Nicotinic Acetylcholine Receptor

Cholinergic Systems

• Acetylcholine binds to its receptor which is either a nicotinic or muscarinic Ach receptor which have differential paharmacology 

• Nicotinic receptors activated by nicotine, muscarinic receptors activated by muscarine 

• Alpha-bungarotoxin is an inhibitor of the nicotinic Ah receptor, atropine binds to the muscarinic receptors and acts as an inhibitor 

• Nicotinic Rs are ionotrophic found primarily in the sympathetic and parasympathetic NS, allowing them to have a rapid response time (microsecond/millisecond) 

• Muscarinic Rs are G protein coupled receptors found in the parasympathetic NS and act a lot slower (milliseconds to seconds)  

Strategies for isolating and studying receptors

• In history receptors needed to be abundant to study which is why these receptors are good targets  

• Receptors initially defined by their pharmacology, so people able to identify high affinity ligands which can help them purify them  

• Need to find a rich source of receptor which allows you to label ligand with radioisotopes, label the tissue that the receptor is present in which allows you to localise it in the body 

• Once the receptor has been labelled allows you to purify it and assess the binding properties 

• Purification of the receptor allows you to understand the molecular change, create agonists for the receptor and to obtain a partial sequence  

Nicotinic Ach receptor isolation

• Can find a rich source of protein in electric fish organs called torpedo nobiliana (can obtain several bands) 

• Has this electroplax tissue on the back of its head which allows it to shock its prey  

• Can very carefully stimulate this tissue and homogenise the tissue then load it onto a sucrose gradient  

  • Layers of different concentrations of sucrose, as you go down the tube the concentration of sucrose increases  

  • Samples placed on top and spun in a centrifuge and the receptor membranes form a band partway down the tube depending on the amount of protein in the sample  

  • Allows you to obstain bands of receptor membrane  

• Can take the receptor fraction and run it on an SDS gel which will give us bands for the alpha, beta, gamma and delta subunits → expect a thick band of alpha as the receptor comprised of 2 alpha subunits, one beta, one gamma and one delta  

Protein based sequence

• Before we could sequence the whole genome, we could obtain a sequence of the protein via Edman degradation  

• Involves labelling the AA at the N terminus with a colourful molecule, then cleave it off and work out what it is, then cleave the next amino acid to resolve the N terminal sequence 

• Now, instead of this technique we would excise the protein from the gel and do mass spectrometry analysis which involves fragmenting the protein and comparing the mass spectrometry to a database 

• If we know what the protein sequence then we can work backwards and work out the DNA sequence  

From partial agonist to full sequence

• The sequence of Aas can be mapped across to the different codon possibilities 

• We can then screen cDNA libraries from humans and sequence the positive clones 

• This allows us to pull out the sequence for each of the NicAchR subunit

Cloning the receptor

• Take the mRNA from the tissue or organism, convert this into cDNA and insert the cDNA into bacterial plasmids 

• Insert plasmids into bacteria and plate the bacteria, lyse them and hybridise the DNA using a labelled probe which allows you to work out which bacteria contain the cDNA of the receptors that we’re interested in  

• Sometimes the numbering of Aas can vary considerably in the older literature because they haven’t worked it out from the gene sequence but from the protein sequences

What can we learn from the sequence

• We realise from the sequence that all of the different subunits are relatively homologous and structurally the subunits are likely to be similar  

• Can also predict what post translational modifications like glycosylation and phosphorylation from the sequence —> this allows us to predict where the protein might be situated within the cell (glycosylated will be outside the cell) 

• Can also do a hydropathy plot analysis which allows us to identify hydrophobic regions → hydrophobic regions 21-24 residues in length indicate a transmembrane pass which is what we can see from the hydropathy plot and can see 4 TM passes  

Topology and orientation of nAchR subunits

• So far had predicted: an N terminus region rich in beta strands, 4 alpha helical TM domains and a glycosylation site with an EC disulphide bond 

• However, this doesn’t help you understand the pharmacology of the receptor 

• Now start to think about expressing this protein in model organisms such as E. coli, xenopus, insect cell culture or tissue culture → opens up more types of analysis 

Homologous expression in Xenopus oocytes

• Inject the 4mRNAs that you have into the eggs 

• Allows you to: 

  • follow this with electrophysiology  

  • Observe how the subunits come together  

  • Identify binding sites for drugs that we’ve identified  

  • Mutate the mRNA inserted to observe the functional relevance of certain parts of the protein  

Electrophysiological studies of nAchR in xenopus

• 2 electrodes into the egg and either plant the current or the voltage and look at the response (change in voltage or current) to certain pharmacological stimuli 

• Can then remove parts of the subunits, mutate parts of the mRNA and observe the response to see how these perturbations affect the pharmacology 

Eukaryotic membrane protein expression in E.coli

• However, expression of these proteins in bacteria can be problematic because they don’t have the same machinery that humans have 

• Start to encounter problems with:  

  • Different codon usage 

  • Limited membranes (no ER) 

  • Different membrane protein insertion machinery compared to eukaryotes  

  • No post translational modifaction  

  • Lipid composition differs  

  • Over expression of heterologous protein may lead to inclusion bodies  

  • Multi-subunit proteins difficult to assemble 

Detergent solublisation

• Need to be able to take them out the membrane and purify them  

• Was initially problematic, but now we’ve been able to come up with solutions around these problems:  

  • Detergent micelles  

  • Lipid/detergent bicelled  

  • Amphipoles 

  • Phospholipid nanodiscs  

  • Liposomes/bilayer  

• All can be used to extract the receptor out of the membrane so that it can be studies  

Separation of membrane proteins

• Once isolated, needs to be purified  

• Involves chromatography and binding a ligand to a chromatography column and passing the sample over this  

  • Affinity chromatography: may involve his tags or maltose binding protein 

  • Ligand affinity chromatography: may involve using an agonist or antagonist like alpha bungarotoxin  

  • Combination: uses a tag-based system for crude purification and functional selection with an agonist-based column 

• Modern approaches may utilise the his tags attached to maltose binding protein, ligand affinity chromatography older technique and requires you to know what can bind to the receptor 

Reconstruction into lipid vesicles

• If you want to study the function of the protein you need to return them to the bilayer (reconstitution)  

• This allows us to understand how fast things are transported across the membrane to understand the accessibility and the binding sites 

• Challenges involve removing the detergent molceules 

• Do this by mixing protein with lipids, then can do  

  • Dialysis: if you have low conc of detergent its now a monomer (not in a micelle) and monomers leave the dialysis bag → protein is taken up into the lipid vesicle  

  • Biobeads: polystyrene balls with ahigh affinity for the detergent molecules and pull them out of solution  

  • Dilution 

Chemical labelling for the identification of binding sites

• Can look at what pharmacophores involved: 

  • ammonia group often onvolved in binding  

  • Has acetylcholine group which contains oxygen. Oxygen great at binding so suggests that this part may be involved in binding  

• Developed a class of different compounds like  

  • Ach mustard which has a ring under tension → can bind this with the receptor to identify which part of the molecule is interacting with the receptor 

  • BromoAch which has a bromine group which can be reacted with cysteine to identify where the cysteines are in the protein  

  • MBTA: has an ammonium group which localises the reactive functional group to the binding site 

  • DDF has a triple bonded nitrogen which are highly reactive and will react with the binding site 

• Now that we can localise the binding site, we can label them with 14C or tritium and can now track which subunits these molecules are binding to by running them on a gel and tracking the radioactivity 

• Found out that all of these different compounds are preferentially reacting within different groups within the alpha subunit 

• Assuming that all of these different molecules are binding to the alpha subunit, suggests that the alpha subunit forms the primary binding site 

Protease digestion

• Digest this with various subunits on the gel and get these different fractions 

• Can now localise this radioactivity (due to binding of ligand) to a band of 20kD 

• We now know where the protease will cleave this protein and can do Edman sequencing  

• The benefit if this is that if you keep on cleaving off individual Aas at some point you will reach your amino acid which you initially radiolabelled and will indicate that you’ve reached the binding site 

• Can obtain sequence specific information about the localisation of the binding site —> allows you to know the specific Aas involved in binding 

• However, to avoid using radioactivity, we can take the same approach with mass spec → fragment the protein and work out which protein has an elevated Mw 

Identification of Ach binding sites

• You would expect the quaternary ammonium group to bind to a negative charge and things that can form hydrogen bonds  

• But in fact it bonds to tryptophan and tyrosine (aromatic residues) → this is useful because if you want to design a drug, you would know to include these aromatic AAs  

• Did confirm that the alpha subunits is the primary binding site, however, also showed that there was binding in the gamma and the delta subunits → part of the binding site is facing towards this complementary subunit 

• We now know that the binding site is between the alpha and the gamma and the alpha and the delta subunit 

Structural characterisation of the nAchR

• Observed tubular structures if left at -80 for 2 years → these turned out to be an array of AchRs  

• Gave all the information to reconstitute the structure of the AchR  

• Used electron diffraction where you defocus the beam of electron microscopy and get a diffraction pattern which can be used to generates an electron density  

• Then managed to generate a structure of the AchR containing:  

  • Ligand binding domain (extracellular) 

  • Transmembrane domain 

  • Intracellular domain  

Intracellular domain

• This is important because the function of the receptor is modulated by the lipid environment 

• Intracellular domain is involved in localising with the receptor through interaction with the cytoskeleton  

Transmembrane domain

• 5 subunits all come together and contribute their second TM domain to form this pentameric structure which is a channel going through the membrane 

• M2 domain conveys selectivity: possesses a negative charge which selects for positive ions 

• M1 and M3 surrounding this in a ring 

• M4 TM domain makes the interaction with the lipid bilayer. This receptor is very sensitive to lipid environments and cholesterol allows it to be activated 

• No activation of the receptor in membranes devoid of steroids because there’s no interaction of the M4 TM domain and so prevents activation  

• Anaesthetics are hydrophobic and hypothesised that they bind to sites at the lipid protein interface which highlights the functional importance of these domains 

Ligand binding domain

• The residues involved in binding the Ach referred to as the beta sandwich: 2 beta sheets packing together on top of each other, C loop with Cysteine involved  

• Within this ligand binding domain the interface region made up of a cys loop and a beta1 and beta2 loop are very important for receptor activation once the ligand has been bound (perturbing this rotation) 

• Chemical labelling data proved that the binding site was located between the alpha and the delta and the alpha and the gamma subunit as hypothesised  

  • Also Confirmed that there were 4 TM domains per subunit  

• C loop suggested through chemical labelling that this region interacts with the ligand when bound  

• Because it’s a channel, a structural change must be involved In fluxing ions → suggested that this is the role of the beta sandwich  

• Binding site made up of main subunit and complementary residues 

• C loop fold over the molecule binding  

Pharmacophores for mapping Ach binding

• Ach is a very flexible molecule that will fit the shape of the receptor and so is not very useful for understanding the binding of this receptor 

• However, can use other molecules which are a lot less flexible to study binding site 

• So far we know that Ach has quaternary ammonium group which is +vely charged and the Acetly group which has potential for hydrogen bonding 

  • Strychnine has O atom which is separated by 4.97A from the N atom and matches the distance for binding to the Ach receptors  

  • Pancuronium binds with high affinity to Ach receptor and suggests that there's a hydrogen bond acceptor which is separated by 5.06A from the quaternary ammonium group 

A model of the Ach binding pocket

• Solid state NMR allows you to determine the structure of the receptor when Ach is bound based off the resonance of particular atoms → Tells you about the local electrostatic environment 

• We have a positive charge and so it was expected that the binding site would have a negative charge but all they found is a big shell of aromatic groups 

• NMR studies labelling the molecule (13C) when bound to the receptor caused a shift in peak of the quaternary ammonium group 

• Position of the group moved by 1.5 ppm and showed that it was binding to this quaternary ammounium group → Ach being bound by this aromatic cage 

• NMR is very sensitive to aromatic groups and results in these ring current effects 

• Identified one of the most common binding interactions: extensive cation-pi interactions 

  • Aromatic rings made up of double bonds are highly polarisable due to their fluctuating ring structure and form the basis of cation-pi interactions 

  • As the positive charge approaches, one side of the ring acquires a positive charge and the other acquires a negative charge 

Docking studies

• Usually rely on a flexible structure binding to a rigid protein but this time held the structure of the ligand rigid and allowed the protein to flex around it 

• Found the C-loop folding over the ligand 

• Docking studies made use of an Ach binding protein which is a soluble ligand binding protein which can be isolated from snails 

• At the time thought that carbonyl group was important for high binding however, was consistently found net to this leucine molecule incapable of H bonding but Ach binding protein mimics the pharmacology of the NachRs in the brain, not the ones that are found in torpedo nobilliana or the NMJ  

• Lower affinity for Ach in the receptors in the brain due to lack of H bonding between the carbonyl group and the  

• Snail Ach binding proteins which have a high degree of similarity but instead of leucines, NMJ receptors have threonines which are capable of H bonding which gives higher affinity binding 

Ach binding proteins

• Once the crystal structure had been defined, tried to exploit this to generate crystal structures of when antagonists and agonists are bound 

• When the antagonist is present, the C loop seems to be pushed back from the binding site  

• But in the presence of an agonist, we see the C loop wrap around the agonist 

• Lots of aromatic groups involved in binding which promotes the idea of cation-pi interactions and confirms the correct binding sites 

• Gives insight into the type of interactions that these receptors make → useful for trying to optimise interactions between these scaffolds and receptor binding site when developing drugs 

How does movement of the C-loop propagate to the change in structure of the ion channel?

• Activation involves ligand binding sequentially of 2 Ach molecules (microsecond timescale), the channel remains open for a few milliseconds and rapidly, the receptor becomes desensitized and the channel closes and so difficult to study the open structure as its very transient  

• Receptor dropped from a height and atomised Ach sprayed as its falling to activate it, before falling into liquid ethane which freezes it and keeps it in this open conformation  

• Movement of the C-loop in the ligand-binding domain acts like a lever and pulls the rigid beta sandwich across  

• The Cys loops and the beta1 and 2 loops form the interface between ligand binding and TM domain —> are pulled away from the top of the TM domains  

• The AchR acts in a prestressed conformational state, which is being held under tension in the closed state  

• The moment the loop  regions are pulled as a result of agonist binding, then you can release the energy of the Tm doamins and can open up the channel of the bilayer 

• These loops interact with the M2 TM domains which are involved in lining and opening the centre of the channel 

• In the closed conformation all the Lecuines are facing the centre, but when the channel opens, they get replaced with these serine molecules which allow the flux of ions 

Summary

Prokaryotic ligand-gated ion channels

• Ligand gated ion channels in bacteria which are more amenable to crystallography  

• Don’t bind Ach  

• ELIC and GLIC:  

  • ELIC has a closed state conformation, while GLIC has an open state conformation  

  • Allows us to get a crystal structure associated with the open/closed LGIC 

• Proposed a revision to the original model and suggested the opening and closing was driven by the beta1 and 2 loop as opposed to the Cys loop  

•  Causes a rotation of beta sandwich and results in the release f the M2 TM domain and interaction f the channel 

• Instead of the rotation of the M2 TM domain, there was a pulling apart of the M2 domains of 12A to allow the flux of ions