Acute synthesis of CPEB is required for plasticity of visual avoidance behavior in Xenopus

Abstract

Neural plasticity requires protein synthesis, but the identity of newly synthesized proteins generated in response to plasticity-inducing stimuli remains unclear. We used in vivo bio-orthogonal noncanonical amino acid tagging (BONCAT) with the methionine analog azidohomoalanine (AHA) combined with the multidimensional protein identification technique (MudPIT) to identify proteins that are synthesized in the tadpole brain over 24 hr. We induced conditioning-dependent plasticity of visual avoidance behavior, which required N-methyl-D-aspartate (NMDA) and Ca(2+)-permeable α-Amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptors, αCaMKII, and rapid protein synthesis. Combining BONCAT with western blots revealed that proteins including αCaMKII, MEK1, CPEB, and GAD65 are synthesized during conditioning. Acute synthesis of CPEB during conditioning is required for behavioral plasticity as well as conditioning-induced synaptic and structural plasticity in the tectal circuit. We outline a signaling pathway that regulates protein-synthesis-dependent behavioral plasticity in intact animals, identify newly synthesized proteins induced by visual experience, and demonstrate a requirement for acute synthesis of CPEB in plasticity.

Figure 1

Conditioning-dependent plasticity of visual avoidance behavior A. Diagram of visual avoidance behavior. When a stimulus approaches a freely swimming tadpole, the animal rapidly changes its swim trajectory. B. The avoidance index (AI) remains constant over 7 hours and is unchanged after 24 hours. Dotted line is the average AI over the first three tests. C. VC for 30 minutes during the period marked with the grey bars significantly enhanced AI for at least 24 h. D, E. AI significantly increased compared to baseline when tested up to 24 h after 2 h or 4 h of VC. **P<0.01, *P<0.05. N= 6–12 animals per group. See also Figure S1.

Figure 2

Behavioral plasticity requires calcium-dependent signaling, gene transcription and protein synthesis. A,B. CPP (25 μM) during but not after VC blocked visual avoidance plasticity. C, D. Joro Spider Toxin (JST, 500 nM) during or after VC blocks behavioral plasticity. E, F. KN93 (5 μM) but not KN92 (5 μM) blocks behavioral plasticity. G,H. Anisomycin (ANI, 25 μM) during or after VC blocks plasticity. I. Actinomycin D (Act D, 25 μM) during VC blocks behavioral plasticity. J. AI (mean ± SEM) from 3–4 hours after VC. N=8–16 animals per group. **P<0.01, *P<0.05. See also Figure S2.

Figure 3

Rapid labeling of newly synthesized proteins in vivo A. Diagram of AHA labeling of newly synthesized proteins and detection by Western blot or MudPIT. B. Newly synthesized proteins detected 2 h after ventricular AHA injection. PBS injection or click chemistry without copper produce minimal labeling. C,D. Quantification of AHA labeling. Top: Dot blots of AHA-biotin label and C4-actin, for normalization. Bottom. Relative changes in AHA labeling. Whole brains were dissected 3 h after AHA injection. C. AHA-labeling increases over 3.5 hours after ventricular AHA injection. D. ANI (25 μM in rearing solution) significantly decreases AHA labeling and PTZ (15 mM in rearing solution) significantly increases AHA labeling compared to controls. Relative intensity of biotin labeling: PTZ =1.40 ± 0.07; ANI = 0.28 ± 0.10. N=3–5. Ventricular injection of ANI or cycloheximide also decreased AHA labeling compared to PBS-injected controls (Figure S3). **P<0.01, *P<0.05.

Figure 4

Mass spectrometric analysis of AHA incorporation in vivo identifies newly synthesized CNS proteins. (A) Annotation of AHA-biotin labeled proteins according to sub-cellular compartments and organelles using STRAP. Proteins in each compartment are listed in Table S1. (B) Annotation of AHA-biotin labeled proteins according to function in neurons. Proteins are listed in Table S2.

Figure 5

Identification of proteins synthesized in response to visual conditioning A. Analysis of AHA-labeled β-tubulin relative to total β-tubulin in tecta from control and VC tadpoles. Left panel: Total β-tubulin in control and VC samples is comparable (VC/control = 0.98 ± 0.09). Middle and right panels: VC increases AHA-biotin labeled β-tubulin (VC/control=1.5±0.2; N=3). AHA-biotin labeled β-tubulin is not detected when copper is omitted from the reaction. B. VC increases protein synthesis. Protein candidates were measured in the total input and as biotinylated proteins from control and VC animals. C,D. Quantification of data in B. Relative labeling intensities (VC/control) of total input (C) or AHA-biotin (D). N=3 independent experiments for each candidate. * P<0.05, ** P<0.01.

Figure 6

CPEB is required for plasticity of avoidance behavior. A. CPEB-MO (blue) blocks VC-induced visual avoidance behavior seen with control morpholino (Ctrl-MO, red). Morpholinos were electroporated 2 days before VC and behavioral testing. B. Co-electroporation of CPEB-MO and full length Xenopus CPEB-CFP increases baseline AI and occludes VC-induced plasticity. C. Co-electroporation of mutant Xenopus CPEB-GFP, lacking CaMKII phosphorylation sites with CPEB-MO fails to rescue the decrease in VC-induced plasticity seen with CPEB MO alone. D. AI (mean ±SEM) 3–4 h after VC. E. Baseline AI (mean ±SEM) before VC. F. CPEB-MO electroporated 2 h before VC blocks behavioral plasticity for at least 18 h. N=8–16 animals per group. **P<0.01. See also Figure S5.

Figure 7

Conditioning-dependent synaptic and structural plasticity requires acute CPEB synthesis A. Visual stimulus-induced excitatory compound synaptic currents (eCSCs) from control, VC and VC + CPEB-MO animals at light intensities of 10, 20, 250 cd/m². B. VC-induced increases in early (<100ms) and late (>100ms) components of CSCc are significantly inhibited by CPEB-MO. Control and Ctrl-MO data were combined because they are not significantly different (Early eCSCs (nA*s): 10 cd/m²: Control = 0.88±0.24nA; Ctrl-MO = 0.87±0.21nA, P=0.97; 20 cd/m²: Control = 1.91±0.68nA; Ctrl-MO = 1.02±0.19nA, P=0.26; 300 cd/m²: Control = 2.47±0.73nA; Ctrl-MO = 1.15±0.22nA, P=0.14. Late eCSC’s: 10 cd/m²: Control = 4.74±1.63nA; Ctrl-MO = 5.57±1.04nA, P=0.66; 20 cd/m²: Control = 20.34±8.9nA; Ctrl-MO = 7.18±1.32nA, P=0.12; 300 cd/m²: Control = 29.45±12.33nA; Ctrl-MO = 10.52±3.39nA, P=0.12.). MO’s were electroporated 2 h before recording. N=7–8 neurons/group) C. Drawings of representative neurons imaged in vivo over 4 h in animals exposed to ambient light, VC, VC+Ctrl-MO and VC+CPEB-MO. MO’s were electroporated 2 h before imaging. D. VC significantly increases total dendritic branch length (TDBL) in control and Ctrl-MO neurons, but not in CPEB-MO neurons. *P<0.05; **P<0.01.