Substrate Phosphorylation in the Protein Kinase C g Knockout Mouse*

The phosphorylation state of three identified neural-specific protein kinase C substrates (RC3, GAP-43/B-50, and MARCKS) was monitored in hippocampal slices of mice lacking the g -subtype of protein kinase C and wild-type controls by quantitative immunoprecipitation following 32 P i labeling. Depolarization with potassium, ac- tivation of glutamate receptors with glutamate, or direct stimulation of protein kinase C with a phorbol ester increased RC3 phosphorylation in wild-type ani-mals but failed to affect RC3 phosphorylation in mice lacking the g -subtype of protein kinase C. Our results suggests the following biochemical pathway: activation of a postsynaptic (metabotropic) glutamate receptor stimulates the g -subtype of protein kinase C, which in turn phosphorylates RC3. The inability to increase RC3 phosphorylation in mice lacking the g -subtype of protein kinase C by membrane depolarization or glutamate receptor activation may contribute to the spatial learning deficits and impaired hippocampal LTP observed in these mice. Mice mild learning long by trichloroacetic acid precipitation as described previously, and 32 PO 4 incorporation into RC3, GAP-43/B-50, and MARCKS was normalized accordingly.Allexperiments were performed blind with regard to the genotype of the animals. Statistical analysis were carried out using a Student’s t test.

The phosphorylation state of three identified neuralspecific protein kinase C substrates (RC3, GAP-43/B-50, and MARCKS) was monitored in hippocampal slices of mice lacking the ␥-subtype of protein kinase C and wildtype controls by quantitative immunoprecipitation following 32 P i labeling. Depolarization with potassium, activation of glutamate receptors with glutamate, or direct stimulation of protein kinase C with a phorbol ester increased RC3 phosphorylation in wild-type animals but failed to affect RC3 phosphorylation in mice lacking the ␥-subtype of protein kinase C. Our results suggests the following biochemical pathway: activation of a postsynaptic (metabotropic) glutamate receptor stimulates the ␥-subtype of protein kinase C, which in turn phosphorylates RC3. The inability to increase RC3 phosphorylation in mice lacking the ␥-subtype of protein kinase C by membrane depolarization or glutamate receptor activation may contribute to the spatial learning deficits and impaired hippocampal LTP observed in these mice.
Mice lacking the ␥-subtype of protein kinase C (PKC␥) 1 show mild spatial learning deficits and exhibit impaired hippocampal long term potentiation (LTP), suggesting that PKC␥ is a key regulatory component in LTP and spatial memory (1)(2)(3). However, the biochemical pathways that are perturbed in these mice have not yet been identified. The acidic, calmodulinbinding PKC substrate RC3 (also called neurogranin), is a likely substrate for PKC␥ because the two proteins colocalize in the dendrites of excitatory neurons of the cerebral cortex, hippocampus, and striatum and are expressed at the same stages of development (4 -6). Additionally, RC3 is phosphorylated during LTP and dephosphorylated during LTD, and antibodies to RC3 interfere with the induction of LTP (7)(8)(9). Here, we examine the incorporation of 32 PO 4 into RC3 and two other major PKC substrates; GAP-43/B-50 and MARCKS by quantitative immunoprecipitation (8) before and after treating hippocampal slices from PKC␥-deficient and wild-type mice with potassium, glutamate, or the phorbol ester 4␣-phorbol 12,13dibutyrate (PDB). We show that stimuli that readily increase RC3 phosphorylation in wild-type mice fail to affect RC3 phosphorylation in PKC␥-deficient mice.

EXPERIMENTAL PROCEDURES
The effects of depolarization, glutamate receptor stimulation, and direct PKC activation were determined in PKC␥ knockout and litter mate control mice (bred into a C57/Bl background for six generations and kindly provided to us by Dr. J. M. Wehner).
For Western blotting, the forebrain of each mouse was homoginized in 10 ml of buffer containing 50 mM Tris (pH 7.5), 100 mM NaCl, 2 mM EDTA, 1 mM EGTA, 50 mM dithiothreitol, 0.6 mM phenylmethylsulfonyl fluoride, and 1% SDS. We equalized the protein concentrations of the homogenates based on three independent protein assays performed in triplicate prior to adding SDS and dithiothreitol using the BCA method and then immuno-blotted serial dilutions of each homogenate. The blotts were probed with polyclonal rabbit anti-RC3 (Affinity Research Products Ltd., Exeter, UK) at a dilution of 1:1000 and then developed with a horseradish peroxidase-based, enhanced chemiluminescence protocol.
All experiments were performed blind with regard to the genotype of the animals. Statistical analysis were carried out using a Student's t test.

RESULTS
Protein levels of RC3 were not different between wild-type, heterozygous, and PKC␥ knockout mice (Fig. 1A), showing that there was no up-or down-regulation of RC3 levels induced by the PKC␥ knockout. Basal in situ phosphorylation of RC3 and MARCKS in hippocampal slices from mice lacking PKC␥ did not differ significantly from those observed in wild-type littermate controls (102.3 Ϯ 8.1% (mean Ϯ S.E.) and 99.3 Ϯ 9.2% of basal phosphorylation in controls for RC3 and MARCKS respectively, p Ͼ 0.1, n ϭ 8, and n ϭ 4). However, increased basal phosphorylation of GAP-43/B-50 was observed in the null mutant (148.9 Ϯ 17.8% (mean Ϯ S.E.) of basal phosphorylation in controls, p Ͻ 0.05, n ϭ 8). As expected, depolarization with potassium, excitation with glutamate, or activation of PKC with PDB induced phosphorylation of RC3 in wild-type controls (Fig. 1, B and C). Strikingly, none of the treatments induced an increase in RC3 phosphorylation in slices from PKC␥-deficient mice. Thus, RC3 is a highly specific substrate for PKC␥ during excitation. PDB increased the phosphorylation of MARCKS and GAP-43/B-50 in slices derived from knockout (147.4 Ϯ 6.6% (n ϭ 4) and 162.7 Ϯ 14.2% (n ϭ 8)), as well as the wild-type (173.7 Ϯ 14.5% (n ϭ 4) and 288.2 Ϯ 35.2% (n ϭ 8)) mice (Fig.  1B). However, GAP-43/B-50 phosphorylation was significantly attenuated in the former compared with the latter (43.5 Ϯ 4.4% of wild type, p Ͻ 0.01, n ϭ 8). Decreased incorporation of 32 PO 4 into GAP-43/B-50 in knockout mice upon direct stimulation of PKC was probably due to higher initial levels of phospho-GAP-43/B-50 because the two appear to offset each other, although the possibility that GAP-43/B-50 could be a substrate for PKC␥ cannot be ruled out. DISCUSSION The experiments described here demonstrate that depolarization with potassium, activation with glutamate, or direct stimulation of PKC with a phorbol ester leads to phosphorylation of RC3 solely by PKC␥. Thus, the results unequivocally delineate the following biochemical pathway: activation of a postsynaptic (metabotropic) glutamate receptor stimulates PKC␥, which in turn phosphorylates RC3. Basal levels of phospho-RC3 appear to be dictated by a calcium and diacylglycerolindependent atypical isoform of PKC, possibly or . Basal levels of phospho-GAP-43/B-50 are higher in the PKC␥ knockout mouse, and this might be a presynaptic mechanism to compensate for the inability to phosphorylate RC3 in dendrites, perhaps by increasing neurotransmitter release in response to decreased postsynaptic gain (6, 10 -13). Inability to phosphorylate RC3 in the PKC␥ knockout mouse by either membrane depolarization or by activation of postsynaptic glutamate receptors may contribute to the electrophysiological and behavioral phenotypes of the PKC␥ knockout mouse.