Article Figures & Data
Figures
- Fig. 1.
Induction of LTD at CA1 individual inhibitory synapses. A, IR-DIC images of double patch pipette tips on a lacunosum-moleculare (LM) interneuron and a pyramidal cell in a hippocampal slice (left). Shown are the morphology (middle) of a synaptically connected LM interneuron and a pyramidal cell pair labeled with biocytin and the arrangement (bottom) of the electrodes at a LM interneuron (i) paired with a pyramidal cell (ii) and an extracellular stimulating electrode at the Schaffer-collateral fibers (iii). B, Induction of LTD of the unitary GABAAR–IPSCs.b1, Single action potentials (top) and 10 consecutive single (center) and averaged GABAAR–IPSCs (bottom) at −60 mV are taken before (i) and after (ii) tetanus.b2, A representative recording (top) and the averaged amplitudes (bottom) of the unitary GABAAR–IPSCs are plotted. b3, Amplitude distribution histograms for the unitary GABAAR–IPSCs before (Baseline) and 20 min after tetanus (LTD) are plotted with bin sizes of 4 pA.Inset, Summarized coefficient variance (CV−2= M2/ς2;n = 10 cells/5 control mice). C, The number of open channels of the synaptic GABAA receptors was reduced during LTD. c1, Ten to 90% rise time (RT) and time constants (τ) of decay before and after the induction of LTD were unchanged. c2, Current–variance relationships for the unitary GABAAR–IPSCs are plotted before (Baseline) and during LTD. The data points are fitted by a parabolic function (ς2 = iIm −Im2/No), where ς2 is the variance, Imis the mean current, i is the single channel current, and No is the number of synaptically activated channels.
- Fig. 2.
LTD recruits CaN-A to form a complex with GABAA receptors. A, B, Immunoprecipitation of the CA1 slices, 30 min after control stimulation (A) or 5 and 30 min after the induction of LTD (B), with nonspecific (N.S.) mouse IgG or a polyclonal mouse anti-CaN-A with (+) or without (−) 10 μg of immunizing antigen (a peptide corresponding to residues 457–482 of CaN), and a previous cross-linked anti-α1 (see Materials and Methods) with (+) or without (−) 10 μg of immunizing antigen (a peptide corresponding to residues 1–15 of α1 subunit). Blots were probed with monoclonal rabbit anti-CaN-A or rabbit anti-α1, anti-γ2, or anti-β2, as indicated. In the lane markedInput, 50 μg of proteins without immunoprecipitation was loaded. The molecular size is marked at the right of the each panel. C, Immunoprecipitates were quantified for anti-CaN-A (filled bars) and anti-α1 (open bars). Each immunoprecipitated band was normalized as a percentage of the respective input. Error bars are ±SEM (n = 4).
- Fig. 3.
Direct binding of CaN-A to the second intracellular domain of γ2 subunit. A, Affinity precipitation of the CA1 extracts, 30 min after control stimulation, or the induction of LTD, with GST-α1, GST-β2, GST-γ2S, or GST alone, and blots were probed with anti-CaN-A. B, Direct binding of CaN-A to GST-γ2. CaN420 (1 μg) or CaN-B (1 μg) was incubated with 10 μg of GST-α1 or GST-β2, GST-γ2S, *GST-γ2S, or GST-γ2S + 10 μg γ2-peptide, or GST-γ2L (long form) or GST alone, and blots were probed with anti-CaN-A or anti-CaN-B, as indicated. C, CaN-A binds to GST-γ2317–332. D, Affinity precipitation of the LTD-CA1 extracts with 10 μg of GST-α1, GST-β2, GST-γ2S, or GST alone, in the presence of 10 μg of γ2-peptide, and blots were probed with anti-CaN-A.E, Affinity precipitation of the CA1 extracts with GST-α1, GST-β2, GST-γ2S, or GST alone (top) or in the presence of γ2 peptide (bottom), and blots were probed with anti-Src, as indicated. F, Affinity precipitation of the CA1 extracts with GST-α1,GST-β2, GST-γ2S, or GST alone (top) or in the presence of γ2 peptide (bottom), and blots were probed with anti-α-adaptin, as indicated. Similar results are observed in each of four experiments (n = 4). The molecular size is marked at theright of the each panel.
- Fig. 4.
Blockade of CaN-A–GABAA receptor complex formation prevents the induction of LTD. A, Immunoprecipitation of the CaN mutant CA1 slices, 5 and 30 min after tetanus, with an N.S. IgG or anti-CaN-A or anti-α1. Blots were probed with anti-CaN-A or anti-γ2. In the lane marked Input, 50 μg of proteins without immunoprecipitation was loaded. Similar results are observed in each of the four experiments. B, LTD is abolished in the CaN mutant mice. b1, Single action potentials (top) and averaged unitary GABAAR–IPSCs at −60 mV (bottom) are taken before (i) and after (ii) tetanus (arrowhead). b2, The averaged amplitudes of the unitary GABAAR–IPSCs are plotted for the experiments with CaN control (open circles; the same as in Fig. 1B) or CaN mutant mice (filled circles). C, Immunoprecipitation of the CaN control mice CA1 slices, after tetanus in the presence of 50 μm AP-5, with an N.S. IgG or anti-CaN-A or anti-α1. Blots were probed with anti-CaN-A or anti-γ2. D, Induction of LTD depends on NMDA receptors. d1, Single action potentials (top) and averaged unitary GABAAR–IPSCs at −60 mV (bottom) are taken before (i) and after (ii) tetanus (arrowhead). d2, Normalized unitary GABAAR–IPSCs are plotted for the recordings with the control (filled circles;n = 5) or AP-5 (open circles;n = 6). E, γ2-peptide blocks the induction of LTD. e1, Single action potentials (top) and averaged unitary GABAAR–IPSCs at −60 mV (bottom) are taken before (i) and after (ii) tetanus (arrowhead). e2, Normalized unitary GABAAR–IPSC amplitudes are plotted for the experiments with γ2-peptide (open circles;n = 8) or scrambled γ2-peptide (filled circles; n = 7).
- Fig. 5.
CaN-A and LTD dephosphorylate GABAAreceptors. A, Protein immunoblot of phosphorylated GABAA receptors using anti-γ2pS327. Top, The GST-γ2S (lanes 1, 3) or the mutant GST-γ2S (Ser327-Ala) (lanes 2, 4) were subjected toin vitro phosphorylation for 1 min (lanes 1, 3) or 5 min (lanes 2,4). Bottom, Blot of SDS-PAGE by anti-GST antibody. B, Immunoblot of SDS-PAGE after precipitation by previous cross-linked polyclonal rabbit anti-γ2 (see Materials and Methods) from CA1 slices with either a rabbit polyclonal anti-γ2, as indicated (lane 1), or anti-γ2pS327 without (lane 2) or with phosphopeptide antigen (lane 3) or nonphosphopeptide (lane 4). C, Anti-γ2pS327 immunoblots of SDS-PAGE after precipitation by N.S. IgG (lane 1) or anti-γ2 from CA1 slices 5 and 30 min after control stimulation (lane 2) or the induction of LTD (lane 3). The precipitates were quantitated for γ2 subunit phosphorylation. The levels of γ2 subunit phosphorylation after induction of LTD (filled bars) were normalized to their respectivelane 2 (control stimulation; open bars). Error bars are ±SEM (n = 4; *p< 0.01). D, The anti-γ2pS327 (top) and anti-γ2 (bottom) immunoblots of SDS-PAGE after precipitation by N.S. IgG (lane 1) or anti-γ2 from the CaN mutant slices without tetanus (lane 2) or with tetanus (lane 3) or the CaN control mice with tetanus in the presence of AP-5 (lane 4) or calyculin-A (lane 5). Data were normalized to lane 2 from the same gel in bar graph. *p < 0.01 (n = 5); paired Student's ttest.
- Fig. 6.
CaN-A and LTD dephosphorylates GABAAreceptor γ2 subunit. 32P-labeled GABAA receptor GST fusion proteins (see below) were exposed to 10 μg/ml iCaN420 (heat-inactivated control; lane 1) or CaN420 (lane 2), CaNctr (control; lane 3), or CaNltd (lane 4) and analyzed by SDS-PAGE and autoradiography. The extent of 32P labeling for the experiments with iCaN420 (black bars), CaN420 (open bars), CaNctr (cross bars), or CaNltd (hatched bars) were quantitated, and normalized to respective controls (n = 4; *p < 0.01).
- Fig. 7.
Interaction of CaN-Aα and GABAAreceptors reduces the number of open channels of the synaptic GABAA receptors. A, Affinity precipitation of the CN98 control, or the mutant CA1 slices, with GST-α1, GST-β2, GST-γ2S, or GST alone and blots were probed with anti-CaN-A. In the lane marked Input, 50 μg of proteins without immunoprecipitation was loaded. Similar results are observed in each of the four experiments. B, The number of open channels of the synaptic GABAA receptors was reduced in CN98 mutant mice. b1, Sample traces of mIPSCs in the presence of 1 μm TTX are taken from the experiments with CN98 control or mutant mice. b2, The number of events in CN98 control and mutant mice is binned at 4 pA.Inset, Summarized coefficient variance (CV-2; n = 10 cells). b3, Ten to 90% rise time and time constants (τ) of decay in mutant mice were unchanged. b4, Current–variance relationships for the mIPSCs are plotted for the experiments with CN98 control or mutant mice.
- Fig. 8.
CaN-A-induced depression and LTD occluded each other. A, Effect of tetanus on spontaneous IPSCs. Single (a1) or averaged (20 consecutive responses) (a2) traces are taken from the experiments with the CN98 control or mutant mice. a3, Current–variance relationships for spontaneous IPSCs are plotted before (filled symbols) and 20 min after tetanus (open symbols) from the experiments with the CN98 control (circles) or the mutant mice (triangles). The data points are fitted by a parabolic function. B, CaN420 reduces amplitudes of the unitary GABAAR–IPSCs. b1, The traces are single action potentials (top) and averaged 10 consecutive unitary GABAAR–IPSCs (bottom) taken at the time indicated by the letters. Normalized unitary GABAAR–IPSCs for the recordings with 10 μg/ml iCaN420 (filled circles; n = 5) or 10 μg/ml CaN420 (open circles; n = 6) are plotted. b2, Single action potential and superimposed unitary GABAAR–IPSCs at −60 to +60 mV (40 mV increment) recorded with high Cl−(top). Current–voltage relationships (n = 4) are taken from 5 min (Baseline; open circles) and 20 min after starting recording (filled circles).C, Effects of tetanus on CaN420 action. Single action potentials (top) and averaged 10 consecutive unitary GABAAR–IPSCs (bottom) are taken at the time indicated by the letters. Normalized amplitudes of the unitary GABAAR–IPSCs when tetanus (arrowhead) was delivered 10 min after the start of recording (c1) or without tetanus (c2) are plotted. CaN420 (10 μg/ml) was actively perfused during the period indicated by the horizontal bars.
