0:00

At the center of our culture war

0:02

lies a single word, woe. I'm Kai

0:04

Wright. Join me on Notes from America,

0:06

live from Harlem's Apollo Theater, to explore

0:08

the word's origins as we celebrate Dr.

0:10

Martin Luther King Jr. Listen

0:12

wherever you get your podcasts. We're

0:19

one step closer to understanding the

0:21

complexity of the human brain. Instead

0:24

of characterizing cells on the basis

0:26

of their shape or who they

0:28

connect with or their firing properties,

0:30

instead one can characterize them on the basis of the

0:33

sets of genes that they use. I'm

0:35

sci-fi producer Shoshana Bucksbaum. It's Thursday,

0:37

January 18th, and we've

0:39

got our brain cells firing on all

0:42

cylinders because today is Science Friday. Late

0:46

last year, scientists released an impressively

0:48

detailed map of human brain cells.

0:51

Considering the human brain has a little over 170 billion

0:53

cells, it's a monumental task made possible

0:58

by an international group of

1:00

scientists. They've identified 3,000 different

1:03

types of cells. Ira Flatow and sci-fi

1:06

producer Kathleen Davis talk with one of

1:08

the scientists who worked on the project

1:10

and take listener calls. Dr.

1:13

Ed Lean, senior investigator at the

1:16

Allen Institute for Brain Science based

1:18

in Seattle. Welcome to Science Friday.

1:21

Good afternoon. Thank you for having me. Dr.

1:24

Lean, just how big an advance is

1:26

this cell atlas for the field of

1:28

neuroscience? Are we talking about a human

1:30

genome project level paradigm shift here? Yes,

1:34

I believe that's a really good analogy

1:36

for the advance made through this work

1:38

as the first installment really. One

1:42

of the problems with neuroscience is

1:44

the extreme complexity of the brain.

1:47

It's just very difficult to study for

1:50

the human brain the scale and the

1:52

inaccessibility of studying the brain are

1:55

serious barriers. We really

1:57

needed a technology breakthrough to be able

1:59

to Handle this complexity. And

2:02

a breakthrough am very it up appropriate

2:04

for the genome. References actually come from

2:06

the feel genomics. Where. Instead

2:09

of characterizing cells on the basis

2:11

of their shape, Or. Who

2:13

they connect with or they're firing properties.

2:16

In. Said one can characterize them on the basis of

2:18

the sets of genes that they use. For.

2:20

Every selling in the body has

2:22

all genes in their dna. By.

2:25

Any given Cel only uses a

2:27

subset of those teams, and that

2:29

molecular fingerprint of itself is a

2:31

really strong way to be able

2:33

to classify cells. As as

2:36

the field of To Know Max has advanced,

2:38

what's happened is that sequencing has become cheaper

2:40

and cheaper. and cheaper. And has

2:42

been miniaturised to be able to in

2:44

to analyze individual cells. And

2:46

now it's possible to analyze all the

2:49

genes being actively use by millions of

2:51

cells. And this now let's

2:53

you. Take. A much broader scope

2:55

of trying to categorize the complexity of

2:57

the brain. And so this

2:59

first installment was really trying to. Get.

3:02

A first draft by sampling about a hundred

3:04

regions of the brain and and then ask

3:06

how many types of cells are there and

3:08

you know I think. For. For

3:11

really over a century we understood that

3:13

the complexity is high but not quite

3:15

as high in other get see that

3:17

thousands of types of cells. So three

3:19

thousand. As you mentioned here, I'm.

3:22

Even since the time that publication, a comprehensive

3:24

analysis of the smaller mouse brain has come

3:27

a. Discovery or Thousand Celta had sexually.

3:29

We're going to get into that. After the break

3:31

we have to take a break so stay with.

3:34

Seductively. In there are, as you

3:36

were saying before the break, there

3:38

are one hundred and seventy one

3:40

billion cells in the human brain

3:42

Cells you possibly. Go about categorizing all

3:44

of them. Yet,

3:47

so it's true that there is that many

3:49

cells, and about half of those are neurons.

3:52

And but they don't come in that many different

3:54

types of cells. and so that

3:57

the key advance here is to be

3:59

able to category the cells into

4:02

groups of cells that have

4:05

different parts of the brain, that have different properties, and

4:08

make a catalog of these and a

4:10

map of these. And so these

4:12

new technologies, as I was describing, are

4:15

allowing us to create these so-called cell atlases.

4:17

And on the one hand, this is a

4:19

classification of the types where

4:22

we can define how many types of cells there

4:24

are, what their relative proportions are in different parts

4:26

of the brain. We can describe

4:28

their spatial organization now, how

4:31

they're organized in local areas

4:34

and also globally in the three-dimensional structure of

4:36

the brain, and begin to characterize

4:38

their properties and their function. And

4:40

so in many ways, this is really

4:42

like the genome project in the sense

4:44

of the genome characterized all of the

4:46

genes and their locations on the chromosomes.

4:49

Here we're characterizing all of the cells and

4:51

their locations and eventually their function in

4:53

the brain. One of the surprises of

4:55

the Human Genome Project was just about

4:57

how many genes there were, a surprising

5:00

less number than people thought. Are

5:03

you being surprised by how many different types of

5:05

cells you're finding in the brain? Most

5:09

definitely. These sort

5:11

of molecular approaches to classify

5:13

cells are revealing a

5:15

whole different level of complexity, probably an

5:18

order of magnitude more than we had realized

5:20

before. And importantly, one

5:23

or two orders of magnitude, so 10 to 100 times

5:25

more complex than any other organ in

5:28

the body. I was just going

5:30

to ask, I mean, how does this go to

5:32

other parts of our body? I can't imagine that,

5:34

like our biceps, for example, have 3,000

5:36

different types of cells. No,

5:39

nowhere near as many types of cells. So

5:41

the complexity is really very high. But

5:43

importantly, actually, these techniques,

5:46

by using genes as the way to define

5:48

cells, this technology can

5:50

be applied to any organ system. And so

5:52

in addition to the brain, in

5:54

parallel, there are efforts happening using these same

5:56

technologies in every other organ system, any

5:59

of the compiled. for example, in a project

6:01

called the Human Cell Atlas to try to put

6:03

all of this together. So we finally have a

6:05

common language to talk about the basic units of

6:07

life, the cells that make up every organ in

6:10

the body, but the complexity factor is

6:12

much higher in the brain. How

6:15

many new kinds of brain cells have you

6:17

discovered that we didn't know existed? And

6:20

the second part would be how many more do you think

6:22

are out there? Yes,

6:26

so these are somewhat difficult questions

6:28

to answer because

6:30

we have a new way of looking at these cells.

6:33

And so when we describe 3,000 types of

6:35

cells, in those parts of the brain that

6:37

we understand well, we can see that this actually

6:41

maps very well to what was known before,

6:44

but adds another level of resolution on this.

6:46

And so maybe you might have

6:48

thought that there were 50 types of cells in

6:51

the part of the neocortex. Now we see

6:53

there may be 150 types of cells. But

6:56

in other parts of the brain that we don't understand as well,

6:59

a lot of this is brand new knowledge. And

7:01

so we don't really know what this means yet.

7:04

But we have the framework now where

7:06

investigators across the whole community can come

7:08

and begin to add information to this.

7:11

So again, very much like the genome where at

7:13

first the genes were mapped and then function was

7:15

laid on top of that, this is

7:17

what's going to happen now. Now that we've defined

7:19

a blueprint of the cell types, now

7:21

we can start to understand what they are and what they

7:23

do. Great. That's great to hear.

7:26

Allison in Erie, Pennsylvania, welcome to Science

7:28

Friday. I'm inspired by

7:30

the limits of things. I know

7:32

Einstein's all things with thought experiments.

7:34

And lately I've

7:36

been appreciating how we can go

7:39

farther when we acknowledge what we don't. My

7:42

metaphor to start is just if

7:44

we smash open a radio, of

7:46

course the machine parts don't necessarily

7:50

give us all of the answers as to

7:53

the content that's coming through the radio, for

7:55

example. I just

7:57

wondered what this inspires in your

7:59

work. you and you're thinking about neuroscience and

8:02

I know we're talking about the cells but

8:04

what are the implications for you? I've

8:06

heard that this is paradigm shift. I'm

8:09

just curious what this inspires you,

8:11

whatever you found, how it's

8:14

inspiring your creativity in the

8:16

space in neuroscience, what it's

8:18

inspiring you to think about

8:20

or consider and it might

8:22

be a sensitive question because you're a scientist so

8:24

often you're not going to talk about things until

8:26

they're proven but I kind of wanted to challenge

8:28

you to talk about that. What

8:30

are some things that this inspires you

8:33

to think about or

8:35

even consider? Again, I'm not to challenge

8:37

the science of what's proven or not

8:39

proven but I know it's important

8:41

to be creative in the space and

8:43

metaphors help me a lot but yeah

8:45

it's exciting what we don't know and

8:48

just curious what it's inspiring you to

8:50

consider or think about. Good question. Let's get

8:52

the answer to that. What do you say

8:54

to that? Yes, that's

8:56

a very interesting question. First of

8:58

all, let me acknowledge

9:01

that this is really a reductionist approach to

9:03

the brain. In your

9:05

analogy of a radio, we take

9:08

it, we deconstruct its parts and we try to

9:10

understand its parts and very

9:12

much like that analogy, this

9:15

is just the beginning. Now

9:18

we know that the neurons

9:20

of the brain form circuits,

9:22

complicated circuits which are by

9:24

definition the connections between the

9:26

different kinds of cells. What

9:29

this really sets the stage

9:31

for is beginning to take

9:33

the next step. If

9:35

we don't understand the basic components, we can't

9:37

understand how they connect together. But

9:40

now that we understand that part of it,

9:42

we can move to the next stage. Even

9:45

that may not be the end of the day. There

9:48

may be software

9:50

on top of that that

9:52

actually dictates the function of this.

9:54

I think that this

9:57

is really just beginning but it is

9:59

an essential component. component of the process

10:01

here and one that we haven't had

10:03

before. If you don't understand the basic

10:05

building blocks, you can't understand how

10:07

it all fits together to work. And

10:11

I think that what's

10:13

important to consider here is that we've

10:15

had this sort of dearth of understanding

10:17

of the fine detailed structure of the

10:19

brain and that's dramatically hampered

10:22

our understanding of disease. And

10:25

this type of resource can now

10:28

allow us to bring this level of

10:30

resolution to understanding exactly what happens in

10:32

disease. And so to give a concrete

10:35

example and something I'm quite passionate about is

10:38

with this information, we can

10:41

think of cells as things that are

10:43

affected in disease that may in fact

10:45

be targets for treatment of disease. So

10:47

we have another project focused on

10:50

Alzheimer's disease called the Seattle Alzheimer's

10:52

Disease Brain Cell Outlets or CAD

10:55

where the idea is let's now look

10:57

at the brain of individuals that have Alzheimer's

10:59

disease and try to understand what

11:02

kinds of cells are affected in disease.

11:05

And we can map against this reference

11:07

and then ask questions about what the

11:09

real basis of disease is. Whereas

11:11

in the past, when we think

11:14

about pathological proteins, plaques and tangles

11:16

that everyone knows about that just

11:18

haven't worked as a way of

11:20

treating disease. And what we find is

11:22

that we can find all of these types

11:24

of cells and individuals with disease, we

11:27

can then see that certain types of

11:29

cells are differentially affected or vulnerable. And

11:32

so this is a kind of a different way

11:34

of thinking of things. It's thinking of the brain

11:36

as a cellular organ where specific

11:38

components do different things, they're differentially

11:40

affected by disease and

11:42

they become actual targets themselves. And

11:45

something else that comes from these atlases

11:47

is the ability to develop tools to

11:49

target particular cells and potentially deliver genetic

11:52

therapies to them. So

11:54

I view this as it's not a

11:56

stem collecting exercise. This is really foundational

11:59

work. that defines the system so that

12:01

we can understand

12:04

and treat disease. So how

12:06

would that work practically if you are trying

12:08

to treat something like Alzheimer's? I

12:11

mean how would that treatment potentially

12:13

work with this knowledge? Yeah,

12:17

so Alzheimer's is maybe a difficult example. It happens

12:19

to be one that I know a lot about.

12:22

But let me give a more

12:25

easy example perhaps of epilepsy.

12:28

So epilepsy is of course

12:31

an imbalance of excitation and

12:33

inhibition and often affects the

12:35

inhibitory cells. There's not enough

12:37

inhibition and you get

12:39

runaway excitation that's a seizure activity. One

12:43

of the things that's come out of this

12:45

cell atlas work is that we

12:48

can not only understand what genes are

12:50

active in what cells, but

12:52

how they're regulated to be so be

12:55

active only in those cells. So we

12:57

can identify the regulatory regions of the

12:59

genome that are responsible for turning on a

13:02

gene in only a certain kind of cell. That

13:05

regulatory element can actually be put

13:07

into a virus as a means

13:10

of gene delivery that's commonly

13:12

used in gene therapies now, such

13:14

as adenocessiated virus. These regulatory

13:17

domains can be put in to

13:19

turn on expression of a particular

13:21

gene, say a gene replacement of a

13:23

genetic epilepsy. And so you

13:25

can infect cells in people and

13:29

deliver a gene just to the right kinds

13:31

of cells that may be able to

13:33

correct those seizure phenotypes.

13:37

And this is just one example. Many types

13:39

of brain diseases will affect specific kinds of

13:41

cells. And what I'm

13:43

trying to convey is we can now

13:45

harness this information of this descriptive atlas

13:48

to build a tool to target the right cell type and

13:50

hopefully correct the genetic deficiency of some

13:52

sort without having side effects or off-target

13:55

effects that happen by hitting the wrong

13:57

kinds of cells. Really interesting.

14:00

technology. Let's go to the phones to

14:03

Ry in Houston. Hi, Ry. Yeah,

14:06

how are you, Ira? Thanks for taking my call. Yeah,

14:08

you're welcome. Go ahead. Yeah,

14:11

so my question is, has there

14:13

been any study of the connection

14:15

between the gut and the brain,

14:17

and is there any genomic similarities

14:19

that might, you know, create a

14:22

relationship or suggest one? Microbiome, our

14:24

favorite topic here at Science Friday.

14:26

Good question. Yeah,

14:28

that is an excellent question that

14:30

I'm afraid I'm somewhat ill-equipped to

14:33

answer. But let

14:35

me just say that I think that the data

14:38

are now becoming available to allow that

14:40

kind of question to be asked. For

14:43

example, in this human cell outlets that

14:45

I was mentioning earlier, there are efforts

14:47

to profile all the cells in the

14:49

gut and the immune system. And

14:52

so by virtue of

14:54

using these same types of technologies

14:56

to look at genetic similarities among

14:58

different kinds of cells anywhere

15:00

in the body, it will be possible to do

15:02

this and we can begin to start to see

15:04

where these functional links may be. But

15:07

I have to speak in generalities because I'm not aware

15:09

of studies that have done that today. I

15:11

mean, sometimes big projects like this raise more

15:14

questions than they answer. How much

15:16

of this project helps you better

15:19

understand the brain versus just opened

15:21

up more questions? Oh,

15:25

this is a huge advance in

15:27

understanding the brain. You know,

15:30

as I mentioned sort of before, you

15:32

know, this really forms a scaffold. It

15:34

forms a cellular framework. But

15:37

at the moment, these are cells that are

15:39

defined by genes with these kinds of methods.

15:41

And that's a very

15:43

powerful way of understanding a cell because the

15:45

set of genes that are selectively used by

15:48

a cell are the genes that

15:50

are responsible for the properties of those cells. So

15:52

this is much more information than simply saying,

15:55

you know, here's the shape of a cell.

15:57

We're going to categorize based on the shapes.

16:00

You can't do very much with that information. But

16:02

if you have all of the genes being asked,

16:04

now you can use this in a thousand different

16:06

ways. What drugs might

16:08

act on molecules that are

16:10

expressed only in certain kinds of cells, for

16:12

example? Up first achieves the

16:15

rare one-two punches of being

16:17

short and thorough, national and

16:19

international, fact-based and personable. Every

16:22

morning, we take the three biggest stories of the

16:24

day and explain why they matter. And we do

16:27

it all in less than 15 minutes. So

16:29

you can start your day a little more in the know

16:31

than when you went to sleep. Listen now

16:34

to the Up First podcast from NPR. What

16:38

about cells that are not actually in the

16:40

brain? For example, I've heard that your retina

16:43

is really part of your brain. We

16:45

have neurons that go through our spinal cord. We

16:47

talked about the gut a little bit, solar plexus.

16:50

Will those cells be included in

16:52

this catalog? So

16:55

at the moment, this particular effort is focused

16:58

on the central nervous system. Right.

17:00

Minus the retina. So the retina is actually part of

17:02

the central nervous system, but it's not being included. However,

17:05

it does profile or characterize

17:09

any kind of cell that's present in the brain,

17:11

whether or not this is an actual brain cell

17:14

or a circulating cell in the vasculature,

17:16

for example. And so actually,

17:18

one of the things to come out of this is that

17:21

there's a greater diversity of cells that

17:24

aren't neurons that are in the brain as

17:27

well. And so this really does give

17:29

you a very comprehensive view of the

17:31

overall makeup of the organ. And

17:34

I'm sorry, I've neglected to complete my

17:36

answer on the last days about whether we understand the brain.

17:42

With this scaffold, now you can have a very targeted approach

17:44

to start to characterize the properties of all these cells. So

17:47

the gene expression part of it is very informative, but

17:49

the next stage is what do these cells look like? Who

17:53

do they connect to? What are their firing

17:56

properties? What are their functions? And so now we're going to

17:58

talk about the brain cells. Now

18:00

that you are able to pin identities on

18:02

these cells, you can make very targeted inquiries

18:04

to start to annotate this or interpret your

18:06

functional experiments in light of what

18:09

kinds of cells are actually active in

18:11

a behavior. If you could, I'm going to give you

18:13

my blank check, if you had an enormous amount of

18:15

money to do or buy or create some kind of

18:17

tool to study what you want to know, what would

18:19

it be? I

18:24

think that I would actually invest

18:26

in the utility

18:28

of creating these tools to

18:30

target particular kinds of cells

18:33

that can help to understand the brain and treat

18:36

disease. I think there's an

18:38

enormous new field of precision medicine

18:40

that's being opened up by this

18:42

information. This

18:45

has already actually become, it

18:48

has become a huge priority for

18:50

the NIH that

18:52

is investing in another program as

18:54

part of its brain initiative that's

18:56

called the armamentarium or sort of

18:58

an arsenal of tools to be able

19:01

to genetically target and manipulate different kinds

19:03

of cells in the nervous system. I

19:06

think this is going to be enormously

19:08

important for medicine where we

19:10

now get a cellular understanding of

19:13

different diseases and can actually

19:15

target and try to correct those diseases.

19:17

Wow, wow, this is certainly exciting for

19:19

personalized medicine. Thank you, Dr. Aleen, for

19:21

the work that you do. Thank

19:25

you very much. That's it for today.

19:27

A lot of folks helped make the show

19:29

happen, including Jordan Smudgik, Charles

19:31

Bergquist, George Harper, John

19:33

Dankoski, and many more. Tomorrow, a

19:36

roundup of the top science news

19:38

of the week. I'm Sci-fi

19:40

producer Shoshana Bucksbaum. Catch you next

19:42

time.