WEBVTT

1
00:00:00.737 --> 00:00:01.777
Hello, everybody.

2
00:00:01.797 --> 00:00:02.638
OK, here we go.

3
00:00:02.658 --> 00:00:04.679
I'm so glad to see you all here.

4
00:00:04.899 --> 00:00:11.761
Very welcome to Nobel Forum for the announcement of this year's Nobel Prize in Physiology.

5
00:00:11.821 --> 00:00:13.222
It's the same guy as last year.

6
00:00:13.422 --> 00:00:14.902
My name is Thomas Perlman.

7
00:00:15.603 --> 00:00:18.824
I'm the Secretary General of the Nobel Assembly and the Nobel Committee.

8
00:00:18.844 --> 00:00:19.844
What's it going to be?

9
00:00:20.124 --> 00:00:26.747
I will first read the announcement in Swedish and then immediately followed in English.

10
00:00:27.107 --> 00:00:29.648
And we will then present the background

11
00:00:30.375 --> 00:00:32.737
To the prize and open up for questions.

12
00:00:51.452 --> 00:00:52.312
Peter Hegemann.

13
00:00:52.332 --> 00:00:54.894
Oh, wait, wait, wait, wait, bro.

14
00:00:54.954 --> 00:00:58.796
Carl Diceroth, Peter Hegemann, and... Bro, I called it.

15
00:00:58.816 --> 00:01:00.416
The first two... Bro, I called it.

16
00:01:00.456 --> 00:01:01.197
This is so good.

17
00:01:01.397 --> 00:01:01.937
I called it.

18
00:01:01.997 --> 00:01:02.798
This is so good.

19
00:01:03.078 --> 00:01:04.298
Optogenetics.

20
00:01:04.318 --> 00:01:05.119
What did I say?

21
00:01:05.139 --> 00:01:06.079
Optogenetics.

22
00:01:06.099 --> 00:01:06.940
What did I say?

23
00:01:07.020 --> 00:01:09.141
That's crazy.

24
00:01:09.461 --> 00:01:10.261
Maybe a year late.

25
00:01:10.281 --> 00:01:11.402
Yeah.

26
00:01:11.462 --> 00:01:13.503
That's good.

27
00:01:13.883 --> 00:01:19.466
Jointly to Carl Diceroth, Peter Hegemann, and Georg Nagel.

28
00:01:20.481 --> 00:01:22.082
Georg Nagel.

29
00:01:22.122 --> 00:01:24.163
So not Misenbach.

30
00:01:24.183 --> 00:01:24.504
Yes.

31
00:01:24.744 --> 00:01:29.106
And not genetics.

32
00:01:29.146 --> 00:01:30.847
The other MIT guy, Ed Boyden.

33
00:01:30.987 --> 00:01:31.868
Yeah, yeah, yeah.

34
00:01:31.908 --> 00:01:32.608
Okay, interesting.

35
00:01:32.708 --> 00:01:37.351
Okay, we're going to dig into this, and we will be diving deep finally.

36
00:01:37.431 --> 00:01:37.911
Yeah.

37
00:01:38.132 --> 00:01:38.872
On optogenetics.

38
00:01:40.190 --> 00:01:59.230
pond algae two biophysicists in germany and a psychiatrist at stanford together they gave biology a light switch a protein that opens the moment light hits it chlamydomonas swims toward light nagel and haegemann found its sensor channel rhodopsin

39
00:01:59.991 --> 00:02:04.953
Blue light opens the channel, ions rush in, and the cell gets an electrical signal.

40
00:02:05.814 --> 00:02:09.115
In 2005, Diceroth put that gene into neurons.

41
00:02:09.595 --> 00:02:12.977
A flash of blue light and the neuron fires, on command.

42
00:02:13.597 --> 00:02:16.118
By 2007, it worked in living mice.

43
00:02:16.938 --> 00:02:22.561
Today, it's how neuroscientists map the circuits behind memory, emotion, and behavior.

44
00:02:23.281 --> 00:02:25.242
And it's being tested to restore sight.

45
00:02:26.022 --> 00:02:32.467
This is From First Principles, and this is the 2026 Nobel Prize in Physiology or Medicine.

46
00:02:34.288 --> 00:02:35.008
Hello, Internet.

47
00:02:35.028 --> 00:02:45.031
This is your captain speaking, Lester Nare, joined as always by my co-host and our resident PhD and predictor of the future, Krishna Chowdhury.

48
00:02:45.151 --> 00:02:52.933
We are here for the first of our three-part Nobel Prize special, and we're talking about medicine and physiology today.

49
00:02:53.293 --> 00:02:58.414
First, we want to thank all of you who voted in our Instagram poll.

50
00:02:58.434 --> 00:02:59.414
34% of you got it right.

51
00:03:02.155 --> 00:03:02.575
Well done.

52
00:03:02.815 --> 00:03:05.037
Yeah, something like 500 plus people voted.

53
00:03:05.377 --> 00:03:08.739
And I was actually very surprised at how few people voted something else.

54
00:03:09.139 --> 00:03:09.459
Yeah.

55
00:03:09.719 --> 00:03:10.920
People have a lot of faith in us.

56
00:03:11.180 --> 00:03:11.841
That's crazy.

57
00:03:13.321 --> 00:03:16.123
So again, thank you all so much for those votes.

58
00:03:16.583 --> 00:03:21.526
We are going to talk about the science from the ground up today as always.

59
00:03:21.766 --> 00:03:27.810
And this will be for the Nobel Prize in Medicine from First Principles.

60
00:03:28.528 --> 00:03:28.928
That's right.

61
00:03:28.948 --> 00:03:47.402
The Nobel Assembly at Karolinska Institute has awarded the 2026 Nobel Prize in Physiology or Medicine jointly to Carl Dyseroth at the Howard Hughes Institute of Medicine and Stanford University, Peter Hegemann at the Humboldt University of Berlin, and Georg Nagel.

62
00:03:48.122 --> 00:03:54.564
at the University of Würzburg in Germany for their discoveries concerning light-gated ion channels and optogenetics.

63
00:03:55.085 --> 00:03:57.646
This is a prize that I have been calling since last year.

64
00:03:57.666 --> 00:03:58.846
It didn't go last year.

65
00:03:59.246 --> 00:04:02.727
I recycled the prediction this year, and voila, it's here.

66
00:04:02.767 --> 00:04:06.309
So that's two years in a row that I've gotten something right.

67
00:04:07.550 --> 00:04:13.276
What makes this prize particularly interesting is that it represents a remarkable convergence of fields.

68
00:04:13.316 --> 00:04:27.111
We've got biophysics, molecular biology, microbial physiology, genetics, electrical engineering, neuroscience, and they're all coming together to solve one of the most fundamental problems in biology, which is namely...

69
00:04:28.141 --> 00:04:42.497
How do you establish a causal relationship, a causality between single cells and their activities in the brain or in neural tissue, and as we'll see in other parts of the body, and the behavior of the entire organism?

70
00:04:43.213 --> 00:04:46.935
That's a crazy link to go from a single cell all the way to behavior.

71
00:04:47.735 --> 00:04:51.637
And that is the central question that this technology is trying to answer.

72
00:04:52.117 --> 00:05:00.641
This is also a great example of how fundamental discoveries are seemingly coming from obscure biological systems, and they can totally transform medicine.

73
00:05:00.741 --> 00:05:08.185
And this, as always, the Nobel Committee puts out artwork, and this is the particular artwork that is their flagship sort of thing,

74
00:05:08.627 --> 00:05:32.156
thing that they're putting out to the press it depicts um a woman who's playing chess there's a light switch that goes into a particular neuron in her brain and it's got something to do with chess behavior by the end we should be able to understand everything about this image it's giving queens gambit yes um but i think also this dovetails with what we talked about with the golden goose awards where where you start and where you end are

75
00:05:32.816 --> 00:05:35.617
there can be a large gap in between those two spaces.

76
00:05:35.977 --> 00:05:45.179
And I think this is going to be an interesting additional proof point to that idea about basic research and then ultimately clinical-type outcomes.

77
00:05:45.239 --> 00:05:46.719
Yeah, yeah, yeah.

78
00:05:46.759 --> 00:05:49.260
This is a flagship example of exactly that.

79
00:05:49.600 --> 00:05:50.360
It's pretty amazing.

80
00:05:50.760 --> 00:05:53.421
So the whole point of neuroscience, right, is to understand the brain.

81
00:05:53.901 --> 00:05:56.902
These are some famous author quotations about the brain.

82
00:05:57.182 --> 00:06:01.223
The Nobel Institute in their press release quotes Virginia Woolf,

83
00:06:01.863 --> 00:06:10.606
with her quote, my own brain to me is the most unaccountable of machinery, always buzzing, humming, soaring, roaring, driving, and then buried in mud.

84
00:06:10.966 --> 00:06:11.526
And why?

85
00:06:11.746 --> 00:06:13.027
What is the passion for?

86
00:06:13.707 --> 00:06:14.487
They quote that one.

87
00:06:14.787 --> 00:06:17.008
I brought up two more that I quite like.

88
00:06:17.348 --> 00:06:20.730
The second one is from Arthur Conan Doyle, the author of Sherlock Holmes.

89
00:06:21.030 --> 00:06:23.430
Sherlock Holmes says, I am a brain, Watson.

90
00:06:23.450 --> 00:06:29.312
The rest of me is a mere appendix, which is quite interesting for him to say all the way back then, that really

91
00:06:30.973 --> 00:06:34.535
the self, the illusion of reality, everything around us.

92
00:06:34.975 --> 00:06:37.517
Finally, Maya Angelou, this one's quite funny.

93
00:06:37.857 --> 00:06:38.818
The brain is wonderful.

94
00:06:38.938 --> 00:06:43.841
It starts working the minute you're born and never stops until you get up to speak in public.

95
00:06:44.021 --> 00:06:56.088
I quite like that last quote by Maya Angelou because Senator John Kennedy, who's very famously funny, he adopted it on the floor of the Senate and said, the brain is an incredible organ.

96
00:06:56.268 --> 00:06:56.408
Yeah.

97
00:06:57.396 --> 00:07:01.638
It starts working when you're born, and then it stops working when you get elected to Congress.

98
00:07:03.219 --> 00:07:06.820
Well, it was one of those highlight videos of all the funny things he says.

99
00:07:06.880 --> 00:07:08.421
But it's really quite amazing, right?

100
00:07:08.441 --> 00:07:17.525
How can this organ that only weighs 1.3 kilograms, like if you put your hands like this, that is about how big your brain is, if you put your hands together like this.

101
00:07:17.945 --> 00:07:21.907
And that 1.3 kilograms of fat and neural tissue is,

102
00:07:22.799 --> 00:07:26.862
That holds childhood memories, daydreams, creativity.

103
00:07:27.263 --> 00:07:30.766
It gives us joy, love, jealousy, all of the emotions.

104
00:07:31.086 --> 00:07:35.950
It also controls the pace of heartbeats, the rhythm of breathing, the sleep cycle.

105
00:07:36.691 --> 00:07:43.997
Literally everything that has to do with being alive is controlled by 1.3 kilograms that are in your skull.

106
00:07:44.557 --> 00:07:45.298
And to give you a

107
00:07:46.610 --> 00:07:49.071
sense of how complex this problem is.

108
00:07:49.551 --> 00:07:53.312
This is a cubic millimeter of the mouse's visual cortex.

109
00:07:53.752 --> 00:07:55.893
This is by the Allen Brain Institute.

110
00:07:56.173 --> 00:07:57.773
They're part of the Microns project.

111
00:07:58.133 --> 00:08:04.215
They took a cubic millimeter of the mouse visual cortex, sliced it up, and then stained it, and you can now...

112
00:08:06.249 --> 00:08:11.713
trace the hundreds of thousands of neurons that are in that one cubic millimeter.

113
00:08:12.153 --> 00:08:17.637
Each of those neurons can be traced using machine learning to like stitch everything together.

114
00:08:17.677 --> 00:08:22.099
And you can see the amazing diversity of the cells there.

115
00:08:22.119 --> 00:08:22.500
Yeah.

116
00:08:22.520 --> 00:08:22.780
Yeah.

117
00:08:23.501 --> 00:08:28.306
There's 500 million synapses in this one cubic millimeter.

118
00:08:28.326 --> 00:08:28.806
That's so unbelievable.

119
00:08:28.826 --> 00:08:36.054
And the length of axons is four kilometers worth of axons, like neural fiber, in this one cubic millimeter.

120
00:08:36.294 --> 00:08:38.457
Now, that's a cubic millimeter of the mouse brain.

121
00:08:38.797 --> 00:08:44.022
For the human brain, the whole human brain, that's 100 billion neurons, 500 trillion synapses.

122
00:08:44.917 --> 00:08:48.703
So the question is, if we want to study this thing, how do we establish causality?

123
00:08:48.863 --> 00:08:49.343
Yeah, right.

124
00:08:49.484 --> 00:08:49.684
Right?

125
00:08:49.704 --> 00:08:49.864
Right.

126
00:08:50.225 --> 00:08:55.492
Because for most of the 20th century, neuroscience was fundamentally an observational science.

127
00:08:55.893 --> 00:08:56.133
Okay?

128
00:08:56.374 --> 00:09:00.279
You observe what the neuron is doing and you try to correlate it to behavior.

129
00:09:00.299 --> 00:09:00.520
Mm-hmm.

130
00:09:01.312 --> 00:09:04.834
And researchers could record electrical activity of neurons.

131
00:09:04.854 --> 00:09:07.135
They could stimulate brain regions using electrodes.

132
00:09:07.495 --> 00:09:09.216
They could lesion particular structures.

133
00:09:09.296 --> 00:09:11.477
And this gives you a good sense of what is going on.

134
00:09:11.917 --> 00:09:14.878
But these methods have significant limitations.

135
00:09:15.579 --> 00:09:21.681
I can imagine because part of it's like you have to wait for the things to do what you're looking them to do and then hope and then track.

136
00:09:22.062 --> 00:09:22.922
And it's not quite –

137
00:09:24.578 --> 00:09:27.780
It's not quite like a two-way street of information.

138
00:09:27.920 --> 00:09:29.381
It's a one-way street of information.

139
00:09:29.701 --> 00:09:31.082
And it's done amazing things, right?

140
00:09:31.102 --> 00:09:33.183
Like consider electrophysiological recordings.

141
00:09:33.623 --> 00:09:35.104
So here you take an electrode.

142
00:09:35.124 --> 00:09:40.427
You can either put it in the vicinity of the cell or sometimes you can use a patch clamp and go inside the cell.

143
00:09:41.107 --> 00:09:45.312
and see what the voltage is across the membrane, from the inside to the outside.

144
00:09:45.612 --> 00:09:54.842
Neurons fundamentally use the transport of ions to create a kind of electricity that they use to talk, and that's how you get these very, very fast responses.

145
00:09:55.123 --> 00:09:58.987
Now, with electrophysiology, you can learn a lot about how action potentials

146
00:09:59.127 --> 00:10:01.569
How neurons propagate from one end of the neuron to the other?

147
00:10:01.870 --> 00:10:03.631
How neurons aggregate information?

148
00:10:03.671 --> 00:10:08.756
Like if a bunch of neurons are talking to the single neuron, how do you actually get this particular neuron to fire?

149
00:10:09.036 --> 00:10:13.601
Turns out there's ions that are coming in that change the membrane potential, right?

150
00:10:13.621 --> 00:10:18.125
Because if you dump a bunch of positive ions on the inside, then that's going to change the electrical environment.

151
00:10:18.465 --> 00:10:24.367
That electrical environment is then going to change the proteins that are nearby, which is going to let in more ions.

152
00:10:24.587 --> 00:10:25.767
That's going to let in more ions.

153
00:10:25.787 --> 00:10:29.068
You have this kind of runaway effect that creates action potentials.

154
00:10:29.469 --> 00:10:42.373
All of this was discovered using electrophysiology, meaning electrodes going into the physiology for us to measure precisely what the electrical environment is around a neuron or even inside of a neuron, right?

155
00:10:42.393 --> 00:10:42.553
Right.

156
00:10:43.375 --> 00:10:44.055
Amazing stuff.

157
00:10:44.115 --> 00:10:44.315
Yes.

158
00:10:44.456 --> 00:10:47.437
Several Nobel Prizes have happened because of this, okay?

159
00:10:47.817 --> 00:10:51.559
But you reach a kind of ceiling on the kinds of science that you can do.

160
00:10:51.699 --> 00:10:51.879
Right.

161
00:10:52.199 --> 00:10:58.322
Because suppose, for example, you're recording from a neuron in the hippocampus while the animal is learning a spatial task.

162
00:10:58.662 --> 00:11:10.227
This was very famously awarded the 2014 Nobel Prize in Medicine and Physiology for the discovery of place cells and grid cells in the hippocampus and in the entorhinal cortex.

163
00:11:10.647 --> 00:11:12.348
And there on the lower right...

164
00:11:12.868 --> 00:11:15.269
is my first paper that was in Nature.

165
00:11:15.309 --> 00:11:16.569
I was a minor author in there.

166
00:11:16.869 --> 00:11:24.331
But the point of that paper was that, you know, originally it had been thought that these place cells in the hippocampus only respond to place.

167
00:11:24.971 --> 00:11:34.234
But the paper from my encumente lab that I was a part of, it showed that actually those same place cells can respond to visual cues.

168
00:11:34.854 --> 00:11:35.794
Only visual cues.

169
00:11:36.314 --> 00:11:36.974
No behavior.

170
00:11:37.715 --> 00:11:37.915
Right?

171
00:11:38.135 --> 00:11:38.575
So now...

172
00:11:39.617 --> 00:11:44.758
You have a relationship between place and this activity, the neurons activity.

173
00:11:44.838 --> 00:11:55.740
But now we've just shown that visual cues are enough to elicit the same cells to do the same kind of firing pattern that you see when the animal is moving around, right?

174
00:11:56.080 --> 00:12:02.142
I remember we covered this briefly and it was very popular because the concept was very surprising for people.

175
00:12:02.202 --> 00:12:04.982
Yeah, yeah, because they thought that place cells are place cells and that's it.

176
00:12:05.002 --> 00:12:05.762
They respond to place.

177
00:12:10.163 --> 00:12:12.145
creating the activity of this particular neuron.

178
00:12:12.205 --> 00:12:16.988
And I just want to shout out one author of that particular paper, Shonali Dhingra.

179
00:12:17.008 --> 00:12:18.589
She was one of the first authors.

180
00:12:18.909 --> 00:12:20.010
She recently passed away.

181
00:12:20.130 --> 00:12:21.531
She was a huge mentor for me.

182
00:12:21.831 --> 00:12:24.053
She was a postdoc when I was a graduate student.

183
00:12:24.313 --> 00:12:27.635
And a lot of the scientists that I became was because of her.

184
00:12:27.735 --> 00:12:35.341
So, you know, in memory of Shonali, I'd like to dedicate this whole episode because she was a huge influence in my life as a scientist.

185
00:12:35.841 --> 00:12:36.922
The point of that paper was...

186
00:12:37.542 --> 00:12:39.924
Correlation doesn't establish causation, right?

187
00:12:40.484 --> 00:12:44.387
The place does not establish why this particular place cell is firing.

188
00:12:44.427 --> 00:12:48.431
And perhaps even the name place cell might not be that good of a name.

189
00:12:48.751 --> 00:12:52.754
It might be a multimodal sensory cell in the hippocampus.

190
00:12:52.954 --> 00:13:02.702
And so the idea being we've been doing this observational science, which has given us a lot of advances, but we can't necessarily jump all the way to the conclusion that we can define causation.

191
00:13:03.222 --> 00:13:07.144
just using that one-way street that we've been using thus far.

192
00:13:07.444 --> 00:13:08.284
Exactly, yeah.

193
00:13:08.544 --> 00:13:15.067
And so you might say, okay, well, let's try to establish causation by stimulating the neuron electrically, right?

194
00:13:15.087 --> 00:13:16.207
You've got an electrode in there.

195
00:13:16.467 --> 00:13:18.088
Why don't I just pump electricity?

196
00:13:18.508 --> 00:13:22.390
The problem there is electrical stimulation is spatially imprecise.

197
00:13:22.830 --> 00:13:24.310
You've got an electrode, right?

198
00:13:24.390 --> 00:13:28.732
Unless it's targeting a single cell, then maybe you can make that cell fire.

199
00:13:28.752 --> 00:13:30.353
But then it's only going to be that cell, right?

200
00:13:30.493 --> 00:13:30.693
Right.

201
00:13:31.333 --> 00:13:35.015
And how many cells are you going to try to elicit some kind of behavior?

202
00:13:35.455 --> 00:13:38.857
Oftentimes, behavior depends on networks of cells, not just a single cell.

203
00:13:39.137 --> 00:13:40.958
Although you can get lucky sometimes.

204
00:13:41.278 --> 00:13:50.523
There's the very famous Jennifer Aniston cell that responds only to when, like, the animal was shown a picture of Jennifer Aniston.

205
00:13:50.743 --> 00:13:52.604
And it was literally called the Jennifer Aniston cell.

206
00:13:52.744 --> 00:13:55.325
But they got extremely lucky somehow, right?

207
00:13:55.826 --> 00:14:00.448
Where it's like they showed exactly the right actress and then they elicited a response.

208
00:14:00.788 --> 00:14:01.969
So fine, there's causality.

209
00:14:01.989 --> 00:14:06.493
But then most of behavior is because of networks of cells, right?

210
00:14:06.994 --> 00:14:13.600
And sure, the electrical stimulation that you're putting an electrode in, that's going to cause some kind of activity.

211
00:14:13.640 --> 00:14:15.201
But it's going to be spatially imprecise.

212
00:14:15.662 --> 00:14:23.169
If it's outside the cell, those electrons are going to leak everywhere because of the fluids in the brain, the Navier-Stokes equation come back again.

213
00:14:23.409 --> 00:14:23.709
And so...

214
00:14:25.148 --> 00:14:28.350
You're not going to get a nice one-way street.

215
00:14:28.550 --> 00:14:37.896
The resulting behavior could be caused by the neuron around there or it could be caused by single neurons somewhere else that are connected to some other part of the brain that is doing all of this nonsense.

216
00:14:37.956 --> 00:14:49.603
This reminds me of the discussion around precision we talked about when we looked at histotripsy versus other ultrasound therapies in terms of using thermal ablation, which has more of a spread impact.

217
00:14:49.623 --> 00:14:51.925
Yeah, because again, the brain is conductive in heat.

218
00:14:52.265 --> 00:14:53.786
It's also conductive in electricity.

219
00:14:53.966 --> 00:14:55.947
Versus having a more mechanistic approach.

220
00:14:55.967 --> 00:14:58.789
So the precision really matters, especially when you talk about the brain.

221
00:14:58.829 --> 00:14:59.550
Yes, exactly.

222
00:14:59.590 --> 00:15:02.892
Because single neurons can elicit crazy stuff, right?

223
00:15:03.092 --> 00:15:04.833
The other option you have is pharmacology.

224
00:15:05.273 --> 00:15:16.100
Pharmacology meaning like you inject some kind of receptor agonist or antagonist, and then you try to like block some type of receptor that blocks the neuron's activity.

225
00:15:16.320 --> 00:15:17.201
But again, same thing.

226
00:15:17.646 --> 00:15:18.427
There's a problem, right?

227
00:15:18.447 --> 00:15:22.592
Because you inject something that's going to spread around because of diffusion.

228
00:15:22.952 --> 00:15:26.276
You're not sure whether you targeted that specific circuit or not.

229
00:15:26.576 --> 00:15:30.281
The other problem with pharmacology is the response time is quite slow.

230
00:15:31.142 --> 00:15:32.584
You're relying on diffusion.

231
00:15:32.804 --> 00:15:34.005
So you're injecting something.

232
00:15:34.045 --> 00:15:35.927
That thing has to spread, go to the neurons.

233
00:15:36.903 --> 00:15:38.163
Neurons are fast.

234
00:15:38.304 --> 00:15:40.404
So we've got to find a different way to do this, right?

235
00:15:40.504 --> 00:15:45.226
Now, genetic approaches offer a way to get specific, right?

236
00:15:45.306 --> 00:15:54.109
Because you can target specific neurons because you can be like, well, the pyramidal neurons in layer three have this specific type of genetic expression.

237
00:15:54.189 --> 00:15:57.210
And if I can hone in on that, then I can like turn it on or off.

238
00:15:57.710 --> 00:16:01.592
But that's not really reversible, one thing, right?

239
00:16:01.692 --> 00:16:04.873
Once you turn it on, it's kind of hard to take out that kind of stuff.

240
00:16:06.321 --> 00:16:11.669
And the brain also has time to compensate, reorganize, develop alternate pathways.

241
00:16:15.148 --> 00:16:16.409
Again, not great.

242
00:16:17.590 --> 00:16:31.039
We're trying to get to a place where we can create a two-way street where there's a very well-understood discrete control on our input into the system that does not have this, like, combinatorial effect in other ways that is unintended.

243
00:16:31.079 --> 00:16:37.043
So we can isolate saying we only touched this thing and it only had this sort of next step outcome.

244
00:16:37.163 --> 00:16:40.085
And that way we can start to really probe what we can do.

245
00:16:41.205 --> 00:16:45.870
with an understanding that when we look for causation, we know that we didn't input more than we need.

246
00:16:45.950 --> 00:16:47.731
We didn't put too many ingredients into the recipe.

247
00:16:47.791 --> 00:16:48.592
Exactly, exactly.

248
00:16:48.892 --> 00:16:49.893
We just changed one thing.

249
00:16:50.313 --> 00:16:54.797
Okay, so there are three things that we need for this perfect, ideal technique.

250
00:16:54.938 --> 00:16:55.178
Okay.

251
00:16:55.238 --> 00:16:55.498
Okay?

252
00:16:55.798 --> 00:16:56.739
Here are the three things.

253
00:16:57.280 --> 00:16:59.962
First, we need genetic specificity, right?

254
00:17:00.042 --> 00:17:04.546
We need to be able to target specific types of neurons in a very specific location, right?

255
00:17:05.908 --> 00:17:09.871
Second, we need temporal precision, meaning timing needs to be extremely fast.

256
00:17:10.291 --> 00:17:16.116
There you see the electrical signature of an action potential recorded with electrophysiology.

257
00:17:16.196 --> 00:17:23.341
One of the great things with electrodes is that you can record at 40 kHz, so 40,000 data points per second.

258
00:17:24.162 --> 00:17:27.184
So that's 40 data points in a single millisecond, right?

259
00:17:27.244 --> 00:17:29.646
In a thousandth of a second, you'll get 40 data points.

260
00:17:29.666 --> 00:17:33.869
So you can really look at the waveform and see what the waveform of an action potential looks like.

261
00:17:34.690 --> 00:17:40.074
you can see that the width of an action potential is like a millisecond.

262
00:17:40.094 --> 00:17:43.496
So we need whatever tool we're developing to be extremely fast.

263
00:17:44.017 --> 00:17:49.220
And the second, and the third actually, I couldn't find a visual for it, so I just used the Uno Reverse card.

264
00:17:49.240 --> 00:17:49.901
Uno Reverse.

265
00:17:50.021 --> 00:17:53.063
But you want it to be experimentally fast.

266
00:17:53.323 --> 00:17:54.404
reversible.

267
00:17:54.544 --> 00:17:59.168
You want the ability to switch the neural activity on and off repeatedly.

268
00:17:59.188 --> 00:18:01.811
That's how you really establish causation.

269
00:18:02.071 --> 00:18:03.392
You turn it on, something happens.

270
00:18:03.432 --> 00:18:05.414
Well, is that because you turn something else on?

271
00:18:05.634 --> 00:18:08.817
Or if you turn this off, do I recover the old behavior?

272
00:18:09.077 --> 00:18:10.058
Do you go back to baseline?

273
00:18:10.158 --> 00:18:10.779
Exactly, yeah.

274
00:18:11.319 --> 00:18:11.499
And...

275
00:18:12.220 --> 00:18:24.549
One of the people who dreamed about this more precise tool to explore the brain was Francis Crick, the Nobel Prize winner in 1962 for the DNA double helix.

276
00:18:25.210 --> 00:18:29.633
He, you know, he discovered the double helix in 1953.

277
00:18:29.773 --> 00:18:33.516
And then decades later, he started investigating human consciousness.

278
00:18:33.716 --> 00:18:36.058
He became obsessed with this idea about what is consciousness.

279
00:18:42.207 --> 00:18:44.848
activate individual neurons in a living brain.

280
00:18:45.528 --> 00:18:45.768
Okay.

281
00:18:46.248 --> 00:18:46.909
And in...

282
00:18:47.409 --> 00:18:47.849
Ambitious.

283
00:18:48.349 --> 00:18:49.290
It is quite ambitious.

284
00:18:49.350 --> 00:18:50.470
And he said it was quite ambitious.

285
00:18:50.590 --> 00:18:56.892
And in 1999, he wrote this article where he forecasted that maybe we could use light.

286
00:18:56.972 --> 00:18:58.973
He said light would be ideal, right?

287
00:18:59.513 --> 00:19:03.795
If we could somehow manipulate nerve cells and get them to react to light...

288
00:19:04.803 --> 00:19:10.705
That might give us the genetic specificity, because we can maybe only target certain neurons.

289
00:19:11.325 --> 00:19:18.747
It could give us the temporal specificity, because light you can turn on and off at an arbitrarily fast time regime.

290
00:19:19.167 --> 00:19:21.968
And also reversibility, because I can turn light on and off.

291
00:19:22.528 --> 00:19:30.690
And he admits in that paper, it's like a perspective paper, and he admits that the idea sounds very far-fetched, but perhaps not impossible.

292
00:19:31.503 --> 00:19:31.783
Okay.

293
00:19:31.863 --> 00:19:32.263
Perhaps.

294
00:19:32.423 --> 00:19:34.184
Perhaps not impossible.

295
00:19:34.204 --> 00:19:38.246
And that vision of the future is what this Nobel Prize is about.

296
00:19:38.546 --> 00:19:38.926
So good.

297
00:19:39.026 --> 00:19:40.507
The new field of optogenetics.

298
00:19:40.667 --> 00:19:43.028
It was the nature method of the year in 2010.

299
00:19:43.408 --> 00:19:45.529
It's been all over scientific journals.

300
00:19:45.989 --> 00:19:51.892
Using light, researchers are now able to switch individual neural circuits on and off.

301
00:19:52.572 --> 00:19:54.214
They can bring memories to life.

302
00:19:54.735 --> 00:19:55.976
They can create feelings.

303
00:19:56.016 --> 00:19:57.317
They can drive behaviors.

304
00:19:57.918 --> 00:20:03.384
They can study the types of neurons that are involved in psychiatric and neurological disorders.

305
00:20:04.297 --> 00:20:08.880
Optogenetics is, I mean, fundamentally transforming our understanding of the brain.

306
00:20:09.261 --> 00:20:13.223
And every year, countless papers come out using this technique.

307
00:20:13.484 --> 00:20:20.208
It's the hammer of the toolbox of neuroscience in the modern, in the context of being a tool with a lot of uses.

308
00:20:20.308 --> 00:20:20.989
Exactly, yeah.

309
00:20:21.529 --> 00:20:23.250
So how do we get there, right?

310
00:20:23.431 --> 00:20:24.051
How do we get here?

311
00:20:24.611 --> 00:20:28.934
The idea of using light to manipulate biological systems was already known.

312
00:20:29.695 --> 00:20:32.357
This might be a good time to talk about a particular individual who,

313
00:20:34.263 --> 00:20:34.864
Misenbach.

314
00:20:34.904 --> 00:20:35.464
Misenbach.

315
00:20:35.484 --> 00:20:36.606
Garo Misenbach.

316
00:20:37.848 --> 00:20:40.572
In his Wikipedia article, it says...

317
00:20:41.186 --> 00:20:44.107
Misenbach is known as the founder of optogenetics.

318
00:20:45.167 --> 00:20:46.647
Well, this is a bit awkward.

319
00:20:47.527 --> 00:20:50.388
This is why I originally thought he might be the third.

320
00:20:50.488 --> 00:20:50.728
Yeah.

321
00:20:50.808 --> 00:20:51.048
Right.

322
00:20:51.628 --> 00:20:52.788
They give it to Nagel instead.

323
00:20:52.828 --> 00:20:53.548
And we'll see why.

324
00:20:54.108 --> 00:20:54.628
Funny thing.

325
00:20:54.728 --> 00:20:58.309
I was watching the whole press conference right after they announced.

326
00:20:58.649 --> 00:21:02.890
The very first question was, why was Misenbach not awarded?

327
00:21:03.550 --> 00:21:07.771
I'm sure the guy knows him or something and was like, yo, what's going on?

328
00:21:07.791 --> 00:21:08.571
So...

329
00:21:10.291 --> 00:21:19.454
The response was, well, we're writing a forward description about the scientific justification, and we do not comment on why someone got it or not.

330
00:21:19.494 --> 00:21:23.875
And then the follow-up was, if there were four recipients, would Misenbach have gotten it?

331
00:21:23.955 --> 00:21:25.475
And the guy was like, mate, come on.

332
00:21:25.576 --> 00:21:27.636
All right, let's go to the next question.

333
00:21:27.956 --> 00:21:28.596
Next question.

334
00:21:28.636 --> 00:21:29.557
Yeah.

335
00:21:29.897 --> 00:21:31.577
So here's what Misenbach was known for.

336
00:21:32.498 --> 00:21:34.878
He developed this thing called the charge system.

337
00:21:35.078 --> 00:21:37.299
It utilized this multi-gene strategy.

338
00:21:38.703 --> 00:21:44.668
to co-express three distinct proteins from the Drosophila visual cascade.

339
00:21:44.689 --> 00:21:45.990
Drosophila is the fruit fly.

340
00:21:46.450 --> 00:21:51.675
He took proteins from that visual cascade and he expressed that in neurons.

341
00:21:52.190 --> 00:21:55.671
And he, crucially, he expressed this in mammalian neurons.

342
00:21:56.191 --> 00:21:58.072
The problem is the following, though.

343
00:21:58.332 --> 00:22:07.155
If you look at figure A, and this is the paper that came out in 2002 where he did this, it's already kind of optogenetic-y in mammalian neurons already.

344
00:22:07.596 --> 00:22:12.117
But the problem is, in part A, that's the response of this neuron.

345
00:22:12.197 --> 00:22:16.399
So the light turns on when the gray bar ends.

346
00:22:17.313 --> 00:22:19.054
Look at the response time of this thing.

347
00:22:19.374 --> 00:22:21.734
The scale bar down there, that's not milliseconds.

348
00:22:21.755 --> 00:22:22.455
That's seconds.

349
00:22:22.995 --> 00:22:23.215
Okay?

350
00:22:23.455 --> 00:22:27.276
This is on the order of thousands of times the timescale of a neuron.

351
00:22:27.717 --> 00:22:29.697
We want things at the millisecond timescale.

352
00:22:29.717 --> 00:22:30.918
This is happening at seconds.

353
00:22:31.198 --> 00:22:32.058
So this is good, right?

354
00:22:32.098 --> 00:22:35.979
You can turn on neurons and turn off neurons, but you don't get the temporal precision.

355
00:22:36.279 --> 00:22:38.580
Furthermore, you get a lot of variability.

356
00:22:38.860 --> 00:22:43.522
On the right-hand side, you're seeing the gray is when the light was off, and then it's turned on.

357
00:22:44.222 --> 00:22:47.487
Sometimes the response is very quick, half a second to a second.

358
00:22:47.887 --> 00:22:52.434
Sometimes it's very, very long, all the way to 10 seconds later, right?

359
00:22:52.494 --> 00:22:58.462
It took a long time for all of this to accumulate and then that neuron to turn on in panel B. Yeah.

360
00:22:59.383 --> 00:23:07.225
And so this is the – it's like a diet version of ultimately what happened in terms of it didn't have the precision from a time perspective.

361
00:23:07.285 --> 00:23:09.145
It didn't also have for the reversibility piece.

362
00:23:09.425 --> 00:23:12.206
It wasn't quite what it needed to be.

363
00:23:12.306 --> 00:23:12.866
Exactly, yeah.

364
00:23:12.886 --> 00:23:20.168
Because if you take this long to turn something on, then like when you turn it off, presumably it's going to take very long time to turn off as well, right?

365
00:23:20.448 --> 00:23:23.609
And the other thing, it required three proteins to go in.

366
00:23:24.049 --> 00:23:28.090
It also required cofactors, which are like things that bind to the protein to enable –

367
00:23:28.470 --> 00:23:31.854
Like there's all these working parts that are going in to create it.

368
00:23:32.174 --> 00:23:33.075
It's the first time though.

369
00:23:33.095 --> 00:23:33.656
It's 2002.

370
00:23:33.836 --> 00:23:35.278
So this is very early, right?

371
00:23:35.338 --> 00:23:35.878
It's quite early.

372
00:23:35.898 --> 00:23:38.581
And that's why he's been given many awards.

373
00:23:38.962 --> 00:23:40.083
It's not the Nobel this time.

374
00:23:40.163 --> 00:23:40.644
It's so tough.

375
00:23:40.684 --> 00:23:41.645
You know, that is tough.

376
00:23:41.665 --> 00:23:42.265
That's tough.

377
00:23:42.546 --> 00:23:43.066
That is tough.

378
00:23:43.407 --> 00:23:43.627
Um,

379
00:23:44.017 --> 00:23:53.403
And so the challenge now is to create a light control that is genetically targetable, and it's fast to operate at that timescale of neural computation.

380
00:23:54.364 --> 00:23:57.426
At this point, you're saying there's a chance.

381
00:23:57.847 --> 00:23:58.427
We now knew.

382
00:23:58.727 --> 00:24:01.809
We now knew there's a chance, but we got to find the right tool, right?

383
00:24:02.049 --> 00:24:03.691
We want a single protein to do the job.

384
00:24:03.971 --> 00:24:05.672
We don't want these three different proteins.

385
00:24:05.952 --> 00:24:07.633
Plus, I need to inject cofactors.

386
00:24:07.913 --> 00:24:13.097
So that means in order to keep this thing going, I need to keep injecting cofactors that are going to help this protein out.

387
00:24:14.561 --> 00:24:15.742
It's not great, right?

388
00:24:16.882 --> 00:24:21.846
And the solution comes from a very unexpected source, microbial photoreceptors.

389
00:24:22.466 --> 00:24:28.751
So to really understand optogenetics, we need to understand how certain microorganisms detect light.

390
00:24:28.791 --> 00:24:33.514
The relevant proteins belong to a broader family called microbial rhodopsins.

391
00:24:33.554 --> 00:24:38.278
These are proteins that have a nice architecture that lets them...

392
00:24:39.461 --> 00:24:47.908
Absorb light in some kind of retinal chromophore that's like a central chemical, you can imagine, on the inside of this protein that absorbs light.

393
00:24:50.190 --> 00:24:55.795
transforms that energy into a kind of kinetics that makes the outer protein change shape.

394
00:24:56.195 --> 00:25:00.618
And then that outer protein changing shape is going to create some kind of activity in the cell.

395
00:25:01.139 --> 00:25:13.169
The historical foundation of this is from halophilic archaeon, halophilic bacteria, by Dieter Osterhelt.

396
00:25:14.248 --> 00:25:17.869
In 1971, he publishes this paper that shows a pump.

397
00:25:18.209 --> 00:25:29.753
It's a photon-mediated pump, meaning there's a pump on the membrane that takes in light and then uses that energy to pump protons one way or the other.

398
00:25:29.853 --> 00:25:30.473
Across the membrane.

399
00:25:30.493 --> 00:25:31.153
Across the membrane.

400
00:25:31.353 --> 00:25:32.094
So we're getting close.

401
00:25:32.494 --> 00:25:38.456
We're getting to a protein that takes in light energy and then uses that to pump ions.

402
00:25:38.716 --> 00:25:40.716
In this case, it's just protons, right?

403
00:25:40.876 --> 00:25:41.537
H plus ions.

404
00:25:41.557 --> 00:25:41.957
Okay.

405
00:25:42.237 --> 00:25:42.457
Now...

406
00:25:43.509 --> 00:25:45.851
This is great, but this is a pump.

407
00:25:46.212 --> 00:25:48.534
A pump meaning like one single photon is coming in.

408
00:25:48.874 --> 00:25:51.717
That's going to cause one ion to move, right?

409
00:25:51.737 --> 00:25:53.279
Like that work is getting transformed.

410
00:25:53.999 --> 00:25:57.022
What we'd really want is one photon to come in.

411
00:25:57.383 --> 00:25:58.784
That opens a kind of gate.

412
00:25:59.785 --> 00:26:04.530
And then whatever ion concentration is already present, let's say there's a lot of,

413
00:26:05.298 --> 00:26:09.021
Some kind of ion, sodium, potassium, calcium on one side, less on the other side.

414
00:26:09.041 --> 00:26:10.142
There's already a gradient.

415
00:26:10.422 --> 00:26:11.903
All we want to do is open the dam.

416
00:26:12.504 --> 00:26:20.090
Instead of it being a toll booth with cars where each car has to stop, we want a drawbridge where multiple boats can go by because the drawbridge gets open.

417
00:26:20.130 --> 00:26:20.550
Yeah, yeah.

418
00:26:20.610 --> 00:26:23.733
And it's all because there's more boats on this side, so they're just like trying to leave.

419
00:26:23.853 --> 00:26:24.013
Yep.

420
00:26:24.173 --> 00:26:24.393
Right?

421
00:26:24.613 --> 00:26:25.374
That's what we'd like.

422
00:26:25.474 --> 00:26:25.974
Yeah, right.

423
00:26:27.095 --> 00:26:31.399
So let's consider the unicellular green algae system.

424
00:26:32.345 --> 00:26:35.266
Chlamydomonas reinhardti.

425
00:26:35.867 --> 00:26:40.429
I always have a hard time with these taxonomical names.

426
00:26:40.629 --> 00:26:42.970
They're obsessed with Latin, the biologists.

427
00:26:43.390 --> 00:26:46.372
But in any case, this is a green unicellular algae.

428
00:26:46.792 --> 00:26:50.074
And like other photosynthetic organisms, it needs to detect light.

429
00:26:50.674 --> 00:26:50.834
Right?

430
00:26:50.854 --> 00:26:55.156
Because it depends on light to actually live and eat.

431
00:26:55.556 --> 00:26:56.717
And so it's got a little eye.

432
00:26:57.995 --> 00:26:58.615
of some kind.

433
00:26:58.895 --> 00:27:04.056
It's not the eye that we have where we can like make images, but it's certainly got something that is receptive to light.

434
00:27:04.517 --> 00:27:04.757
Okay.

435
00:27:05.017 --> 00:27:10.198
It's got an eye spot and that eye spot senses light and it creates phototaxis.

436
00:27:10.238 --> 00:27:11.738
For example, you put it in a Petri dish.

437
00:27:12.098 --> 00:27:14.019
We've got light on one side, dark on the other side.

438
00:27:14.259 --> 00:27:16.879
You'll see all the algae migrate towards the light side.

439
00:27:17.300 --> 00:27:20.360
So there's got to be some kind of sensitivity in there that is making them sense.

440
00:27:20.660 --> 00:27:21.861
The light is in this direction.

441
00:27:21.961 --> 00:27:23.421
I'm going to swim in this direction.

442
00:27:23.461 --> 00:27:24.361
This is where my food is.

443
00:27:24.381 --> 00:27:24.981
We're going to go there.

444
00:27:25.241 --> 00:27:25.661
Exactly.

445
00:27:28.322 --> 00:27:32.464
And it possesses this kind of light-sensing architecture that can tell the flagella which way to swim.

446
00:27:32.984 --> 00:27:38.607
Quick aside, as I was doing research on this stuff, people use this algae for all different sorts of stuff.

447
00:27:38.667 --> 00:27:42.409
This is the Shoji Takeuchi Research Group at the University of Tokyo.

448
00:27:42.429 --> 00:27:47.231
They're making little tiny machines that are propelled by these algae.

449
00:27:47.711 --> 00:27:49.632
Like, the algae get trapped in these baskets.

450
00:27:50.193 --> 00:27:54.935
And then the flagellar motion, like, creates pinwheels and, like...

451
00:27:55.848 --> 00:27:57.269
movement in some sense.

452
00:27:57.349 --> 00:28:00.050
I just thought that was kind of cool that people are like doing this.

453
00:28:00.170 --> 00:28:03.192
I'm sure there are plenty of use cases, right, for this kind of stuff.

454
00:28:03.592 --> 00:28:05.093
I just thought it was kind of cool.

455
00:28:05.213 --> 00:28:11.236
So the point is researchers have been studying this algae for well over a century because it's really, really cool.

456
00:28:12.097 --> 00:28:14.518
The fact that a unicellular organism can respond to light

457
00:28:16.178 --> 00:28:18.279
Just from a fundamental biology perspective.

458
00:28:18.379 --> 00:28:22.360
There's so many different questions and experiments you'd want to do with that, knowing that that's there.

459
00:28:22.560 --> 00:28:23.021
Exactly.

460
00:28:23.201 --> 00:28:39.227
Now, the quest to uncover the molecular mechanism underlying this phototactic behavior, that takes a turn when Peter Hegeman, our first Nobel laureate that we're going to talk about, he recently graduated with his PhD from Dieter Osterholtz's lab at the Max Planck Institute.

461
00:28:40.567 --> 00:28:42.149
The guy who did the first proton pump.

462
00:28:42.309 --> 00:28:42.789
Okay.

463
00:28:42.949 --> 00:28:45.051
So graduated with a PhD from that lab.

464
00:28:45.091 --> 00:28:48.934
He joins the laboratory of Kenneth Foster in Syracuse University.

465
00:28:49.798 --> 00:28:54.060
The lineage thing is always so interesting when we talk about this story because it really does kind of matter.

466
00:28:54.260 --> 00:28:55.421
It really does matter, right?

467
00:28:56.702 --> 00:29:03.445
He goes to get his PhD from the guy who is the premier guy for bacterial phototaxis.

468
00:29:03.585 --> 00:29:08.147
And then he starts worrying about, okay, this eukaryote, this algae, how does it do it?

469
00:29:08.508 --> 00:29:12.650
He wants to understand the biophysical basis for these photoreceptor currents.

470
00:29:20.606 --> 00:29:23.991
the photoreceptor currents in wall deficient algae mutants.

471
00:29:24.332 --> 00:29:28.498
So there you've got like a pipette that kind of is stuck to the algae.

472
00:29:28.878 --> 00:29:34.847
And from that now you can record the activity of the algae and you can see like how fast is the light response.

473
00:29:35.614 --> 00:29:38.058
And look at the timescale on that light response.

474
00:29:38.078 --> 00:29:39.180
That's what we're talking about.

475
00:29:39.200 --> 00:29:39.841
That's one millisecond.

476
00:29:39.901 --> 00:29:40.862
That's what we're talking about.

477
00:29:40.882 --> 00:29:41.824
That's what we're talking about.

478
00:29:42.004 --> 00:29:42.344
Okay?

479
00:29:42.645 --> 00:29:45.369
Immediately when you see that, you go, nice.

480
00:29:45.489 --> 00:29:46.010
We got him.

481
00:29:46.150 --> 00:29:46.931
We got him, right?

482
00:29:47.112 --> 00:29:48.394
This is in 1996.

483
00:29:49.275 --> 00:29:50.537
Hegeman comes out with this paper.

484
00:29:50.597 --> 00:29:50.897
Okay.

485
00:29:51.659 --> 00:29:59.341
Now, this one millisecond timescale, it's hard to reconcile this with what we normally know about rhodopsins.

486
00:29:59.861 --> 00:30:06.343
For example, the vertebrate rhodopsin, the visual rhodopsin that's in our rods and cones in our eye, here's how it works.

487
00:30:06.843 --> 00:30:14.765
There's the opsin, which is the optical element, the protein that has the optical ability to...

488
00:30:15.720 --> 00:30:20.824
taken light, that thing is coupled to a G-protein receptor.

489
00:30:21.865 --> 00:30:23.867
So it's called a G-protein coupled receptor.

490
00:30:24.848 --> 00:30:28.591
And the signaling happens in a cascade.

491
00:30:29.131 --> 00:30:29.372
Okay.

492
00:30:29.512 --> 00:30:29.792
Okay?

493
00:30:30.012 --> 00:30:41.222
So the light absorption initiates some kind of biochemical signaling cascade that ultimately then regulates a cyclic GMP-gated ion channel that then causes depolarization of...

494
00:30:41.902 --> 00:30:43.544
To then give us the sensation of light.

495
00:30:44.265 --> 00:30:51.593
And so this is a multi-stage process where light is an instigator in now three other downstream steps.

496
00:30:51.733 --> 00:30:53.215
Yeah, and then there's an ion channel.

497
00:30:53.235 --> 00:30:55.197
Then there's an ion channel that the drawbridge opens.

498
00:30:55.297 --> 00:30:55.497
Yeah.

499
00:30:55.598 --> 00:30:57.099
And there's like these two intermediate steps.

500
00:30:57.239 --> 00:30:57.620
Exactly.

501
00:30:57.920 --> 00:31:00.263
Now, presumably, this is not going to be that fast, though.

502
00:31:01.170 --> 00:31:01.690
Right, right.

503
00:31:01.710 --> 00:31:03.050
Because there's a lot going on.

504
00:31:03.070 --> 00:31:04.731
There's a lot going on there in between.

505
00:31:04.791 --> 00:31:06.131
And so it's going to take time.

506
00:31:06.171 --> 00:31:09.852
Not that much time, but not a millisecond, right?

507
00:31:10.292 --> 00:31:21.215
A millisecond is telling you that the left part, the opson part, which is the light detector, and the ion channel part are somehow a single entity.

508
00:31:22.408 --> 00:31:23.769
So I think I get what you're saying.

509
00:31:24.610 --> 00:31:39.880
By the fact that we could see that this was happening at a response time that was at millisecond timescales, there is some other relationship between the opsin and the ion channel that we're not seeing that is not necessarily related to this G-protein cascade that was the starting point.

510
00:31:40.120 --> 00:31:40.560
Exactly.

511
00:31:40.620 --> 00:31:40.981
Okay.

512
00:31:41.101 --> 00:31:42.762
That's the point, right?

513
00:31:42.982 --> 00:31:44.623
So in the early 2000s,

514
00:31:46.064 --> 00:31:51.109
Hegeman and his colleagues, they start investigating what is the molecular basis for that phototaxis, okay?

515
00:31:51.330 --> 00:31:53.772
In the 90s, he's established that this thing is fast.

516
00:31:54.413 --> 00:31:57.976
The 2000s, he gets his own lab, and he's like, I'm going to go after it, okay?

517
00:31:57.996 --> 00:31:59.838
I'm going to see what exactly is happening.

518
00:32:00.179 --> 00:32:05.284
Now, traditional biochemical approaches, they're difficult because these proteins are present in very low abundance.

519
00:32:05.644 --> 00:32:08.687
They're only in the eye spot, and even then, it's not a lot of them.

520
00:32:09.508 --> 00:32:12.730
And it's difficult to purify in the traditional sense.

521
00:32:13.110 --> 00:32:17.473
So the research team pursues a different molecular approach using genes.

522
00:32:17.913 --> 00:32:19.354
They want to identify candidate genes.

523
00:32:19.974 --> 00:32:38.725
And one thing that they piggybacked on is the fact that the algae's expressed sequence tag database, this is by Asamizu and other authors in 2000, they had already published the sequence tag database for this gene.

524
00:32:40.044 --> 00:32:41.346
algae.

525
00:32:41.706 --> 00:32:46.812
So the genetic data is now available for people to try and access.

526
00:32:47.273 --> 00:32:48.835
And they didn't need to create that from scratch.

527
00:32:49.456 --> 00:32:50.877
So they would be able to piggyback again.

528
00:32:52.379 --> 00:32:54.402
Science is interdisciplinary.

529
00:32:54.482 --> 00:32:56.364
Only three people got the Nobel Prize, but

530
00:32:57.517 --> 00:33:00.658
There are multiple people that are involved in this research, right?

531
00:33:00.898 --> 00:33:04.699
And it wouldn't have happened without the entire community going after this problem.

532
00:33:05.239 --> 00:33:06.179
I wanted to make that clear.

533
00:33:06.579 --> 00:33:18.802
So in the early 2000s, three research groups, one led by Peter Hegeman, but another one by John Spudich and then Toshio Takahashi, they independently identify the DNA sequences that encode for these proteins.

534
00:33:19.142 --> 00:33:22.403
What they look for is a protein that looks kind of like

535
00:33:23.163 --> 00:33:25.785
the rod protein in our eyes.

536
00:33:26.065 --> 00:33:27.346
And that's exactly what they find.

537
00:33:27.786 --> 00:33:30.628
On the left is our vertebrate rhodopsin, right?

538
00:33:30.648 --> 00:33:31.889
This is the stuff that's in our rods.

539
00:33:32.229 --> 00:33:35.711
And on the right-hand side is the protein structure of what they found.

540
00:33:35.751 --> 00:33:38.393
You can see that there's immediate similarities, right?

541
00:33:38.493 --> 00:33:45.138
All these alpha helices that are clustered around a central chromophore, right?

542
00:33:46.238 --> 00:33:48.340
And so what they can look for is analogs.

543
00:33:49.267 --> 00:33:53.209
You already know the sequence over here from the Human Genome Project and all that other stuff.

544
00:33:53.510 --> 00:33:56.171
And then you can look for homologues in the algae.

545
00:33:56.651 --> 00:33:59.993
And so we already had this database that was done previously.

546
00:34:00.073 --> 00:34:01.414
We know how it works in the humans.

547
00:34:01.434 --> 00:34:05.617
And we're basically saying, okay, in this database, what looks like the thing that we have on the human side?

548
00:34:05.817 --> 00:34:06.257
Exactly.

549
00:34:06.857 --> 00:34:09.539
And as a way to identify where to do the work, where to start.

550
00:34:09.579 --> 00:34:12.901
Like instead of doing a blind, like starting and where do you begin?

551
00:34:12.921 --> 00:34:14.402
What do you even look for, right?

552
00:34:14.442 --> 00:34:15.883
The genetic sequence is massive.

553
00:34:15.903 --> 00:34:16.063
Right.

554
00:34:16.503 --> 00:34:20.224
So from this, they identify two segments of DNA.

555
00:34:20.764 --> 00:34:24.245
They call it channel Opsin 1 and channel Opsin 2.

556
00:34:25.346 --> 00:34:29.407
These are then put into the genetic databank.

557
00:34:30.027 --> 00:34:36.949
And now these are like maybe the thing that is creating the photo ability of these algae, right?

558
00:34:36.969 --> 00:34:38.670
The idea that these guys can maybe...

559
00:34:39.901 --> 00:34:41.142
Let the algae respond to light.

560
00:34:41.762 --> 00:34:42.122
Okay.

561
00:34:42.262 --> 00:34:47.744
The decisive functional breakthrough emerges when Peter Hegeman contacts George Nagel.

562
00:34:48.684 --> 00:34:49.865
Georg Nagel, I should say.

563
00:34:51.205 --> 00:34:56.267
Georg Nagel, strong electrophysiological expertise in studying ion channels and pumps.

564
00:34:56.747 --> 00:35:08.471
He has already used his expertise with pumps to look at how the original, you know, the proton pumps that I was talking about from Dieter?

565
00:35:08.651 --> 00:35:08.851
Yeah.

566
00:35:09.570 --> 00:35:15.732
He put these inside the Xenopus laevis oocytes, which are frog embryos, effectively.

567
00:35:16.012 --> 00:35:18.013
And he had figured out that these are proton pumps.

568
00:35:18.213 --> 00:35:19.753
That's how we know that they're proton pumps.

569
00:35:19.773 --> 00:35:20.513
This is the guy who did it.

570
00:35:20.854 --> 00:35:27.276
He genetically engineered frog embryos to express that particular protein.

571
00:35:27.396 --> 00:35:30.757
And then you can actually characterize that this is a light-driven proton pump.

572
00:35:31.017 --> 00:35:32.657
So he's already got expertise in doing this.

573
00:35:33.057 --> 00:35:38.139
So Peter Hegeman goes to him and he's like, look, I got these two pieces of DNA, chop one and chop two.

574
00:35:39.579 --> 00:35:42.180
And I want you to express them and study their function.

575
00:35:42.900 --> 00:35:50.283
In 2002, Georg Nagel, Ernest Bamberg, and Peter Hegeman and their colleagues, they published this seminal paper.

576
00:35:51.103 --> 00:35:56.185
This is the first of several papers that we're going to go through that earns them the Nobel Prize.

577
00:35:56.925 --> 00:36:00.206
Channel rhodopsin won a light-gated proton channel in green algae.

578
00:36:00.246 --> 00:36:01.067
This is in science.

579
00:36:01.387 --> 00:36:02.447
Shout out to AAAS.

580
00:36:02.527 --> 00:36:02.807
Yes.

581
00:36:03.887 --> 00:36:08.009
So he showed that green light illumination can induce inward photocurrents

582
00:36:09.211 --> 00:36:11.293
that are primarily carried by protons.

583
00:36:11.333 --> 00:36:12.554
This is, again, still protons.

584
00:36:12.854 --> 00:36:17.718
This thing is opening up, and if there's a gradient of protons, this thing is going to let protons in.

585
00:36:18.039 --> 00:36:21.161
Somehow the algae maintains a gradient of protons.

586
00:36:21.201 --> 00:36:22.162
That's for another story.

587
00:36:22.502 --> 00:36:28.888
But if there's already a gradient of protons, there's more protons on one side than the other, you open up this dam, the protons are going to go through.

588
00:36:29.148 --> 00:36:30.549
And not just a single one this time.

589
00:36:30.589 --> 00:36:32.291
It's going to have the flow of multiple.

590
00:36:32.351 --> 00:36:32.751
Exactly.

591
00:36:32.771 --> 00:36:33.372
It doesn't need...

592
00:36:33.872 --> 00:36:34.673
light to do work.

593
00:36:34.873 --> 00:36:36.054
It's just opening up a dam.

594
00:36:36.074 --> 00:36:36.254
Yep.

595
00:36:36.334 --> 00:36:36.534
Right?

596
00:36:36.954 --> 00:36:43.739
And this biophysical fingerprint provides the first direct evidence of light-gated ion channels.

597
00:36:43.779 --> 00:36:44.079
Okay.

598
00:36:44.219 --> 00:36:48.502
And this is the first time we get a formal designation channel rhodopsin 1.

599
00:36:48.522 --> 00:36:49.382
Okay?

600
00:36:49.402 --> 00:36:51.724
Because now we've established that this is the protein that's doing it.

601
00:36:52.024 --> 00:36:53.325
This is amazing already, right?

602
00:36:53.425 --> 00:36:55.446
This thing can function as a light-activated ion channel.

603
00:36:55.486 --> 00:36:56.487
Yeah, yeah, yeah.

604
00:36:56.507 --> 00:36:57.147
It's already amazing.

605
00:36:57.167 --> 00:36:57.627
Yeah, yeah, yeah.

606
00:36:57.687 --> 00:37:02.991
But it's not the complete story because the electrical response...

607
00:37:03.824 --> 00:37:07.960
isn't particularly spectacular, and it's doing protons.

608
00:37:08.846 --> 00:37:09.066
Right.

609
00:37:09.587 --> 00:37:15.452
We would like something that messes around with sodium, potassium ions, things like that.

610
00:37:15.712 --> 00:37:17.353
The things that are in the brain.

611
00:37:17.473 --> 00:37:17.934
In the brain.

612
00:37:18.014 --> 00:37:18.214
Right.

613
00:37:18.254 --> 00:37:19.435
That stuff that the neurons use.

614
00:37:19.535 --> 00:37:20.636
Neurons don't use protons.

615
00:37:20.716 --> 00:37:24.219
Mitochondria use protons, but I want to turn on a whole cell.

616
00:37:24.419 --> 00:37:25.801
This is cool.

617
00:37:25.861 --> 00:37:26.962
We want to go to the big leagues.

618
00:37:27.062 --> 00:37:27.242
Yeah.

619
00:37:27.382 --> 00:37:27.562
Yeah.

620
00:37:27.822 --> 00:37:33.067
And so the next year, subsequent landmark paper by Nagel and his colleagues, channel rhodopsin 2.

621
00:37:33.347 --> 00:37:34.548
This is the second landmark.

622
00:37:34.648 --> 00:37:36.811
piece of cDNA that they had identified.

623
00:37:37.131 --> 00:37:41.516
This is a directly light-gated cat-ion selective membrane channel.

624
00:37:41.897 --> 00:37:45.080
That word, cat-ion selective, is massive here.

625
00:37:45.300 --> 00:37:47.623
Cat-ions are positively charged ions.

626
00:37:47.944 --> 00:37:50.767
This is a general-purpose ion channel.

627
00:37:51.780 --> 00:37:54.102
You give it sodium, it'll let in sodium.

628
00:37:54.122 --> 00:37:56.304
You give it potassium, it will let in potassium.

629
00:37:56.624 --> 00:37:56.884
Okay?

630
00:37:56.924 --> 00:37:57.745
This is a big deal.

631
00:37:57.805 --> 00:37:58.426
This is a big deal.

632
00:37:58.446 --> 00:37:59.026
This is a big deal.

633
00:37:59.166 --> 00:37:59.326
Right?

634
00:37:59.366 --> 00:38:01.108
This is the second protein, channelrhodopsin 2.

635
00:38:01.448 --> 00:38:05.732
And the other thing that I want to show you, this is, again, he used the oocyte embryos to do this.

636
00:38:07.265 --> 00:38:13.167
In panel B, what you're seeing is the response time of this guy, right?

637
00:38:13.327 --> 00:38:15.827
Immediately it turns on and it actually plateaus.

638
00:38:15.867 --> 00:38:17.848
It's not like it turns on and then goes to baseline.

639
00:38:18.328 --> 00:38:26.190
The plateau, it still keeps producing a current, even when you keep it on for a whole second, right?

640
00:38:26.430 --> 00:38:26.810
Mm-hmm.

641
00:38:27.310 --> 00:38:31.334
So the illumination is directly opened by this ion-conducting pathway.

642
00:38:31.354 --> 00:38:31.894
This is good.

643
00:38:32.114 --> 00:38:39.040
And it's fundamentally different from the conventional G-protein-coupled stuff that we have in our eye, because now this is directly light-gated.

644
00:38:39.220 --> 00:38:39.440
Yes.

645
00:38:39.540 --> 00:38:42.783
So this is a protein that absorbs light and then immediately just turns on.

646
00:38:42.823 --> 00:38:43.324
There's no...

647
00:38:44.124 --> 00:38:49.048
There's none of those two steps that we talked about earlier.

648
00:38:49.548 --> 00:38:52.611
This is more that direct relationship from the initial image.

649
00:38:52.671 --> 00:38:52.991
Exactly.

650
00:38:53.011 --> 00:38:56.554
And that's why we're getting that really fast timescale of work.

651
00:38:56.654 --> 00:38:57.034
Okay.

652
00:38:57.134 --> 00:38:57.494
Right?

653
00:38:57.594 --> 00:38:58.675
Okay.

654
00:38:59.035 --> 00:39:03.199
This was an amazing paper that was out in Proceedings of National Academy of Sciences.

655
00:39:03.959 --> 00:39:08.482
And the researchers immediately recognized a potentially much broader implication.

656
00:39:09.383 --> 00:39:21.505
If this protein is expressed in an unrelated cell, and if that cell could then be made to be electrically responsive to light, then you can use it to control cells in general.

657
00:39:21.545 --> 00:39:22.105
Yeah, right.

658
00:39:22.406 --> 00:39:22.626
Right?

659
00:39:22.746 --> 00:39:22.926
Right.

660
00:39:22.946 --> 00:39:23.586
It's a platform.

661
00:39:23.726 --> 00:39:26.786
And in that work, the last line is actually quite amazing.

662
00:39:27.066 --> 00:39:29.007
This is the last line of their paper in 2003.

663
00:39:29.307 --> 00:39:35.388
Additionally, we have shown that expression of channelrhodopsin 2 in oocytes or mammalian cells is...

664
00:39:35.902 --> 00:39:45.428
may be used as a powerful tool to increase cytoplasmic calcium 2 plus concentration or to depolarize the cell membrane simply by illumination.

665
00:39:45.448 --> 00:39:46.349
That's so crazy.

666
00:39:46.389 --> 00:39:46.589
Right?

667
00:39:46.809 --> 00:39:49.551
So you are literally mimicking synapses in some sense.

668
00:39:49.631 --> 00:39:55.335
Like when a bunch of neurons synapse onto a particular neuron, they release calcium 2 plus ions.

669
00:39:55.655 --> 00:39:59.778
And then those calcium 2 plus ions go inside or they release a bunch of neurotransmitters

670
00:40:00.104 --> 00:40:05.827
That induces the calcium 2 plus ions to go inside the cell, depolarize the cell, and create an action potential.

671
00:40:06.247 --> 00:40:09.849
Here, you're opening up the calcium 2 plus ions, the channel.

672
00:40:10.109 --> 00:40:15.251
And the calcium 2 plus ion goes in, depolarizes the cell membrane simply by illumination.

673
00:40:15.732 --> 00:40:24.616
We now have this understanding of this cellular channel of communication that we can then now use to go back to our two-way street.

674
00:40:25.216 --> 00:40:26.097
We could always read.

675
00:40:26.377 --> 00:40:26.597
Yeah.

676
00:40:27.217 --> 00:40:32.900
We were trying to figure out what language, how can we write with those three elements we talked about at the beginning.

677
00:40:32.920 --> 00:40:33.060
Yep.

678
00:40:33.361 --> 00:40:38.323
The time responsiveness, the experimental reproducibility, and there was a third one that's escaping me.

679
00:40:38.404 --> 00:40:41.765
Reversibility, time, and genetic specificity.

680
00:40:41.785 --> 00:40:42.446
Specificity.

681
00:40:42.686 --> 00:40:43.947
And we kind of have now hit.

682
00:40:44.007 --> 00:40:44.907
Yeah, we've hit them all.

683
00:40:44.967 --> 00:40:47.989
All three of those in that 2003 paper.

684
00:40:48.029 --> 00:40:48.950
Yes, exactly.

685
00:40:49.050 --> 00:40:49.250
Yeah.

686
00:40:49.310 --> 00:40:49.510
Yeah.

687
00:40:49.950 --> 00:40:56.073
And so now the next step is how do we go from a microbial ion channel to now optical control of neurons, right?

688
00:40:56.093 --> 00:40:57.194
You said that that could work.

689
00:40:57.514 --> 00:40:57.714
Right.

690
00:40:57.994 --> 00:41:00.756
But now we need to make this kind of a general purpose thing, right?

691
00:41:00.816 --> 00:41:03.277
How do we do it for like anything?

692
00:41:03.297 --> 00:41:04.218
It theoretically could work.

693
00:41:04.238 --> 00:41:04.878
Yes.

694
00:41:04.918 --> 00:41:06.219
That's what they said, 2003.

695
00:41:06.519 --> 00:41:07.079
Theoretically.

696
00:41:07.319 --> 00:41:09.881
So let's talk about neurons just in general real quick.

697
00:41:09.921 --> 00:41:12.902
Neurons are electrically excitable cells.

698
00:41:12.922 --> 00:41:13.263
Right.

699
00:41:14.083 --> 00:41:19.908
They work with ion channels that open and close and lead in sodium and potassium.

700
00:41:20.268 --> 00:41:23.751
Now, crucially, these things are voltage-gated ion channels.

701
00:41:23.871 --> 00:41:24.752
Here's what I mean by that.

702
00:41:25.353 --> 00:41:27.715
The protein in its native state is closed.

703
00:41:28.863 --> 00:41:40.311
And now let's say I get a bunch of calcium 2 plus coming in because of either channel rhodopsin or because synapses put in a bunch of neurotransmitters and then now that leads to a bunch of calcium 2 plus coming in.

704
00:41:40.591 --> 00:41:42.672
That's going to change the voltage across the cell, right?

705
00:41:42.972 --> 00:41:49.276
If you change the voltage across the cell, you've got an electrical force, an electrical field, and that literally changes the shape.

706
00:41:49.857 --> 00:41:55.580
So it's a voltage-gated ion channel because it's an ion channel that turns on or off depending on the voltage.

707
00:41:56.041 --> 00:41:57.742
And if I can control the voltage by...

708
00:41:58.302 --> 00:42:05.825
letting in calcium 2 plus ions, then I can control these voltage-gated ion channels, and that is going to create my action potential that goes from one place to the other.

709
00:42:07.125 --> 00:42:17.329
The key question now is whether the relatively small photocurrent that is produced by the channel rhodopsin is enough to trigger all of the rest.

710
00:42:17.509 --> 00:42:25.732
Because usually the synapses, they dump a bunch of neurotransmitters, a bunch of calcium 2 plus goes in, and then that causes the membrane potential to change that causes this cascade effect.

711
00:42:26.072 --> 00:42:27.373
Can channel rhodopsin do this?

712
00:42:27.814 --> 00:42:29.135
Right, across this.

713
00:42:29.195 --> 00:42:30.035
Yeah, yeah, yeah.

714
00:42:30.075 --> 00:42:33.636
Can it trigger the runaway cascade, right?

715
00:42:33.956 --> 00:42:38.438
You can imagine neurons are like in this unstable equilibria of like an inverted pendulum.

716
00:42:38.618 --> 00:42:38.798
Okay.

717
00:42:38.838 --> 00:42:42.180
And it needs a kind of kick to like go down and fire.

718
00:42:42.260 --> 00:42:42.460
Yep.

719
00:42:42.540 --> 00:42:42.740
Right?

720
00:42:42.980 --> 00:42:45.121
But is the channel rhodopsin kick enough?

721
00:42:45.341 --> 00:42:45.861
To have it.

722
00:42:45.881 --> 00:42:46.041
Right?

723
00:42:46.081 --> 00:42:46.241
Yep.

724
00:42:46.301 --> 00:42:46.981
That's the question.

725
00:42:47.001 --> 00:42:47.762
That's a great visual.

726
00:42:47.942 --> 00:42:56.225
So in 2005, Ed Boyden, Bang Zhang, George Nagel, and Ernst Bangberg, and Carl Dyseroth

727
00:42:56.664 --> 00:42:59.065
They publish one of the field's defining papers.

728
00:42:59.085 --> 00:43:02.065
This thing has like 5,000 citations by now.

729
00:43:02.446 --> 00:43:03.466
It's in Nature Neuroscience.

730
00:43:03.866 --> 00:43:08.687
Millisecond timescale, genetically targeted optical control of neural activity.

731
00:43:08.807 --> 00:43:09.908
And this is where we get all three.

732
00:43:10.088 --> 00:43:15.329
This is the translation now from the algae to now neurons.

733
00:43:15.389 --> 00:43:15.609
Yeah.

734
00:43:15.949 --> 00:43:18.610
This paper demonstrates that you can take channelrhodopsin.

735
00:43:18.990 --> 00:43:27.294
And you can express it in mammalian neurons and use that to trigger action potentials with millisecond timescale precision.

736
00:43:27.674 --> 00:43:30.195
So I want to talk about those two second authors.

737
00:43:30.375 --> 00:43:30.635
Okay.

738
00:43:30.835 --> 00:43:32.356
Carl Diceroth is on the left.

739
00:43:32.696 --> 00:43:33.757
He's the one who won the prize.

740
00:43:34.517 --> 00:43:36.398
Ed Boyden was a PhD student at the time.

741
00:43:37.761 --> 00:43:40.842
He kind of was at the start of this whole thing.

742
00:43:41.022 --> 00:43:43.683
He did a lot of the work, and he hasn't gotten the prize.

743
00:43:43.703 --> 00:43:44.824
He's at MIT right now.

744
00:43:44.864 --> 00:43:47.945
There's going to be a lot of drama out there about why he didn't get it.

745
00:43:48.365 --> 00:43:55.968
The third guy, Feng Zhang, this is the second time that he's... Because the first time we talked about was CRISPR.

746
00:43:56.108 --> 00:43:56.848
Oh, that's right.

747
00:43:56.908 --> 00:43:58.409
That's the same Feng Zhang.

748
00:43:59.409 --> 00:44:03.291
He's the same guy who used CRISPR to finally...

749
00:44:04.978 --> 00:44:10.041
to finally create a type of programmable genetic scissor in mammalian cells.

750
00:44:10.101 --> 00:44:10.481
Remember?

751
00:44:10.901 --> 00:44:13.623
We had the whole drama between Berkeley and MIT.

752
00:44:13.663 --> 00:44:15.523
Berkeley was Jennifer Doudna.

753
00:44:15.603 --> 00:44:17.524
This is the MIT guy who has the patent.

754
00:44:17.885 --> 00:44:18.925
I don't know.

755
00:44:19.105 --> 00:44:20.406
And this is the second time.

756
00:44:20.466 --> 00:44:21.226
That's tough.

757
00:44:21.386 --> 00:44:22.227
It's tough, bro.

758
00:44:22.687 --> 00:44:26.849
It's crazy that he's been at the forefront of both CRISPR and optogenetics.

759
00:44:27.289 --> 00:44:28.409
That's actually pretty crazy.

760
00:44:28.589 --> 00:44:29.830
Yeah, I mean, the guy's a beast.

761
00:44:29.870 --> 00:44:31.210
Yeah, that's, yeah, okay.

762
00:44:31.230 --> 00:44:36.111
So, but, you know, MIT Broad Institute just getting shafted.

763
00:44:36.211 --> 00:44:36.971
I plead the fifth.

764
00:44:37.211 --> 00:44:37.871
Yeah.

765
00:44:37.891 --> 00:44:39.492
You know who didn't get shafted last year?

766
00:44:39.712 --> 00:44:40.052
Princeton.

767
00:44:40.212 --> 00:44:41.132
Yeah, that's right.

768
00:44:42.012 --> 00:44:44.553
And perhaps coming this year, tomorrow or the day after.

769
00:44:45.093 --> 00:44:47.294
In any case, so here's what the experiment was.

770
00:44:47.374 --> 00:44:50.914
It was conceptually quite simple, but it's technically very, very important.

771
00:44:51.034 --> 00:44:53.635
This is a Nobel Prize committee project.

772
00:44:54.812 --> 00:44:56.333
reproduction of some of their work.

773
00:44:56.693 --> 00:45:13.604
They took hippocampal neurons in a Petri dish, so from the hippocampus, mammalian hippocampus, they expressed channelrhodopsin in those hippocampal neurons and showed that the hippocampus can actually affect the firing of the neurons.

774
00:45:13.624 --> 00:45:15.625
The light can actually affect the firing of the neurons.

775
00:45:16.085 --> 00:45:16.486
Also,

776
00:45:17.567 --> 00:45:25.932
With suitable illumination, the depolarization can be controlled like the amount of depolarization can be controlled, right?

777
00:45:25.972 --> 00:45:31.276
The intensity of the light will tell you sort of how much firing happens.

778
00:45:31.356 --> 00:45:38.260
You turn the volume dial to 11, the depolarization is at a discrete point that's similar to that volume 11.

779
00:45:38.280 --> 00:45:39.101
You turn it to 5, 5.

780
00:45:39.861 --> 00:45:40.382
It looks like...

781
00:45:40.682 --> 00:45:50.850
So you now have a granularity of control where the intensity of light actually drives the level... Like how the depolarization happens.

782
00:45:51.231 --> 00:45:56.575
Which then gives you more degrees of movement when you're trying to do this.

783
00:45:56.715 --> 00:46:04.762
Yeah, and you can get reliable spiking frequency of tens of hertz, which is on the order of what normal neurons would do in a brain.

784
00:46:05.062 --> 00:46:08.605
Now, this is extraordinary because, one, it shows the molecular machinery...

785
00:46:08.905 --> 00:46:10.267
that was discovered in the green algae.

786
00:46:10.447 --> 00:46:12.509
You can now express it inside mammalian neurons.

787
00:46:12.849 --> 00:46:13.690
Yeah.

788
00:46:14.090 --> 00:46:15.592
The protein is genetically encoded.

789
00:46:15.652 --> 00:46:18.935
So once you put the gene into the neuron, the cell can produce...

790
00:46:19.687 --> 00:46:21.768
the light-sensitive channel by itself.

791
00:46:21.908 --> 00:46:23.088
You don't need to keep pumping stuff in.

792
00:46:23.208 --> 00:46:23.568
Exactly.

793
00:46:23.588 --> 00:46:25.648
It'll just make the channel rhodopsin on its own.

794
00:46:26.029 --> 00:46:31.310
And the channel rhodopsin will go and be expressed on the membrane because the cell kind of knows what to do with it.

795
00:46:31.450 --> 00:46:31.990
This is quite nice.

796
00:46:32.030 --> 00:46:32.810
In some sense, right?

797
00:46:33.270 --> 00:46:36.891
And the other thing is the chromophore that's in the middle of this whole thing.

798
00:46:37.411 --> 00:46:43.593
The protein surrounds, it's a bunch of alpha helices that surround a chromophore that is the actual thing that absorbs the light.

799
00:46:44.853 --> 00:46:48.294
That chromophore, retinol, is naturally found.

800
00:46:49.582 --> 00:46:52.906
in mammalian tissue at exactly the right concentration.

801
00:46:53.126 --> 00:46:56.170
So you don't need to keep pumping this cofactor into it, right?

802
00:46:56.370 --> 00:46:57.631
Like we talked about earlier.

803
00:46:57.651 --> 00:46:59.113
Uh-huh.

804
00:46:59.153 --> 00:47:01.055
There's an abundance of the source material.

805
00:47:01.195 --> 00:47:01.596
Exactly.

806
00:47:01.656 --> 00:47:05.860
So all you have to do is somehow figure out how to encode the gene and express that gene.

807
00:47:06.461 --> 00:47:07.302
That's all you got to do.

808
00:47:07.382 --> 00:47:07.662
Okay.

809
00:47:07.682 --> 00:47:07.883
Right?

810
00:47:08.003 --> 00:47:08.684
That's a big deal.

811
00:47:09.004 --> 00:47:10.285
The other thing, that third thing,

812
00:47:10.986 --> 00:47:11.747
Is it reversible?

813
00:47:11.947 --> 00:47:12.187
Yes.

814
00:47:12.628 --> 00:47:14.329
The stimulation is reversible.

815
00:47:14.589 --> 00:47:19.414
So on the left-hand side, we're seeing the thing turn on for a particular amount of time in the blue.

816
00:47:19.654 --> 00:47:20.515
Boom, boom, boom, boom, boom.

817
00:47:20.535 --> 00:47:20.935
It fires.

818
00:47:20.955 --> 00:47:21.556
You turn it off.

819
00:47:21.996 --> 00:47:22.477
It goes back.

820
00:47:22.697 --> 00:47:23.337
It doesn't fire.

821
00:47:23.678 --> 00:47:26.100
On the right-hand side, we actually see that other neurons...

822
00:47:27.451 --> 00:47:36.896
that are not related to the specific one, other neurons can actually stop firing because this particular neuron that started firing is maybe inhibiting that one.

823
00:47:36.916 --> 00:47:37.797
Yep, yep.

824
00:47:37.817 --> 00:47:37.997
Right?

825
00:47:38.017 --> 00:47:42.679
So you can have second order effects and actually start looking at how do networks behave.

826
00:47:42.699 --> 00:47:44.180
This is, yeah, now we can, yeah, exactly.

827
00:47:44.200 --> 00:47:44.360
Right?

828
00:47:44.420 --> 00:47:50.323
You can be like, well, this neuron was definitely affected by this one, but I never expressed channel rhodopsin in this one.

829
00:47:50.403 --> 00:47:53.005
So it's got to be tied together as part of the same network.

830
00:47:53.545 --> 00:47:53.725
Right.

831
00:47:54.526 --> 00:47:55.528
And that same year.

832
00:47:55.548 --> 00:47:57.851
So this was in 2005.

833
00:47:58.231 --> 00:47:59.153
That very same year.

834
00:48:00.775 --> 00:48:03.198
Paper after paper about channel rhodopsin.

835
00:48:03.358 --> 00:48:04.580
Dicerats was the first one.

836
00:48:05.581 --> 00:48:15.047
But that same year, we've got independent studies published within a year that confirms the efficacy of channelrhodopsin in controlling the activity of hippocampal neurons.

837
00:48:15.147 --> 00:48:15.948
That's Lee et al.

838
00:48:16.028 --> 00:48:16.948
in 2005.

839
00:48:17.469 --> 00:48:21.411
Lee is in Stefan Herlitz's lab at Case Western.

840
00:48:21.711 --> 00:48:26.715
We also have motor activity in the embryonic chick spinal cord, also by Lee.

841
00:48:27.655 --> 00:48:30.217
Behavior in C. elegans, which is huge.

842
00:48:30.517 --> 00:48:48.156
see elegans as the model organism for so much developmental studies and reinforcement learning in drosophila by shroll so we're seeing this being applied to different types of organisms living organisms right it's happening this is and this is what we talk about when you say you know

843
00:48:50.561 --> 00:48:52.563
usually get a Nobel.

844
00:48:52.983 --> 00:49:00.388
You want to see all the labs have all the universities have a lab that is doing work in this area.

845
00:49:00.769 --> 00:49:06.413
You're seeing multiple unlocks from some fundamental discovery point.

846
00:49:07.133 --> 00:49:13.814
And then usually you'll then want to see how it then has positive impact at some endpoint.

847
00:49:13.874 --> 00:49:19.636
But at first it has to be something that starts to spread and becomes a fundamental basis for some area of study.

848
00:49:19.656 --> 00:49:20.136
Yeah, yeah.

849
00:49:20.516 --> 00:49:26.177
And Dysorot's lab really took that to the max because he's the first guy who had – he had Ed Boyden.

850
00:49:26.197 --> 00:49:27.017
He had Feng Zhang.

851
00:49:28.678 --> 00:49:35.039
He kind of got lucky in being at Stanford and having these amazing scientists who were postdocs and grad students at the time.

852
00:49:35.059 --> 00:49:35.239
Mm-hmm.

853
00:49:36.330 --> 00:49:40.453
So now the term optogenetics captures the central idea of this technique, right?

854
00:49:40.473 --> 00:49:45.736
You're combining genetic targeting with optimal control, with optical control, right?

855
00:49:45.976 --> 00:49:48.358
Genetics provides that specificity.

856
00:49:48.838 --> 00:49:56.423
Which is such a not like as for me, it just seems so unintuitive or non-intuitive that you would be able to have genetic control.

857
00:49:57.043 --> 00:49:57.303
Yeah.

858
00:49:57.343 --> 00:49:57.624
Yeah.

859
00:49:57.864 --> 00:49:58.224
Yeah.

860
00:49:58.244 --> 00:49:58.344
Yeah.

861
00:50:13.819 --> 00:50:30.681
what's happening at the micro level of biology to the end behavior human behavior that we well we haven't gotten to behavior but like that's like the idea like that's where we're going we're going it's so i'm like it's already so crazy yeah it's already so crazy and we haven't even gotten to the second piece yeah yeah

862
00:50:31.182 --> 00:50:40.627
The second piece is now we want to make this thing better, both in terms of genetic specificity and the general use case, and in terms of we want to make the channelrhodopsin itself better.

863
00:50:40.827 --> 00:50:41.028
Okay?

864
00:50:41.108 --> 00:50:41.308
Okay.

865
00:50:41.328 --> 00:50:47.791
So the first piece is how do we express this in living organisms at a general level?

866
00:50:47.811 --> 00:50:50.353
Like suppose I want to target this particular neuron, right?

867
00:50:50.393 --> 00:50:50.553
Right.

868
00:50:51.073 --> 00:50:51.654
How do I do that?

869
00:50:51.714 --> 00:50:55.717
Well, I can set up a viral vector or I can create transgenic animals.

870
00:50:56.098 --> 00:50:57.299
And that's what these papers did.

871
00:50:58.079 --> 00:51:02.503
This was in 2006 by B. et al., Ishizuka, and Zhang.

872
00:51:02.804 --> 00:51:04.605
These are the three papers that I highlighted.

873
00:51:04.965 --> 00:51:17.577
Specifically what you do is you create a viral vector, a kind of genetic element that has a part that recognizes a specific type of neuron and another part that has the genetics for channel rhodopsin.

874
00:51:18.408 --> 00:51:22.269
This thing goes into the specific neuron that it recognizes.

875
00:51:22.569 --> 00:51:28.190
And then that specific neuron, because it has the other part, the channel rhodopsin part, it's going to express channel rhodopsin.

876
00:51:28.490 --> 00:51:32.130
And now I can very specifically target those neurons.

877
00:51:32.211 --> 00:51:36.811
The beauty of this approach is that the light itself is not specific, right?

878
00:51:36.871 --> 00:51:38.612
I'm lighting up the entire tissue.

879
00:51:39.432 --> 00:51:45.793
But only the neurons that I targeted genetically are the ones that have the ability to actually respond to the light.

880
00:51:46.770 --> 00:51:53.336
It's the delivery mechanism that allows a singular outside light source, but then targeted internal activation.

881
00:51:53.436 --> 00:51:53.656
Yeah.

882
00:51:54.277 --> 00:51:57.339
And that becomes super important in some of the later studies that we're going to see.

883
00:51:57.660 --> 00:52:03.085
Now, the next thing we want to do is also make channelrhodopsin a bit more general.

884
00:52:03.365 --> 00:52:05.807
Because right now, it's cation-specific, right?

885
00:52:05.867 --> 00:52:06.648
It's positive ions.

886
00:52:06.968 --> 00:52:11.432
Positive ions mean that the neuron is going to fire because it's going to...

887
00:52:12.257 --> 00:52:15.519
create the correct type of voltage to make that runaway effect.

888
00:52:15.800 --> 00:52:17.281
What if I want to silence the neurons?

889
00:52:17.301 --> 00:52:25.647
Then I need cholinergic activation, meaning like chlorine ions, negative ions, need to go through so that the voltage becomes opposite and it shuts off the neuron.

890
00:52:26.007 --> 00:52:37.395
So Feng Zhang comes out with this paper, Multimodal Fast Optical Interrogation of Neural Circuitry, again with Carl Diceroth in 2007.

891
00:52:37.696 --> 00:52:40.478
And this is where he introduces the chlorine version

892
00:52:41.363 --> 00:52:42.624
They were going off in the mid-2000s.

893
00:52:42.644 --> 00:52:43.384
They were going off, dude.

894
00:52:43.404 --> 00:52:44.044
Dice Rod's lab.

895
00:52:44.064 --> 00:52:45.345
They were going off.

896
00:52:45.485 --> 00:52:46.926
Yeah, this is two years later.

897
00:52:46.966 --> 00:52:49.847
Yeah, that's the reason I feel like 2007.

898
00:52:49.867 --> 00:52:54.910
Yeah, which means that they had already, like by the time that they had published 2005, they were already in on this.

899
00:52:55.010 --> 00:52:55.330
Right, right.

900
00:52:55.390 --> 00:52:56.851
And they're like, this is clearly what's next.

901
00:52:57.111 --> 00:52:59.933
And over there you can see the cation version.

902
00:53:01.533 --> 00:53:03.514
You turn it on, boom, lots of spiking.

903
00:53:04.115 --> 00:53:06.536
The chlorine version, there's a lot of spiking.

904
00:53:06.636 --> 00:53:08.297
You turn it on, spiking stops.

905
00:53:08.477 --> 00:53:08.717
So good.

906
00:53:09.197 --> 00:53:09.417
Right?

907
00:53:10.078 --> 00:53:10.478
Amazing.

908
00:53:11.618 --> 00:53:20.885
And so that was halorhodopsin, which is a microbial light-driven chlorine pump that they've now used for this exact purpose.

909
00:53:22.447 --> 00:53:22.627
Right?

910
00:53:22.787 --> 00:53:31.894
And now the last thing that Dicerat's lab did that – I mean it's not the last thing, but the last thing in terms of increasing the efficacy of this stuff is they developed cheetah.

911
00:53:32.114 --> 00:53:34.416
That's what we're seeing over here.

912
00:53:34.836 --> 00:53:35.757
Ultra fast cheetah.

913
00:53:37.363 --> 00:53:38.964
optogenetic control in 2010.

914
00:53:39.644 --> 00:53:43.166
On the left-hand side, what you're seeing is normal channel rhodopsin.

915
00:53:43.566 --> 00:53:43.826
Okay?

916
00:53:44.327 --> 00:53:46.208
You see the little blue ticks?

917
00:53:46.408 --> 00:53:46.588
Yep.

918
00:53:46.808 --> 00:53:49.089
That shows when channel rhodopsin went on.

919
00:53:51.050 --> 00:53:55.072
And you can see that, you know, there's a bunch of spikes, but there's extra spikes.

920
00:53:55.092 --> 00:53:56.133
You don't just get one spike.

921
00:53:56.493 --> 00:54:06.598
You get extra spikes, and then there's kind of a trailing off of the membrane potential because, like, the thing depolarizes, but then it kind of just stays because maybe the channel rhodopsin itself...

922
00:54:07.108 --> 00:54:09.135
has a kind of relaxation time, right?

923
00:54:09.456 --> 00:54:11.845
That it opens really quickly, so that's good.

924
00:54:12.356 --> 00:54:14.578
But I want to be very, very precise.

925
00:54:14.898 --> 00:54:20.823
And so this is where we're seeing that kind of like that little hill here on the left in the top charts.

926
00:54:21.084 --> 00:54:25.848
But we basically effectively only want to have a spike with as minimal amount of drop-off as possible.

927
00:54:25.948 --> 00:54:27.829
I don't want a prolonged depolarization.

928
00:54:27.909 --> 00:54:30.151
I want a boom, and then I want you to turn off.

929
00:54:30.211 --> 00:54:32.073
And that's cheetah on the right-hand side.

930
00:54:32.173 --> 00:54:32.473
You see?

931
00:54:32.814 --> 00:54:36.797
Same timescale, but now I'm affecting single spikes.

932
00:54:36.857 --> 00:54:38.939
Isn't that insane?

933
00:54:38.959 --> 00:54:40.080
Yeah, with no prolonged depolarization.

934
00:54:40.420 --> 00:54:43.721
Yeah, and it's also, this is all just...

935
00:54:43.841 --> 00:54:44.021
Right?

936
00:54:44.962 --> 00:54:45.442
It's amazing.

937
00:54:45.482 --> 00:54:50.884
And then on the lower, you can see that, like, sometimes the cell gets, the channel rhodopsin gets tired.

938
00:54:51.024 --> 00:54:51.204
Yeah.

939
00:54:51.244 --> 00:54:52.305
And you get mist spikes.

940
00:54:52.685 --> 00:54:56.506
Because it's been firing for so much, sometimes it gets tired, the cell gets tired.

941
00:54:56.806 --> 00:54:58.407
With cheetah, because it's so precise...

942
00:54:59.825 --> 00:55:00.646
You can just go.

943
00:55:01.207 --> 00:55:02.048
It's like a metronome.

944
00:55:02.469 --> 00:55:04.432
You don't miss any spikes, right?

945
00:55:05.113 --> 00:55:10.361
This is the kind of stuff that is required to really now dig in to how neurons are working.

946
00:55:11.440 --> 00:55:16.842
These faster variants allow more precise control of high-frequency neurons if we want to do that.

947
00:55:17.483 --> 00:55:21.644
It's made possible the production of sustained changes in neural activity.

948
00:55:22.125 --> 00:55:23.345
It's huge, right?

949
00:55:23.385 --> 00:55:30.368
So the evolution of channelrhodopsin, there's a larger point about the nature of technological breakthroughs in biology.

950
00:55:30.448 --> 00:55:36.811
You start with something and you just keep getting better and better because you start seeing the efficacy of the tool.

951
00:55:37.331 --> 00:55:40.901
And you start honing in on exactly what you want for each use case.

952
00:55:41.867 --> 00:55:43.247
And it is a tool.

953
00:55:43.467 --> 00:55:46.348
And then different people are going to use that tool for different purposes.

954
00:55:46.928 --> 00:56:07.673
And then ultimately, we get this better understanding because like everything, like humanity in general, we've just continued to make better tools which have allowed us to then have better outcomes because we better understand the world around us at an increasingly large scale and increasingly small scale.

955
00:56:07.693 --> 00:56:07.793
Yeah.

956
00:56:08.418 --> 00:56:09.279
Absolutely incredible.

957
00:56:10.000 --> 00:56:11.081
This is good.

958
00:56:11.181 --> 00:56:11.941
Yeah.

959
00:56:12.081 --> 00:56:13.503
We've gotten to single neurons.

960
00:56:13.583 --> 00:56:13.783
Yes.

961
00:56:13.903 --> 00:56:14.804
Now let's do circuits.

962
00:56:14.924 --> 00:56:15.124
Yes.

963
00:56:15.244 --> 00:56:15.505
Okay.

964
00:56:15.825 --> 00:56:24.012
So this is what makes optogenetics particularly powerful for dissecting human circuits, for dissecting neural circuits, I should say.

965
00:56:24.333 --> 00:56:29.718
So here in this paper, this is in 2007, what they're showing is...

966
00:56:32.962 --> 00:56:34.643
an optical neural interface.

967
00:56:35.083 --> 00:56:38.125
Here they're expressing it in a living mouse.

968
00:56:39.046 --> 00:56:50.233
They've used that viral vector type thing to target the part of the cortex that controls the whiskers, the mouse.

969
00:56:50.653 --> 00:56:55.236
We're trying to go from now neural circuits to behavior.

970
00:56:55.436 --> 00:56:56.076
This is the link.

971
00:56:56.096 --> 00:56:57.437
This is the paper that links the two.

972
00:56:58.017 --> 00:57:02.100
The idea is if I target the rodent...

973
00:57:03.410 --> 00:57:07.551
motor system that is involved with the whisker deflection, can I make the whiskers deflect?

974
00:57:07.751 --> 00:57:10.272
The whiskers are how a lot of rodents sense their environment.

975
00:57:10.612 --> 00:57:12.973
If I turn these neurons on, can the whiskers deflect?

976
00:57:13.313 --> 00:57:14.933
That's a behavior response.

977
00:57:14.973 --> 00:57:17.474
This is the first time that this happens, okay?

978
00:57:17.794 --> 00:57:18.614
In 2007.

979
00:57:18.874 --> 00:57:23.416
And now at last, we finally have Francis Crick's theoretical vision.

980
00:57:23.636 --> 00:57:23.976
Yeah.

981
00:57:24.276 --> 00:57:32.458
Scientists can now control specific genetically ordained neurons in an alert animal.

982
00:57:33.313 --> 00:58:01.723
using only light right alert awake behaving animal using light i i plead the fifth but and we've talked about this so many times and every time we talk about it i have the same reaction because i we're just so we're so brilliant and i think we're able to do things that part of the benefit of talking through this amount of detail is again going to the specificity of well how is it that you know what's going on in the brain like how like you know or like

983
00:58:03.508 --> 00:58:04.589
science is not a black box.

984
00:58:04.609 --> 00:58:04.809
Yeah.

985
00:58:05.710 --> 00:58:05.930
Right?

986
00:58:06.290 --> 00:58:08.172
And you can literally trace the lineage of this.

987
00:58:08.252 --> 00:58:09.453
Yeah.

988
00:58:09.693 --> 00:58:13.476
And it kind of starts with Crick's kind of far-fetched.

989
00:58:13.496 --> 00:58:13.777
Yeah.

990
00:58:14.137 --> 00:58:15.078
He said it's far-fetched.

991
00:58:15.298 --> 00:58:15.818
At the time.

992
00:58:15.978 --> 00:58:16.879
Yeah.

993
00:58:17.179 --> 00:58:24.065
But I know we still have a little bit more to go, but I just, this is so, the behavior piece is so interesting because it's a little bit,

994
00:58:24.946 --> 00:58:38.413
It's like different than the biological piece in terms of there is a huge leap here to go to behavior and ascribing the behavior to a very particular underlying change.

995
00:58:38.493 --> 00:58:39.453
Yeah, it's causality.

996
00:58:39.613 --> 00:58:39.813
Right.

997
00:58:39.854 --> 00:58:40.854
It's that causality point.

998
00:58:41.114 --> 00:58:46.477
And you want to be very sure when you're saying that's what's happening.

999
00:58:46.637 --> 00:58:48.698
And this tool lets you say that.

1000
00:58:48.718 --> 00:58:49.158
Be very sure.

1001
00:58:49.718 --> 00:58:52.780
It offers a way to selectively manipulate specific populations.

1002
00:58:52.980 --> 00:58:58.101
So I'm going to just real quick go through some of the work that has come about in optogenetics.

1003
00:58:58.421 --> 00:59:01.942
There's obviously a huge repertoire of work that we cannot get to.

1004
00:59:02.443 --> 00:59:07.044
Perhaps another episode that is based on what optogenetics has given us.

1005
00:59:07.424 --> 00:59:08.984
But a few that I'll go into.

1006
00:59:09.825 --> 00:59:10.965
First one, Ilana Witten.

1007
00:59:12.140 --> 00:59:13.921
Carl Diceroth and colleagues.

1008
00:59:14.001 --> 00:59:18.884
Ilana Witten is Ed Witten's daughter, who's now at Princeton University.

1009
00:59:19.224 --> 00:59:20.365
She's a professor there.

1010
00:59:21.065 --> 00:59:32.692
In this particular paper, she used optogenetic approaches to investigate reward-related circuitry and dissected the role of cholinergic neurons that make up only 1% of the local neurons.

1011
00:59:34.093 --> 00:59:40.817
Turns out they play a very specific and significant role in modulating driving behavior and

1012
00:59:41.283 --> 00:59:45.945
like addictive behavior, like in this particular case, cocaine conditioning.

1013
00:59:46.085 --> 00:59:46.445
Okay.

1014
00:59:46.505 --> 00:59:46.705
Right?

1015
00:59:47.065 --> 00:59:55.709
So we can now start targeting specific neurons and be like, yeah, this is actually involved in the kind of addiction and things like that, right?

1016
00:59:56.569 --> 01:00:02.652
You can also do optical manipulation of hippocampal circuits for learning and memory.

1017
01:00:03.153 --> 01:00:04.113
This one's pretty crazy.

1018
01:00:04.614 --> 01:00:12.457
One profound application is the search for the memory engram, which is the physical cellular substrate for encoding a specific memory.

1019
01:00:12.517 --> 01:00:13.798
Like, I've got a specific memory.

1020
01:00:14.058 --> 01:00:15.679
That lives somewhere in the brain, right?

1021
01:00:16.019 --> 01:00:18.200
Very famously, Tsusumu Tonegawa...

1022
01:00:18.800 --> 01:00:26.063
who is a Nobel Prize winner in his own right in 1987 for the discovery of genetic principles underlying antibody diversity.

1023
01:00:27.224 --> 01:00:31.386
He's seeking his second Nobel by searching for the memory engram.

1024
01:00:31.446 --> 01:00:32.086
Good on you.

1025
01:00:32.226 --> 01:00:32.886
Good on you.

1026
01:00:32.946 --> 01:00:33.527
Yep.

1027
01:00:33.567 --> 01:00:34.087
He's trying.

1028
01:00:34.427 --> 01:00:38.149
What he wants to find is the cellular basis for like stored memories.

1029
01:00:38.469 --> 01:00:38.649
Yeah.

1030
01:00:38.749 --> 01:00:39.289
Okay.

1031
01:00:39.429 --> 01:00:42.851
In 2012, he published a paper with Liu and colleagues.

1032
01:00:43.251 --> 01:00:45.852
It's a landmark paper in nature that,

1033
01:00:46.258 --> 01:00:50.341
demonstrates optogenetic activation of a neural ensemble.

1034
01:00:51.422 --> 01:00:55.825
It's associated with a particular experience that triggers the expression of fear.

1035
01:00:56.646 --> 01:00:56.906
Okay?

1036
01:00:57.146 --> 01:00:58.327
This is pretty crazy.

1037
01:00:58.507 --> 01:00:59.388
This is pretty crazy.

1038
01:00:59.628 --> 01:01:06.133
It's a striking demonstration of the possibility of, like, manipulating memory-related neural representations.

1039
01:01:06.433 --> 01:01:09.295
Obviously, a lot of weird implications.

1040
01:01:09.496 --> 01:01:11.317
Also, a lot of clinical implications.

1041
01:01:11.597 --> 01:01:14.739
For example, with PTSD, things like that, right?

1042
01:01:15.080 --> 01:01:15.340
So...

1043
01:01:15.759 --> 01:01:18.401
It is worth being precise about what these experiments establish, though.

1044
01:01:18.721 --> 01:01:21.242
Like, you're activating a neural ensemble.

1045
01:01:21.262 --> 01:01:25.785
You've targeted a bunch of neurons, and you've activated them using this channelrhodopsin trick.

1046
01:01:26.645 --> 01:01:32.248
Now, that does not mean that a complete, fully formed memory is literally stored in that population, okay?

1047
01:01:32.549 --> 01:01:33.829
Because memory is a weird thing.

1048
01:01:33.869 --> 01:01:40.473
It depends on distributed networks, specific synaptic modifications, other brain networks that you might not have targeted.

1049
01:01:40.964 --> 01:01:49.373
What this optogenetics tool is making possible is a causal test that, yeah, this particular set of neurons elicits a behavioral response.

1050
01:01:49.753 --> 01:01:50.014
Okay.

1051
01:01:50.234 --> 01:01:51.655
So I do want to make that distinction.

1052
01:01:51.775 --> 01:01:52.636
That's actually an important one.

1053
01:01:52.656 --> 01:01:52.897
Yeah.

1054
01:01:53.197 --> 01:01:53.377
Yeah.

1055
01:01:53.637 --> 01:01:56.761
And then finally, the last one I'll talk about is a causal link.

1056
01:01:57.281 --> 01:02:07.027
between manipulating dopaminergic activity and prediction and reward and the error from that prediction.

1057
01:02:07.067 --> 01:02:08.127
Like I predict something.

1058
01:02:08.428 --> 01:02:09.408
If I get it, that's good.

1059
01:02:09.468 --> 01:02:10.649
If I don't get it, that's bad.

1060
01:02:10.689 --> 01:02:14.251
The brain fundamentally is a kind of predicting machine, right?

1061
01:02:14.371 --> 01:02:17.173
Oh, are you saying we're just the next token prediction machine?

1062
01:02:17.233 --> 01:02:17.993
Yeah, exactly.

1063
01:02:18.213 --> 01:02:21.015
Well, this particular paper kind of says that.

1064
01:02:27.921 --> 01:02:44.676
corroborated by some of these optogenetic experiments, where what you can do is specifically target these dopaminergic neurons, make them fire when errors happen or when the prediction comes true, and see how the circuit adapts, right?

1065
01:02:44.696 --> 01:02:48.479
So you can literally now causally test this model of prediction and error.

1066
01:02:49.142 --> 01:02:50.383
in this neural circuit.

1067
01:02:50.863 --> 01:02:51.783
That's very nice.

1068
01:02:51.923 --> 01:02:53.864
You can study much more complicated behavior.

1069
01:02:53.904 --> 01:02:57.346
Diceroth, for example, has gone even beyond neuroscience.

1070
01:02:57.726 --> 01:02:57.986
Okay.

1071
01:02:58.186 --> 01:02:58.426
Okay?

1072
01:02:58.726 --> 01:03:08.811
Diceroth has demonstrated that the heart rhythm can affect our emotions because he's developed another type of extra-sensitive opsin called kermine.

1073
01:03:09.031 --> 01:03:10.912
I think it's channelrhodopsin-mine.

1074
01:03:11.832 --> 01:03:17.695
And that thing is so sensitive that you can put it deep within the body and it'll still respond to the light.

1075
01:03:18.637 --> 01:03:27.530
So you can stick it in the heart, in the neural circuitry in the heart that controls heart rhythm, and now you can shine light on it through the muscles and everything.

1076
01:03:27.730 --> 01:03:28.972
Oh, that's really, yeah.

1077
01:03:28.992 --> 01:03:33.939
And then the heart is going to now start beating faster, and that has an effect on the brain in terms of anxiety.

1078
01:03:33.959 --> 01:03:34.240
Mm-hmm.

1079
01:03:34.975 --> 01:03:39.196
It used to be thought that, no, the brain is doing the anxiety part down.

1080
01:03:39.596 --> 01:03:41.837
But actually, there's something going up as well.

1081
01:03:41.897 --> 01:03:43.077
It's a two-way street.

1082
01:03:43.298 --> 01:03:48.199
It's starting to sound like that street in Hollywood where it's like six.

1083
01:03:48.539 --> 01:03:50.100
It's a circular six entryway.

1084
01:03:50.220 --> 01:03:51.980
Oh, yeah, yeah.

1085
01:03:52.020 --> 01:03:52.600
In Beverly Hills.

1086
01:03:52.620 --> 01:03:53.141
In Beverly Hills.

1087
01:03:53.161 --> 01:03:53.581
Yeah, yeah.

1088
01:03:53.661 --> 01:03:54.341
In Beverly Hills.

1089
01:03:54.361 --> 01:03:55.941
That was the worst stop sign in LA.

1090
01:03:56.001 --> 01:03:56.702
Yeah.

1091
01:03:56.782 --> 01:03:57.022
Yeah.

1092
01:03:58.122 --> 01:03:58.542
Exactly.

1093
01:03:58.622 --> 01:04:00.923
I mean, the body is a complicated thing, right?

1094
01:04:00.943 --> 01:04:01.643
100%.

1095
01:04:02.324 --> 01:04:06.745
Optogenetics is most closely associated with neuroscience, but it's got a much broader significance.

1096
01:04:07.085 --> 01:04:10.886
As we just showed, it can be used to talk about the heart.

1097
01:04:11.487 --> 01:04:17.709
Light-sensitive proteins can be engineered to regulate intracellular signaling pathways inside the cell.

1098
01:04:18.289 --> 01:04:26.412
They can be used to control light-sensitive systems like Roe family GTP aces, which are another set of

1099
01:04:27.287 --> 01:04:49.993
cytoskeletal proteins that like can make the cell move and contort in development biology light sensitive tools can be used to investigate how you know spatially localized tissue becomes what they are as the animal grows up right you've got a ball of tissue how does this know to become the head and this become the tail we talk about all the time we talk about that all the time that's a that's a problem that can be used

1100
01:04:50.722 --> 01:04:52.464
that optogenetics can be used for.

1101
01:04:52.784 --> 01:04:54.466
So it's, it's huge, right?

1102
01:04:55.186 --> 01:04:58.389
And finally, um, there is hope for new medical treatments.

1103
01:04:58.690 --> 01:05:03.514
I mean, beyond just the fundamentals of understanding how the brain works, right?

1104
01:05:03.534 --> 01:05:07.318
That is going to clearly give us a better understanding about depression, depression,

1105
01:05:07.678 --> 01:05:12.601
anxiety, PTSD, Alzheimer's, dementia, Parkinson's, you name it, right?

1106
01:05:12.761 --> 01:05:21.185
At the end of the day, understanding the brain at a fundamental level is going to give us insights into all of these ailments, right?

1107
01:05:21.485 --> 01:05:23.366
Because fundamental science, I think, happens first.

1108
01:05:23.726 --> 01:05:26.488
But in a very direct sense, it's also...

1109
01:05:27.930 --> 01:05:33.592
used as the first step towards an optogenetics treatment for blindness.

1110
01:05:34.052 --> 01:05:42.674
So there's ongoing clinical trials that are happening to restore vision in people who have become blind due to retinitis pigmentosa.

1111
01:05:43.895 --> 01:05:45.915
It's a disease that destroys the rods and cones.

1112
01:05:46.595 --> 01:05:52.717
Here, they've inserted channelrhodopsin-like proteins into the eye of a blind person,

1113
01:05:53.445 --> 01:05:56.968
Those neurons then express channel rhodopsin.

1114
01:05:57.289 --> 01:05:59.571
And now the person has regained some kind of vision.

1115
01:06:00.251 --> 01:06:05.496
This is a paper that shows like special glasses that emit light into the person's eye.

1116
01:06:05.977 --> 01:06:08.960
And then that person is able to discern and grasp objects.

1117
01:06:09.400 --> 01:06:11.502
Someone who was, you know, previously blind.

1118
01:06:13.312 --> 01:06:24.617
I mean, again, both things are now concurrently happening, which is the expansion of the basic and fundamental research across all of the areas of the human body, not just the brain.

1119
01:06:25.237 --> 01:06:38.703
But we are in the clinical trial phase for applied clinical or medical outcomes that are impacting people's real lived experience.

1120
01:06:38.723 --> 01:06:38.903
Yeah.

1121
01:06:39.723 --> 01:07:05.199
uh and like obviously we're human beings our eyesight is so such a core part of what how we experience the world exactly if you have a very like effective way to restore that for people um what's not to love what's not to love right and when we predicted when we predicted optogenetics there were a lot of comments that were saying that this is going to clearly lead to zombies and mind control now i'd like to

1122
01:07:06.632 --> 01:07:08.994
Brace that type of speculation.

1123
01:07:09.134 --> 01:07:16.219
I don't think optogenetics is going to lead to zombies because I don't think, you know, we're going to get viral vectors in our brain anytime soon.

1124
01:07:16.780 --> 01:07:19.742
On the other hand, the more we understand about the brain...

1125
01:07:20.638 --> 01:07:23.239
the more nefarious applications can become.

1126
01:07:23.699 --> 01:07:28.321
So, you know, there is some concern here that, like the memory engrams, right?

1127
01:07:28.361 --> 01:07:28.661
Yeah.

1128
01:07:28.961 --> 01:07:39.385
Clearly, of course, it's going to be used for PTSD and clinical approaches to understanding trauma and maybe eliminating trauma, past trauma of an individual.

1129
01:07:39.625 --> 01:07:40.425
That would be the dream.

1130
01:07:40.725 --> 01:07:46.348
But at the same time, you know, with the advent of tools like Neuralink and AI,

1131
01:07:48.633 --> 01:07:54.738
compounding on tools like Neuralink, you can imagine a world where things get very scary.

1132
01:07:54.818 --> 01:07:54.998
Right?

1133
01:07:55.058 --> 01:08:01.603
We've discussed on this podcast several brain-machine interfaces, like the reading inner thoughts.

1134
01:08:01.763 --> 01:08:02.704
Yep.

1135
01:08:02.764 --> 01:08:03.925
That was a good one.

1136
01:08:04.025 --> 01:08:04.325
Yeah.

1137
01:08:04.485 --> 01:08:10.730
So, you know, putting this in the context of all of the neuroscience that we've looked at, it's both...

1138
01:08:11.390 --> 01:08:25.619
something that gives us hope because optogenetics is an incredible tool that is going to let us understand very nitty-gritty causal links between brain neural circuits and behavior and perhaps even maybe consciousness.

1139
01:08:25.639 --> 01:08:28.480
That was another question that was asked during the Nobel Press Conference.

1140
01:08:28.560 --> 01:08:32.603
And the person said, well, not yet, but maybe in the future, maybe in the future, right?

1141
01:08:32.623 --> 01:08:33.543
We'll get some idea.

1142
01:08:33.803 --> 01:08:37.426
But at the same time, you know, we have to be vigilant of

1143
01:08:38.452 --> 01:08:44.774
all of the other weird things that could happen because we learn more and more about the brain because the brain is really the organ that makes us us.

1144
01:08:45.314 --> 01:08:45.454
Yeah.

1145
01:08:45.474 --> 01:08:45.594
Right.

1146
01:08:45.734 --> 01:08:52.656
And if we lose control of that, if we lose sovereignty over the brain, who are we in some sense?

1147
01:08:52.676 --> 01:08:52.856
Right.

1148
01:08:52.916 --> 01:08:53.096
Right.

1149
01:08:53.897 --> 01:09:00.198
So, um, I, I mean, I've been calling optogenetics for a while, even before this podcast.

1150
01:09:00.358 --> 01:09:05.220
So the first year that we did this podcast, I said optogenetics, I recycled this year and I got it right.

1151
01:09:05.240 --> 01:09:05.420
Right.

1152
01:09:05.803 --> 01:09:09.526
And so anyone who said, oh, but you recycled, it's because it was correct.

1153
01:09:10.667 --> 01:09:13.269
Sometimes you got to recycle.

1154
01:09:14.150 --> 01:09:21.276
We have a lot of great previous episodes as well that touch on a variety of different areas in neuroscience.

1155
01:09:21.316 --> 01:09:23.958
And again, it's interesting how these things all connect.

1156
01:09:23.978 --> 01:09:25.940
The myelin sheath stuff we talked about, about...

1157
01:09:26.160 --> 01:09:33.064
how it then create the degradation of that, and then the repair in order to have these action potentials and these ion channels continue to be able to operate.

1158
01:09:33.444 --> 01:09:35.185
We did whole deep episodes on that.

1159
01:09:35.505 --> 01:09:43.029
I think the Nobel Prize medicine deep dive last year and this year are two of my favorites because I think it's –

1160
01:09:44.050 --> 01:09:48.554
Sometimes for other subjects, it's hard for people to get how it relates to themselves.

1161
01:09:49.175 --> 01:09:51.978
And this one can feel very personal.

1162
01:09:52.399 --> 01:09:54.181
So we are going to keep this tight today.

1163
01:09:54.721 --> 01:10:01.328
We will be back again tomorrow early morning for one of the favorites.

1164
01:10:01.949 --> 01:10:04.531
which is going to be the physics Nobel.

1165
01:10:04.591 --> 01:10:10.615
For those of you who are joining us for the first time, welcome to the best science show on the planet.

1166
01:10:10.775 --> 01:10:17.260
You can follow us on all of our socials at FFPpod, as well as check out our episodes and all of the research papers.

1167
01:10:17.560 --> 01:10:18.721
that we covered in this episode.

1168
01:10:18.761 --> 01:10:22.283
If you want to actually take a look at them, you can go to ffppod.com.

1169
01:10:22.363 --> 01:10:26.586
Each one of our episodes always catalogs all of the research papers we cover.

1170
01:10:27.086 --> 01:10:32.669
And if you would like to watch, we're available on YouTube, all the podcast networks to listen as well.

1171
01:10:33.090 --> 01:10:35.591
We appreciate you as always.

1172
01:10:35.631 --> 01:10:36.772
My favorite week of the year.

1173
01:10:37.672 --> 01:10:41.635
How are you feeling about already being on the board after day one?

1174
01:10:41.935 --> 01:10:46.518
I'm so happy because now there's no pressure for like physics and chem, you know?

1175
01:10:46.678 --> 01:10:46.838
Yeah.

1176
01:10:47.485 --> 01:10:47.945
Already done it.

1177
01:10:48.126 --> 01:10:48.766
Already did it, yeah.

1178
01:10:48.786 --> 01:10:49.226
Already done it.

1179
01:10:50.367 --> 01:10:52.009
Maybe we'll get Princeton on the board tomorrow.

1180
01:10:52.029 --> 01:10:53.029
We'll see.

1181
01:10:53.330 --> 01:11:01.756
My name is Asunari, joined as always by my co-host and our resident PhD and the Oracle, Krishna, the Oracle, Chowdhury.

1182
01:11:02.777 --> 01:11:06.380
We will see you all tomorrow for the Physics Nobel Prize.
