 Actually, I'm going to talk about lipid droplets which are very peculiar organelles in the cells because they are an organelle that is not surrounded by a membrane but instead by a phospholipid mono layer and a lot of proteins have to localize to lipid droplets and a lot of them use an athletic helix to target lipid droplets. So what you want to understand is why a protein goes to lipid droplets not to another organelle that's surrounded by a bilayer and so in order to answer this question. So I started looking at this protein family called the perilipins which are really like hallmark mammalian lipid droplet proteins. So there is three perilipins that have been studied a lot. They localize to lipid droplets and all of them they contain in their sequence a predicted amphithelix region. But there's another protein in the family that's really striking. So it's called perilipin 4. It hasn't been studied very much but this one has a predicted amphithelix region of almost 1000 amino acids in the human sequence. So it seemed like a really good candidate to study targeting to lipid droplet by an amphithelix and so not only is it very long but it's also extremely repetitive. So the sequence is composed of these 33 male repeats and in the human version you can identify about 29 repeats and here like I'm showing you a plot. So this is an alignment of the repeats from the human protein. So you can see how in many positions of the repeats you always have the same amino acid and so that when you plot this one repeat on a helical wheel you can see that it could form an amphithelix so it has a hydrophobic side and it has a polar side. And actually it's not a very, so you can see that the hydrophobic side is not very strong so you don't have any large hydrophobic residues. And we can purify these peptides and look at the structure by a CD spectroscopy and when you look at the protein in solution it really has this typical signature unfolded motif. But then when we increase the concentration of lipids in this mixture we get a really nice helical signature and this is over 400 amino acids actually we have gone up to 660 amino acids. So it's really by far the longest consecutive amphithelix, not amphithelix, any helix that we know of. So what you're saying is that it doesn't have a helical structure? No it's completely unfolded in solution and actually this is how we purify it. So then we can ask okay does it go to lipid droplets and so we express this protein in Hila cells fused to M-cherry and you can see that there's a lot of cytosolic signal but then it also surrounds lipid droplets that are here labeled in green and the localization to lipid droplets really depends on the length of the protein. So if we have a short sequence of like 2-33 maripetes or 66 amino acids it doesn't go to lipid droplets but as we increase the length the targeting to lipid droplets becomes really efficient. So then we can also look at the sequence of the helix to see what are the parameters that are important for targeting and as I've told you so it's really not a very hydrophobic helix but it has a lot of free-on-in for example. So the first thing I wanted to ask is like how does hydrophobicity affect targeting? So because if you imagine so this is a helix that should interact with lipids over a long interaction surface. So we don't want to make big mutations but instead we make small mutations and repeat them along the length of the helix. So here for example I'm mutating 3-on-ins into valines or serines that are a little bit more or a little bit less hydrophobic and if you increase the hydrophobicity just by a little bit now the helix starts to localize very efficiently to lipid droplets whereas if you decrease it by mutating 3-on-ins into serine you lose all lipid droplet targeting and then but then another thing happens so okay so if it's more hydrophobic it goes to lipid droplets better but you can see also that the cellular pool of the helix changes and in fact we can see like if it's more hydrophobic it starts to invade very efficiently other cellular membranes. So here in a cell that is expressing the protein more highly you can see very strong endoplasmic reticulum signal but probably goes to all sorts of membranes whereas where you have low level expression you see primarily lipid droplets so from this you can conclude that both length and hydrophobicity improve binding to lipid droplets but in fact if you're more hydrophobic then you lose specificity you become more mischievous for other membranes so there's something about the surface of lipid droplets that make them really sticky so that empathy helix can bind very easily so we wanted to understand what is that so for that we now do experiments in vitro with purified protein so we can label the protein with MBD so that it's fluorescent and the fluorescence depends on whether it's in a hydrophobic environment and test how it binds to bilayer liposomes so we have prepared liposomes of where we vary composition for example we increase the monounsaturation of phospholipids which which interferes with packing of phospholipids we increase the charge we increase curvature we added diacylglycerol and actually the wild-tap empathy helix really doesn't want to bind to bilayers whereas if we use this mutant that has morphine jahobic now we have very promiscuous binding to all sorts of liposomes and in fact there was only one composition that we could find for the liposomes that was efficient to recruit this this empathy helix and this is an artificial acyl chain that you can you can buy like in phospholipid that contain mason groups so you have metal group every four carbons so you can see here so this is this diffetanoil so these metal groups prevent like efficient packing of phospholipids so again you get a surface that is not very well packed and so this surface is good for binding to empathy for binding of our empathy helix but obviously what does that have to do with lipid droplets so one thing that that I've told you about lipid droplets so they don't have a bilayer they have a monolayer so like a bilayer so in bilayer you have the two leaflets coupled whereas here like the monolayer can spread so this phospholipids can spread on the surface and this this will increase the surface tension of the lipid droplet so such lipid droplets are going to become unstable and they will fuse so we thought that maybe this characteristics characteristic of the lipid droplet would be important for targeting life of the empathy helix so for this we do like we decided to do an extreme case experiment so let's imagine that we have only neutral lipids and we don't have any phospholipids what will happen so it's a very crude simple experiment so we have we have a solution of protein we add the droplet of oil we vortex really hard and as you can see as you as you increase the the the concentration of protein in the solution you get after vortexing you get increase in turbidity and if we put this mixture on under electron microscope you can see that small droplets small all droplets have formed and you can also look at them by dynamic light scattering and they have quite uniform size so in the range of a few hundred nanometers and if we use this in the experiment this mutant protein that didn't go to lipid droplets in the cells now we don't see any formation of droplets by dynamic light scattering and also we can look by we can we can label the protein fluorescently and look under fluorescent microscope and you can see in the bigger droplets that you have the protein fluorescent protein very nicely surrounding the the core of the the neutral lipid core and so so so it looks like this empathy helix can in fact replace the phospholipids so it's acting like a like an emulsifier in place of phospholipids so obviously this is a very artificial in vitro experiments so we wanted to see like if there's evidence for any evidence for that in the cell so for this we use them drosophila cells because drosophila has been have been used a lot in screens to determine protein factors that are important for regulating the size or the distribution of lipid droplets and one protein that came out of the screens is this protein called cct1 which is an enzyme that is catalyzing the rate limiting step in the synthesis of of phosphatidylcholine so so the phenotype that you get when you deplete the cct1 is that lipid droplets get bigger and so their explanation for that is that because you don't have enough phospholipids the lipid droplets are fusing so if empathy helix can can do the same thing as a phospholipids that means that would which would be able to rescue the size of the lipid droplets under these conditions and this is indeed what we see in the experiments of here I'm showing the experiment where we deplete cct1 and you can see that in the cells that are not expressing the protein the lipid droplets get very big but when we have the protein expressed the lip the size of the lipid droplets is rescued which you can see quantified here compared to the control experiment so from this we conclude that the spray lip in 4 is an empathy helix that is really optimized for interacting with neutral lipids over a long surface and it can it can act as a coat to form this droplets and so this could be important in the cell under conditions for example we don't when you don't have enough phospholipids you could quickly snip on the protein to stabilize lipid droplets so I like to imagine this protein like as a as a millipid because you have lots of legs and interacts weekly with the surface but actually I'm just going to show you one piece of data that this millipid model is really not very good because so if you go back to the this nutritional analysis so one thing that I didn't talk about is the charge of an empathy helix so in the polar side of the helix you have quite a lot of charge residues so actually throughout the sequence you have the net charge is always plus one so I tried mutating residues to get to reverse the charge or to increase the charge in all cases we decrease the binding to lipid droplets but one strange thing is also that the charge is always asymmetrically distributed which is kind of unusual because if you want to interact with the surface you would want to have charge close to the surface so this is the case that I'm showing you here so so we asked like what will happen if we keep the composition the same but we kind of redistributed the M&S to make them as symmetrical as possible so so in this case we also decreased the binding of the protein to lipid droplets but we can also express these proteins in yeast which I'm very happy about because I'm yeast person by heart so in yeast also it goes to lipid droplets so these targeting mechanisms are very conserved but actually it goes to the plasma membrane quite efficiently and so if we if you compare the targeting of this wild type protein and the protein that has this charge swapped we can we can we can change the ratio between plasma membrane and lipid droplets signal which would correlate with the distribution of charge and so we propose that in this case in the protein actually the charge is mediating inter helical interactions to kind of form of mesh work on the on the surface of the lipid droplets to stabilize them so so I'm I this work was I worked as a researcher in the lab of Kathy Jackson and so they're sorry I have shown you work of three people so Manuel is a PhD student and two engineers have done a lot of the experiments that I have shown you and we have done this work in collaboration with the group of Bruno Antony so all the liposome experiments that I have shown you and I see the spectroscopy has been done by Bruno and Marco Mani and thank you very much for for your attention so and you mentioned the changes in the droplet size as a consequence of either phospholipid metabolism or expression are there changes to the metabolism of neutral lipids in these droplets do you see changes in them in cluster hydrolysis for example are very nonspecific in some way are there is a rate of cholesterol mobile so you mean when we have the infaticulix present we haven't done these experiments we have looked a little bit if the composition would affect targeting and we see some differences but but yeah that we don't know what do you think is the in is functioning the cell of these lipid droplets I mean of I know that there are I think that I know that they like and inside they have many lipids that maybe are sequestered but what is the function of having proteins outside is this like a parking lot for proteins that are supposed to be in membranes or so so they're really important lipid droplets for lipid metabolism so one thing is like if you have too many so for example fatty acids are toxic so you need to store them into triglycerides to get them into and then when you need energy you recruit lipases that degrade the triglycerides and the esters of the some proteins getting between them and membranes in of the proteins who are which are on the lipid so you have a lot of enzymes and then these perlipines for example they don't have enzymatic activity so they have been proposed to act like as a I mean they can recruit other for philippine one for example recruits lipases so they they should regulate in metabolism I notice their tip 47 it's also supposedly has a role in protein trafficking and so is the why is it in both places do I mean why their proteins in both places what are not lipid metabolism well the the the role of tip 47 in trafficking is a little bit under question yeah so it's originally proposed to be involved in trafficking but actually it seems to localize really well on lipid droplets and so all the function correlates with the contradictory maybe yeah it's just as I said parking lot for yeah but actually I mean lipid droplets are really like connected with I mean so they they're connected with the endoplasmic reticulum they have a lot of contact sites with other organelles so they're really like an integral part of the cells you have a lot of trafficking through the lipid droplets to their organelles so maybe I missed it in what you said but do you have a mechanism observed or in mind whereby then the lipid droplets are some say you get rid of the so how do how do you get the protein off yeah we don't know so one thing one speculation would be that like you can imagine by phosphorylation because we know that if when the charge changes like it it it doesn't bind and that it has all these three and in serine so you could imagine that if you phosphorylated you could very quickly take it off but that's completely speculation we have no evidence for that so it's something that's very like under study so there are some so for example there is some there are some some suggestions in literature that you have difference in the composition at least in some cell types you seem to have the difference in the composition of the core of lipid droplets and that some perillipines prefer like cores with more triglycerides the other one with cholesterol esters we have tested this a little bit in yeast and we don't see a difference and then there also you have you have some lipid droplets for example that are in close contact with peroxisomes or like more in contact with endoplasmic reticulus and there seems to be like some speciation but this is not a well understood but it's very studied a lot right so thank you very much