Zinc transporter 3 is involved in learned fear and extinction, but not in innate fear

Abstract

Synaptically released Zn²+ is a potential modulator of neurotransmission and synaptic plasticity in fear-conditioning pathways. Zinc transporter 3 (ZnT3) knock-out (KO) mice are well suited to test the role of zinc in learned fear, because ZnT3 is colocalized with synaptic zinc, responsible for its transport to synaptic vesicles, highly enriched in the amygdala-associated neural circuitry, and ZnT3 KO mice lack Zn²+ in synaptic vesicles. However, earlier work reported no deficiency in fear memory in ZnT3 KO mice, which is surprising based on the effects of Zn²+ on amygdala synaptic plasticity. We therefore reexamined ZnT3 KO mice in various tasks for learned and innate fear. The mutants were deficient in a weak fear-conditioning protocol using single tone-shock pairing but showed normal memory when a stronger, five-pairing protocol was used. ZnT3 KO mice were deficient in memory when a tone was presented as complex auditory information in a discontinuous fashion. Moreover, ZnT3 KO mice showed abnormality in trace fear conditioning and in fear extinction. By contrast, ZnT3 KO mice had normal anxiety. Thus, ZnT3 is involved in associative fear memory and extinction, but not in innate fear, consistent with the role of synaptic zinc in amygdala synaptic plasticity.

Figure 1.

Expression of ZnT3 RNA and protein in the brain. (A) RNA in situ hybridization shows ZnT3 presence in the amygdala, hippocampus, TE3 area of the auditory cortex, and perirhinal cortex but not in the MGm/PIN of the auditory thalamus in the wild-type (WT) mouse brain. (B) RNA in situ hybridization shows ZnT3 absence in the brain of ZnT3 KO mice. (C) Expression pattern of ZnT3 protein detected by anti-ZnT3 antibody. Brain section on the left shows ZnT3 expression in the PFC and includes an inset showing prelimbic (PrL) and infralimbic (IL) areas. (D) Detection of synaptically released zinc by Timm's method.

Figure 2.

ZnT3 KO mice are deficient in fear conditioning in the single tone–shock-pairing protocol but are normal when five pairings are used. (A–C) Results of fear conditioning in the single-pairing protocol. (A) During acquisition, ZnT3 KO mice froze significantly less than WT mice immediately after the shock. **P < 0.01. (B,C) Retention data, context test, and tone test, respectively. In both tests, ZnT3 KO mice froze less than their WT littermates. **P < 0.01. (D–F) Results of fear conditioning in the five-pairings protocol show that ZnT3 KO mice are normal during acquisition and retention for both context and tone. Results are presented as mean ± SEM.

Figure 3.

ZnT3 KO mice are normal in fear conditioning in the five-pairing protocol when a continuous tone is used as the CS, but are deficient when a discontinuous tone is used. (A) Procedures for the five-pairing protocol using continuous or discontinuous tone. Both tones are at 85 dB and last 8 sec. For the discontinuous tone, each pulse lasted 500 msec and the interstimulus interval was 70 msec. (B) There are no differences between the genotypes during the acquisition phase. (C) Testing for context fear conditioning showed no differences between the genotypes regardless of the type of tone used during acquisition. (D) Testing for cued fear conditioning showed normal freezing in ZnT3 KO mice for the continuous tone; however, ZnT3 KO mice froze significantly less than the WT mice for the discontinuous tone. **P < 0.01. Results are presented as mean ± SEM.

Figure 4.

ZnT3 KO mice are deficient in trace fear conditioning. (A) ZnT3 KO mice show normal rate of acquisition of the task. (B,C) During testing for memory for the context (B) and tone (C) ZnT3 KO mice froze significantly less than the control mice. *P < 0.05; **P < 0.01. Results are presented as mean ± SEM.

Figure 5.

ZnT3 KO mice are deficient in extinction of fear conditioning. ZnT3 KO mice extinguished faster during cued fear conditioning. (A) Percent freezing to the tone is shown by five blocks of four tone presentations during the 4 d of extinction. ZnT3 KO and WT mice started at the same level and reached the same level of extinction but ZnT3 KO extinguished faster than WT littermates. Results are presented as mean ± SEM. (B) Renewal tested 15 d after the end of the extinction session in the context of acquisition. No difference was detected between ZnT3 and WT mice. Results are presented as mean ± SEM.

Figure 6.

ZnT3 KO mice have normal shock sensitivity, innate fear, and depression-like behavior. (A) No difference was found between ZnT3 KO and control mice in sensitivity to the electric shock, as measured by movement, vocalization, and jump in response to the gradual increase of the mild shock. (B) The basal level of anxiety/innate fear as analyzed in the elevated plus maze is normal in ZnT3 KO mice. (C) No difference was detected between ZnT3 KO and control mice in the open field. (D) The general level of depression, as analyzed in the Porsolt's forced-swim test, is normal in ZnT3 KO mice. Results are presented as mean ± SEM.