 Thanks very much indeed Tommy and Hazer and everybody for organizing this, it's been fun putting this presentation together and it did remind me I think my first publication with Andrew Miller in neutrons was in 1984, I was using ISIS nearby here at Oxford, but of course I've been heavily involved in ILL over the years and the complementation between deuteraminomar and deuteration in neutrons was clear to me very early on in my career and a number of papers have come out with both methods in the same publication and it's been really fun to do, I have to say. I have to admit now that I've now mandatory retirement in Oxford, you have to leave at the 70s so I've been gone from Oxford for about six months now so well, but it was fun putting this together and thank you for inviting me along. I did look at the profiles of everyone and I think I saw NMR in two of them and I don't know and Tommy you might be able to tell me if you're close to this whether everyone is familiar with NMR or not or should I just give a brief overview of NMR to help people. What do you think? I think a very brief overview would be helpful. Okay a brief overview then right now I've seen some of the earlier talks so I know what to talk about. Okay good so NMR what can it give us? It can give us molecular structure and identification of unknown chemical substances, it's used for quantitative analysis, it's used in analysis of mixtures in particular in food industry for example, it's important there. We can get dynamics and the dynamic time scan I will address because that's very important in biology, it doesn't move, it often doesn't do anything so that's important and I put this in on top of this slide from jail because many of you may well have had or no people who've had MRIs and scans and so on and this is exploding proton NMR this is where it all came from. Dynamics in terms of chemical reactions, speed of reactions, finding sites and drug discovery and so on so it's important here but also diffusion coefficients, complication polymers I know some people work on biomaterials and it's really quite prominent in there indeed a lot of the wide line and solid state NMR methods came from looking at block copolymers. So that's a general outline of what it can do, what is it? Well it's simply a form of absorption spectroscopy so it's no different from optical spectroscopy except that you need a high magnetic field or a magnetic field at least to get the spectral lines. Only certain nuclei can be observed, natural abundance nuclei for example protons, N14 phosphorus all important in biology and I will tilt this towards biology always of course and the detectable nuclei all have a so called magnetic moment which means they are NMR active so you can't see everything but you can see some nuclei and that can be a big help or also it can be a big disadvantage. So the natural abundance nuclei that are important for biology the silence are for example carbon 12 natural abundance carbon and isotope and also option 16 but isotopes are detectable for example for protons we have deuterium which is going to be the focus here tritium which is very high sensitivity carbon 13 for example N15 is also quite important because N14 is quite different in its NMR characteristics from N15 and also O17 I won't mention much about that today because O17 is only visible really with solid state NMR it's a special methodology and we published certainly in that area as well so bear that in mind that you're only looking at specific nuclei and very often placing those specific nuclei gives us very direct and specific information as I mentioned the samples are placed in a strong magnetic field you can actually do zero magnetic field NMR but typically for the kind of applications in biology in life sciences this is anywhere from five to about 25 tesla and we're now getting up to a little bit higher than that now with magnet technology which is a major limitation in the field and then these samples are irradiated usually radio frequency in the megahertz range and by putting the sample in a superconducting magnet such as shown here in this cartoon here is a kind of spectrum it was set up here's the magnets and then this is where all the electronics are and we send in some radiation to the sample here and then the coil measures the absorption spectrum that comes out and the absorption spectrum here is a kind of way of looking at which of the nuclei you can observe for protons of high magnetic field and then you can tune your spectrometer to almost any other nuclei that's visible and this just shows the relative positions of some of the lines fluorine is a pretty good nucleus I mean I think 40 percent all drugs are fluorinated for example so fluorine is a very good methodology for looking by NMR at drug action there's phosphorus here of course in eukotides and phospholipids I won't talk much about that because we've got deuterium is the focus here lithium you've got carbon 13 you've got silicon and you've got deuterium down here this is going to be important because you can tune your animal machine to look at deuterium if you like but you can also tune your machine to look at protons but deuterium can help you but I'm going to show you how that happens so this is the kind of spectrum that you would tune your NMR machine to and it's within the proton lines here that you would get that information about proton NMR or carbon lines where you might get the information here so focusing in on this particular frequency range will give you the information you need by proton NMR so just bear in mind that NMR machine is simply a way of tuning to get absorption spectra at different wavelengths much like you tune a regular at home with different frequencies what can we learn well here are some basic parameters I won't dwell on these during the talk but I have certainly alluded to them chemical shift is very important this tells you information about the composition of atomic groups within a molecule it can tell you bond angles around peptide bonds which is fundamental to structure determination and very importantly something that's missed by almost every other methodology is the ionization state of the pk so NMR is very sensitive to the electronic configuration around the nucleus and you can plot out pka's of individual residues within a protein or a bimolecular complex very precise with single electron changes are uniquely uh sensors um detected by NMR we also have something called spin spin coupling so this is when two nuclei come together and the information about adjacent atoms is then resolved and that's what we use to pull out structure you know distance between two atoms then you can work out the structure relaxation time this is um how a nucleus responds to a pulse of radiation in the words how it returns to equilibrium gives us information about molecular dynamics and there is overlap now between the kinds of things Elko were talking about earlier today and molecular dynamics in terms of time scales and we should never forget the fact that biology is a highly dynamic system and NMR can probe the biologically relevant dynamic time scales required for function and signal intensity um there's a lot of complications about this but i won't go into them in detail but it gives quantitative information for example the ratios of the molecule of me helpful for determining molecular structure and proportions of different compounds in a mixture and this is in a biotechnology context absolutely vital in industry and a major activity this is a cartoon of a structure just simply ethanol and you can see the coupling that occurs in these lines you can now resolve exactly what is in your molecule and this is a proton spectrum so it's these hydrogens that we look at in this particular spectrum so all spectra are much like this or much more complicated and i will show you so so why would you use NMR if you can't get a crystal for example you have fibers amyloid fibers if you have a disordered system spider cells which we worked on for example or just polymers i know some people are working on polymers and so it's ideal to look at these systems you may have a heterogeneous mix multi-component systems a drug plus a target and so on you might want to work totally in solution and cellular NMR is now becoming viable you may want to vary conditions such as pH salt and so on and you may want to look at binding kinetics and energies to other molecules and the dynamic aspect of NMR helps you without a lot and you might want to understand stability folding and unfolding and most of protein folding work has come from NMR methods i won't dwell on that but deuteramide for the exchange is one way of doing that and you might want to measure fast dynamic processes NMR relaxation times can get you down to the nanosecond type time scale now you've been listening to diffraction and scattering methods already in this course and each data point in any of your diffraction patterns contains information about each atom within the asymmetric unit so that means diffraction methods scattering methods are long-range methods and that's important to bear in mind that different whereas in NMR each data point contains information only about one single atomic distance or angle process so NMR is a short-range method very good for short-range interactions typically over the one to ten angstrom or one to two to one nanometer time scale length scale so it's a short-range method that's quite important and people often don't distinguish between those two differences of the methodologies the size of a magnet matters how big the field is you get better spectra the higher the field the better resolution better the data as usual so technologically pushing the high field strength is really quite difficult we've now got to something like 1.2 tesla but field strength is is technically very very difficult to achieve just the engineering and the way of producing these superconducting magnets so that is important and then also the size of the biomolecule matters the larger the molecular mass usually the worst the spectral resolution and this actually is where deuterium comes in and helps so it's predominantly a solution state method in other words purified macromolecules insolvers with isotope isotopic substitution so if you mistake that it's typically what we do when you take a global view here is a protein in solution produces lots and lots of spectral lines with absorption spectrum they're all characteristic of their environment we use multiple methods and people like Kurt Bertrick for example got Nobel Prize for measuring the form resolving these multi-dimensional ways of taking these spectra and then reassessing them then from this we get to distance constraints and we get to structural determination and one thing that is quite different from rigid atom diffraction patterns in NMR is as you can see here you see conformational spaces being explored by structural elements within this protein in other words you see the dynamics you see that there's lots of motion within these kind of these proteins rather than the rigid atom position which you see in x-ray depraction which you then may see diffuse in electron density which is which is what NMR is telling you here and this is done in solution not in a crystal of course and you may get distance constraints this is for introducing one beta but you can also then get rms and then of course you can put that in the protein database in exactly the same way as you've had for depraction so short range distances bond angles and charged state dynamics and you need good spectral resolution membrane proteins though and I've been asked to focus on membrane proteins are large so they are fighting against all of the things I've been saying about getting nice high resolution because when you have lipids present as well these complexes are now mega dolgins they're huge and these large molecules and large complexes cause broadening due to slow tumbling and if you have slow tumbling in solution of these large complexes you get loss of resolution just to show you this is ubiquitin but much like the interleukin I showed you a moment ago there's 76 myos is tumbling freely in solution if I now compare that with a membrane transporter in a membrane which is only 100 kilodolms but in bilayers this is the kind of spectrum you get so this looks hopeless it looks as though there is no way that you can get any spectral resolution from this from a carbon-13 absorption spectrum of this particular system what I'm going to tell you is that there are methods and deuterium can help us get that and this is by using something called solid state NMR I said it's a solution state so everything is aqueous or in solution but solid state NMR looks at large macromolecular complexes still in solution they're not solid although you can use the same methods to look at solids block copolymers and so on so solid state is a way of extending the molecular weight range of NMR into a useful system and it's a system it's a methodology that's been around since about the mid-1990s in biology and people like stanopalabography and so on have been leaders in this particular field method development you need some real technical developments and it is a technical development and we've been involved in those two and there are two approaches one is to actually exploit these broad lines with wide line NMR so exploit the wide lines to try and get information out of these I'll show you those in a moment the other is to spin the biomelecular complex at high speed so all the magnetic interactions mainly from protons which cause this broadening I mentioned here there that is possible to average those out by spinning the sample and normally the spinning speed if you keep kilohertz and again technology people going faster and faster we average at the isotropic magnetic interactions with many of you will know about from diffraction scattering data but at the magic angle now why is the magic angle important well all of the interactions magnetic interactions of course have a spherically average dependence and the spherical averaging is given by three cos square theta minus one as always that's just geometrical but at feature equals 54.7 degrees the magic angle this reduces to zero so if you take a complex like this you spin it to the magic angle in the rotor in the NMR machine you can start to get high resolution light spectrum and that's the principle of magic angle spinning it's not that simple it's there's lots more going on in terms of pulse sequences methodology this it is a way of looking at large complexes to get high resolution type information so that's very schematic it's very basic but that's the principle of the methodology for those of you who are interested in the magic angle intuitively if you think of a tube of unit size so every axis is one and then you draw the diagonal from one corner to the opposite corner it makes an angle of 54.7 degrees with every axis so that means every position of coordinates x y z then are reduced to the same angle and so that is the magic angle explained in an intuitive and non rigorous way but it's a nice easy way to think about it what does that do to our NMR spectrum well here is a typical powder pattern NMR spectrum like I showed you before it's actually a phosphorus in the membrane if we now start to spin that sample in the membrane in the NMR machine what happens is the lines start to get narrower and narrower in fact the distance between these lines is exactly the spinning speed and until you're spinning faster than the spectral width of this it hurts you get a single line so that is exactly what happens for an individual particular system so this is very cartoon like but I hope it helps right so protons is the most sensitive NMR nucleus I showed you earlier but it's too abundant there are too many of them the spectrum can be really complex complicated and complex to analyze and flag information and this is a common feature of NMR there's too much data and we don't know how to analyze it completely the large biomolecules in particular they are then broadened due to the slow motion I was telling you within the magnetic field unless you use magic angle spinning so we need ways of simplifying spectrum or detecting other nuclei and that's where the tricks come in that's where the expertise comes in of using NMR successfully in in the life sciences so isotopic substitution I alluded to earlier can be used through expression to put in labels of specific sites of interest for example chemically you can introduce into a ligand or a protein as a ligand by expression into side chains for example or you can put it uniformly into a backbone of a protein and usually you you need multiple samples and you can add and subtract and convoluted different spectra to pull out the information that that is required and then multiple labelling techniques with different carbon compounds and sources can help you and common isotopic substitution to C13 to C12 and 15 to N14 or deuterium for protons uniformly in D2O or specifically if you want to label methyl groups and this is a very active area of labeling methyl groups with deuterium to simplify the proton spectrum from the methyl groups left behind and here's a way of doing that for example using D-glucose here we have labeled deuterium here on the alpha carbon and then you can label the side groups with deuterium as well so what you're doing by adding deuterium is diluting out protons and that's where the deuterium comes in and that's where it's powerful so if you do a proton NMR spectrum now you can only see these protons on the nitrogen and not the C-alpha protons and you can do the opposite experiment as well so that's one way of using deuterium in NMR deuterium is also used to suppress large solvent signals in proton NMR now if you're dissolving your protein in water the vast majority of your signal is going to be water protons and so one way is to dissolve in D2O it needs to be a very high level of D2O because even a small amount of H2O will give a signal and then here you can see that you can reduce this signal here which is predominantly water in the sample so you can start to get the narrow spectral lines without water being there and this is just for small molecule caffeine for example which has been dissolved in water technically it's becoming less of an issue these days NMR machines are now being fitted with ways through software of suppressing the water signal but it certainly has been a very valuable method of using D2O in the past it's getting less important but there is one other important aspect to this NMR machines can drift if you're looking at a biological sample you might need to take a spectrum over many hours or even days and so the field can drift the electronics can drift but the D2O the deuterium in the sample you add some D2O can produce a signal which the NMR machine can look at I said it's NMR active and you can look at this and produce it and use it as a so-called LOX signal so that means the NMR machine goes back to this LOX signal and fixes on that one unchanged signal and then adapts the software so that the spectra are all corrected for any instrumental drift for example in the magnetic field or the electronics you know if the room heats up then the electronics work differently and because the change in the positions of these lines are extremely minute they're very small indeed this is still a very important way of keeping the machine on field you can also use deuterium to increase spectral resolution these exchangeable protons I mentioned can be reduced and here is a particular system where amide and carboxylic acids here can change out the this is a simple antibiotic can then increase the spectral resolution as you can see and also in lines here for example this has got much better resolution than in water in this position here and so by exchanging out using deuterium it can be a major advantage and then also having used deuterium to help the machine and to help resolution of other nuclei or protons we can also exploit deuterium as a nucleus in its own right it has a magnetic moment spin one it is detectable and I won't go into the details but it's a so-called quadrupole nucleus it has two detectable spectral lines now you've been listening to oriented membranes you're listening to systems where you have lipids and so on and this orientational dependence of the two lines has been very very important indeed there's thousands of publications exploiting this particular feature of deuterium NMR so a typical deuterium NMR spectrum of a single rapidly rotating deuteron it could be a deuterated lipid in chloroform for example or it could be just D2O is an isotropic everything's averaged due to fast spinning rapidly typically faster than millisecond phase like a solvent and that's a single line because all of the anastrophic interactions that you have associated with this spin one are averaged if you have fast motion it's isotropic this is the deuterium NMR spectrum you get from wherever the deuterium is in your biomolecule but if you have a partially ordered system for example a liquid crystal or a bina membrane where you've been listening to such as this then you can get two spectral lines the deuterium is still undergoing fast motion but now it has an orientation with respect to the magnetic field north south in your magnet and that is really quite important and very very informative and I'll give you some examples of that so just to give you an example here this is from a nice earthquake in the early days from Michael Sattler he's an expert in this area proton proton magnetic interactions are very strong they give the broad spectrum to dipole a coupling so-called 100 kilohertz in energy and I'll come back to that so it results in broad spectral lines but partial deuteration as I mentioned really suppresses the spin diffusion between all these protons which is broadening the spectrum and so if you do that then you get the narrow lines so the signal to noise the magnet in the spectrum the remaining protons is significantly reduced now this is one way in which the tumbling small molecules gives narrow lines but large molecules give broad lines this is one way of extending the molecular weight range and it's good for large proteins so that's pushing the molecular weight limit as we call it for solution state and a mark by giving narrowness spectrum and here's just one of these so-called two-dimensional spectrum this is carbon 13 against protons I won't go into detail but I think you can qualitatively see here that the lines in this region are quite broad but here they've now been resolved into a number of individual lines and that's just due to this proton deuterium exchange in this system partial deuteration through expression has been used so I just need probably to allude to some more recent and very recent approaches just to acknowledge that people are still looking at these problems are trying to make the method available for much larger complexes and you can simply it's not simply but a lot of effort has gone into making different so-called pulse sequences a feature of NMR is that pulse sequences are always given these acronyms names and this is just a new one that's come out relatively recently and what they're doing here is changing the way in which they pulse method in other words the way the radiation is pulsed into the sample has been changed so that we don't need deuteration in the sample which is what this says here but one of the reasons is that so far there's no successful production of deuterated proteins from a mammalian cell line so if you want to produce a protein in a mammalian cell line bacterial works and a lot of people are doing that but in a mammalian cell line therefore that limits NMR applications of these kind of proteins which can't be deuterated so the aim is to record the signal of a central methyl group in this model system which is C13 which is black label here and by diluting out the proton signal is nearby which are making a broad spectrum and this also fast so-called methodology is sending radiation from local nuclei in a particular way to give a high-resolution spectrum without the need to dilute half the spins from the deuterium if anybody's interested the reference is there and it also increases the sensitivity of experiments as well and if you just want to have a qualitatively this is a kind of absolutely standard first experiment you might do by NMR and there's a line that's been looked at here is another type of pulse sequence here's another one and then here is the most recent one and I think you can see from here to here the resolution of this just one line that they have picked out gets much better with this new pulse sequence so just so that I'm paying tribute to people doing instrumental developments and technological developments without deuterium in difficult systems there are alternative approaches so this introduction just to summarize deuterium can be a nucleus in wide line MR to be exploited to give information secondly deuterium is a nucleus that can be used to improve the data from other nuclei so that's those two kind of bottom lines from this introduction so what I'm going to do now is to give you some examples all of you are familiar with membrane systems so this is focused on membrane proteins and membranes and also I'll give you some technological input at the end so the first one I'm going to talk about is lipid in bilam membrane we're all well aware you've just been hearing talks looking at lipids in bilam membranes what can deuterium NMR do for us the lipids in bilam membranes well let's go back to this slide I told you that deuterium has a spin one it's a quadrupole moment it has two spectral lines which are sensitive to orientation and often we want to know the orientation and also dynamics of particular groups of particular parts of the membrane it's in low abundance so enrichment is necessary and what we're going to do now here is our NMR machine we can tune it over all these nuclei and here is deuterium way down here and the further you go down to zero field in this particular spectrum the lower the sensitivity so deuterium is first of all low in natural abundance but also low in sensitivity so you might think it's not worth bothering with but indeed it is so the quadrupole nucleus means that unlike protons and spin one spin half nuclei which have a completely isotropic distribution of nuclear charge deuterium has a non-spherical positive charge distribution and so this is like going from a soccer ball to a rugby football soccer being quite popular at the moment of course everyone's watching the matches and they're not watching these lectures rugby is pretty important in in Britain and in France and some other countries as well so I always imagine this kind of nucleus like a like a rugby football so that means the electric field gradients across this molecule are asymmetric if you look at it along one axis it's got a larger interaction than across the other axis and that's what gives the anisotropy and the orientational dependence so the magnetic interaction applied field is now not isotropic but anisotropic you are all anisotropic you're tall you're anything from you know meter 15 two meters whatever and you're also narrow across your shoulders in your body so you are fully anisotropic in a system if I threw you at the back of an aeroplane spinning very quickly in all orientations you would become isotropic so spinning malecule motion in this case can give you an isotropic light spectrum and that's why deuterium in solution gives you a single line but in an orange system where it's restrictive it gives you anisotropy and this is really quite powerful so here is the deuterated liquid in oriented membranes exactly the kind of systems I was looking at in the 1980s by neutron diffraction and also by deuterium NMR which is very new in those days it was quite a new approach in methodology and we complemented the neutron and the NMR things are together so I've got a magnetic field in my magnet like any magnet it's got a direction along parallel to the normal it will give you two spectral lines and this is deuterated for example in the chain or for example in the head group if we now have a static field parallel to the membrane normal the membrane normal is here so I just turned my membrane from one orientation to another the lines have a different separation so in other words we can work out the orientation of the cd group in our liquid from this quite precisely given that kind of spectrum and there's a huge amount of literature on liquid crystals doing exactly that if we now smash our membrane up into so-called liposomes or powder pattern we get spherically averaged pattern now don't forget each of these lines will have a spherical distribution starting off here in the parallel orientation through to here in the perpendicular orientation and that is a powder pattern and that's typically what we would get from a liposomal system that's not tumbling quickly so that's large liposomes not small small liposomes will give you an isotropic isotropic spectrum if they're tumbling faster if they're smaller than about a hundred nanometers it's large liposomes large membranes natural membranes can give you that kind of spectrum if they are deuterated or even the very similar from phosphorus which is a spin half nucleus which is half of that pattern so this can be quite important to look at the morphology of the system we're looking at and just to show you how that works this is a glass plate and a bilayer and you can see that the lines crossover and that crossover point there for spectral lines is actually the magic angular game and then comes through to here so you can do an orientational dependence in your oriental system if you want so that is what's happening to the spectral lines but this system going to this system that is the trajectory that the spectrum would take what can we get from this well this is the NMR parameter it's this parameter here which is called the quadripole splitting for deuterium is the distance between the lines and the quadripole splitting which is typically in hertz or megahertz is very large compared to normal NMR spectra it can be up to about 200 kHz and some quadripole nuclei are in 8 megahertz wide option 17 for example so to do NMR you don't need high resolution for these systems in fact you you can use a machine that doesn't have very high spectral resolution because these lines are easily measured so the quadripole splitting is explicitly given by these constants here times this parameter which I've come across before it's three cos square d minus one so that's theta gives a measure of the amplitude of motion of our cd group which could be here at the head groups or in the chains the line shape of this spectrum and relaxation properties gives us information about dynamics so here for example there's a rigid aromatic residue here is a methyl group rotating around a c3 axis and here are phenyl groups flipping in the protein and all of these weaves observed and others in amino acids in proteins to give some information about specific residue orientation and also motion within a protein and that there's a whole literature on those but I won't go into more detail just to say that deuteration can give you quite detailed information both in solid biomaterials but also in biological systems so here we have dynamics and we have amplitude amplitude of motion of course gives you an order parameter and dynamics gives you the rate of motion and these two are quite important you can have order parameters with slow motion and fast motion and you can have dynamics slow motion fast motion so you can pull these out for deuterium quite explicitly so lipids in bilam membranes as an example order parameters of lipid chains now what we're doing is putting the deuterium at all the different places in the chain here is dppc dichromatophosphate and alcholene as we go down the chain and these are the lines from each of the powder spectra that I showed you a minute ago this is not oriented and here are the order parameters plotted out this was first done by spin labels and electron spin resonance um by us in fact but then also we did it and people like Joe say they did it for deuterated membranes and you can see that there's a lot of order in the upper half of membranes this is about where water penetrates within bilayers but then it drops off with very significant amplitude of motion and this is the order parameter amplitude of motion down to the center of the bilay and each of these groups is undergoing rapid motion because these are narrow spectra these two lines in the middle are from the methyl groups right at the center of the bilay and this happens to be for dmpc with cholesterol and without cholesterol so you can see that cholesterol increases the order which is exactly what's being shown by other spectroscopic methods as well so that's just pure bilayers this is dmpc this happens to be dpbc but it's the same exactly and this also reveals some interesting subtleties this is dopc many of you know diominal with a double bond here so this is unsaturated what does that do to it and what that does to it is it changes significantly the order parameter around the center of the bilay so I said this is where water penetrates and this was identified by Joe Staley as I said so here you now have the precise position of this cd bond and it's quite interesting that most saturated liposympiology are unsaturated or unsaturated around the center of the and these are the deuteration sites of this system so what about looking at how proteins can interact or intercalate into bilayers well here is namaloid polypeptide which has been looked at and they're looking at the penetration into the upper half of the membrane so here's our lipid here's our bilayers and this is the model they have and here what they're doing is this is the pure liquid bilayers this is the deuterium spectrum and as we add more and more peptide the order profile changes dramatically or more so in the upper half of the bilayers than it does in the lower half and this is the same we've done this the toxins for example a number of membrane penetrating peptides and typically these will penetrate to about halfway down the half of the bilayers and this was followed up by lots of other biophysical evidence as well so this is an interesting and sensitive way of looking at the intercalation of peptides and this is a relatively recent study here of lipids by deuterium enema into bilayers so that's a nice biological example well lots of molecules bind to the surface of membranes you know that too so pka of intercalating species and it's something you can measure i've told you enema is sensitive to charge but in this case this is the ionization state of the small local anaesthetic in this case that integrates into bilayers by looking at the change in the structural parameters in the membrane so here we are this is a phospholipid it's just a choline deuteration is at these two positions in the choline head group and they are going to be reporting on us to us on the ionization state of this coternar ammonium here in the tetra cane and here we go from the pure lipid to adding tetra cane and you can see there are changes here and those changes are reflecting the local structural perturbations that are occurring in the membrane surface due to ionization of tetra cane so here is just the lipid over a pH range from 3 to 11 now we add tetra cane and you're pulling out rather precisely the pka of tetra cane in the bilayer and this is a notoriously difficult as partition coefficient and notoriously difficult to measure for small molecules into lipid bilayers here you are looking specifically at the bilayer perturbation by the ionization state here a non-perturbing way of determining surface titration well a lot of lipids in nature of course are anionic there's no cationic but anionic lipids cytochrome c binding against something we and many others have been looking at what we find is cytochrome c is not a rigid crystalline molecule when it binds to membranes it undergoes very significant dynamics that's another story by looking at phosphorus relaxation but if we want to look at the the binding here we have phosphorus in our glycerol here we have the head groups of the lipids which we can resolve this is deuterated in the glycerol head group and now we can look at the fast motions the relaxation properties of our deuterium on binding of cytochrome c not much happens but the slow motions here we're in the millisecond time range millisecond to microsecond time range we're seeing that there are perturbed by the binding of cytochrome c so this is a very subtle identification of dynamics that are going on with membrane surfaces as cytochrome c which is known to bind to these violets bind and I just saw a talk earlier this week from a meeting with how is it in Portugal and this kind of thing is is actually explicit in the way in which cytochrome c is an electron okay so that's lipid in violets there's many many hundreds of papers in the literature on digital what about ligands and prosthetic groups in that well I think many of you know this system these are retinal rod cells at the back of your eye here are the discs that are in there and the redopsin of course is the photoreceptor protein in mammalian eyes it's what gives you essentially red eye in the back of your eye from in a photograph and it's this chromophore of vitamin a derivative retinaldehyde which is now in the redopsin which is a gpcr and many of you know about all this of course and the nice thing here is that the membrane is pretty high density of this it's about 60 or 70 lipids per one redopsin and redopsin is the the predominant molecule in these disc membranes and it's possible now to uh this is a an early electron diffraction pattern there is now crystal structures for redopsin um from Chris Balchowski but then two or three others now in the database and the most important thing to notice is that the resolution in these kind of electron diffraction patterns is nowhere near sufficient to observe the retinal within the redopsin indeed the crystal structures very often the retinal is modeled back into those crystal structures using solid state NMR data from the retinal itself because the resolution in the diffraction is not higher and in fact this is probably the most important part of the molecule but you can't see it so NMR has been able to to to resolve the confirmation precisely so um the most important part is the chromophore not well resolved to the grand state or activated states nice thing about NMR is that you can put a membrane into the NMR machine you can look at it in the dark then you can shine light on it and you can do an experiment in the light so whereas crystallography you need lots of different crystals at different forms and we're doing a lot of that kind of work as well so just to kind of show you an example here is the bond vectors here we deuterated specifically the retinal at three positions along the the polyene chain here we've also deuterated other positions but these are the important ones and this is the confirmation is the bond vectors that we can measure by the oriented membranes by deuterium NMR in the dark and then in the light and so now we can see um how the chromophore changes during the isomerization that occurs when you put light onto this protein and we can feed that back into the protein crystal structure a couple of interesting things the methyl groups here are undergoing very fast rotation in this system and that we can measure by NMR and also we have a 6s transferring so in other words we've got information about the betaion over here and that of course came out of these publications something else that modeling a crystal structure retinal into a crystal structure the protein will not tell you but NMR does tell you is that the polyene chain is curved and this was interesting to see and now that fits the electron density profiles of high resolution structures so the retinal in a crystal is is is extended but in the protein it's actually curved and that's important for the activation of the redoxin um in the system so you know you get new surprises by using NMR approaches which are relatively non-perturbing non-crystallizing and in natural membranes and all the natural lipids in the system so let me give you another ligand example many of you will know this is an electric fish it's torpedo monorata it's a good model for the astral coating receptivities of high density in the neural membranes of this here is some low-resolution electron microscopy for nice lung wind but the activating one of the most abundant neurotransmitters in our brain is acetylcholine and here is acetylcholine it binds to two of the monomers of this pentamus somewhere around here we now have a high resolution crystal structure of the extra membranous parts and what we've done is we've deuterated it in the paternary ammonium group of the acetylcholine and looked at it binding to the protein in these kind of semi-purified membranes with all the natural lipids and all the other natural proteins the very specific binding of the acetylcholine and we've used deuteramin NMR to look at the orientation in particular of the binding site this is quite important because this was the electron density resolved by large lung women as colleagues and it was not clear where the binding site was or what orientation it was within the protein and the deuteramin NMR spectra gave orientational dependence and the line shape was sensitive to molecular motion so we could use those spectra and if you look here in oriented membranes of acetylcholine is bromelated in this case to fix it in the site but that's a technicality we can see the spectrum here the membranes which are oriented with the magnetic field parallel to the membrane normal compared to over here on the right perpendicular to the membrane normal the spectral shape changes very dramatically indeed these are computer simulations of the spectrum and you can each one of these orientations we could measure menu from 0 to 90 is a unique piece of information and then this is plus five degrees minus five degrees so we can look at the error on this on the um the precision of this measurement and what we pulled out was the acetylcholine binding site is at 42 degrees with respect to the membrane normal and these methyl groups that we have deuterated are undergoing very fast rotational motion within the binding site and that again has been used computationally related by people for um tertiary my umerium drugs which compete for acetylcholine in a medicinal context and now that's a big area but this was quite important by some of those studies so um sorry I'm going backwards here so here's a final example this is membrane bound ATPases those who have breakfast this morning this protein is pumping proteins of protons into your your guts the gastric ATPase and then the sodium ATPase is an analogous protein we now have some high-resolution structures in fact in Denmark which were not available when we did this this work which was together with drug companies and what they wanted to know is whether inhibitors of these ATPases this is one particular one from Roche and this is Wellbane which is a poison which will kill you where do they bind in this protein how do they bind and uh crystal structures now show that in fact the determinant of our information was correct so let me just show you an example with kidney ATPase here we are looking at the Wellbane which is a cardiac glycoside so the ramnose sugar is here and then the sterile nucleus is here and what we've done is we have put in NMR nuclei on here and worked out the conformation of that molecule in the protein so that was a confirmation we got that and then crystallography subsequently some years later came along but what they could not resolve in the crystallography was the ramnose group here there's no electron density associated with that sterile parts here and we used deuterium to try and resolve that question what the problem was the way we did that was to again come back to the Wellbane the small molecule we labeled with deuterium and also carbon-13 fluorine as I said the conformation but we labeled with deuterium here and here so this is going to be telling us about the motion of the steroid nucleus this is going to be telling us about the ramnose now you might think well why is that important well it turns out that the spectrum from the steroid nucleus is this broad simulated energy broad pattern which told us that the steroid part is fixed down in the protein this was all done before crystallography and then the ramnose when we put the deuterium here gives this isotropic spectrum which I showed you earlier comes from a very freely mobile deuterium so we have differential mobility of Wellbane when bound to a kidney ATP so crystallography is not going to help you here you can see electron density about the steroid but not about the ramnose NMR is telling you why differential mobility is really very very important for drug action you make this molecule and you restrict that chemically between the steroid and the ramnose in other words you make it rigid the thing doesn't work it is just totally ineffective that differential mobility is vital for function wouldn't advise looking at Wellbane without very serious health and safety requirement because it is a lethal poison as I said earlier but this work was was quite instrumental in starting a whole new range of differential mobility drug design and that's something else that came from deuterium okay so that's this bit what about now membrane protein structure to finish off I'm now going to tell you about some very recent work because the technology deuteration facilities the way in which we can do deuteration is only now becoming really realized and here is a timeline if you like by increasing spectral resolution using exchangeable protons as I told you and reduce spectral broadening as I told you earlier both in protons and C13 and it's these large 100 kilohertz I mentioned that intentionally else comes in the moment causes spectral broadening so what possible solution here replace protons by deuterons in membrane proteins can this be done one yes this is proton content okay so this is 100% protons 100% deuterons we only ever get to about 10% deuterium here and so that's a kind of a timeline a line of improvements and that usually is with time how we're doing better and better in deuterating membrane proteins and along this rate are the kind of resolution you're getting for spectra and I think if you look at some of these here the resolution we're getting now this is one of our examples is getting really rather good for spectra here for bacteridopsis happens in a model system but there are others that have been used on pecs for example and beta barrel protein here so as deuteration is improved spectra have improved but also this magic angle spinning technology I mentioned to you is getting better so here's 10 to 15 kilohertz this is going up to 40 to 60 and this was the first example 60 kilohertz we're now up to 100 kilohertz magic angle spinning so now the resolution is really getting rather good and it's becoming comparable to what we have for solution soluble protein and this is on pecs in a membrane with lipids that's the important thing membrane proteins like their membrane lipids in fact they need to have a rank of the fully functional competence so this is the way the field has been going over the last 10 years or so and it's particularly encouraging and I'm quoting here from this article I showed you here that's our paper that's the review article on it now we're now getting up to 300 residues and here are some of the systems you will recognize immediately the menu don't have alpha helices or beta barrels structure elements which are relatively rigid within the membrane and that's what we're looking at here and here is an example it's not paid crystallography shows us a crystal structure has been done for some while but what it tells us in a membrane an NMR tells us and deuterium helps us it's not a rigid protein membrane proteins are not rigid in membranes they have lots of flexibility and that explains of course why loops have poorly resolved in most diffraction patterns of membrane protein but what it tells us is these beta barrels have a rocking motion here they are and they they they can rock within the membrane they're rocking around and the NMR is giving us an indication of that from the order parameters from the different labeled residues within the protein we have flexible loops we know that we've identified loops crystallography distorts roots we've shown that we've done that by NMR by labeling their loops and then comparing it to the highest resolution crystal structures and show that crystal packing does distort loops in membrane protein so you really need to do this in a membrane without crystal packing and then protons detected in solid state NMR 60 megahertz that's what this one was gigahertz spectrometers so going up in field as well and they are becoming available 1.2 may well go to these places that can afford them like E.J. Zurich for example and then this one was reconstituted back into lipids I think it was probably a DMPC or a DOPC a GPCR has been resolved in nanodisks by Stanopoulos group and here are some of the NMR spectra you can see that they did structural resolution what I've done here is compared his structure by NMR with an X-ray crystal structure this side kind of sector and here are the backbones and you can see that in green we've got the crystal structure from crystallography using lipid phase crystallography and then from sorry XRD crystallography is in pink and NMR is green so the crystallography was done in lipid phase this was done in membrane bilanes in biceps impact and you can see there are subtle differences and those differences almost certainly are due to the lipid environment giving a different confirmation of flexibility and also structure compared to a crystal so where would you find these structures well there are about 240 membrane protein structures in the PDB from NMR and though I haven't checked everyone but it wouldn't surprise me if all are deuterated and they are usually in these kind of membrane memetics going from these smokes which we've worked on to use extensively but also nanodisks and biceps and even bilanes so deuterium is vital for this kind of area and here are two particular web pages I've given me but it's here from Drogfarszowski and Steve White and others showing all of those and their conditions and the resolution and the lipid variant because the lipid is important by these methods just very recently a couple of very nice primers have come out on how to do a solid-state NMR right the way from experimental flow design protein production sample how it gets into the NMR rota into the magic angles spinning for example sample packing and evenly given you where to get and which kind of reagents you need for deuteration to grow proteins in bacterial systems of course ready for deuterium NMR and also of course I shouldn't I need to mention the deuteration of our tree in ILL not least because they have a very nice web page of lots of protocols, the lipids and so on we've given some protocols to that page and there's a lot of deuteration expertise in the world around and people like Michael Hartlein of course drove to site whoever he knows involved in this deuteration lab which was one of the first set up and now there are many around and it's not trivial for membrane proteins and they have some fantastic expertise and Juliette in particular for protein production so there's a lots of results out there that people need and just to finish there is a conference coming up biophysics neutrons NMR will be there it's a hybrid meeting it's fully flexible council and so on and I will be there I hope in July if the Austrians allow Britain people from the UK to go and then I will be there and thank you very much Tony thank you so thank you very much Tony for this fantastic overview of all the exciting thing you can obtain with NMR on membrane protein and lipids so really appreciate it so is there some questions well it's a lot of clapping of hands is there a question for somebody yeah can I yeah thank you very much for your presentation from my previous experience it's very difficult to infer the order parameter from per deuterated molecules because these splitting sometimes they overlap and it's difficult to assign to a special CD bond in the molecule and I'm wondering if there are any good now the convolution like softwares or it is still better to do partial deuteration to see exactly which yeah yes thank you very much that's the important question and certainly in the early days this is one that we tried to address so you're right partial and specific deuteration is a way forward but there are depaking programs which you can use to computer simulate your spectrum and then try to optimize the fitting of the spectrum to the depaking program and that is the best way forward but you are right there could be overlap especially between carbon 2 and carbon 8 where all the order parameters are very similar in the upper half of the bilayer trying to decongolute that it is quite difficult as you get further down the chain to the center of the bilayer then it's easier because the quadruple splitting is more separated but you're absolutely right the other difficulty with this kind of system and people have made mistakes here is that if you have slow molecular motion then you're not fully averaging the quadruple around esotropy and you therefore get broad lines and so a way of improving that is to increase the temperature if you can in your system so I always have a bit of an issue with order parameters for gel phase bilayer for example I think you have to know what you're doing and you have to be very specific about dynamic range or you're in biology is totally fluid as we all call so but you're right it is an important problem there are depaking programmers out there and the experts still in this field people like Michael Brown for example in Tucson he's the personal approach for example okay I hope that helps yeah yeah thank you it's still very difficult to do this yeah okay well he does a question yes thank you so much it was a very interesting presentation I have a question I'm here I'm a study interaction between for example boy salt and the bilayer phospholipids if I prepare for example a liposome from deuterated partial or fully deuterated the liposome and then add a boy salt to them my question is that can I from this NMR technique can I get information especially about distribution of the boy salt inside the liposome for example is it can I see any difference if there is a difference between outer and inner leaflets can I get those kind of information from NMR point of view oh so now you're asking something about membrane asymmetry and as you know that is a pretty tough thing to try and determine that can you make asymmetric these things I assume that I assume that they it's distributed randomly but I assume that the boy salt it goes mostly to the outer leaflet and I just want to somehow prove that that if it's this hypothesis is true yeah so what you're asking is if you can distinguish the outside bilayer from the inner bilayer of a liposome now we have been making and London has been making asymmetric lipid bilayer and if you can make an asymmetric lipid bilayer then you could put your deuterium in the outside and do a measurement put your deuterium on the inside deuterating the bits and then do a measurement so that you could do if you predetermine the physical asymmetry of the liposome to begin with and it's quite tough to do but these asymmetric bilayers certainly in biology all membranes are asymmetric but to achieve it technically has been quite difficult but you you you can start to do that and maybe you can try and do that and you should be able to do that with liposomes now whether you can do it with straightforward symmetric liposomes and distinguish both leaflets that that is something that that is also of interest I guess there have been ways of using lanthanide shifts and phosphorus NMR now that you can do if you can look at the phosphorus signal from the outside leaflet and perturb that with a lanthanide that does not penetrate the bilayer that was done many years ago by BSTRO then you can distinguish what's happening in the outer lipid compared to the inner lipid so if you use something else to help you distinguish the outer lipid but maybe physically making asymmetric bilayers would be the most convincing way forward okay so biosomes you will certainly see the perturbation of the surface of the bilayer you will be able to see whether they penetrate the upper half you'll be able to see the packing you'll be able to see titration behavior because these are presumably ionizable compounds so yes you can see all of that for sure and there's a good literature on that thank you pleasure any more questions so I'm sure professor what's we'll be happy to receive emails and with questions and so on even after when you have studied the slides that are revisited lecture so so with this I would really like to thank Tony for doing such a wonderful lecture and pointing out the differences and complementarity between scattering and NMR I mean you can get a lot of additional information by combining the two techniques yeah and in many of these systems that are deuterated for NMR you can apply deuterium and neutron experiments too as we have done many times so I didn't ask me to talk about that but that's a whole new lecture on how to compare deuterium NMR and deuterium in neutron experiments but that that is certainly a very viable way forward and you only have to make the deuterium labeled once and you can use it in two big facilities so you can get a lot back oh it's been a pleasure and thank you very much for asking me I'm really sorry I can't be in Sweden I love Sweden I love Scandinavia clearly I have some routes somewhere in Scandinavia in my past generations and I wish you luck with the sweetness course great and good to see you Tommy and everybody yeah