 OK, let's start. Welcome, everybody. This, my friends, is a fourth presidential lecture here at OIS this year. And I'm happy to see people coming in to the door and just quite full in the room. It's great. And as you might remember, the last presidential lecture we had from Professor Fujita on self-assembly molecular structures, that sparked many follow-up discussions and interdisciplinary dialogues. And I am pretty sure that this one will be at least as many, because this is a very interesting lecture you will hear here today. And really, in the intersection between disciplines, that really fits well into OIS culture and the way that we are working over here. So today, we are very honored to have Professor Hiroaki Suga, who will guide us into his world of biological chemistry. And as Professor Suga is an internationally renowned scientist whose research spans across the fields of chemistry, biochemistry, medicinal science, and fundamental biology, maybe even some more. We will hear about that later. So among the major achievements in the development of RNA-based catalysts, which have revolutionized the discovery of bioactive peptides. Yes, I have always also been interested about the intersection between chemistry and biology, and had a very nice discussion with Professor Suga here before today. So I'm sure that you all will continue to take this opportunity later on. So as a great scientist, Professor Suga is also dedicated to education and mentoring, and is also an active promoter of entrepreneurship and internationalization in Japan. In fact, we also talked about the leadership and importance of how universities should think about this for the future and how we are active all the way from the local to the global level. In 2012, he was awarded the Presidential Award for New Inventions in the Government Industry Academy Relations by the Science Council of Japan. And he's the co-founder of companies, a startup company, Peptidream, and is the current president of a chemical society of Japan. And so you will hear. I shouldn't really say all these things because you will be introduced, but I'm sorry about that. But I couldn't refrain from saying a little bit of that. Today, we are honored to have Professor Suga at OIST for a presidential lecture, and in the series of our presidential lectures, which is bringing some of the brightest minds on the planet, a world-renowned scientist and thinkers right here to our beautiful campus at OIST. And so we want them to share not just their groundbreaking discoveries, but also their shared love and passion for science. We want them to nurture interdisciplinary interactions. We all have been attracted to OIST for its creative interdisciplinary research as part of its foundation principles and base to pursue innovative ideas. The OIST presidential lecture serves as a catalyst for these interdisciplinary interactions where the yet unknown science is to be found. So my friends, open your mind and get nurtured by the love of science that extends beyond our walls to the benefit for the broader public. And I want you all to welcome Professor Hiroaki Suga. But first, before you applaud for him, please. Valar, please take on. Good afternoon, everybody. I give you a few other details. About Professor Suga's background, Professor Suga was born in Oka-Yama city, is a distinguished scientist with an impressive journey in chemistry and chemical biology. After completing his bachelor and master at engineering at Oka-Yama University, he pursued his academic efforts at the MIT for his PhD studies and conducted postdoctoral research in RNA biochemistry with Professor Shostak. Starting his independent career, Professor Suga became a tenure track assistant professor at the New York State University at Buffalo in 1997. Later in 2003, he relocated to the University of Tokyo, where he's currently served as a full professor in the Department of Chemistry. Professor Suga's contribution have been transformative, particularly in pioneering the development of RNA-based enzyme, such as flexi-zyme, reprogramming the genetic code. His background work also led to the creation of rapid system, utilizing direct evolution for producing a selective microcyclic peptides. Beyond his scientific accomplishment, Professor Suga is a recognized leader in scientific community, currently serving as a president of the Chemical Society of Japan and the founding editor-in-chief of RSC Chemical Biology. He actively promote education, trains PhD students, and encourage entrepreneurship. Professor Suga diverse talent extend to his passion for playing and collecting electric guitars, showing a multi-faceted individual dedicated to advancing science and foresting academic and cultural connections. Through his creatives and ambitious inter-study disciplinary work, Professor Suga exemplified the spirit of passion, curiosity, and rigor of the distinguished list of our previous speakers at the OIST presidential lecture. Thank you for coming and please. Thank you. First of all, I would like to thank the president, Marokides, and it's a great opportunity to come here and it's really a great pleasure to see how OIST evolved in the last 10 years. Actually, I haven't been in 10 years almost and this is really a big change from 10 years ago to now. And really it's great to see you and thank you for Bora to communicate with me and also a very kind introduction. So let me speak about my research today. These days I use the iPhone to control everything. So, okay, so this is going to be. Today's title is like this, but every time I changed a little bit and often I don't know what title was, but today the pseudo-natural peptides and products and the neurobiotics. It's covering all a bunch of different modality of the therapeutics and for the application. So, this is the first slide I often show to the people that were interested in developing drugs. On the left, there's a small organic molecule that pharmaceutical industry have really made effort to develop the drugs. And this is a molecular waste size of about 500 Doton and the reason of this is very often we can develop the oral available drugs. But of course, the small molecule has also the pros and cons. You can get to the oral available. You can have a very cheap to synthesize, but very often you have suffering from side effects. On the right, this is a biologics represented by antibodies. And this molecule, molecular weight is 150,000 Doton. So it's a huge molecule. But of course, the biologics is a very good to define the molecules that specifically interact with target and giving a very long lifetime in your body that you can circulate around in the blood in over a few weeks. So the pharmacokinetics are very good. On the other hand, it is never be oral available because it's too big. So when I started this project quite some time ago, it's over 20 years ago, that we decided to work on something different from anybody else. And then that's what the macrocyclic peptides are. So at the time, there is even not popular talking about the macro cycle as a new modality of the molecule, new type of the molecules. But it's really the idea is actually the macro cycle would be a very interesting scaffold because of the size of the molecular weight. It's probably less than 2,500 Doton and the chirality rich and the confirmation rich. And of course, it becomes a three-dimensional rich. So it's as a drug candidate, it's probably a very interesting molecule. The major inspiration, I mean, before that, sorry. This is a view of the size that you're looking at. This is the left one is the cyan blue is the typical size of the protein target. And the yellow one is the molecules that I'm interested in and then the right one, the smaller molecules. And you will see the whole entire antibody structure. You can see how the difference between the one molecules to the other, the size-wise. So the inspiration of the macro cycle comes from natural product. And this is, it's called the cyclosporine A and isolated from fungus. And turned out to be, of course, the fungus doesn't produce the molecules for human. It's basically fungus wants to kill the environment, other competitors around the fungus. But turned out to be this molecule is the immunosuppressive agent for human accidentally. But this molecule is a very interesting, the old peptidic molecule. However, the very interesting structure features around. So this is a normal peptide bond is NHCO, but this has a lot of N-methylated CO. And then plus, there are some side chains that which is quite unique, but also here the alanine, but this is a dealanine. And then there's a structure of the macro cycle. So all of these structure features together, this, the cyclosporine A becomes a very strong potent, the potent molecules bind to the cyclophilins to recruit the carmodulin that makes it suppressive activity. But also the structure features gives you the membrane permeability and also the protease resistance. So this is really the inspiration that we wanted to discover this type of the molecules. But the fact is that looking for natural product is not so easy to do. And now it's a people are trying to do a more data, the genome data mining to discover the natural product. But at the time, 20 years ago, I was interested in studying it, there is not such a technology not available, but instead I want to make it something physically synthesize cyclosporine A like library and then you screen effectively active molecule as against any drug target of the human. So the goal is what I set it up is that I want to have molecule to synthesize as a library is over a million, a trillion molecules, not a million, trillion different molecules. The reason is this 10 to the 12 idea comes from the antibody diversity. Antibody diversity is 10 to the eighth of the germ line and then hyper mutation gives you about 10 to the 11th. So the 10 to the 11th of different molecules can be synthesized to then you should be able to get the molecules are very strongly bind to the target because antibody took care of most of the drug, the protein target to be to remove the from our body. So this is a very, very nice evolution system. So you need to have over 10 to the 11th different molecules. So I set the goal with the 10 to the 12, 10 times more. And then I set the goals to screen the molecule to discover the molecules within a month. So in a month, you have to get solution after the use starting the project. So I start developing the methods. First, we have to synthesize the molecules and how we're gonna synthesize the treated in different sequences. And I decided to use the protein synthesis machinery. It's a ribosome translation system. But translation system won't take the N-methylamin acid or D-amin acid that which I mentioned in the cyclosponiae. So I have to come up different ways to overcome the problem. And then also the screening method that which I'm going to tell you about how we can do achieve the goal. So there are many efforts that I had to make this to be realized. This is called the flexesign. And this is one of my most important invention. This is RNA catalyst isolated from completely random sequence. But this sequence was isolated by many experiments that are which I don't discuss today. But outcome is this, the 45 nucleotide and 46 nucleotide lengths of the flexesigns are able to charge a variety of different aminases on the transfer RNA. So I show you here the three prime end of the sequence is complementary to the three prime end of the transfer RNA. And this transfer RNA when it binds the flex time created cavity. And this goes to bind into the substrate and then bring it into the amino acidation on the transfer RNA. The reason we need to make amino acidation on transfer RNA this enables us to change the genetic code which is I tell you in a minute. So this tool enables to charge different aminases and on the transfer RNA of the three prime end of the hydroxyl group. But the recognition of this flex time happens only this dinitrobenzyl ester group. So doesn't care about the side chain, doesn't care about the amino group, doesn't care about the chirality. So you can change the completely different side chain or N-methylated aminases, acylated aminases and D-aminases on the even beta aminases. So there's a kind of all the aminases that usually translation system one tank into the peptide chain. But if you charge this on the transfer RNA, we can change the genetic code. So I will show in the next slide. So under this, I show you the on the left, sorry, on the left is the formyl methionines as an initiator. And generally the translation system enables to elongate formyl methionine to the rest of the amino acid chain. And on the right, this is the chain amino acid chain coding the genetic code. So we have a 20 different aminases there. So from here, we remove formyl methionine in reconstituted in vitro translation system. So this is a cell free translation system that you are enabled to manipulate gradient of the translation system. So I removed first the formyl methionine. Okay, so now you lose the methionine in a genetic code. Therefore you cannot initiate translation. However, if you place this chloroacetyl D tryptophan, this is anything you can put it in, but just as an example, then this is one molecule that we really like it. This will be available for the initiation since that we use a flexis line to charge this on the initiator TRNA. And that automatically this molecule are assigned to the initiator. And then translation takes place after this chloroacetyl D tryptophan using on the right the genetic code. But then we remove, we set it up the NNU combination sequence. NNU is the first two bases are random sequence. So all the possible UCAG will be appearing. And the second also, but then last base we chose as a U. So now it's all black letters can be still usable for this genetic code. But then we remove the father of the certain amino acids. As an example, I removed the phenylalanine, leucine, isoleucine, and alanine. And then we just replace with N-methyl amino acids. Or beta amino acid, gamma amino acid, D amino acid, which I'm going to discuss today, father. So let's say that this is the first example that we published in 2011 that we placed N-methyl amino acids into the genetic code. So in this way, you can actually use this genetic code and now it's the five different non-standard amino acids are in the genetic code. And then you can express by the ribosome. So we synthesize the DNA, temperate, and then you transcribe in vitro and then to form the messenger RNA. Now this messenger RNA starts from the initiation and then followed by random sequence of N-n-u repeat. And we actually, length has a 10, 11, 12, 13, 14, 15 repeat. So we're a mixture of the lengths. And then the last three we placed the assisting codon. So when the transition takes place, you will see D tryptophan with the chloracetyl group at the N-terminus and you have a peptide sequence. And if you hit that N-methyl amino acid codon boxes by the random sequence and you incorporate N-methyl amino acid and then lastly we put the cysteine on it. So this was synthesized peptides and the N-terminus, this is from very much chemistry, the N-terminus is a chloracetyl group and assisting of the cytogenes, the thiol group and this spontaneously deact to form the macro cycle. So you form the thioetha bond. Thioetha bond is a non-reducible bond, stable in vivo. So this is really a lot stable bond that is cyclized. We actually use this thioetha bond formation at that time wasn't very popular at all in the peptide field. People tried to make it macro cycle with a normal peptide bond formation, it's very difficult. So I changed the mind to make it thioetha bond and this makes a huge impact because we get over the quantitative independent from the sequences. Now this method becomes extremely popular. Everybody do macro-cyclization by doing this similar way. So that really changed everything for macro-cyclization. So we made this by translation system but you have to screen because we make 10 to the 12 different sequences. So truly on different sequences cannot be analyzed by independently because it's almost one molecule, single molecule in the tube. So we attach to the messenger RNA. So I'm gonna go through this messenger RNA display system. So this DNA originally developed by Jack Shostack, my former postdoc mentor. He developed concept. So we tweak this ideas in combination of the genetic code reprogramming. So the messenger RNA has a constant sequence that we design and then we chemically synthesize DNA and the three prime end that has a peg linker and then we have a C-C pure omicin. Pure omicin is the antibiotics to terminate translation system. So when this is annealed together and then we often the ligates at the end and then you subject it to the translation system and you start from choral acetyl d-tryptophan and you random sequence and assisting and then you make a translation to take place and the ribosome will go through the translation to the end. We usually have the release factor one comes into this and then creeps the ester bond to release the peptide from messenger RNA. But this is not what we want. We want to keep the peptide and the messenger RNA together. So that's why we use the pure omicin system. We remove the release factor one from the translation system. So all the methods that Jack Schostack developed, they don't do it. So the fusion efficiency was very, very poor. But we remove the release factor one, ribosome does not know how to creep the ester bond so there we can cheat the better with the pure omicin to come in and then ligate it. So now at the end you have all peptide and peptides through the messenger RNAs are connected. Okay, so this peptide sequence is encoded on this messenger RNA. So spontaneous macrocyclicization takes place and then we do the bus transcription to make sure that we have CDNA together and this thioethan macrocyclic peptide is sprayed on the messenger RNA CDNA complex and then we apply to the drug target. So the drug target was immobilized on the bees and then you select a strong binders. Now you need to recover is only the CDNA. So you take the CDNA off and then you do PCR to amplify it. So now we can make a 1000 copy or even 10,000 copy of the molecules depending on what the selection pressure you gave. But then you made a 10,000 copy starting from one copy to the 1000 copy and then you're going to repeat this cycle. Okay, even though you have a trillion different sequences you do go around like four round, five rounds you just could basically you get very good candidates of the pool. So you do a second generation sequencer. You get many sequence to read the same sequence and that's the answer. So we're going back to what the sequence was and what the genetic code was reprogrammed and you look at what the sequence supposed to be and we do chemical synthesis. So you can provide a huge amount of the molecules by chemical synthesis. So this is very often people ask, can you use this to synthesize a molecule? We can synthesize a molecule small amount, not a huge amount because very expensive to reprogram the genetic code. But once you know the answer its best solution is the chemical synthesis. So we do chemical synthesis. Okay, so this is a rapid platform system summary. So I just wanted to make sure that the very good way is that the high cost performance is we usually only use the hundred micro liter translation system, tiny bit. And the later round is only less than five micro liter. So it's a very, very small amount of translation system that you need to use it. And then so you can actually very cheaply screen the drug candidates. And the rapid turnover, I told you that I want to do the one month, it turned out to be less than two weeks. We can actually get two weeks. Now you get to the drug candidate to know what you have to synthesize. And basically the rest of the part is the high affinity to ligand, the very low nanomolar to sub-nanomolar and the success rate is very, very high too. So I just wanted to tell you about the problem that I faced. So this was a 2011, we got worked out everything in the endometrial amino acids into the peptide chain and we demonstrated that we can do the selection. But we had a big issue to present it. Very often the audience ask, can you put in the D amino acids into the peptide chain? So I show you, 2013, we had this result. We tried to make it, which one amino acid, D amino acid can be incorporated. We screened all 19 D amino acid. The glycine is a chirol, so it's 19 amino acids. And we show you here that all the blue bars that are appearing on this figure are the possible that you can incorporate into the peptide chain. So previously a lot of people couldn't get it and we actually able to do quite well. But not all of them, but some of them are very poor and some of them are okay. But the biggest problem is, I'm gonna skip this, biggest problem is the consecutive D amino acid incorporation. So if you consecutively D amino acid try to incorporate it, we have a zero success. And if you have a D, L, D, still very poor. So virtually in this case, so I'm gonna do this, the lighting, this word. So I said the double incorporation of the D-aluminum, D-fianilum, which is a very good amino acid to incorporate a single amino acid, but we can't do double or alternative, alternating incorporation of the D amino acids. So it's basically, you can't synthesize as a library. So we came up a solution by studying a basic study of the tRNA. So this called the protein tRNA. And a protein is a very poor amino acid for normal protein synthesis in cells. So then they just solved problems based on developing another protein called the EFP. So EFP basically stopped the drop off during the slow process of the peptide transfer reaction. So what happens is, I show you the D amino acid here, but this is the same as the protein amino acid. So we supposed that D amino acid is so slow to do that. So before getting the peptide bond formed, the peptide tRNA drop off takes place. So the nature D amino, sorry, the protein amino acid often happens this, so the EFP evolved and that stopped the peptide tRNA drop off. So we applied to the D amino acid and this will help to make it the D amino acid bond formation. So we made this Chimera tRNA that has the effective recruitment of the EFTU. It's a very technical details for translation system, but please listen it, I'm sorry. EFP can recruit the proteam and this is the way really solve the problem. And it took a long time for us to understand that why this didn't work and then we now has a solution. So eventually we are able to succeed even the D 3D consecutive peptide bond formation in the ribosome to make a macro cycle like this. So this is really the only the tRNA which we use, this engineer the tRNA, it's called pro E2 is the possible for incorporation of the D consecutive amino acid. But this didn't stop to the D amino acid. We can even expand to the beta amino acid. Again, the beta amino acid was a big problem that we can't incorporate consecutive beta amino acid. But having this new tRNA, we are able to add three different cyclic beta amino acid into the genetic code. So in this genetic code, we have one, two, three or these are three type of the beta amino acids and then we assign a starting from D tyrosine and D cysteine to close the bond. So this is a design that we made for the new genetic code. So we took the selection against the human factor 12A. This is one of the protease drug target. And then we find the molecule. So this is the slider which I show you. There are a little bit too busy slides so I'm gonna expand it only two peptides which we are interested in, F3 and F4. These peptides have two beta amino acids in here and then KDE against the human factor 12A is a single digit nanomolar. And KI inhibitory activity is almost same, one nanomolar. So it's a very strong inhibitor and this shorter version is about 10 fold less active but still 10 nanomolar of the KI which is a very potent inhibitor. And we also show that this nature chemistry paper that how selective it is. It's a really selective. We don't see any other protease inhibitory activity using these two peptides. We are able to get the solution of the structure. This red part is the beta amino acid. You can see the very nice turn structure appearing here and digitify the whole entire turn structure and this also. But we also see the happy to see this as the cyclohexane ring is interacting with the protein surface. That means that this hydrophobic molecules contribute to the binding. So the stability of the serum is very important. Now we traced already the other examples that we have is very often we incorporate this cyclic beta amino acid becomes very stable against the human serum proteases. So in this case the F4 shorter version was over 10 days. We don't see any, we don't see more than half of the molecule remained intact. And F3 is a little less stable but still you can have something several days to remain in the serum. So it's really the beta amino acid gave you stability, secure. So it's a better to drug candidate. And very recently, let's see, yeah, this is a beta one. So we recently published, this project was started in the middle of the pandemic. So we wanted to put the gamma amino acid. This is another difficult amino acid that you can place in, but this mirror has a really made effort to make it, these molecules can be incorporated into the peptide chain. And we, because of the middle of the pandemic we built this library and then we took against to the COVID-19 drug target is the main protease. So this target is very hot at the time. And so we last year, but it was took long time to get the paper out. It's almost nature chemistry hold it one year to publish it. So I was very upset about that. But anyway, we finally published. So the molecules I'm gonna just discuss is the GM4. And this molecule has a one gamma amino acid in the near the middle of the sequence. And the inhibitory activity is about the five nanomolar, 50 nanomolar and a five nanomolar KD. And we have some improvement of the mutated by the mutations 10 nanomolar KD, IC50. So this is a non-covalent inhibitor. Any inhibitors known in the literature now is all covalent inhibitor. And this non-covalent inhibitor with a 10 nanomolar range and it is probably the most potent inhibitor ever made. So we had also extra structure. This was collaboration with Oxford University people in the Crisco field. And we have a very good collaborations with them continuously. And this is a structure that we solved. And then this is a peptides three dimensional structure. We have a gamma amino acid sitting here and there. And this is the whole entire structure that I showed you. So this is a gamma amino acid in the sitting and then very close by to the cleavage side of the substrate. And that actually prevent the cleavage because of the gamma amino acid. And then this is the wall head, the getting to the active site and then prohibit the peptide bond cleavage. So it's a really unexpected structure but the gamma amino acid playing a very important role in this. So we have a more improvement of this. Now it's about 100 picomolar IC50 molecules up here. So we're probably going to repeat in the near future. Okay, one quick example before changing a topic. There's a Prixin B1. And this is a very important target molecule involving in a bone homeostasis. And the semaphore D, which is a ligand and the Prixin B1 is a membrane bound protein. And when they inject each other, they form the dimer and then this will send a signal that stop the differentiation. So the osteoblastocytes generally stop the differentiation so that bone formation is stopped. The bone homeostasis means that you actually break the bone, forming a bone, break the bone, the forming a bone. So that means that this bone differentiation stops that basically bone doesn't repair. So we decided to work on the inhibitor of this interaction, protein-protein interaction. Then we found this molecule that which is an allosterically inhibit with the KD of the 3.5 nanomolar gains a very strong binder and then nicely wrap around the part of the protein. And that makes it allosterically inhibit interaction. So we set the collaboration with another scientist in the Tokyo medical dental school. And this group worked on this semaphore D structure. So this is something that which I just talked with you, but that's not the work. This is a different work. But this is one that which I can tell you because we have all the patent disclosed. So this molecule works with the animal model. So this animal model is a female mice which we remove the ovary. And this is very well established animal model that when this type of the animal has a developing a very weak bone. So they actually can't keep maintaining bone strong. So we took this animal model and then inject our molecules over 13 weeks four times and then see how the bone can recover. So I show you here that this is a wild type normal mice that which has a very good dense bond. And this is the mice, the model that ovary was removed. You see the very weak bone inside. But then when you administrate the molecules then you can get to this dose. You basically covered entirely the white the normal mice bone strengths. And this is a little bit maybe over those and it makes it even stronger bond. So it's really interesting that this type of the smaller molecule you inject you can change the strength of the bone. Okay, so I'm gonna skip explanation of the cyclosporine A that of which I told you many different modifications but in natural product there are some molecules that which we can't synthesize by genetic codally programming. Particularly for instance this dehydral alanine and oxazole or thiazole these are difficult for us to do by the incorporation of the genetic codally programming. So we became interested in how we're gonna install this type of the molecules into the peptide chain. And this particular lactazole A natural product has a pyridine ring to cyclize the macro cycle. So it's a very unique molecule. So how this does work is actually using enzyme setting thrown in cysteine enzyme modification takes place to cyclize the backbone to form the oxazole, oxazoline and oxazole eventually. And then the other one is to simply eliminate water to form the dihydral alanine. So this is a great enzymes enzymes that remove the water in the water. So it's really nice enzymes. So we look at the biosynthesis of the cyclic molecules that is called lactazole A. So called the thiopeptide, thiazole containing peptides. So this molecule, lactazole A, that we studied all these bunch of different homologues and the lactazole A is a little unique so we decided to work on this lactazole A. So there are two people involved, Yuki Goto and Alex Bikinotofi in my group. And the collaboration with Onaka's group in the original person that we discovered lactazole A. So we took all enzymes, genes and express and then purify and then we tried to reconstitute all entire biosynthesis machinery in vitro. So the goal is that we wanted to synthesize this natural product in one tube, in one part and within a day. So that's our goal. We always set the goal, something that we have to achieve. So I'm going to step-wise explain this. So the biosynthesis takes place from the DNA to transcribe the messenger RNA and then translate to form this long chain of peptide. This has an extra sequence called leader sequence. This sequence is important to recruit first enzyme to start modifying the core sequences. And the second step is the following. It's a very important part. Las D, las E, two enzymes, depending on what the cysteine or selenase and then it's cyclized back bone to the azurine and then las F is oxidative enzyme to make it las O. And then the next step is just acylation occurs to the serine residues and then las F, which is the same enzyme as this, will eliminate to the dihydroalanine. And then last step is a very interesting for chemist. The two dihydroalanine, one of them are totimerized and having a two, four cycle addition takes place to give you pyridine ring. But at the same time, they cleave this bond. So this cleave the bond and then you form the macro cycle and get rid of this leader sequence away. So we decided to perform everything this in one part. So we took every single enzyme to study how long we have to incubate, how long we have to do what the conditions you want to do and then we figured out everything. And at the end, we have done with this kind of conditions, the 16 hours incubation and the six hours incubation later with two sets of the different enzymes. And that gives you the exact natural product in one part, okay? So this is a probably ultimate natural product synthesis in a day and so we can achieve that. But that's not only our purposes. We wanted to change the sequence. We want to change the structure of the natural product. So in this case, we have tried to scan everything alanine to define which residues are critical for making a macro cycle to complete the natural product like molecule. And we found that any peaks here, the form, the product, you're missing a peak that's critical. So we go through all these things and then you're missing some of the part in here. So that's the answer that's one here and then there's four other residues are important. And then that tells you that you can do a macro cycleization. Okay, so now we are interested in seeing that ring sizes, because we can change the ring sizes. It's a very interesting. So we try to do that. So we have different ring sizes we all go through. The fact is most cases that we can still cyclize it. So the next slide is a summary of what we get. So this ring size is different and different sequences as well. So you see here the very small sizes like this. So this is one of the large sizes that we made. And this is a typical size of the ring size. So in this case, we can actually see that a variety of the sequences, variety of the ring sizes can be synthesized by the constituted biosynthesis machinery. In vivo, in the cellular system, you can't do this very well because you don't know how stable they are and so on. But this is all possible in vitro system. So you can combine together with genetic code reprogramming. You can create a completely different natural product. So now we want to do a selection. So it's basically we want to screen a new activity based on what we get. So the system you can probably imagine that you can combine together with the mRNA display system. So now instead of the sequence that we made, this is going to be random sequence. So we made a random sequence and we tried to do the selection. So this case that this is the kind of the design. I don't go through this in the steps. This is a little bit time consuming to do that. I'm going to skip this, but anyway, that we can do the selections very similar to what we did in the previous three. And then this time we actually made a target with the TINIC, which is intracelular target. This is a kinase that downstream of the VINTO signal. So it's related to cancer development. We had quite a few candidates that we against the TINIC, but one of them TIP 50, TP 15 is the very potent molecule. This is single digit nanomolar KDE, 14 nanomolar YC50. And we found extremely selective to the TINIC. There is another one, but this is one that which is exactly the same catalytic site. So it's definitely the inhibitor. So we're happy to see this selectivity and then we took the X-ray structure. So the X-ray structure tells us that the peptide region is really binding into the substrate region and then very well to inhibit the kinase. And there are some specific interaction that we are monitoring. And here is a very interesting. This pyridine-link-flat structure also makes the interaction with the aromatic cluster in the enzyme. So it's already unexpected, but this is the kind of things that we expected to obtain by a trillion different sequences in the library. Nowadays, we are very interested in cell activity because this is a remaining challenge for the peptide people to work on whether or not this molecule can go in a cell and inhibit the cell activity. And this is the case that we're looking at the membrane permeability. It's called Kappa assay, one of the assay, the Chloroquine Penetration Assay developed by Yoshikuritsu in the Taft University. And we monitor this about 10 nanomolar or five nanomolar, I'm sorry, five micro-molar penetration activity, which is the same as TAT. And we have a molecule that now it's getting a better membrane permeable molecules, but now today I can't tell you in a minute today. But at least we see the intracellular activity. This is what we're monitoring downstream of the signaling cascade and we are able to see inhibitory activity. So again, this is a specific inhibitor. And so that's the comparing with the non-specific small molecule. Of course, they are really potent molecules, but they're non-specific. So we're very happy to see that it's not too far away from what the small molecule inhibitors are. So the final topic I tell you about the lasso grafting. This is a really big collaboration between my group and the Osaka University Junji Takagi's group. And very recently that we have a very good collaboration with Matsumoto's group for a long time, but it is also he's really working on together with us for a new Neobiologics project. So briefly, this repeat, so I won't talk about anything details, but difference what I told you is we have a lot of different interesting amino acids, non-standard amino acid incorporation in the middle. But if you don't do that, you're just a normal peptide. And it just, as far as we can cyclize it, this also gave you a very good binder using a rapid selection system. So basically macrocycler and then you have a peptide chain that is all natural amino acid. It's still very potent binder. So you can, I showed you one example, this is a peptide against the met and this will get you, I don't go to details of this, but it's forming the induction of the dimerization, then you can send the signal to cell growth and differentiation. So this is the molecules that we isolated, the three, four peptides and the three peptides we studied a lot. These are all a single digit nanomolar peptide and then we link it together with the dimers. So the idea itself is if you make this peptide to incubate with the met, met can be dimerized so you can send the signal downstream. The fact is the following, it's like we can send all signal downstream. It's very, activity is very simple to HGF, which is natural protein ligand. So selectivity is also same, HGF specifically activate the met, our peptide dimers also able to select, selectively activate met and then we show that the phenotypic behavior is exactly the same. So you can use the RP tech, this particular cell lines that you can use to form those tube-like structure. So we're very excited this, but then is we stop developing further because this is really you can't do the human application. So we came up new ideas about the raster grafting and the reason we wanted to do is we wanna completely cover the rest of the molecular weight size by this new method. So the molecule which I told you we have all these extra structure solved in the macro cycle with all natural amino acid. And we, I show you the six examples. Now all shows a very different structure, beta sheet like structure, alpha helical structure, or completely cycle of structure. And, but only the thing is the same commonly shared is this red bond. The between the red atoms to atoms. So this is a thioethobond. So thioethobond is always exposed to the solvent, which makes sense because this is connecting to the messenger RNA. So basically the peptide bind to use this middle part, not the thioethobond part. But we know that if you remove the thioether, we lose the activity by 100 fold. So virtually inactive. We consider one micromolar is inactive. So if you don't get one narrow range, it's inactive. So this start out to be a not appreciable activity. However, if we can place in a part of the loop of the protein, now you mimic macro cycle because the protein can fold it. And the peptide itself is very is going to be fold by themselves as far as the end is close each other. Close each other, proximity. So Takagi came up idea that can we do this experiment? Then I said, well, interesting, let's do it. And so we inserted our pharmacophore peptide sequence in the genome. So this is a messenger RNA that the containing the loop replaced with our macro cycle peptide pharmacophore sequence. So it's no longer macro cycle, but it's a loop. So we encoded everything in the messenger RNA expressed. So this idea, we do rapid selection using my technology. And once you know the sequence, the water you want to synthesize, we usually go to the chemical synthesis to study it, but instead take this information of the sequence and then you encoding into the messenger RNA and then you express protein and then you see whether or not it can bind to the target. So as an example, I show you here is a FC domain of the antibody. So antibody, you know that this FAB fragment of the CDR region is responsible for binder. However, FC domains are also very important because this is the molecule that which allows you to circulate the antibody. So the molecule itself has a role to bind to the FC RNA. And so you can actually have a longer pharmacokinetics. So this is a very good molecule, but doesn't have any binding ability to the different targets. However, this has a loops. So now you're actually ready to insert pharmacophore molecule that which we use our rapid system to find that new peptide, de novo peptide sequence targeting to the particular protein and then you insert it to this FC domain. So you can have, you have a six loops so you can actually place it where you want and you don't need to restrict it by yourself. It's a single molecule, it's a different kind of molecules that can be inserted. So this is a three dimensional view. It's a topology is a little differences. It's always dimetic. So you have a two different, three different topology if you look at these six loops location. So this example that I told you about the met interaction of the peptide and you dimerized it and then it actually made a active for activating a met. So for the FC, of course, doesn't bind to the met. So it doesn't really do any binding to the met. However, once you grafted to the loop, now this FC becomes a binder to the met. Now all the location that we made in this bottom part all become binder. But binding does not necessarily gives you the agonistic activity because sometimes important how the molecule can assemble together. So we made all combinations. So in the case, I don't explain all of them but this site was one of the most active site B3 and B1, B2, B3, and then these are kind of things that are also different locations. But all these peptide sequence makes it the agonistic activity. So we're very excited to see particularly with this particular peptide sequence with the B3 site and makes a really nice agonistic activity. So the selectivity and downstream signalling it's all rely on the peptide sequence. So we see the exact to the same behavior. So regardless this FC or peptide dimer, it does work. The reason we became interested in doing this way is the circulation, pharmacokinetics. So now it's the HGF natural ligand is in a clinical trial. But this is only for a spinal cord injury. So you have to topologically inject and then cure the cells to regenerate. But you can't inject it for the bloodstream because it's too unstable. So you see this example, if you put the HGF inject immediately it's gone. Because the instability of the proteins, basically you lose everything. So you can't really reach the patient, the tissues by the injection. You have to go to topologically directly injected to the region. On the other hand, you see here our FC of course, but the grafted FC, we call Mirabody, this will stay in the body without any problem. So this is over 200 hours, 300 hours. The exact to the same behaviors as FC molecules like antibody, it stays long, right? 10 days still we see the effective concentration of the met can be detained. So it's a very long pharmacokinetics life. So we're looking at the model that humanized the liver was prepared to this transgenic mice and then we injected to see that we can activate the liver by this Mirabody. And so you see here that we're looking at the messenger RNA expression level and then you see all of the cases as you see here that the control versus the Mirabody that makes it activation of the liver. So this enables us to inject to activate the liver. HGF is actually expressed from the liver, but of course there are the diseases like Nash, for example, that's what we really working on for this as a drug. So you try to repair the liver by the molecule to inject. So this is one thing which we are trying to do, but another thing is we are interested in delivering this molecules into the brain. The reason is that there is Alzheimer's disease and ARS and MS and many things that people have suggested. Maybe if you can bring this activations into the brain, maybe you can cure certain diseases in the brain, but there is a not effective way to deliver the molecules to the brain. So we decided to work on this way. Maybe next slide. This is the way. So we made a molecules that FAB molecules that which has the binding to transferrin receptor. So this is what we call adobody that antibody plus antibody against the transferrin receptor and then we inserted the met molecules in here. So now we're expecting that molecule like this is a bispecific antibody. It binds to the met, binds to the transferrin. So we inject it and then eventually gets to the, this we're looking at the interactions with the met and also transferrin. So this is the experiment we're looking at how much of the molecule went to the brain and in the turn out to be this one arm antibody, adobody, it gave a really nice accumulation into the brain and the delivery. So it's less than much amount of the molecules that you retain in the bloodstream and it goes to the brain more effectively. So this concept, LASO graft can apply to the variety of different proteins, regardless of proteins that we demonstrated all these type of the molecules, protein molecules can be utilized as a scaffold and then we graft into it. So this is one of the published data and then plus unpublished data. So this is the one that we published. We use this concept to the applied to the AAV. So it's the beaker to deliver the gene. And so we demonstrate that in this case is one of the loop we inserted targeting to maybe this is a good ones, the met. And this is a different Plexin B1 that which I mentioned for the bone. And you generated the molecules that AAV to specifically deliver the genes to the cells that express target molecule. So we don't touch the any cells but it's only the cells that expressing target protein. So this was utilized. So this is a little bit unpublished data but this is data that I showed you for delivering the genes to the brain more specifically. So that we took AAV nine, wild type. This is a control experiment that we demonstrated day seven, day four T. You see that the delivery is brains almost nothing but it's a very well-delivered to the body. This is a literature known mutant of the AAV which reported as a very good delivery system to the brain. So we repeat the experiment. Yes, it does go slowly over 14 days but we actually did another target than the receptor. I can't really leave the target but the grafted new AAV which makes it better the gene delivery. If you combine together with this type of things now becomes very cleaner the delivery to the brain. So this method has a lot of potential. I can't really tell you too much about it but it is really a lot of potential to design a vehicle to deliver specific tissues. So this again that's a grafted technology. It's very easy to work on and it's a plug and a pray and so I formed the company about five years ago maybe six years ago, Mira Biologics and this is Mira comes from Miracle and Mirai in the future of Japanese and we have been working together with the big farmers as well as the bio ourselves. And I would like to thank to the attention and the people that my group is a really fantastic group of the people as international as this is school my group is really made a lot of international people from Europe, from Australia, from America, from Canada and all a bunch of different places, Australia. And so it's really a great number of the people and I have so many collaborators as well and thank you for your attention. Thank you for the wonderful lecture. Any comment or question from the audience? Thank you for impressive talks. So I was wondering the binding model of the selected peptides, so you showed us some of the structures is I think it's generally to bind to the active site or of enzymes, but there are any example to the like aroesthetically inhibition or something? Yes, so I show you the Plexin B1 that's not active site, that's aroesthetic site. And the met is also probably aroesthetic site. The experience we have, if you take enzyme it goes to the active site because enzyme will design it to bind it to the peptide often. And so usually when you have the site that are available for peptide, it goes to there. But if you take a membrane protein, they don't have a particular binding site, they have a surface to interact with each other, right? So then you actually don't have a particular site. You have a lot of different peptide binding in different sites. So depending on the target, we are not designing a peptide, right? We are designing a library, but we are not designing peptide so that you don't restrict it to where it goes. So for enzymes, so when the selection steps is just to select the peptide to the binding, so after that you will test it to the... Yeah, but as I told you, if you take enzyme, 100% you get enzyme binding to the active site. Interesting. Thank you. Can you talk a little bit more about how you modified the A-Vine? I mean just general strategy. I think I missed some key details in this slide. Sorry that I skipped. So the sites that which are available, S1, S2 loops, I'm sorry that I really quick go through this, I'm sorry. So S1 site, S2 site is a possible site, but we took S2 site because once you change the S1 to something arbitrary sequence, it won't change anything. A little bit weaker infectious ability, but virtually the same. But when you touch the S2, completely lost. That means the S2 is a very important to interact. I don't know what the targets are, but it does important for infectious ability. So if you took S2, basically you lose, let's say I inserted, this case is we inserted the peptide binding to the Plexin B1, S2 site. And if it's a cells that doesn't have the Plexin B1, it won't infect because there's no ways to bind, right? But then you insert it to the Plexin B1 peptide into the S2, start to infecting two Plexin B1 cells. So same thing you can demonstrate it here. So the one that which, so this is actually a particular example is the background zone, so zero. So you see a really huge enhancement of the infectious ability. But in this case, we xx target, so that's, I won't say the target, but if you grafted this to the peptide sequence to bind to the this xx target in the brain, these are able to make it better delivery. This one is a little more complicated ways to design it, this to yyy2. And so it's a really important, have a control that whether or not this is actually, so this one thing is I have to say, this actually goes to there, to the BBB, and then transport it into the brain vision, and then you have to infect it, right? But this has also ability to comes back. So there's a way that which we did of yyy, this has ability to bind to the brain cells. So we go effectively go to the infection over there. So it's transport and infect. The BHBB already has a small peptide modification, you preserve that and you add onto that another modification. Yes, okay. Exactly. Any other question? Thank you for your presentation. I'm interested in actually how efficient are those microcyclic peptides comparing to small molecules in terms of their ADMT characteristics in general, are they less like way less effective or like still are pretty competent with them in this aspect? What effective, was that said again? I'm sorry that I missed that. ADMT characteristics like all these toxicity excretion. How is it? Yeah. How different are they from small molecules, parameters? Like I expect them to be a bit less good in that, but are they still competitive in this? Yeah. So the macro cycle is a very poor, well actually it's a very good selectivity against the target and it doesn't have any off target effect. And small molecule actually very hydrophobic, right? So they have a very weak interactions with off target and that sometimes becomes a very problem. But macro cycle has a good balance of the hydrophilicity and hydrophobicity so that doesn't have any off target interaction. So far that we have done animal studies as well as the big farmers has started already phase one, phase two and they don't have the side effect much. So I think now it's the, to be honest with you, it's a, we had a peptidry informed and then it's really huge impact in the industry. And I think now it's like Merck, Genentech, Novartis, J&J, Liddy, all these company has a technology that which I developed. They actually got sub licensing from peptidry. So they are all using this technology for their own drug screen candidates. They realize that how good it is, how the molecules have trustable, how easy to go through the phase one, phase two, but it's probably phase three that becomes the challenging a little bit. But so far so good. So in other words, there are more data accumulating and that really helps a lot. And so it happens really in the last five years and it came down a little cool down, but there is one big success happens recently from the Merck, all of a very peptide. Now it becomes a so big hit now. It's like all the venture capital wants to start the small biotech for focusing on the macro cycle or available drug. So I think this is another big wave that's coming. And so I think when it's coming that these waves and eventually get this drug to be more popular and it's very standard drug development process to occur. And then it's probably we get to know more about such a drug property, the critical drug property. Thank you, Professor Suga. Any other question? If not, thank you. Good.