synapse assembly and plasticity!

so i went to a lecture this morning given by Ann Marie Craig, who is currently at UBC, talking about the work she's been doing there and previously at WashU on synapse assembly and plasticity. most of her work was focused on neurexins and neuroligins, which i incidentally did a project on last summer quarter for bio 401. i'm not going to go into detail about these molecules right here and now, but the short of it is that neurexins are expressed on presynaptic neurons and neuroligins on postsynaptic neurons, and the interaction between NXs and NLs is a key interaction in synapse assembly. what makes these guys so difficult to study is the fact that each one has several splice isoforms - especially the neurexins. the long (alpha) isoform of neurexins has 5 splice sites, and the short (beta) isoform has 2 (confusingly called S4 and S5, since they are homologous to the 4th and 5th splice site in alpha-neurexins). so the Craig lab (as i understand from her talk this morning) has shown among other things that the "insert" in splice site 4 (S4) of neurexin 1-beta makes it bind specifically to neuroligin-2, which is specific to GABAergic synapses (which are generally inhibitory), whereas the form without the insert is much more likely to bind to NLs 1, 3 or 4, which are more specific to excitatory/glutamatergic synapses. they also showed that fibroblasts (basic, non-neural cells) co-cultured with hippocampal neurons, and ectopically expressing NX, can induce dendrites to form what she termed "hemipostsynapses" onto the fibroblast, whereas fibroblasts expressing ectopic NL can induce axons to form "hemipresynapses". pretty cool work.

but by far the coolest thing (IMHO) in her talk was some time-lapse imaging they did on cultured neurons with a fluorescently tagged version of CaMKIIa, or calcium/calmodulin-dependent protein kinase II-alpha. (Protein kinases are proteins which add a phosphate group to other proteins, and they are crucial for many intracellular signal transduction events.) CaM kinases have a special domain or subunit, calmodulin, which binds to calcium ions and becomes active. this protein is important in transducing signals from Ca ion concentration into phosphorylation signals... which is key to promoting synapse assembly and potentiation. so the cool data that she showed in the talk was that if a cultured hippocampal neuron with this fluorescently tagged CaMKIIa is stimulated with a "puff" of glutamate/glycine solution, within seconds of application, the (previously uniformly located) CaMKIIa clusters at synapses, and this wave of CaM movement propagates (in some cases) across the neuron, from dendrite through soma to axon. so what is going on? Calcium binding to the CaMK is somehow triggering it to relocate. what exactly is going on, she couldn't offer any ideas... but it is a cool result and i'll be interested in seeing what else they find out.

NMDA receptors and hypoxia

(Note to my regular readers: A while back I toyed with the idea of starting a science blog, where I'd muse about the scientific questions and problems that float around in my head. I quickly realized that I do not have time to keep up a new blog, and given the infrequency with which I post to this one, I figured I could just kill two birds with one stone, and publish the occasional geekiness. If this post makes your eyes glaze over, I apologize.)

My scientific interests are pretty far-flung; I'm getting my bachelor's degree in Cellular, Molecular and Developmental biology this June, but I'm also interested in biochemistry, biophysics and physical chemistry, and neuroscience (from ion channels to systems neuro to the nature of consciousness). I read a lot for fun; today, I'm reading Hypoxia Enhances S-Nitrosylation-Mediated NMDA Receptor Inhibition via a Thiol Oxygen Sensor Motif, by Hiroto Takahashi et al down at UCSD, published in the January 4 issue of Neuron. Hypoxia is another of my random interests; any of you who have heard about my experience last summer on White Mountain know that. That said, this paper isn't exactly about high-altitude physiology, it's about the brain's mechanisms for protecting neurons in the hypoxic conditions that come about during stroke. in hypoxic conditions, it is logical that a system would want to attenuate NMDA receptor activity. Why is that, you ask?

NMDA receptors are one of the brain's types of glutamate receptors; glutamate, though it's less well-known than some other neurotransmitters, like dopamine or serotonin, is actually the central nervous system's major excitatory transmitter. The NMDA receptor has been extensively studied as a model for other neurotransmitter receptors; it's one of the best-characterized of the ligand-gated ion channels. It's also expressed in almost every region of the brain, indicating global importance in CNS function.

So why would the brain want to shut down NMDA receptors during hypoxic conditions? Consider that molecular O2 is absolutely necessary for brain function, but it also creates an oxidizing environment, leading to free radicals that can damage proteins and cause brain damage. The brain thus keeps its oxygen levels carefully rationed and controlled. PO2 in ambient air is around 140 mmHg; pO2 in the brain is closer to 10 mmHg. So there's very little room for error; O2 use must be controlled. In stroke, regions of the brain are cut off from the blood supply and thus have limited O2 resources. Oxygen is critical for ATP formation; ATP is consumed by the Na/K ATPase, or sodium/potassium pump, the membrane ion transporter responsible for establishing the transmembrane potential that is required for all ion flux and thus any neural activity whatsoever. NMDA receptors are ligand-gated ion channels; they create a path for ions to move down the concentration gradient established by the pump, thus requiring more ATP to re-establish the gradient and allow for further neural activity. If oxygen supplies are low, ATP cannot be produced in excess, and so the ion concentration gradient must be preserved in order for neurons to survive. It's a bit like how when the power is out, you keep the refrigerator door shut, unless you want all your food to spoil. It takes energy to establish the gradient (in this case, electricity establishes a temperature gradient across the door of the refrigerator), and by creating a path for the transfer from one side to the other, you destroy the gradient.

So what did Takahashi et al discover? They knew from previous work that there was a pair of cysteine residues in the NMDA receptor structure that formed a disulfide bridge under oxidizing conditions, but existed in a dithiol form under reducing conditions. So they solved the crystal structure of this part of the protein, and studied it under nitrosylating conditions, and discovered that the dithiol can be nitrosylated, but that in and of itself does not inhibit the enzyme. However, nitrosylating these two cysteine residues does make another cysteine residue more prone to nitrosylation as well, and it is this third site that actually inhibits the enzyme.

Interesting paper - I love to see structural biology tied in with ion channel physiology. It's also interesting to learn more about the importance of nitric oxide (NO) in short-range, short-term neural signalling. Considering that 20 years ago, we thought that NO was just a component of pollution, we have come to discover that it is critically important for everything from regulating blood pressure to neurotransmission.