WEBVTT
1
00:00:05.009 --> 00:00:07.519
My name is Kyle Lowe. I'm an associate professor at
2
00:00:07.540 --> 00:00:08.160
Stanford Medicine.
3
00:00:08.900 --> 00:00:11.480
Can you describe the two sources of our brain, but
4
00:00:11.539 --> 00:00:13.160
from a developmental point of view?
5
00:00:13.939 --> 00:00:16.339
So we often think of the brain as a single
6
00:00:16.420 --> 00:00:19.660
organ that comes from a common source in embryonic development.
7
00:00:20.339 --> 00:00:22.760
But we found that the brain is actually built from
8
00:00:22.839 --> 00:00:25.579
two different sources. One that makes the front half of
9
00:00:25.600 --> 00:00:28.739
the brain, the forebrain and midbrain. Another source that makes
10
00:00:28.760 --> 00:00:32.049
the back half of the brain, the hindbrain. And this
11
00:00:32.859 --> 00:00:34.939
New view of the brain as not just one organ,
12
00:00:35.020 --> 00:00:37.560
but really one organ built from two different sources has
13
00:00:37.579 --> 00:00:40.539
really interesting implications for making different kinds of human neurons
14
00:00:40.560 --> 00:00:42.979
in a petri dish and for the evolution of the brain.
15
00:00:43.920 --> 00:00:46.350
As you know, motor neurons control your muscles, and most
16
00:00:46.390 --> 00:00:49.189
people think of the motor neurons in your spinal cord
17
00:00:49.509 --> 00:00:52.210
that control your arms, legs, help you do lots of things.
18
00:00:52.950 --> 00:00:56.210
But there's also a second, less well-studied group of motor
19
00:00:56.229 --> 00:00:59.289
neurons that sit in your hindbrain or your brainstem. And
20
00:00:59.350 --> 00:01:01.850
as medical professionals, you know that they project through the
21
00:01:01.890 --> 00:01:06.170
cranial nerves to control your face, jaw, eye, tongue, and
22
00:01:06.189 --> 00:01:09.040
neck muscles. So while you probably appreciate why those are
23
00:01:09.060 --> 00:01:10.659
so important, when I give a talk to a basic
24
00:01:10.700 --> 00:01:13.480
science audience, I have to remind people, why do we
25
00:01:13.519 --> 00:01:16.280
care about these high-end motor neurons? Well, they're critical for
26
00:01:16.379 --> 00:01:18.819
eating and swallowing, which are, of course, essential for life,
27
00:01:19.560 --> 00:01:24.230
and for speech and facial movements. I actually think, well,
28
00:01:24.250 --> 00:01:26.469
the first one's more clinically relevant. The second one's actually
29
00:01:26.489 --> 00:01:29.430
very interesting because, you know, in evolution, Speech and facial
30
00:01:29.469 --> 00:01:31.609
movements allow us to conduct this interview right now. It
31
00:01:32.030 --> 00:01:34.090
allows us to connect the outside world and in evolution,
32
00:01:34.129 --> 00:01:36.230
distinguish us from other species. And I think that's actually
33
00:01:36.250 --> 00:01:39.689
very cool. So there have been a lot of successes
34
00:01:39.849 --> 00:01:43.140
in turning human stem cells to make spinal cord motor neurons.
35
00:01:43.939 --> 00:01:45.739
And many of you have read about these, and these
36
00:01:45.819 --> 00:01:50.200
obviously have opened up many avenues to model neuromuscular diseases,
37
00:01:50.900 --> 00:01:53.120
think about how to treat spinal cord injury and other things.
38
00:01:54.079 --> 00:01:56.239
But there's been less attention to turn stem cells to
39
00:01:56.420 --> 00:01:59.030
hindbrain motor neurons. So that's how we got into this.
40
00:01:59.959 --> 00:02:02.640
So just to take a step back, while many people
41
00:02:02.659 --> 00:02:06.579
have turned stem cells into forebrain and midbrain, there's been
42
00:02:06.599 --> 00:02:09.479
maybe less attention to make the hindbrain. And for a
43
00:02:09.500 --> 00:02:11.789
long time, we thought, that's a bit of a curiosity. Like,
44
00:02:11.810 --> 00:02:15.270
why are people so good at making forebrain and midbrain neurons? Like,
45
00:02:15.310 --> 00:02:18.370
for instance, dopaminergic neurons, loss in Parkinson's in the midbrain,
46
00:02:18.750 --> 00:02:20.969
but less in the hindbrain. So when we found that
47
00:02:21.009 --> 00:02:25.240
the brain actually comes from these two different progenitors, we
48
00:02:25.259 --> 00:02:28.930
were like, aha, at the moment, it just made sense.
49
00:02:29.889 --> 00:02:32.750
Because we realized if the hindbrain comes from its own
50
00:02:32.930 --> 00:02:36.509
unique founder cell or progenitor cell in development, we have
51
00:02:36.530 --> 00:02:40.240
to basically know that to turn stem cells into those
52
00:02:40.289 --> 00:02:42.500
cells to make hindbrain motor neurons. So the analogy I
53
00:02:42.520 --> 00:02:44.639
like to use is, let's say you're navigating down a
54
00:02:44.680 --> 00:02:47.699
main road into the forebrain and midbrain. But if you
55
00:02:47.740 --> 00:02:49.740
know some of those rural country roads, there's a sign
56
00:02:49.759 --> 00:02:52.490
that says, oh, warning, concealed side entrances. If you never
57
00:02:52.530 --> 00:02:54.530
knew there was that concealed side entrance to turn right
58
00:02:54.560 --> 00:02:56.680
to go to the hindbrain, you just would miss that turn.
59
00:02:56.780 --> 00:02:58.659
And the stem cell, you just shoot by. You'd never
60
00:02:58.680 --> 00:03:00.400
be able to get on that hindbrain for.
61
00:03:01.340 --> 00:03:04.900
Can you describe how you generated motor neurons from the
62
00:03:04.919 --> 00:03:05.560
stem cells?
63
00:03:06.139 --> 00:03:07.379
Here, I want to give a lot of credit to
64
00:03:07.430 --> 00:03:10.229
my MD PhD student, Brian Jokai, who led this. And
65
00:03:10.250 --> 00:03:13.430
I learned a lot. So he basically said, well, Kyle,
66
00:03:13.449 --> 00:03:15.539
it's kind of academic. that we've discovered that there are
67
00:03:15.560 --> 00:03:17.780
these two progenitor cells, one with the forebrain moving, the
68
00:03:17.800 --> 00:03:19.710
other with the hindbrain, but he said he really wanted
69
00:03:19.719 --> 00:03:22.090
to make those hindbrain neurons because they're so controlled with
70
00:03:22.110 --> 00:03:25.669
life-sustaining functions. So to walk you through how these experiments work,
71
00:03:26.389 --> 00:03:28.229
in a mouse embryo, when a brain is growing up,
72
00:03:28.909 --> 00:03:31.949
we found the genes that identified the two different progenitor cells.
73
00:03:32.969 --> 00:03:35.150
So the genes, I don't know if it's that relevant.
74
00:03:35.629 --> 00:03:38.469
One is called OTX2. It marks the first progenitor cell
75
00:03:38.490 --> 00:03:42.159
that makes the forebrain midbrain. Another is called GBX2. And
76
00:03:42.219 --> 00:03:47.520
that gene marks this future hindbrain progenitor. So essentially, Ryan
77
00:03:48.319 --> 00:03:52.129
and Carolyn, co-first author, created methods to turn stem cells,
78
00:03:52.189 --> 00:03:55.689
those two different kinds of brain progenitors, the OTX2 positive
79
00:03:55.710 --> 00:03:58.830
and the GBX2 positive. Then Ryan went on to turn
80
00:03:58.870 --> 00:04:04.789
these GBX2 positive hindbrain progenitors into actual hindbrain. Then different
81
00:04:04.810 --> 00:04:07.189
parts of the hindbrain, the so-called dorsal versus ventral part,
82
00:04:07.509 --> 00:04:10.060
and the ventral part made the hindbrain motor neurons. So
83
00:04:10.080 --> 00:04:13.680
there's actually many sequence of steps from that incipient hindbrain
84
00:04:13.699 --> 00:04:16.060
progenitor to make the hindbrain motor neuron. And our lab
85
00:04:16.079 --> 00:04:18.379
specialty is really working out the nuances of these steps,
86
00:04:18.939 --> 00:04:22.860
understanding what signals direct them and what genes they express
87
00:04:22.899 --> 00:04:25.480
to really recapitulate their formation.
88
00:04:26.790 --> 00:04:30.649
What are the potential implications for developing therapies for motor
89
00:04:30.689 --> 00:04:32.329
neuron disorders such as SMAs?
90
00:04:34.449 --> 00:04:36.839
But for us, I think what we're really excited by
91
00:04:36.860 --> 00:04:40.240
is the fact that we can now study how these
92
00:04:40.279 --> 00:04:43.620
hindbrain motor neurons are killed. And actually, it's very, very
93
00:04:43.699 --> 00:04:46.339
understudied in the SMA and ALS field. So to the
94
00:04:46.360 --> 00:04:48.899
best of my knowledge, there's a huge body of literature
95
00:04:48.939 --> 00:04:52.779
on how SMA and ALS kill spinal cord motor neurons. Clinically,
96
00:04:52.800 --> 00:04:56.899
very important to understand why there's difficulty in posture, arms,
97
00:04:56.939 --> 00:05:01.870
and legs. But there's much less attention making hindbrain motor neurons.
98
00:05:01.889 --> 00:05:05.689
In fact, sometimes I tell people, people, clinicians about hambemodernization, oh,
99
00:05:05.730 --> 00:05:08.050
we didn't even like think of that. They really existed, right?
100
00:05:08.709 --> 00:05:12.170
So we obviously, we can't access, we cannot access the
101
00:05:12.209 --> 00:05:15.649
brainstem of these SMA or ALS patients to see like
102
00:05:15.670 --> 00:05:18.910
what's happening. So the experience that we're doing now that
103
00:05:18.930 --> 00:05:22.540
we're taking from Stanford Hospital with John Day, who's a
104
00:05:22.569 --> 00:05:26.139
pioneer in neuromuscular research and at Joe Wu's labs, we've
105
00:05:26.240 --> 00:05:30.180
created human-induced plapone stem cell lines from patients with type
106
00:05:30.180 --> 00:05:32.329
1 SMA. But these are stem cells in a petri
107
00:05:32.350 --> 00:05:36.329
dish that are genetically identical to the type 1 SMA
108
00:05:36.370 --> 00:05:38.879
patient they have, and they have the causal mutation in SMN1.
109
00:05:39.420 --> 00:05:42.259
And then we're converting them into hybid motor neurons side
110
00:05:42.300 --> 00:05:46.379
by side with stem cell lines from control patients. And
111
00:05:46.399 --> 00:05:50.689
this very controlled reductionist system in a petri dish can
112
00:05:50.699 --> 00:05:53.029
now allow us to see what is SMA really doing
113
00:05:53.069 --> 00:05:54.829
to these hybid motor neurons in a way that we
114
00:05:54.910 --> 00:05:58.750
obviously cannot see in a patient. And secondly... because we
115
00:05:58.759 --> 00:06:00.990
can make massive numbers of hind-brain motor neurons, we can
116
00:06:01.050 --> 00:06:04.790
now test different drugs and model their effects, which we're
117
00:06:04.829 --> 00:06:05.730
very excited by too.
118
00:06:07.529 --> 00:06:11.889
Can you elaborate on the pathophysiology of SMA and some
119
00:06:11.910 --> 00:06:15.569
of the research you're doing to better understand it?
120
00:06:15.569 --> 00:06:18.430
So the first is, of course, the holy grail, which
121
00:06:18.610 --> 00:06:22.209
may be very difficult, is can we take a healthy
122
00:06:22.250 --> 00:06:26.149
stem cell with the normal SN1 gene into a healthy
123
00:06:26.189 --> 00:06:28.379
hind-brain motor neuron and transplant it? because that would be
124
00:06:28.420 --> 00:06:33.620
very definitive. The issue is once we transplant those motor neurons,
125
00:06:34.079 --> 00:06:36.529
they would have to learn how to project their axons
126
00:06:36.629 --> 00:06:38.930
to the skeletal muscle they innervate. So that is a
127
00:06:38.949 --> 00:06:42.189
big challenge that will probably entail the work of many
128
00:06:42.230 --> 00:06:44.509
years and decades to see if that's even feasible. So
129
00:06:44.550 --> 00:06:48.980
that's the most ambitious long-term goal. The second goal that's
130
00:06:49.000 --> 00:06:51.740
already being done is the gene therapy in small molecules
131
00:06:52.319 --> 00:06:57.350
to essentially supply functional SMM. So my understanding is that There's,
132
00:06:57.370 --> 00:07:01.889
of course, great clinical efficacy, but it's not perfect. And
133
00:07:01.910 --> 00:07:04.189
we don't fully understand why it's not perfect. It could
134
00:07:04.230 --> 00:07:07.610
be because we're not restoring the levels of SMN sufficiently.
135
00:07:08.029 --> 00:07:11.069
What's not getting to all the cells. So I think
136
00:07:11.089 --> 00:07:13.709
those are the leading theories. But the third thing, which
137
00:07:13.730 --> 00:07:18.149
I think is important, is understanding what SMA really does.
138
00:07:18.209 --> 00:07:20.939
And it's actually quite a mystery, which I'm happy to
139
00:07:20.980 --> 00:07:23.379
go into. And I think that is where our system
140
00:07:23.420 --> 00:07:25.399
to create high brain motor neurons will allow us to
141
00:07:25.439 --> 00:07:28.420
watch what SMN deficiency is really doing to these high
142
00:07:28.439 --> 00:07:31.060
brain motor neurons and test for drugs, which I think
143
00:07:31.100 --> 00:07:31.839
is very important.
144
00:07:33.860 --> 00:07:37.040
What is the missing SMN1 gene doing?
145
00:07:38.420 --> 00:07:42.110
Yes. So it is a huge mystery. So the SMA,
146
00:07:42.370 --> 00:07:45.519
like many diseases, is obviously a complex disease affecting multiple
147
00:07:45.550 --> 00:07:48.230
kinds of cells. But we know from the name Survival
148
00:07:48.250 --> 00:07:51.209
of motor neuron 1, SMN1, it's particularly important in motor neurons.
149
00:07:52.709 --> 00:07:54.850
What I think is shocking, and of course, I am
150
00:07:54.889 --> 00:07:58.610
a novice in this field, is SMN1 is ubiquitously expressed.
151
00:08:00.709 --> 00:08:03.939
Every cell in the body has it. It's also ancient
152
00:08:03.980 --> 00:08:07.779
in evolution across many kinds of all animals. So it's
153
00:08:07.839 --> 00:08:10.939
very surprising why a gene expressed in all cells of
154
00:08:10.959 --> 00:08:16.899
the body, when lost, impacts a relatively small subset of them.
155
00:08:17.860 --> 00:08:21.639
The mystery even further deepens because SMN1 is required for
156
00:08:21.670 --> 00:08:25.170
a very fundamental process in a cell, gene splicing. As
157
00:08:25.189 --> 00:08:28.370
you know, genes have exons and introns. These introns are
158
00:08:28.550 --> 00:08:31.709
cut out to basically get the exons. So you have
159
00:08:31.730 --> 00:08:33.490
a coding sequence of a gene. It turns out you
160
00:08:33.509 --> 00:08:36.549
have two types of machines that do the splicing. One
161
00:08:36.590 --> 00:08:40.279
called a major spliceosome that splices most genes. And a
162
00:08:40.340 --> 00:08:43.460
second machine called a minor spliceosome that splices a much
163
00:08:43.580 --> 00:08:47.629
smaller set of genes. It turns out SMN1 is essential
164
00:08:47.649 --> 00:08:50.769
for the assembly of the minor spliceosome. So this is
165
00:08:50.809 --> 00:08:54.889
the mystery. Gene splicing is such a fundamental process. Why,
166
00:08:54.929 --> 00:09:00.509
when disrupting it, would you impact a fairly small yet
167
00:09:00.730 --> 00:09:04.049
exclusive and important type of cell, right? What are the
168
00:09:04.169 --> 00:09:06.919
genes that are not being spliced correctly when you take
169
00:09:06.960 --> 00:09:09.799
away SMN1 in the minor spliceosome? So of course we
170
00:09:09.840 --> 00:09:12.179
know some of the answer and there's many labs working
171
00:09:12.200 --> 00:09:15.169
on these questions. but we certainly don't know the whole answer.
172
00:09:15.190 --> 00:09:16.909
And I think that's what we're really excited to find
173
00:09:17.029 --> 00:09:20.169
out because you can't really analyze a hybrid motor neuron
174
00:09:20.509 --> 00:09:22.929
in a patient to figure out these issues. And analyzing
175
00:09:22.970 --> 00:09:24.710
a hybrid motor neuron in a mouse is a great,
176
00:09:24.789 --> 00:09:28.629
but certainly not identical to a human. And I think
177
00:09:28.669 --> 00:09:32.009
that a deeper understanding of what is really going on
178
00:09:32.029 --> 00:09:35.220
is very important. And that's the understanding part. The treatment
179
00:09:35.259 --> 00:09:39.799
part is, well, it's very difficult to test these SMA
180
00:09:39.840 --> 00:09:42.429
and therapies in a mouse model. And there's a huge
181
00:09:42.470 --> 00:09:44.610
debate over which mouse model is the best for SMA
182
00:09:44.629 --> 00:09:48.740
and a very active debate is ongoing, right? But if
183
00:09:48.779 --> 00:09:51.120
you can make these hindbrain neuronal neurons at Petri dish,
184
00:09:51.139 --> 00:09:54.220
you can test thousands, millions of drugs to find ones
185
00:09:54.259 --> 00:09:57.049
that can treat them. And we are very excited about
186
00:09:57.090 --> 00:09:59.690
that too, that you can finally really look at a
187
00:09:59.710 --> 00:10:05.139
human like very specific biology of these human cells essentially
188
00:10:05.860 --> 00:10:08.220
and how to correct them. So we were really excited.
189
00:10:09.080 --> 00:10:11.360
We didn't touch on this interview much on ALS, but
190
00:10:11.379 --> 00:10:15.799
the same logic, same rules apply there. ALS is even
191
00:10:15.840 --> 00:10:19.309
more complicated because genetic basis is much more complex than SMA.
192
00:10:20.159 --> 00:10:22.799
And we are very actively also trying to move into
193
00:10:22.919 --> 00:10:25.980
ALS as well. Another ambition for us will be to
194
00:10:26.059 --> 00:10:29.120
also create different kinds of spinal cord motor neurons. Because
195
00:10:29.389 --> 00:10:31.889
although we regard them as one entity, there's a huge
196
00:10:31.929 --> 00:10:35.389
diversity in spinal cord motor neurons. Disease like SMA actually
197
00:10:35.429 --> 00:10:37.789
kill some but spare others, which is very mysterious. And
198
00:10:37.809 --> 00:10:41.990
we want to understand why. One thing that the SMA
199
00:10:42.029 --> 00:10:45.169
Foundation has taught me that I'm very excited by is
200
00:10:45.830 --> 00:10:51.299
that very closely related motor neurons... have totally opposite responses
201
00:10:51.320 --> 00:10:54.480
to SMA. So SMA is often taught as, quote, a
202
00:10:54.559 --> 00:10:59.830
motor neuron disease. But actually, it doesn't kill all motor neurons.
203
00:11:00.629 --> 00:11:02.549
So for instance, if you look in, as you know,
204
00:11:02.710 --> 00:11:05.330
very sadly, a patient with SMA often die because of
205
00:11:05.389 --> 00:11:08.029
respiratory failure because their respiratory muscles stop working. But it
206
00:11:08.049 --> 00:11:12.200
turns out that two motor neurons are responsible. One innervates
207
00:11:12.220 --> 00:11:14.460
the diaphragm, which provides a lot of the breathing force,
208
00:11:15.039 --> 00:11:17.470
called the phrenic motor neuron. The second motor neuron called
209
00:11:17.490 --> 00:11:20.450
the hypoxyl motor neuron innervates the accessory breathing muscles that
210
00:11:20.470 --> 00:11:24.139
provide less of the breathing force. Turns out SMA, the
211
00:11:24.200 --> 00:11:27.000
phrenic motor neuron is mostly fine. It doesn't really care.
212
00:11:28.259 --> 00:11:31.460
The hypoxyl motor neuron, a very closely related spinal cord
213
00:11:31.480 --> 00:11:35.320
motor neuron, is the one that dies. That's why these
214
00:11:35.360 --> 00:11:39.269
patients probably die. die within roughly one year of birth
215
00:11:39.330 --> 00:11:42.009
if untreated. Because if you lost your diaphragm, you probably
216
00:11:42.029 --> 00:11:44.129
just couldn't breathe at all upon birth. But if you
217
00:11:44.169 --> 00:11:47.360
have these accessory motor muscles that kind of weaken over time,
218
00:11:47.379 --> 00:11:49.480
then you can breathe for a while, but eventually you
219
00:11:49.500 --> 00:11:52.159
get exhausted. So I believe one of the great mysteries
220
00:11:52.200 --> 00:11:55.059
in SMA is it's not just like why the motor
221
00:11:55.080 --> 00:11:57.259
neuron affected more so than other cells in the body.
222
00:11:57.279 --> 00:12:00.370
Like there's like dozens of kinds, or maybe let's say
223
00:12:00.870 --> 00:12:03.669
a handful or maybe a dozen or dozens of kinds
224
00:12:03.730 --> 00:12:06.610
of motor neurons affected And some are spared and some
225
00:12:06.649 --> 00:12:08.909
are destroyed. And I feel like that's really the key.
226
00:12:09.409 --> 00:12:12.809
Because right now, I feel like as human beings, we're
227
00:12:12.870 --> 00:12:16.009
artificially trying to treat SMA by thinking like, what's wrong? Okay,
228
00:12:16.110 --> 00:12:18.730
SN1 is gone. Let's replete it with gene therapies or
229
00:12:18.750 --> 00:12:22.049
with molecules. But in our bodies, the secret is already there.
230
00:12:23.500 --> 00:12:26.700
Because clearly, the phrenic motor neuron doesn't really care about
231
00:12:26.820 --> 00:12:31.159
SMA and the hyperactron is vulnerable. There are motor neurons
232
00:12:31.240 --> 00:12:33.480
in the body of an SMA patient that are mostly intact.
233
00:12:33.899 --> 00:12:35.820
We could figure out what makes them different from a
234
00:12:35.840 --> 00:12:38.259
very closely related kind of motor neuron, that kind of resilience.
235
00:12:38.720 --> 00:12:40.240
That could be a totally new way of thinking about
236
00:12:40.259 --> 00:12:42.500
SMA therapy that our lab is very excited to pursue.
237
00:12:42.759 --> 00:12:45.100
It's like the natural treatment. It already exists in the body.
238
00:12:45.980 --> 00:12:47.679
Is there anything else you want to add to this conversation?
239
00:12:48.620 --> 00:12:50.379
I know there's a medical audience and people already know this,
240
00:12:50.519 --> 00:12:52.279
but whenever I meet like a stranger, I tell them
241
00:12:52.299 --> 00:12:55.370
type 1 SMA is the number one genetic killer of
242
00:12:55.409 --> 00:12:58.230
babies less than one year of age. And that should
243
00:12:58.269 --> 00:13:02.929
make people sad. And these kids die because they can't breathe.
244
00:13:03.590 --> 00:13:06.250
Or they choke and they get pneumonia and they die.
245
00:13:06.529 --> 00:13:09.840
And I really hope that people will work on that.
246
00:13:09.860 --> 00:13:11.620
And that's what drew me to work on it. Although
247
00:13:12.360 --> 00:13:14.840
my background clearly isn't in this area, right? More people
248
00:13:14.860 --> 00:13:17.779
should know about this. It's really sad that little infants
249
00:13:17.860 --> 00:13:19.799
are subject to this. It's terrible.
1
00:00:05.009 --> 00:00:07.519
My name is Kyle Lowe. I'm an associate professor at
2
00:00:07.540 --> 00:00:08.160
Stanford Medicine.
3
00:00:08.900 --> 00:00:11.480
Can you describe the two sources of our brain, but
4
00:00:11.539 --> 00:00:13.160
from a developmental point of view?
5
00:00:13.939 --> 00:00:16.339
So we often think of the brain as a single
6
00:00:16.420 --> 00:00:19.660
organ that comes from a common source in embryonic development.
7
00:00:20.339 --> 00:00:22.760
But we found that the brain is actually built from
8
00:00:22.839 --> 00:00:25.579
two different sources. One that makes the front half of
9
00:00:25.600 --> 00:00:28.739
the brain, the forebrain and midbrain. Another source that makes
10
00:00:28.760 --> 00:00:32.049
the back half of the brain, the hindbrain. And this
11
00:00:32.859 --> 00:00:34.939
New view of the brain as not just one organ,
12
00:00:35.020 --> 00:00:37.560
but really one organ built from two different sources has
13
00:00:37.579 --> 00:00:40.539
really interesting implications for making different kinds of human neurons
14
00:00:40.560 --> 00:00:42.979
in a petri dish and for the evolution of the brain.
15
00:00:43.920 --> 00:00:46.350
As you know, motor neurons control your muscles, and most
16
00:00:46.390 --> 00:00:49.189
people think of the motor neurons in your spinal cord
17
00:00:49.509 --> 00:00:52.210
that control your arms, legs, help you do lots of things.
18
00:00:52.950 --> 00:00:56.210
But there's also a second, less well-studied group of motor
19
00:00:56.229 --> 00:00:59.289
neurons that sit in your hindbrain or your brainstem. And
20
00:00:59.350 --> 00:01:01.850
as medical professionals, you know that they project through the
21
00:01:01.890 --> 00:01:06.170
cranial nerves to control your face, jaw, eye, tongue, and
22
00:01:06.189 --> 00:01:09.040
neck muscles. So while you probably appreciate why those are
23
00:01:09.060 --> 00:01:10.659
so important, when I give a talk to a basic
24
00:01:10.700 --> 00:01:13.480
science audience, I have to remind people, why do we
25
00:01:13.519 --> 00:01:16.280
care about these high-end motor neurons? Well, they're critical for
26
00:01:16.379 --> 00:01:18.819
eating and swallowing, which are, of course, essential for life,
27
00:01:19.560 --> 00:01:24.230
and for speech and facial movements. I actually think, well,
28
00:01:24.250 --> 00:01:26.469
the first one's more clinically relevant. The second one's actually
29
00:01:26.489 --> 00:01:29.430
very interesting because, you know, in evolution, Speech and facial
30
00:01:29.469 --> 00:01:31.609
movements allow us to conduct this interview right now. It
31
00:01:32.030 --> 00:01:34.090
allows us to connect the outside world and in evolution,
32
00:01:34.129 --> 00:01:36.230
distinguish us from other species. And I think that's actually
33
00:01:36.250 --> 00:01:39.689
very cool. So there have been a lot of successes
34
00:01:39.849 --> 00:01:43.140
in turning human stem cells to make spinal cord motor neurons.
35
00:01:43.939 --> 00:01:45.739
And many of you have read about these, and these
36
00:01:45.819 --> 00:01:50.200
obviously have opened up many avenues to model neuromuscular diseases,
37
00:01:50.900 --> 00:01:53.120
think about how to treat spinal cord injury and other things.
38
00:01:54.079 --> 00:01:56.239
But there's been less attention to turn stem cells to
39
00:01:56.420 --> 00:01:59.030
hindbrain motor neurons. So that's how we got into this.
40
00:01:59.959 --> 00:02:02.640
So just to take a step back, while many people
41
00:02:02.659 --> 00:02:06.579
have turned stem cells into forebrain and midbrain, there's been
42
00:02:06.599 --> 00:02:09.479
maybe less attention to make the hindbrain. And for a
43
00:02:09.500 --> 00:02:11.789
long time, we thought, that's a bit of a curiosity. Like,
44
00:02:11.810 --> 00:02:15.270
why are people so good at making forebrain and midbrain neurons? Like,
45
00:02:15.310 --> 00:02:18.370
for instance, dopaminergic neurons, loss in Parkinson's in the midbrain,
46
00:02:18.750 --> 00:02:20.969
but less in the hindbrain. So when we found that
47
00:02:21.009 --> 00:02:25.240
the brain actually comes from these two different progenitors, we
48
00:02:25.259 --> 00:02:28.930
were like, aha, at the moment, it just made sense.
49
00:02:29.889 --> 00:02:32.750
Because we realized if the hindbrain comes from its own
50
00:02:32.930 --> 00:02:36.509
unique founder cell or progenitor cell in development, we have
51
00:02:36.530 --> 00:02:40.240
to basically know that to turn stem cells into those
52
00:02:40.289 --> 00:02:42.500
cells to make hindbrain motor neurons. So the analogy I
53
00:02:42.520 --> 00:02:44.639
like to use is, let's say you're navigating down a
54
00:02:44.680 --> 00:02:47.699
main road into the forebrain and midbrain. But if you
55
00:02:47.740 --> 00:02:49.740
know some of those rural country roads, there's a sign
56
00:02:49.759 --> 00:02:52.490
that says, oh, warning, concealed side entrances. If you never
57
00:02:52.530 --> 00:02:54.530
knew there was that concealed side entrance to turn right
58
00:02:54.560 --> 00:02:56.680
to go to the hindbrain, you just would miss that turn.
59
00:02:56.780 --> 00:02:58.659
And the stem cell, you just shoot by. You'd never
60
00:02:58.680 --> 00:03:00.400
be able to get on that hindbrain for.
61
00:03:01.340 --> 00:03:04.900
Can you describe how you generated motor neurons from the
62
00:03:04.919 --> 00:03:05.560
stem cells?
63
00:03:06.139 --> 00:03:07.379
Here, I want to give a lot of credit to
64
00:03:07.430 --> 00:03:10.229
my MD PhD student, Brian Jokai, who led this. And
65
00:03:10.250 --> 00:03:13.430
I learned a lot. So he basically said, well, Kyle,
66
00:03:13.449 --> 00:03:15.539
it's kind of academic. that we've discovered that there are
67
00:03:15.560 --> 00:03:17.780
these two progenitor cells, one with the forebrain moving, the
68
00:03:17.800 --> 00:03:19.710
other with the hindbrain, but he said he really wanted
69
00:03:19.719 --> 00:03:22.090
to make those hindbrain neurons because they're so controlled with
70
00:03:22.110 --> 00:03:25.669
life-sustaining functions. So to walk you through how these experiments work,
71
00:03:26.389 --> 00:03:28.229
in a mouse embryo, when a brain is growing up,
72
00:03:28.909 --> 00:03:31.949
we found the genes that identified the two different progenitor cells.
73
00:03:32.969 --> 00:03:35.150
So the genes, I don't know if it's that relevant.
74
00:03:35.629 --> 00:03:38.469
One is called OTX2. It marks the first progenitor cell
75
00:03:38.490 --> 00:03:42.159
that makes the forebrain midbrain. Another is called GBX2. And
76
00:03:42.219 --> 00:03:47.520
that gene marks this future hindbrain progenitor. So essentially, Ryan
77
00:03:48.319 --> 00:03:52.129
and Carolyn, co-first author, created methods to turn stem cells,
78
00:03:52.189 --> 00:03:55.689
those two different kinds of brain progenitors, the OTX2 positive
79
00:03:55.710 --> 00:03:58.830
and the GBX2 positive. Then Ryan went on to turn
80
00:03:58.870 --> 00:04:04.789
these GBX2 positive hindbrain progenitors into actual hindbrain. Then different
81
00:04:04.810 --> 00:04:07.189
parts of the hindbrain, the so-called dorsal versus ventral part,
82
00:04:07.509 --> 00:04:10.060
and the ventral part made the hindbrain motor neurons. So
83
00:04:10.080 --> 00:04:13.680
there's actually many sequence of steps from that incipient hindbrain
84
00:04:13.699 --> 00:04:16.060
progenitor to make the hindbrain motor neuron. And our lab
85
00:04:16.079 --> 00:04:18.379
specialty is really working out the nuances of these steps,
86
00:04:18.939 --> 00:04:22.860
understanding what signals direct them and what genes they express
87
00:04:22.899 --> 00:04:25.480
to really recapitulate their formation.
88
00:04:26.790 --> 00:04:30.649
What are the potential implications for developing therapies for motor
89
00:04:30.689 --> 00:04:32.329
neuron disorders such as SMAs?
90
00:04:34.449 --> 00:04:36.839
But for us, I think what we're really excited by
91
00:04:36.860 --> 00:04:40.240
is the fact that we can now study how these
92
00:04:40.279 --> 00:04:43.620
hindbrain motor neurons are killed. And actually, it's very, very
93
00:04:43.699 --> 00:04:46.339
understudied in the SMA and ALS field. So to the
94
00:04:46.360 --> 00:04:48.899
best of my knowledge, there's a huge body of literature
95
00:04:48.939 --> 00:04:52.779
on how SMA and ALS kill spinal cord motor neurons. Clinically,
96
00:04:52.800 --> 00:04:56.899
very important to understand why there's difficulty in posture, arms,
97
00:04:56.939 --> 00:05:01.870
and legs. But there's much less attention making hindbrain motor neurons.
98
00:05:01.889 --> 00:05:05.689
In fact, sometimes I tell people, people, clinicians about hambemodernization, oh,
99
00:05:05.730 --> 00:05:08.050
we didn't even like think of that. They really existed, right?
100
00:05:08.709 --> 00:05:12.170
So we obviously, we can't access, we cannot access the
101
00:05:12.209 --> 00:05:15.649
brainstem of these SMA or ALS patients to see like
102
00:05:15.670 --> 00:05:18.910
what's happening. So the experience that we're doing now that
103
00:05:18.930 --> 00:05:22.540
we're taking from Stanford Hospital with John Day, who's a
104
00:05:22.569 --> 00:05:26.139
pioneer in neuromuscular research and at Joe Wu's labs, we've
105
00:05:26.240 --> 00:05:30.180
created human-induced plapone stem cell lines from patients with type
106
00:05:30.180 --> 00:05:32.329
1 SMA. But these are stem cells in a petri
107
00:05:32.350 --> 00:05:36.329
dish that are genetically identical to the type 1 SMA
108
00:05:36.370 --> 00:05:38.879
patient they have, and they have the causal mutation in SMN1.
109
00:05:39.420 --> 00:05:42.259
And then we're converting them into hybid motor neurons side
110
00:05:42.300 --> 00:05:46.379
by side with stem cell lines from control patients. And
111
00:05:46.399 --> 00:05:50.689
this very controlled reductionist system in a petri dish can
112
00:05:50.699 --> 00:05:53.029
now allow us to see what is SMA really doing
113
00:05:53.069 --> 00:05:54.829
to these hybid motor neurons in a way that we
114
00:05:54.910 --> 00:05:58.750
obviously cannot see in a patient. And secondly... because we
115
00:05:58.759 --> 00:06:00.990
can make massive numbers of hind-brain motor neurons, we can
116
00:06:01.050 --> 00:06:04.790
now test different drugs and model their effects, which we're
117
00:06:04.829 --> 00:06:05.730
very excited by too.
118
00:06:07.529 --> 00:06:11.889
Can you elaborate on the pathophysiology of SMA and some
119
00:06:11.910 --> 00:06:15.569
of the research you're doing to better understand it?
120
00:06:15.569 --> 00:06:18.430
So the first is, of course, the holy grail, which
121
00:06:18.610 --> 00:06:22.209
may be very difficult, is can we take a healthy
122
00:06:22.250 --> 00:06:26.149
stem cell with the normal SN1 gene into a healthy
123
00:06:26.189 --> 00:06:28.379
hind-brain motor neuron and transplant it? because that would be
124
00:06:28.420 --> 00:06:33.620
very definitive. The issue is once we transplant those motor neurons,
125
00:06:34.079 --> 00:06:36.529
they would have to learn how to project their axons
126
00:06:36.629 --> 00:06:38.930
to the skeletal muscle they innervate. So that is a
127
00:06:38.949 --> 00:06:42.189
big challenge that will probably entail the work of many
128
00:06:42.230 --> 00:06:44.509
years and decades to see if that's even feasible. So
129
00:06:44.550 --> 00:06:48.980
that's the most ambitious long-term goal. The second goal that's
130
00:06:49.000 --> 00:06:51.740
already being done is the gene therapy in small molecules
131
00:06:52.319 --> 00:06:57.350
to essentially supply functional SMM. So my understanding is that There's,
132
00:06:57.370 --> 00:07:01.889
of course, great clinical efficacy, but it's not perfect. And
133
00:07:01.910 --> 00:07:04.189
we don't fully understand why it's not perfect. It could
134
00:07:04.230 --> 00:07:07.610
be because we're not restoring the levels of SMN sufficiently.
135
00:07:08.029 --> 00:07:11.069
What's not getting to all the cells. So I think
136
00:07:11.089 --> 00:07:13.709
those are the leading theories. But the third thing, which
137
00:07:13.730 --> 00:07:18.149
I think is important, is understanding what SMA really does.
138
00:07:18.209 --> 00:07:20.939
And it's actually quite a mystery, which I'm happy to
139
00:07:20.980 --> 00:07:23.379
go into. And I think that is where our system
140
00:07:23.420 --> 00:07:25.399
to create high brain motor neurons will allow us to
141
00:07:25.439 --> 00:07:28.420
watch what SMN deficiency is really doing to these high
142
00:07:28.439 --> 00:07:31.060
brain motor neurons and test for drugs, which I think
143
00:07:31.100 --> 00:07:31.839
is very important.
144
00:07:33.860 --> 00:07:37.040
What is the missing SMN1 gene doing?
145
00:07:38.420 --> 00:07:42.110
Yes. So it is a huge mystery. So the SMA,
146
00:07:42.370 --> 00:07:45.519
like many diseases, is obviously a complex disease affecting multiple
147
00:07:45.550 --> 00:07:48.230
kinds of cells. But we know from the name Survival
148
00:07:48.250 --> 00:07:51.209
of motor neuron 1, SMN1, it's particularly important in motor neurons.
149
00:07:52.709 --> 00:07:54.850
What I think is shocking, and of course, I am
150
00:07:54.889 --> 00:07:58.610
a novice in this field, is SMN1 is ubiquitously expressed.
151
00:08:00.709 --> 00:08:03.939
Every cell in the body has it. It's also ancient
152
00:08:03.980 --> 00:08:07.779
in evolution across many kinds of all animals. So it's
153
00:08:07.839 --> 00:08:10.939
very surprising why a gene expressed in all cells of
154
00:08:10.959 --> 00:08:16.899
the body, when lost, impacts a relatively small subset of them.
155
00:08:17.860 --> 00:08:21.639
The mystery even further deepens because SMN1 is required for
156
00:08:21.670 --> 00:08:25.170
a very fundamental process in a cell, gene splicing. As
157
00:08:25.189 --> 00:08:28.370
you know, genes have exons and introns. These introns are
158
00:08:28.550 --> 00:08:31.709
cut out to basically get the exons. So you have
159
00:08:31.730 --> 00:08:33.490
a coding sequence of a gene. It turns out you
160
00:08:33.509 --> 00:08:36.549
have two types of machines that do the splicing. One
161
00:08:36.590 --> 00:08:40.279
called a major spliceosome that splices most genes. And a
162
00:08:40.340 --> 00:08:43.460
second machine called a minor spliceosome that splices a much
163
00:08:43.580 --> 00:08:47.629
smaller set of genes. It turns out SMN1 is essential
164
00:08:47.649 --> 00:08:50.769
for the assembly of the minor spliceosome. So this is
165
00:08:50.809 --> 00:08:54.889
the mystery. Gene splicing is such a fundamental process. Why,
166
00:08:54.929 --> 00:09:00.509
when disrupting it, would you impact a fairly small yet
167
00:09:00.730 --> 00:09:04.049
exclusive and important type of cell, right? What are the
168
00:09:04.169 --> 00:09:06.919
genes that are not being spliced correctly when you take
169
00:09:06.960 --> 00:09:09.799
away SMN1 in the minor spliceosome? So of course we
170
00:09:09.840 --> 00:09:12.179
know some of the answer and there's many labs working
171
00:09:12.200 --> 00:09:15.169
on these questions. but we certainly don't know the whole answer.
172
00:09:15.190 --> 00:09:16.909
And I think that's what we're really excited to find
173
00:09:17.029 --> 00:09:20.169
out because you can't really analyze a hybrid motor neuron
174
00:09:20.509 --> 00:09:22.929
in a patient to figure out these issues. And analyzing
175
00:09:22.970 --> 00:09:24.710
a hybrid motor neuron in a mouse is a great,
176
00:09:24.789 --> 00:09:28.629
but certainly not identical to a human. And I think
177
00:09:28.669 --> 00:09:32.009
that a deeper understanding of what is really going on
178
00:09:32.029 --> 00:09:35.220
is very important. And that's the understanding part. The treatment
179
00:09:35.259 --> 00:09:39.799
part is, well, it's very difficult to test these SMA
180
00:09:39.840 --> 00:09:42.429
and therapies in a mouse model. And there's a huge
181
00:09:42.470 --> 00:09:44.610
debate over which mouse model is the best for SMA
182
00:09:44.629 --> 00:09:48.740
and a very active debate is ongoing, right? But if
183
00:09:48.779 --> 00:09:51.120
you can make these hindbrain neuronal neurons at Petri dish,
184
00:09:51.139 --> 00:09:54.220
you can test thousands, millions of drugs to find ones
185
00:09:54.259 --> 00:09:57.049
that can treat them. And we are very excited about
186
00:09:57.090 --> 00:09:59.690
that too, that you can finally really look at a
187
00:09:59.710 --> 00:10:05.139
human like very specific biology of these human cells essentially
188
00:10:05.860 --> 00:10:08.220
and how to correct them. So we were really excited.
189
00:10:09.080 --> 00:10:11.360
We didn't touch on this interview much on ALS, but
190
00:10:11.379 --> 00:10:15.799
the same logic, same rules apply there. ALS is even
191
00:10:15.840 --> 00:10:19.309
more complicated because genetic basis is much more complex than SMA.
192
00:10:20.159 --> 00:10:22.799
And we are very actively also trying to move into
193
00:10:22.919 --> 00:10:25.980
ALS as well. Another ambition for us will be to
194
00:10:26.059 --> 00:10:29.120
also create different kinds of spinal cord motor neurons. Because
195
00:10:29.389 --> 00:10:31.889
although we regard them as one entity, there's a huge
196
00:10:31.929 --> 00:10:35.389
diversity in spinal cord motor neurons. Disease like SMA actually
197
00:10:35.429 --> 00:10:37.789
kill some but spare others, which is very mysterious. And
198
00:10:37.809 --> 00:10:41.990
we want to understand why. One thing that the SMA
199
00:10:42.029 --> 00:10:45.169
Foundation has taught me that I'm very excited by is
200
00:10:45.830 --> 00:10:51.299
that very closely related motor neurons... have totally opposite responses
201
00:10:51.320 --> 00:10:54.480
to SMA. So SMA is often taught as, quote, a
202
00:10:54.559 --> 00:10:59.830
motor neuron disease. But actually, it doesn't kill all motor neurons.
203
00:11:00.629 --> 00:11:02.549
So for instance, if you look in, as you know,
204
00:11:02.710 --> 00:11:05.330
very sadly, a patient with SMA often die because of
205
00:11:05.389 --> 00:11:08.029
respiratory failure because their respiratory muscles stop working. But it
206
00:11:08.049 --> 00:11:12.200
turns out that two motor neurons are responsible. One innervates
207
00:11:12.220 --> 00:11:14.460
the diaphragm, which provides a lot of the breathing force,
208
00:11:15.039 --> 00:11:17.470
called the phrenic motor neuron. The second motor neuron called
209
00:11:17.490 --> 00:11:20.450
the hypoxyl motor neuron innervates the accessory breathing muscles that
210
00:11:20.470 --> 00:11:24.139
provide less of the breathing force. Turns out SMA, the
211
00:11:24.200 --> 00:11:27.000
phrenic motor neuron is mostly fine. It doesn't really care.
212
00:11:28.259 --> 00:11:31.460
The hypoxyl motor neuron, a very closely related spinal cord
213
00:11:31.480 --> 00:11:35.320
motor neuron, is the one that dies. That's why these
214
00:11:35.360 --> 00:11:39.269
patients probably die. die within roughly one year of birth
215
00:11:39.330 --> 00:11:42.009
if untreated. Because if you lost your diaphragm, you probably
216
00:11:42.029 --> 00:11:44.129
just couldn't breathe at all upon birth. But if you
217
00:11:44.169 --> 00:11:47.360
have these accessory motor muscles that kind of weaken over time,
218
00:11:47.379 --> 00:11:49.480
then you can breathe for a while, but eventually you
219
00:11:49.500 --> 00:11:52.159
get exhausted. So I believe one of the great mysteries
220
00:11:52.200 --> 00:11:55.059
in SMA is it's not just like why the motor
221
00:11:55.080 --> 00:11:57.259
neuron affected more so than other cells in the body.
222
00:11:57.279 --> 00:12:00.370
Like there's like dozens of kinds, or maybe let's say
223
00:12:00.870 --> 00:12:03.669
a handful or maybe a dozen or dozens of kinds
224
00:12:03.730 --> 00:12:06.610
of motor neurons affected And some are spared and some
225
00:12:06.649 --> 00:12:08.909
are destroyed. And I feel like that's really the key.
226
00:12:09.409 --> 00:12:12.809
Because right now, I feel like as human beings, we're
227
00:12:12.870 --> 00:12:16.009
artificially trying to treat SMA by thinking like, what's wrong? Okay,
228
00:12:16.110 --> 00:12:18.730
SN1 is gone. Let's replete it with gene therapies or
229
00:12:18.750 --> 00:12:22.049
with molecules. But in our bodies, the secret is already there.
230
00:12:23.500 --> 00:12:26.700
Because clearly, the phrenic motor neuron doesn't really care about
231
00:12:26.820 --> 00:12:31.159
SMA and the hyperactron is vulnerable. There are motor neurons
232
00:12:31.240 --> 00:12:33.480
in the body of an SMA patient that are mostly intact.
233
00:12:33.899 --> 00:12:35.820
We could figure out what makes them different from a
234
00:12:35.840 --> 00:12:38.259
very closely related kind of motor neuron, that kind of resilience.
235
00:12:38.720 --> 00:12:40.240
That could be a totally new way of thinking about
236
00:12:40.259 --> 00:12:42.500
SMA therapy that our lab is very excited to pursue.
237
00:12:42.759 --> 00:12:45.100
It's like the natural treatment. It already exists in the body.
238
00:12:45.980 --> 00:12:47.679
Is there anything else you want to add to this conversation?
239
00:12:48.620 --> 00:12:50.379
I know there's a medical audience and people already know this,
240
00:12:50.519 --> 00:12:52.279
but whenever I meet like a stranger, I tell them
241
00:12:52.299 --> 00:12:55.370
type 1 SMA is the number one genetic killer of
242
00:12:55.409 --> 00:12:58.230
babies less than one year of age. And that should
243
00:12:58.269 --> 00:13:02.929
make people sad. And these kids die because they can't breathe.
244
00:13:03.590 --> 00:13:06.250
Or they choke and they get pneumonia and they die.
245
00:13:06.529 --> 00:13:09.840
And I really hope that people will work on that.
246
00:13:09.860 --> 00:13:11.620
And that's what drew me to work on it. Although
247
00:13:12.360 --> 00:13:14.840
my background clearly isn't in this area, right? More people
248
00:13:14.860 --> 00:13:17.779
should know about this. It's really sad that little infants
249
00:13:17.860 --> 00:13:19.799
are subject to this. It's terrible.
