Figure 1. mTOR and its downstream targets are important for dendritic branching. A) Negative rapamycin effect on PI3K induced branching can be rescued by Rapa-resistant mTOR mutant. Representative micrographs of hippocampal neurons transfected at DIV7 with the indicated constructs and treated with vehicle or rapamycin (10 nM) for 6 days. p110* - constitutively active PI3K; mTOR (RapaR) – rapamycin resistant mTOR B) Dendritic branching is impaired in S6K siRNA transfected neurons. Representative micrographs of hippocampal neurons transfected at DIV7 with control vector (pSuper) or S6K siRNA construct (pSuper-S6K1293) for 3 days; neuronal morphology was visualized by cotransfected GFP.

mTOR is a serine/threonine protein kinase that controls cell size in both non-neuronal (Fingar et al., 2002; Shioi et al., 2002; Faridi et al., 2003) and neuronal cells (Kwon et al., 2003). In addition, mTOR activity has been implicated in long-term synaptic plasticity (Raymond et al., 2002; Tang et al., 2002; Cammalleri et al., 2003; Hou and Klann, 2004). mTOR is thought to act primarily by phosphorylating eIF-4E binding protein (4E-BP) and p70 ribosomal S6 protein kinase (p70S6K), which are important regulators of protein translation. Moreover, recent findings have shown mTOR involvement to be important for local protein synthesis in neuronal dendrites (Takei et al., 2004). In the context of these findings our recent data describing mTOR-4EB-P1 and p70S6kinase involvement in dendritic branching raises the interesting question whether local or general mTOR signaling is required for dendrite morphogenesis. It serves as a starting point for studying the more general question of the potential role of local protein synthesis in dendritic tree development. However, “chemical genomics” performed on yeast identified 400 rapamycin-dependent mutants, the analysis of which suggested that mTOR might be involved in cellular functions other than translation such as transcription, ubiquitin-dependent proteolysis and microtubule stability (Chan et al., 2000; Xie et al., 2005). That raises general question that should be answered first - what are mTOR dependent proteins and cellular processes involved in dendritogenesis process?

To answer these question we want to:

1) Identify mTOR-regulated proteins in neurons by a proteomic approach

2) Design siRNA library against mTOR-regulated proteins expressed in neurons. and perform systematical screen for these mTOR-dependent proteins that are involved in process of dendritic branching.

3) Establish a link between local protein translation and physiological dendritic arbor development. This task requires first establishing strategy for specific inhibition of local protein synthesis. This unique technology will help us to check how local protein production in dendrites contribute to their development.

4) Characterize both, mTOR-regulated cellular processes and local protein synthesis in dendritic arbor pathologies observed in MR, stress and schizophrenia.

Zinc homeostasis in neurons in vitro and in vivo.

It has been known for years that zinc plays an important role in regulating central nervous system function (for review see: Colvin et al., 2003). Zinc is crucial for proper nervous system development (Prasad et al.,1997). Zinc deficiency in adult mammals leads to memory deficits and higher susceptibility to stress (Sandstead et al., 1984). In living cells there are two different pools of zinc. One bound to the proteins and the second pool of free, ionic zinc. The latter stored mostly in the presynaptic terminals of neurons can be released upon the “arrival” of action potentials. The released zinc can bind and modulate the activity of postsynaptic receptors or crosses the postsynaptic membrane to trigger intracellular signaling processes (Frederickson and Bush, 2001; Li et al., 2003). Moreover excessive Zn²⁺ influx into postsynaptic neuron has been proposed to play role in excitotoxic cell death of neurons induced by ischemia, traumatic brain injury and seizures (Choi et al., 1988; Frederickson et al., 1989; Calderone et al., 2004). Along these lines zinc chelators have been shown to be neuroprotective. Nevertheless, the mechanism of zinc related neurotoxicity is controversial (see for example: Takeda, 2000; Frederickson et al., 2004). One proposed explanation suggests that zinc kills neurons by entering mitochondria, inducing prolonged superoxide production and mitochondria dysfunction (Sensi et al., 2002; Bossy-Wetzel et al., 2004). On the other hand, mitochondria have also been proposed to be a physiologic store for zinc (Colvin et al., 2003; Sensi et al., 2003). Furthermore, there is molecular evidence that ionic zinc in non-neuronal cells can be pumped into other organelles such as Golgi apparatus and endosomes by zinc transporters of the SLC30 family (for review see: Colvin et al., 2003; Palmiter and Huang, 2004). Of particular interest - Znt6 has also been detected in neurons (Huang at al., 2002).
Our collaborative research with Prof. Lippard’s group (Massachusetts Institute of Technology), suggests that zinc can be stored in the Golgi apparatus of neuronal cells and that Golgi zinc content can be increased upon cell stimulation with glutamate, (see Figure below).

We intend to continue investigating the regulation of zinc in neurons with high and low endogenous zinc content, derived from dentate gyrus and CA1-3 fields of hippocampus, respectively. We plan to use novel organelle-specific zinc sensors designed by Prof. Lippard’s lab to trace zinc changes selectively in different cell compartments. It will be of great interest to check whether - Golgi or mitochondria serve as zinc stores under physiologic pathological conditions. The ultimate goal of these studies will be to determine if increasing the zinc buffering capacity by genetic manipulation of zinc transporters in cells vulnerable to zinc toxicity (CA1, CA3) will increase their survival rate after glutamate treatment in vitro and after ischemic or seizure episodes in vivo.

PSD-95:FP transgenic mice – tool to study synapses in the living brain
Transgenic mice carrying variants of fluorescent proteins (FP) are very useful tools to look at processes of neuronal plasticity in vivo, as they allow direct visualization of changes in motility and shape of spines, the anatomical equivalent of excitatory synapses (Majewska and Sur, 2003). Recently, PSD95:GFP protein was used to image synapses in intact brain of living zebrafish (Niell et al., 2004). To study the behavior of synapses in mammalian brain during development as well as in response to sensory stimulation in collaboration with Dr. Morgan Sheng laboratory (Massachusetts Institute of Technology) generated mice which express a fusion protein of PSD95 and the fluorescent protein - DsRed2 under the control of the forebrain specific – CamKIIα promoter. For further insight into recently activated synapses, we also generated a line of transgenic mice where PSD95:YFP gene expression is controlled by the promoter for c-fos, an immediate early gene highly upregulated by neuronal activity. Therefore the fusion protein should be expressed only in a subpopulation of activated neurons (see Figure below). Currently, generated mouse lines are being characterized.