1
00:00:00,000 --> 00:00:03,759
Imagine for a moment an entire universe of tiny

2
00:00:03,759 --> 00:00:06,459
incredibly complex machines operating within

3
00:00:06,459 --> 00:00:08,679
your body right now. These aren't you know the

4
00:00:08,679 --> 00:00:10,919
cogs and levers you might picture but rather

5
00:00:10,919 --> 00:00:14,960
intricate biological origami constantly folding

6
00:00:14,960 --> 00:00:17,239
and unfolding performing millions upon millions

7
00:00:17,239 --> 00:00:20,079
of tasks every single second. They're involved

8
00:00:20,079 --> 00:00:22,300
in everything from say the subtle twitch of your

9
00:00:22,300 --> 00:00:24,420
finger to the incredibly complex thought processes

10
00:00:24,420 --> 00:00:26,679
humming away in your brain. You quite literally

11
00:00:26,679 --> 00:00:29,929
rely on these molecular marvels for every single

12
00:00:29,929 --> 00:00:32,469
process that keeps you alive, thinking, and,

13
00:00:32,490 --> 00:00:35,729
well, thriving. So let's dig into this. Today,

14
00:00:35,789 --> 00:00:38,450
our deep dive takes us into the astonishing world

15
00:00:38,450 --> 00:00:41,570
of proteins. Our guiding text, Introduction to

16
00:00:41,570 --> 00:00:43,570
Proteins, Structure, Function, and Motion, makes

17
00:00:43,570 --> 00:00:45,350
it really clear. These aren't just important

18
00:00:45,350 --> 00:00:47,289
molecules. They're absolutely fundamental to

19
00:00:47,289 --> 00:00:49,929
life itself. It's no exaggeration to say that

20
00:00:49,929 --> 00:00:51,729
protein science sits at the very epicenter of

21
00:00:51,729 --> 00:00:53,770
modern biological research. We're talking applications

22
00:00:53,770 --> 00:00:56,130
spanning medicine, agriculture, biotechnology.

23
00:00:56,350 --> 00:00:58,450
It's everywhere. Our mission in this deep dive

24
00:00:58,450 --> 00:01:01,829
is to really unravel how the precise, intricate,

25
00:01:02,049 --> 00:01:04,349
three -dimensional shapes of these proteins dictate

26
00:01:04,349 --> 00:01:06,689
their functions, and why understanding what scientists

27
00:01:06,689 --> 00:01:09,069
call the protein folding problem is one of the

28
00:01:09,069 --> 00:01:11,469
most crucial frontiers in science today. It's

29
00:01:11,469 --> 00:01:14,099
truly a puzzle with massive implications. What's

30
00:01:14,099 --> 00:01:16,680
fascinating here is that our mission goes beyond

31
00:01:16,680 --> 00:01:19,900
just grasping the basics. We aim to give you

32
00:01:19,900 --> 00:01:22,060
a really comprehensive understanding of protein

33
00:01:22,060 --> 00:01:24,319
folding. And yeah, we'll delve into its technical

34
00:01:24,319 --> 00:01:27,519
intricacies, but hopefully in a way that's accessible

35
00:01:27,519 --> 00:01:29,739
and engaging for everyone. Then we're going to

36
00:01:29,739 --> 00:01:32,739
connect those sort of nitty gritty details to

37
00:01:32,739 --> 00:01:35,569
some of the most exciting. real -world applications

38
00:01:35,569 --> 00:01:38,450
that are happening right now, like how this knowledge

39
00:01:38,450 --> 00:01:41,189
is helping design groundbreaking drugs, discover

40
00:01:41,189 --> 00:01:43,829
powerful new antibodies, and even create the

41
00:01:43,829 --> 00:01:46,569
next generation of AI tools that could, frankly,

42
00:01:46,750 --> 00:01:49,920
revolutionize biology. At its very core, this

43
00:01:49,920 --> 00:01:52,459
whole pursuit of understanding structure, dynamics,

44
00:01:52,500 --> 00:01:54,659
and function right down at the atomic level,

45
00:01:54,859 --> 00:01:56,500
well, that's precisely what we call structural

46
00:01:56,500 --> 00:01:59,120
biophysics. It's kind of like deciphering the

47
00:01:59,120 --> 00:02:01,140
universe within us, you know, one molecule at

48
00:02:01,140 --> 00:02:03,459
a time. The marvel of proteins. Yeah. Structure

49
00:02:03,459 --> 00:02:05,920
and function unveiled. Okay, let's start right

50
00:02:05,920 --> 00:02:08,639
at the beginning then. The preface of our source

51
00:02:08,639 --> 00:02:11,539
material lays it out pretty starkly. Proteins

52
00:02:11,539 --> 00:02:13,960
are highly complex molecules that are actively

53
00:02:13,960 --> 00:02:16,300
involved in the most basic and important aspects

54
00:02:16,300 --> 00:02:19,900
of life. When you really stop and think about

55
00:02:19,900 --> 00:02:21,639
that, it's an incredible statement, isn't it?

56
00:02:21,759 --> 00:02:24,020
What exactly are we talking about when we say

57
00:02:24,020 --> 00:02:27,500
basic and important aspects of life? Well, metabolism,

58
00:02:27,759 --> 00:02:30,599
all chemical reactions, just keeping things running,

59
00:02:30,719 --> 00:02:33,439
movement too, right? Everything from tiny cellular

60
00:02:33,439 --> 00:02:35,620
transfer mechanisms all the way up to like...

61
00:02:35,840 --> 00:02:38,639
Your muscles contracting. Proteins form the very

62
00:02:38,639 --> 00:02:41,300
backbone of your body's defense system, fighting

63
00:02:41,300 --> 00:02:43,719
off invaders constantly. They act as cellular

64
00:02:43,719 --> 00:02:46,120
communicators, sending vital signals, and they're

65
00:02:46,120 --> 00:02:48,120
responsible for molecular recognition, knowing

66
00:02:48,120 --> 00:02:50,099
exactly what to interact with and what to ignore.

67
00:02:50,259 --> 00:02:52,759
It's just an astounding array of duties for,

68
00:02:52,900 --> 00:02:55,300
well, a single class of molecule. Yeah, and if

69
00:02:55,300 --> 00:02:58,020
we connect this to the bigger picture, it becomes

70
00:02:58,020 --> 00:03:00,060
abundantly clear that protein functions aren't

71
00:03:00,060 --> 00:03:03,500
just random, right? They are direct, unavoidable

72
00:03:03,500 --> 00:03:07,300
results of their very specific, exquisitely crafted

73
00:03:07,300 --> 00:03:09,240
three -dimensional structure. And not just the

74
00:03:09,240 --> 00:03:11,580
structure, but the subtle dynamic ways those

75
00:03:11,580 --> 00:03:13,400
structures move and change. We call these movements

76
00:03:13,400 --> 00:03:16,860
structural dynamics. So to truly understand how

77
00:03:16,860 --> 00:03:19,639
a protein performs its function, maybe it's catalyzing

78
00:03:19,639 --> 00:03:21,879
a reaction or transporting something or recognizing

79
00:03:21,879 --> 00:03:24,960
a pathogen. We absolutely have to characterize

80
00:03:24,960 --> 00:03:27,400
the physical forces acting on these proteins,

81
00:03:27,539 --> 00:03:29,240
and we need to do that right down at the atomic

82
00:03:29,240 --> 00:03:32,680
level. This profound insight, this... fundamental

83
00:03:32,680 --> 00:03:34,840
realization that structure dictates function.

84
00:03:35,080 --> 00:03:37,520
That's precisely what gave rise to the whole

85
00:03:37,520 --> 00:03:40,599
field of structural biophysics. Our source really

86
00:03:40,599 --> 00:03:42,340
emphasizes this with what it calls an energy

87
00:03:42,340 --> 00:03:44,960
-oriented approach, which basically means we're

88
00:03:44,960 --> 00:03:47,300
not just asking what proteins do, but how the

89
00:03:47,300 --> 00:03:49,400
intricate dance of molecular interactions and

90
00:03:49,400 --> 00:03:51,379
the thermodynamic change as the energy shifts

91
00:03:51,379 --> 00:03:53,639
actually provide the power that drives their

92
00:03:53,639 --> 00:03:55,939
actions. It's all about understanding their unique

93
00:03:55,939 --> 00:03:58,449
energy landscape. You know, the hills and valleys

94
00:03:58,449 --> 00:04:01,169
that guide their every single move. OK, so to

95
00:04:01,169 --> 00:04:03,509
really navigate that energy landscape, we have

96
00:04:03,509 --> 00:04:05,610
to start with the fundamentals, the building

97
00:04:05,610 --> 00:04:08,009
blocks. Our book calls them nature's alphabet,

98
00:04:08,270 --> 00:04:10,370
which I kind of like. And that alphabet is found

99
00:04:10,370 --> 00:04:13,750
in Chapter 2, the 20 natural amino acids. Every

100
00:04:13,750 --> 00:04:16,430
single protein, no matter how big or small, is

101
00:04:16,430 --> 00:04:19,149
just a chain made up of these 20 unique links

102
00:04:19,149 --> 00:04:21,389
strung together like beads on a necklace, basically.

103
00:04:21,930 --> 00:04:24,290
Each amino acid shares a common core structure.

104
00:04:24,759 --> 00:04:27,120
a central carbon atom the alpha carbon bonded

105
00:04:27,120 --> 00:04:30,279
to a hydrogen an amino group alpha amino a carboxyl

106
00:04:30,279 --> 00:04:32,800
group alpha carboxyl and then crucially a unique

107
00:04:32,800 --> 00:04:35,740
side chain And it's this side chain, and only

108
00:04:35,740 --> 00:04:38,019
this side chain, that gives each amino acid its

109
00:04:38,019 --> 00:04:40,500
distinct chemical personality, its role. Right,

110
00:04:40,560 --> 00:04:42,139
and these properties are absolutely critical

111
00:04:42,139 --> 00:04:44,199
for how proteins eventually fold up and do their

112
00:04:44,199 --> 00:04:46,420
jobs. While the central backbone, all the atoms

113
00:04:46,420 --> 00:04:48,139
except that side chain, is chemically identical

114
00:04:48,139 --> 00:04:50,819
across all amino acids, it's that unique side

115
00:04:50,819 --> 00:04:52,800
chain that really differentiates them, makes

116
00:04:52,800 --> 00:04:55,040
them special. Now, you mentioned those alpha

117
00:04:55,040 --> 00:04:57,699
-carboxyl and alpha -amino groups. At physiological

118
00:04:57,699 --> 00:05:00,819
pH, so around 7 in most of your body, the alpha

119
00:05:00,819 --> 00:05:03,100
-carboxyl group, it has a low pKa. Which means

120
00:05:03,100 --> 00:05:05,699
it tends to give up a proton easily, right? It's

121
00:05:05,699 --> 00:05:08,259
negative. Yep, it's protonated to negative. And

122
00:05:08,259 --> 00:05:10,779
conversely, the alpha amino group has a high

123
00:05:10,779 --> 00:05:13,459
pKa, so it tends to hold on to its protons, stays

124
00:05:13,459 --> 00:05:15,819
protonated, positive charge. This charge behavior

125
00:05:15,819 --> 00:05:18,379
is super important for individual amino acids

126
00:05:18,379 --> 00:05:21,160
just floating around. But when they link up in

127
00:05:21,160 --> 00:05:23,740
the protein chain... These backbone charges mostly

128
00:05:23,740 --> 00:05:25,639
get neutralized when they form those peptide

129
00:05:25,639 --> 00:05:28,680
bonds, except, crucially, right at the very beginning

130
00:05:28,680 --> 00:05:30,720
and the very end of the protein sequence, the

131
00:05:30,720 --> 00:05:32,839
termini. Exactly. And that side chain diversity

132
00:05:32,839 --> 00:05:35,420
you just highlighted, that's precisely what enables

133
00:05:35,420 --> 00:05:38,339
such an astonishingly vast range of functional

134
00:05:38,339 --> 00:05:40,560
variation. It really does raise an important

135
00:05:40,560 --> 00:05:42,939
question, doesn't it? How does this relatively

136
00:05:42,939 --> 00:05:46,279
small alphabet, just 20 letters, manage to create

137
00:05:46,279 --> 00:05:49,439
millions of distinct, highly specialized proteins,

138
00:05:49,759 --> 00:05:52,589
each with its own job? Well, it's like nature's

139
00:05:52,589 --> 00:05:55,089
molecular Lego set, isn't it? A brilliantly efficient

140
00:05:55,089 --> 00:05:57,269
system. These subtle chemical differences in

141
00:05:57,269 --> 00:05:59,209
the side chains act like personalities, allowing

142
00:05:59,209 --> 00:06:01,610
proteins to adopt this almost infinite array

143
00:06:01,610 --> 00:06:04,550
of shapes and functions, from rigid scaffolding

144
00:06:04,550 --> 00:06:07,470
to dynamic enzymes. And these side chains, they're

145
00:06:07,470 --> 00:06:09,730
broadly categorized into five basic types based

146
00:06:09,730 --> 00:06:12,230
on their chemical properties. non -polar, polar

147
00:06:12,230 --> 00:06:15,029
in charge, polar charge, aromatic, and then glycine,

148
00:06:15,110 --> 00:06:17,050
which is kind of its own thing. Each type brings

149
00:06:17,050 --> 00:06:20,250
unique properties to the table. Take cysteine,

150
00:06:20,310 --> 00:06:22,269
for instance. It's unique because its thiol group

151
00:06:22,269 --> 00:06:24,990
has sulfur, and it can form disulfide bonds,

152
00:06:25,209 --> 00:06:27,370
these strong covalent links. Super important

153
00:06:27,370 --> 00:06:29,170
for stabilizing proteins, especially outside

154
00:06:29,170 --> 00:06:31,870
the cell. It can also bind metals. Methionine,

155
00:06:31,990 --> 00:06:34,470
another sulfur one, is more involved in weaker

156
00:06:34,470 --> 00:06:36,829
polar interactions, and it can even act as an

157
00:06:36,829 --> 00:06:39,089
antioxidant. Then you have serine and threonine.

158
00:06:39,310 --> 00:06:41,740
They've got hydroxyl groups, and they're absolutely

159
00:06:41,740 --> 00:06:43,519
vital for regulation through phosphorylation,

160
00:06:43,579 --> 00:06:45,779
adding that phosphate group in glycosylation,

161
00:06:45,899 --> 00:06:48,259
which is adding sugar chains. It affects solubility,

162
00:06:48,339 --> 00:06:51,279
recognition, lots of stuff. Tyrosine, with its

163
00:06:51,279 --> 00:06:53,339
aromatic ring and hydroxyl group, it can also

164
00:06:53,339 --> 00:06:55,959
be phosphorylated, sulfated, both key modifications.

165
00:06:56,399 --> 00:06:58,600
And its ring helps stabilize this phenylite ion,

166
00:06:58,899 --> 00:07:01,899
giving it a peak high around 9 .8. Makes it subtly

167
00:07:01,899 --> 00:07:04,620
basic. And what's even cooler, I think, is how

168
00:07:04,620 --> 00:07:07,500
the local environment inside a protein can dramatically

169
00:07:07,500 --> 00:07:09,800
tweak these p -kyan values. You know, we often

170
00:07:09,800 --> 00:07:11,899
think of p -kyan as a fixed property, but inside

171
00:07:11,899 --> 00:07:13,899
the cell, or even in different compartments like

172
00:07:13,899 --> 00:07:17,360
lysosomes, which are really acidic, pH 4 .7 versus

173
00:07:17,360 --> 00:07:20,000
a cytosol at 7 .2. Well, the local pH, or just

174
00:07:20,000 --> 00:07:22,540
having other charged residues nearby, can massively

175
00:07:22,540 --> 00:07:24,959
shift a residue's p -kyan. It can cause a residue

176
00:07:24,959 --> 00:07:27,560
to become ionized, lose its proton when it normally

177
00:07:27,560 --> 00:07:29,720
wouldn't. This is a critical molecular trigger

178
00:07:29,720 --> 00:07:32,779
for loads of enzymes, like in serine and cysteine

179
00:07:32,779 --> 00:07:34,959
protease, there's often a nearby histidine residue.

180
00:07:36,199 --> 00:07:38,079
static interactions, Coulomb interactions, this

181
00:07:38,079 --> 00:07:40,560
histidine helps pull the proton off the serine

182
00:07:40,560 --> 00:07:42,959
or cysteine, which makes that residue a potent

183
00:07:42,959 --> 00:07:45,699
nucleophile, makes it reactive, ready to attack

184
00:07:45,699 --> 00:07:48,120
another molecule and start the reaction. Or,

185
00:07:48,180 --> 00:07:50,420
you know, in some redox enzymes, nearby positive

186
00:07:50,420 --> 00:07:52,860
lysines or arginines can stabilize the negative

187
00:07:52,860 --> 00:07:55,420
charge on a deprotonated cysteine, helping it

188
00:07:55,420 --> 00:07:58,060
transfer electrons. It's just elegant, really

189
00:07:58,060 --> 00:08:00,769
fine -tuned by nature. Yeah. Elegant is a good

190
00:08:00,769 --> 00:08:02,949
word for it. And speaking of those aromatic side

191
00:08:02,949 --> 00:08:05,769
chains, phenylalanine, tyrosine, tryptophan,

192
00:08:05,870 --> 00:08:07,750
histidine too, they have some really special

193
00:08:07,750 --> 00:08:10,569
tricks, don't they? Those unique electron clouds

194
00:08:10,569 --> 00:08:13,050
they have above and below the ring, they allow

195
00:08:13,050 --> 00:08:15,990
for distinct interactions, like pi -pi interactions.

196
00:08:16,370 --> 00:08:18,930
Short -ranged, but powerful, right? Lose strength

197
00:08:18,930 --> 00:08:21,069
really fast with distance, like tiny magnets

198
00:08:21,069 --> 00:08:23,350
almost. Responsible for forming these aromatic

199
00:08:23,350 --> 00:08:25,689
clusters inside proteins, important for binding.

200
00:08:29,899 --> 00:08:32,120
positive side chains like lysine or arginine

201
00:08:32,120 --> 00:08:35,919
plus maybe surprisingly aromatic rings can even

202
00:08:35,919 --> 00:08:38,659
act as hydrogen bond acceptors weaker ones but

203
00:08:38,659 --> 00:08:41,379
still it's interesting the book notes that aromatic

204
00:08:41,379 --> 00:08:43,759
amino acids showed up relatively late in evolution

205
00:08:43,759 --> 00:08:46,200
maybe that explains their sophistication and

206
00:08:46,200 --> 00:08:47,820
why they're common in things like ion channel

207
00:08:47,820 --> 00:08:50,620
gates where that precise geometry is needed for

208
00:08:50,620 --> 00:08:53,019
opening and closing okay so here's where it gets

209
00:08:53,019 --> 00:08:55,220
really interesting how do these individual amino

210
00:08:55,220 --> 00:08:58,279
acids link up and then fold into this incredible

211
00:08:58,279 --> 00:09:00,320
functional machinery. It's a hierarchy, right?

212
00:09:00,419 --> 00:09:02,899
Yeah. Step -by -step thing. From a simple linear

213
00:09:02,899 --> 00:09:06,120
chain to a complex working 3D molecule. First

214
00:09:06,120 --> 00:09:07,899
up, primary structure. That's just the sequence,

215
00:09:07,980 --> 00:09:11,440
the order of the amino acids. Like DNR, CNHL,

216
00:09:11,500 --> 00:09:13,179
that sequence is the primary structure determined

217
00:09:13,179 --> 00:09:15,620
by your DNA. Then we move to secondary structure.

218
00:09:15,860 --> 00:09:19,299
This is about local repeating folds. The common

219
00:09:19,299 --> 00:09:22,440
ones are alpha helices and beta sheets. The secret

220
00:09:22,440 --> 00:09:24,940
here is the peptide bond linking the amino acids.

221
00:09:25,480 --> 00:09:28,500
It's planar, rigid, because of electronic resonance,

222
00:09:28,700 --> 00:09:31,419
limits rotation. So rotation is mainly restricted

223
00:09:31,419 --> 00:09:34,559
to just two angles along the backbone. Phi. and

224
00:09:34,559 --> 00:09:37,299
psi, these constraints are absolutely key. And

225
00:09:37,299 --> 00:09:39,220
to visualize these allowed rotations, scientists

226
00:09:39,220 --> 00:09:41,139
use this amazing tool called the Romachandran

227
00:09:41,139 --> 00:09:43,720
plot. It basically maps out all possible fip

228
00:09:43,720 --> 00:09:45,860
-psi combinations, showing which shapes are okay

229
00:09:45,860 --> 00:09:47,320
sterically and which ones are just impossible

230
00:09:47,320 --> 00:09:49,440
because atoms would crash into each other. The

231
00:09:49,440 --> 00:09:51,379
right -handed alpha helix and beta sheets, they

232
00:09:51,379 --> 00:09:53,539
show up as these dominant, stable regions on

233
00:09:53,539 --> 00:09:55,860
the plot. Glycine, though, is special. Its side

234
00:09:55,860 --> 00:09:57,720
chain is just a hydrogen, so it's way more flexible.

235
00:09:57,899 --> 00:09:59,799
It can occupy more areas on the plot, great for

236
00:09:59,799 --> 00:10:02,259
turns, alpha helices, they're like molecular

237
00:10:02,259 --> 00:10:04,769
spiral staircases. stabilized by hydrogen bonds

238
00:10:04,769 --> 00:10:07,350
inside the helix, between residues that are four

239
00:10:07,350 --> 00:10:09,309
positions apart in the sequence. And they can

240
00:10:09,309 --> 00:10:12,230
be amphipathic, one side water -loving, one side

241
00:10:12,230 --> 00:10:14,570
water -fearing, perfect for sitting in membranes

242
00:10:14,570 --> 00:10:16,350
or bridging different environments. There are

243
00:10:16,350 --> 00:10:18,870
variations too, right? Like the tighter 310 helix,

244
00:10:18,929 --> 00:10:21,450
the wider pi helix, the weird polyprolene helix.

245
00:10:21,570 --> 00:10:23,970
Beta sheets, on the other hand, are more extended,

246
00:10:24,149 --> 00:10:26,990
like a flat zigzag ribbon, stabilized by hydrogen

247
00:10:26,990 --> 00:10:29,529
bonds between adjacent strands. and their side

248
00:10:29,529 --> 00:10:32,070
chains stick out perpendicular to the sheet they

249
00:10:32,070 --> 00:10:34,950
can run parallel or anti -parallel and they often

250
00:10:34,950 --> 00:10:37,629
get this characteristic twist gives them rigidity

251
00:10:37,629 --> 00:10:40,509
okay moving up another level in this amazing

252
00:10:40,509 --> 00:10:43,450
architecture tertiary structure this describes

253
00:10:43,450 --> 00:10:47,029
the unique overall 3d fold of a single polypeptide

254
00:10:47,029 --> 00:10:49,710
chain formed by combinations of those secondary

255
00:10:49,710 --> 00:10:52,759
bits helices and sheets and loops This specific

256
00:10:52,759 --> 00:10:55,159
fold is held together by a whole bunch of forces.

257
00:10:55,700 --> 00:10:58,539
Non -covalent interactions, like van der Waals,

258
00:10:58,600 --> 00:11:00,840
forces those tiny fleeting attractions between

259
00:11:00,840 --> 00:11:03,720
atoms. Electrostatic interactions, two plus and

260
00:11:03,720 --> 00:11:05,700
minus charges attracting. And the big one, the

261
00:11:05,700 --> 00:11:07,159
hydrophobic effect, which we'll definitely get

262
00:11:07,159 --> 00:11:10,220
into more. And sometimes you get strong covalent

263
00:11:10,220 --> 00:11:12,519
disulfide bonds between two cysteines locking

264
00:11:12,519 --> 00:11:15,059
things together, especially important for proteins

265
00:11:15,059 --> 00:11:17,179
outside the cell. What's really remarkable is

266
00:11:17,179 --> 00:11:18,899
how tightly packed the protein's core usually

267
00:11:18,899 --> 00:11:22,399
is, almost crystalline. Optimizes those van der

268
00:11:22,399 --> 00:11:24,379
Waals forces, makes them strong collectively.

269
00:11:24,740 --> 00:11:26,840
But even in that dense core, you can have these

270
00:11:26,840 --> 00:11:30,200
small, empty spaces, packing defects, maybe 30

271
00:11:30,200 --> 00:11:33,440
to 100 cubic angstroms. And far from being flaws,

272
00:11:33,740 --> 00:11:35,879
they can actually be functional, like creating

273
00:11:35,879 --> 00:11:38,279
a path for oxygen and myoglobin to get to its

274
00:11:38,279 --> 00:11:41,220
binding site deep inside. Or they can help mediate

275
00:11:41,220 --> 00:11:43,019
conformational changes in the whole protein.

276
00:11:43,450 --> 00:11:45,649
It's kind of counterintuitive, isn't it? Shows

277
00:11:45,649 --> 00:11:47,490
how what looks like a flaw can actually be a

278
00:11:47,490 --> 00:11:49,789
clever design feature, essential for function.

279
00:11:50,090 --> 00:11:52,549
These secondary elements, they combine into recurring

280
00:11:52,549 --> 00:11:54,889
patterns called motifs, like beta, alpha, beta.

281
00:11:55,230 --> 00:11:58,389
And motifs assemble into larger folds. Some are

282
00:11:58,389 --> 00:12:00,529
so common and vital, they're called superfolds.

283
00:12:00,909 --> 00:12:02,990
like the immunoglobulin fold, the Rossman fold,

284
00:12:03,149 --> 00:12:05,509
the P -loop fold, Tim Beryl, you see them over

285
00:12:05,509 --> 00:12:07,669
and over again across evolution, often doing

286
00:12:07,669 --> 00:12:10,269
really important jobs like binding nucleotides

287
00:12:10,269 --> 00:12:12,690
or forming enzyme active sites. And then finally,

288
00:12:12,769 --> 00:12:14,549
at the very top of this structural hierarchy,

289
00:12:14,990 --> 00:12:18,190
we have quaternary structure. This describes

290
00:12:18,190 --> 00:12:21,190
the precise arrangement when multiple polypeptide

291
00:12:21,190 --> 00:12:23,750
subunits, multiple folded protein chains, come

292
00:12:23,750 --> 00:12:25,750
together to form a larger functional complex.

293
00:12:26,210 --> 00:12:29,250
Think of it like a highly coordinated team. Each

294
00:12:29,250 --> 00:12:31,970
member is an individually folded protein, but

295
00:12:31,970 --> 00:12:33,549
they work together to achieve something bigger,

296
00:12:33,629 --> 00:12:36,830
often more complex. These multi -subunit complexes

297
00:12:36,830 --> 00:12:39,389
are mainly stabilized by non -covalent interactions,

298
00:12:39,730 --> 00:12:42,350
which is key, actually, because it allows for

299
00:12:42,350 --> 00:12:44,990
a dynamic equilibrium. The complex can assemble

300
00:12:44,990 --> 00:12:46,909
and disassemble, which is really important for

301
00:12:46,909 --> 00:12:49,830
regulation, the driving forces, again largely

302
00:12:49,830 --> 00:12:52,629
the hydrophobic effect in electrostatics, often

303
00:12:52,629 --> 00:12:54,570
forming these intricate cooperative interaction

304
00:12:54,570 --> 00:12:57,679
networks. Arginine, with its positive charge,

305
00:12:57,879 --> 00:13:00,279
often acts as a key bridging element, linking

306
00:13:00,279 --> 00:13:02,480
subunits through strong hydrogen bonds and salt

307
00:13:02,480 --> 00:13:04,940
bridges. Evolutionarily, these complexes often

308
00:13:04,940 --> 00:13:07,580
arise through clever tricks like gene duplication.

309
00:13:07,639 --> 00:13:09,980
You copy a gene, the copies diverge a bit, and

310
00:13:09,980 --> 00:13:11,480
now you have slightly different subunits that

311
00:13:11,480 --> 00:13:14,139
can interact. Or through domain swapping, where

312
00:13:14,139 --> 00:13:16,620
bits from two identical proteins actually reach

313
00:13:16,620 --> 00:13:19,259
out and grab onto the other copy, stabilizing

314
00:13:19,259 --> 00:13:22,539
a larger complex. It's efficient. Subtle mutations

315
00:13:22,539 --> 00:13:24,620
in the linker regions can then fine -tune these

316
00:13:24,620 --> 00:13:27,120
interactions, making them stable and evolutionarily

317
00:13:27,120 --> 00:13:29,460
cheap to produce, leveraging existing folds.

318
00:13:29,740 --> 00:13:32,019
Okay, so the 20 standard amino acids are amazing,

319
00:13:32,220 --> 00:13:34,860
no doubt. But they alone can't explain the sheer

320
00:13:34,860 --> 00:13:37,259
diversity millions of unique proteins out there.

321
00:13:37,399 --> 00:13:39,059
That's where post -translational modifications,

322
00:13:39,639 --> 00:13:42,159
PTMs, come in, adding entirely new layers of

323
00:13:42,159 --> 00:13:43,980
complexity and function after the protein is

324
00:13:43,980 --> 00:13:46,639
made. Over 200 different types are known. It

325
00:13:46,639 --> 00:13:49,299
dramatically expands the protein alphabet, especially

326
00:13:49,299 --> 00:13:52,190
in complex organisms like us. Huge source of

327
00:13:52,190 --> 00:13:54,549
diversity in regulation. Phosphorylation is a

328
00:13:54,549 --> 00:13:57,009
fantastic example. Such a massive impact from

329
00:13:57,009 --> 00:13:59,509
a seemingly small change. It involves adding

330
00:13:59,509 --> 00:14:01,750
a bulky, negatively charged phosphate group,

331
00:14:01,850 --> 00:14:05,889
usually to serine, threonine, or tyrosine, using

332
00:14:05,889 --> 00:14:09,110
enzymes called kinases. This really raises an

333
00:14:09,110 --> 00:14:11,389
important question. How does this tiny chemical

334
00:14:11,389 --> 00:14:14,350
addition have such a profound effect? The answer

335
00:14:14,350 --> 00:14:16,830
lies in the electrostatic field. Adding that

336
00:14:16,830 --> 00:14:19,330
big negative charge fundamentally perturbs the

337
00:14:19,330 --> 00:14:22,070
local electrostatics. It changes the protein's

338
00:14:22,070 --> 00:14:24,090
entire energy landscape, basically pushing it

339
00:14:24,090 --> 00:14:26,409
towards new, often lower energy shapes it wouldn't

340
00:14:26,409 --> 00:14:29,230
normally adopt. This can trigger huge conformational

341
00:14:29,230 --> 00:14:31,830
changes. I mean, in glycogen phosphorylase, adding

342
00:14:31,830 --> 00:14:34,129
one phosphate can cause parts of the protein

343
00:14:34,129 --> 00:14:36,730
to shift by like... 50 angstroms. That's enormous

344
00:14:36,730 --> 00:14:39,409
on a molecular scale. And these shifts then alter

345
00:14:39,409 --> 00:14:41,250
the protein's activity, change its binding sites,

346
00:14:41,330 --> 00:14:43,070
or tweak the alignment of catalytic residues

347
00:14:43,070 --> 00:14:46,009
in an enzyme. It's a molecular switch. A tiny

348
00:14:46,009 --> 00:14:47,970
change fundamentally redesigning the protein's

349
00:14:47,970 --> 00:14:50,549
behavior. And phosphorylation is just one. The

350
00:14:50,549 --> 00:14:52,450
cell has this whole toolbox of modifications,

351
00:14:52,610 --> 00:14:55,779
like merestoilation and... palmitoylylation.

352
00:14:56,000 --> 00:14:59,080
These involve sticking on long fatty acid chains,

353
00:14:59,379 --> 00:15:03,059
myrcic acid C14, palmitic acid C16. These act

354
00:15:03,059 --> 00:15:05,059
like molecular anchors, right? Allowing the protein

355
00:15:05,059 --> 00:15:07,879
to embed in or associate tightly with cell membranes.

356
00:15:08,139 --> 00:15:10,399
Crucial for getting proteins to the right place

357
00:15:10,399 --> 00:15:13,679
in the subcellular localization. Palmitoylylation

358
00:15:13,679 --> 00:15:15,899
can also regulate active cysteine residues and

359
00:15:15,899 --> 00:15:18,100
enzymes, kind of dynamically controlling their

360
00:15:18,100 --> 00:15:20,779
function. And these lipid modifications pop up

361
00:15:20,779 --> 00:15:23,000
in disease too. Cancer, diabetes, Alzheimer's,

362
00:15:23,000 --> 00:15:24,460
Huntington's shows how important important they

363
00:15:24,460 --> 00:15:27,299
are for health. Then there's ubiquitination or

364
00:15:27,299 --> 00:15:30,659
isomoyolation, sticking small proteins ubiquitin

365
00:15:30,659 --> 00:15:33,620
or isumo onto target proteins. Often this tagging

366
00:15:33,620 --> 00:15:35,720
is like a molecular kiss of death, marking the

367
00:15:35,720 --> 00:15:38,480
protein for destruction by proteasomes. Proteasomes

368
00:15:38,480 --> 00:15:40,200
are like the cell's recycling centers, breaking

369
00:15:40,200 --> 00:15:42,659
down old, damaged, or misfolded proteins, so

370
00:15:42,659 --> 00:15:45,250
it's a vital... quality control mechanism, ensures

371
00:15:45,250 --> 00:15:47,450
only good proteins stick around, also regulates

372
00:15:47,450 --> 00:15:49,269
protein levels based on the cell's needs. And

373
00:15:49,269 --> 00:15:51,870
we also see glycosylation sticking complex carbohydrate

374
00:15:51,870 --> 00:15:56,120
chains, sugars, onto proteins. Very common, especially

375
00:15:56,120 --> 00:15:58,220
for proteins on the cell surface or secreted

376
00:15:58,220 --> 00:16:01,100
outside the cell. Glycosylation is crucial for

377
00:16:01,100 --> 00:16:02,820
making them soluble, protecting them from being

378
00:16:02,820 --> 00:16:05,159
chewed up by enzymes, and really importantly

379
00:16:05,159 --> 00:16:08,720
for cell recognition, how cells talk to each

380
00:16:08,720 --> 00:16:11,159
other. And just like phosphorylation, adding

381
00:16:11,159 --> 00:16:13,659
these bulky sugar groups profoundly influences

382
00:16:13,659 --> 00:16:16,340
the protein's energy landscape, affects stability,

383
00:16:16,620 --> 00:16:19,600
folding, interactions. These diverse PTMs are

384
00:16:19,600 --> 00:16:21,820
basically the cell's incredibly clever way of

385
00:16:21,820 --> 00:16:23,960
getting maximum functional bang for its genetic

386
00:16:23,960 --> 00:16:26,740
buck, creating millions of distinct protein forms

387
00:16:26,740 --> 00:16:28,679
from a relatively small number of genes. It's

388
00:16:28,679 --> 00:16:31,500
just amazing biological efficiency. The art of

389
00:16:31,500 --> 00:16:34,639
folding. A dance of stability and dynamics. Okay,

390
00:16:34,700 --> 00:16:37,019
so what does all this intricate detail, amino

391
00:16:37,019 --> 00:16:39,580
acids, bonds, PTMs mean for one of the biggest

392
00:16:39,580 --> 00:16:43,240
mysteries in biology? The protein folding problem.

393
00:16:43,559 --> 00:16:46,120
This problem, fundamentally, is about figuring

394
00:16:46,120 --> 00:16:48,860
out exactly how a protein's linear amino acid

395
00:16:48,860 --> 00:16:52,000
sequence dictates its unique functional 3D structure.

396
00:16:52,200 --> 00:16:54,120
It's like, how does this long, seemingly random

397
00:16:54,120 --> 00:16:56,419
string of beads automatically tie itself into

398
00:16:56,419 --> 00:16:58,480
a perfect, intricate knot that can then do a

399
00:16:58,480 --> 00:17:00,940
specific job? It's baffled scientists for ages.

400
00:17:01,299 --> 00:17:03,559
Yeah, and what's truly fascinating here is that

401
00:17:03,559 --> 00:17:06,059
proteins achieve this specific native state,

402
00:17:06,240 --> 00:17:09,039
their working, folded shape, that is only marginally

403
00:17:09,039 --> 00:17:11,329
stable. meaning it's just slightly more stable

404
00:17:11,329 --> 00:17:13,950
energetically than all the unfolded or partially

405
00:17:13,950 --> 00:17:16,430
folded states it could be in. This delicate balance

406
00:17:16,430 --> 00:17:18,910
is absolutely crucial for function. If it were

407
00:17:18,910 --> 00:17:21,190
too stable, it might be too rigid, right? Couldn't

408
00:17:21,190 --> 00:17:24,109
move and interact properly. Too unstable, and

409
00:17:24,109 --> 00:17:26,849
it would just fall apart. The stability is determined

410
00:17:26,849 --> 00:17:30,000
by its free energy. Lower G means more stable,

411
00:17:30,099 --> 00:17:32,440
more energetically preferred. And that free energy

412
00:17:32,440 --> 00:17:36,039
change comes from the famous equation Liege -AAA

413
00:17:36,039 --> 00:17:38,700
-OS. Liege is about changes from bonding and

414
00:17:38,700 --> 00:17:40,920
interactions, and the not -a -query term is all

415
00:17:40,920 --> 00:17:44,160
about changes in order or disorder entropy. The

416
00:17:44,160 --> 00:17:46,500
dominant forces driving this whole folding process

417
00:17:46,500 --> 00:17:48,740
are really remarkable, and they all interplay.

418
00:17:49,170 --> 00:17:51,190
The primary driver is widely considered to be

419
00:17:51,190 --> 00:17:53,309
the hydrophobic effect. This is largely entropy

420
00:17:53,309 --> 00:17:55,509
-driven. It's about increasing the overall disorder

421
00:17:55,509 --> 00:17:58,069
of the water around the protein. See, when nonpolar

422
00:17:58,069 --> 00:17:59,829
water -fearing residues are exposed to water,

423
00:17:59,910 --> 00:18:01,930
they force the water molecules nearby into these

424
00:18:01,930 --> 00:18:04,549
ordered cage -like structures, like tiny icebergs.

425
00:18:04,829 --> 00:18:08,150
By burying these nonpolar bits inside the protein

426
00:18:08,150 --> 00:18:10,210
during folding, you release those ordered water

427
00:18:10,210 --> 00:18:13,230
molecules back into the bulk solvent. They become

428
00:18:13,230 --> 00:18:15,809
free, disordered, and that increases the overall

429
00:18:15,809 --> 00:18:18,309
entropy of the whole system. Which is favorable.

430
00:18:18,730 --> 00:18:21,309
There's also a favorable enthalpy bit, too, as

431
00:18:21,309 --> 00:18:23,369
the released water can now make better hydrogen

432
00:18:23,369 --> 00:18:26,650
bonds in the bulk. But folding isn't free. There's

433
00:18:26,650 --> 00:18:29,490
a big configurational entropy tunnelty. As that

434
00:18:29,490 --> 00:18:31,710
flexible chain folds into one specific shape,

435
00:18:31,910 --> 00:18:34,230
it loses a ton of freedom of movement, becomes

436
00:18:34,230 --> 00:18:37,009
much more ordered. This penalty is huge, maybe

437
00:18:37,009 --> 00:18:39,750
1 .2 to 1 .9 kilocalories per mole per residue,

438
00:18:39,910 --> 00:18:42,670
roughly. But it's beautifully balanced by the

439
00:18:42,670 --> 00:18:44,670
gain from the hydrophobic effect, leading to

440
00:18:44,670 --> 00:18:46,549
that characteristic marginal stability we talked

441
00:18:46,549 --> 00:18:49,630
about. Beyond the hydrophobic effect, other interactions

442
00:18:49,630 --> 00:18:51,829
are definitely key. Van der Waals interactions,

443
00:18:52,109 --> 00:18:54,190
those weak fleeting attractions, they're optimized

444
00:18:54,190 --> 00:18:56,670
in the tightly packed core. Individually weak,

445
00:18:56,769 --> 00:18:58,650
but add them all up in that dense core and they

446
00:18:58,650 --> 00:19:01,549
become collectively very strong. Contribute a

447
00:19:01,549 --> 00:19:04,630
lot to stability, though maybe secondary to hydrophobic

448
00:19:04,630 --> 00:19:07,089
effect as a driving force. Hydrogen bonds are

449
00:19:07,089 --> 00:19:09,329
absolutely critical, especially for forming and

450
00:19:09,329 --> 00:19:12,289
stabilizing secondary structures like alpha helices

451
00:19:12,289 --> 00:19:15,549
and beta sheets and for overall stability. It

452
00:19:15,549 --> 00:19:17,829
costs a lot of energy to bury a polar backbone

453
00:19:17,829 --> 00:19:20,210
group in the core if it doesn't form a hydrogen

454
00:19:20,210 --> 00:19:23,660
bond. Makes the protein unstable. But when they

455
00:19:23,660 --> 00:19:25,920
form those nice, well -ordered H -bonds inside

456
00:19:25,920 --> 00:19:28,599
the core, shielded from water, that cost drops

457
00:19:28,599 --> 00:19:31,940
dramatically. Makes them highly favorable. And

458
00:19:31,940 --> 00:19:34,819
finally, electrostatic interactions plus -minus

459
00:19:34,819 --> 00:19:37,299
attractions. Most common on the surface, interacting

460
00:19:37,299 --> 00:19:39,740
with water and other polar groups. But they can

461
00:19:39,740 --> 00:19:41,559
be important inside the core too, especially

462
00:19:41,559 --> 00:19:43,599
if shielded from water by other groups, or even

463
00:19:43,599 --> 00:19:46,039
trapped water molecules. And their strength really

464
00:19:46,039 --> 00:19:47,859
depends on the local environment, the dielectric

465
00:19:47,859 --> 00:19:50,619
constant. How well the surroundings dampen electrical

466
00:19:50,619 --> 00:19:53,430
forces. It's also really important to distinguish

467
00:19:53,430 --> 00:19:55,650
between thermodynamic stability as the folded

468
00:19:55,650 --> 00:19:58,549
state, the lowest energy state, and kinetic stability.

469
00:19:58,910 --> 00:20:01,289
Kinetic stability is more about resistance to

470
00:20:01,289 --> 00:20:03,650
unfolding. How long does it take for the protein

471
00:20:03,650 --> 00:20:06,049
to fall apart? Kinetically stable proteins often

472
00:20:06,049 --> 00:20:08,450
tend to be more rigid, maybe lots of beta sheets,

473
00:20:08,549 --> 00:20:11,130
which are very sturdy, or extensive long -range

474
00:20:11,130 --> 00:20:13,589
interactions locking things down. These features...

475
00:20:13,890 --> 00:20:16,150
physically block solvent molecules from getting

476
00:20:16,150 --> 00:20:18,089
to weak spots and starting the unfolding process.

477
00:20:18,630 --> 00:20:21,690
Also, things like oligomerization forming complexes

478
00:20:21,690 --> 00:20:24,950
and certain PTMs like glycosylation can significantly

479
00:20:24,950 --> 00:20:27,529
boost kinetic stability, reinforce the structure,

480
00:20:27,690 --> 00:20:30,309
make it last longer. Now, folding inside a cell

481
00:20:30,309 --> 00:20:33,029
in vivo, that presents a whole different set

482
00:20:33,029 --> 00:20:35,109
of challenges, right? The cytoplasm where lots

483
00:20:35,109 --> 00:20:37,970
of proteins are made is incredibly crowded, dense.

484
00:20:38,700 --> 00:20:41,200
packed with molecules. So it's really easy for

485
00:20:41,200 --> 00:20:43,420
unfolded or partially folded proteins to just

486
00:20:43,420 --> 00:20:45,599
bump into each other nonspecifically. Get tangled

487
00:20:45,599 --> 00:20:47,829
up. Aggregate, these aggregates, these clumps,

488
00:20:47,970 --> 00:20:50,369
they can actually be energetically more stable

489
00:20:50,369 --> 00:20:51,990
than the correctly folded protein sometimes.

490
00:20:52,250 --> 00:20:54,910
But they're usually inactive. And often they're

491
00:20:54,910 --> 00:20:56,970
outright toxic to the cell. It's a big hurdle.

492
00:20:57,109 --> 00:20:59,609
Exactly. And this is precisely where chaperones

493
00:20:59,609 --> 00:21:02,390
come in. They are the cell's essential, incredibly

494
00:21:02,390 --> 00:21:05,349
sophisticated folding assistants. Chaperonins,

495
00:21:05,430 --> 00:21:08,950
like the famous Groel Hsp60 complexes, are these

496
00:21:08,950 --> 00:21:11,869
large cage -like protein machines. Their main

497
00:21:11,869 --> 00:21:14,930
job is to help refold unfolded or misfolded proteins.

498
00:21:15,470 --> 00:21:17,349
They basically grab these substrate proteins,

499
00:21:17,549 --> 00:21:19,910
the ones needing help, and encapsulate them inside

500
00:21:19,910 --> 00:21:22,490
this protected cage -like structure. By doing

501
00:21:22,490 --> 00:21:24,769
that, they provide this isolated safe space where

502
00:21:24,769 --> 00:21:26,650
the protein can try to fold or refold without

503
00:21:26,650 --> 00:21:28,329
getting tangled up with all the other sticky

504
00:21:28,329 --> 00:21:30,869
molecules in the crowded cytoplasm. Prevents

505
00:21:30,869 --> 00:21:34,029
aggregation. And it's not passive. It's an active

506
00:21:34,029 --> 00:21:37,089
process. Requires energy from ATP hydrolysis

507
00:21:37,089 --> 00:21:39,650
to power its conformational changes. The asymmetrical

508
00:21:39,650 --> 00:21:42,089
model for Grohl -Grohes shows this. beautifully.

509
00:21:42,470 --> 00:21:44,690
Unfolded protein goes into the GroEL cavity,

510
00:21:44,950 --> 00:21:47,750
then ATP binding in the GroES cap. protein trap

511
00:21:47,750 --> 00:21:50,069
it inside gives it a chance maybe a second or

512
00:21:50,069 --> 00:21:52,329
third chance to fold correctly in isolation now

513
00:21:52,329 --> 00:21:53,970
there's this interesting ongoing debate about

514
00:21:53,970 --> 00:21:56,430
chaperones do they just prevent aggregation letting

515
00:21:56,430 --> 00:21:58,970
the protein fold passively on its own time or

516
00:21:58,970 --> 00:22:00,849
do they actually accelerate the folding process

517
00:22:00,849 --> 00:22:03,849
itself some studies suggest they can maybe speeding

518
00:22:03,849 --> 00:22:06,269
it up by as much as 10 times compared to folding

519
00:22:06,269 --> 00:22:09,509
alone even without aggregation the idea is that

520
00:22:09,509 --> 00:22:11,930
by confining the protein inside the cage they

521
00:22:11,930 --> 00:22:14,829
reduce the entropy penalty of folding favor more

522
00:22:14,829 --> 00:22:17,710
compact native -like states over randomly unfolded

523
00:22:17,710 --> 00:22:19,329
ones. It's a really cool example of how cells

524
00:22:19,329 --> 00:22:21,750
optimize a fundamentally tricky physical process,

525
00:22:22,029 --> 00:22:23,990
making sure proteins get to their functional

526
00:22:23,990 --> 00:22:26,549
shapes efficiently. Yet it's not just about the

527
00:22:26,549 --> 00:22:28,569
final shape, is it? It's the intricate dance

528
00:22:28,569 --> 00:22:31,309
it performs. Which brings us to protein dynamics

529
00:22:31,309 --> 00:22:34,660
and lostry. We really need to shift from thinking

530
00:22:34,660 --> 00:22:36,859
of proteins as rigid statues to seeing them as

531
00:22:36,859 --> 00:22:39,099
incredibly dynamic, constantly shifting between

532
00:22:39,099 --> 00:22:41,180
different shapes or conformational substates.

533
00:22:41,420 --> 00:22:43,640
Driven by the thermal energy just buzzing around,

534
00:22:43,859 --> 00:22:46,130
these motions are diverse too. Hinge motions,

535
00:22:46,410 --> 00:22:48,829
like parts swinging open and closed, rotations

536
00:22:48,829 --> 00:22:51,309
of whole domains, translations, parts sliding,

537
00:22:51,490 --> 00:22:53,609
even bits of secondary structure folding and

538
00:22:53,609 --> 00:22:56,609
unfolding locally, like myosin, the motor protein.

539
00:22:56,829 --> 00:22:58,930
Hinge motions are literally how it converts ATP

540
00:22:58,930 --> 00:23:01,569
energy into mechanical force for muscles. Many

541
00:23:01,569 --> 00:23:04,210
enzymes use hinge motions too, opening and closing

542
00:23:04,210 --> 00:23:06,930
access to their active sites, regulating catalysis.

543
00:23:07,369 --> 00:23:10,049
There are even databases, like MoleMoveDB, dedicated

544
00:23:10,049 --> 00:23:12,529
to visualizing these amazing movements. Shows

545
00:23:12,529 --> 00:23:15,190
proteins are truly living machines. And what's

546
00:23:15,190 --> 00:23:17,250
truly fascinating here is how these built -in

547
00:23:17,250 --> 00:23:20,410
dynamics lead to allostery. This is a key regulatory

548
00:23:20,410 --> 00:23:23,069
thing where binding a ligand, any molecule that

549
00:23:23,069 --> 00:23:25,609
binds at one spot, the allosteric site, influences

550
00:23:25,609 --> 00:23:27,890
what happens at a distant spot, often the active

551
00:23:27,890 --> 00:23:30,809
site. Our understanding here has evolved. From

552
00:23:30,809 --> 00:23:33,569
early, more rigid models, like monads, we've

553
00:23:33,569 --> 00:23:35,730
moved toward the conformational selection model.

554
00:23:35,990 --> 00:23:38,769
This model suggests proteins aren't just static

555
00:23:38,769 --> 00:23:40,690
waiting for a ligand to bind and force a change.

556
00:23:40,829 --> 00:23:43,349
Instead, they're constantly, spontaneously sampling

557
00:23:43,349 --> 00:23:45,779
different shapes. like a dancer trying out poses.

558
00:23:46,099 --> 00:23:49,299
The ligand then preferentially binds to and stabilizes

559
00:23:49,299 --> 00:23:51,819
one of those existing shapes, shifting the energy

560
00:23:51,819 --> 00:23:54,480
landscape to favor that state. Koshland's induced

561
00:23:54,480 --> 00:23:56,980
fit model still plays a role too, describing

562
00:23:56,980 --> 00:23:59,259
how sometimes slower adjustments happen to optimize

563
00:23:59,259 --> 00:24:01,759
binding, maybe involving latch residues that

564
00:24:01,759 --> 00:24:04,460
click into place. Hemoglobin is the classic example,

565
00:24:04,619 --> 00:24:06,839
right? Carrying oxygen in your blood. Oxygen

566
00:24:06,839 --> 00:24:08,859
binding makes it switch from a low affinity T

567
00:24:08,859 --> 00:24:11,509
state tense to a high affinity R state. Relax.

568
00:24:11,730 --> 00:24:14,049
When oxygen binds the iron in the heme group,

569
00:24:14,130 --> 00:24:16,549
the iron moves slightly. This pulls on a key

570
00:24:16,549 --> 00:24:19,069
amino acid, the proximal histidine, which then

571
00:24:19,069 --> 00:24:21,450
pulls on a whole helix segment, the F -helix.

572
00:24:21,890 --> 00:24:24,029
This chain reaction changes the whole central

573
00:24:24,029 --> 00:24:26,430
cavity, making it easier for more oxygen to bind.

574
00:24:26,950 --> 00:24:29,529
Cooperative binding. And there's another histidine,

575
00:24:29,609 --> 00:24:32,029
the distal one, that helps stabilize the bound

576
00:24:32,029 --> 00:24:34,289
oxygen and actually makes it harder for poisonous

577
00:24:34,289 --> 00:24:37,349
carbon monoxide to bind as tightly. Clever design.

578
00:24:37,859 --> 00:24:40,359
These allosteric changes have huge functional

579
00:24:40,359 --> 00:24:42,799
effects everywhere in biology. They can open

580
00:24:42,799 --> 00:24:45,980
or close active sites, like we said, or subtly

581
00:24:45,980 --> 00:24:48,480
reshape them to change binding affinity. For

582
00:24:48,480 --> 00:24:51,019
instance, in the enzyme DHFR, loops that aren't

583
00:24:51,019 --> 00:24:53,200
even part of the active site transmit these small

584
00:24:53,200 --> 00:24:55,619
coordinated motions across about eight angstroms

585
00:24:55,619 --> 00:24:58,160
to the active site. This tiny motion actually

586
00:24:58,160 --> 00:25:00,799
stabilizes the transition state, making the enzyme

587
00:25:00,799 --> 00:25:02,980
way more efficient. Another amazing example,

588
00:25:03,180 --> 00:25:06,799
the SH2 domain of a kinase. A tiny flip, a cis

589
00:25:06,799 --> 00:25:09,319
-trans isomerization of just one peptide bond

590
00:25:09,319 --> 00:25:11,880
completely redesigns the protein surface, changes

591
00:25:11,880 --> 00:25:14,700
which ligands it prefers to bind. It just hammers

592
00:25:14,700 --> 00:25:16,339
home that function isn't just about the static

593
00:25:16,339 --> 00:25:18,119
shape. It's deeply tied to dynamic flexibility

594
00:25:18,119 --> 00:25:20,440
and responsiveness. And adding yet another layer

595
00:25:20,440 --> 00:25:23,460
of... frankly mind -bending complexity we have

596
00:25:23,460 --> 00:25:27,680
intrinsically unordered proteins iups or intrinsically

597
00:25:27,680 --> 00:25:30,539
disordered regions idrs these are proteins or

598
00:25:30,539 --> 00:25:32,240
bits of proteins that just don't have a stable

599
00:25:32,240 --> 00:25:35,220
3d structure on their own in isolation instead

600
00:25:35,220 --> 00:25:38,240
they exist as these dynamic ensembles of rapidly

601
00:25:38,240 --> 00:25:40,920
interconverting shapes like a constantly shifting

602
00:25:40,920 --> 00:25:44,339
cloud This disorder comes from a really flat

603
00:25:44,339 --> 00:25:46,660
energy land state. No single shape is strongly

604
00:25:46,660 --> 00:25:48,819
preferred. What's super intriguing is that these

605
00:25:48,819 --> 00:25:51,460
IUPs often do fold into a stable structure, but

606
00:25:51,460 --> 00:25:53,799
only when they bind to a specific partner molecule.

607
00:25:54,160 --> 00:25:56,119
Now, this binding might have an entropy cost,

608
00:25:56,299 --> 00:25:58,200
maybe leading to lower affinity than structured

609
00:25:58,200 --> 00:26:00,720
interactions, but it also allows for incredibly

610
00:26:00,720 --> 00:26:03,640
rapid binding and release. This fast on -off

611
00:26:03,640 --> 00:26:06,279
rate is crucial for complex processes like signaling

612
00:26:06,279 --> 00:26:08,400
cascades or regulatory networks where things

613
00:26:08,400 --> 00:26:10,980
need to happen quickly and transiently. Medically,

614
00:26:11,140 --> 00:26:13,190
IUPs are hugely relevant. They're involved in

615
00:26:13,190 --> 00:26:15,509
loads of diseases, cancer, neurodegenerative

616
00:26:15,509 --> 00:26:18,230
diseases like Alzheimer's and Parkinson's, cardiovascular

617
00:26:18,230 --> 00:26:21,289
disease, diabetes, makes them important but also

618
00:26:21,289 --> 00:26:23,910
really challenging drug targets. Their flexibility

619
00:26:23,910 --> 00:26:26,950
presents unique problems and opportunities for

620
00:26:26,950 --> 00:26:29,710
drug design. But sometimes this whole intricate

621
00:26:29,710 --> 00:26:32,799
dance of folding just goes terribly wrong. with

622
00:26:32,799 --> 00:26:35,000
devastating results. And here's where it gets

623
00:26:35,000 --> 00:26:37,319
really interesting in a rather grim way, because

624
00:26:37,319 --> 00:26:39,460
the consequences of protein misfolding can be

625
00:26:39,460 --> 00:26:42,319
severe, leads to these debilitating, often fatal

626
00:26:42,319 --> 00:26:45,019
diseases. We group these under the term amyloidosis.

627
00:26:45,119 --> 00:26:47,700
Over 40 human diseases characterized by proteins

628
00:26:47,700 --> 00:26:49,940
clumping together, aggregating into these toxic,

629
00:26:50,039 --> 00:26:52,420
fine fibers called amyloids. Yeah, think about

630
00:26:52,420 --> 00:26:55,200
really devastating conditions like Alzheimer's

631
00:26:55,200 --> 00:26:58,240
disease that involves beta amyloid and tau proteins

632
00:26:58,240 --> 00:27:01,099
misfolding and clumping in the brain, or Parkinson's

633
00:27:01,099 --> 00:27:03,319
disease characterized by aggregates of alpha

634
00:27:03,319 --> 00:27:07,210
-synuclein. Huntington's disease, another neurodegenerative

635
00:27:07,210 --> 00:27:10,589
one, involves the Huntington protein. The general

636
00:27:10,589 --> 00:27:13,069
idea is that misfolded proteins often expose

637
00:27:13,069 --> 00:27:15,930
more hydrophobic, sticky patches on their surface.

638
00:27:16,170 --> 00:27:19,130
This makes them prone to clumping together, aggregating,

639
00:27:19,130 --> 00:27:21,109
and precipitating out as these highly ordered

640
00:27:21,109 --> 00:27:24,670
toxic amyloid fibrils. Exactly how they're toxic

641
00:27:24,670 --> 00:27:26,910
isn't always fully clear. It probably varies.

642
00:27:27,109 --> 00:27:28,950
But it's generally thought of as a harmful gain

643
00:27:28,950 --> 00:27:30,930
of function the aggregates themselves do damage

644
00:27:30,930 --> 00:27:32,730
rather than just losing the original protein's

645
00:27:32,730 --> 00:27:35,720
function. But this raises a really provocative

646
00:27:35,720 --> 00:27:39,079
question. Are amyloid proteins always bad? There's

647
00:27:39,079 --> 00:27:41,319
this fascinating hypothesis from Nobel laureate

648
00:27:41,319 --> 00:27:44,539
Eric Kandel about a protein called CPEB. In sea

649
00:27:44,539 --> 00:27:46,420
slugs, it seems to be involved in long -term

650
00:27:46,420 --> 00:27:48,740
memory, and it aggregates in a controlled way,

651
00:27:48,880 --> 00:27:50,779
forming a functional amyloid structure that seems

652
00:27:50,779 --> 00:27:52,480
crucial for maintaining the synaptic connections

653
00:27:52,480 --> 00:27:55,480
that store memories. So... It highlights a potential

654
00:27:55,480 --> 00:27:58,259
evolutionary advantage for controlled aggregation

655
00:27:58,259 --> 00:28:01,660
in specific contexts. Maybe it's not always pathological.

656
00:28:01,920 --> 00:28:04,519
It's a powerful reminder that even the dark side

657
00:28:04,519 --> 00:28:06,660
of protein folding might have unexpected evolutionary

658
00:28:06,660 --> 00:28:09,099
roots and even serve vital functions sometimes.

659
00:28:09,519 --> 00:28:12,079
Designing the future. Protein folding's impact

660
00:28:12,079 --> 00:28:14,779
on technology and health. Okay, so we've journeyed

661
00:28:14,779 --> 00:28:16,900
through this incredible complexity amino acids.

662
00:28:17,339 --> 00:28:20,720
folding dynamics even misfolding now let's shift

663
00:28:20,720 --> 00:28:23,220
gears let's look at the impact of all this knowledge

664
00:28:23,220 --> 00:28:26,119
how is this fundamental understanding of protein

665
00:28:26,119 --> 00:28:28,759
folding and dynamics letting us design the future

666
00:28:28,759 --> 00:28:31,400
especially when it comes to making better more

667
00:28:31,400 --> 00:28:36,000
targeted drugs 3 .1 drug design and action tailoring

668
00:28:36,000 --> 00:28:38,420
molecules to targets right so at the heart of

669
00:28:38,420 --> 00:28:40,460
pretty much all drug action is protein ligand

670
00:28:40,460 --> 00:28:43,400
interaction that intricate dance often called

671
00:28:43,400 --> 00:28:45,599
lock and key but we know it's way more dynamic

672
00:28:45,599 --> 00:28:48,380
than that right ligands small molecules peptides

673
00:28:48,380 --> 00:28:50,380
even bigger macromolecules they bind to proteins

674
00:28:50,380 --> 00:28:53,539
to kick off or regulate countless processes catalysis

675
00:28:53,539 --> 00:28:57,319
regulation transport signaling you name it and

676
00:28:57,319 --> 00:28:59,539
the driving forces they largely echo what we

677
00:28:59,539 --> 00:29:01,700
talked about for folding itself The hydrophobic

678
00:29:01,700 --> 00:29:04,180
effect is huge, pushing water out of the binding

679
00:29:04,180 --> 00:29:06,099
site when the ligand fits in. That's favorable

680
00:29:06,099 --> 00:29:08,420
entropy, plus favorable enthalpy from the released

681
00:29:08,420 --> 00:29:10,519
water making better bonds elsewhere. Van der

682
00:29:10,519 --> 00:29:12,500
Waals interactions contribute a lot to the tight

683
00:29:12,500 --> 00:29:14,819
packing, the snug fit, ensuring good affinity.

684
00:29:15,079 --> 00:29:17,960
And electrostatic interactions, plus minus attractions,

685
00:29:18,200 --> 00:29:20,779
they're absolutely crucial for specificity. Making

686
00:29:20,779 --> 00:29:23,400
sure the drug hits the right target. Highly geometry

687
00:29:23,400 --> 00:29:26,240
dependent. When we talk drug design, two things

688
00:29:26,240 --> 00:29:29,359
are paramount. Affinity and specificity. Affinity

689
00:29:29,359 --> 00:29:32,539
is how strongly the ligand binds. Kuntz and colleagues

690
00:29:32,539 --> 00:29:34,660
estimated an upper limit for typical affinity,

691
00:29:34,880 --> 00:29:38,079
maybe around 13 kilomole. Enzymes, though, can

692
00:29:38,079 --> 00:29:40,140
achieve way stronger binding, but specifically

693
00:29:40,140 --> 00:29:42,319
to the transition state of the reaction they

694
00:29:42,319 --> 00:29:44,960
catalyze. A brilliant idea from Linus Pauling.

695
00:29:45,079 --> 00:29:47,440
By binding super tightly to that high -energy

696
00:29:47,440 --> 00:29:49,539
transition state, they speed up the reaction

697
00:29:49,539 --> 00:29:51,400
immensely. Then the products release quickly.

698
00:29:52,259 --> 00:29:53,740
Specificity, on the other hand, is all about

699
00:29:53,740 --> 00:29:55,859
hitting the right target and only the right target.

700
00:29:55,940 --> 00:29:58,240
It comes down to the number of precise, geometry

701
00:29:58,240 --> 00:30:01,259
-dependent interactions. But what's truly fascinating

702
00:30:01,259 --> 00:30:03,299
here, and actually a massive challenge for drug

703
00:30:03,299 --> 00:30:06,259
design, is this thing called binding promiscuity.

704
00:30:06,819 --> 00:30:08,940
You know, we often think of proteins as being

705
00:30:08,940 --> 00:30:12,240
super specific, but the reality is many proteins

706
00:30:12,240 --> 00:30:14,420
can actually bind multiple different ligands.

707
00:30:14,970 --> 00:30:17,569
This can happen in a few ways. Different ligands

708
00:30:17,569 --> 00:30:20,230
might use the same binding site residues, just

709
00:30:20,230 --> 00:30:22,210
with different strengths. Or they might bind

710
00:30:22,210 --> 00:30:24,789
to totally different residues nearby. Or maybe

711
00:30:24,789 --> 00:30:26,930
use a mix of overlapping and distinct sites.

712
00:30:27,230 --> 00:30:29,670
And it can even involve different dynamic shapes,

713
00:30:29,809 --> 00:30:32,569
different conformations of the binding site itself.

714
00:30:32,970 --> 00:30:35,369
Maybe certain residues are accessible only in

715
00:30:35,369 --> 00:30:38,730
one shape, or their charge changes. This promiscuity.

716
00:30:39,650 --> 00:30:42,119
It's tricky. On one hand, it makes designing

717
00:30:42,119 --> 00:30:45,140
hyperspecific drugs hard, could lead to off -target

718
00:30:45,140 --> 00:30:47,200
effects. But on the other hand, it suggests proteins

719
00:30:47,200 --> 00:30:49,599
might be quite tunable. Maybe we can design novel

720
00:30:49,599 --> 00:30:52,200
molecules to bind them therapeutically. And maybe

721
00:30:52,200 --> 00:30:54,240
some drugs work by hitting multiple targets weekly.

722
00:30:54,619 --> 00:30:57,400
It complicates the simple picture, but it's probably

723
00:30:57,400 --> 00:31:00,059
closer to reality in a complex cell. Yeah, that

724
00:31:00,059 --> 00:31:02,039
complexity is key. And this deep understanding

725
00:31:02,039 --> 00:31:04,440
isn't just for drugs targeting existing proteins.

726
00:31:04,579 --> 00:31:07,019
It's also revolutionizing protein engineering

727
00:31:07,019 --> 00:31:10,319
for industrial enzymes. Enzymes are vital industrial

728
00:31:10,319 --> 00:31:13,039
workhorses, right? Used in food production cheese,

729
00:31:13,400 --> 00:31:16,019
beer textiles, making pharmaceuticals precisely.

730
00:31:16,319 --> 00:31:18,859
Basically tiny, efficient biological factories,

731
00:31:19,059 --> 00:31:22,059
super precise, super fast. When we want to tweak

732
00:31:22,059 --> 00:31:23,859
these enzymes for industry, there are two main

733
00:31:23,859 --> 00:31:26,180
ways. Rational design or directed evolution.

734
00:31:26,660 --> 00:31:28,720
Rational design uses our knowledge of structure.

735
00:31:28,839 --> 00:31:31,440
and function plus computer modeling to make specific

736
00:31:31,440 --> 00:31:35,079
targeted amino acid changes, like protecting

737
00:31:35,079 --> 00:31:37,559
sibilucin, an enzyme, and detergents from oxidation

738
00:31:37,559 --> 00:31:39,960
by swapping a key methionine for something like

739
00:31:39,960 --> 00:31:43,220
serine or alanine makes it last longer. Directed

740
00:31:43,220 --> 00:31:45,380
evolution is more like accelerated natural selection

741
00:31:45,380 --> 00:31:47,619
in a test tube. You make random mutations in

742
00:31:47,619 --> 00:31:49,500
the gene and screen like crazy for the improved

743
00:31:49,500 --> 00:31:51,799
trait you want. Let evolution do the heavy lifting.

744
00:31:52,019 --> 00:31:54,220
And when it comes to engineering enzymes, especially

745
00:31:54,220 --> 00:31:57,160
for industry, a really common goal is improving

746
00:31:57,160 --> 00:32:00,329
thermostability. making them tougher, able to

747
00:32:00,329 --> 00:32:02,970
withstand heat. Based on everything we know about

748
00:32:02,970 --> 00:32:05,349
folding, several clever strategies have emerged.

749
00:32:05,950 --> 00:32:09,190
One is to regenify the unfolded state. Make it

750
00:32:09,190 --> 00:32:11,369
harder to unravel in the first place. You can

751
00:32:11,369 --> 00:32:13,609
do that by putting in rigid prolines, getting

752
00:32:13,609 --> 00:32:16,410
rid of flexible glycines, adding bulky aromatics

753
00:32:16,410 --> 00:32:19,490
for better packing, or even engineering new disulfide

754
00:32:19,490 --> 00:32:22,269
bonds to staple things together. Another way

755
00:32:22,269 --> 00:32:25,000
is to strengthen the hydrophobic core. Optimize

756
00:32:25,000 --> 00:32:27,240
the packing of nonpolar residues, make it more

757
00:32:27,240 --> 00:32:29,720
resistant to water getting in. You could also

758
00:32:29,720 --> 00:32:32,299
boost electrostatic interactions, add more salt

759
00:32:32,299 --> 00:32:34,599
bridges, those plus -minus attractions inside

760
00:32:34,599 --> 00:32:37,539
the protein. Promoting alpha helices, which are

761
00:32:37,539 --> 00:32:39,400
intrinsically stable structures, can help too.

762
00:32:39,680 --> 00:32:42,359
And finally, optimizing interactions on the surface

763
00:32:42,359 --> 00:32:45,059
with the solvent can prevent irreversible aggregation,

764
00:32:45,099 --> 00:32:47,619
which is a big problem under stress. These are

765
00:32:47,619 --> 00:32:49,619
all rational strategies derived directly from

766
00:32:49,619 --> 00:32:51,720
understanding the forces that hold proteins together.

767
00:32:52,170 --> 00:32:54,529
Our deep grasp of protein -ligand interactions

768
00:32:54,529 --> 00:32:56,950
also underpins the whole drug discovery pipeline,

769
00:32:57,210 --> 00:32:59,490
which is, you know, a long, tough road from idea

770
00:32:59,490 --> 00:33:02,150
to medicine. It involves several key stages.

771
00:33:02,309 --> 00:33:04,549
First, identifying targets proteins involved

772
00:33:04,549 --> 00:33:07,710
in disease. GPCRs, G -protein coupled receptors,

773
00:33:07,829 --> 00:33:10,210
are huge targets here because they do so much

774
00:33:10,210 --> 00:33:13,049
signaling. Then finding hits molecules showing

775
00:33:13,049 --> 00:33:15,829
some initial activity. Then leads more promising

776
00:33:15,829 --> 00:33:19,569
compounds and tons of lead optimization. Computational

777
00:33:19,569 --> 00:33:21,390
drug design is absolutely indispensable now.

778
00:33:21,690 --> 00:33:24,130
One key approach is the pharmacopher hypothesis

779
00:33:24,130 --> 00:33:26,630
that's ligand -based. You look at known drugs

780
00:33:26,630 --> 00:33:29,170
for a target, identify common features, H -bond

781
00:33:29,170 --> 00:33:32,049
donor susceptors, aromatic rings, charges, and

782
00:33:32,049 --> 00:33:34,589
use that pattern to design new molecules. Then

783
00:33:34,589 --> 00:33:36,910
there's docking, which is receptor -based. You

784
00:33:36,910 --> 00:33:39,150
computationally try to fit potential drug molecules

785
00:33:39,150 --> 00:33:41,589
into the protein's binding site, predict how

786
00:33:41,589 --> 00:33:43,970
they bind, how strongly. Docking is tough, though.

787
00:33:44,009 --> 00:33:45,849
You have to sample all the possible shapes of

788
00:33:45,849 --> 00:33:48,670
the flexible ligand in the protein. For millions

789
00:33:48,670 --> 00:33:51,329
of ligands. That's trillions and trillions of

790
00:33:51,329 --> 00:33:54,089
possibilities. Huge computational cost. Once

791
00:33:54,089 --> 00:33:56,170
you have leads, lead optimization refines them

792
00:33:56,170 --> 00:33:57,869
using more accurate free energy calculations,

793
00:33:58,029 --> 00:34:00,730
things like alchemical methods or MMPB, GBSA,

794
00:34:00,750 --> 00:34:03,250
trying to boost affinity, improve selectivity,

795
00:34:03,369 --> 00:34:05,730
and crucially, reduce toxicity by minimizing

796
00:34:05,730 --> 00:34:09,809
off -target binding. Now, how do drugs actually

797
00:34:09,809 --> 00:34:12,210
interact with enzymes, often as inhibitors, stopping

798
00:34:12,210 --> 00:34:14,730
them working? Competitive inhibitors bind right

799
00:34:14,730 --> 00:34:16,690
in the active site, blocking the natural substrate.

800
00:34:17,119 --> 00:34:19,679
They increase the apparent Km, but Vmax stays

801
00:34:19,679 --> 00:34:21,380
the same if you flood the system with substrate.

802
00:34:21,760 --> 00:34:24,059
Non -competitive or allosteric inhibitors bind

803
00:34:24,059 --> 00:34:26,380
somewhere else. They change the enzyme's shape

804
00:34:26,380 --> 00:34:29,099
or activity, reducing Vmax or affinity, and substrate

805
00:34:29,099 --> 00:34:31,260
concentration doesn't really overcome it. Let's

806
00:34:31,260 --> 00:34:34,480
look at examples. Ibuprofen, the NSAID. It inhibits

807
00:34:34,480 --> 00:34:37,400
the cyclooxygenic COX by binding its active site,

808
00:34:37,579 --> 00:34:40,000
mimicking arachidonic acid, stops inflammatory

809
00:34:40,000 --> 00:34:43,079
signals. ACE inhibitors like Captopril, Lisinopril,

810
00:34:43,159 --> 00:34:45,860
crucial for blood pressure. They inhibit angiotensin

811
00:34:45,860 --> 00:34:47,780
-converting enzyme by mimicking the transition

812
00:34:47,780 --> 00:34:50,119
state of its substrate. Blocks the enzyme, lowers

813
00:34:50,119 --> 00:34:52,019
blood pressure. Really clever mimicry. And speaking

814
00:34:52,019 --> 00:34:54,059
of prime targets, G -protein -coupled receptors,

815
00:34:54,380 --> 00:34:57,280
GPCRs, are absolutely central. These amazing

816
00:34:57,280 --> 00:34:59,820
membrane proteins, like 20 -30 % of our genome,

817
00:34:59,920 --> 00:35:02,559
involved in so much. Our senses, smell, taste.

818
00:35:02,880 --> 00:35:05,280
Internal signaling via hormones, neurotransmitters.

819
00:35:05,300 --> 00:35:07,340
Their structure is classic. Seven alpha helices

820
00:35:07,340 --> 00:35:09,380
weaving through the membrane. And the mechanism

821
00:35:09,380 --> 00:35:11,980
is fascinating. An agonist ligand binds outside,

822
00:35:12,219 --> 00:35:17,739
causes a shape change inside. Exactly. And this

823
00:35:17,739 --> 00:35:20,420
complex dynamic mechanism makes them incredibly

824
00:35:20,420 --> 00:35:24,000
hot targets for drugs. Orthosteric drugs bind

825
00:35:24,000 --> 00:35:26,199
where the natural ligand binds, like adrenaline

826
00:35:26,199 --> 00:35:29,920
binding to its receptor. Aripiprazole, the antipsychotic,

827
00:35:30,079 --> 00:35:33,300
is a partial agonist at the D2 dopamine receptor's

828
00:35:33,300 --> 00:35:35,920
orthosteric site. But the really exciting area

829
00:35:35,920 --> 00:35:39,340
now, arguably, is allosteric drugs. These bind

830
00:35:39,340 --> 00:35:40,960
somewhere else on the receptor, not the main

831
00:35:40,960 --> 00:35:43,349
site. and they modulate activity, maybe make

832
00:35:43,349 --> 00:35:46,210
the natural ligand bind better or worse, or stabilize

833
00:35:46,210 --> 00:35:49,329
an active or inactive state. This raises an important

834
00:35:49,329 --> 00:35:51,610
question, though. Why bother with allosteric

835
00:35:51,610 --> 00:35:53,610
drugs if we have orthosteric ones? Well, several

836
00:35:53,610 --> 00:35:56,130
reasons. They offer an alternative, potentially

837
00:35:56,130 --> 00:35:57,889
allowing for more subtle fine -tuned control.

838
00:35:58,329 --> 00:36:00,929
And crucially, allosteric sites tend to be less

839
00:36:00,929 --> 00:36:02,869
conserved across different receptor subtypes

840
00:36:02,869 --> 00:36:05,369
than the main orthosteric sites. This means allosteric

841
00:36:05,369 --> 00:36:08,030
drugs can often be more subtype -selective. Hit

842
00:36:08,030 --> 00:36:09,989
only the receptor involved in the disease, hopefully

843
00:36:09,989 --> 00:36:12,829
fewer side effects. Maraviroc, the HIV drug,

844
00:36:12,889 --> 00:36:15,829
is a great example. It targets CCR5, a coreceptor

845
00:36:15,829 --> 00:36:18,730
HIV uses to get into cells. It binds allosterically

846
00:36:18,730 --> 00:36:21,590
and stabilizes an inactive shape of CCR5. It

847
00:36:21,590 --> 00:36:23,590
doesn't block the main binding site, just locks

848
00:36:23,590 --> 00:36:26,579
the receptor so HIV can't use it. blocks a key

849
00:36:26,579 --> 00:36:29,579
transmission switch movement. Powerful example

850
00:36:29,579 --> 00:36:31,760
of manipulating protein dynamics therapeutically,

851
00:36:31,840 --> 00:36:34,340
locking the door on the virus. Wow. Okay. This

852
00:36:34,340 --> 00:36:36,440
deep dive into folding also connects directly

853
00:36:36,440 --> 00:36:38,619
to our immune system, doesn't it? Especially

854
00:36:38,619 --> 00:36:41,360
in antibody discovery and engineering. Antibodies

855
00:36:41,360 --> 00:36:44,260
themselves are amazing proteins. About 10 % of

856
00:36:44,260 --> 00:36:46,239
our genome is involved in making them and related

857
00:36:46,239 --> 00:36:48,980
immune molecules. They, along with T -cell receptors,

858
00:36:49,119 --> 00:36:51,219
are the immune system's precision targeting systems,

859
00:36:51,579 --> 00:36:54,960
recognizing specific invaders. T -cells recognize

860
00:36:54,960 --> 00:36:57,840
specific peptide fragments presented by MHC molecules

861
00:36:57,840 --> 00:37:00,699
like a molecular wanted poster. And B -cells

862
00:37:00,699 --> 00:37:02,539
produce these incredibly specific antibodies

863
00:37:02,539 --> 00:37:08,159
to... It really is. And what's truly fascinating

864
00:37:08,159 --> 00:37:10,280
here is this kind of dual nature of antibodies.

865
00:37:10,699 --> 00:37:12,840
They're famous for being highly specific to their

866
00:37:12,840 --> 00:37:15,440
target antigen. But their binding sites also

867
00:37:15,440 --> 00:37:18,039
show functional versatility. They can sometimes

868
00:37:18,039 --> 00:37:20,940
bind multiple related things. Promiscuity again.

869
00:37:21,099 --> 00:37:24,019
So how can they be both specific and promiscuous?

870
00:37:24,360 --> 00:37:26,980
It's a great question. Research suggests certain

871
00:37:26,980 --> 00:37:29,440
residues like bulky tyrosines contribute a lot

872
00:37:29,440 --> 00:37:30,960
to the overall binding energy of the affinity.

873
00:37:31,340 --> 00:37:33,559
Makes the interaction strong. But it's often

874
00:37:33,559 --> 00:37:35,940
the shorter chain polar residues, serines, threonase,

875
00:37:36,139 --> 00:37:38,820
aspartates, asparagens, that provide the exquisite

876
00:37:38,820 --> 00:37:42,139
specificity. Through precise, short -range, electrostatic

877
00:37:42,139 --> 00:37:44,519
interactions and hydrogen bonds that really depend

878
00:37:44,519 --> 00:37:46,539
on the exact shape of the antigen fitting perfectly.

879
00:37:46,739 --> 00:37:49,659
So it's this interplay. Strong general binding

880
00:37:49,659 --> 00:37:52,639
driven by some residues, fine -tuned specificity

881
00:37:52,639 --> 00:37:55,340
driven by others. And understanding this is critical

882
00:37:55,340 --> 00:37:58,510
for antibody maturation and engineering. Your

883
00:37:58,510 --> 00:38:00,530
immune system naturally matures antibodies over

884
00:38:00,530 --> 00:38:03,030
time through somatic mutations, tiny changes

885
00:38:03,030 --> 00:38:05,030
in the antibody genes that improve binding affinity.

886
00:38:05,389 --> 00:38:08,070
But interestingly, most of these beneficial mutations

887
00:38:08,070 --> 00:38:10,369
don't directly touch the antigen. Instead, they

888
00:38:10,369 --> 00:38:12,829
cause subtle global conformational changes within

889
00:38:12,829 --> 00:38:16,070
the antibody structure itself. These shifts slightly

890
00:38:16,070 --> 00:38:19,170
reposition other residues, effectively increasing

891
00:38:19,170 --> 00:38:21,409
the number of productive interaction sites with

892
00:38:21,409 --> 00:38:23,909
the antigen, making the binding tighter overall.

893
00:38:24,539 --> 00:38:27,920
Like, in one antibody studied, 48G7, maturation

894
00:38:27,920 --> 00:38:31,880
involved residues R96 and TR33, forming new interactions

895
00:38:31,880 --> 00:38:34,159
with the antigen after these conformational shifts,

896
00:38:34,340 --> 00:38:37,059
even though they didn't interact initially. Boosted

897
00:38:37,059 --> 00:38:40,079
affinity hugely. This dynamic adaptability is

898
00:38:40,079 --> 00:38:42,659
incredible. It means we can now potentially custom

899
00:38:42,659 --> 00:38:44,559
-design therapeutic antibodies with enhanced

900
00:38:44,559 --> 00:38:47,199
specificity and affinity for all sorts of targets.

901
00:38:47,500 --> 00:38:49,960
Imagine engineering antibodies for specific cancer

902
00:38:49,960 --> 00:38:52,900
cells or neutralizing specific viruses or toxins

903
00:38:52,900 --> 00:38:55,440
with pinpoint accuracy. It's a massive frontier

904
00:38:55,440 --> 00:38:57,880
in precision medicine. Engineering our own defenders.

905
00:38:58,219 --> 00:39:00,699
Okay, let's unpack this even further then. Beyond

906
00:39:00,699 --> 00:39:02,579
therapeutics, the real cutting edge seems to

907
00:39:02,579 --> 00:39:05,420
be where protein science meets AI. artificial

908
00:39:05,420 --> 00:39:08,239
intelligence especially in finally cracking that

909
00:39:08,239 --> 00:39:11,340
protein folding puzzle how are these new ai tools

910
00:39:11,340 --> 00:39:13,320
and prediction methods changing the game for

911
00:39:13,320 --> 00:39:16,880
understanding and even designing proteins 3 .3

912
00:39:16,880 --> 00:39:20,099
new ai tools and prediction solving the folding

913
00:39:20,099 --> 00:39:23,380
puzzle yeah historically figuring out protein

914
00:39:23,380 --> 00:39:25,659
structures experimentally it's been incredibly

915
00:39:25,659 --> 00:39:28,639
hard laborious expensive think about membrane

916
00:39:28,639 --> 00:39:31,219
proteins vital but super tricky to work with

917
00:39:31,219 --> 00:39:33,980
only about 3 .5 percent of known structures back

918
00:39:33,980 --> 00:39:36,119
in 2017 were membrane proteins because they're

919
00:39:36,119 --> 00:39:38,460
so hard to crystallize for x -ray methods this

920
00:39:38,460 --> 00:39:40,340
bottleneck really spurred the development of

921
00:39:40,340 --> 00:39:43,059
computational prediction methods ab initio approaches

922
00:39:43,059 --> 00:39:44,760
try to predict structure just from the amino

923
00:39:44,760 --> 00:39:47,239
acid sequence alone Note templates. Traditionally

924
00:39:47,239 --> 00:39:49,559
very tough, computationally heavy, but useful

925
00:39:49,559 --> 00:39:51,920
for refining experimental data or for dynamic

926
00:39:51,920 --> 00:39:54,860
simulations. Comparative modeling, or homology

927
00:39:54,860 --> 00:39:56,900
modeling, has usually been the go -to. Find a

928
00:39:56,900 --> 00:39:59,039
similar protein with a known structure, a template,

929
00:39:59,219 --> 00:40:01,539
and use that as a starting point. Accuracy really

930
00:40:01,539 --> 00:40:03,619
depends on how similar the sequences are. Over

931
00:40:03,619 --> 00:40:06,179
50 % identity. Usually gives a pretty good model.

932
00:40:06,480 --> 00:40:08,880
But it still struggles with flexible loops, often

933
00:40:08,880 --> 00:40:11,070
important for function. and getting the side

934
00:40:11,070 --> 00:40:13,929
chain positions exactly right. Fold recognition

935
00:40:13,929 --> 00:40:16,539
or threading is slightly different. Tries to

936
00:40:16,539 --> 00:40:18,539
match a sequence to known folds in the database

937
00:40:18,539 --> 00:40:20,699
based on shared properties, secondary structure

938
00:40:20,699 --> 00:40:23,340
propensities, solvent exposure, etc. Sees if

939
00:40:23,340 --> 00:40:25,880
the sequence fits a known fold. Right. And if

940
00:40:25,880 --> 00:40:27,739
you connect this to the bigger picture, the field

941
00:40:27,739 --> 00:40:29,760
is rapidly moving towards these incredibly powerful

942
00:40:29,760 --> 00:40:32,460
integrated experimental computational approaches.

943
00:40:33,039 --> 00:40:35,280
It's a really exciting trend. You take lower

944
00:40:35,280 --> 00:40:37,559
resolution experimental data, maybe from cryo

945
00:40:37,559 --> 00:40:41,820
-EMs, access NMR, CD, free T stuff, that doesn't

946
00:40:41,820 --> 00:40:43,260
give you the full atomic picture on its own.

947
00:40:43,659 --> 00:40:45,800
And you combine it with computational modeling.

948
00:40:46,119 --> 00:40:48,800
The experimental data provides crucial geometric

949
00:40:48,800 --> 00:40:51,380
constraints, narrows down the search space dramatically.

950
00:40:51,659 --> 00:40:53,860
This massively improves prediction accuracy,

951
00:40:54,079 --> 00:40:56,300
especially for big floppy complex things like

952
00:40:56,300 --> 00:40:58,980
the yeast nuclear pore complex. And cryo -EM

953
00:40:58,980 --> 00:41:01,360
itself has had a revolution. Single particle

954
00:41:01,360 --> 00:41:03,480
cryo -EM can now get near atomic resolution,

955
00:41:03,800 --> 00:41:06,900
like 4 -5 angstroms sometimes. Unimaginable just

956
00:41:06,900 --> 00:41:09,760
a decade ago. Molecular dynamics, MD simulations

957
00:41:09,760 --> 00:41:12,579
are also playing a huge role. Simulating atoms

958
00:41:12,579 --> 00:41:14,739
moving over time. Gets amazing insights into

959
00:41:14,739 --> 00:41:17,340
dynamics, folding pathways, function. Early MD

960
00:41:17,340 --> 00:41:20,199
studies, like on the small protein BPTI, had

961
00:41:20,199 --> 00:41:22,320
to use super simplified models, like two spheres

962
00:41:22,320 --> 00:41:24,159
per amino acid. Just not enough computer power.

963
00:41:24,360 --> 00:41:26,940
But now, distributed computing like folding at

964
00:41:26,940 --> 00:41:29,579
home, specialized hardware like the Anton supercomputer.

965
00:41:30,000 --> 00:41:32,480
We can simulate millisecond long folding events

966
00:41:32,480 --> 00:41:35,119
for small proteins. Extract real thermodynamic

967
00:41:35,119 --> 00:41:37,719
data. MD relies on these force fields parameter

968
00:41:37,719 --> 00:41:40,300
sets describing how atoms interact. Developing

969
00:41:40,300 --> 00:41:42,929
accurate force fields is ongoing work. Especially

970
00:41:42,929 --> 00:41:45,210
tricky for non -protein molecules, like drugs,

971
00:41:45,349 --> 00:41:47,090
and there's always the trade -off with solvent.

972
00:41:47,329 --> 00:41:49,849
Model every single water molecule explicitly.

973
00:41:50,269 --> 00:41:53,110
More accurate but way slower. Use implicit solvent

974
00:41:53,110 --> 00:41:56,670
models. Faster but less precise. Scientists constantly

975
00:41:56,670 --> 00:41:59,219
juggle these choices. Which brings us right up

976
00:41:59,219 --> 00:42:02,380
to the current revolution, AI's role in smashing

977
00:42:02,380 --> 00:42:04,980
through these old barriers. Artificial intelligence,

978
00:42:05,219 --> 00:42:06,960
especially deep learning, is just dramatically

979
00:42:06,960 --> 00:42:09,659
accelerating structure prediction, hitting accuracy

980
00:42:09,659 --> 00:42:12,139
levels that honestly seem like science fiction

981
00:42:12,139 --> 00:42:14,579
not long ago. Breakthroughs came from realizing

982
00:42:14,579 --> 00:42:16,639
you can predict structure from sequence variation

983
00:42:16,639 --> 00:42:19,639
across species. By looking at which residues

984
00:42:19,639 --> 00:42:22,239
co -evolve change together, you can infer which

985
00:42:22,239 --> 00:42:24,519
ones must be close together in the 3D structure.

986
00:42:24,940 --> 00:42:28,099
provides powerful constraints. And the CASP competition

987
00:42:28,099 --> 00:42:30,340
critical assessment of protein structure prediction

988
00:42:30,340 --> 00:42:33,059
running since 94 has been a huge driver, blind

989
00:42:33,059 --> 00:42:35,960
tests pushing the algorithms forward. The progress

990
00:42:35,960 --> 00:42:38,619
in the last few CASP rounds, particularly with

991
00:42:38,619 --> 00:42:40,219
things like AlphaFold from DeepMind, has been

992
00:42:40,219 --> 00:42:42,420
absolutely staggering. It's fundamentally changed

993
00:42:42,420 --> 00:42:44,900
structural biology. Yeah, it really has. So this

994
00:42:44,900 --> 00:42:46,840
raises an important question. What does this

995
00:42:46,840 --> 00:42:48,960
future of increasingly accurate, increasingly

996
00:42:48,960 --> 00:42:52,039
fast protein structure prediction, powered by

997
00:42:52,039 --> 00:42:55,500
AI, actually mean for us? The implications are,

998
00:42:55,539 --> 00:42:58,420
well, they're vast, transformative, really exciting.

999
00:42:58,539 --> 00:43:01,360
In drug design, it means much faster, much more

1000
00:43:01,360 --> 00:43:03,980
accurate identification of targets and lead compounds.

1001
00:43:04,460 --> 00:43:06,880
We might be able to design drugs against targets

1002
00:43:06,880 --> 00:43:09,179
previously considered undruggable because their

1003
00:43:09,179 --> 00:43:11,860
structures were unknown or too complex. Opens

1004
00:43:11,860 --> 00:43:14,269
up whole new therapeutic avenues. For antibody

1005
00:43:14,269 --> 00:43:16,369
engineering, imagine custom -designing antibodies

1006
00:43:16,369 --> 00:43:18,949
with unprecedented precision for specific pathogens

1007
00:43:18,949 --> 00:43:22,170
or specific cancer markers, accelerating immunotherapy

1008
00:43:22,170 --> 00:43:24,969
and vaccine development hugely. In synthetic

1009
00:43:24,969 --> 00:43:26,849
biology and protein engineering, we could start

1010
00:43:26,849 --> 00:43:29,309
creating novel proteins with functions that don't

1011
00:43:29,309 --> 00:43:31,769
even exist in nature, optimizing enzymes for

1012
00:43:31,769 --> 00:43:34,349
biofuels, creating new biomaterials. The possibilities

1013
00:43:34,349 --> 00:43:36,750
are wild. And fundamentally, it means getting

1014
00:43:36,750 --> 00:43:38,789
much deeper insights into misfolding diseases.

1015
00:43:39,579 --> 00:43:42,340
Understand exactly how proteins go wrong, paving

1016
00:43:42,340 --> 00:43:44,280
the way for targeted therapies that fix the root

1017
00:43:44,280 --> 00:43:47,280
cause, not just symptoms. being able to predict

1018
00:43:47,280 --> 00:43:49,960
structure accurately and quickly, unlocking biological

1019
00:43:49,960 --> 00:43:52,760
design at its most fundamental level. So we've

1020
00:43:52,760 --> 00:43:54,760
really gone on a journey, haven't we? From the

1021
00:43:54,760 --> 00:43:57,699
basic building blocks, the amino acids, through

1022
00:43:57,699 --> 00:44:00,539
that intricate dynamic dance of protein folding,

1023
00:44:00,639 --> 00:44:03,599
and now to how this fundamental biology is being

1024
00:44:03,599 --> 00:44:06,940
harnessed to reshape medicine, biotech, even

1025
00:44:06,940 --> 00:44:09,840
AI itself. It's just a testament to the incredible

1026
00:44:09,840 --> 00:44:11,760
power of understanding things at the molecular

1027
00:44:11,760 --> 00:44:16,039
level and, well, human ingenuity. Indeed. What

1028
00:44:16,039 --> 00:44:18,119
we've unpacked today, understanding this precise

1029
00:44:18,119 --> 00:44:20,880
language of protein structure and dynamics right

1030
00:44:20,880 --> 00:44:22,900
down at the atomic level, it isn't just academic

1031
00:44:22,900 --> 00:44:25,820
curiosity, is it? It genuinely unlocks incredible

1032
00:44:25,820 --> 00:44:27,960
potential, potential to tackle some of the biggest

1033
00:44:27,960 --> 00:44:30,500
challenges we face in health, maybe environment,

1034
00:44:30,579 --> 00:44:32,960
technology. It gives us the tools to not just

1035
00:44:32,960 --> 00:44:35,280
observe life, but potentially to engineer it

1036
00:44:35,280 --> 00:44:37,119
with a level of control we've never had before.

1037
00:44:37,360 --> 00:44:39,880
So what does this all mean for you listening

1038
00:44:39,880 --> 00:44:42,920
right now? If every living cell really is this

1039
00:44:42,920 --> 00:44:45,380
complex symphony orchestrated by millions of

1040
00:44:45,380 --> 00:44:47,639
perfectly folded proteins all working in harmony,

1041
00:44:48,559 --> 00:44:50,579
What new symphonies might we start to compose

1042
00:44:50,579 --> 00:44:52,920
now that we're finally beginning to properly

1043
00:44:52,920 --> 00:44:56,000
read nature's molecular sheet music? Think about

1044
00:44:56,000 --> 00:44:58,599
the mind -boggling possibilities. Designing life

1045
00:44:58,599 --> 00:45:00,559
-saving drugs that seemed impossible just years

1046
00:45:00,559 --> 00:45:03,280
ago. Creating novel biological systems straight

1047
00:45:03,280 --> 00:45:05,739
out of science fiction. All by mastering this

1048
00:45:05,739 --> 00:45:08,280
intricate, beautiful art of molecular origami.

1049
00:45:08,320 --> 00:45:10,460
It's a future that's unfolding incredibly rapidly

1050
00:45:10,460 --> 00:45:13,039
right before our eyes. And it's a future you're

1051
00:45:13,039 --> 00:45:15,380
now, hopefully, a little better equipped to understand.