The laboratory investigates the signal transduction mechanism involved in the enhancement of synaptic transmission and plasticity. Studies of these neural processes are essential to understanding the complex processes of learning and memory. Our approach is to generate genetically modified mice by deletion or expression of a gene specifically expressed in the brain. Our efforts have led us to generate a strain of mice devoid of a neural-specific protein, neurogranin (Ng). The protein is normally expressed at high levels in the neurons within the cerebral cortex, hippocampus, and amygdala and has been implicated in the modulation of synaptic plasticity. Ng binds calmodulin (CaM) in a Ca2+-sensitive manner such that its binding affinity for CaM decreases with increasing Ca2+ concentration. It is believed that, at basal levels of Ca2+, all the CaM are sequestered by Ng, with its cellular concentration at least twice as high as that of CaM. Upon synaptic stimulation, the influx of Ca2+ displaces Ng from the Ng/CaM complex to form Ca2+/CaM. The buffering of CaM by Ng serves as a mechanism to regulate neuronal free Ca2+ and Ca2+/CaM concentrations. The concentration of free Ng fluctuates as a result of phosphorylation by protein kinase C (PKC) and oxidation by nitric oxide (NO). Both the phosphorylated and oxidized Ng exhibit lower affinities for CaM than does the unmodified protein. Synaptic responses triggering long-term potentiation (LTP) or long-term depression (LTD) depend on the amplitude of Ca2+ influx and the sensitivity of the transduction machinery to amplify the signal. Based on our previous findings that the autophosphorylation of Ca2+/CaM-dependent protein kinase II was enhanced by Ng, we speculate that neurogranin up-regulates other Ca2+- and Ca2+/CaM-regulated pathways. The diverse effects of Ng could account for its positive impact on the behavior of mice. Investigation of the regulatory mechanism of synaptic transmission by Ng will help us design novel therapeutic approaches to improve memory in humans.

Involvement of Neurogranin in Learning and Memory of Spatial Tasks
F. Huang, Li, K-P. Huang
Ng is highly concentrated in the cell bodies and dendrites of select neurons within the cerebral cortex, hippocampus, and amygdala. To examine the role of Ng in neural function, we generated mutant mice devoid of Ng. The Ng knock-out mice did not exhibit any obvious developmental and neuroanatomical abnormalities, but the knockout did cause an impairment in spatial learning and memory. The deficits in the knock-out mice were accompanied by a defective mechanism in the activation of Ca2+/CaM-dependent protein kinase II, whose activation by autophosphorylation has been linked to postsynaptic induction of LTP and storage of long-term memory. In the hippocampus-dependent spatial tasks, the rank order of performances was wild-type (WT), heterozygous (HET), and knock-out mice. Further analysis of the performances of these animals showed that the hippocampal levels of Ng varied broadly between individual WT and HET mice and that the performances of the HET mice, but not the WT mice, correlated positively with their hippocampal Ng levels. The results suggest that the WT mice contain levels of Ng above the threshold required for optimal performance of the spatial tasks. The knock-out mice performed poorly in the same tasks and exhibited a right-shift in the tetanus-frequency response curve in favor of LTD at frequencies (5 to 10 Hz) exhibiting in animals during exploration of a novel environment. We conclude that Ng enhances the synaptic responses by up-regulating Ca2+- and Ca2+ /CaM-mediated signaling pathways.

Attenuation of PKC and PKA Signal Transduction in Neurogranin Knock-Out Mice
Wu, Li, K-P. Huang, F. Huang
We investigated the Ng-regulated signaling pathways by using hippocampal slices from the WT and Ng knock-out mice treated with activators of PKC and PKA. Activation of PKC and PKA is required for the induction and maintenance of LTP and for formation and storage of memory. PMA, an activator of PKC, caused phosphorylation of Ng in the WT mice and promoted a greater extent of translocation of PKC from the cytosolic to the particulate fraction as compared with the knock-out mice. Phosphorylation of the downstream targets of PKC, including mitogen-activated protein kinases, 90-kDa ribosomal S6 kinase, and cAMP-response element-binding protein, was significantly attenuated in the knock-out mice. Stimulation of the hippocampal slices with forskolin, a PKA activator, also resulted in greater stimulation of PKA in the WT mice as compared with the knock-out mice. Similarly, phosphorylation of several down-stream targets of PKA, including those shared with the PKC pathway, were attenuated in the knock-out mice. The results suggest that Ng potentiates both the PKC and PKA signaling pathways and consequently contributes to the enhancement of synaptic transmission as well as to learning and memory of spatial tasks in mice.

Glutathione Disulfide S-Oxide Is a Potent Glutathiolating Agent Generated after Oxidative Stress
Li, F. Huang, K-P. Huang
Protein S-glutathiolation can be induced in cells by mild oxidative stress. Several compounds, including glutathione disulfide, superoxide, peroxynitrite, nitric oxide (NO), nitrosothiol, and, in particular, S-nitrosoglutathione (GSNO), are thought to be involved in protein S-thiolation. In mammalian cells, a relatively high concentration of GSH (0.5-10 mM) serves as an NO sink to form GSNO, which can undergo transnitrosylation with protein sulfhydryl groups to form S-nitrosoprotein and GSH or protein-GSH mixed disulfides and nitroxyl (HNO). GSNO can also release NO in the presence of cuprous ion, ascorbate, or thiols and serves as a possible source of nitrsonium (NO+) or nitroxyl (NO-) ions. In addition, GSNO is unstable in aqueous solution and undergoes decomposition, probably by homolytic cleavage of the S-N bond, to yield NO and a thiyl radical. We identified glutathione sulfonic acid (GSO3H), glutathione disulfide S-oxide (GS(O)SG), glutathione disulfide S-dioxide (GS(O2)SG), and glutathione disulfide (GSSG) as the major decomposition products of GSNO. We tested each of these compounds and GSNO for their efficacies in modifying rat brain Ng and neuromodulin/GAP-43 (Nm). We found that GS(O)SG was the most potent in causing glutathiolation of both proteins; four glutathione residues were incorporated into the four Cys residues of Ng and two into the two Cys residues of Nm. Ng and Nm are two in vivo substrates of PKC; their phosphorylation by PKC attenuates the binding affinities of both proteins for CaM. When compared with their respective unmodified forms, the GS-Ng was a poorer and GS-Nm a better substrate for PKC. Glutathiolation of these two proteins caused no change in their binding affinities for CaM. Treatment of [35S]cysteine-labeled rat brain slices with xanthine/xanthine oxidase (X-XO) or a combination of X-XO with sodium nitroprusside resulted in an increase in the cellular level of GS(O)SG. These treatments as well as those by other oxidants all resulted in an increase in thiolation of proteins; among them, thiolation of Ng was positively identified by immunoprecipitation. The results show that GS(O)SG is one of the most potent glutathiolating agents generated upon oxidative stress.