Genetic interference to study amygdala function in mice

Abstract

In this PhD thesis some molecular mechanisms underlying fear conditioning are addressed by genetic manipulation of a key determinant of synaptic plasticity, namely the AMPA receptor subunit GluR-A. GluR-A is critically involved in longterm potentiation at hippocampal CA3-to-CA1 synapses and is necessary for the formation of spatial working memory. To elucidate whether GluR-A, within the lateral nucleus of the amygdala (LA), is required for the acquisition of fear memories, procedures to generate LA-specific GluR-A depletion, either by generating amygdala-specific transgenic mice, or by employing stereotactic virus delivery, were implemented. First, transgenic mouse lines were generated by expressing enhanced green fluorescent protein (EGFP) under control of the promoter for the Lypdc1 gene, for which in situ hybridization studies showed specific activity in the basolateral amygdala. Unfortunately, in all transgenic lines the Lypdc1-promoter driven EGFP expression was not restricted to the amygdala but was also detected in additional brain regions. Therefore, the Lypdc1-promoter is not useful for manipulating the GluR-A gene specifically in the amygdala. As an alternative strategy, recombinant adeno-associated-Cre virus (rAAVhSyn- Cre/IRESven) was stereotactically delivered into the LA of mice with loxPflanked exon 11 of the Gria1 gene (GluR-A2lox/2lox) prior to fear conditioning. In twothirds of the injected animals, Cre recombinase expression, which was accompanied by loss of GluR-A signal, could be detected in 10-30% of the LAneurons. This level of ablation had been previously shown by others to be sufficient to evoke a phenotype in fear conditioning. As an essential step, different paradigms for fear behavior in wildtype (WT) and global GluR-A knockout (KO) mice were established. The GluR-A KO mice showed a prominent impairment during the acquisition of conditioned fear, demonstrated by the absence of tone-shock induced freezing behavior. Since the sensory systems of GluR-A KO mice were not impaired, this observation suggested that the short-term association of tone and shock is GluR-A dependent. When challenged 24 hours later, the GluR-A KO mice exhibited reduced, although still detectable, memory of the conditioned tone. Thus, it is possible that efficient shortterm association of tone and shock is not necessary for the formation of long-term memory of the aversive stimulus. It seems that GluR-A dependent plasticity mechanisms are operative during the acquisition phase and that GluR-A independent mechanisms can be used for long-term fear memory formation. GluR-A in the LA might be necessary for the immediate tone-shock association during fear acquisition, since the rAAV-Cre mediated LA-specific GluRA KO mice showed a trend to exhibit less freezing during the acquisition phase than uninjected control animals. However, in order to substantiate this finding it would be necessary to optimize virus injection to achieve more efficient Cre targeting within the LA.