 But thank you very much for the organizers for inviting me and I really enjoyed yesterday's program and this digital format that worked surprisingly well. I was particularly happy to hear that this was the Lund Institute of Advanced Neutron and X-ray Sciences and particularly the neutron that allowed me to pull up some work that I did during my PhD and that was on a beta-barrel protein and after using X-rays to get a high resolution structure we used neutron scattering at the ILL in Grenoble together with contrast variation to determine the structure of the detergent in these crystals and that is shown here on this in this picture and it shows this intricate three-dimensional network of the micelles in these crystals and that illustrates clearly one of the of the challenges that we have in membrane proteins where you have your interface between the hydrophobic and the hydrophilic parts of the protein and that's in many techniques can play really tricks on us. But then again that interfaces is where most of the excitement happens. So I'm working as Karis said at AstraZeneca in a group that basically does the protein reagent protein science generation for early drug discovery in AstraZeneca. We have a daughter group in the UK that is mainly supporting the oncology areas and we from Gothenburg are supporting cardiovascular and renal and metabolic diseases and respiratory and inflammation diseases mainly and protein science is really key to many early drug discovery projects and it's important to deliver the right quality at the right time and many projects in the beginning when they don't have proteins simply they can't progress their projects and we need to be fast, innovative, agile and productive and clearly this is a managerial sentence put in there and that basically needs you need to work harder, smarter and deliver more in less time with less people and that is what we aim to do. Now the proteins that we generate and that's not only membrane proteins obviously but many different types of proteins they are used in biochemical essays, they're used for affinity screening, we supply to biochemical high throughput screening projects, biophysics and mode of action studies when we've progressed a bit longer in projects. Structure is a very important component of what we deliver proteins to an antibody generation, bioanalyst analysis and also we supply proteins for example in vivo PK studies and each of these applications put different demands on the proteins that we deliver and if we then look at specifically membrane proteins at AstraZeneca we've had a long-standing interest in membrane proteins clearly because membrane proteins are a very important and fruitful class of drug targets. Already in 2005 we established a structural system for the MPGS the human MPGS recombinantly expressed in insect cells and possibly different from academia. We then continued not only with one structure but solved over 100 structures of MPGS in complex with different compounds and in this way driving the chemistry forward. So it's clearly the first structure is of most important but then the iterative structure support to guide the chemistry is something that we focus a lot on. Now we've also looked at some methodology development and FSEC was mentioned already a few times. We further developed that using a his specific that we then use and we don't we are not dependent on GFP anymore. We're looking at single molecule microscopy as an essay methodology particularly for membrane proteins. My director Nick Decker has had a collaboration with Rosalind around these smalls and more lately we've worked with DNA encoded library screening and did some methodology development around protein purification and the latter is what I'll present in the first part of my presentation and in the second part of the presentation we'll be looking at a specific example the part two receptor that we work together with have tires on as well as touch upon the DNA encoded library screening. So if we start with the first part it's methodology development that we've done on membrane protein purification and then we need to look at the T-beck and the T-beck that was invented by two ladies in the US in 903. A picture from the patent is shown here where they have a cotton mess in a metal frame in which you can put your tea bags and that you can put in your hot water making tea that never really caught on but then a few years later a salesman he accidentally reinvented the T-beck as he sent samples of tea to customers soon in small sealed bags and the customers used that to make tea and then when he re-sent larger packages of this tea he caught complaints that why didn't he supply them in these nice tea bags. Now when we thought about protein purifications and ways of making that more efficient at a certain moment we thought why can't we put the resin in tea bags and then use that as a purification method. So that is what we actually developed and here you see a picture where we purified a Histec GFP in a tea bag so the resin is enclosed in a tea bag and it's nicely colored and on UV you see the fluorescence. Now actually this was not only a reinvention of the tea bag but it turned out that it was actually a reinvention of a method that Porat and Sundberg already in 1971 published. It was a group in Uppsala and they used also the concept of a tea bag for purifying proteins. It never caught on really and was forgotten and we had to re-find it ourselves but we believe that due to development both in resin but also in expression and over expression the tea bag method can be a very powerful way of purification. Now we explored different materials and in the end we ended up with a P-tex, CFAR P-tex material which is a mesh and you can get it in different mesh sizes and we ended up predominantly with the 40 micrometer mesh size which gives an optimum between the resin particle size and fluid exchange. We explored different shapes and more exciting ones like pyramids, tetrahedrals and cylinders but in the end we ended up with squares or rectangles as most useful and the method can be applied in many labs. It's low tech. Basically you fold the mesh then you heat seal on two sides. You can fill it with a resin of choice and then you heat seal or you clamp down the last side so that your resin is enclosed in the mesh. Then you just take the tea bag and you throw it into your solution with the protein of interest. Incubate it, shake it and then you can fish out your tea bag, wash it and elute. Very, very convenient, simple and fast. Now here's an example. We initially developed this for secreted proteins. Here's an example of human embryonic kidney cell culture. On the right hand side, conventional purification. Here you see it using the cleared media and in the first example on the left hand side it's actually the fermentation brought with the cells and the tea bag so you can use it immediately in your culture with cells. Now what do we save? We save quite a lot of time so you don't have to spin down and get your cells from your growth. You just add it at the end of your growth a few hours depending on your binding kinetics and then you pull it out the bag and you basically have your purified protein. So you can save a lot of time compared to conventional purifications. When we applied that in projects what we saw is that in some cases particularly for sensitive proteins we saw improvements in the quality and actually projects which protein targets in projects that completely rely on tea bag purification to get functional protein and that prompted us to think about application to membrane proteins as well. And this is a collaboration with a group of Pontus Gurdon as a lunch university and in the University of Copenhagen with a student of Pontus Julie and together with Jenny Heering a student at AstraZeneca supervised at Gothenburg University. And membrane proteins and it's known to everybody if you have them expressed in the cells what you typically do is you break the cells and you isolate the membranes and possibly apply a few washes then you go to solubilization you centrifuge away your insoluble material you apply it to purification either a column or in batch and eventually you would leave it to protein. Now we believe that the tea bag can be applied and that many of these steps to speed up the process and what you can do is you can do whole cell solubilization and in this whole cell solubilization add your tea bag incubate wash and purify you can apply it also at the membrane protein stage sort of the membrane stage where you solubilize the membranes or you can also apply it at the stage where you have your cleared solubilization and this all delivers quite some time game. We applied that to a range of different proteins from different protein classes and also looked at different expression hosts so the CLC1 and the acroporin are expressed in yeast the part 2 and the KCC2 are expressed in insect cells and MRAY is a bacterial protein is expressed in bacteria and what we what we observed is that in most of the cases you get quite comparable purification in some cases you see that the purity may be slightly slightly less for KCC2 and CLC1 and interestingly for acroporin 10 it looks like tea bag purification even gives an added benefit on the purity of the of the protein but surprisingly and to our pleasant surprise in many cases we see slightly improved size exclusion profiles and particularly for KCC2 we see a much sharper peak here compared to the conventional purification and less of this larger form and we know from SPR that this larger form is is ligand binding incompetent whereas this the smaller and the sharper peak is ligand binding incompetent. So what we what we feel is we we definitely have a advantage with respect to the time that purification takes but also the hands-on time that allows us then to do other things in between we believe that the quality is is equal and in some cases exceeds actually conventional methods and it's very scalable you can basically just add more tea bags and and also you can very simply parallelize this so if you have multiple multiple preparations you can just do the multiple preparations and and add tea bags to all these in different bottles what we did see was somewhat reduced yields but then again if we go to cryoman cryoem who cares about yields and to illustrate that this is not only that something that we that we do for a publication it is actually something that we apply in in live projects and recently I've applied it to a an orphan gpsr where tea bag purification gave a very nice pure sample and I realize this is just a band on a gel but it's also accompanied by a nice size exclusion profile and with that I want to switch a bit gears and look at a specific project that we've been working on and that is a protease activated receptor protease activated receptors there's class a gpcr there are four family members and they have a extended n-terminus which is sensitive to proteases and once these proteases have cleaved at the specific site in the n-terminus the n-terminus actually auto activates the receptor and leads to downstream g-protein signaling as well as beta resting mediated signaling now it's these receptors can also be activated by externally added peptides so the the same sequence this SLIGKV peptide if you add that even in the absence of the the cleaved n-terminus you also get g-protein mediated signaling and beta resting signaling the picture is is is more complicated as I I paint it here because there are other proteases that can actually deactivate uh also the n-terminus there there are also proteases that can cleave at the different sites and the actual cleavage without the n-terminus also induce some signaling but the main mechanism is via this n-terminal cleavage and then auto activation now these Part IIs are interesting because they're related to many inflammatory processes and hence they are involved in many different diseases and so in 2001 AstraZeneca started a collaboration with Heptires company and Heptires has this star technology which is originating and in pioneering Chris Lab and so in that process through systematic point mutations and carefully analysis of thermal stabilization Heptires can generate stabilized receptors and it's shown here in the graph on the on the left left bottom here you see in blue the wild type it's uh poorly very poorly expressed and has a low thermal stability and by introducing five point mutations through a careful recombination of individual point mutants you see an increase in expression level and also an increase in instability and finally the construct that was used for crystallization that included 10 point mutations you see a very good expression and a very high thermal stability now on top of these point mutations there was also further engineering required so introduction of first of all cleavage of the n-terminal domain and or truncation of cleavage and introduction of fusion proteins the four lysosim and the n-terminus as well as the strategy of replacing the intracellular loop in this case with BRIL as well as truncation of the C-terminus so Chris has a talk titled whipping GPCRs into shape and this is clearly one of the examples where considerable whipping was required to get a well or decently behaving protein that could be used in in various essays and applications to illustrate how we use I will introduce a molecule this is the AZ8838 and this molecule that originated from a high throughput screen a cellular high throughput screen based on elevation of intracellular calcium in a flipper assay and from the ACS there were well it was moderately successful I would describe it but there was a molecule in this series a single ton which means that there was not multiple proteins that had the same core or scaffold and it was purely inhibiting and in the absence of biochemical reagents to probe target engagement or to confirm target engagement and to look at structures these single tones and particularly if then there is a poor or a very steep sorrow are for chemists really difficult to to develop but if you can confirm target engagement and you can look at target binding and eventually also at structures then suddenly there is a lot more scope for developing compounds and for evolving compounds and that is where this biochemical stable and well behaving part two protein comes in so what we could do is we could sorry what was interesting with this compound is that depending on the essay how the essay was set up we could see shifts in the potency so with a very short incubation or with co-application this molecule in the cellular assay showed a very modest potency if you pre-incubated the cells with this compound for 15 30 60 minutes what you see is that your potency apparent potency improves so the compound becomes more potent now what could that be and how could that where could that originate from and for that we looked at a biocore SPR essay where the stabilized protein was immobilized on on the chip and small molecules are then flowed over and the mass change is measured and from that what we observed was that this specific compound has a low on-rate and a very low off-rate so the residence time of the compounds is in the order of two hours and particularly considering the modest affinity of this compound that is surprisingly surprisingly long now how how can that be explained and for that we we looked at a structure that was solved for this for this molecule and we used lapidic cubic phase crystallization in the presence of this compound so copure vacation and then crystallization and could solve the structure to 2.8 angstroms and what you can see is that the the molecule binds quite deeply inside the gpcr and it's it's basically fully excluded from the solvent it it binds under this ecl2 and really nicely fits into a small pocket which is formed in the in the middle of well almost in the middle of this part 2 receptor we confirmed this binding pocket by point mutations and what you also see is that it really is quite nicely fitting there inside this pocket and there's little room to to expand so this this binding mode suggests clearly conformational changes and locking of this molecule inside and underneath the ecl2 probably explains why the both the on-rate and the off-rate are are slow now it also highlights that further developing of this compound or extending the compound with additional groups wasn't successful simply because the pocket is is not large enough to to allow additional chemical groups to be to be present so the the the star the structure activity relationship as chemists use it was very very steep so there's there's modest opportunity for further optimizing this compound and as there's there was there was little additional opportunities for optimizing we also looked at a different heat finding strategy and again there is the biochemical reagent the purified stabilized receptor was really key and that is DNA encoded library screening and that's a methodology that we have last year internalized we have collaborated with xcam company specialized in this and we've licensed in and and internalized their methodology this is based on libraries which are synthesized on a piece of of DNA a barcode of DNA and by smart combining and doing chemistry splitting recombining you can get huge libraries in the order of billions of compounds of molecules in a single tube and each of these molecules they can be identified through their DNA barcode and so if you compare our high throughput screening compound libraries in the order of two million compounds here here we are we have a number of libraries and in total I think we have in the order of 10 billion compounds in these DNA encoded libraries so you have really a huge opportunity of screening chemical space now what you then do is you you combine these libraries with your protein of interest and what will happen is that some of these molecules that that have affinity for your target that will bind to the protein you then immobilize your protein you wash away everything that is is non-bound and you repeat that process a few times and basically then by a heat elution step you elute finally all the compounds that were bind to the protein of interest and then you apply your next generation sequencing to look at what is the DNA sequence of all these tags all these barcodes and from that you can identify which compounds had high affinity for your protein of interest then the subsequent step is to make a selection of those compounds do the resynthesis but then off DNA and then confirm with a functional assay that actually the the binders that you identify in this process actually also have a functional effect on your on your protein and we apply that for for part two as well using the stabilized reagents and there we identified a series Benzimidazole named after the group here in the middle Benzin and Amidazole and what you can see is that that on SPR these molecules were quite potent and also in a number of the cellular assays these were quite potent interestingly in the assay where the cell-based assay where the part two is activated by the trypsin and not by its peptide we see a a drop-off in potency and that that highlights really the need for using your most relevant assay to drive your chemistry forward. Now also by SPR the binding kinetics they are they are much more in in line with expectation so fast on fast off and the residence time in the order of of minutes which is expected for this this KD and what we then also looked at so what happens if we have our 8-8-3-8 compound as well and that hardly affected the binding of this this this molecule so that indicates that these molecules they bind at different pockets and and binding of one doesn't exclude the binding of the other compound. Now also here we we solved the structure this structure was to somewhat less resolution but sufficient to to model the molecule in the in the density and what we saw is that this molecule surprisingly binds actually on the outside of the receptor between here this is 2, 3 and 4 and and there is some sort of a of a pocket formed and again by mutagenesis we could confirm that this pocket indeed is the the the binding pocket for this molecule and interestingly then looking in more detail at the trypsin activation assay so for the 8-8-3-8 it it's not overly potent but it's nice monofasic for the 3451 compound what what we saw is is a biphasic profile which is obviously highly intriguing and this biphasic profile was only present when the protein was activated by trypsin and not by by the exogenous peptide and then looking in detail at some of these point mutations so for a mutated phenylalanine or this 157 residue the interesting thing is that this high affinity or site was lost but the low affinity inhibition transition was still present so from that data it suggests that actually this this molecule has multiple binding sites on the on the protein this binding site being the high affinity binding sites and the highly potent binding site and somehow there must be an additional binding site with a lower potency which presumably is only present in the part 2 form which is activated by the trypsin and where the n-terminus auto activates we don't fully understand it but the data suggests that there is an additional binding site and with that you obviously want to know whether actually these molecules do what they what they are supposed to do and for that we used a red model where you look at pore swelling and so these the reds are injected in the pores with protease for the activation of the of the receptor or or the exogenous peptide and that triggers the swelling that can be shown here in the black lines so after after 30 minutes you have 100 swelling and then after a few hours it sort of subsides again and then what you do is you pre-treat these animals subcutaneous for this molecule or oral administration for the 8838 molecule and then you activate the receptor or you aim to activate the receptor with trypsin or with the peptide and what you then look at it's swelling and here you can see that for the 8838 you get approximately a 40% reduction in the swelling for the 3451 there is about a 60% reduction in in the swelling and that is not only the swelling but also the histological analysis show that there is less inflammatory signals in the tissue of the of the pores so these molecules they actually are doing what's what they're supposed to do which is of course very promising now what I hope to have given you an idea on how these purified membrane proteins how they can be used and what the value is in our direct discovery processes and and clearly DNA encoded library screening and and structure and wouldn't have been possible with this high quality reagents having said that it is still a challenge and for part two the number of different compound structures that we could determine was relatively relatively low and cost quite some some efforts but these structures really do give a lot of input to to the chemistry and understanding of the mode of action so then looking forward so what what do we experience and what do we expect in in the near future so we see an increase in complexity of the of the targets and from genomics initiative for example we get more targets where there is less known which which puts additional challenges on us we definitely see a a lot of demand and opportunities for further application of DNA encoded library screening and again also if you think about novel targets where there is little precedence there are no tool compounds then DNA encoded library screening might also be a very rapid tool to find your initial tools but also as a hit finding method and particularly combining different hit finding strategies like ACS and Dell together with structure based drug discovery you can hybridize series and that is a very fruitful way of of going forward and we've been touching up that on that already yesterday clearly the cryo em resolution is something that we we see happening and we are applying that as well and here are 2d class averages of a gpcr that that we're working on and I just find it amazing that you actually can see here your g protein and your some indication of helices in your in your my cell and that is we see that particularly for the larger membrane proteins that is really the way to to go and crystallization will remain important for many projects but cryo em will just grow and I'm convinced that it will deliver more more quickly and richer source of information and there I think the chemists and well basically everybody we also need to develop our understanding of how can we use for example the dynamic information that we get from cryo em structures for the design of novel chemistry so I think it's a very exciting time for the field of membrane proteins and clearly the work that I presented is based on many many different people and I'm just a small piece in the puzzle for part 2 haptires who who has basically generated the protein reagents and then transferred the technology for us to generate additional protein x cam for the screening university of Queensland for the animal model and then many people involved in AstraZeneca as well in different areas of the of the company the tea bag purification that has been a very nice and fruitful collaboration between University of Copenhagen, Lund, Gothenburg and in the groups at AstraZeneca and with that I would like to thank you for your attention and I think we can open up for questions thank you so much Arjan I think we all totally agree that we are in a very exciting time so we are not a little bit late but we can take just a couple of questions and then maybe we can keep some of the questions for the general so I'll just take a few of the first one that came in and one question was about tea bags if they are reusable yes they are reusable so you can you can basically wash them after elution so for example flag tech we we do the the glycine low pH and then a re-equilibration in all pH and then washing in your detergent buffer again and then you can reuse them also if you can use other resins than iMac or have you used other resins than iMac yes so we we've used affinity resin like map select sure for fc flag we have used a range of different iMacs we haven't been very successful with the iron exchange resins but the the high affinity resins they they work well for us have you used for membrane proteins have you used sma smokes for in the tea bag system as well no no we haven't we we'd love to to do that we haven't really uh gotten the the small routine and protocols in in our labs and i'm i'm looking forward to do that but there is only so much you can do okay just a few more questions concerning your the second project or say presentation part so the mutations how were they selected yeah so that is that's based on on systematic mutations so starting from basically well not not one but residue residue for residue point mutations carefully looking at the pharmacology and and the thermal stability identifying those that are more stable and then recombining that and uh so that is uh technology and and that was uh done at at haptares and i'm sure chris will show more about that in in his talk but it's a systematic changes careful thermal stability assessment and and recombining those just one last question that i take there is the at this n-term real life design and does that affect the on and off rate for the for your compounds i suppose also that goes along with the mutations how how do you know that those do not affect your uh the on and off rate that uh that is independent on the that is in largely independent on the t4 life design and that is based on the cell uh essay data that we have which in which we use uh clearly non modified uh part two and we see this pre incubation effect that if you incubate longer you get a um higher apparent potency