 Thank you very much and thank you for inviting me to give this short talk. So I'm going to give a talk about a bit similar to Joe's yesterday, but looking into another P-type ATPase that's a bit atypical. So just quickly, we went over this yesterday, but there are three groups of membrane transporters that are named the flipases, flopases, and scramblases, and the names really suggest what they do. So the scramblases are bidirectional and the lipids between the leaflets and have a very broad specificity. The flopases and the flipases are ATP dependent, where the flopases translocate towards the exocytostolic leaflets, whereas the flipases transport towards the cytostolic leaflet. They both have high substrate specificity and the flipases are mainly P4 ATPases, which is what I'm going to focus on today. So yeah, the P4 ATPase is a P-type ATPase subfamily, and as Joe showed you yesterday with the beautiful structures, it mostly functions as this heterodimer with a P4 ATPase and a CDC50 accessory subunit that together make the lipid flipase. They have very high substrate specificity and they are very often auto inhibited, which makes them a bit challenging to both purify and study. We have 14 mammalian P4 ATPases and five encyclomuses Cerviziae, which is the organism I'm working in primarily. So I said that most of them function was a CDC50 subunit, but some don't, and these are the ones I'm focusing on. So I work mainly with the yeast nir1, which should look something like this. It has no interaction with the CDC50 or any other identified subunits, and in general the CDC50 independent subgroup, they are linked to flipase activity, but it hasn't been demonstrated clearly yet. So they are a bit of a mystery so far. Specifically for nir1, it's essential in its host organism, which is why knockout studies are a bit tricky. So we can't really find out anything that way. They are located in the Golgi network and the early endosomes, and they have been linked extensively to this endosomal remodeling complex with mon2, r1, and top1 to name a few. And we have some indirect published evidence of PE translocation, but nothing clear yet, whether this is because it has another function or because we have testing the different wrong substrates or we have some kind of auto inhibition, we don't know yet. But we are working on it. So after all the talks yesterday, I looked back into my very, very early purification trials, also after hearing how popular the HISTAC still is, just to show that this can not necessarily work for all projects. So when I started around five years ago on this project, we had a HISTAC cleavable by Flonbin. But as you can see over here, I had no specific defined band but more like a smear. After switching out the protease, it turned out that Flonbin was very trigger happy with NINIA1, likely also because it was very unstable at that point. But even when switching the protease out, we could never really get rid of this HISTAC protein and endogenous yeast protein that kept co-purifying for no physiological relevant reason that we can identify at least. So instead, now five years later, I think, we switched to the expression strategy that Joe also mentioned yesterday. So we have NINIA1 with the biosin acceptor domain that can be cleaved by TEV in this PYED P60 vector that we got from our collaborator Guillaume Benoît at Paris-Seclée. So we express in cyber-mesosavizia, Delta-Pep4 cells in rich media and can do a double induction. Afterwards, we harvest the cells and we open them in this beat beater, which I normally described as a glorified blender, which is essentially what it is, by just blending the cells with a lot of beats in cycles to not overheat the protein too much. Afterwards, we isolate the target membranes, which are what we call the P3 membranes, since the protein is mainly in the early endosomes and Golgi. As for the purification, this has been through a lot of iterations, trying to get a stable protein, but basically we solubilize in a large excess of TDM and we bind it through the striptavidin beats by the bad tag. Afterwards, we do a very neat detergent exchange with a large volume, column volume washes, and we also do a 500 millimolar sodium chloride wash to get rid of any of the few contaminants that are of endogenous biotinylated proteins. Due to the specificity of the bad tag, we have to cleave it off the beats, so we do that overnight with Tev. And just to show you, this is an expensive approach for sure, but we also get after the cleavage, the eluded protein is basically just my protein and Tev. So it's super pure, something that I would never be able to accomplish with the HISTAC. Then we concentrate a lot and I'll get back to why this is interesting, and we inject it on a standard SuperDex 200, and I manually collect the peaks for cryoEM collection. As you can see in this slightly overloaded gel, the protein is super pure after size exclusion, and when screening grids, it also looks very nice. We can see some helical indications. At this point, we get about a quarter to half a milligram per liter of culture, which is a lot compared to what we started out with, so we're really happy about this and where the protein has gone. So this chromatogram is actually from something I did very recently. It's the second run, and I'll show you what the first run looked like. So this is my first run, very recent, and it caused me to look into LMNG a bit because it's not as easy to work with as you would think because the aggregation number, so the number of detergent molecules in the micelle actually changes dramatically based on the LMNG concentration. This was all published in a very nice article by Christine Ebel's group a few years ago. So just as an example, so I concentrate my protein 20-fold prior to SAC, so that means I have 10 CMC LMNG present in my elution buffer, so I go from having 0.1 millimolar to 2 millimolar. So that means I go from having an aggregation number of 63, where I have 1.2 micelles per protein, to actually only having half a micelle per protein with an aggregation number of 160. So this pattern I've seen it many times, and the first one here is a dimer, and the second one is my monomer, but based on the aggregation number of LMNG and the available micelles, it's, I mean, it makes total sense that the protein shows or make these dimers. But as an experiment, I concentrated the double peak, and I incubated it with LMNG for a target of 1.5 micelles per protein to have a small excess, and incubated that 30 minutes at 20 degrees and reran it on sec. And as you can see, it's a very dynamic process, so I can actually resolve most of my dimers this way. And it also shows that just even though I thought I had plenty of LMNG present, this three-fold increase in the aggregation number really plays a role when playing with this detergent, that I'm sure we have many people using due to its very low CMC. Yes. And that was a bit of a small snapshot from what I'm doing. I would like to thank my two supervisors, Paul Nissen and Joe Lyons, and we have a small P4-ATPA group that share successes and small frustrations, so especially thanks to previous member, Jacob Ullstrup, who really introduced me to the project, and I took over from him and through me, Lina and Tipo. And thank you very much for listening. Thank you for this interesting talk, Lina. I really enjoyed how you actually removed those double peaks in your size exclusion. Yeah, I was surprised it was that easy, but it's nice sometimes the science can be easy. Yeah. It was quite interesting. So I think we don't have any questions so far. So I'll go with the next talk, which is by Julie Ticker. Sorry, Swati. There's one question, I think, for Lina, before we move on. I could not see it. Oh, it just came, sorry, sorry, Lina. It just came to my, something happened. I'm really sorry about it. So someone asked, like, can you use LMNG below its CMC? So that's a great question. Yes, for size exclusion steps, we have, so for size exclusion, we normally have around five CMCs present, but multiple members of a group has used around three CMCs. I don't know of anyone who used lower than that, but in principle, LMNG should be relatively sticky detergent, so it could be possible. So we have another question. Can you provide the reference for LMNG paper? Yeah, I'll just post it in the chat in a second. And as for Bruno asking the galactode concentration I use, I think it's 2%, but I'll have to check back on that. But we see a clear increase in the yield when we do the double induction instead of just a single induction. Okay, so thanks once again, Lina.