SF - AMPA/kainate Flashcards
Evidence for glutamate as a NT
- Synaptic VGLUT transfers cytosolic glutamate into vesicle
(VGLUT = vesicular transporter on vesicles in pre-synaptic terminals- conc. Glu into vesicles) - NT released in Ca-dependent manner
- Specific protein targets (ionotropic & metabotropic receptors)
- Postsynaptic: all 3 ionotropic & mGluRs
- Pre-synaptic: mGluRs, some evidence that KARs & NDMARs also expressed presynaptically (reasonably convincing but less strong)
- Glial cells (astrocytes): mGluR - surrounds synapses - Rapid removal of NT from synapse: glutamate transporters present on pre- and postsynaptic nerve terminals (EAAT = excitatory AA transporters
- (pre: EAAT2? unsure if exists, post: EAAT3)
- also on astrocytes (EAAT1/2)- variety of transporters expressed on glial cells, take glutamate into astrocytes - Process for glutamate synthesis: glutamate-glutamine shuttle and metabolic processes within presynaptic terminal
- enzyme pathway within pre-synaptic neuron where glutamine is converted into glutamate via glutaminase; then glutamate packaged into vesicles
- glutamate also cycled back from the astrocytes to the neuron so it can be used again
- glutamate not just a transmitter; also triggers cell death so can’t have high levels in brain (think stroke mechanism)
- to get glutamate from astrocyte → neuron, it’s converted to glutamine by the enzyme glutamine synthetase as this won’t activate cell death pathways
- glutamine then pumped out by glutamine transporter, shuttled across to neuron, taken up by glutamine transporter again, then converted back to glutamate (via glutaminase)
Main structure of AMPARs
Tetramers (4 subunits) assembled from GluA1, GluA2, GluA3 and Glu4:
Homomers (one type subunit) or heteromers (different subunits - any combination)
- GluA1-GluA2 relatively common type of heteromer
Differences in subunit composition in different brain areas (or neurons within brain region) leads to diversity in the functions and properties of AMPARs and in their sensitivity to pharmacological agents e.g. GluA3 highly expressed in some brain regions & less in others
Even in same regions, different types of neurons can express different combinations: drugs can be made to selectively target some subunits
AMPA and Kainate Receptor Subunit Topology?
Topology = geometrical properties/spatial relations
Consists of two dimers: e.g. 2 GluA1-GluA2 dimers come together to form tetramer
Amino terminal domain (ATD)
- extracellular
- contains clam-shell like domain (but nothing in particular binds it)
Ligand binding domain (LBD)
- Contains S1S2 domain, contains clam shell domain where GLUTAMATE binds
- S1 domain (extracellular peptide loop from TM1 → N terminal)
- S2 domain (extracellular peptide loop joining TM3 + TM4 & therefore the carboxyl terminals)
Transmembrane domain (TMD)
- contains pore of ion channel & intracellular carboxyl terminals
- note TM2 doesn’t go fully across the membrane; it forms a RE-ENTRANT loop which forms lining of PORE
TM2 has Q/R site very close to pore, determines calcium permeability
((All AMPA receptors: Na⁺ in & K⁺ out, some receptors will allow calcium in but not all))
What is RNA editing in AMPA receptors?
GluA2 DNA encodes for glutamine (before transcription)
Q/R switch: after mRNA produced, the triplet coding for glutamine (Q) is switched to arginine (R) in the mRNA
Edited gluA2 mRNA has arginine (R) at residue 607 (the Q/R site). This change is expressed at a site in RE-ENTRANT LOOP 2 in GluA2 that controls Ca²⁺ ion permeability through the ion channel.
RNA EDITING causes calcium IMPERMEABILITY: arginine doesn’t allow calcium through the pore
AMPA receptors containing EDITED GluA2(R), = Ca impermeable
AMPA receptors containing UNEDITED GluA2 (Q) OR LACKING GluA2 = Ca permeable
EDITED (GluA2(R)-containing receptors) also have lower single channel conductance
(how much charge channel will carries when activated). Therefore, editing means less current passes through
What kind of subunits do most AMPAs show? In what situations are different types found?
Most AMPARs in adult rodent brain = EDITED(R) GluA2, so Ca²⁺ IMPERMEABLE (most research = rodent but also seems true of adult humans). Ca²⁺ important second messenger- tightly controlled, therefore most AMPARs are Ca²⁺-impermeable
GluA2-LACKING AMPARs expressed:
- following induction of some forms of LTP
- chronic stress (↑proportion of AMPARs without GluA2, replaced with other subunits)
- early during development (first 2 postnatal weeks in rodents; synapses formed, more plasticity, AMPARs probably involved)
- some GABAergic interneurons
In disease e.g.
- after stroke, down-regulation of GluA2 - (unlikely to be helpful as calcium can trigger cell death?!)
- MND patients show decrease in GluA2 EDITING in motor neurons: may contribute to the death of the motor neurons (NOTE: this is less EDITING rather than less GluA2)
What are examples of AMPA antagonists?
CNQX: AMPA, Kainate (& NDMA) antagonist
- NBQX: AMPA selective
- COMPETITIVE: compete at LBD, not selective for specific subunits
Blake et al (1998): Electrophysiology: glutamate synapses between CA3 + CA1 in rat hippocampus slice (record post-synaptic cell, stimulate axons, record response).
- Control condition: depolarisation (glutamate-mediated. CNQX applied: block glutamatergic synaptic response. Wash drug away; response returns. (same happens with NBQX).
What kind of transmission at AMPARs?
Fast excitatory synaptic transmission mediated by AMPA receptors in most areas of brain (hippocampus and many other regions)
AMPA mediates majority of FAST BASAL synaptic transmission (everyday communication between brain regions)
(some regions: KainateRs also involved, AMPA key in hippocampus and other areas)
What is an antagonist at AMPARs WITHOUT EDITED GluA2?
*IEM-1460: antagonises AMPA receptors without edited GluA2
Magazanik et al (1997): recorded from GluA3 tetramers or GluA3/GluA2(R) tetramers (only receptors expressed on cell). Used kainate as agonist of AMPARs - causes inward current, partially reversed by 1um of IEM-1460.
Increase conc of IEM-1460 (2µm), more of a block - reverse when wash off antagonist
With GluA3, 1-2µm gives large ~50% block, but with GluA2(R) subunit, 100µm only causes about 10% inhibition
What is PTX?
Philanthotoxin (PhTX-74) from wasp venom
Antagonist in absence of GluA2
Enters pore & interact with Q/R site to block the ion channel, but won’t fit when GluA2 present
Poulson et al (2014):
Concentration-response curve, increasing concentrations of PhTx-74: for GluA1 homomer and GluA3 homomer, PhTx quite effective at reducing size of response (EC50 about 0.3mM)
If GluA2(R) present (with either GluA1 or GluA3), drug much less effective, about 100x change in potency (two log units shift to right)
*Therefore PhTX-74 can distinguish between AMPA receptors that contain GluA2 and those that don’t, useful in determining when we see these receptors
Main structure of KARs?
Tetramers (homomeric or heteromeric):
GluK1, GluK2, GluK3, GluK4 & GluK5
GluK1-3 = LOW affinity for kainate
GluK4-5 = HIGH affinity for kainate
(but less than 5x difference - not big diff)
GluK4 & GluK5 have an RXR RETENTION MOTIF in C-TERMINAL tail: when protein expresses motif, gets RETAINED in ENDOPLASMIC reticulum, never gets trafficked to cell surface
HOMOERMIC assemblies of GluK4 or GluK5 not surface expressed (also tetramers of GluK4+GluK5). However, GluK1-3 can combine with GluK4 or GluK5 to produce tetramers that are surface expressed (retention is overcome & receptor trafficked to surface)
*subunit composition varies in different brain areas, therefore differences in function/pharmacology in different brain regions
What is the function of KARs in CA1 of hippocampus?
CA1 (and other areas): KARs = presynaptic, DEPRESS glutamate transmission
Depolarise axon terminals, causing INITIAL SPIKE of glutamate release (AMPA EPSC amplitude increases)
BUT VG CALCIUM channels INACTIVATE with prolonged depolarisation → depolarisation block → less glutamate released → DEPRESSION in the AMPAR response (AMPA EPSC amplitude decreases)
*NS-102: kainate receptor antagonist, blocks this effect of kainate (AMPA EPSP amplitude is maintained despite kainate application)
What is the function of KARs in CA3 region of hippocampus?
CA3 = high density of POST SYNAPTIC KARs
Allow sodium influx (+maybe calcium) - depolarises cell contributing to EPSP
TRAINS of stimulation (rather than 1 stimulation causing 1 AP), e.g. 5 stimulations in short period → BIGGER GLUTAMATE RELEASE→ trigger a SLOW KAINATE RESPONSE (slow depolarisation lasting ~100ms (in AMPA, response over by 10-15ms)
This slow depolarisation is blocked by CNQX (AMPA/kainate antagonist)
LY294486: first selective(ish) kainate receptor antagonist - also suppresses this response
* EPSC = excitatory post-synaptic currents
How can Xray crystallography be used to visualise ionotropic receptors?
1) Grow part of the protein in solution; crystals start to form as protein precipitates & increases in size. Grow crystal of AMPAR.
2) Pass X-rays through crystal structure: results in diffraction of X-rays.
3) Measuring angles and intensities of the beams = = 3D picture of the electron densities (electron density map)
4) Mean positions of atoms can be determined from this, and their chemical bonds
- can be used to develop drugs that will selectively bind to different regions of the receptor (most common way is to take LBD with the agonist bound, grow a crystal of that and look at the structure)
Whole receptor much bigger protein, more difficult to do
Isolated ligand bind domain (LBD) structures of glutamate receptors are more common than those of the full tetramer - structures of membrane bound proteins difficult to obtain: correct protein folding outside the membrane difficult to achieve due to hydrophobic transmembrane domains
What is the crystal structure of antagonist bound GluA2 tetramer?
ATD has a clamshell structure and has a LOW SEQUENCE HOMOLOGY with BACTERIAL PERIPLASMIC proteins and mGluRs - it is involved in receptor assembly and trafficking.
LBD has a clamshell structure and is made up of S1 & S2 segments. The TMD has similar structural features to an inverted K⁺ channel pore.
- Structure of the C-terminal tail was not resolved (easier to get the crystals without the C-terminal, and not particularly long anyway, therefore don’t know structure of the tail)
What is the position of the LBD in ionotropic glutamate receptors? How do we know?
1) Bind agonist to LBD, then cut using RESTRICTION ENZYMES; cut particular regions e.g. top of transmembrane regions
2) Add a LINKER so that it stays as one protein, then CRYSTALLISE the region (very simplistic view because isolated bit of protein, but hopefully will relate to structure when it is in the complete protein)
- working out exact LBD structure/exactly where agonists bind - develop drugs that bind some receptors + not others?
- using molecular biological techniques, cDNA encoding for only the S1S2 REGION was isolated (gave access to large amounts of protein which could be crystallised, in some cases with agonists/antag. bound to S1S2 region
How can agonists and antagonists stabilise different LBD conformations?
If grow crystal with agonist bound, e.g. AMPA with GluA2, S1/S2 shut, CLAM SHELL CLOSES, causes ACTIVATION of receptor
If grow crystal with antagonist bound e.g. UBP282 with GluA2, clam shell domain is WEDGED OPEN, prevents activation/shutting of clam shell
*technique not dynamic - can’t see process of change between open/shut
What are the binding sites of AMPAR NAMs?
- NAMs antagonists but not bindings to LBD
- Bind close to ion channels at all 4 subunits; stop activation of receptor (binding of agonist to AMPAR/KARs causes closing of shell which causes TWISTING of receptor to open the ion channel: GYKI STABILISES the receptor to STOP ion channel opening.
GYKI & PERAMPANEL bind in TMD region adjacent to the PEPTIDE linker connecting the TMD to the LBD (bind to a site formed by pre-M1, M3 & M4 in each subunit - and to residue to an adjacent subunit - prevents movement of transmembrane segments necessary for pore opening
(i.e. 4 NAM binding sites per tetramer)
Prevents the movement of transmembrane segments necessary for ion channel pore opening.
What is perampanel?
Perampanel: AMPAR NAM
Anticonvulsant activity and is approved for treatment of patients with partial onset seizures (these affect only part of the brain at onset)
How have crystal structures aided design of potent GluK1 KAR antagonists? What has this provided insight into?
Insight into structural basis for KAR vs AMAPR selectivity
Willardine = AMPAR agonist
UBP310 = exactly the same, but with a thiophene ring = kainate selective agonist (very potent, 18nM = high affinity)
- For GluA2 + GluK2, affinity is over 100µM, therefore much more potent for some kainate receptors than others
- Most potent → least potent: GluK1 → GluK3 → GluK2 + GluA2 (AMPA)
- not good at all for GluK2/AMPA subunit
How is UB310 selective?
The THIOPHENE ring forms favourable interaction with a VALINE residue in GLUK1 to enhance binding affinity at the LBD
Valine residue in GluK1 replaced by leucine residue in AMPARs (e.g. GluA2)
-This larger residue in AMPARs impedes antagonist binding (leucine residue takes up some of the space where UB310 usually binds) - explains GluK1 vs AMPAR selectivity
