robbins

T. W.

Everitt

B. J.

(2000). Dissociable roles of the central and basolateral amygdala in appetitive emotional learning. European Journal of Neuroscience, 12, 405–413.

40.

Pavlov

I. P.

(1927). Conditioned reflexes. London, UK: Routledge and Kegan Paul.

41.

Pearce

J. M.

Hall

(1980). A model for Pavlovian learning: Variations in the effectiveness of conditioned but not unconditioned stimuli. Psychological Review, 87, 532–552.

42.

Pitkanen

(2000). Connectivity of the rat amygdaloid complex. In Aggleton

J. P.

(Ed.), The amygdala (pp. 31–115). Oxford, UK: Oxford University Press.

43.

Poulos

A. M.

Sterlace

S. S.

Tokushige

Ponnusamy

Fanselow

M. S.

(2009). Persistence of fear memory across time requires the basolateral amygdala complex. Proceedings of the National Academy of Sciences, 106, 11737–11741.

44.

Quirk

G. J.

Repa

J. C.

LeDoux

J. E.

(1995). Fear conditioning enhances short-latency auditory responses of lateral amygdala neurons: Parallel recordings in the freely behaving rat. Neuron, 15, 1029–1039.

45.

Repa

J. C.

Muller

Apergis

Desrochers

T. M.

Zhou

LeDoux

J. E.

(2001). Two different lateral amygdala cell populations contribute to the initiation and storage of memory. Nature, 4, 724–731.

46.

Rescorla

R. A.

(1988). Behavioral studies of Pavlovian conditioning. Annual Review of Neuroscience, 11, 329–352.

47.

Romanski

L. M.

Clugnet

M.-C.

Bordi

LeDoux

J. E.

(1993). Somatosensory and auditory convergence in the lateral nucleus of the amygdala. Behavioral Neuroscience, 107, 444–450.

48.

Setlow

Gallagher

Holland

P. C.

(2002). The basolateral complex of the amygdala is necessary for acquisition but not expression of CS motivational value in Pavlovian second-order conditioning. European Journal of Neuroscience, 15, 1841–1853.

49.

Stuber

G. D.

Sparta

D. R.

Stamatakis

A. M.

van Leeuwen

W. A.

Hardjoprajitno

J. E.

Cho

Bonci

(2011). Excitatory transmission from the amygdala to nucleus accumbens facilitates reward seeking. Nature, 475, 377–382.

50.

Wallace

K. J.

Rosen

J. B.

(2000). Predator odor as an unconditioned fear stimulus in rats: Elicitation of freezing by trimethyl- thiazoline, a compound in fox feces. Behavioral Neuroscience, 114, 912–922.

51.

Wilensky

A. E.

Schafe

G. E.

Kristense

M. P.

LeDoux

J. E.

(2006). Rethinking the fear circuit: The central nucleus of the amygdala is required for the acquisition, consolidation and expression of Pavlovian fear conditioning. Journal of Neuroscience, 26, 12387–12396.

52.

Zimmerman

J. M.

Rabinak

C. A.

McLachlan

I. G.

Maren

(2007). The central nucleus of the amygdala is essential for acquiring and expressing conditional fear after overtraining. Learning & Memory, 14, 634–644.