 Back to Madison, former associate of our department, former grad student, postdoc, faculty. Dr. Charles Dabert is senior scientist at the Research Institute of the Hospital for Six Children in Toronto, very much a center of excellence in biomedical research. He's also professor of biochemistry at the University of Toronto. He obtained his Bachelor of Science in Chemistry from the Polytechnic Institute of Brooklyn. That's where he's from. You don't expect a Canadian accent. He has a PhD in organic chemistry from MIT, working with Arthur Cope. And he was then a postdoctoral fellow at Harvard with Alkan Bloud. After that, he spent a period of one or two years, I believe, here in Madison, working at the Anzheim Institute in collaboration with Harry Lardi and then left to take his position as an assistant professor then in Toronto. He has a long list of honors, award, and professional achievements. He has over 250 publication international journals. He has been president of the American Peptide Society. He has been the editor-in-chief of Peptide Science and in the editorial board of many other journals. He's a fellow of the Royal Society of Canada. He was awarded the Vincent Duvenal Award for Outstanding Achievement in Peptide Research by the American Peptide Society and the Moray Gold One Award for Scientific Excellence and Mentorship in Peptide Science. And I could go on. If I had to define Dr. Diber in a nutshell, for those that don't know him very well, he is an outstanding peptide chemist. And he is somebody who applies this to understanding folding and structure of membrane protein, a field that he has pioneered. He has been studying a factor that drives association and insertion of peptide in lipid bilayer and detergent, studying motifs and physical factor that drive this phenomenon, like van der Waals, hydrogen bonds, aromatics, chytion pi, and so on, laying very much a foundation for very much everything that we do in my lab. He has done most of these studies, primarily in the concept of biological important molecules, such as CFTIR, which is the chloride channel that causes remutation, cystic fibrosis, or viral cold proteins. Another interesting fact that I learned from Dean that he is a recurrent contributor to the New York Times crossword puzzle. He's done that right, which I found very interesting. I think he contributed a couple of years ago, one of the 50th anniversary of the Beatles, right? Yeah, recently, yeah. OK. Today, Dr. Liebe will present some work, recent work, on a new class of peptide that have some very outstanding properties, especially in light of antibiotic resistance. Without taking more of his time, I would like you to join me to welcome him back to Madison. Well, thanks, Alessandro. That's a beautiful introduction, and we'll try not to let my Brooklyn accent distract you too much. Going to tell you today about, as the title indicates, something about antibiotic resistance. And it's a topic that really came out of all the transmembrane segments and all the peptides and membranes and so on. It's one of these things, curiosity-driven. It isn't something like we started out to do, but you will see where it ended up. I do want to say a couple of things first. It is very honored to be the Gladys Everson lecturer. From what I read about her on the website, it seems she was a pioneer in the importance of exercise and the importance of good nutrition. And those topics, of course, are very current today, and I'm sure she would relate to that very favorably. As far as our experiences in Madison, my wife and I did come here from Boston, myself from Harvard, she from MIT. She was assistant professor in the political science department, and I was a research associate with Henry Lardy in the Research Institute. We were here overall two or three years. She is Canadian, and eventually we got two permanent jobs in Toronto. But we certainly remember our stay here extremely fondly, and it's just great to be back here. Well, most of you know that bacteria are very tough customers, and there's been a lot of negative things in the news about our fight to control them. The number of new antimicrobial drugs, these blue bars here have been heading down. And meanwhile, the number of resistant strains of these bacteria, including ones that are quite well known, enterocucchi, the MRSA, pseudomonas aeruginosa, which is the main pathogen and cystic fibrosis. I'll come back to that. So the question is, of course, what are we going to do about it? So bacterial antibiotic resistance is a major challenge, and clinicians are beginning to run out of options actually with the classical conventional antibiotics. Majority of bacterial infections involve biofilms, sort of colonized bacteria. And I'm going to get into that in a little while. And very few antimicrobial agents are in the pipelines, as we just saw. So approaches with novel mechanisms of action can certainly be used. So why are bacteria so good at this? How do they fight off all the drugs that we give them? Well, we can look at this kind of schematically. If we have an antibiotic, for example, it gets across the bacterial membrane. It combines with an enzyme, and it inhibits the action of that enzyme. So all of a sudden, the bacteria aren't inhibited anymore. They just make a slightly different enzyme. There are other factors and enzymes in the bacteria that have power to just chew up just about anything we do. And there are bacterial pumps, these so-called small multi-drug-resistant proteins that are in the membranes of the bacteria. And if you give the bacteria a drug, it just pumps it right out, pretty efficient. So what are we going to do to combat these types of mechanisms? So since our work in the lab is focused on peptides and proteins and membranes, we're going to look at the bacterial membrane from a couple of different points of view. Now, in order to fight bacteria, it's not just the antibiotics that we give them. Our body also tries to fight off the bacteria with antibiotics. And most organisms, everything from butterflies, to frogs and birds and mammals and so on, they all secrete some kind of native peptides in conjunction with the immune system to try to fight off the bacteria. And most of these peptides, which have names like macanin and sycropin, indulcide and protagrin, some of these are names you may have seen. And these are peptides that are basically membrane active. They just go to the membrane of the bacteria. And they don't have any specific target. And the idea is to get into the membrane and just disrupt the bacterial membrane. So what I'm going to tell you today is two themes, two strategies that we've tried against bacterial resistance. The first one is the development of membrane active peptide antimicrobials, as I just mentioned, that act directly on the membranes, rather than on some sort of protein or receptor target. This is less likely to evoke bacterial resistance. And even though all the organisms do this, it's obvious that whatever are the secreted peptides, they haven't been enough to kill all the bacteria. So maybe there's some new twist on it that we could think of. And the second thing we're going to talk about in the last part of the seminar today is the design of membrane-based peptide inhibitors that actually inhibit the ability of the bacteria to pump out the drugs. So we'll look at those two things and we'll see what we can do in the laboratory to get at that. So if we look just for a minute at a structural comparison between so-called conventional antibiotics and what antimicrobial peptides might be like, you can see just at a glance that some of the well-known antibiotics like Tobromycin and aminoglycoside, Cyprophloxacin is a quinolone drug. This one called meropenem is a penicillin derivative, one of these beta-lectam drugs. They each have some sort of target in the bacteria. And these are the drugs that bacteria develop the resistance to. So on the other hand, meganin, as I mentioned before, is an antimicrobial peptide from frog skin. And it acts by a mechanism of physically disrupting the bacterial membranes. So it's the meganin is the one that we're going to try to model ourselves after. So if we want to make peptides like meganin that get into bacterial membranes, there is a practical side to all of this. They are very insoluble in water. And they should be because they're supposed to be in membranes and not in water. So we have to think of some way to make membrane active peptides but still make them soluble in water and be able to make them purifier and characterize them. So transmembrane helixes, as they're called, or TM for transmembrane, highly hydrophobic. It's challenging to synthesize them. They're not amenable to characterization. And we required a solution in the lab to overcome what's been called the insiability hurdle. So we borrowed basically a feature from natural transmembrane segments. You can see these ones from native proteins like M13, glycophorin, and CFTR, is a cystic fibrosis protein. When they exit from the membrane, they usually exit with a bunch of positive charges. Often lysines, sometimes arginines. Sometimes they're on one side concentrated here. Sometimes they're on both sides. So basically the idea is we would make hydrophobic peptides and we would make sure they have enough positive charge in order to help them be water soluble. And then when they approach the membrane, they use their hydrophobic portion to do the interaction with the bacterial membrane. So this is what it looks like in a little more detail. We call these cationic antimicrobial peptides or CAPS. So I'm gonna use that name from now on. So natural CAPS like meganin, it's the one you just saw from the frog skin, amphipathic helix. And it means that the charges, the positive charges and the hydrophobic parts are interdispersed all throughout the sequence. So when you fold this up into a helix, you've got one face that has just positive charges, the other face that has just hydrophobic residues. This is the sequence meganin. You can see the blue residues or the lysines, histidine. So when this folds up into a helix, you can see the positive charges are on one face and the green hydrophobic residues are on the other face. So if this is attracted to a bacterial membrane, the positive charges go to the phosphate groups on the head group, and then the other residues like phenylalanine or leucine point down into the membrane. We design and synthesize a sort of a novel version of that, but what we did is we separated the charges and the hydrophobic residues. And this is designed anyway to facilitate the penetration and destruction of the bacterial membranes. So here you see a model of this. You notice all the lysines are all grouped together on one side, and then you have an uninterrupted hydrophobic stretch. So the best ones are these. We found to be 17 residues, six lysines and 11 hydrophobes. And the best ones that I'm gonna talk about today have this kind of general structure, AX, Fax, W. We always put tryptophan in as a fluorescent marker, AX, Fax. So the F comes, the 6F17, are we calling it that? Here, I don't see that here. We calling it 6KF17, and there's 17 amino acids and X can be a mixture of alanine or leucine, and that depends on how hydrophobic you wanna do it. So we thought that the advantage of this would be, in contrast to the natural amphipathic peptides, if you take a bacterial surface, just as I mentioned that the peptide can be gripped into the surface, but it's held there by the lysines and it can't get down too far. This is the idea. Whereas with our peptides, which we call instead of the grip, we call it the grip and dip mechanism where we have the peptide in the surface of the bacterial membrane, but then the lysines stay up here and we think that the hydrophobic part can penetrate and be more effective at disrupting the membrane. This is of course a dramatization and it's at least the thought behind the experiments. So coarse segments, the 11 residues can vary and the hydrophobicity, if you're talking about inserting something into a membrane is a huge factor. It has to do with how it interacts with the lipids. It has to do with propensity to partition from the water into the membrane. Several factors going on. So this is the chemistry of this, what these peptides actually look like if you put the lysines off to the side. You can see you can have alanines in this position, phenylalanines, tryptophan. So what I'm calling 6KF17 is this one AAFAA, WAAFAA, so that's this one. And you can improve the hydrophobicity by putting leucines in instead of alanines. So in some positions, we put in four leucines in this particular example. So these are where they substituted for alanine. So the 4L version of this is gonna be considerably more hydrophobic. So even on this simple design really, you can make millions of different combinations and derivatives. You can put the 6Ks on one side or if you want to, you can mix three and three just to test out the importance of gathering the charge together. You can put no leucines or two or three or four. You can put D residues as in this one in the lower case letters and you get different hydrophobicities for these different examples. We make each one, do it by a solid phase peptide synthesis. We purify them by HPLC. We check them out by mass spectrometry to make sure we have exactly the right compounds. So these are all relatively straightforward and easy to make. Another thing that's very important is you have to show that these will insert into a membrane and be helical because that's the trick. And in fact, if you try this in a detergent like sodium dodecyl sulfate, most commonly used in the lab for this SDS, you can find that the 6KF17 gives this nice helical value when it's in SDS, but it gives a random coil that's the green dotted line here when it's in a water phase. So it's water soluble, but when you put it in the presence of SDS, it enters the membrane and it becomes helical. So that's exactly what we want to happen. And in fact, the forlue version of this does exactly the same thing. So now this is one of these one day in the lab things this goes back quite a few years now. We were aware of all the natural caps that humans and animals secrete and we realized some sort of relationship between our peptides and those. And in those days, we had longer ones and shorter ones and all. So we decided to test them against bacteria just to see if they had any antimicrobial activity. And this is without reading all the numbers is just to point out these ones like 6KF17 and so on, if you test them in E. coli, you get these so-called MICs, they're called minimum inhibitory concentrations. That's what it takes to fully inhibit bacterial growth in these very established protocols that carry this out. The main thing is you want very low numbers. If you have two and four and one and 0.5, you're doing pretty well because in this assay, the macanin is up at 16. So we are, in many cases, even better than the natural one as in all these different, these are different strains. PA stands for pseudomonas, so we're gonna know. So I'm gonna come back to that. These are the some of the strains that you get from the cystic fibrosis lung tissue. So caps physically disrupt bacterial membranes. They cause cell lysis and death. The advantages are no specific protein targets, less likely to evoke resistance. But here's another very important thing. They're highly selective against the negatively charged bacterial membranes. And when you have negative phospholipid head groups and lysine-rich peptides, they are attracted to each other sort of like an electrostatic complex. And although the mechanism isn't really 100% proved, people talk about this in terms of the carpet mechanism, where the peptides accumulate on the surface, and then they eventually just overwhelm the membrane and disrupt it. So this is sort of the carpet mechanism in combination with the grip and dip mechanism. So it's interesting to point out that bacterial membranes are highly anionic but mammalian membranes, for example, erythrocytes in our blood cells, for example, those are neutral. They are what's called zwiderionic. It means they're net-neutral. They have a positive charge in the head group and some anion from the phosphate, but they end up being neutral. So this is a very good thing when you're trying to devise new drugs. The mammalian outer leaflet is 100% neutral lipids. The bacterial intermembrane, 75%, 25%, 25% anionic. They are also different terms. The mammalians have cholesterol. There was no cholesterol in the bacterial membranes. But because they're positively charged, our caps are highly selective against anionic membranes. So you can see this in a couple of different ways. If you just do fluorescence studies on the compounds and just look for the blue shifts, and here what we did is we incorporated or sonicated our peptides into liposomes that are made up of just the mammalian lipids. So these are all the net-neutral lipids. And you look at some of the compounds I showed in that list before, you see the 6K F17 has no blue shift. The blue shift is characteristic of putting tryptophan in a more hydrophobic environment. So these ones down here, 6K F17 and so on, no blue shift. So that tells you that they're not entering the membrane in the mammalian types. And whereas if you start increasing and putting more leucines, these 4Ls into here, all of a sudden you get huge blue shifts. And that means that the peptides are directly entering the membrane. So if this happened to an erythrocyte, that wouldn't be a good thing, because if you want to drug against bacteria, you don't want it to hit the host also. There's a couple of ways to look at that here. Look what happens if you add caps to some violators that have been painted onto a face here. Just the tiniest little bit, 1.5 micromolar, and you've already completely disrupted the bacterial membrane. And the 4L compound does that also. And I don't have the slide here, but if you try this same experiment with a mixture of mammalian lipids instead of bacterial lipids, it looks like this. It's like two things that just simply just don't see each other. So this is a very, very powerful cap at disrupting bacterial membranes. And this is kind of the biology proof here, if you will. The best caps are non-hemolytic to human red blood cells up to 320 micromolar. So you mix the caps with suspensions of red cells. If they get hemolyzed, the hemoglobin comes out and you see the supernatant gets red. If the cells are not disrupted, then it just stays clear. And you can see the stark contrast here, which fits the blue shift work that I just described, that the 6KF17 with just alanines has absolutely no hemolysis. The 4-lucine one is really hemolytic. So you can immediately rule out anything with lucine like that as being a possible antibiotic because it's dangerous to your own cells. And this number, which is not sure visible on the charts here, goes up to 320. 320, that's a huge high number for hemolysis. If it was meganin, it would be down around 10 or something much lower. So all the properties of these peptides is in the so far so good category. But there's one twist on all this work. And the twist is that bacteria are not always free-swimming around. It was a called planktonic bacteria that was free-swimming. Most of them that infect us end up being colonized into what are called biofilms. I don't know if any of you have heard of this word. I mean five or six years ago the word biofilm was not so common. Then I heard it recently on a toothpaste commercial that you have biofilms on your teeth. So these things are basically the same bacteria but they become colonized. They adhere to your tissues or to your teeth, your prosthetic devices and so on. And then once they form their colonies they start secreting mainly big long chains of polysaccharides. The ones from Pseudomonas are called alginate. It's a major type of what are called exopolysaccharides and they're formed in these biofilm colonies of the bacteria. They are basically sort of derivatives of mannose and this one gulose. And the difference is they just have a carboxyl group on it instead of another OH. But basically they're just huge sugar chains. They come in mixtures of M and G blocks and depend on what kind of alginate. They're actually found in seaweed. The word algae comes from seaweed. We've never been able to establish the connection between Pseudomonas and Regonosa and seaweed but it is point to ponder. And the sugars themselves have conformations just like proteins have helices, right? The sugars have things that look pretty organized like this. The ribbons structure is almost helical. And just to note in passing, some of them have well displayed CH groups. The CHs on sugars can occur in different combinations. In mannose, for example, all three of them are on the same face. And that can play a role if you're trying to interact peptides with polysaccharides. So biofilms, by definition, then are surface-attached microbial colonies enclosed in this sugar-polymeric substance. And that makes them way more resistant in the same organism in the free floating stakers now. Whatever antibiotic you give it has to penetrate and get to the actual bacteria. Pseudomonas is one of these biofilm-forming bacterias and many cystic fibrosis patients develop these infections. And even though CF is a disease where the salt balance is not maintained, the mortality is mostly from these Pseudomonas infections, actually. So here's the situation, right? You've got to say the bilayer of a bacteria. You've got a cap that's coming and it's going to interrupt the membrane. But all of a sudden there's a polysaccharide layer between them. So what is the peptide going to do when it hits the polysaccharide? Got to get through. We found, and this is just one example of a lot of different studies, that the caps that we're working with, here's the example that happens with the forlue peptide that we did, but it works with all of them, that the peptides actually bind the polysaccharides. This is an understudied area, I have to say, when we first googled, you know, polysaccharide antimicrobial peptides, virtually nothing. And the whole study of how peptides get together with sugars is quite an interesting thing just in itself. And considering that you have anionic polysaccharides and positive peptides, it's really the same thing. They're offering each other the same interactions that the bacterial membrane and the cap would offer to each other. So in effect, our peptides really bind these polysaccharide layers, and that can be used to advantage. So here's a chart, I don't mean you to read all the numbers here, but I want to show you what happens when you have a biofilm instead of a planktonic. So the MICs and the biofilm inhibitory concentrations, which we call BICs, we're comparing them with one of our design caps and the conventional antibiotic Tobromycin, which is widely used against biofilms. So you can see some of the numbers here I'm highlighting just a few of them here. But for example, let's take this resistant strain. The MIC is 8 to 16 for the MIC, and then it goes up to 64 to 128 for the biofilm. So you see the biofilm is much more resistant. That's what that tells you. Now if you go to Tobromycin, which is the drug that we're actually using in the clinic for these things, you can see the MIC is around 32 and greater than 128 for the biofilm. So what this means is that our cap has lowered both the MIC and the BIC even lower than the one that's being used in the clinic. So this is a very good result. Here's another one. This one in the MIC is way lower. So with this in mind, we undertook some combination therapy experiments. And basically this is a grid where you add your peptide, let's say, in one dimension and you add your conventional antibiotic like Tobromycin in the other direction. You mix them together so each well has different concentrations and ratios of these things going from the lowest combination. So there'll be one micro molar, say of each, or one microgram each here and up to 128 micrograms each up here. And you look to see how they work alone and then how they would work in combination. So here's an example of how we graph these things. Tobromycin alone, if you use it against this particular resistant strain of pseudomonas in the planktonic form, no matter how much you add from 0 up to 128, it's green, which means that Tobromycin cannot fight this strain in that range because they grow no matter how much you put. If you try the peptide alone going along this axis, when you get up to about 32 micrograms, it starts to kill the bacteria. So anything above here is all no growth. So this is what it would look like if you had done these independently and just graphed them, you would see something like this. So now what would happen if we mix them together? So here we're testing against the planktonic resistant strain. Here's Tobromycin alone. It's still green here. The peptide alone is still red above the 32. But now the growth has been stopped in all of these wells. So in other words, the combination of these two is highly synergistic. And now between the two of them together, they're able to kill the bacteria in much lower doses. So this is a very encouraging thing here. I really must back this up one of these days or years. Here's another interesting example of this. Here's a resistant strain, and now we got the biofilm. So this is the big enchilada here, right? Biofilm. Okay, so on this strain, Tobromycin alone, no effect. Arpeptide, one of these 6KF-17s alone, no effect. So if we use either one of these individually, neither one of them can touch the biofilm form of this resistant strain. And this is the toughest ones in the lungs of the CF patients. I should say that these combination therapy experiments have been tested on clinical isolates. So this is from the sputum of CF kids at the Hospital for Sick Children. So they each have a different strain of pseudomonas. So we're testing them in a live way. This is one of the good things about working at the hospitals without the clinicians and the scientists kind of work together. This is a really good example of what could be accomplished here. And this is a very good result because now we have a considerable region of red, which means that in combination now these two drugs have the potential to kill the pseudomonas in the biofilm form at relatively low concentrations. 32 micrograms is still pretty low as these things go for doses. So very pleased about all of this. And we did just for sort of controls. We did the three antibiotics that are used most often in CF therapy. And Tobromycin, Ciprofloxin, I call that Cipro, Meropenem, and basically just to summarize all of this, mostly they're just additive and there's very, very little synergy. So even when doctors use sort of a cocktail, as they call it, of different combinations of these, the combination really doesn't enhance the situation. Of course these are all in vitro studies and everybody, every patient is different, but it's all very promising. So to sum up basically what I've been telling you about all these caps and the combination therapy, we think there's a dual action mechanism. We think the first thing that happens is that the peptides can bind the alginate or the polysaccharide layer. If you get enough peptides, it can basically disrupt the alginate and some of the peptides actually precipitate the alginate. You can see this in the test tube. So once the alginate layer is disrupted, one of the conventional antibiotics, so this is the combination therapy idea, can get past the alginate and get into the cell. So it's almost like changing the bacteria back to a platonic, planktonic. But that's not the whole story because now, remember the caps can coat the bacterial membrane surface like a carpet mechanism and weaken the membrane that way. So now the conventional antibiotic has an easier path into the bacteria. I can watch this all day actually. Okay, so right now, a lot of interest around our place is in the caps and we hope in the longer range people always ask about commercialization of academic discoveries and we hope this may be an example of it. So for the last few minutes here, I'm going to give you a survey of something that's happening relatively recently in the lab and that is the idea of trying to stop bacteria from pumping out the drugs that we add to them. Now as I mentioned right at the beginning, the bacteria have small multi-drug resistant proteins as depicted by that blue cylinder if you remember and that's what this thing is. They call SMRs. A lot of bacteria have them. This one is called EMRE. This is one that happens to be from E. coli. Now bacteria have like five different categories of these pumps that pump drugs out. The two of major interest to us will be these small SMRs and then this big guy here which is called ACRB, ACRA, Tulsi complex. This part of it in here in the membrane is very similar to this part of this protein and then it goes into a long and cooperative pathway ending up in a sort of a beta-barrel structure that enables the bacteria to pump the drug out and pass the periplasm and out past the cell wall. There is one and finally all out. So actually some people believe that the small ones pump it to here and it gets about this far and then continues the rest of the way. So between these two, they are in the major sort of drug pumping complex. But now you may well ask here doesn't this look a little bit small to be pumping out some big drugs? So let's look a little more closely at these SMRs. They usually just over a hundred amino acids. They have four transmembrane helices, about half of the common bacteria have these things including E. coli and Staph aureus and HSMR which is called H. selenarum. It's one of these archaebacteria that's used in lab studies and I could add to that list pseudomonasuriginosa, PA also has these. There's a mechanism where protons go in and drug molecules get driven out and interestingly with these things there's a brunt specificity because you've got aromatic antibiotics, antiseptics like disinfectants, dyes and so on all be pumped out. So how does a small molecule pump out all of these things? So the X-ray structure and just other evidence suggests that the minimum structure of these SMRs has to be a dimer. You can't have one thing with four helices pumping out big, you need some kind of channel or something. And it's believed to be formed by two molecules coming together with six transmembrane helices in this configuration and then the substrate goes through the middle. This one shown here is a tetrafenyl phosphonium. It's not a high resolution structure but it does give the idea that this is the main business part of the molecule but it still has to be held together as a dimer. So the dimer is held together up here the TM4, the fourth transmembrane segment is like away from the other three and in combination then you have six helices here and then two here and these two bind each other through a cross section of residues to TM4, TM4 interaction that is basically holding the dimer together. So you can look at the sequence TM4 and see what's going on with there. The X-ray evidence suggests that this is the cytoligomerization. We used an H-cellenarum, HSMR because it shows the dimers on SDS page and we made a bunch of mutations to the TM4 in one of these SMRs just to show that the TM4, TM4 interaction is the main thing and also to determine what residues are actually involved in that. You look at some of the actual sequences you see some sites are conserved all throughout different bacteria evolutionarily conserved so this gives us an idea that somewhere right in here between about 88 and 99 somewhere in there this is most likely to be where the TM4 and TM4 interaction is a little bit like a scissor or cross section through the middle of the TM. So we made mutants of all residues and TM4 focusing on the... these are just some of the main results here and basically what we did we made some small to large some large to small like L to A or V to A if we had G we made it G to V and so on just with the idea of disrupting any motifs that might be dependent on the size or shape of the side chains and we got this pretty remarkable result here that if you graph these residues on a helical wheel so you're looking at it from the top and you see all the residues half of the residues had absolutely no effect really on the ability of the proteins to pump out a toxin it's the one we use is called the Thidium bromide it's a laboratory model on the other hand if we use these ones given in red which maps to this face here you can see the distinction all the disruptive mutants are on one face and all the green or non-disruptive mutants are on the other face it's pretty well telling you that this is the face that's going to be involved in the TM4-TM4 interaction and in greater detail this is what it looks like here you have G90 to V98 you have this GG7 motif which is similar to the GG4 two turns of the helix that allows the helices to have a close approach you have a couple of hydrophobic residues with leucine and isoleucine and valine and then a couple of other these X residues that can be one of many different things so this is basically the dimerization site and this is what's holding the the SMR into a dimer so to summarize what I've said so far here you have the dimer you have these three helices, TM1, 2 and 3 there is a glutamic acid that's conserving their glue 14 that has to do with a transient interaction site as the substrate passes through and then you have the motif G90 to 98 and this is the minimal dimerization motif almost like two separate parts of the dimer so this protein can pump out ethidium bromide which is a toxic dye that has certain fluorescent properties that will be useful and it can also pump out something that's called benzalconium if you recognize that, that's the disinfectant that's used in hospital surfaces when you swab down surface so actually this protein can pump that right out and that's not good because you want it to kill the bacteria so here's the mechanism of this dimer here and now instead of things going into the bacteria like before, this time they're heading out this is the tetracycline and one of the substrates for the SMRs so what are we going to do about that if we want to stop it and we have some understanding of these membrane active peptides we'll make a synthetic peptide based on TM4 that seems logical place to start and we'll hope it'll have some sort of transmembrane insertion and then we'll head to the TM4, TM4 interaction bind the TM4, compete the normal TM4, TM4 dimer out and create in effect two monomers and these monomers will not be able to pump the drug at least that's the concept so we hope to inhibit tetracycline transport so in order to do that we had several iterations before we hit on one that really worked some prerequisites, first of all you have to have positive residues as I mentioned there else the peptides are going to be insoluble but in this case you have to have one end that's neutral because if you want the peptide to sort of tunnel into a membrane like a corkscrew kind of thing you can't have lysines on that end because it's not going to enter the membrane so what we have anyway is if we could design such a peptide we hope that we would be able to inhibit the export of the ethidium and basically in this case our idea is to just break the two up into two monomers well it works so if you make a peptide that is just TM4 that's the exact sequence of the wild type TM4 you put three lysines in over here just for signability and on this side we used after a lot of trials and errors sarcosine if you know what it is it's an N-methylglycine it's something that's called a peptoid so proline would be an example of the natural thing where the N is substituted by another carbon bond in this case it's just N-methyl so it's actually a very simple glycine derivative but because you put the methyl on it gets much more water soluble that's what happens actually you get a secondary amine function and right on the end of it is acetylalanine so what happens here if you take some E. coli cells you suspend them and you add ethidium bromide which is a toxin the cells just pump it out so the ethidium is a DNA binding dye that is pumped out and it changes its fluorescence depending on whether it's bound to DNA or just out so this is the thing you normally see the fluorescence signal drops so what happens if we do the same experiment in the presence of our TM4 derivative here then we're very pleased to see that the E. flux slows down considerably in other words the TM4 to a great extent actually about 60% is inhibiting the bacteria from pumping out the ethidium there's always some background stuff because there's other transport mechanisms and so on but this we thought was a very good start so now if you want to make this a useful inhibitor that might actually be used in real life something of course it's way too long so the natural thing is to see how short can we make it so that it would still work and what you're thinking of now is that heptad and the center part of the TM4 interaction so that's exactly what this sequence is here and we can make shorter peptides with just TM4 90 to 98 keeping the sarcocene and lysine we can make it too longer we can make it too longer so keep in mind the numbers here the shortest one is 90 to 98 the longer one is 88 to 100 so first of all we showed an interesting thing that the 90 to 98 and the 89 to 99 are helical in the SDS but not as much as the natural TM4 in red but the 88 to 100 suddenly jumps to the high helical value so in other words by making it just exactly that long we're able to have the same amount of helicity so that's a good sign that it sounds like this thing is penetrating into the membrane so here's these bar graphs they're very hard to read I'll just tell you in like one minute what this all means in order to test whether these things are very useful we did four types of experiments and you'll see what each one first of all we did the ethidium efflux assay that's just the one you saw with the two things coming down so here's our SMR set at 100% so we saw the TM4 inhibits quite a bit then we found the 90 to 98 and the 89 to 99 as we found now don't do much of anything but at least the 88 to 100 is down to about 50% so that means in principle at least it's working so now we did what's called the resensitization assay and all you do here instead of doing the kinetic thing with the efflux you just incubate the bacteria overnight with your peptide of interest and you see if the bacteria are dead in the morning and if they're dead and they're down here because it only takes 1.2 micromolar from the TM4 to kill them and the 88 to 100 work pretty well the other two nothing the bacteria just go on living so you can see right away from this the 88 to 100 is really the minimum life that's going to work here then we got a real fun day in the lab here we found in a hemolysis assay remember that thing with the red things and our TM4 the original one that I described killed every red cell touched really poisonous the first sense as well we tried that was not a good thing at all but the other three including the 88 to 100 have no effect on the red cells so at least got to reprieve and in another assay you can study these things just as bacteria the way you do the MIC assay you just see if it's an anti-microbial peptide or not despite the mechanism of disrupting the SMR and in our case it looks like the 88 to 100 works pretty well just as an antibiotic that's not necessarily what you want because if it's an antibiotic mechanism it could be evolving resistance at some point but that's a side thing so we did one more thing that I'm going to tell you about here which is where this work is really at right now last couple of months or so there is a reaction that is called the Grubbs metathesis and my organic chemistry is long behind me and I'm looking at Chuck here but this catalyst basically it takes in summary it takes two unsaturations let's say R2 here and R1 here and it goes through a cycle which creates an R1, R2, olefin between the original two separate ones and this leads to the introduction of what is called a helix staple into a peptide and Greg Verdeen pioneered the use of this for peptides not just the organic molecules but peptide specifically and what you have to do during the synthesis you have to get this commercially available pentenol alanine which has not only the methyl group but a side chain that comes out like this with a double bond on the end so you see you got the R1 double bond and the R2 double bond that's the idea of this and what you want to do is use this Grubbs ruthenium catalyst 2 here and you can try it we tried it first with the 90 to 98 and you get a peptide that looks like this and now you have an olefinic bridge in the middle and actually this peptide is now more hydrophobic than it was before because you have all of this CH2, CH2 all of this stuff that makes a very interesting compound but the idea is to freeze out the geometrically crucial TM4 peptide so that it basically if you choose the positions of this and this, the two alanine derivatives you go one turn of the helix you lock it in on one side and then the backbone over there is exactly right for the helix that's going to interact with the TM4 in your competition so that's the idea these have been used in pharmaceutical circles now to do just that to freeze out the pharmacophore of interest first time it's been applied to peptides to our knowledge so now we know the minimal motif needs to be 80 to 100 and we can operate on some of the residues within that heptad section because those are not the ones that are evolutionarily conserved so we can for example take an X residue here and change it from the wild type to veil or alloglycer basically anything we want to do and we made some of those stapled ones without going into too much detail they are suitably alpha helical but you'll notice here that they're starting to take on a little bit of helical character even in water which gives you an example that the the conformation is stabilized and now we got some good results with here this one here the stapled one I'm calling V95X inhibits 67% in that efflux assay but now it's down very very low over here so this is a great one for killing the bacteria resensitization and here's the good news this one is not hemolytic it's hard to predict which ones will or won't sometimes and it has a little bit of bactericidal activity so we actually have an inhibitor here that fulfills essentially almost all four of the criteria that we're after and we think we're on the right track to make a really good inhibitor here so this is something like they say in the movies you know it just might work the most recent data on this shows that the helix staple protects the peptide against proteolytic attack and in blood plasma samples a liver homogenates the lines stay right at the top it means that they're not being metabolized so that's very preliminary so to summarize all of what I've been telling you here we have a new category of designed antimicrobial peptides that are highly active against a variety of bacteria unlikely to evoke resistance no hemolysis we tested different types of African green monkey cells with no effect and it obviously improves the activity of conventional antibiotics and the conventional therapy assays so now we have to test the efficacy of this and CF model mice to see if they can inhale it and kill an infection we have to test their efficacy in human sputum samples more further than a couple of trials that we did and this is something which is a big deal, very expensive you have to evaluate the safety in the pharmacokinetics and animal models before it can go further as far as the helix staple is concerned I think we've identified a threshold sequence from a TM4 that shows helicity and inhibitory activity both the helix staple then retains the activity the increase in hydrophobicity leads to bacterial and hemolytic activity so we can put hydrophilic amino acids in the X positions and find just the right balance and the staple increases peptide and metabolic stability so these are two examples of things that we think can actually confront bacteria and with the appropriate amount of R&D they might in some way some future time actually prove useful this work is a lot of people behind it I get the chance to present it but I'm really very thankful to the members of the lab and the co-founder just to mention Catherine did the staple work that I just described and Brad Poulsen did the original TM4 work with the SMRs Ariana Rath is involved with the CAHPS work and Lois Yen did a lot of the testing and the combination therapy especially so I appreciate their contributions enormously and especially on the CAHPS project as you can see from this list we had collaborators that came from many different walks of life I'm certainly not a card carrying microbiologist to say the least so when you start working with bacteria you learn a lot of things you didn't know for example lysine got its name from the word lysis which means that it blows up everything when I first told my microbiology collaborator Lori Burrows I had peptides that were good antimicrobial she says does it have lysine I said yeah it's got six of them like I was good she says that's not good so it turned out from what I showed you that if you choose the peptide right and you don't make it 200 phobic it doesn't lyse the human erythrocytes so that was part of the R&D that went into this but we've worked with the computer engineering guys and the molecular structure and function the infectious disease guys the atomic force microscopy guys all trying to bear the maximum kind of information on the project and I thank these funding sources and I thank you for your patience thank you very much question yeah we did several combinations of those and they were all put through these like four panel graph things so the one I showed you was just the best one of those but you are right that it might be possible to make a shorter one yeah you can also make a longer one to put this staple one and seven it might be able to capture just the heptad with that there's a lot of theme of variations possible with that it's a good idea I don't cure our the question is about a cure our I don't know what that is okay that is possible we haven't really done much on that we did the one of the 6KF17s in a mouse model CD1 mice this is not my expertise the person that did it said that up to 10 times MIC there was no detrimental that's all we have but that's way too soon to we got to try all the D residues and you know there's a lot of a lot of stuff that's a good point yeah so the question is about some bacteria have like a cell wall and then a membrane yeah it for some reason known only not me anyway everybody ignores this cell wall whether bacteria gram positive or gram negative they just look at the plasma membrane of the bacteria that's where all the these pumps are and everything and it is kind of enigmatic in its own way how all these incoming peptides how they get through the outer part but it seems to be oh yeah they'll get through it's maybe it could be that the outer wall is intended to be porous because I guess there's an exchange of that I'm not you mean which in the gram the gram negatives yeah they just make it oh there what was the other part of your question that was okay oh here Sam yeah yeah we got 20 or 30 different versions of these things some of them with the lysines in the middle and some of them I showed with the lysines on two ends those are all amphipathic sequences yeah I haven't done that that would be a good control to see if the principle is right yeah you mean in the combination therapy thing there's a thing that's called the Calgary method it's some sort of a thing with a little looks like a little Eppendorf tube with a cap with a bacteria made to grow on the inside of the cap and then it's dipped into the fluids and it's some dynamic thing with the fluids contained in the peptides pass by this thing you have to see it it's a kind of a little mini apparatus you're starting with them you wait till it gets to the mucoid yeah I guess it might just be an OD measurement or something that sort or just you could or you could just visually read the the these caps the little things that look like little triangular they can probably be read in an OD form or something that sort it's back there I see someone yeah you do three I must say all these types of questions a lot of this work is in its infancy when it comes to these you know if these things get if these things are tough on mitochondria we're all we're cooked you know yeah we haven't tried that at something to move toward yep exactly yeah okay there was a question here yes never actually did the NMR on it simply because like you said the CD is helical and all peptides like this over the years are always helical but what we wanted to do is make a one to one complex let's say of one of the peptides with a defined segment of the alginate like you know five sugars and one peptide and see you know try to get some idea from the NMR about how that you know forms a specific complex and how it might aggregate the sugars that is something that's it's not that easy to get the defined segments of the sugar no it didn't do that specifically one more that would be a good thing I've been talked about in order for us to test whether we're really looking at a dissociation of the NMR or whether some other non-specific thing or other pumps have taken over yeah yep that's a good one to do yep one more in the early going when we first made these I realized they might be anti-microbial we tried many versions of this we have arginine instead of what you saw in the lysines arginines are very dangerous to red cells so you know there are some arginine derivatives that go under the name of TAT peptides and there's something about arginines that seem to attract I don't know phospholipids and make themselves into little magic bullets that just go zipping into membranes it's an interesting story on to itself arginine ones are not amenable to this it's a good question