Effects of context-drug learning on synaptic connectivity in the basolateral nucleus of the amygdala in rats

Abstract

Context-drug learning produces structural and functional synaptic changes in the circuitry of the basolateral nucleus of the amygdala (BLA). However, how the synaptic changes translated to the neuronal targets was not established. Thus, in the present study, immunohistochemistry with a cell-specific marker and the stereological quantification of synapses was used to determine if context-drug learning increases the number of excitatory and inhibitory/modulatory synapses contacting the gamma-aminobutyric acid (GABA) interneurons and/or the pyramidal neurons in the BLA circuitry. Amphetamine-conditioned place preference increased the number of asymmetric (excitatory) synapses contacting the spines and dendrites of pyramidal neurons and the number of multisynaptic boutons contacting pyramidal neurons and GABA interneurons. Context-drug learning increased asymmetric (excitatory) synapses onto dendrites of GABA interneurons and increased symmetric (inhibitory or modulatory) synapses onto dendrites but not perikarya of these same interneurons. The formation of context-drug associations alters the synaptic connectivity in the BLA circuitry, findings that have important implications for drug-seeking behavior.

Keywords: addiction; associative learning; basolateral amygdala; electron microscopy; glutamic acid decarboxylase; physical disector.

Figure 1

Low magnification images depicting GAD+ neurons and the boundaries of the BLA. a, A photomicrograph of a section stained with cresyl violet. The boundaries of the BLA are outlined by the black line. The scale bar equals 1 mm. b, A photomicrograph of a section immunoreacted for GAD 65+67 then processed for electron microscopy. The arrows point to GAD+ neurons in the BLA. The boundaries of the BLA are outlined by the black line. The scale bar equals 500 μm.

Figure 2

Behavioral conditioning with AMPH produced AMPH CPP. Animals treated with AMPH and immediately paired with a specific chamber spent more time in the AMPH-paired than the unpaired chamber during the CPP test (n=6 rats). Animals treated with the delayed pairing regimen (n=5 rats) or with saline alone (n=4 rats) did not exhibit a preference for one chamber over the other during the CPP test. Data are expressed as the mean ± S.E.M. * P < 0.05 versus the delayed pairing and saline groups.

Figure 3

AMPH CPP increased the number of excitatory synapses. a, An electron micrograph depicting a GAD− terminal forming an asymmetric synapse with a GAD− spine. b, An electron micrograph depicting a GAD− terminal forming an asymmetric synapse with a GAD+ spine. c, An electron micrograph depicting a GAD− terminal forming an asymmetric synapse with a GAD− dendrite. d, An electron micrograph depicting two GAD− terminals forming asymmetric synapses with the same GAD+ dendrite. Also depicted is a GAD− terminal forming an asymmetric synapse with a GAD− spine. e, The total number of asymmetric synapses and the number of asymmetric axospinous and axodendritic synapses were greater for animals in the AMPH CPP group compared to animals in the delayed pairing and saline groups. There were no differences in the total number of asymmetric synapses and the number of asymmetric axospinous and axodendritic synapses for rats in the delayed pairing group compared to the saline group. Data are expressed as the mean ± S.E.M. f, The number of GAD− terminals with asymmetric synapses contacting GAD− spines was significantly increased in the AMPH CPP group compared with the delayed pairing and saline control groups. There was no difference in the number of GAD− terminals with asymmetric synapses contacting GAD− spines for rats in the delayed pairing group compared to the saline group. g, The number of GAD− terminals with asymmetric synapses contacting GAD− dendrites and the number of GAD− terminals with asymmetric synapses contacting GAD+ dendrites was significantly increased in the AMPH CPP group compared with the delayed pairing and saline control groups. There was no difference in the number of GAD− terminals with asymmetric synapses contacting GAD− dendrites and the number of GAD− terminals with asymmetric synapses contacting GAD+ dendrites for rats in the delayed pairing group compared to the saline group. Data are expressed as the mean ± S.E.M. * P < 0.05 versus the delayed pairing and saline groups. Abbreviations: as, asymmetric synapse; sp−, GAD− spine; sp+, GAD+ spine; d−, GAD− dendrite; d+, GAD+ dendrite. The scale bar in d is valid for a-d and equals 250 nm.

Figure 4

AMPH CPP increased the number of inhibitory or modulatory synapses. a, An electron micrograph depicting a GAD− terminal forming a symmetric synapse with a GAD− dendrite. Also depicted is a GAD− spine protruding from a GAD− parent dendrite and a GAD− terminal forming an asymmetric synapse with a GAD− spine. b, An electron micrograph depicting a GAD− terminal forming a symmetric synapse with a GAD+ dendrite. c, An electron micrograph depicting a GAD+ terminal forming a symmetric synapse with a GAD− dendrite. d, An electron micrograph depicting a GAD− terminal forming a symmetric synapse with a GAD+ dendrite. e, An electron micrograph depicting a GAD− terminal forming a symmetric synapse with a GAD+ soma. f, An electron micrograph depicting a GAD+ terminal forming a symmetric synapse with a GAD+ soma. g, An electron micrograph depicting a GAD− terminal forming a symmetric synapse with a GAD− soma. h, The total number of symmetric synapses and the number of symmetric axodendritic but not axospinous or axosomatic synapses were greater for animals in the AMPH CPP group compared to animals in the delayed pairing and saline groups. There were no differences in the total number of symmetric synapses and the number of symmetric axospinous, axodendritic, or axosomatic synapses for rats in the delayed pairing group compared to the saline group. Data are expressed as the mean ± S.E.M. i, The number of GAD− terminals with symmetric synapses contacting GAD+ dendrites was significantly increased in the AMPH CPP group compared with the delayed pairing and saline control groups. There was no difference in the number of GAD− terminals with symmetric synapses contacting GAD+ dendrites for rats in the delayed pairing group compared to the saline group. Data are expressed as the mean ± S.E.M. * P < 0.05 versus the delayed pairing and saline groups. j, AMPH CPP had no effect on the number of GAD+ terminals with symmetric synapses contacting GAD+ and GAD− dendrites. Data are expressed as the mean ± S.E.M. k, AMPH CPP had no effect on the number of GAD− terminals with symmetric synapses contacting GAD+ and GAD− somata or the number of GAD+ terminals with symmetric synapses contacting GAD+ somata. GAD+ terminals with symmetric synapses contacting GAD− somata were not detected. ‘0’ in the graph is used to indicate that the group mean is equal to zero. Data are expressed as the mean ± S.E.M. Abbreviations: as, asymmetric synapse; ss, symmetric synapse; sp−, GAD− spine; d−, GAD− dendrite; d+, GAD+ dendrite; som−, GAD− soma; som+, GAD+ soma. The scale bar in g is valid for a-g and equals 250 nm.

Figure 5

AMPH CPP remodeled existing synapses. a, An electron micrograph depicting a GAD− MSB contacting two GAD+ targets. b, An electron micrograph depicting a GAD− MSB contacting two GAD− targets. c, The total number of MSBs and the number of MSBs contacting GAD+ and GAD− targets were greater for animals in the AMPH CPP group compared to animals in the delayed pairing and saline groups. There were no differences between the delayed pairing and the saline group for any of these three categories. Data are expressed as the mean ± S.E.M. * P < 0.05 versus the delayed pairing and saline groups. Abbreviations: GAD+, glutamic acid decarboxylase positive; GAD−, glutamic acid decarboxylase negative; MSB, multisynaptic bouton. The scale bar in b is valid for a and b and equals 250 nm.

Figure 6

Context-drug learning alters the synaptic landscape of BLA pyramidal neurons and GABAergic interneurons. a, Behavioral output is dictated by the interaction between glutamatergic inputs (single gray circles) contacting distal spines of BLA pyramidal neurons (Pyr), dopaminergic (DAergic) input (gray triangles) originating in the ventral tegmental area and contacting proximal dendrites of pyramidal neurons, and GABAergic inputs (gray rectangles) which provide a dense innervation of the perisomatic region of pyramidal neurons. Also depicted is the dense innervation of the perisomatic region of neighboring GABAergic interneurons by the axon collaterals of BLA pyramidal neurons. b, After repeated context-drug associations, there is an increase in the number of excitatory and MSB (double gray circles) synapses contacting BLA pyramidal neurons and an increase in excitatory, inhibitory or modulatory, and MSB synapses contacting GABAergic interneurons. The net result is approach behavior to the drug-associated context.