Figure 2.

Blocking cytoplasmic, including newly synthesized, CREB1 impaired LTF and associated increase in nuclear CREB1. A1, Protocol for control IgG or anti-tCREB1 antibody injection and electrophysiological testing. Antibodies were injected in the cytosol of sensory neurons 1 h after the end of treatment with 5-HT. A2, Confocal images of a neuron injected with both anti-tCREB1 antibody and Alexa 488-dextran and of a noninjected neuron. Neurons were incubated with secondary antibody only. In the injected neuron, tCREB1 immunoreactivity was limited to the cytoplasm, suggesting that the antibody did not cross the nuclear membrane. Minimal autofluorescence and background staining were observed in the noninjected neuron. Scale bar, 20 μm. A3, EPSPs recorded from sensorimotor cocultures before (pre) and 24 h after (post) treatment with 5-HT or vehicle. In some experiments, the facilitation was so large that it led to the initiation of a spike. In these cases, the EPSP was assigned a value of 65 mV as was done in previous studies (Zhao et al., 2003). A4, Summary data. One-way ANOVA followed by post hoc tests showed that injection of anti-tCREB1 antibody blocked LTF without significantly affecting basal transmission over 24 h. B1, Cytoplasmic injection of anti-tCREB1 antibody blocked 5-HT-induced increase in nuclear tCREB1. In the top panel of representative images, the cytoplasmic locus of fluorescein indicated antibodies were injected into the cytoplasm. In the bottom panel, tCREB1 levels in the nucleus 2 h after 5-HT treatment was measured by immunofluorescence staining using anti-tCREB1 antibody. Scale bar, 20 μm. B2, Anti-tCREB1 injection blocked 5-HT-enhanced expression of tCREB1 in nucleus (n = 4 independent experiments, representing 22 IgG-injected plus vehicle, 21 IgG-injected plus 5-HT, 21 anti-tCREB1-injected plus vehicle, and 23 anti-tCREB1-injected plus 5-HT neurons). C, Model of CREB1 feedback loop for the induction and consolidation of long-term memory in Aplysia (see Discussion for description).