Controlling the Machinery of Life with Synthetic Photoswitches | Dirk Trauner, NYU

12:18PM Jul 19, 2021

Speakers:

Keywords:

receptor

molecule

glutamate

photos

pharmacology

proteins

molecular machines

instance

synthetic

nanometers

nice

tethered

membrane

glutamate receptor

ion channel

light

insulin receptor

control

push

electrophile

Hello, everyone, welcome to the latest in our monthly seminars that we have here in the foresight molecular machines group. So for those of you who don't know me, I'm James Cooper. So I'm the chair of this group. And for those watching on YouTube later on also, welcome. So this evening, we have one speaker, which is Professor dog trainer from NYU, who is going to discuss controlling the machinery of life for photo pharmacology. So the floor is yours. Well,

thank you so much, James, for your kind introduction. And thanks to the foresight team for inviting me. And thanks to all of you who are attending at what I would consider past my bedtime for some of you. So stay with me, I hope I'll keep you awake, with my take on molecular machinery, because I have a sort of fascination as all of us with the molecular machinery of life. And My take is that instead of building molecular machinery from scratch, which of course has made incredible advances in recent years and decades, I would rather hijack the evolved over billions of years molecular machinery of life and endow it with new functional properties put on new ignition elements, control elements, which is to be able to turn on and off and shape its its performance. And we have done this mostly with light, I should say that we are also interested in using magnetic fields, for instance, to do this in ultrasound, but this is not the topic of today's lecture of this talk, rather. And please interrupt me at any point to sort of Formula a bit off track. What I want to tell you about today is the use of synthetic photoswitches and their merger with some of the best studies and best understood molecular machines in sales in order to in order to control cell functions. But of course, we're also interested in the individual molecules themselves. What I like about this approach is that it's relatively easy to address these machines, right, because they're kind of wired up for expression in the male brain. And then you can interrogate them for instance, with electrophysiology and other methods. Let me see what I can advance my slides here. Yes. So here is a sample of the machinery we have been working on. We started our work some 1518 years ago, and I actually with voltage gated ion channels, which of course make these wonderful movements across the membrane in response to changes in voltage opening a gate letting ions flow. And then we have progressed from voltage gated ion channels. Over the years to ligand gated ion channels such as ionotropic glutamate receptors such as pentameric ligand gated ion channels, for instance, GABA a receptors nicotinic acetylcholine receptors. Today, I will tell you quite a bit about G protein coupled receptors, because they have been a wonderful platform for photo pharmacology where we put synthetic photos which is on I won't tell you much about arguably the best studied, most famous, at least, molecular machine at present phase some years ago, we published a small molecule photoswitchable molecule that functions like a photoswitchable range, and was able to control the f1 fO atpase with light in that way. But we've only done biochemical studies and the sort of mitochondrial targeting versions is still not ready for publication. So I want to wait a little bit before I can say show you some Sarah work. But we have also worked on, for instance, on exciter amino acid transporters. And we have actually very nice paper in development with Paul Neeson on zerker on the sarco, endoplasmic reticulum, calcium atpase, which we can also switch with a photoswitchable version of topical toxic algal control. And then we move over the years a little bit more into Sailele I have really fallen in love with the cytoskeleton with its highly dynamic nature and its fascinating movements, it allows and enables. And this dynamic instability, dynamic of the of acting and tooling for instance, we could control with photos which will molecules and we'll throw a little bit into there about an attempt to control a Kenny's in spindle Kaneez in called egg five with the photo switchover inhibitor, which is not published yet but which we are soon going to publish that allows us to control mitosis with light. And then it is a work on transcription factors, nuclear hormone receptors, a lot of work recently has been done on photoswitchable lipids because these so benzene photos, which is you guessed it, which I'm going to focus on today, they are absolutely a match in heaven with with fatty acids. There are lipophilic molecules that can have cysts, double bonds, if you treat them with light, that's what you'll find in lipids, for reasons and since I can impart this with light, we can do some pretty cool lipid physiology, but also some very nice membrane biophysics with photoswitchable lipids and there's a whole series of papers in there, already published, I can tell you that we have been able now to feed photoswitchable fatty acids to certain cells. And we find that up to 40% of the endoplasmic reticulum of the phosphatidyl

molecules of the endoplasmic reticulum contain these photoswitchable fatty acids. And therefore we can control for instance exocytosis with light in the forthcoming Rick, but today, as I said, really want to focus mostly on ion channels and gpcrs. And for a little bit in on, on Kenny's in, and then maybe, if we get around, also talk about aefi. Like is is these are the proteins that mark our proteins for destruction and we can turn on them we kind of honor of light as well. So, yeah, a little photos which occurred, I think, towards this audience, I don't have to emphasize what is wonderful, but so benzenes from my perspective, the fact that they switch so fast that there is no time to fall into triplet states. And therefore, there is no single oxygen production in biological systems, which actually distinguishes them from other wonderful switches, which are not as useful in biology because you do fall into prepared states occasionally, but a nasal benzene just flips within two Pico seconds and therefore there is no time for that. And as you know, we can greatly vary them some 12 years ago already, we made Hato suitcase of benzenes. So you can change the photophysical properties device to be referral base the beauty greatly you can change the photo stationary states but trying to pull the absorption spectra off this is in the transformer path as much as you can of course, many others named Don here have made amazing and made amazing contributions to that. Also, in terms of sine inversion, Rana haggis is daisies for us were born as well. These are the days of pains and derivatives the more stable feminine amicably in the dark in the beam form, and then you can push them towards these elongated form as opposed to regular or semi regular is amazing. And that of course allows us to pharmacologically sign invert in many cases and I have a whole series of papers in the world published but also in the making the this was really crucially they actually it's a beautiful way to sign in. But the logic often so benzene and rocks, there are some limitations in terms of actual spectra but other than that, they're wonderful focus which is and we have contributed to the synthesis as well to make them as synthetic accessible. Yeah, let's classify for the pharmacology the simplest form of photo pharmacology is some ligand. For instance, ad blocker shown here an ion channel blocker that changes its efficacy with light which for instance in the dark is a good blocker it crams itself into the into the poor, whereas upon photoswitching it becomes this dial Pentagon. This is my cartoni language. Now, a cute star means active Gao Pentagon means no means inactive and an ion slow. And there could be a photoswitchable blocker of ion channel for instance, who have made them and they had actually gone very far they are now evaluated in the company for vision restoration. But much earlier than that the summit earlier that we actually initially tethered the blocker to the channel of interest and that has one big advantage namely that the tethering and core requires some form of bioconjugation and this bioconjugation can put on a genetic control. The simplest form we can put the system in there and you do some system Olamide chemistry, but as you will see, they are much cleaner and more exciting ways to correlate Li attach something to a protein of interest or nearby. And in this very early version from 2004 we tethered an extra cellular block a tetraethylammonium ion to the surface of a voltage gated potassium channel and with that, we could unblock it if the ace of angels in the transformer block it when the ac urbaines was in this form. This is shown here more structurally this was this very simple molecule mulato ca le and that's a nice nomenclature I've become better scenes with acronyms. Malema quaternary ammonium and easily in between. And that could be tailored and engineered since then on the distorted loop of a voltage gated potassium channel and then could reach the poor or you could withdraw it. This was done together with rich Kramer in Berkeley and have to emphasize that this is the second paper in the history of optogenetics. So of course then the channeled options came along to kind of wipe this out. But conceptually, at least historically. This is a very early version of optogenetics. And we could make hippocampal neurons fire. excitable cells fire in a light dependent fashion by controlling these voltage gated machine with an extra light gate. So little bit of history here. It's long, long, long ago. And then over the years, we have refined this idea of tailored for the photo pharmacology, as opposed to simple photo pharmacology that is the parent one, the terror one requires some call it attachment and that typically are always because you can also do affinity labeling, but typically that requires some form of

modification here, which you can put on a genetic control. So, of course, biologists love that word genetic control, in addition to light control. So here's the general scheme ketones came and he has some molecules here for instance, glutamate derivative, which which with which we control ligand gated ion channel, so called Yana tropic euromod receptor with light. These are truly fascinating molecular machines. They this catone. While nice, doesn't do it justice, you have this transmembrane domain, I'm sorry, this is a mistake, they should be TMD transmembrane domain here. And then you have a ligand binding domain, which looks like a clamshell like a venus flytrap it shows down on the neurotransmitter on the ligand on glutamate. And in doing so this is mechanically coupled with the opening of a of a gate through which then ions can flow in this case, so to mean and that, of course, depolarizes, the membrane leads to a firing of an action potential and that's the major excitatory neurotransmitter glutamate when you have it postsynaptic released presynaptic released on the postsynaptic side, it opens these channels. Through this mechanical event I just described them. And we couldn't resist tethering a version of glutamate to the clamshell to this clamshell domain It is called in this world in such a way that the local concentration changes dramatically from well as inaccessible to fitting right into the binding site. And then the clamshell closes and the channel can open. And this is also ancient history now, because this required system ultimate chemistry and you refine this chemistry somewhat, for instance, remain from this Ace of veins in the ASO aniline. To this is a means in which is a push pull system, which has a electron donating subsequent here and sort of neutral more or less neutral subsequent here, slightly eloquent, withdrawing perhaps. And then we've been doing so you can greatly shift the action spectrum from a lambda max of in the UVA range to one in the blue range. And in fact, when we look at the action spectrum, so it is an occurrence plotted as a function of light or darkness, because this thing also turns itself off automatically. Since it's a thermally unstable. One, it's fast relaxing. So benzene, you can see that we have very nice action spectrum. And I like this because this action spectrum pretty much matches the action spectrum of the blue calm in, in the retina. And in trichromatic vision, it's pretty much almost identical actually. And yes, indeed, these things have been used to restore vision, the systems, but not for very long, because something better came along. And there's something better I want to illustrate with the metabotropic glutamate receptor. The metabotropic glutamate receptor now is a G protein coupled receptor. But it's time Eric. And it has also this extra sailor, Venus flytrap we're like ligand binding domains LBD, which is actually quite similar to be one in your topic receptors. This is one of these motifs in nature, that has been picked up again and again and again. This sort of chewing down on it on the ligand has been adapted to a wide variety of, of receptors, ion channels, gpcrs, also receptor tyrosine kinase is transporters. Some ABC transporters have this motif. It's a typical case of Mr. Mason matching evolution, I believe, that leads to giant

leaps in, in functional sophistication, changes at least. So and it's time Eric so there's there's the seven helix transmembrane domain and then the ligand binding domains actually linked together by a dicer offer so it's actually dimeric GPC is a family CJ PCR, if you want to classify it and that shows down on glutamate and then this leads to intracellular signaling, but not flow of ions rather dissociation of a heterotrimeric G protein, but ions to flow in the end, because you also activate correct channels G protein coupled in with rectifier channels and starches is one of the major modulatory metabotropic glutamate metabotropic receptors and we have initially banned some c pharmacology not so interesting. We've done some closely tailored for the pharmacology not so interesting because this system Malamud chemistry works nice on sales, and in some simple tissues perhaps also actually also in the retina, but it is due to the hydrolysis of of the maleimide. Not great chemistry for bioconjugation as shown here, so, this is the bioconjugation. But it turns out that this Malema is of course, not indefinitely stable in serum. But there's some a much better way to do this. And this is what I want to mostly focus on. This is bioconjugation through bioconjugation tags. These are essentially suicidal enzymes, proteins that self label that are engineered to correlate the attach electrophile with the cleanliness and the speed of an answer elven enzymatic reaction. And this electrophile is not advised directive does not react with water or with the firearm. It's a very poor electrophile. But once it's captured by this enzyme, it will lead to suicide or in some labeling, so to speak. And the arguably the most famous ones are the snap tag, and the clip tag in the halo tag, which racked with pain, slick warnings, Article highlights, respectively, and have been greatly engineered and can be fused to all kinds of proteins when nearby place nearby proteins to now allow for the corporate attachment of some tethered molecule. And as opposed to the previous closely tailored version, this terror has to be pretty large and flexible. Because these are not so small, it's a sort of 20 to 24 kilodaltons these domains. And therefore, you don't really change the local concentration much with the photos, which you move the photo switch into the pharmaco for as indicated here. And then you change the efficacy of the pharmaco for but not so much the length of the table and the orientation of the data. Because this is a long polyethylene glycol chain, we call this portlaw for photoswitching orthogonally remotely Terra laggin. You see we got a little bit better with our acronyms too. And it turns out this portal concept is indeed very broadly applicable. We'll see a lot of varieties in a moment. But the important feature really is that the switch is in the cannot pharmaco for all of us compare this with a dog on a leash, where the leash kind of restricts the action radios of the dog and tailors it to an owner of interest, but bioactivities in this nap, right? That's what happens as opposed to free diffusible talk. This by the way was in May 2020. Just to give you an idea about New York look like in these early days of the pandemic was eerily beautiful in a way and we have been able to apply this portal concept now to have a very broad variety of, of receptors. We've done it with ion channel now this is unpublished we've done it with glutamate receptors, we've done it with dopamine receptors we've done it with serotonin receptors, you'll see a little bit about this also unpublished. So family a Chico protein coupled receptors and family c G protein coupled receptors. And here's the metabotropic glutamate receptor version

that will be n terminally extend the ligand binding domain with a snapback or some tech and then we attach the photos which on a leash. And if the photos which gets activated with light it activates first the ligand binding domain and this is mechanically coupled to change an interface here, this dissociation of the heterotrimeric G prothane. And this can be an itchy alpha if for instance in the glutamate receptor metabotropic glutamate receptor inhibitory itchy Beta Gamma then travels and and activates G protein coupled imod rectified potassium channels which also have an inhibitory effects. So, these are mostly inhibitory receptors. And, and that actually worked beautiful a little bit of chemistry. This is the table of glutamate. Here's the photos which here's the glutamate with a specific stereochemistry here configuration router here should be precise. And here we have the long model as long flexible data we have also long around shorter ones. We put this together. You don't need to have click chemistry This is just killed. But of course you can just make an MIT because you make it in vitro anyway. And then you edit and here's the painful guanine and everything that is on this pain so group gets transferred to sustain in this net deck. But only two that says they know rails because this is more or less unreactive also with Google file, and is a spatial electrophile in that way, and we have made one that maximum activates for 380 nanometers with UVA light, which is maybe not so good for in vivo work. But it is very nicely compatible with fluorophores. So as that, and we also made some that is activated with blue light for 60 nanometers maximally. And the only change, let's go back Look carefully here was the removal of one carbon rule here. So this one carbonyl here makes all the difference in terms of the action spectrum. And here's a model of what this looks like. Before rotated, I just want to explain this array to seven helix transmembrane domains. In the meantime, by the way, there is a cryo em structure of, I think Angular five, and a few other emulators. And then he has an adapter. So we'll switch the rich domain, this is just a lever. This is a very stiff adapter mechanical adapter. And here is the clamshell, the two clamshell, sort of back to back when I wrote it, this maybe it is located, you will see better the two clamshells and as the clamshell closes, this lever arm here pulls on the seven helix transmembrane portion, which then changes the interface and kicks out literally the heterotrimeric G protein that is attached here which then travels and does stop stressing signaling. And let's go back here. And in orange I have modeled in the snap text you get the sense how small or larger snap tag is. And in green we have modeled in the tether and the photoswitching here in red This is actually glutamate which was for the crystal structure. So get a good sense for how big this is. This fear the five nanometers fear here we put in the sort of measure and calculate the local concentration, effective concentration of this of this glutamate, which is very high, we are always in this sort of high millimolar range and mid mid to high millimolar range. And this construct was expressed that woody isikoff and pet john Flannery with a viral transfection method with an AV virus in the retina of blind mice. And then Mike Berry, the greatest student

with Udi with night urges you used till I believe he actually measured with multi electrode array recordings, this electrical activity is from a client, right and so this NML was blind. This is a client mouse model. And you can expand the retina, sacrifice it expand the retina and then measure x the sale of field potentials. These are these little sort of ticks here, which we can sum up when you get a sense for electrical activity of the retina at the level of retinal ganglion cells, which are the sort of output neurons that project and it was opticals. And you'll see when we express this, like you are and this portal version and add our molecule, we see very nice inhibition when we radiate before 60 nanometers. So there is clear inhibition. And this happens within one second. But it also happens at much shorter timescales. 500 millisecond ups, let me let me try to make a nice laser point that for some reason, I always lose this 250 milliseconds down all the way down to 25 milliseconds, you see very clean signal. And I like this, because that's kind of the time unit of psychological processing of visual processing in humans, we, if something happens at a rate of 50 hertz, 20 milliseconds, we see this as a continuous picture. The older colleagues here and will remember self refreshing screens red, which were refreshing at a rate of 20 hertz, you could still see a movie on them. And and I don't know what the frame rate now is in the movie theater, but the old days it was about that as well. So that's kind of the unit and I find this very cool because in 25 milliseconds, this means that the photo switch switches, that's a very fast process already said that it's about two Pico seconds. And then the now activated ligand finds the clamshell or the let's call it venus flytrap, they're flat rep closes this leads to dissociation travels all the way down here to signals leads to dissociation or heteromeric, Cipro pain, all this stuff happens that leads to inhibition. And as soon as the light is gone, this kind of spring loaded system pops back into an inactive form and turns itself off and all that then 25 millisecond, which of course, biochemical is an eternity but I still find this cool from a machinery point of view, so to speak, that this is so fast, but it reflects really high low concentration, the likelihood that is the activated form immediately finds its landing set and then as soon as It's relaxing thermally it dissociates very fast. So the key, the key off must be very fast as well, once this gets into the inactive 100x is in the inactive form. Maybe got just lucky, but it turns out to work quite well. Okay, let me take a quick sip from a coffee because I'm talking fast. So how can we increase the effect of this? Well, my student yohannes bo Houghton, who, who pioneered this recommend group, they moved on and refined it and came up with this very nice idea to make branched

tethered portals, we I like to call them the cracking right or the Hydra. Because that further increases the local concentration, maybe let's let me go to that molecule right away. This is one that your hand is made with four of these photoswitchable molecules now on and I always like to compare this to the Hydra, it's almost impossible to beat these down anymore when you are receptor and they're going after you. And that's what you see here right because now they are response upon irradiation is as strong as saturated glutamate. This is one millimolar saturated glutamate. And you see that with this, we can now do it. And you can multiplex it, you can have different photo switches with different light sensitivity, this is all in development, we can add fluoro force to also see where this stuff goes. You can add attainers and on and on. Because this bioconjugation chemistry is very robust and is still compared to proteins, at least the nano bodies or something like that relatively small molecules that easily diffuse through authority system. The other thing that we thought about is how can we do this with native receptors. So one problem you have when you do genetic manipulation, and is that you always express something unnatural, often towards the background of the natural receptor, and they're complicated analysis and controlling these expression levels is not so easy. And, and they have it would be great if you could actually control native receptors with like, going away now from genetic control, because we are thinking about therapy now. And of course, simple photo pharmacology works, right? This is clear that you can do this. But if you want to do it a little bit more sophisticated, you can think about nanobody photos, which conjugates NPCs, which we already have pioneered and more in development, because then you have a very nice long cover and a strong interaction here. You can also place the poet or bioconjugation tag onto a property nearby, we had a paper with woody isikoff last year where we call this map wet all the membrane anchored portal. And this actually works beautifully. Because the lateral diffusion of these systems is fast and the local concentration is still high enough. And you still have the possibility to anchor the pharmacology to a sale that you have defined. So this is a great way to do select the pharmacology with native receptor expression levels. Here's a version of this nanobody one that you have published already, we teamed up with your highness bro hug and George levy. It's in New York. And Josh has made this GFP fused metabotropic glutamate receptor. And that was in nano body, snap tech conjugate already published. So we just had to make this make nanobody sorry, published was the crystal structure of GFP with these nanobots a very good nanobody for GFP and which has n terminal receipt ml extended to make the receipt, snap take on. And although this looks unwieldy and complicated, this work actually beautifully. We could control metabotropic glutamate receptors, in this case lightly modified or GFP tagged with his nanobody update this this has long been published. Alright, let me now dive a little bit deeper into sort of machinery considerations and what I like to call synthetic biology. I already told you what I love about nature's molecular machines is that they're modular and you can mix and match. This is very often the case that a certain motifs that has been functionally use can be combined in different ways to confer new functions. And we decided to do this tool. We decided to turn the human insulin receptor into a photoreceptor but first input glutamate receptor, The Omen insulin receptor is arguably the best or the receptor tyrosine kinase. This binds to its alpha beta subunits here to insulin and then This leads to cross phosphorylation kinase activity. And then you have downstream signaling usually for the MAP kinase pathway. So, classic Of course in signal transduction widely studied how this exactly works, but it's dependent on a conformational change on the exosome. Aside upon upon

insulin binding, and Yanis Bo hog man, Philip Lappe, a Miko in Munich still decided to just cut this whole thing out and replace it with a venus flytrap domain of metabotropic glutamate receptor. And the system rich domain By the way, I'm not showing it here. But this lever the stiff lever was also the was actually very important. And then a fusion protein that now has the glutamate binding site as opposed to the insulin binding site. And amazingly, this actually could be turned on with glutamate, not very well, we used Of course, 200 micro molar one milli molar glutamate, so high concentrations, but they could actually turn on the insulin receptor. And of course, the downstream pathways of the insulin receptor, very well studied. And you can all kinds of antibodies against the phosphorylated proteins, including the receptor itself available, and the liquid studied. And once we established that, that was actually the hard part, I think your hand has made about 12 versions and two of them have worked 1212 constructs. And two of them. Brexit also teamed up with Shin amin on in, in, in Lille. And once we had this down, of course, we knew how to turn this into a photoreceptor because that's the again mix and match. This is our modular system. So we put the snap tag on on the photos which, and lo and behold, this is not human insulin receptor, that can be turned on with light, at least in syllables or sights. But more yet, for sure, but it's good for but as a sort of basic scientist and molecular bio molecular machine aficionado, I think that this is quite interesting. And here the data so, sorry for the information richness, but let me point out that only when we have here in the insulin receptor, we call this lie here the light gate that human insulin receptor, only if this thing is expressed, and the molecule is present, we can do a lot of controls, then we see cross phosphorylation of the receptor itself. And then downstream phosphorylation of akt and downstream phosphorylation of aurkc. Here with fully human insulin receptor, we can also lose with an added receptor tyrosine kinase called met. The same thing this equals for inspiration offers for select Auto phosphorylation with immunoprecipitation of the receptor itself. And then phosphorylation of akt. And of work with this force for antibodies. So it was published together with Sean B. Khan and quality Sue, last year, late last year. So an exercise I would say in synthetic biology. Let me wrap this up, at least this part, and you tell me how much time I've left by telling you a little bit about family hepc rs. These are the somewhat simplest one because they don't have this large extra set of domains. The most famous one is the photoreceptor namely the opskins rhodopsin, which has a call entity that inverse agonist bound to photoswitchable one bound right, but that's the only photoreceptor that and they made an obscene so let me look, the options are the only photoreceptors in the human genome as far as we know. But it's pretty clear now that they're the only ones and we were very interested in imparting the logic of imparting the logic of options to other gpcr such as other receptors metabotropic receptors, acetylcholine receptors, for instance, and opioid receptors, you name it, Lipitor receptors, and the simple ways again, simple sort of pharmacology, we make something acute and then it activates the receptor, or you can tether and again, the huge advantage of tethering is that you can now untangle pharmacology and in unpublished work that's going to come out very soon. We have now finally found a good way to do this with the serotonin receptor five hydroxy tryptamine receptor, five HTR. serotonin is a very complicated neuro transmitter neuromodulator because it has a lot of receptor subtypes. And this is very difficult to untangle. On the other hand, they are usually important in in psychiatry, right five HD to our for instance is the one that responds mostly to LSD. psychedelic effects and they are right now Before from began for depression and anxiety and all kinds of

diseases, but they are really difficult to selectively address with selective pharmacology. However, if you can tailor the pharmacology to your receptor to a subtype of their, whatever 14 subtypes that are, then you will be able to gain more insights in what the exact role of this is an ad can put into the lead control. When it's otherwise silent, it's even more powerful. So we have extended the serotonin receptor five htt er, I should have put this down here. Fact to a five HDR to a the sub tab to a with a snap tag, and we have tethered the photoswitchable version of seratonin to this receptor. And as you can see here, from this calcium traces, we can very nicely turn this receptor on and off, it's written out five ht to a R, that's the correct way to say it with the snapback and painful guanine version of our photoswitchable seratonin. So this is a tailored version out on a leash. And we can very nicely as seen from these calcium prices. These are GQ coupled so you can look for calcium traces. We can see that we can irradiate with creating nanometers to turn this on or off. And here is saturating one micro molar of FHD found one micro molar serotonin. So stay tuned for this. This will probably be put online next week. But they're very excited because again the untangling of serotonin receptors has been a nightmare and has been very difficult. But pay that photoswitches provided a solution. How much more time do I have? You probably got my 10 minutes. Let me let me briefly talk about another unpublished work which really Brooks motors, right. So molecular motors I haven't talked too much about and of course in this audience. This is unforgivable, and some of the most famous motors that can easons. And in terms of pharmacology, or medical importance, one of the most important counties in his limitata counties in egg five, this is the one that pushes the centrosomes apart in mitosis in a phase of mitosis, the centrosomes need to be pushed apart, and then the chromosomes need to be lined up with equitorial plate before they can be pulled apart. And this, this movement is conferred by egg five by the mitotic kinase and egg five, which is a tetra mer and links to this anti parallel microtubules on both sides, and then literally pushes them apart. And therefore the mitotic spindle, the bipolar mitotic spindle forms. If ik five is inhibited or depleted, cells have this phenotype here, these smaller Aster like phenotype, a mono polar phenotype microtubules sort of project from a central, I could say tetramethyl double dimer of two centrosomes. And therefore, it's very clear what happens this is a typical phenotype of of IQ, five inhibition or depletion. And there are some famous molecules such as monistrol is immortal Aster that conferred is and since this happens in dividing cells, of course, this has been explored as a as an anti cancer strategy. Unfortunately, so far, nothing is on the market. But it has certainly been looked into. And we decided to make photoswitchable inhibitors of the metastatic cancers in f5. To complement our work on the optimal control of tubulin. And octane, you may be aware that we've published a couple of papers now on tubulin, and acting with photoswitchable versions of taxall. And of comparators, that in and of just blinking light, and so on, have latrunculin.

But these are the tracks, right. And here are the motors that rock on there on those. And we finally after actually very long systematic search, this was a tough one, we found a molecule that has this property, that is a photoswitchable inhibitor of egg five indices of the age of benzene. This is a classical aser benzenes of 70 nanometers to paint this form, and you can push it back through the Transform. And we initially always do live dead essays just to see whether this works. These are not super informative, but at least to get a sense for whether it's an effect on sales depends on whether you can push them ultimately in the upper poses or not. And you'll see very nicely and this is important for us that the compound is more Active is an inhibitor indices form, then you transform. And then we do me to sell. So I think for racing sale, so I think we see this really beautifully what happens because when we add this molecule and the radiated with 370 nanometers, we push pretty much for almost all of the cells into the CI to M phase of the sales cycle. Whereas if the molecule is in the transform in the dark, it's a normal distribution that you have with PMS or era in the dark upon irradiation. So we can very nicely push this into cell cycle arrest in the G to M phase. And we can also do imaging. And we see here for instance, that with a parent molecule called MD with a number, which doesn't have the photos, which you see these monastrol phenotypes, so this is clearly an EB, five inhibitor rata. With our Acer version, we see that only with the 70 nanometers, we push them into this monastrol phenotype. Whereas in the dark, the chromosomes line up nicely. And the bipolar spindle is one, these are the chromosomes here this white bubbles, and demons, of course, it's just the it's just the control. And together with Henry Hayes, I wrote this lead half an hour ago. So Henry, here's his name, it should be here, who of course, is very prominent in our community. We showed that the gliding essays, that we can also look at it biochemically. So here is the inactive form the Transform, let me actually show this again, the Transform, you'll see that these trashing this early, free mobile, and we capture many of them. Whereas if we go, we go who this is from, we capture very few. And those that are captured or that are actually in the plan are very little active. And there is the statistics, which I'm going to disclose in the in the paper. And I think I'll stop here because I could tell you now about e three ligase is and their fascinating similarity to alien spaceship machines. But that's maybe something for a different day, we can control the degradation of proteins with light for photoswitchable protect that we call for tuck and just leave it at this T the T cell level.

Thanks very much Dirk for that it was like a crash course in biology and synthetic biology, at least for me, there was all the structures and everything that was was going on this question in the chat from Thomas Schroeder. What we've done in the past is had you engage with the chat? Because it seems to be easier than me misunderstanding the question if that's why there's only one and then people can unmute themselves and I might ask a few was as well afterwards,

I'll do this again another one and half years, they'll ask how do I

like can I can just read I will break with tradition. I can just read it to you. So the question was, am I right in understanding that the advantage of remote tethering over close tethering is a matter of being able to selectively address a molecular machine of interest over others in the system via genetic engineering?

Absolutely above closed and remote tethering, I will say, the close tethering, let's forget about it. I'm not doing this anymore. The system ultimate chemistry is yesterday's news. Because the what I call remote tethering actually works well, right. And you either tailor it to the machine of interest. So you can, I don't know, at some point, fuse something with a sub unit of F one F, all right, and then throwing that range in a light dependent fashion. And you can only do this now with versions of that machinery of genetically tack them, or you're doing nearby either with it with a membrane anchored protein that is close. And that has a pretty high local concentration because it has only lateral diffusion. But you can also think about scaffolding proteins for instance, that you use as your docking station for attaching the terror. We haven't shown this Yes, but I think this is something especially in synaptic cleft, one should explore. So and yet, the real advantage is is that we that we can determine which subtype or at least which cell type we are addressing.

I don't know if other people have questions maybe they're still

the glading is is again because it is correctly said that what you want to see again. Is that what you want to see again Yes, please. So attempt Tell us a little bit more about what's happening in these two movies. So this is immobilized egg five, which, which moves tubulin that is recently labeled. These are these gliding aces that are in the community of motor proteins quite common that you immobilize a motor protein. And there's some sophistication in how to do this correctly. It turns out with egg five, this is not so simple, because egg five functions as a tape, right, so there is some investigation and then you like in a mosh pit, you know, you literally move your, your, whoever your rock style on this immobilized one, but you first have to capture them, and then you can move them. And, and what we see on the left side is that, we first of all, we capture a lot of these microtubules, and then they are somewhat mobile. Whereas on the right side, we actually capture very few and those who are captured, the ones that move so fast, actually not captured, but those who get captured, they are, they are not moving very fast. And I have to I have to admit, I have to defer to Henry Hess for the fully professional explanation. That's my understanding. But the statistics in the paper, and, and Henry does this for a living and all this is a layman. That's the gist of it. And you, you stabilize you make these microtubules, but also adding Taxol to sort of to prevent them from depolymerization and eflora. For and then you, you can actually visualize those.

So one of the main questions this group has for most of our speakers is to sort of speculate a little bit on like five, I think it's 515 and 30. Goals inside, maybe it could be in the general field, but I suppose in this it will be your sort of specific area of research, perhaps theme towards know you've discussed a bit about synthetic biology, do you have some idea of where you think work in this field is, is going? Our

medium term goals already talked about possibility to control things with our input signals, then light? Light is great, right? But light doesn't go far? Who is? Well, and if you can do this with magnetic fields, or with ultrasound, then perhaps, I think, is Harold here? I think so Herald, the Herald Herald here, can you can answer this a little bit more, I certainly more farther along with adding additional control elements, I would love to put molecular motors on proteins. So you know, I don't see a reason why one cannot pack a firing a motor on snap pack, and then move things around in a sensible way. Right, then for me, sandable always means they had some interesting biological function. But you know, if you have a speculating here, if you're if you have a membrane anchored snap tag, and you, you make a pencil cone in offering a motor and you pack it on, I'm pretty sure you can move things along the membrane, we'd like. And that would be interesting, right? The main forces have been generated by making gases, right? This is well known. But other than superoxide, dismutase, I don't know too much first generation that has been used in, in, in in biological functions, artificial first generation.

And then one thing that I would ask, I'm sort of I'm a synthetic supermarket chemist, I've background in things that operate in membranes, and that's where my independent career is currently sort of focused. And so I think about problems in this area from a totally synthetic point of view. So my question would be do so your talk today has been focused on integrating synthetic elements with biological constructs? Do you think it's all leading on because do you think that's the real future of molecular machines? Or do you think there's a scenario where we can have a totally synthetic analog of what biology can do?

Absolutely. And it's just not my personal take. I absolutely believe it will appear as the timescale of Kitty Hawk up to Boeing 747 we will see right.

If that if that happens, I'll be 100 before we finish, which sounds about right, so

right. So we will see but, you know, it's, it's clear that we have made already enormous strides. Could you design something as complex as f one f all from scratch with current knowledge? I'm not sure what I again, what I really like it's not so much that As a Oh, nature is 4 billion years in, and how can you compete on the contrary, actually don't believe that is all right, because there are developmental cycles are getting faster and faster. And we're learning more and more. And at some point, we have machine learning in this field, and they already have it. And what I really like about this biological machines is that they are so addressable, because you can link them up with something that you can't address. So the virus the big, big, big problem, in my opinion with synthetic molecular machines is where the wires that you can address them with, and how can you immobilize them in a way that they perform work that is useful and can be multiplied and whatnot. And, and this is all with, with transmembrane proteins, and that's why we started on transmembrane proteins, this is all a given you can you already know how it is oriented, and you know how to visualize it, and you know how to address it. And, and I think that is a that is important to have sort of the supramolecular environment, you have to think about the environment where you place your molecular machines. And this is not done enough in the purely synthetic community, if I may say.

I mean, I think from the from a membrane point of view, it's very rare anyway, just as a general concept, I think that's something that a lot of people in this community are very including myself very interested in, so

that you can do things that membranes capture, you're going to cost and props and things like that. To place stuff. Right. And that is extremely exciting.

So I don't know if there's any other any other questions from anyone in the in the audience? There's certainly not I don't think there are any others in the in the chat. So I don't know if anyone wishes to either just unmute themselves or Hands up. Otherwise we can I feel

I haven't talked enough about the sophistication of the Acer benzenes. Because we actually think a lot about the molecular details of this Acer pens is that just alluded to in the beginning, but we are following this very closely. What can be done with a super engine photos, which is how can you How can you push them into the near infrared, right? We have now so painting photos, which is that you can integrate the liquids that work with 60 nanometers, I just did a paper on lipid nanoparticles that can release that are made up of photoswitchable lipids that you can release cargo with with C six 660 nanometers. So that's molecular or synthetic from them very closely following them. They will be clickable is a benzenes more and more. right this is coming online. Now I first paper is out on using click chemistry and using the modularity of click chemistry to make photoswitches. And I think this is going to be exciting. It's not that I'm not in all of the other photoswitches. But as I've said, we actually worked with some of them aspiro parents, for instance, in in the logical context, there are issues with with bleaching and, and single oxygen productions. So I wonder if I wonder if that's if that's this is the last chance for anyone to ask anything else if they if they want to?

Otherwise, we can wrap up here I think

what I wish I would have more input from is the modeling community, right the molecular because this is kind of getting very strong in molecular machines. And what I showed you as a model is worth just cobbled together if I wish somebody would have more contact with the molecular community. really understand that there are so many equations we don't understand, we don't understand, for instance, whether they photos, which is not written up well enough is referred to switches can holders summarize whilst bound to their receptor will have to come off from the receptor. And I'm not biophysicist enough to systematically Look at that.

As a Yeah, call there for anyone in the audience that might have that level of expertise, I suppose. I think it's something that I certainly would benefit from as well. If that's that's about it, there's nothing else in the chat that I can see. I don't see any raised hands or anyone else wanting to to ask anything. So we've thank you very much again for sharing the work with us, Dirk, it was really exciting and interesting stuff. So we'll call the meeting to a close there. And thanks, everyone, for coming along. I think it's been really interesting today see a bit more biologically Applied Chemistry, which I think we haven't covered quite as much in this or that we've had some in the past. So it'll be the third week of august next month for the next talk. So thanks everyone again, and take care. Bye bye.