Shadowing Practice: Helge Ewers Seminar - Learn English Speaking with Video

So my name is Helge Ebers.

I'm professor for membrane biochemistry at the Free University in Berlin.

And we are very much interested in membrane compartmentalization, something that we share with our host.

And we have been selected for today's journal club for a recent paper from a finishing grad student from the lab,

Jia Ahui Li, that was published earlier this year

and what she did there was she used tiny magnetic fluorescent particles

and pulled them using magnets over the cell surface

and i would like to share with you now why we did this

and what we think we may be able to do with it

and maybe this would be interesting for you or your work

so uh do we just jump right in yeah

so i mean we are very excited about this paper

so just start whenever you want all right

so uh here obviously this is a fluorescence micrograph of a cell here you can see trajectories of this particle

that we can pull all the way to the edge of the cell

and then it sort of travels around the edge of the cell here

and gets gets stuck and this is already one feature

that i really want to drag your attention to is

that we have very high accuracy right

so we can use the extremely high pointing accuracy of source

and particles together with pulling the molecule into a specific direction thereby we want to learn

something about the places where these molecules will get stuck

or where they stop moving where we come from is we

want to understand how membranes are compartmentalized all of you were

obviously familiar with the singer nicholson model from 1972 where singer

nicholson synthesized a number of recent observations on what plasma membranes may be made of

and consist of and how they may be assembled from lipids

and proteins into these fluent fluids and continuous bilates right

so a membrane protein would be embedded in a mosaic of lipids

and proteins in which these molecules are free to move right

and as an illustration i brought to you here a quantum dot

that is bound to a lipid in a supported membrane bilayer

and you can see that this molecule here is of course absolutely free to move

and just wiggles around in a random motion

through this membrane bilayer okay

so it's diffusing already pretty early after the 70s a number of investigators such as mike sheets

and naki kusumi and ken jacobson and and others

used modern video microscopy to investigate what is happening in membranes

and immediately they found

that in plasma membranes proteins are actually not entirely free to

move indeed the diffusion coefficient is like an order of magnitude

or more slower than would be expected in free lipid bilayers

and they started investigating this

and the most elaborate model here stems from decades of work

in the laboratory of akakusumi who has correlated movies with extremely high

frame rate imaging of of lipids or proteins bound to gold beads

or fluorescent particles and what he and others would observe is

that they would sort of seem to jump between domains in

which they would reside for a certain amount of time and when they did platinum replica EN microscopy of the plasma membrane,

they found that the cortical actin at the membrane would form these little corrals.

And that led to a so -called fence model

that Aki Kusumi has been putting forward on how these corrals

of actin filaments at the membrane basically compartmentalize the membrane

and proteins have a hard time jumping over these fences when they diffuse around in the plasma membrane.

So we now try to understand how this works, if that is really indeed the case, and what forms of membrane compartmentalization exist.

So diffusion barriers, like obstacles to the motion of membrane lipids, are ubiquitous in cell biology.

So the sperm head is divided from the tail by a barrier

that separates membrane proteins from diffusing back and forth there is the tight junction

and epithelia separates the apical from the basolateral membrane the cilial

plasma membrane is indeed a different compartment from the apical membrane so it's a different

uh complement of membrane proteins in here than in the remainder

of the membrane neurons have a diffusion barrier in the initial segment

that separates the axonal membrane proteins from the somato dendritic membrane proteins

and dendritic spines as well compartmentalized allowing only selective access to their head in

which the excitatory synapses sit finally dividing cells also have a diffusion barrier around the cleavage furrow

and that seems to be at least at least in yeast cells even be present in the endoplasmetic reticulum

and the nuclear envelope so the the specific regulation

and physical hindrance of membrane protein motion is a ubiquitous phenomenon in cell biology

and we we try to understand how this works

so in the initial segment here this diffusion barrier is one of the first ones

that has been described

and investigated in great detail first by a paper from the EMBL from Toshiri Kobayashi

and Carlos Dorsey's lab who found that

if they would use a pipette filled with fluorescent lipids

and touch the axon with it they would move through the axon

but they would not move out to the um side to the to the cell body okay

a couple of years later bettina winkler and numing poos lab found that this was likely due

to the soma just being so much larger than the axon

that you just couldn't detect this worsen lipids

and she could not define find a diffusion barrier here only

to a couple of years later uh doing some work with iran melman

and paul forscher a great expert on the normal cytoskeleton

to investigate using optical tweezers pulling membrane proteins through the initial segment

that they would get stuck in the initial segment

and they could not move there okay this was then further

investigated by aki kasumi who could correlate the motion of membrane

proteins to the accumulation of an adapter molecule called anchor

and g and that was published in nature biology in 2003

so uh yeah

so we are interested now in this um diffusion barrier here at the axon initial segment

that separates the somato dendritic domain from the axonal domain

and here what we find is a specific complement of adapter molecules cytoskeletal molecules such as spectrins ion channels

and adhesion molecules that make a specific array

that is about 60 to 90 microns alone

and this separates the external membrane proteins from the dendritic membrane proteins

which makes a lot of sense because of course the somatodendritic part integrates

and receives signals and the external part sends out signals so you don't want to mix that up

now it's long known that we have a diffusion barrier here so we got really excited when in In 2013,

the Xiaoyi Zhuang lab published the fantastic observation that along the initial segment,

100

there would be an array of rings of actin circumfering the axon.

101

So every 200 nanometer, you would have this actin ring that goes around the axonal membrane.

102

Now, for us, of course, this was a fantastic way of testing whether these actin rings

103

we find here in the excellent initial segment could be used

104

as a paradigm to finally test the kosumi model right

105

so they could never really look at the actin at the same time as the proteins

106

because the actin cortex obviously is very plastic so it constantly moves

107

and there's so much actin in the cell

108

that it's very very hard even with modern super resolution methods to get a reasonable resolution of these compartments

109

and in the time scale that is comparable to membrane protein motion

110

so here we now have these rings that are persistent over longer time

111

and we were hopeful to correlate that motion to the

112

that structure to the motion of membrane proteins

113

and we could indeed do this

114

and here you could see you can see a single particle

115

tracking data of lipid anchored gfp gpi gfp coupled to quantum dots

116

and we could localize indeed that these stripes are right between these actin rings okay

117

so the membrane molten protein diffuses between these actin rings

118

and has a hard time in this wicked model of jumping over

119

that actin ring now we are very happy to to publish this

120

and we're really excited to prevent this to kusumi

121

and others as a way of investigating membrane protein motion to find direct evidence for the Kusumi model.

122

But of course, uh we always have to be skeptical right so one idea

123

that is prevalent especially in the initial segment field stems from the observation that where we have these actin rings

124

these actor rings are interspaced by spectrum tetramers okay

125

so these elongated spectrum tetramers are about 190 nanometer long

126

and they have binding sites for a large adapter molecule called anchoring g

127

and this anchoring g recruits a lot of huge ion channels

128

like five nanometer diameter eight nanometer they have a huge ion channels

129

and also adhesion molecules and plenty of them

130

so what we actually indeed have in between these actin rings

131

is kind of like a forest of all these ion channels okay

132

so when you talk to initial segment people they say well

133

i'm not so sure that your membrane protein actually bumps into the actin rings maybe it's rather

134

that it sort of gets stuck between all these ion channels and it's just not moving out

135

so um we like uh complicated experiments i have to confess we like we just like really

136

things that are complicated but are to the point so what we thought we should do

137

is we should try and image a single membrane protein

138

and then pull it across this area okay

139

so now we would either just pull it through the membrane

140

and it would kind of get stuck in between all these ion channels

141

and then we could pull it again or

142

when we pull it it will bump into the actin

143

and then we can pull it through to the next ring

144

and it will bump into and we can pull it on okay

145

so and by combining a means to pull the membrane protein with

146

very high resolution single particle tracking we were hopeful that we could achieve this

147

so optical tweezers are you know expensive to set up

148

and we didn't have them so we looked into much simpler means than

149

that and we decided to go for a magnetic nanoparticles okay

150

so we went through quite a few i have to say

151

not all magnetic fluorescent particles you can buy are really very

152

useful and nice so it was really um several months of work just to figure out

153

which ones work well and which ones are mono dispersed

154

and bind without background and you can still move them

155

and they are bright enough and so on

156

but we found some where we have an iron core

157

that is surrounded by a starch shell and

158

that is then coupled to striptavidin and

159

so we can then bind it to biotimulated nanobodies

160

or biotimulated lipids on supported membrane bilays okay

161

so we can throw these nanoparticles into um light scattering uh device

162

and we can see they're nicely in mono dispersed okay

163

so we have a single peak around 200 nanometer in size

164

um this is pretty much in agreement with the electron microscopy images we get of these iron cores

165

so they the real particles will be a little bigger

166

because there's a starch shell around it

167

which we can't see i like this one it looks like a gummy bear a little bit

168

if you like um so hopefully uh this is going to be a little more round

169

when it then finally binds to the cell okay

170

so we then made supported membrane bilers in these uh cover glass bottom dishes

171

and we were thankful to get from friends a micro manipulator we could use to deliver a tiny magnet

172

and attached to the magnet we have a metal wire

173

that we pulled to be very very um uh uh thin at the tip okay

174

so basically we have very strong neodyne magnet and a little wire

175

that you can't see here it's so fine

176

and then we dip it into the um solution where our cell sits

177

and then we try to do microscopy okay so this is basically the setup you have your objective

178

you you have you can bring it into less than 800 micrometer distance and then your magnet

179

is parallel to the cover glass in a couple of microns above the cover so we can't really tell

180

we just kind of try to almost touch it and then move the tip here you can see the tip

181

towards the cell this is the tip of the wire right so the wire is here attached to the magnet

182

this is what it then looks like here you can see this is a supported membrane bilayer

183

we can now pull a couple of these particles you can see some of them are clearly stuck

184

some of them don't really care so much about the magnet

185

but a couple of them really clearly move towards the needle

186

that is to the right here okay we can localize these

187

particles very accurately you can see a couple of hundred subsequent

188

localizations of a single particle as you can see it's just

189

a few nanometers error we can achieve using these particles okay

190

what we of course did was we looked at hundreds of trajectories of particles on supported membrane bilayers

191

and then quantified the force they experienced according to the distance

192

so the closer you go with the magnet the higher the

193

force gets now the force is not extremely high we're talking femton newton here

194

but this is of course good enough to pull it through water

195

and also through the membrane bilayer

196

but you would have a hard time competing against strong cellular forces

197

and this is exactly what we want we don't want to

198

disrupt anything in the cell we just basically want to move it

199

and see if it experiences some kind of drag in some parts of the membrane.

200

So if we go closer we can really see that according to the distance it really rises up to about one femtonut.

201

So this is before there's no apparent force they just wiggle

202

around in x as well as y direction it doesn't really matter

203

but when we add the magnet we have a clear the femtonewton force in X direction, in this direction.

204

So to test how accurately we can measure the motion of the particle, we looked at some defects in our membrane.

205

So you can see a fluorescent supported membrane bilayer.

206

This dark area is a patch on the cover glass where bilayer could not be formed.

207

And we could see when we now pull these particles, they move

208

towards the edge of the bilia and then they move along

209

that bilia and you can see already here

210

that from the trajectory you can see these very well defined little areas of membrane

211

and if we now zoom into this little trajectory here you can see

212

that we have very fine detail and a pretty straight line okay

213

so this gives you already a little bit of a hint

214

and we're going to further look into this one here this

215

gives you a little bit of a hint as to how

216

uh how accurately we can localize our particles here this is two microns

217

and you can see this is where the fluorescence

218

goes down where the edge of the membrane is and you can see that there's a very well defined edge

219

where the particle approaches and approaches and approaches but can't come across okay

220

so This is basically the error of the measurement of the membrane edge you can see from this particle trajectory.

221

And this made us hopeful that by later correlating the motion of the particle to fixed cellular structures,

222

we would be able to see where the particle then finally would get stuck.

223

The next challenge was, of course, to go for cellular particles so we used our magnetic nanoparticles,

224

coupled GFP nanobodies to them via biotin -striptavidin linkage,

225

and then coupled this to a lipid -anchored GFP molecule, okay?

226

So this is a CV1 cell that expresses GPI -GFP, and we then add the particles to the top of the cell, and then they move around the membrane,

227

and we can pull them clearly in the direction of the magnet.

228

You've seen this went actually pretty fast, right?

229

Now, we did this for a number, so you can see the movie again.

230

So we did this for a number of particles and a number of membrane proteins.

231

So the simplest one would be the lipid -anchored GPI -GFP with the fluorescent magnetic nanoparticle attached to it.

232

But we also had a standard LYFP -GT46.

233

This is a standard transmembrane probe which has the transmembrane domain of the LDL receptor.

234

coupled to a GFP on the outside and the transferrin receptor, which is also an often used probe for single particle tracking,

235

where we also have a GFP on the outside here, the YFP.

236

And no matter what protein we used, all of them, GFP, GPI, YFP, GD46, or transferrin receptor GFP,

237

they would exhibit random motion before.

238

We tracked them when we approached the sample with a magnet, they would move into the direction of the magnet.

239

And then when we removed the magnet, they would resume random motion in the plane of the membrane.

240

And we could show that this was specific binding.

241

Here you can see this is in transfected cell.

242

This is a neighboring untransfected cell.

243

You can see that these particles, they only bind to the transfected cell.

244

Using confocal microscopy, we can see that they are nicely on top of the cell

245

and don't bind to the cover slip or anything.

246

And when we then quantified the difference in diffusion coefficient of many, many particles and many, many cells before and after,

247

we could see that globally speaking, there's no difference in particle diffusion before and after they were pulled.

248

So it doesn't seem that we introduce any greater effects just by pulling them a little bit through the membrane.

249

So this is a freely diffusing, normally behaving membrane molecule before you pull it.

250

Then you pull it through the membrane and it can just resume its lateral motion again.

251

Our aim, of course, is now to correlate the motion of these particles to obstacles inside the cell membrane

252

and to do so we perform storm imaging of the actin cortex so we have a 3d capable

253

storm microscope so we can distinguish the dorsal sides of the back of the cell from the ventral side

254

here you can see ventral actin cables

255

and then the dorsal um filaments you can see this is dorsal this is ventral actin

256

and of course when we look at molecules

257

that bind to the top surface of the cell we want to correlate motion to these dorsal actin fibers right

258

so here you can see again gpi gft expressed in a

259

cd1 cell we can then see as we pull this molecule

260

through the membrane it kind of hits a barrier slides along

261

that barrier and then so it hits the barrier it slides along the barrier

262

and then it seems to get stuck okay um and when

263

so of course this is where the magnet is right

264

so it tries to move towards the magnet

265

and then it somehow hits a barrier and it finally gets stuck

266

so now if you look at the same area in wide field actin imaging we can see

267

that there's some prominent actin structures there this is the dorsal actin

268

and the 3d storm

269

and here's the correlation here's the correlation of the trajectory with the actin filaments

270

and you can see that it kind of moves between a couple of actin bits

271

and then it gets stuck on this longer filament tries to find a way moving along it

272

but finally gets caught in between the actin cortex

273

so i hope i could convince you that using this method

274

that we can detect obstacles to the motion of membrane proteins in the plasma membrane.

275

So this is how far we got in this story.

276

And we hope that in the future, that we will be able to use this to further investigate diffusion barriers in the plane of the membrane of cells.

277

So our first aim is of course now to go back to the diffusion barrier in the neurons

278

and then see if we can see any compartmentalization using this method.

279

in neurons that correlate with these 200 nanometer distance actin rings.

280

But of course there's other essays one can imagine that this could be useful for.

281

Of course, one of the biggest problems that we face in membrane biology is quantifying molecular events such as receptor dimerization, right?

282

So a lot of people try to do this by blinking

283

or by co -localization studies or by clustering or by all kinds of statistics.

284

However, if you imagine you would have two subunits of a dimerizing receptor,

285

one of them would be fluorescent and the other one would carry a magnetic tag, and you would now add a magnet to it,

286

you could see that the true dimers, so molecules that are really interacting, they would move.

287

whereas other dimers apparent dimers

288

that only appear to co -localize within a certain distance would tear apart the magnetic part of the dimer

289

and we could move so we could calculate a fraction of molecules

290

that would be dimers of course like all these dimer this would still be dependent on quantitative labeling,

291

but this would be something we would have to figure out like in all assays.

292

So with that, I am already finished with the work from this paper.

293

This work has been done by Gio Hui Li, who is pictured here when she's just about to win the poster prize at the biophysical society meeting this year,

294

so that was very exciting for us.

295

She is soon to be starting, hopefully, her postdoc, and we look forward for a very exciting PhD defense.

296

She has done this work together with the Paula Santos -Otte, an undergrad student who's been spearheading the initial experiments in the lab.

297

Brady and Au has been computing software to overlay trajectories and super resolution data.

298

Stefan Block helped us with calculation of forces in supported membrane bi -layers.

299

And Jakob helped as well with super resolution imaging.

300

And with that, I want to thank our funding agency, the Deutsche Forschungsgemeinschaft, who has been very generous.

301

And I'm open for your questions.

302

Thanks so much, Helge.

303

Such a fantastic story.

304

Thank you really really lovely and congratulations for gia

305

and you as well um so i'd like to start with my own questions by tradition

306

so you said you are not really changing the diffusion coefficient

307

of the molecule by pulling it you're just directing it right

308

yeah can you actually change the diffusion coefficient of the molecule can you apply more force let's say

309

um magnetic field so we could get closer so there are people

310

so when i looked around into this

311

if it would make sense to go in

312

that direction i found a guy who actually has a needle

313

that hasn't that is like an electromagnet right

314

that would of course be much superior to what we do

315

and that would be really exciting to do

316

because you could tune the the strength of the field and thereby more accurately influence membrane protein motion.

317

But this was like a defense project.

318

So this is outside of our budget.

319

So I think that would be, of course, very exciting, but it's beyond our technical capabilities.

320

So of course you detect, right?

321

As soon as you pull, you detect an apparent difference in diffusion coefficient, but that is because you have two components.

322

You have the diffusion and the pull.

323

And related to this, now you're pulling things towards the side of the membrane.

324

Can you pull the things towards the center of the membrane so that you actually collect some molecules towards the center to,

325

let's say, induce some curvature or induce some crowding in the cell?

326

I guess one could attack something from the top.

327

Yeah, I think that should be possible.

328

We never tried really to go like this with the needle.

329

I guess it could be possible because the forces are weak enough, right?

330

So this is based on, of course, indebted to the late Maxime Dahan and Jacob Pila and others,

331

Fred Etock, who have been using intracellular magnetic particles, and they have been covering them with activated CDC42,

332

so they could pull them around in the cell, this clump of magnetic particles inside the cell, and wherever they would then let it be,

333

it would have a burst of actin polymerization and stuff like

334

that, right? so i think that could be possible especially if you compare it with like i don't know holding

335

a guv in direction of the magnet

336

or something this of course one could imagine yeah yeah i

337

think i mean this will be this is super super cool

338

technique you can do endless things you can induce whatever in the cell you can induce a domain

339

and you see you know

340

if it recruits some other molecules okay i still have more questions

341

but i'll start with the odd questions by the audience

342

so ed ed jenkins from oxford is asking this is really

343

cool can you apply this setup to achieve repulsion instead of pulling

344

and have you tried to apply it in a perpendicular manner

345

yes i think perpendicular we just talked about so we have not done

346

that repulsion so the thing is with these tiny particles they are not actually magnetic like a magnet right

347

so it's just too little iron in there and say they are paramagnetic

348

or super paramagnetic so basically they can become attracted by the magnet but they themselves

349

cannot act like a magnet and be repulsed by the respective pole of the magnet

350

so this unfortunately doesn't work in these very very tiny magnetic particles

351

so you need more mass to to achieve that to really have a permanent magnet

352

thank you okay so the next question is by xin jiao very interesting story

353

and fantastic tool can it be applied for intracellular membrane protein study for example er membrane proteins

354

yeah i mean all you have to do is to get the magnetic particle inside

355

um so you know it's possible um it may not be all

356

that physiological i i was first i thought about this very in the beginning the first time

357

when i heard about this uh thing called magnetogenetics right

358

so there are people used ferritin which is a iron harvesting molecule

359

and basically couple that to an ion channel and then they claimed

360

that they could pull on the ferritin

361

which basically has a tiny iron cluster to open the ion channel now that um has been very very controversial

362

because some physicists that know a lot about magnetism could quickly demonstrate mathematically

363

that this is entirely impossible because there's not enough iron in there.

364

So you really need like 50, 100 nanometer iron core to even be possible to be able to move it at all.

365

So it's unfortunately so far, I haven't found a way to genetically encode a magnetic particle.

366

That would of course be wonderful.

367

The closest that gets to this may be these bacterial

368

um iron capsules so some bacteria make capsules that can enrich iron inside

369

and these can be big enough for sensing of magnetic fields

370

and so on so these are like 100 nanometer like like a virus kind of capsules

371

and they can have iron inside so this would be something

372

but it's not very much explored thanks elga

373

so next question is by ej she's asking can these nanoparticles

374

be studied in receptor signaling could these nanoparticles for instance show how they activate the catalytic domain in receptor dimerization

375

which you touched upon a bit yeah i mean

376

so i think uh what what would be an idea that

377

that we would have is that basically you you pull on a protein

378

and then you would think that

379

if the protein is dimerized to another receptor you would pull it with right

380

so so this could then

381

if the other molecule would also be fluorescent you would then see all right i'd pull on the red protein

382

and when i pull my 50 red proteins to the side 20 green proteins come with

383

so i guess two out of five of my receptors at this time are engaged as a dimer right

384

so this is a conclusion that that i could imagine would be accessible experimentally um

385

but of course one would have to have a very well suited system

386

and as quantitative as possible labeling

387

which i think are the the as always the challenges right i mean we know that

388

thank you i guess our next question is by tara sitch

389

from here excellent talk thank you very much i have two

390

questions the first one is how such barrier enforces face control is it mainly size based

391

or are there any specific mechanism what passes and what is not

392

yeah that's a really a good question

393

so we don't really know akikosumi has two models for how his diffusion barrier works right

394

so the the i guess the question you're referring to is

395

how does a molecule in the right

396

so the membrane has two leaflets how does the membrane molecule

397

in the outer leaflet like a lipid how does it even sense something

398

that is below the bottom leaflet how does that work

399

and the idea from akikosumi is that you basically have your actin

400

and bound to your actin are a number of membrane proteins

401

that stick through the membrane and they are kind of like a pillar so

402

if you have several of them your molecule whenever it wants

403

to cross the actin even though it doesn't interact directly with the actin it can encounter one of these pillars

404

and you can then do mathematical modeling and you can calculate

405

that if you have 20 30 percent coverage of your actin filament by pillars this is already leads to measurable

406

confinement in these these zones one could also imagine other means

407

so for example there could be interaction with the cytoskeleton via um phosphatidylserine

408

so there's like charge interaction of the cytoskeleton with the bottom leaflet

409

and then lipids in the bottom leaflet interact with lipids in the top leaflet

410

and this could be some sort of lipid interdigitation

411

or lipid phase phase dependent local confinement so there's some evidence suggesting

412

that this could be possible as well thank you

413

so second part of the question is does actin create an obstacle

414

or a trap in other words if you pull particle back

415

when it's at the obstacle will it be trapped

416

or will it be trapped in the obstacle

417

or it will go back yeah so um

418

if it's just a straight line like this i would expect it to just go back right

419

so um if you imagine having these little pickets then of course

420

if you pull it into an area where there's a lot

421

of pickets until it can't move anymore it may be confined in an area

422

and they kind of have a hard time getting out of there right

423

so um this i guess will be dependent on the circumstances

424

and that's why we are

425

so excited about this axon model where we have just these actin rings every 200 nanometer

426

because it's it's like it's designed for this experiment right

427

so it's very uh clearly defined we know where there's actin

428

where there's no actin um we can fairly straightforward image

429

that after the experiment so hopefully um this will allow us to ask questions exactly like

430

that in the future

431

so can you actually turn off the magnetic field in the middle of the trajectory to see

432

if it starts free diffusion again yeah i think this is what uh gia did

433

so he has this micromanipulator and i mean she's not microsecond you know

434

so so basically you just have to turn the screw

435

and then it moves out um so so this works

436

and we we actually did this

437

so you can see i think there's even this time tags

438

no not here um i think we have a time tag in the paper let me see if i can open it

439

what do i share do i share my screen

440

or do i share that i don't share anything

441

so let me just share this share

442

so in the paper we have a time scale so you can see this is seconds

443

can you can you see that yeah so this is seconds up here

444

so for 24 seconds we observe it wiggling around

445

and then moves in the magnet and it moves for 24 seconds in

446

that direction and then she removes the magnet

447

and then she needs to find the particle again

448

and then she lost some time

449

so it's i you know it's a couple of seconds

450

or a second that you need to remove it

451

but whenever we did this you know we almost all of

452

the time we were able to look at the trajectory before then move it

453

and then have it wiggle around this was kind of our control to make sure that

454

you can remove it and the particle resumes diffusion right

455

because what we wanted to evade was

456

that we just sort of introduce a big artifact in the protein and just destroy something inside the cell

457

this is very good so i have two questions by dig allah from here as well

458

so taking asking how do the different membrane anchors perform in

459

terms of pulling force do longer transmembrane anchors get stuck more

460

often that's the first part of the question about the tm second part is the intracellular part

461

and does the length of the intracellular part have an influence on

462

which kind of structures they get stuck

463

and i myself expand the question a bit let's say you are looking at a cd44

464

which according to sergio greenstein that it interacts with actin cytoskeleton

465

but but according to aki most of the membranes actually interact with actin cytoskeleton directly

466

or indirectly forming the barrier if you look at different proteins

467

that would in the intracellular part directly in track with actin

468

or not would it make a difference

469

or would it be an interesting question yeah

470

so this is uh this would be very very interesting

471

so this is definitely something where we we want to pursue um it is hard work

472

so uh you have to you know this is like a single particle methods.

473

So We are at it, but we can't really tell you anything about it yet.

474

So now we are focusing first on the, in the remaining time we have GIA.

475

We're focusing on the neuron story, but of course we would be very excited to see, you know, what is with the huge ion channel.

476

You know, we would expect it more likely to get stuck.

477

What is, if you have a long transmembrane domain, is it more likely to interact with the actin? um this is of course questions

478

that we would like um to investigate further using this

479

one extension of this question is about the tm i mean it would be probably also interesting

480

if you check a um let's say the raft preferring ordered domain preferring tm versus disordered preferring tm yeah um

481

and maybe one thing also i don't know probably you have done this

482

because this would be obvious thing

483

that would come to a person who is doing all this phase separation

484

and if you pull one thing from ordered membrane to this ordered phase to disordered phase

485

or disordered phase to ordered phase to to calculate the force

486

that is needed absolutely i think i think one one thing

487

that would also be very interesting just on the very basic

488

side on this whole membrane lipid domain field would be to pull a lot of

489

raftophile molecules on one edge of the cell

490

and then do some lardin staining and see if indeed you have a change in, yeah, this is something.

491

Don't tell me if you don't see any difference, okay?

492

Okay, so the next question is by Abhi from Berlin.

493

asking I wonder that have you also tried this in combination with expansion microscopy?

494

Hmm.

495

No. Can you do expansion microscopy with live cells where there's still diffusion?

496

Not really.

497

I guess that would be an instant kill like electron microscopy that's still we're waiting for movies.

498

um we haven't done it i guess as a postdoc imaging

499

method why not right it might work as well as storm does

500

so we might as well do

501

that i don't see a reason why we should not um thanks again

502

so next question is by betul from turkey the three -dimensional

503

structure of the proteins may change due to the magnetic field would

504

that would that lead to any breaking or reorganization of the bonds

505

and interactions would it change actually the force that yeah yeah i think

506

so that's a very good question i think it would not

507

so one should not underestimate um the forces that keep a protein together

508

so hydrophobic effect is an extremely strong thing

509

so it would need really many many piconewtons to pull a

510

single lipid out of a membrane for example right this is something

511

that is many orders of magnitude beyond this i think we are here in the femton obviously,

512

all membrane proteins, they are never really the same, right?

513

So they all kind of like wiggle around and fold and unfold a little the whole time.

514

So I guess you could sort of say you may favor one confirmation a little bit.

515

I can imagine that.

516

But in terms of structural changes that are accompanied by strong binding,

517

I would think that we are below the energies

518

that this can be right so

519

if you imagine like a virus binding to its specific receptor these are significant amounts of energy

520

that get released in such a process so i would think we are below

521

that kind of level thanks again related to this next question is by arta

522

and he is asking with this magnetic manipulation would it be possible to modulate an ion channel

523

yeah i mean so this people try to do this using uh what they called magnetogenetics um

524

but this i as i said was highly doubted i think

525

people have in animals used magnetic beads to do meccano uh

526

like this piezo channels the the meccano sensitive channels this i think this has been done.

527

but it has not been combined with tracking.

528

So I think magnetic beads are used in a lot of different applications already in humans,

529

or you can also use it for molecular imaging and so on.

530

And I think it has been used to pull open these mechanosensitive channels.

531

So earlier there was one question asking if you can do intracellular imaging

532

and as you said the challenge would be getting the magnetic nanoparticle inside the cell

533

now we're talking about cellular level but if you think about an organismal level

534

do you think this would be possible to somehow label with magnetic nanoparticles the extracellular

535

proteins and somehow move them in a larger scale so

536

that you can create let's say a signaling morphogen gradient

537

or something um do you think this would be possible in a larger scale

538

i guess yeah why not um yeah so you think

539

that maybe in an embryo you would sort of inject them in an embryo

540

and then try to artificially enrich in one part of the embryo your magnetic particle coupled i guess

541

that should be possible especially because the forces are not

542

that strong so depending on the distance right so the

543

so you'd have to adjust probably very carefully

544

because with magnets the distance dependence is extremely high right so

545

so force goes crazy so one would have to be be careful about

546

that but it basically compares to i mean

547

if you tried these magnetic beads to pull down proteins right if you put the magnet on the side

548

So imagine that in a little weaker, and I guess it should be possible, yeah, why not?

549

Thanks, so next, usually we have a lot of questions.

550

No, it's wonderful, I think it's really, congratulations to you.

551

I mean, this is an amazing format that everybody is so open to discuss, that you really create a great atmosphere here,

552

and I'm very happy that it's appreciated so much by the students and the audience.

553

I mean, this is a compliment for you and your setup.

554

Well, I told you earlier, I'm going to join in the future.

555

Yeah, this is great.

556

Yeah, actually, I'm not doing anything.

557

I'm just sending you a link and reading the questions.

558

That's my only job.

559

Thanks to you.

560

You are doing all the job, Helgen.

561

It's the style and the atmosphere that makes this happen.

562

That's really nice.

563

So next question is by Christina, and she's asking, are structural changes observed between the Acton ring

564

and the encode protein during stimulation between axon and dendrite

565

and may these changes be observed with magnetic nanoparticles i wish

566

um yeah i mean this is this is very very exciting

567

so there's myosin motors that basically change the the these rings um uh there's membrane molecules

568

that dynamically become recruited to these rings and then execute some signaling.

569

Right now, I have to say the biggest problems we face is actually cleaning up the particles,

570

getting really nice particles that bind well, that behave well.

571

So this is something that just isn't fantastic at the time, right?

572

So this is still not really out of the box particles.

573

You can buy them from many companies

574

but it has been a very very frustrating journey to get some particles

575

that really work well so so they have not achieved a level

576

that maybe quantum dots have right so i think 20 years ago

577

when the quantum dots were first around there was only one of them

578

that you could use only the 605 ones were any good

579

and and now the 655 ones are still the best

580

so I think there's still some way to go for these

581

to be really really nice hopefully I've been now approached by some companies

582

so hopefully we can work together with them to make this a more homogenous

583

and more useful and more tractable sample because this is I have to say right now our biggest problem

584

to get them onto cells

585

in a way

586

so we have to get rid of a lot of them to have the good ones left over

587

and then put them on the cell.

588

So we need to do quite some preparation to get them going well.

589

And is it challenging to actually label

590

the proteins on the surface with biotin tag so

591

that you can label the proteins and pull the proteins in

592

and out no no that's that's that's very straightforward

593

so we have the the nanobody that is bound to a biotin

594

so the nanobody gfp interaction is in the nanomolar range and

595

by controlling the amount of biotin nanobody we add we can control

596

a little bit that you bind monovalently right so

597

if you add less nanobodies then you assume to be surface proteins then

598

or if you pre -coat your particles with nanobodies then you can

599

have the equilibrium shift towards having mostly only one molecule bound to your nanoparticle

600

because of course the cross -linking would also be a worry

601

if you have these 200 nanometer size bits but that is something

602

that has been done by many labs for many different particles it's a very common coupling chemistry.

603

Okay

604

so then a related question comes from Hussein Emre can neuronal

605

stimulation be modulated by moving NMDA channel let's say in neurons

606

out of the synaptic space using fluorescent magnetic nanoparticles um yeah uh

607

well you have to ask okay so uh um

608

so daniel okay in his lab he has been able to show

609

that um in the post -synaptic density of excitatory synapses there are glutamate receptors

610

and once these are opened they close and then they are desensitized for a while so they can't reopen and this is why

611

if you do two quick pulses one after the other the

612

first pulse is going to be higher than the second pulse it's called paired pulse depression

613

and what indeed happens is that these desensitized receptors they diffuse out of synapses and can become replaced by

614

fresh receptors so if you abolish diffusion inside the synapse the pelt balls depression is much stronger

615

so the difference between the first peak

616

and the second peak is much higher than if you have diffusion

617

so diffusion is actually important for the exchange of used receptors

618

for for fresh ones to achieve high cadence high frequency signaling

619

so i could imagine that indeed if you would forcefully remove molecules

620

or keep them out of the synapse you could have a similar effect i do not think

621

that if you have a receptor

622

so receptors are linked to the post -synaptic density via stargazin

623

and pst95 i do not believe that you can pull using the magnetic force

624

an anchored ion channel out of the postsynaptic density for

625

that there's just too many binding sites

626

and it's not gonna i don't think that's gonna work

627

so one one thing for me to understand is rings the

628

actin rings um are they like a like a spring

629

or they are rings next to each other so then they are not continuous i don't think they are continuous um

630

now that it has been shown that myosin actually crosslinks them

631

and seems to pull them together and they can widen

632

and become closer again um i think it's really even an exciting question

633

if they if they have actin always in one direction right

634

so if there's only plus directed act in one way

635

that would be weird because that would mean an axon would have a left

636

and a right right because there would be the plus side this way always this way

637

so it would always go up

638

with the plus end on one side of the axon

639

and down on the other side of the axon i can't imagine

640

that so i would assume that you have them go plus

641

and minus in both directions every single ring

642

so i would think it's rather ring structures than one continuous spiral also that would be a very long spiral.

643

Yeah, yeah.

644

So last question, Helge, by Jan Su.

645

Is it possible to temporarily block the ion channels between the actin rings with biochemicals

646

and reduce the electrical activity of the ion channels to reduce membrane protein stacking possibility?

647

Okay, I do not feel qualified to answer that.

648

That's, can you pose that question again?

649

It sounds very interesting.

650

Is it possible to temporarily block the ion channels between the actin ring with the biochemicals

651

and reduce the electrical activity of the ion channels to reduce the membrane protein stacking possibility?

652

Okay.

653

Well, there are pharmacological inhibitors of voltage -gated channels in the initial segments.

654

So that should be possible um i could imagine calcium having

655

or a calcium influx having an influence on membrane protein motion

656

because calcium of course for example um uh interferes with pip2 your charged lipid cytoskeleton interaction

657

so that's for example one important uh process in dendritic spines

658

when calcium flows in the the membrane relaxes from the cytoskeleton and you can sort of like

659

through new polarization you can make a bigger spine um but i haven't thought about that and

660

that would be a hard experiment to do i guess

661

but uh yeah yeah i could imagine it i'm not saying it's happening

662

but uh yeah okay thanks thanks

663

so much as always it is such a pleasure to yeah

664

likewise this is really great i'm going to join you

665

when is the next one next week yeah next week every

666

week wednesday that's really nice very nice format i i thoroughly enjoyed it thank you i'm sorry

667

that i'm the only person that you see whole talk

668

but um you know there are always downsides yeah no no

669

it's great i really enjoy it it's very uh nice atmosphere

670

here it doesn't feel as weird as it should um sitting

671

in front of your uh your computer the only thing i can't take is

672

that you have a talk sitting down right i think that's

673

so strange um yeah i've seen actually people giving talks standing

674

up in front of the computer as well maybe i should do that next time

675

all right yeah thank you much for having me i hope

676

you enjoyed the paper i everyone enjoyed the paper a lot i really get good feedback on this paper

677

and thanks very much for joining us having time for us today for a fantastic presentation

678

and congratulations again gia

679

and you for this great paper thank you gs in the audience please feel free to write us anytime

680

if you have questions we'd be happy to help thanks a lot helge

✨ Recommend Video

146 sentences

The Most Complex Language in the World

IELTS PrepTOEFL PrepUniversity Students

83 sentences

SOFI: Sequence-based, cross-sectoral, One-health surveillance of food-borne infections | IMMEM XIV

IELTS PrepTOEFL PrepUniversity Students

72 sentences

First reactions | Francis Halzen, Nobel Prize in Physics 2026 | telephone interview

IELTS PrepTOEFL PrepUniversity Students

219 sentences

Review of Musculoskeletal System (Bones & Muscles) - Biology Lectures Series

IELTS PrepTOEFL PrepUniversity Students

32 sentences

What if quantum physics could eradicate illness? | Jim Al-Khalili for Big Think

IELTS PrepTOEFL PrepUniversity Students

Vocabulary and speaking notes for this lesson

This video has 680 sentences and 8384 words to shadow. The speech runs for 57:31. The speaker talks at a steady 146 words per minute, a comfortable pace for shadowing. Only 78% of the words are among the 3,000 most common in English, so the vocabulary is demanding.

Key vocabulary in this video

15 words from the video worth learning, with pronunciation and meaning:

Word	Pronunciation	Meaning
membrane noun	/ˈmɛm.bɹeɪn/	A mechanical, thin, flat flexible part that can deform or vibrate when excited by an external force.
particle noun	/ˈpɑɹtək(ə)l/	A very small piece of matter, a fragment; especially, the smallest possible part of something.
molecule noun	/ˈmɑ.lɪ.kjul/	The smallest particle of a specific element or compound that retains the chemical properties of that element or compound; two or more atoms held together by…
magnetic adjective	/mæɡˈnɛtɪk/	Of, relating to, operating by, or caused by magnetism.
magnet noun	/ˈmæɡ.nɪt/	A piece of material that attracts some metals by magnetism.
barrier noun	/ˈbæɹi.ə/	A structure that bars passage.
bind verb	/ˈbaɪ̯nd/	To tie; to confine by any ligature.
receptor noun	/ɹɪˈsɛp.tə/	A protein on a cell wall that binds with specific molecules so that they can be absorbed into the cell in order to control certain functions.
segment noun	/ˈsɛɡmɛnt/	A length of some object.
domain noun	/ˌdəʊˈmeɪ̯n/	A geographic area owned or controlled by a single person or organization.
investigate verb	/ɪnˈvɛs.tə.ɡeɪt/	To inquire into or study in order to ascertain facts or information.
plasma noun	/ˈplæzmə/	A state of matter consisting of partially ionized gas and electrons.
obstacle noun	/ˈɑbstəkəl/	Something that impedes, stands in the way of, or holds up progress, either physically or figuratively
trajectory noun	/tɹəˈd͡ʒɛktəɹi/	The path an object takes as it moves.
label noun	/ˈleɪ.bəl/	A small ticket or sign giving information about something to which it is attached or intended to be attached.

Phrasal verbs you will hear

Word	Meaning
figure out verb	To come to understand; to discover or find a solution; to deduce.
go back verb	To return to a place or state after having been there at a previous time.
move along verb	To continue without being distracted; to keep going.
move out verb	To leave one's present location; to leave somewhere for another place.
clean up verb	To make an area or a thing clean; to pick up a mess; to tidy.
come across verb	To change sides; to cross over to work for the opposition.
get out of verb	To leave, exit or escape from; to become free of.
go around verb	To move or spread from person to person.

Grammar in this video

The structures the speaker uses most, with the exact words from the video:

Structure	In the video
Conditional sentences if + clause, will/would + verb — a condition and its result	if you imagine you would have · if the other molecule would also · if you have questions we'd be
Present perfect continuous have/has been + -ing — an action that started earlier and is still going on	has been putting · has been computing · have been using
Present perfect have/has + past participle — a past action that still matters now	have been selected · has correlated · has been described
Passive voice be + past participle — the focus is on what happens, not who does it	have been selected · was published · are compartmentalized

Pronunciation to watch

The speaker uses 65 contractions and reduced forms, such as don't, I'm, can't. Say them the short way, as you hear them.

The “th” sounds: thereby /ðɛɹˈbaɪ/, thin /ˈθɪn/, thoroughly /ˈθʌɹ.ə.li/
The “sh” and “zh” sounds: interaction /ˌɪn.təˈɹæk.ʃən/, congratulations /kənˌɡɹæt͡ʃəˈleɪʃ(ə)nz/, shell /ʃɛl/, structural /ˈstɹʌk(t͡)ʃəɹəl/, observation /ˌɑbzɚˈveɪʃn̩/
Long words — get the stress right: investigate /ɪnˈvɛs.tə.ɡeɪt/, trajectory /tɹəˈd͡ʒɛktəɹi/, experiment /ɪkˈspɛɹ.ə.mənt/, interaction /ˌɪn.təˈɹæk.ʃən/, biology /baɪˈɑ.lə.d͡ʒi/

How to practise with this video

Listen to the whole video once without speaking and note the words you do not know.
Shadow it sentence by sentence at normal speed, repeating each one until your rhythm matches the speaker.
Record yourself and compare with the original, paying attention to words like membrane, particle, molecule.

What is the Shadowing Technique?

Shadowing is a science-backed language learning technique originally developed for professional interpreter training and popularized by polyglot Dr. Alexander Arguelles. The method is simple but powerful: you listen to native English audio and immediately repeat it out loud — like a shadow following the speaker with just a 1–2 second delay. Unlike passive listening or grammar drills, shadowing forces your brain and mouth muscles to simultaneously process and reproduce real speech patterns. Research shows it significantly improves pronunciation accuracy, intonation, rhythm, connected speech, listening comprehension, and speaking fluency — making it one of the most effective methods for IELTS Speaking preparation and real-world English communication.

Shadowing technique: read the full step-by-step guide →

Shadowing Practice: Helge Ewers Seminar - Learn English Speaking with Video

The Most Complex Language in the World

SOFI: Sequence-based, cross-sectoral, One-health surveillance of food-borne infections | IMMEM XIV

MRI-Negative Epilepsy

The Uncertainty Principle: What Does It Mean, How Does It Work?

Why the First Cities Might Just Be Fan Clubs - Prof. Jiang Xueqin

Candy Corn In Space

Nanotechnology and Its Applications

Superfluid Dark Matter

The Professor Who Taught People How To Think (1962)

First reactions | Francis Halzen, Nobel Prize in Physics 2026 | telephone interview

Review of Musculoskeletal System (Bones & Muscles) - Biology Lectures Series

What if quantum physics could eradicate illness? | Jim Al-Khalili for Big Think

Vocabulary and speaking notes for this lesson

Key vocabulary in this video

Phrasal verbs you will hear

Grammar in this video

Pronunciation to watch

How to practise with this video

What is the Shadowing Technique?