0:00

Before the show starts, I'd like to ask

0:02

you to consider subscribing to News from Science.

0:04

You've heard from some of our editors on

0:06

here, David Grimm, Mike Price. They

0:09

handle the latest scientific news with accuracy

0:11

and good cheer, which is

0:13

pretty amazing considering it can sometimes be over 20

0:15

articles a week. And

0:18

you hear from our journalists. They're all over the

0:20

world writing on every topic under the sun, and

0:22

they come on here to share their stories. The

0:25

money from subscriptions, which is about 50 cents

0:27

a week, goes directly

0:29

to supporting non-profit science

0:31

journalism, tracking science policy,

0:33

our investigations, international news,

0:35

and yes, when we

0:37

find out new mummy secrets, we report on that

0:40

too. Support non-profit

0:42

science journalism with your subscription

0:44

at science.org/news. Scroll down and

0:46

click subscribe on the right

0:48

side. That's

0:51

science.org/news. Click

0:53

subscribe. This week's episode

0:55

is brought to you in part by Science Careers.

0:58

Looking for some career advice, wondering how to

1:00

get ahead or how to strike a better

1:02

work-life balance? Visit our site to

1:04

read how others are doing it. Use

1:07

our individual development plan tool,

1:09

access topic-specific article collections, or

1:12

search for an exciting new job. Science

1:15

Careers, produced by Science and AAAS, is

1:17

a free website full of resources to

1:19

help get the most out of your

1:21

career. Visit sciencecareers.org

1:24

today to get started.

1:32

This is a science podcast for January 12th, 2024. I'm

1:36

Sarah Crespi. First up, we

1:38

have journalist Rich Stone. He's back with news

1:41

from his latest trip to Ukraine. This

1:43

week he shares stories about environmental damage

1:46

from the war, particularly the grave consequences

1:48

of a dam explosion. Next

1:50

on the show, producer Kevin

1:52

McClain talks with researcher Nardi

1:55

Gomez-Lopez about signaling between fetus

1:57

and mother during childbirth and

1:59

how understanding... Understanding this crosstalk might

2:01

one day help predict premature labor.

2:04

Finally, in a sponsored segment from our custom

2:06

publishing office, Director of Custom

2:09

Publishing Erica Berg talks with researcher

2:11

Andrew Postefilik about how epigenetics

2:14

stabilizes particular gene expression patterns

2:16

and how these patterns impact our risk

2:18

for disease. We've

2:24

talked about the war in Ukraine a number of times on

2:26

the show with the science focus.

2:28

We talked about how the war has

2:31

affected science in Ukraine, can scientists keep

2:33

doing their jobs, how facilities fared. We

2:36

also had a segment on how Russian scientists

2:38

are leaving that country because of disagreements with

2:40

government policy and the prosecution of the war,

2:42

as well as about the state of science

2:44

and scientists who have remained in Russia. This

2:47

week we're going to hear about the environmental

2:49

toll this war has had on Ukraine. Contributing

2:52

correspondent Rich Stone visited in the fall.

2:55

Hi Rich, welcome back to the Science Podcast.

2:58

Hi Sarah, glad to be back. The focus,

3:00

as I mentioned in the intro, is about

3:02

the environment impact of the war and the

3:04

scientists. But we should really just mention

3:07

that this war has had an immense impact

3:10

on the lives of people in Ukraine, the infrastructure,

3:12

you know, there's been many deaths.

3:14

So we can't just ignore that and

3:16

only talk about science. For sure. Hundreds

3:19

of thousands of people have died during the

3:21

war. That's been a tragedy

3:24

for Ukraine and certainly

3:27

the ecological consequences that we're going

3:29

to talk about also affect people's

3:32

daily lives. Yeah, for sure. So

3:34

we're going to mainly talk about the loss

3:36

of this dam in Ukraine that was holding

3:38

back a lot of water and formed a

3:40

reservoir. It was destroyed by

3:43

a pair of explosions over the summer.

3:45

Can you set the scene for us here? The D

4:00

of the country into the Black

4:02

Sea. And it's one

4:05

of the major rivers of Europe, not just Ukraine.

4:07

During Soviet times, there

4:10

were six dams, hydroelectric dams

4:12

built on the Nipro River.

4:14

The Kolkata was the final

4:16

dam and the one furthest

4:18

downstream. And that's the one

4:20

that was breached. On

4:22

June 6 of this year,

4:24

two explosions, damage to dam.

4:26

There was a massive breach around

4:29

400 meters wide and

4:31

flood waters surged down into

4:33

the lower Deepro Basin inundating

4:36

hundreds of square kilometers of land.

4:39

Yeah, so there was a huge reservoir behind this. 400

4:41

meters is a huge breach. And

4:44

all this water came out. What were the immediate

4:46

effects? What do we know about the consequences of

4:48

all this water flooding down like

4:50

that? Well, let's start in

4:52

the former reservoir, which mostly

4:55

emptied out. That left a lot

4:57

of dead fish on desiccated riverbed.

5:00

It also left around 500,000 tons

5:02

of zebra mussels rotting

5:06

in the summer sun. These are

5:08

the bad guys in the Great Lakes in the

5:10

US, but they're not the bad guys in

5:13

Ukraine. They are pretty notorious invasive

5:15

species in the US. In Ukraine,

5:17

they occur naturally. And they're very

5:19

important ecologically. They filter the water

5:22

column. A lot of the

5:24

water that was reaching the Kukovka reservoir

5:27

is runoff from agriculture, sewage, and

5:29

so forth. And

5:31

the zebra mussels were performing

5:33

incredibly valuable ecosystem service in

5:36

filtering this water before it

5:38

flowed downstream. So

5:40

immediately, you lost that ability

5:42

to filter water entering the

5:44

Deepro River north of the

5:46

former dam site. That's a big loss of

5:49

ecosystem service. And

5:51

there are other effects as well.

5:54

The emptying of the

5:56

reservoir created this marshy

5:59

area femoral ponds and

6:01

lots and lots of mosquitoes. So

6:03

it was a big mosquito

6:05

breeding area. So it

6:07

was not a pleasant place to be

6:09

this past summer and it's going to

6:11

take a long time for these dead

6:13

organisms to decompose. Right. What about

6:15

downstream from the reservoir? You know, where

6:18

this water went? What do we

6:20

know about the impact of the release there?

6:22

Yeah, so it flooded around

6:24

80 settlements downstream, including a

6:27

pretty major city of Harsone,

6:29

which the Ukrainians had recaptured

6:32

from Russia last fall.

6:35

One of the scientific facilities that

6:37

was in harm's way was

6:40

an experimental sturgeon breeding center

6:42

that would breed sturgeon

6:45

that are adapted to the

6:47

Nipro River environment and release

6:49

the sturgeon into the Nipro

6:52

reservoirs. The station

6:54

took a direct hit before a meter

6:56

of wall of water came in. Wow.

6:59

Really devastated it and wiped out the

7:01

sturgeon, which are very sensitive

7:03

to toxicants. There was a

7:05

lot of pollution in the water, heavy metals.

7:07

One good thing is that the staff of

7:10

the station evacuated in time. This

7:12

is a very dangerous environment for study if

7:15

people want to go check on

7:17

the water on the reservoir. Some of

7:19

this river is affront in the war.

7:21

Were you able to talk to people about that experience? The

7:24

Nipro River forms more than 300 kilometers

7:27

of the front line, actually. So very

7:29

dangerous to

7:32

go to the Ukrainian side of the

7:34

river and take samples. I

7:37

interviewed Balodomir Osace of

7:39

the Ukrainian Meteorological Institute.

7:42

He and a colleague

7:44

had gone down to take

7:46

water samples during a break in the shelling,

7:49

a predictable break in the shelling. They

7:51

could see the other side of the

7:53

riverbank less than a kilometer away. That's

7:55

where the Russian troops were. terrifying

8:00

experience, a reconnaissance drone flew right

8:02

over them. They hit

8:04

the ground and they felt very

8:07

lucky that it had passed, but it

8:09

really brought home how dangerous this

8:11

work is. So they got their sample, brought

8:13

it back to the lab, and they could

8:15

study the nutrient profile of

8:17

the water column. You mentioned

8:19

visiting Odessa, which is where this

8:22

water eventually makes its way to the Black

8:24

Sea. Was it safer there?

8:26

Were people able to do some of the

8:28

research that we're talking about on those shores? Yeah,

8:31

Odessa is far enough downstream and

8:33

a little bit away

8:36

from the river. So the Nipro River

8:38

empties into Estrary, and then

8:40

the Estrary empties into Odessa

8:42

Bay. Odessa Bay is

8:44

under Ukrainian control. It's

8:47

a risky place in general because this

8:49

is one of the major ports where

8:52

Ukraine ships grain to Europe

8:54

and other countries, and

8:56

the Odessa port is often attacked

8:59

by Russian drones and missiles. So

9:01

in Odessa, there were frequent air raid

9:04

sirens when I was there, but

9:06

it's away from the front line and

9:08

scientists are able to take samples. They

9:10

observed how the freshwater came into the

9:13

Bay, reduced the salinity

9:15

dramatically, essentially for a little while

9:17

it was freshwater at the mouth

9:19

of the Bay. That's really rough

9:22

on an estuary. That is devastating

9:24

for marine life, acclimated to salinity

9:27

that can't move. Mollusks,

9:29

for example, fish consume

9:31

away. The cetaceans, dolphins and

9:33

corpusces consume away, but the

9:36

benthic life was just devastated by

9:38

that freshwater. That's a good

9:40

segue. You do mention cetaceans in the

9:42

Black Sea here. Has there been an

9:44

impact from the war on dolphins or

9:47

other mammals in the water? It's a

9:49

little bit controversial. So there are

9:51

three subspecies of

9:54

cetaceans that are found endemic in

9:56

the Black Sea. There are two

9:59

kinds of. dolphins and one kind

10:01

of porpoise. Last year, there

10:03

were claims that tens of thousands

10:06

of cetaceans in the Black Sea had

10:08

died because of the war, possibly

10:11

because of acoustic damage due to

10:13

high powered sonar

10:15

by Russian submarines, explosions of

10:17

underwater mines, et cetera. It

10:20

was a claim that has not held

10:22

up under scrutiny. Certainly

10:24

there wasn't uptick in mortality

10:26

last year during the early part

10:28

of the war. That

10:31

seems to have subsided.

10:33

There were much fewer

10:35

strandings reported this year.

10:38

But the cause of the

10:40

excess deaths last year is

10:43

still under investigation. Researchers

10:45

autopsy cetaceans when they can,

10:48

but the preliminary findings have

10:50

not pointed to a single

10:53

factor for the death of the

10:55

cetaceans. Another reason that

10:57

researchers are doing this isn't just

10:59

to monitor the situation with these

11:01

ecosystems, but also to gather evidence

11:03

for environmental war crimes. So how

11:05

would these things fit in that

11:07

category? This is

11:10

being done at the central

11:12

government level. The Ukrainian general

11:14

procurators office is gathering evidence

11:17

of potential war crimes, environmental

11:19

war crimes. So

11:21

this evidence comes from various

11:23

scientific studies. So the cetacean

11:26

autopsies, all that data goes up to Kiev

11:28

to build the case. The

11:31

documentation of the ecological losses

11:33

in the Dipra River and

11:36

the reservoir, the

11:39

losses to irrigation,

11:41

to municipal water supplies,

11:44

et cetera. So Kolkovka has had

11:46

very broad consequences for the

11:48

country. In addition to that,

11:50

there's been destruction of protected

11:52

forests. There's mines that

11:54

have been laid, innumerable mines along

11:56

the front line, and

11:59

the shelling. the incredible impact of

12:01

the shelling and contamination

12:03

from artillery shells in

12:05

farm fields and forests

12:08

is something that is going to both

12:10

be documented as potential

12:12

war crimes, but also will take

12:15

generations to clean up. Right,

12:17

yeah, the remediation is a really big ongoing

12:20

and growing problem for, as

12:23

you say, all these military actions and

12:25

also these ecosystems that have been so

12:27

damaged. What about the dam? Is that

12:29

something that people want to

12:31

put back? Because, you know, dam removal

12:33

is a big issue in other places.

12:36

The National Academy of Sciences of Ukraine

12:38

has come up with a few different

12:40

options that the government can consider. One

12:43

is simply rebuild the dam pretty much

12:45

as it was and

12:47

that would restore water regulation, so

12:49

water releases during

12:51

the dry season. It would maintain

12:53

both the economic activity in the

12:56

region and the ecosystems, the water

12:58

needs of the ecosystems. There's

13:01

a second idea that also involves

13:03

rebuilding the dam and then also

13:05

building an embankment about 50 kilometers

13:07

long that would reduce the size

13:10

of the reservoir, shave

13:12

off a big stagnant area of

13:14

the former reservoir, just allow that

13:16

to recover as step grassland.

13:19

That is one option a lot of scientists

13:21

favor. And then there's a third option

13:24

and that's just let nature take its course. This

13:26

is essentially a dam removal

13:28

that was not planned and

13:30

has that devastating consequences, but

13:33

people say why not let the

13:35

area rewild. There's a smaller number

13:37

of scientists in Ukraine who favor

13:39

that and it appears the government

13:42

is not disposed to allowing

13:44

natural recovery that way. I

13:47

think there is a intention at this point

13:49

after the war to rebuild the dam. Even

13:52

then there could still be the opportunity to

13:54

rewild parts of the lower Nipro Basin even

13:57

with the rebuilding of the dam. So you

13:59

actually... have a sidebar with your

14:01

story that just focuses on environmental

14:04

contamination from radiation, which has been an

14:06

ongoing concern since the very beginning of

14:08

the war. I actually had, you

14:10

were on the show, I don't know if it was the last time

14:12

or the time before, where we just talked about Chernobyl. So

14:14

yeah, like what is the situation

14:17

with all the uranium, the reactors,

14:19

the milling facilities? Has anything happened

14:21

there? Are there still concerns? Well,

14:24

Ukraine's nuclear infrastructure has been under attack

14:26

from the very beginning of the war,

14:29

as you know, we did

14:31

talk about Chernobyl last year, but

14:33

there's operating nuclear power

14:35

plants that have been under

14:38

siege. So the Zaporizhia nuclear

14:40

plant is the largest in

14:42

Europe. It's now under Russian

14:44

occupation and the reactors have

14:46

been shut down, but they

14:48

still pose a potential risk

14:50

if a missile were to

14:52

hit either the reactor vessel

14:55

or the spent fuel that's stored on

14:57

site that could distribute contamination

14:59

in the area. A lot

15:01

of these concerns are dirty bomb sorts

15:03

of concerns. As opposed to

15:05

the ignition of fissile material, it's more,

15:07

you know, this idea of a dirty

15:10

bomb where you have a regular

15:12

kind of bomb that just is

15:14

contaminated with radioactive material and

15:16

it's spread all over the

15:18

place. Exactly. And so some

15:21

of the other nuclear plants in

15:23

Ukraine, there's been missile and drone

15:25

attacks near the plant that raise

15:27

a lot of concerns. And

15:29

there's just a lot of

15:31

concern about the potential for

15:34

what's called false flag operation.

15:36

So Russia has, propaganda has,

15:38

saying that Ukraine actually wants

15:40

to detonate a dirty bomb

15:42

on its own soil. Okay. So there's all

15:44

this propaganda saying, oh, they're going to blame

15:46

Russia if they do that. And

15:49

the Ukrainians and their Western

15:51

allies consider that propaganda as

15:54

arranging the information space such that

15:56

if there was a dirty bomb

15:58

attack perpetrated by Russia. Russia

16:00

could say, you see, I told you so. Ukraine did

16:02

this and blamed us. There's

16:04

that concern has been

16:07

around since early on in

16:09

the war and it has not gone away.

16:11

There's other sites as well that have been

16:13

under the radar like you mentioned, the

16:15

uranium milling site, one of the biggest

16:17

uranium mills during the Soviet Union. It

16:19

has been shut down, but there's still

16:21

a lot of radioactive material on the

16:24

site. If that were hit by

16:26

a missile, it would contaminate the local city. Lots

16:29

of perils on the radioactive landscape

16:31

in Ukraine. Definitely. Thank you

16:33

so much, Rich. Rich Stone

16:36

is a freelance science newsreader. He's

16:38

also the senior science editor for

16:40

the Howard Hughes Medical Institute's Tangled

16:42

Bank Studios, where he oversees science

16:45

content for documentaries and other nonfiction

16:47

productions. You can find a

16:49

link to the story we discussed today

16:51

at science.org/podcast. Stay tuned for

16:53

a chat with producer Kevin McLean and

16:56

researcher Nardi Gomez Lopez on

16:58

reading the transcripts of conversations between

17:00

fetus, placenta, and mom during childbirth.

17:11

For both mother and fetus, there's a

17:13

lot that has to go right in

17:15

order for a healthy, spontaneous labor and

17:17

birth to occur. Tissues

17:19

and cells of multiple bodies have to be

17:21

ready to go through a lot of changes

17:23

in a short amount of time. Throughout

17:26

childbirth, we know there's a

17:28

complex process of cellular communication

17:30

and signaling happening. And

17:33

this maternal fetal crosstalk is

17:35

important, but we don't

17:37

know a lot about how it unfolds

17:39

and the mechanisms that control it. This

17:42

week in Science Translational Medicine, Nardi

17:44

Gomez Lopez and her colleagues used

17:46

single cell RNA sequencing of the

17:49

placenta in order to

17:51

investigate the critical cell types

17:53

and communication pathways active during

17:56

childbirth. Nardi, welcome to

17:58

the Science Podcast. Thank you, Kevin. for

18:00

having me. Great. Well, just to start

18:02

off, childbirth or parturition

18:04

is such a foundational part

18:07

of life. Why is

18:09

this an area that we don't know

18:11

that much about? Is it just that

18:13

it's complicated and tools haven't caught up

18:15

or has there been less research focus

18:17

on it historically? Yeah, that's

18:19

a great question. I think it's

18:21

a combination of multiple

18:23

factors. The first is,

18:26

it's a very complex process. It involves

18:28

multiple dishes from the

18:30

mother and there is

18:32

also signaling from the fetus. The

18:35

placenta allows us to interrogate

18:37

both. That's the reason why

18:39

we utilize the placenta. However,

18:42

the mother has to

18:44

undergo different processes such

18:46

as contraction, cervical ripening

18:49

and there are different processes

18:51

in the interneopic space where

18:53

the placenta is during gestation.

18:56

Now, the technologies

18:59

also have advanced. With single

19:01

cell RNA sequencing, we can interrogate

19:03

every single cell from a specific

19:05

tissue. And in this

19:07

case, the placenta, because it has

19:10

maternal and fetal cells circulating in

19:12

the maternal-fetal interface, then we can

19:14

decipher every cell and how they

19:17

are communicating with each other. Can you

19:19

talk a little bit about

19:21

what this maternal-fetal crosstalk entails

19:23

and some of the challenges

19:25

involved in studying that kind

19:27

of communication? Yes, there

19:29

is maternal-fetal crosstalk from very

19:32

early in pregnancy and

19:34

these communications change throughout pregnancy.

19:36

So, we're speaking about 40

19:39

weeks of gestation in a normal term

19:42

delivery baby. So, those

19:44

communications change. Also, there

19:46

are processes that happen locally like

19:48

the maternal-fetal interface where the placenta

19:51

and the maternal tissues encounter but

19:53

also in the circulation. So, studying

19:57

maternal-fetal crosstalk, it's complicated.

20:00

because it involves different tissues In

20:18

our case, we are focusing on parturition so

20:21

that the reason why we utilize the tissues

20:23

at the time of parturition and we

20:25

collect maternal blood samples before parturition

20:28

of course Collecting samples and coordinating

20:30

with all of those births must

20:32

have been a really complicated process

20:34

in itself but the placenta really

20:36

contains some of the key information

20:38

you need for this work What

20:40

was that whole process like? In

20:44

order to collect samples, it has

20:46

to be thought well and coordinated

20:48

with the physician-scientist and also that

20:51

it's safe for the mother but

20:53

the reason why we focus on the placenta is

20:55

because it's the most important organ during

20:58

pregnancy and because it contains both

21:00

maternal and fetal components The placenta

21:02

is a fetal organ but because

21:04

it has maternal blood circulating

21:06

through specific areas that are

21:09

called intervial space then we

21:11

can contain both maternal and

21:14

fetal components Also the

21:16

placenta is attached to the myometrium

21:19

and the myometrium is

21:21

the muscular wall of the uterus?

21:23

Yes, and the decidualized

21:25

myometrium is maternal tissue

21:28

Therefore, when we collect the placenta, we can

21:30

collect also the decidua which is

21:32

both maternal or it is Okay, so

21:35

you have all these different tissues from

21:37

the placenta, from the decidua, the intervial

21:39

space components from both

21:41

mother and the fetus So what

21:43

happens next? What do you need

21:45

to find in all of those

21:47

tissues? So we wanted

21:50

to first identify what

21:52

cells were present in the

21:54

placenta and to

21:56

do so we utilized different computational

21:58

methods and The first

22:00

challenge was to name those

22:02

cells. What are these cells doing? Right?

22:05

So that was the difficult part to do because

22:07

we didn't have a reference. So we

22:09

utilize information that was

22:12

publicly available as well as

22:14

our own biological concepts. These

22:16

cells looks like a T cell.

22:18

These cells look like a trophoblast.

22:20

You know, so that was a

22:22

very time consuming process. So

22:24

that was the most difficult part, I

22:27

think, to name what cell is

22:29

what and what are that cell is

22:31

specifically doing. So we found

22:33

multiple cell types and there are

22:35

many, many cell types that we

22:37

have filled closer than together because

22:39

we couldn't assign specific limits

22:42

to those cells, but we could say

22:44

these cells look like they are immune

22:46

cells. These looks like they are not

22:48

immune cells. And then we were more

22:51

granular and we identify specific types

22:53

of immune cells, specific types of

22:55

non-immune cells. And of course there were cells

22:57

that we were expecting like trophoblast, because it's

22:59

the main cell in the present. All

23:01

right. So you sort all of these

23:04

cells by cell type and you look

23:06

at the cells individually and that's what

23:08

creates your reference map, your atlas. How

23:11

does that process happen? So

23:14

what we are doing with single-cell

23:16

RNA sequencing is dissociating

23:18

the tissue. Every

23:20

single cell gets encapsulated into

23:23

a droplet and there are

23:25

specific reactions that occur in every cell

23:27

and that will allow us to read

23:29

the transcripts of the gene

23:31

expression in every single cell. And

23:34

through computational methods, then we

23:36

can decipher where each

23:38

transcript came from or what is

23:40

every cell doing in the transcriptome

23:43

and how are these cells communicating, what

23:45

are the mediators that these cells are

23:48

producing? What are the receptors that these

23:50

cells are expressing? And of

23:52

course they will change depending on different processes

23:54

that occur in the specific tissue and then

23:56

we change with the process of label. Right.

23:58

So with this, you're able to

24:01

look at the changes throughout the process

24:03

of labor, what did this help you

24:05

understand? The next thing was,

24:07

okay, now that we have these

24:09

atlas, how are these cells changing

24:11

when we compare tissues

24:13

from women with labor and tissues from

24:16

women without labor? So then

24:18

we could see the majority of the

24:20

cells obviously stay there, but their

24:22

transcriptomic activity, so their gene expression

24:25

change with the process of labor.

24:28

We already know that the

24:30

process of labor involves

24:32

inflammation, involves different types

24:34

of biological processes. What

24:37

we didn't know before this paper is, where

24:39

are these processes coming from? So

24:42

now, with single-cell RNA sequencing and

24:44

by us naming that

24:46

specific cell type, now we know

24:49

that those specific biological processes or

24:51

specific transcripts come from that

24:53

specific cell. Got it. Okay,

24:56

so specific cells change throughout labor,

24:58

and you can see that in

25:00

the transcription activity and gene expression,

25:03

but what about this communication

25:05

part, this coordination between maternal

25:07

and fetal tissues? Now

25:09

we need to investigate whether they are

25:12

talking between them, whether they are

25:14

communicating. So there are

25:17

specific computational analysis that

25:19

is basically works like a cell phone.

25:22

You want to call someone, so then

25:24

you have their phone number, you

25:26

dial the number and that person answers. Cells

25:29

communicate in a similar way. They

25:32

have specific mediators that they are released.

25:34

If they want to call specific cells

25:36

and those other cells that are the

25:38

respondents, they will have to

25:40

pick up the phone, and to do so, they will

25:42

have to express a receptor. So

25:45

we evaluated the cell-cell

25:47

interactions by evaluating the

25:49

ligands that a specific cell is

25:51

producing and that the

25:53

receiving cell is expressed. Does

25:55

that make sense? Yeah, that's so interesting.

25:57

So you're sort of taking look

26:00

at who is calling who and

26:02

then seeing who is receiving that

26:04

call to sort of get a

26:06

sense of this communication. Exactly. Right.

26:10

You know, one of the things that I found

26:12

really fascinating was the fact that you were able

26:14

to detect signals in the

26:16

blood of pregnant women who went on

26:18

to have preterm births. Can you talk

26:20

a little bit about that process and

26:23

how that was possible? It

26:25

is very well known. Pretty many years ago,

26:27

it was demonstrated by multiple smarter

26:30

people than us that there

26:32

were placental particles in the maternal

26:34

circulation of women. So that was

26:37

already known. Our

26:39

thought process was if the

26:41

placenta sheds out into the

26:43

maternal circulation, are we

26:45

able to identify signatures in

26:47

the maternal circulation if we

26:49

know what cell it's

26:51

sending the signal locally? Because

26:54

we already have the identity of

26:56

the cells in the placenta because

26:58

we did single cell RNA sequencing.

27:00

So what we did is to

27:03

take those specific identities and look

27:05

for them computationally in the maternal

27:07

circulation. We first look

27:09

at whether these signatures, we

27:11

call them single cell signatures or

27:13

placental signatures, were present in the

27:15

maternal circulation of normal women. We

27:18

found that that was the case. Then

27:20

we look for whether these

27:23

signatures were somehow different in

27:25

women who have preterm labor.

27:27

So preterm labor is contractions and basically

27:30

activation of labor before the seven weeks

27:32

of gestation. So we

27:34

look at these women who have preterm

27:36

labor and eventually deliver preterm. So

27:38

we saw the signatures there. And

27:40

then when we found our signatures

27:42

to be somehow specifically changed in

27:45

women with preterm labor who ultimately

27:47

deliver preterm, we were very excited

27:49

because we thought we may

27:51

be able to generate a

27:53

biomarker to identify women before they

27:55

come with an episode with both preterm labels.

27:58

And we found that they were specifically. signatures

28:01

that have a modest predicted

28:03

value for pretender. And

28:05

I want to emphasize that this is a modest

28:07

predicted value. And what

28:10

is exciting about our story, I think,

28:12

is the proof of concept. We can

28:14

identify the signatures. They may

28:16

have a predicted value for a specific

28:19

type of pretender because pretender

28:21

is so complicated. It's

28:23

really interesting to know that it's possible

28:25

to even detect that. Like you mentioned,

28:27

this proof of concept, why

28:30

is knowing these signals potentially early

28:32

on really useful? Do you think

28:34

that there would be into the

28:36

future something that could be done?

28:38

If you know somebody may be

28:40

predisposed or they might

28:42

possibly have a preterm labor,

28:45

what's helpful about knowing

28:47

that that is a possibility earlier

28:49

on in the pregnancy? The

28:51

idea is to have a biomarker that

28:53

will allow us to identify women

28:56

who eventually deliver pretender. But

28:58

now we want to identify with

29:01

all these technologies that we have

29:03

a specific subset of pretender. Maybe

29:06

the reason why we have not been

29:08

able to identify women who eventually deliver

29:10

pretender is because we are trying to

29:12

identify all women who deliver pretender. So

29:15

we believe that by dividing and

29:18

there's different neptiologies, then we can

29:20

maybe be able to conquer a

29:22

specific subset of pretenders. So

29:25

the idea is to tailor

29:27

the therapies to a specific

29:29

subset of pretender. Interesting.

29:31

Yeah. And can you put in context,

29:33

I mean, you mentioned preterm birth and

29:36

preterm labor is, there

29:38

are many different things involved, many

29:40

different potential causes, but it's a

29:42

very serious complication that can happen

29:44

and it continues to be a

29:46

big problem worldwide. Is that right?

29:48

Yes. It's in the United

29:51

States, one of every 10 babies is born

29:53

preterm. So that is

29:55

pretty devastating. We have

29:57

been focusing on investigating what pretember

30:00

happens and where the geologists, the

30:02

mechanisms of disease that lead to

30:04

pretember. We know that

30:07

in some cases there are bacteria

30:09

invading from the vagina into the

30:11

cervix, into the amniotic cavity and

30:13

that causes infection and

30:15

that leads to pretember. And

30:18

usually it's associated with very early

30:20

gestation. For those types

30:22

of pretember in which we can identify

30:24

bacteria in the amniotic cavity, there

30:27

are a specific set of antibiotics to

30:29

treat these women. So we know that

30:31

the rights regimen of antibiotics work for

30:33

those women with infection in their meredicalities.

30:36

But it's only a small portion

30:38

of pretenders. The majority

30:40

of pretenders, we don't know the geologists.

30:42

We also know that there is a

30:44

subset of pretenders that have inflammation.

30:47

The rest of pretenders, we don't know

30:49

why they happened. It may be because

30:51

the placenta is not working well. It

30:54

may be because some type of

30:56

immune dysregulation. So the idea is

30:59

to generate biomarkers that would allow

31:01

us to identify this specific type

31:03

of pretember so that we can

31:05

have tailored strategies to treat

31:07

that specific type of pretender. Great.

31:10

Well, thank you so much for joining us,

31:13

Nardi. It was really great talking with you.

31:15

Oh, thank you so much. Nardi

31:17

Gomez Lopez is a professor in

31:19

the Department of Obstetrics and Gynecology

31:21

and Pathology and Immunology in the

31:24

Center for Reproductive Health Sciences at

31:26

Washington University School of Medicine. You

31:28

can find a link to the

31:31

paper we discussed in Science Translational

31:33

Medicine at science.org. Next,

31:36

we have a custom segment sponsored by the

31:38

Van Andel Institute. Custom Publishing

31:40

Director Erica Berg chats with

31:42

researcher Andrew Pospasilik about how

31:44

epigenetic patterns impact our

31:47

risk for disease. The views expressed in

31:49

the custom segments are those of the guests and

31:51

do not reflect the policies of science or AAAS.

31:56

Hello to our listeners and welcome

31:58

to this sponsored interview for from the

32:00

Science AAAS Custom Publishing Office and

32:03

brought to you by Van Andel

32:05

Institute. I'm Erica Berg, Director

32:07

and Senior Editor for Custom Publishing at

32:09

Science. Today, I am

32:12

delighted to welcome Dr.

32:14

Andrew Pospicilik, a

32:16

leader in the field of epigenetics.

32:18

We'll be having a conversation about

32:20

how our epigenetics makes it more

32:23

or less likely that will go

32:25

on to develop diseases such

32:27

as diabetes, Parkinson's, or

32:29

cancer. Thank you so much

32:31

for joining, Andrew. Thanks, great to

32:33

be here. So my first question

32:36

is just to get us all on the

32:38

same page. Can you provide

32:40

us a brief overview of

32:42

what epigenetics is and how

32:44

it regulates gene expression and

32:46

metabolism? So epigenetics is the

32:49

set of molecular processes that I think

32:51

give stability to our gene expression profiles.

32:53

So it's all the bookmarks that tell

32:55

us the cell to be the kind

32:57

of cell it is, whether that's a

32:59

liver cell or a brain cell or

33:02

a beta cell in the pancreas. And

33:04

it's the mechanisms that help make sure that

33:06

those cells, once they've figured out what they

33:09

are, they stay doing what they're supposed

33:11

to be doing and they stay who they are. And

33:14

typically we think of these processes

33:16

and these mechanisms as

33:18

being the set of molecular mechanisms

33:20

that packages DNA. So

33:22

you can package DNA really tight so that

33:25

a certain gene won't ever be expressed. You

33:28

can package it in a way that

33:30

it can be inducibly expressed and turn

33:32

on when needed. Or you

33:34

can package it in a totally open way

33:36

where no matter what, it will

33:38

always be expressed and the gas pedal

33:40

can just be pushed down a little bit

33:42

harder or less. What is

33:44

it that first interested you in the

33:47

study of epigenetics? After my PhD,

33:49

which was focused very much

33:52

on translational endocrinology and the

33:54

idea of generating future

33:56

diabetic medications, I joined

33:58

a mouse lab, so a functional... genetics lab

34:00

where we knocked out certain genes

34:02

and discovered what they did in the body.

34:06

And one thing that I always found interesting

34:08

was that we have these isogenic mice, so we

34:10

have mice that we can get 10, 20,

34:14

30, 100 mice that all have essentially

34:16

an identical DNA sequence. But

34:18

no matter what measurement you do, they're all a little bit

34:21

different. And so it got me

34:23

thinking about variability or stochasticity

34:25

some people would call it, or

34:27

how probabilistically could things go

34:29

a different direction. And

34:32

almost no matter what scientific measurement you

34:34

do, there's always some distribution. And

34:37

sometimes those distributions are odd that you

34:39

see reproducible patterns that the biggest

34:42

mouse is also maybe the most active

34:44

mouse or the most curious one or

34:46

the one that bites the least. And

34:49

this started getting me onto these ideas of

34:52

reproducible phenodepic patterns that

34:54

don't necessarily have so much

34:56

to do with the DNA sequence itself,

34:59

but rather about stabilizing gene expression patterns

35:01

in it. Wow, well

35:04

get me those less bitey mice. We

35:08

talked about how epigenetics was

35:10

related to how a cell becomes

35:12

who it is, but how does

35:14

epigenetics impact our risk

35:17

for diseases like diabetes,

35:19

neurodegenerative diseases, cancer, even

35:22

obesity? So it's

35:24

a good question. I think you can break it down

35:26

in a few different ways. You

35:28

can think of epigenetics at the cellular level, you

35:31

can think of epigenetics at a whole organism

35:33

level, and you can even

35:35

think of it across generations. So

35:37

an epigenetic signal goes from parent

35:40

through the germline, so through the sperm

35:42

and the egg to the next

35:44

generation could have lifelong programming consequences.

35:47

And those have been documented

35:49

repetitively across almost every organism.

35:52

We don't understand exactly how that works for most cases,

35:54

but then we can go right to the other extreme

35:57

and we think of a cell where

35:59

the epigenetic program... is a

36:01

little bit different and then let's

36:03

take a disease like cancer. If a cancer

36:06

requires a certain part of the DNA

36:08

to emerge, so expression of a certain

36:10

locus, epigenetics is what controls

36:12

whether that locus is open or not.

36:15

So the epigenetics could in a

36:17

sense determine whether a cell can

36:20

easily become cancerous or may

36:22

be quite resistant to it. Take that

36:24

analogy and extend it to, you know,

36:26

if we talk about metabolic disease, we've

36:28

recently shown that there's two types of

36:30

beta cells and they're really defined by

36:33

epigenetic differences. We call them high

36:35

cells and low cells because one

36:37

of them has very high levels of a silent

36:39

chromatin mark and the other cell type

36:41

has very low levels. That means

36:43

in those two cell types, the gene expression patterns

36:46

are going to be different, the high cells happen

36:48

to be much more functional and

36:50

if you have more of those functional cells,

36:53

you'll tend to fight off diabetes. From

36:56

your research, from what you've learned, is there a

36:58

way I can get more of the better beta

37:00

cells or is that not sort of why you're doing

37:02

this research? I assume part of it is

37:05

wanting to figure out like, okay, we've got

37:07

these two populations of beta cells, good

37:09

ones, less good ones, how do we

37:11

enrich the good ones or is that

37:14

way off base? No, I

37:16

think that's exactly the motivation. I think it's a little

37:18

bit early for us to be able to say that.

37:21

What we do know is if you manipulate the

37:23

dosage of the epigenetic marks, just a little bit.

37:25

If we don't take out a whole gene, we

37:27

don't knock out the whole gene but we just

37:29

take out one of our two

37:31

copies of the gene, in this case the methylation

37:33

of histone 3-lysine 27,

37:36

and if we increase

37:38

it or decrease the methylation of that, we

37:40

can skew the ratio of kind

37:43

of the better beta cells to the

37:45

average beta cell. So it's definitely possible.

37:47

That then begs the question, well, could

37:49

we use dietary interventions and, you know,

37:52

methionine is a substrate to generate

37:54

methylation? And so in the very

37:56

simplistic terms you could imagine, oh,

37:59

we're getting increasing the thionine in our diet,

38:01

increased methylation, get more of the better

38:04

beta cells. But that's not something we've

38:06

proven. That's the next step. You

38:08

know, this story is still quite new. Your

38:12

lab has made significant

38:14

advances such as identifying

38:16

disease-causing epigenetic defects in

38:18

diabetes and defining pathways

38:20

for healthy versus unhealthy obesity.

38:24

Could you elaborate on these breakthroughs

38:26

and their implications? Yeah,

38:28

so for the question

38:30

about healthy and unhealthy obesity, these

38:32

alternate developmental states that we characterized,

38:35

and so we've characterized them in

38:37

mice and identified some of the

38:39

regulators that are required

38:41

to buffer against turning on

38:44

the alternate state. And

38:46

what's interesting, so then if you look at the whole population

38:48

and we go to really big databases, we find that

38:50

about 50% of obese

38:53

individuals carry this molecular

38:55

signature of an alternate developmental program.

38:59

And all of those individuals have

39:01

very high inflammatory signatures. They

39:03

have hyperinsulinemia and they have lots

39:05

of signatures that are indicative of

39:07

comorbidities or of what we call

39:09

unhealthy obesity. So what most people

39:11

think of as obesity, not a

39:13

healthy condition. What's less known is

39:15

that about a third to a

39:18

half of obese individuals are actually quite

39:20

metabolically healthy. And there's a

39:22

big debate going on about whether they're

39:24

just on the road towards unhealthy. But

39:27

if you look at the data critically

39:29

and objectively, there's quite a degree of

39:31

individuals that are quite overweight that really

39:34

have zero to no comorbidities. How

39:37

we interpret all that data put

39:39

together is that the developmental trajectory,

39:41

if it's switched on, pushes

39:43

you onto a second state that

39:45

regardless of your genetic predisposition for

39:47

obesity, you're always going to be

39:50

in a more unhealthy condition. So

39:52

it's like an epigenetically triggered

39:55

unhealthy state in the context of obesity.

40:00

question. All of this that

40:02

we've been talking about, it sort of

40:04

seems to be leading toward the question

40:06

of how might we use epigenetics in

40:08

a clinical setting? Dr. Michael C.

40:10

Young I think the easy way to conceptualize it

40:12

is precision medicine. Precision medicine has

40:15

received a lot of attention over

40:17

the last 10, 20 years, and it should

40:19

have because we're getting more genetic data

40:21

on more individuals and understanding in a

40:23

more refined way all the disease substates

40:25

that exist out there. Right

40:28

now, it may be a surprise, but

40:30

much of precision medicine is based almost

40:32

exclusively on genetic data. It doesn't

40:35

incorporate any or much environmental

40:37

data, and it doesn't incorporate

40:39

epigenetic states in it. And so

40:42

I think what the future will hold, and I don't think

40:44

it's such a distant future, is the

40:46

appreciation that, you know, let's see

40:48

this very crude figure as about a third

40:51

of who we are is genetic, about a

40:53

third is the environment we are in or

40:55

have grown up in, and then

40:57

another third is a mixture of

40:59

randomness and alternate states and these

41:02

kind of probabilistic processes. And

41:04

so what I see in 10, 20 years is that

41:07

in the clinical space, we'll have enough

41:09

tests and enough big

41:12

data appreciation to be able

41:14

to accurately place any individual

41:16

onto kind of the map

41:18

of disease risk that's informed

41:21

by their genetics, by their

41:23

environment, and by their epigenetics.

41:26

Whereas right now, it's really just that

41:28

genetic part that's fueling that. AMT

41:30

– Andrew, thank you so much for making the

41:32

time to talk with me today. I'm looking forward

41:34

to seeing what comes out of your lab next.

41:36

ANDREW – Thanks so much. Pleasure

41:39

to be here. AMT – And you

41:41

can learn more about Andrew's work at

41:43

vai.org. Our thanks to Van

41:45

Andel Institute for making this conversation possible.

41:47

And a big thanks to you for

41:50

listening. And

41:55

that concludes this edition of the Science

41:57

Podcast. If you have any comments or

41:59

suggestions, Write to us

42:01

at sciencepodcasts at aaas.org. To

42:04

find us on Podcasting Hub, search for

42:07

Science Magazine. Or you can

42:09

listen to the show on our website, science.org. This

42:13

show was edited by me, Sarah Crestie, and

42:15

Kevin McLean with production help from Megan

42:17

Tuck at Podigy. Jeffrey Cook

42:19

composed the music. On behalf

42:22

of Science and its publisher, Triple S, thanks

42:24

for joining us.