Molecular Processing of the Visceral Neuraxis and Conditioned Taste Aversion

Conditioned taste aversion learning (CTA) is a form of associative learning in which an animal learns to reject a previously palatable tastant (conditioned stimulus; CS) after conditioning with a toxic unconditioned stimulus (US) such as i.p. lithium chloride (LiCl). CTA learning is robust and long lasting, making it a good model to use in the study of learning. I used CTA learning as a model of learning to explore molecular mechanisms involved in learning and memory. I focused on a pathway that had previously not been examined in CTA learning, the protein 14-3-3-transducer of regulated cyclic-AMP response element binding protein (CREB) activity (TORC) pathway. Overall I identified that the protein 14-3-3-TORC pathway is critical for CTA learning and regulation of the visceral neuraxis after LiCl injection. Protein 14-3-3 isoforms are key to many cellular processes and are ubiquitous throughout the brain. 14-3-3 is a regulator of ser/thr phospho-signaling by binding and sequestering phosphorylated substrates including kinases, histone deactylases, and transcription factors. The role of protein 14-3-3 in CTA had not previously been examined. I parameterized CTA learning in difopein-YFP transgenic mice, which have widespread expression of the artificial peptide difopein in the forebrain, including the basolateral amygdala and insular cortex, resulting in functional knockout (FKO) of all 14-3-3 isoforms. I found that a single pairing of saccharin or NaCl (CS) and LiCl injection (US) was not sufficient to induce CTA in FKO mice. Multiple pairings of CS and US did lead to CTA acquisition in the FKO mice; however, the CTA rapidly extinguished within 30 minutes to 24 hours after acquisition. Several experiments were conducted to determine if sensory of global learning deficits were responsible for the inability of the FKO mice to learn a CTA. Additionally, I found that 14-3-3 FKO mice have an attenuated visceral neuraxis response to LiCl as measured by c-Fos induction. However the mice show xv normal lying-on-belly behavioral respose to LiCl. The deficit in FKO was not due to an inability to discriminate or avoid tastants, because they showed normal unconditioned taste preferences for both palatable (saccharin, maltodextrin, low concentration NaCl) and unpalatable tastants (quinine, HCl, and high concentration NaCl) and they were able to reduce intake of a maltodextrin solution adulterated with quinine. The FKO did not have a global deficit in ingestive learning, because they were able to form a conditioned flavor-nutrient preference. Thus, FKO of forebrain 14-3-3 appears to disrupt CTA learning leading to forgetting, rapid extinction, or failure to reconsolidate. This further implicates ser/thr phospho-signaling pathways in the regulation of long-term CTA learning. CREB mediated gene expression is critical for learning and memory including learning about visceral stimuli (e.g. CTA learning). However, CREB requires additional co-factors to produce gene expression. One such co-factor is TORC, which binds to CREB to facilitate gene transcription. TORC also binds to protein 14-3-3 when inactive, as part of its regulation. CREB mediated gene transcription can induce c-Fos, a widely used marker of neuronal activity after visceral stimulation (e.g. acute LiCl injection). The nuclear localization of TORC1 (one of 3 types of TORC) had not been examined in the visceral neuraxis. Additionally, if CREB mediated gene transcription is necessary for c-Fos induction in the visceral neuraxis after i.p. LiCl injection, then TORC should be active following stimulation. I examined nuclear (activated) TORC1 in the visceral neuraxis at 30, 60, and 180 minutes after LiCl injection. Consistent with previous studies I found increases in c-Fos 60 min after LiCl injection in all brain regions of the visceral neuraxis. I found increases in nuclear TORC1 in the area postrema, nucleus of the solitary tract and paraventricular nucleus after LiCl injection. In the central amygdala after LiCl injection almost all TORC1 was cytoplasmic whereas control injected rats xvi had almost all nuclear TORC1. No differences in nuclear TORC1 were found in the parabrachial nucleus or supraoptic nucleus. I examined fluorescent cellular co-localization of TORC1 and c- Fos at 60 min after LiCl injection in the visceral neruaxis. I found across the visceral neuraxis total TORC1 and total c-Fos in each region were significantly greater than the number of doublelabeled cells. Thus, I identified a separate (from c-Fos) population of cells (TORC1 cells) that are regulated by LiCl injection. Additionally this suggests that TORC1, and in turn CREB mediated gene transcription, may not be necessary for c-Fos induction of the visceral neuraxis after LiCl injection. Another binding partner of protein 14-3-3 that has been implicated in learning and memory is histone deacetylase 4 (HDAC4). HDACs have been previously shown to be important for CTA learning, however HDAC4 specifically had not been explored with relation to CTA. I examined interaction of HDAC4 and protein 14-3-3 in the amygdalar lobe and dorsal hindbrain, two regions of the brain important for CTA learning. I found that in both regions HDAC4 and protein 14-3-3 interact, suggesting a need for further study to explore the regulation of histone deacetylation during CTA learning.