Skip Navigation
20## Annual Report of the Division of Intramural Research, NICHD National Institutes of Health Eunice Kennedy Shriver National Institute of Child Health and Human Development

Kuo-Ping Huang, PhD, Head, Section on Metabolic Regulation
Freesia L. Huang, PhD, Staff Scientist
Lian Chen, PhD, Adjunct Investigator
Catherine Boucheron, PhD, Postdoctoral Fellow
Zsolt Kovacs, MD, Postdoctoral Fellow
Pavan K. Shetty, PhD, Postdoctoral Fellow

We investigate the signal transduction mechanisms involved in synaptic transmission and plasticity, processes that are essential to learning and memory. We generate genetically modified mice by deletion or expression of a gene specifically expressed in the brain, in particular mice devoid of neurogranin (Ng), a neural-specific protein that is normally expressed strongly in specific neurons in the forebrain and has been implicated in the regulation of Ca2+/calmodulin (CaM)-dependent reactions. Ng binds to apoCaM; its binding affinity is attenuated with increasing Ca2+ concentration, phosphorylation by PKC, and oxidation of its sulfhydryl groups. Ng levels are relatively high in the neuronal cell body and dendrites, and the protein sequesters apoCaM in these compartments at basal physiological Ca2+. Upon synaptic stimulation, the influxed Ca2+ displaces Ng from the Ng/apoCaM complex to form Ca2+/CaM and free Ng. The buffering of CaM by Ng serves as a mechanism to regulate neuronal free Ca2+ and Ca2+/CaM concentrations. Furthermore, Ng is a potent reductant of oxidants generated during synaptic transmission. Our aim is to define the regulatory functions of Ng in neuronal signaling that are important for learning and memory and to design therapeutic approaches to improve cognitive functions in an aging human population and patients suffering from dementia.

Regulation of synaptic plasticity by neurogranin

Huang K-P, Huang F

Neurogranin (Ng) is a neuronal protein that binds to apoCaM under basal physiological conditions and dissociates from CaM upon synaptic stimulation, thus elevating intracellular Ca2+. Ng is localized in the neuronal soma and dendrites and is present in high concentrations in the dendritic spines. It has been suggested that Ng modulates synaptic responses by buffering and sequestering CaM to regulate the level of free Ca2+ and Ca2+/CaM complexes. The binding affinity of Ng for CaM is reduced by increasing Ca2+, phosphorylation by PKC, or oxidation by oxidants. The Ng concentration in the hippocampus of adult mice varies broadly; the level in Ng+/+ mice is one of the highest among all neuronal CaM-binding proteins. Among Ng+/- mice, but less so in Ng+/+ mice, a significant relationship exists between hippocampal levels of Ng and performances on several cognitive functional tests. Ng-/- mice performed poorly in all the tests; they also displayed deficits in high frequency-induced long-term potentiation (LTP) in area CA1 of hippocampal slices while low frequency-induced long-term depression (LTD) was enhanced. Measurements of Ca2+-transients in CA1 pyramidal neurons following weak and strong tetanic stimulations revealed a significantly greater [Ca2+]i-response in Ng+/+ than in Ng-/- mice. The diminished Ca2+ dynamics in Ng-/- mice is a likely cause of their decreased propensity to undergo LTP and their deficits in cognitive function.

To characterize further the specificity of Ng in post-synaptic signal transduction, we subjected hippocampal slices of Ng-/- mice to stimulation by phorbol ester, which is known to enhance presynaptic responses by increasing neurotransmitter release. In contrast to the high frequency-induced LTP, which is mediated by stimulation of post-synaptic NMDA receptors, the phorbol ester PDBu induced robust LTP in Ng-/- mice. The findings showed that Ng is specific for post-synaptic responses and implied that drugs that stimulate the presynaptic responses could ameliorate the behavioral deficits of Ng-/- mice.

To confirm the critical role that Ng plays in the enhancement of synaptic plasticity, we prepared Ng-expressing DNA constructs for injection into the brain of Ng-/- mice. To increase expression, we packaged the transgene in a non-pathogenic recombinant adeno-associated virus (rAAV) driven by a hybrid chicken B-actin/CMV promoter and added the cis-acting woodchuck post-transcriptional regulatory element. The transgene consists of Ng, IRES (internal ribosome entry site), and green fluorescence protein in a bi-cistronic-expressing cassette for co-expression of both proteins. The rAAV constructs will be stereotaxically injected into the hippocampus and amygdala of Ng-/- mice to determine if expression of Ng in these brain regions can correct their electrophysiological and behavioral deficits.

Huang K-P, Huang FL, Jäger T, Li J, Reymann KG, Balschun D. Neurogranin/RC3 enhances long-term potentiation and learning by promoting calcium-mediated signaling. J Neurosci 2004;24:10660-9.

Neurogranin in the regulation of metabotropic glutamate receptor-mediated long-term depression

Huang F, Huang K-P

Calcium entry into post-synaptic sites is essential to trigger the signaling cascade for synaptic plasticity; a higher level of post-synaptic Ca2+ induces LTP, and a lower level induces LTD. Two mechanistically distinct forms of LTD co-exist in the hippocampus. One depends on activation of NMDA receptors and post-synaptic protein phosphatases, and the other depends on activation of post-synaptic group 1 metabotropic glutamate receptors (mGluRs) and the local translation of dendritic mRNA. Low-frequency tetanic stimulation (1 Hz, 900 pulses) induces NMDA receptor-dependent LTD, which is normal in Ng-/- mice. Similarly, brief treatment of hippocampal slices from Ng-/- mice with DHPG, a group 1 mGluR agonist, induces LTD to a similar extent as that in Ng+/+ mice. DHPG is known to activate group 1 mGluR, including mGluR5, in the hippocampal CA1 area. In addition, MPEP, a noncompetitive antagonist of mGluR5, blocks the DHPG-induced LTD. However, Ng-/- mice were insensitive to MPEP-mediated antagonism. Given that group 1 mGluR agonists activate the phospholipase C cascade, including stimulation of PKC, we monitored phosphorylation of Ng. Unexpectedly, we found that Ng was dephosphorylated following DHPG treatment, probably as a result of the activation of calcineurin by low post-synaptic Ca2+ transients. Pretreatment with MPEP, which blocked LTD in Ng+/+ mice, prevented Ng from dephosphorylation induced by DHPG. The signaling cascade of DHPG-induced LTD requires rapid translation of pre-existing dendritic mRNA, which results from the activation of several upstream components, including Gq-type G proteins, MAP kinases, tyrosine phosphatases, PI3-kinase, PDK1, Akt, mTOR kinase, and S6 kinase. While MPEP does not directly inhibit mGluR5 receptors, the antagonistic action of MPEP on DHPG-induced LTD could result from interference with any of the steps leading to translation of dendritic mRNA. Lacking Ng, Ng-/- mice are unresponsive to MPEP-mediated blockade of DHPG-induced LTD, suggesting that Ng is required for the action of MPEP to antagonize group1 mGluR-mediated synaptic plasticity.

Effect of environmental enrichment on the cognitive behaviors of neurogranin knockout mice

Huang F, Huang K-P, Boucheron

Environmental enrichment and exercise (EEE) can increase neurogenesis, alter neural plasticity, and improve cognitive performances. Ng, a specific substrate of PKC, is abundantly expressed in brain regions important for cognitive functions. Deletion of Ng in mice causes severe deficits in learning and memory of spatial tasks, in induction and maintenance of LTP, and in amplification of Ca2+-mediated signaling in the hippocampus. To determine if an increase in the physical activities of these mutant mice would improve their performance on memory tasks, we housed groups of mutant mice and their wild-type littermates of different age groups in roomier cages for three weeks. Each cage had several toys and a running wheel for voluntary exploration and exercise. The control groups were housed in their regular home cages without any toys. For young adults, both Ng+/+ and Ng+/- mice performed significantly better than their controls in the enriched environment on both the hidden platform and probe trial versions of the Morris water maze as well as on the step-down inhibitory avoidance fear conditioning task. However, EEE had minimal effect on the performance of Ng-/- mice on these same tasks. Hippocampal LTP in the CA1 region induced by high-frequency stimulation was also higher in the enriched group than in the controls among Ng+/+ and Ng+/- mice, whereas we observed no significant change among Ng-/- mice. Quantitative immunoblot analyses showed that enriched mice of all three groups had elevated hippocampal levels of CaMKII and CREB, but not of ERK. The increase in hippocampal Ng levels in Ng+/+ and Ng+/- mice caused by enrichment seemingly contributed to their improved LTP and behavioral performances.

For aged mice (older than 18 months), EEE has variable effects on spatial and emotional (fear) learning and memory. The enriched groups of all three genotypes performed better than their control cohorts in the land-based spatial task, Barnes maze, and during training and remembered location better during probe trial. However, the effects of enrichment on the step-down inhibitory avoidance fear-conditioning task for these animals were unexpected. For Ng+/+ mice, the aging process alone, regardless of environmental factors, appeared to increase the animals' awareness of their new environment and their caution in making movements in the training chamber. These psychological traits were less evident in Ng+/- and Ng-/- mice; they stepped down from the platform within a short time. In contrast, the behavior of the enriched Ng+/- and Ng-/- groups was similar to that of the wild type; the enriched groups stayed on the testing platform for a longer period. After training, all genotypes displayed fear memory; however, unlike in the case of the young adults, enrichment greatly enhanced the memory of Ng-/- compared with their controls. The enrichment-induced improvement was most significant for the Ng-/- as compared with Ng+/- and Ng+/+. In fact, for emotional learning, EEE had no effect on Ng+/+ mice. Despite a significant positive effect of EEE on spatial and emotional memory of Ng-/- mice, EEE had no effect on LTP. For Ng+/- and Ng+/+ mice, EEE had only a meager positive effect on the LTP of these aged mice. The findings suggest that Ng plays a smaller role in the EEE-mediated response in aged than in young adult mice. Quantification of the levels of Ng mRNA from the control and EEE groups among young (about three-months-old), adult (three- to six-months-old), and aged (older than 18 months) Ng+/+, Ng+/-, and Ng-/- mice showed that EEE enhanced the hippocampal Ng mRNA levels among Ng+/+ and Ng+/- mice but had negligible effects on CaMKIIα mRNA. Results from in situ hybridization also showed that Ng mRNA levels in frontal cortex, caudate putamen, hippocampal CA1 and CA3, and dentate gyrus were significantly higher in aged EEE groups than in controls. The findings show that environmental enrichment can augment the hippocampal level of Ng; however, its effect on cognitive behaviors is variable and depends on animal age.

Huang FL, Huang K-P, Wu J, Boucheron C. Environmental enrichment enhances neurogranin expression and hippocampal learning and memory but fails to rescue the impairments of neurogranin null mutant mice. J Neurosci 2006;26:6230-7.

Modification of protein by reactive sulfur species: disulfide S-monoxide and S-dioxide

Huang K-P, Huang F, Shetty; in collaboration with Yergey

Accumulation of reactive oxygen and nitrogen species produced by oxidative stress has been linked to aging and many human diseases. In the CNS, neurotransmission under normal and pathological conditions generates a variety of oxidants, which, if not properly neutralized, will be detrimental to cell survival. Numerous neural disorders, including Alzheimer's and Parkinson's diseases, are believed to result in part from excessive damage to neurons by endogenous oxidants. Glutathione (GSH), at its cellular concentrations of 1-10 mM in mammalian cells, is one of the major low-molecular weight reductants to counter the assault from oxidative stress; it also serves as an abundant precursor of reactive sulfur species. Recently, we found that several oxidized forms of GSH with higher oxidative states, including glutathione disulfide S-monoxide (GS-DSMO) and disulfide S-dioxide (GS-DSDO), could be generated in oxidant-treated rat brain slices. The two compounds are much more reactive than glutathione disulfide (GSSG) for the thionylation of protein. Glutathionylation or thionylation of proteins has been shown to regulate the activities of enzymes, transcriptional factors, cell surface receptors, and cytoskeletal proteins. We proposed that these reactive disulfide S-oxides (DSOs) may be involved in the oxidant-mediated signaling cascade to modify proteins at those locations where oxidants are generated. In addition to endogenous thiols, numerous thiol-containing drugs and foodstuffs might participate in this signaling system.

To characterize further the reactions mediated by DSOs, we devised chemical methods to synthesize these compounds in bulk quantity with high purity. Both DSMO and DSDO were synthesized by iron- or methyltrioxorhenium (VII)-catalyzed oxidation of thiols, e.g., glutathione (GSH) and captopril (CPSH), with H2O2. In these metal-catalyzed reactions, thiols were better substrates than their corresponding disulfides in forming DSOs. The DSOs exhibited high reactivity toward thiol to form mixed disulfides. The compounds are highly reactive toward the sulfhydryl group of proteins, including PKC. The IC50 of GS-DSDO in the inactivation of PKC was at least an order of magnitude lower than that of CPS-DSDO, and both GS- and CPS-DSDO were more potent inhibitors of PKC than their corresponding DSMOs. We verified thionylation of PKC by GS-DSOs by immunoblot with antibody against GSH and by incorporation of GS-moiety from [35S]GS-DSOs into the kinase. The stoichiometries of thionylation of PKC beta mediated by [35S]GS-DSMO and [35S]GS-DSDO were approximately 1 and 5 mol/mol, respectively. Thionylation of PKC by GS-DSDO and CPS-DSDO rendered the kinase highly susceptible to proteolysis. Interestingly, GS-DSDO and CPS-DSDO preferentially targeted PKC at the catalytic and regulatory domains, respectively, as revealed by their susceptibilities to limited proteolysis. Furthermore, thionylation of PKC mediated by CPS-DSDO, but not that by GS-DSDO, increased autonomous kinase activity. The results show that the reactive sulfur species, especially DSDOs, generated from various thiols are potent thionylating agents and possess specificity toward PKC and probably other proteins. Thus, DSOs generated from thiols during oxidative stress can modulate the activity of PKC and tag the kinase for proteolysis. The unique specificity and diverse effects of various reactive sulfur species toward PKC may provide a prototype for the design of drugs to target a specific group of proteins for therapeutic application.

Oxidative modifications of CaMKII in the brain

Shetty, Huang F, Huang K-P

Calcium/calmodulin-dependent protein kinase (CaMKII) is one of the most abundant kinases in the brain; its autophosphorylation has been linked to the enhancement of synaptic plasticity. Stimulus-induced autophosphorylation of CaMKIIα at Thr286 converts the kinase into a constitutively active form, which is independent of Ca2+/CaM for activity. Activation of the NMDA receptor by excitatory amino acids causes not only CaMKII autophosphorylation but also its aggregation. The mechanism of the stimulus-induced aggregation of CaMKII is unknown. Given that synaptic stimulation causes an influx of Ca2+ and generation of reactive oxygen species at post-synaptic sites, we predicted that CaMKII aggregation may be induced by an oxidative mechanism. Previously, we showed that treatment of rat brain slices with oxidants generated two reactive sulfur species, glutathione disulfide S-monoxide (GS-DSMO) and S-dioxide (GS-DSDO). Incubation of purified mouse brain CaMKII with these oxidants caused glutathionylation and aggregation of the kinase. Interestingly, GS-DSMO mainly modified the α isoform of CaMKII, whereas GS-DSDO modified both α and β CaMKII. Modification of CaMKII by GS-DSMO caused only a minor inactivation of the enzyme; in contrast, modification by GS-DSDO resulted in dose-dependent inhibition of autophosphorylation of the kinase as well as of the kinase's activity toward exogenous substrate. Treatment of mouse brain synaptosomes with oxidants such as H2O2, diamide, and sodium nitroprusside caused thionylation and aggregation of proteins that were detectable by immunoblot with antibodies against GSH-associated protein. The treatments primarily caused CaMKII to form aggregates with a concomitant reduction of kinase activity. These findings suggest that oxidants generated during normal synaptic stimulation or under pathological conditions could modulate CaMKII activity through thionylation as well as through the formation of aggregates.

COLLABORATORS

Detlef Balschun, PhD, Leibniz Institut für Neurobiologie, Magdeburg, Germany
Alfred L. Yergey, PhD, Laboratory of Cellular and Molecular Biophysics, NICHD, Bethesda, MD

For further information, contact huangk@mail.nih.gov.

Top of Page