Amygdala, Pain Processing and Behavior in Animals

Synonyms

Amygdaloid Complex; nociceptive processing in the amygdala, behavioral and pharmacological studies

Definition

The amygdala is an almond shaped structure in the ventromedial temporal lobe that constitutes part of the brain’s limbic system. It comprises several neuroanatomically and functionally distinct nuclei with widespread connections to and from a variety of cortical and subcortical brain regions.

Characteristics

In a general sense, the amygdala plays a prominent role in the coordination of defense reactions to environmental threats (LeDoux 2003). The hypothesized role of the amygdala in emotional information processing represents one component in this overall role. Clearly, environmental threats are diverse and include the animate (e.g. extraspecies predators, intraspecies rivals) and inanimate (e.g. thorns or spines on plants). Stimuli signaling the presence of threats can be “natural” elicitors of the psychological state of fear such as a sudden, novel sound or the presence of a larger animal. Or previously “neutral” stimuli (discrete sensory cues or distinct environmental contexts) can come to elicit defense reactions following occasions in which they coincided in time with an occurrence of injury or the presence of a natural threat (i.e. through classical conditioning processes). Such “conditioned” stimuli can elicit either acute fear or the qualitatively different state of anxiety, which is a more future-oriented psychological state that readies the animal for a potential environmental threat.

The amygdala is well connected to coordinate reactions to stimuli that signal potential danger. By way of incoming neuroanatomical connections to its central and basolateral subdivisions, the amygdala receives information from the organism’s internal environment (viscerosensation) and information from the external environment consisting of simple sensory inputs and complex multi-sensory perceptions. This information already has already been highly processed by various subcortical and cortical brain structures (e.g. cortical sensory association areas) but the amygdala serves the purpose of attaching emotional significance to the input. By way of its outgoing neuroanatomical connections, the amygdala communicates with brain areas involved in motor preparation / action and autonomic responses. When sensory information arrives relating to environmental danger, the amygdala probably is involved both in the generation of emotional states (e.g. fear, anxiety) and the coordination of appropriate autonomic and behavioral changes that enhance the chance of survival (e.g. defensive fight or flight, subsequent avoidance behaviors, submissive postures, tonic immobilization, autonomic arousal and hypoalgesia or hyperalgesia).

Since pain can signal injury or the potential for injury, it should not be surprising that the processing of nociceptive information by the amygdala can be one of the triggers of these events. Electrophysiological studies show that individual amygdala neurons, particularly in the central nucleus of the amygdala (CeA), respond to brief nociceptive thermal and mechanical stimulation of the skin and or nociceptive mechanical stimulation of deeper (knee joint) tissue (Bernard et al. 1996; Neugebauer et al. 2004). Many CeA neurons have large receptive fields, with some neurons being excited by and others inhibited by nociceptive stimulation. The lateral capsular and, to a lesser extent, the lateral division of the CeA have been termed the “nociceptive amygdala” and receive nociceptive input from lamina I of the spinal and trigeminal dorsal horns. This lamina I input arrives at the CeA via several different routes (Gauriau and Bernard 2002): 1) indirectly, from relays in the lateral and external medial areas of the brainstem parabrachial complex (lamina I → PB → CeA), 2) indirectly, from the posterior triangular nucleus of the thalamus (PoT) to the amygdalostriatal transition area (AStr), which overlaps partly with the CeA (lamina I → PoT → AStr / CeA), 3) indirectly, from the insular cortex by way of the PoT (lamina I → PoT → IC → CeA) and 4) to a much lesser extent, from direct, monosynaptic projections (lamina I → CeA). The basolateral complex of the amygdala also probably receives highly processed nociceptive information from unimodal and polymodal sensory areas of the cerebral cortex (Shi and Cassell 1998).

Human functional neuroimaging studies have supported a role for the amygdala in nociceptive processing by correlating changes in neural activity in the amygdala with the perception of brief painful stimuli. In a manner analogous to the different responses of individual CeA neurons described above, presentation of a painful thermal stimulus to skin of healthy human subjects can result in increases or decreases in neural activity in the amygdala as measured by positron emission tomography (PET) or functional magnetic resonance imaging (fMRI), depending on the stimulation parameters employed. These changes appear to be linearly related to stimulus intensity (Bornhovd et al. 2002; Derbyshire et al. 1997).

In addition to brief pain, neuroplastic changes in amygdala neurons may contribute to the induction and maintenance of chronic pain states. Rodent studies utilizing indirect measures of neuronal activation in the forebrain (e.g. immediate early gene expression or changes in regional cerebral blood flow) have suggested increases in neural activity in the amygdala that correlate with behavioral indices of persistent pain. Several groups have analyzed patterns of Fos protein-like immunoreactivity (Fos-LI) in the rat forebrain after hind paw injection of formalin (i.e. the formalin test). The formalin test involves injecting a small volume of dilute formalin into a hind paw, resulting in an array of pain-related behaviors (paw lifting, licking and flinching) that persists for 11/2–2 h. Behavioral indices of formalin-induced nociception correlate with appearance of Fos-LI in the basolateral amygdala (Nakagawa et al. 2003). Fos-LI also appears in the basolateral amygdala and CeA following stimulation of the trigeminal receptive field in conscious rats with capsaicin (Ter Horst et al. 2001) or after prolonged, nociceptive colonic distension (Monnikes et al. 2003). In a rat model of neuropathic pain (the chronic constriction injury, or CCI, model), a significant increase in regional cerebral blood flow (rCBF) is seen in the basolateral amygdala after 8 or 12 weeks, but not 2 weeks following CCI surgery (Paulson et al. 2002).

The response characteristics of individual CeA neurons have been studied in vivo in rats with or without experimental arthritis in a knee joint (Neugebauer et al. 2004). Prolonged nociception produced by injection of carrageenan and kaolin into the knee joint results in enhancement of both receptive field size and responsiveness to mechanical stimulation of a subset of CeA neurons. Infusion, by microdialysis, of a selective NMDA receptor antagonist (AP5) or an mGluR1 receptor antagonist (CPCCOEt) into the CeA inhibits the increased responses to nociceptive and normally innocuous mechanical stimuli more potently in the arthritic vs. the control condition. By contrast, infusion of a non-NMDA (AMPA / kainate) receptor antagonist (NBQX) or an mGluR5 receptor antagonist (MPEP) inhibits background activity and evoked responses under both normal control and arthritic conditions. These data suggest a change in mGluR1 and NMDA receptor function and activation in the amygdala during pain-related sensitization, whereas mGluR5 and non-NMDA receptors probably are involved in brief as well as prolonged nociception.

In vitro brain slice electrophysiology has provided additional insights (Neugebauer et al. 2004). It is possible to study properties of synaptic transmission (using whole-cell patch-clamp recordings) in brain slices taken from control rats vs. rats with persistent pain. In the nociceptive CeA of such rats, it is possible to study monosynaptic excitatory post-synaptic currents (EPSCs) evoked by electrical stimulation of afferents from the parabrachial complex or from the basolateral amygdala. In rats with experimental arthritis, enhanced synaptic transmission (larger amplitude of evoked monosynaptic EPSCs) is observed at both the nociceptive PB-CeA synapse and the polymodal (including nociceptive) BLA-CeA synapse as compared with control rats. CeA neurons from arthritic rats also develop an increase in excitability. Induction of experimental colitis (by intra-colonic injection of zymosan) produces similar effects, except for the fact that enhanced synaptic transmission is observed only at the nociceptive PB-CeA synapse. In the arthritis model, synaptic plasticity in the amygdala is accompanied by an increase in presynaptic mGluR1 function. Both the selective mGluR1 antagonist CPCCOEt and the group III mGluR agonist LAP4 decrease the amplitude of EPSCs more potently in CeA neurons from arthritic rats than in control animals. The selective group III mGluR antagonist UBP1112 reverses the inhibitory effect of LAP4. During the application of LAP4, paired-pulse facilitation was increased, while no significant changes in slope conductance and action potential firing rate of CeA neurons were observed. These data suggest that presynaptic mGluR1 receptors and group III mGluRs regulate synaptic plasticity in the amygdala in a rat model of arthritis.

Human neuroimaging studies have provided additional supporting evidence by correlating changes in neural activity in the amygdala with the perception of persistent pain. In patients suffering from irritable bowel syndrome (IBS), Wilder-Smith et al. (2005) demonstrated a bilateral decrease in neural activity in the amygdala during episodes of experimentally induced rectal pain.

Neuroimaging techniques, measurement of immediate early gene responses and in vivo electrophysiological studies are useful for identifying brain regions with activity that co-varies with the presence or absence of pain or nociception, but such studies are limited with respect to mechanistic insights and determining cause vs. effect. On the contrary, rodent behavioral studies have been highly informative in this regard. Such studies provide evidence that the amygdala is involved in encoding the affective or aversive component of pain. Hebert et al. (1999) used an alley-shaped apparatus with an array of protruding, sharp pins situated in the middle of the alley to investigate this issue. During 10 min test sessions, the behavioral patterns of normal rats were characterized by voluntary contact with the pins followed by periods of avoidance and risk assessment (referred to by the investigators as “stretch attend” and “stretch approach” behaviors). Of the group of normal rats tested, few actually crossed the array of pins. In contrast, rats with bilateral lesions of the amygdala showed a significant increase in both the number of crossings of the pin array and the amount of time spent on the pins as compared with normal rats. The results suggest that the aversive quality of the painful mechanical stimulation imparted by the pin array is encoded at least partly by the amygdala.

The affective / aversive quality of pain in rodents also has been studied using a variation of the place-conditioning paradigm. In 2001, Johansen et al. introduced the formalin-induced condition place avoidance model (F-CPA). By pairing the experience of formalin-induced pain with a distinct environmental context / compartment within a place-conditioning apparatus, the investigators hoped to establish a behavioral endpoint that is directly related to the negative affective component of pain. After two pairings of formalin-induced pain (1 h) with the compartment, rats learned to avoid the compartment and spend most of their time in the other two compartments of the apparatus. Lesions of the rostral anterior cingulate cortex (rACC) blocked the acquisition of F-CPA but did not affect the expression of acute formalin-induced pain behaviors (paw lifting, paw licking, etc.). The results suggested that the rACC lesions reduced the affective salience, but not the sensory-discriminative component of formalin-induced pain (Johansen et al. 2001). Using the F-CPA model, a similar pattern of results was obtained after bilateral lesions of the either the CeA or basolateral amygdala (Tanimoto et al. 2003). The results provide strong causal data suggesting that the processing of nociceptive information in the amygdala and rACC relates to encoding of the affective component of pain. Furthermore, the results fit with the role in defense reactions ascribed to the amygdala at the beginning of this essay. By attaching emotional significance to a stimulus signaling danger (in this case the pain associated with formalin), the amygdala sets the stage for coordination of appropriate acute and delayed responses to the stimulus by way of its multitude of connections with other brain regions and neural circuitry (Fig. 1). These responses include acute protective behaviors and autonomic responses followed by avoidance of the environment in which the pain was experienced.

Figure 1

A simplified illustration of major nociceptive pathways to the amygdala and possible consequences of stimulation of these pathways. Abbreviations: IC, insular cortex; PB, parabrachial complex; PoT, posterior triangular nucleus of the thalamus.