Glutamate receptors Flashcards
Evidence for multiple receptors
Hestrin et al 1990 - there are two components to excitatory transmission, a fast one, and a slow one seen only at less hyperpolarised holding potentials. Fast one blocked by CNQX
This slow one is NMDA receptors, because they’re blocked by external divalent cations when hyperpolarised. Slow one was blocked by AP-V
A third, even slower one was seen at higher stimulation strengths or increased numbers of stimulation. This was the mGluR.
Criteria for a neurotransmitter
Method for synthesis and storage in presynaptic neuron
–enzymatically, from alpha-ketoglutarate, by GABA transaminase. Glutamine is transported between glia and neurons, converted to glutamate (or GABA) in the neuron–
Released upon stimulation
Method for inactivation and removal from cleft
–Glutamate transporters, on presynaptic neurons and glia–
Where is glutamate found?
Hard to stain for with antibodies because it’s so small. Can fix and then stain for glutamate bound to proteins.
Immuno-gold labelling, where the secondary antibody has gold attached to it, shows that there’s not much glutamate at the cell body layers of the hippocampus, only where there are synapses. Allows labelled EM.
Excitatory synapses are asymmetric (because of the postsynaptic density), and have glutamate at the presynaptic terminal but not postsynaptic.
Inhibitory synapses are symmetric, and have no glutamate at either terminal.
Where are glutamate receptors found?
Staining for GluR2/3 subunits of AMPA receptors showed an equal distribution across the bouton.
But Nair et al 2013 super resolution imaging suggested AMPA receptors are clustered in highly regulated nanodomains that move about, regulated by PSD-95 (which binds to C-terminus of TARPs), depending on LTP and LTD. Strength of synapse is correlated with larger PSDs, more PSD-95 and more AMPA (Chen et al 2011)
Staining for NR1 subunits of NMDA receptor showed more expression in the centre.
PSD-95 can bind the C-terminal of NR2 subunits, tethering them to the centre of the synapse.
Determining the structure
Hollmann et al 1989 - cloned the first of 18 genes encoding members of the iGluR family. They screened a rat brain cDNA library for a kainate-activated channels expressed in Xenopus oocytes. The hydropathy index of it primary sequence allowed us to predict where TM segments might be.
In 1990, four members of the AMPAR family were found from another cDNA library search. They had approx 70% sequence homology. Three had an glutamine and one an arginine in TM2, which forms the channel. The different members (presumably subunits) were widely but differently expressed in different regions of the rat brain. Kainate receptor expression was more restricted.
Ways of generating variation in subunits
Alternative splicing in a region between TM3 and TM4 was found to be important, generating ‘flip’ and ‘flop’ variants, which were activity-dependently switched and differently expressed in the rat hippocampus.
Sommer et al 1991 - the arginine instead of glutamine in TM2 profoundly altered ion channel properties. Whilst mRNA of one subunit contained arginine codon, the genomic DNA of all subunits contained glutamine. They excluded multiple alternative exons as potential sources for this arginine, so concluded it must be RNA editing. This was one of the first examples of RNA editing found. Now known to be via adenosine deaminase, turning A to I.
Verdoorn et al 1991 - The subunit with an arginine in TM2 is called Glu-B, aka GluA2. Homomers of A, C or D subunits showed double rectifying IV curves, while a homomer of B subunits showed a simple outward rectification. The presence of B in heteromers altered their function in a dominant way, due to this arginine (as demonstrated by site-directed mutagenesis - i.e. turning the glutamine in D to arginine made it behave like B, and vice versa in B).
Now we say that GluA2 subunits confer a linear voltage relationship (because v depolarised potentials are not physiological), and are impermeable to calcium. The same Arginine as prevents calcium permeability also prevents spermine entry and block. AMPAs without GluA2 are permeable to calcium, and show inward rectification due to a spermine block.
Isa et al 1995 - Over about 10 minutes, AMPA receptors lost their inward rectification. The rectification could be maintained by application of physiological amounts of spermine to the cytoplasmic side of patch membranes.
Calcium permeability
Iino et al 1990 - found two types of kainate induced current response. Type I was not permeable to calcium and showed outward rectification in the control external solution. Type II was permeable to calcium, and showed inward rectification.
Hence the arginine in TM2 confers calcium permeability., and thus RNA editing is critical.
Calcium impermeability thus requires both expression of GluA2 subunit, and of adenosine deaminase to RNA edit it and put arginine in it
Molecular structure of AMPA
TM2 is more like a loop, doesn’t pass all the way through
Normally exist in tetramers
Molecular structure of GluA2 homomers:
Big N terminal domain, consisting of ligand binding domain and amino terminal domain
Relatively small C terminal
Tall, N shaped cross section, Y shaped coronal section
But not that useful, since GluA2 homomers don’t exist in nature…
GluA2/3 heteromers have an O shaped cross-section/pore, and way shorter, so wouldn’t stick out so much into synaptic cleft - interestingly, the heteromer’s height is the same as an NMDAR.
It’s also been suggested that the extracellular region is flexible enough to move between the Y shaped and O shaped conformations, perhaps affecting the binding of linker proteins like cerebellin, which binds presynaptic neurexin and postsynaptic iGluRs, perhaps to alter receptor clustering.
AMPA Regulation
Auxiliary subunits like stargazin (because mutations in it caused mice to keep looking up)
TARPs and other auxiliary subunits are important for trafficking to the membrane, and for ‘diffusion trapping’, keeping the receptors at the synapse, in association with scaffolding proteins in the postsynaptic density like PSD-95.
Constals et al 2015 - desensitisation causes a loss of the AMPAR-stargazin interaction, increasing AMPAR mobility and allowing it to move away from the synapse so it can be replaced by non-desensitised receptors, allowing fast recovery from postsynaptic depression
CaMKII is activated during synaptic plasticity induction, and phosphorylates stargazin C-terminal serines, causing it to dissociate from the membrane - maybe this is important in plasticity?
Differential expression of TARPs may allow functional heterogeneity of AMPA. gamma-2 can increase AMPAR number at a synapse, while gamma-8 just maintains basal level, yet it’s gamma-8 that is a critical CaMKII substrate in LTP (Park et al 2016)
gamma-2 can also increase single channel conductance, alter kinetics, and affect spermine block (without abolishing inward rectification)
NMDA structure
Think of them as coincidence detectors! Between glutamate release and postsynaptic depolarisation
Has asparagine where GluA1,3,4 had glutamine and GluA2 had arginine
Voltage dependent magnesium sensitivity (when depolarised, block is relieved. This is similar to spermine causing inward rectification, but at increasingly negative voltages the spermine block is removed and there is inward current, whereas magnesium block is never totally removed by hyperpolarisation, only depolarisation, allowing outward current only. Remember spermine blocks from the inside, magnesium from the outside. Magnesium block removed by depolarisation, spermine by hyperpolarisation.)
Obligate heteromers - the prototype is a tetramer formed of two heterodimers, with 2xNR1 and 2xNR2.
Extensive interactions between amino terminal domain and ligand binding domain.
ATD has clamshell structure, just like AMPA ATD.
Two NR1 subunits (8 possible splice variants)
Two NR2 subunits (4 splice variants, but mainly A and B)
NMDA - Subunit variation and functional effects
NR1 is obligatory (synthesised throughout the brain), NR2A/B/C etc have differential expression, changing throughout development and activity dependent. May confer different plasticity tendencies - an reduced NR2A:NR2B ratio is thought to be permissive for LTP (Cho et al 2009)
Applying TTX to block spiking prevents the subunit switch from occuring, and prolongs the period of plasticity. Rearing the animals in the dark had a similar but not quite so dramatic effect, i.e. delaying but not abolishing the decrease in NMDA-mediated EPSPs.
During development, NR2A replaces NR2B. NR2A cannot activate CaMKII, so is less amenable to plasticity
Molecular structure of NMDA
LBD has clamshell structure, just like AMPA ATD.
ATD is much more compact, interdigitating with LBD, creating a short extracellular region like the O shaped AMPA structure
Extracellular domains of different subunits are closely knitted together, with ‘crossover’ just like AMPA. Further crossover in TM domains, because M4 helices interact almost exclusively with TM domains of the neighbouring subunit.
Within the LBD there are two main interfaces, one within the heterodimer, one between dimers.
Binding of agonist causes closure of LBD clamshell
Modulation of NMDA
NMDAR differs from AMPAR in that it has many allosteric modulators.
NR1 binds:
Coagonists glycine and D-serine
Protons - reduce the open channel probability without altering deactivation or desensitisation time course, and without affecting agonist or coagonist binding. Banke et al 2005 suggest that a protonated receptor is identical to an unprotonated one, except that it cannot open.
NR2 binds:
Polyamines - increase currents but also enhance binding of open channel blockers, may affect binding of glutamate and glycine but possibly via a different site
Zinc - accumulates in presynaptic vesicles of glutamatergic synapses. Binds high affinity site on NR2 to reduce open channel probability, binds low affinity voltage-dependent site in pore to block channel. Kinetic modelling suggested that zinc-bound receptors have higher energy barriers to opening and less stable open states.
Heterotrimeric NMDAR
Lu et al 2017 - The most commonly found NMDAR complex is a heterotrimer of NR1/NR2A/NR2B. Using cryo-EM:
NR2A’s clamshell was closed unless zinc is present, NR2B was open unless an allosteric antagonist is present
NR2A’s ATD interacts more extensively with NR1
NR2A’s LBD interacts with NR1’s at both major and minor interfaces
The presence of NR2A reduced the efficacy and affinity of allosteric antagonists such as Ro.
Summarising blocks and permeabilities
All NMDARs are calcium permeable
All NMDARs can be blocked by extracellular Magnesium
The Mg block causes inward rectification
The Mg block is removed by depolarisation
Only AMPARs without GluA2 are calcium permeable
Only AMPARs without GluA2 are blocked by intracellular spermine
The spermine block causes inward rectification
The spermine block is removed at both negative and positive potentials (i.e. there’s a double rectification), but physiologically you’d never be depolarised enough to see that end of the block removal.
