Microbial Bioelectronics: Living Wires & Biosensors - Joshua T. Atkinson

WEBVTT

00:00:00.001 --> 00:00:06.420
Oh yeah, yeah, so it turns out you could express all these pieces together and you could couple sort of ferredoxin and metabolism to

00:00:06.640 --> 00:00:12.700
production of an electrode signal as the output and so we were able to basically have electrons come from NADPH

00:00:13.180 --> 00:00:16.660
Go through our synthetic electron transport chain and then find themselves out at an electrode

00:00:17.290 --> 00:00:22.460
And what we ended up doing is using that to build a biosensor for endocrine disrupting chemicals

00:00:23.920 --> 00:00:29.800
Welcome to Entangled Health, where quantum tech, AI, health, biology and consciousness research overlap.

00:00:30.580 --> 00:00:35.980
I'm Thomas Ehmer and I like to have the conversations that never quite fit into conferences or reports.

00:00:37.080 --> 00:00:41.940
When I sit down with somebody, I usually carry one question and let that be the starting point.

00:00:42.340 --> 00:00:46.720
From there, we follow whatever actually matters to the person in the room.

00:00:48.740 --> 00:00:54.280
About 25 years ago I was in the lab in Heidelberg and putting tiny glass pipettes into single

00:00:54.440 --> 00:00:58.980
cells and listening to the faintest electrical whispers and trying to see how living cells

00:00:59.120 --> 00:00:59.760
talk to each other.

00:01:00.740 --> 00:01:03.980
Back then I thought that was already the strange edge of biology.

00:01:05.080 --> 00:01:11.320
Now my guest today, Josh Atkinson is a professor at Princeton in Civil and Environmental Engineering

00:01:11.620 --> 00:01:12.200
and in bioengineering.

00:01:13.720 --> 00:01:16.880
as a bioengineer and synthetic biologist and part of

00:01:16.900 --> 00:01:21.560
a small global community that builds living electronic devices out of microbes.

00:01:22.680 --> 00:01:27.540
He and his group design microbes that behave almost like circuit components turning chemicals

00:01:27.610 --> 00:01:30.400
in water straight into electrical signals that we can read.

00:01:31.540 --> 00:01:35.920
Today we talk about what actually happens in Josh's lab right now, how these living

00:01:36.100 --> 00:01:41.540
sensors work and then zoom out to his larger vision of smart responsive infrastructure

00:01:41.580 --> 00:01:46.780
From water systems that monitor themselves, to new ways of watching the health of our end.

00:01:46.900 --> 00:01:48.400
environment in real time.

00:01:49.540 --> 00:01:52.780
With the quick disclaimer, these are personal conversations that could create a situation

00:01:53.070 --> 00:01:57.000
between my guest and me and nothing of this is a free statement from any company or from

00:01:57.140 --> 00:01:58.620
my employer or anyone else.

00:02:01.940 --> 00:02:04.480
Yeah Josh, thanks very much and welcome to the show.

00:02:05.080 --> 00:02:05.640
Thank you for having me.

00:02:06.070 --> 00:02:11.460
I think the best is just if you start explaining what happened when you discovered the thing

00:02:11.530 --> 00:02:16.760
you discovered, so what they are and just walk us through one day in the lab.

00:02:17.500 --> 00:02:17.660
Sure.

00:02:18.250 --> 00:02:23.940
Yeah, I can give you sort of a primer on my descent into microbial electrochemistry because

00:02:23.970 --> 00:02:34.880
I didn't start

00:02:24.600 --> 00:02:29.960
there. So I guess if we go back to the beginning, I guess I went to university at the University

00:02:30.200 --> 00:02:33.060
of Michigan, I was coming out of high school, I didn't really have a scope for what research

00:02:33.260 --> 00:02:38.860
was, I was maybe interested in oncology and healthcare because of family relations to cancer.

00:02:39.880 --> 00:02:40.640
And so I got, I was

00:02:40.800 --> 00:02:41.820
basically looking for research.

00:02:41.910 --> 00:02:42.860
And the very first thing I,

00:02:43.720 --> 00:02:46.060
lab that I found myself in was Paul Dunlop's lab

00:02:46.480 --> 00:02:47.340
and he works on

00:02:49.220 --> 00:02:52.420
quorum sensing in the bacteria that live in the light organs

00:02:52.800 --> 00:02:54.580
of deep sea fishes as well as squid.

00:02:54.820 --> 00:02:55.400
So these are some

00:02:55.900 --> 00:03:00.640
sort of famous examples of how bacteria communicate with each other and you know learn how their population

00:03:01.580 --> 00:03:05.720
You know the number of cells that are in a particular location so they can do collective functions, right?

00:03:06.460 --> 00:03:16.300
So that's where I sort of got hooked on microbiology as well as sort of genetic circuitry and regulation and sort of how cells make decisions and go about

00:03:17.120 --> 00:03:19.000
interrogating their own personal environment

00:03:19.060 --> 00:03:19.620
[LAUGH]

00:03:20.560 --> 00:03:24.680
>> And so I got obsessed with metabolism shortly thereafter because synthetic biology was becoming

00:03:25.200 --> 00:03:29.480
a popular field at that moment and that seemed like a way to have societal impact was we're

00:03:29.540 --> 00:03:33.500
going to go and replace petrochemicals with things we can make from CO2 from bacterial

00:03:33.700 --> 00:03:34.140
metabolism.

00:03:35.160 --> 00:03:38.700
I went off to do my PhD at Rice University starting to work.

00:03:39.200 --> 00:03:40.320
solely in synthetic biology.

00:03:40.580 --> 00:03:43.500
So that sort of was the focus of the PhD program

00:03:43.660 --> 00:03:44.520
as well as biophysics.

00:03:45.400 --> 00:03:47.420
And that's where I started to get entangled

00:03:47.540 --> 00:03:49.840
with some electrons and starting to think of the real

00:03:50.400 --> 00:03:51.640
that's happening inside of itself.

00:03:51.760 --> 00:03:55.500
And so I went in just wanting to do pure metabolic engineering,

00:03:55.760 --> 00:03:57.740
but I found myself doing protein engineering instead.

00:03:58.500 --> 00:04:04.540
thinking about how basically you could engineer proteins to act as regulators to control metabolism

00:04:04.670 --> 00:04:09.740
in the cell and try to take all the genetic circuitry, so you know transcription and translation,

00:04:10.100 --> 00:04:14.040
gene expression, remove that and just do everything post-translationally at sort of the protein level.

00:04:15.540 --> 00:04:20.980
And so the thing I got hooked on there were these important redox proteins that deliver electrons in metabolic pathways

00:04:21.620 --> 00:04:24.120
and how these could be used as nodes to try to control metabolism.

00:04:24.910 --> 00:04:27.500
And so I started thinking about these proteins are called ferredoxinins,

00:04:27.680 --> 00:04:32.700
they're important for photosynthesis, so all of the electrons that get excited by the photons in the photosystem

00:04:32.860 --> 00:04:37.419
find their way into one of these ferredoxinins before they get dispensed out into your plant or your cyanobacterial cell.

00:04:38.200 --> 00:04:40.160
working for a bunch of different oxidative

00:04:40.400 --> 00:04:41.180
bacterial metabolisms.

00:04:42.260 --> 00:04:44.320
And so we started just taking these very simple proteins

00:04:44.520 --> 00:04:46.540
and asking the question, can you turn them into allosteric

00:04:46.580 --> 00:04:46.640
switches?

00:04:47.160 --> 00:04:49.880
So you can turn them on and off by adding a chemical input.

00:04:50.940 --> 00:04:52.340
And so we started building some of these.

00:04:52.520 --> 00:04:53.560
We got some switches to work.

00:04:54.080 --> 00:04:55.800
And then I went to a conference to start

00:04:55.920 --> 00:04:56.880
to talk about this work.

00:04:57.940 --> 00:04:58.680
And I had a poster.

00:04:59.220 --> 00:05:01.620
This was SB This is sort of one of the premier synthetic biology

00:05:01.720 --> 00:05:02.440
conferences that happened.

00:05:02.900 --> 00:05:05.460
I had a poster next to a gentleman named Moshe Baruch.

00:05:06.840 --> 00:05:11.280
Moshe was a postdoc scholar in Carolyn Ajo-Franklin's group at Lawrence Berkeley National

00:05:11.440 --> 00:05:17.580
Lab And what they were doing there was taking this redox pathway from this environmental

00:05:18.000 --> 00:05:22.120
microbe called Shewanella oneidensis and they were getting it to work inside of E. coli. So what

00:05:22.220 --> 00:05:25.760
this protein complex does is it sits in the outer membrane or on the surface of the cell.

00:05:26.520 --> 00:05:28.920
and allows electrons to come from inside the cell

00:05:28.960 --> 00:05:30.660
and move outside of the cell to minerals.

00:05:31.020 --> 00:05:32.620
That's what she went L.O. likes to breathe off of.

00:05:33.480 --> 00:05:35.200
But what they were doing was trying to take this protein

00:05:35.460 --> 00:05:37.560
complex, put it in a cell that doesn't normally do this,

00:05:37.840 --> 00:05:38.880
and then hook it up to an electrode

00:05:39.220 --> 00:05:41.560
to see if they could see electrons coming from metabolism

00:05:41.820 --> 00:05:43.620
from any E. coli cell that normally doesn't do this.

00:05:44.720 --> 00:05:48.180
I was standing right next to Moshe through this whole poster session of, "Oh hey, I have this

00:05:48.280 --> 00:05:51.940
electron transfer switch. You have this electron transfer wire that hooks up to an electron."

00:05:52.880 --> 00:05:56.320
These are obviously something we should merge together. And so we actually sort of ignored all

00:05:56.420 --> 00:05:59.460
the people who were coming to our posters and were just having like an ongoing conversation

00:05:59.700 --> 00:06:00.580
most of that poster session.

00:06:01.660 --> 00:06:06.680
That turned into a longer-term collaboration and sort of what, you know, sparked my interest

00:06:06.860 --> 00:06:09.060
in sort of the microbial electrochemical world.

00:06:09.880 --> 00:06:12.240
So, Shewanella is not the only organism that does this.

00:06:12.920 --> 00:06:16.940
It's actually incredibly common in the bacterial world to try to basically get rid of electrons

00:06:17.300 --> 00:06:18.780
by whatever means you can.

00:06:19.099 --> 00:06:22.000
So oftentimes we think about our terminal electronic

00:06:22.259 --> 00:06:24.199
sector, like maybe oxygen or nitrate.

00:06:24.379 --> 00:06:26.780
These are the very famous ones that E. coli will use

00:06:26.789 --> 00:06:28.879
and we understand regulation related to them.

00:06:29.460 --> 00:06:31.139
But they're also hooking themselves up to minerals,

00:06:31.330 --> 00:06:33.240
so rocks that they find in their environment.

00:06:33.379 --> 00:06:34.319
So oftentimes-

00:06:34.419 --> 00:06:39.240
In aquatic environments, you get tons of oxidation of iron as well as manganese and they have these

00:06:39.340 --> 00:06:43.900
manganese and iron oxides that are relatively favorable electron acceptors. And so as long as you

00:06:43.979 --> 00:06:47.639
can deal with the fact that they're insoluble and they're outside the cell and you can come up with

00:06:47.659 --> 00:06:52.180
a way to move an electron on the outside, you can use these as your terminal electron acceptor.

00:06:52.259 --> 00:06:55.759
And so Shewanella does this and other famous organisms, Geobacter sulfur reduces.

00:06:57.160 --> 00:07:03.940
But there are many, many more, both in the bacterial and the archaeal world, that are basically finding these free sinks of electrons and beginning to use them.

00:07:04.259 --> 00:07:09.120
And that's, I've now gotten deeply obsessed with all of the different ways that microbes do.

00:07:10.620 --> 00:07:16.400
Yeah, and for somebody who is not so familiar with, let's say, neither physics or biology.

00:07:17.820 --> 00:07:19.340
So how does this look?

00:07:19.700 --> 00:07:25.560
what you're building? So is it, so I recall you, you mentioned you're working with some

00:07:26.040 --> 00:07:32.440
biological wires or these things. I tried to explain this to my son, I'm not sure if I,

00:07:33.980 --> 00:07:38.540
and the question then is, but how do they look? How long are they, what are they doing?

00:07:39.780 --> 00:07:43.620
Can I, can I buy them somewhere? Yeah, yeah, these are great questions. So,

00:07:44.840 --> 00:07:49.840
As with anything in the microbial world, there's an incredible diversity in the molecules that cells used to do this.

00:07:50.720 --> 00:07:54.560
But to give you some notion of sort of ones we work with as well as some of the most famous ones,

00:07:54.740 --> 00:07:59.320
so the organism we work Shewanella oneidensis, it doesn't make long wires that are made out of proteins.

00:07:59.800 --> 00:08:04.900
does is rather is it extends part of its membrane out away from the cell and then

00:08:05.060 --> 00:08:08.199
all that membrane are basically you could think about them as pores they're sort

00:08:08.259 --> 00:08:12.419
of circles and then within that a pore there's one of these small proteins that

00:08:12.479 --> 00:08:15.780
has a few hemes in it and so electrons hop across the hemes and then they

00:08:15.840 --> 00:08:17.720
just find themselves right on the surface of the cell.

00:08:18.200 --> 00:08:21.760
And so the way Shewanella tries to reach out to a mineral is it actually blebs its membrane

00:08:21.930 --> 00:08:26.060
and extends it out. Maybe if you're familiar with like pseudopodia or something, it looks vaguely

00:08:26.400 --> 00:08:30.820
similar to that. This long sort of hair-like thread that comes away from the cells entirely

00:08:31.060 --> 00:08:32.780
made out of lipids, so it's just the cell membrane.

00:08:33.300 --> 00:08:35.440
with those proteins decorated all over it.

00:08:35.880 --> 00:08:37.839
And this is millimeter size or?

00:08:38.110 --> 00:08:39.620
So it's typically-

00:08:39.740 --> 00:08:41.500
It's like millimeters, milligrams in length.

00:08:42.080 --> 00:08:44.300
So maybe one to two microns,

00:08:44.519 --> 00:08:46.180
they're about the same size as the cells.

00:08:46.420 --> 00:08:47.000
Okay, okay.

00:08:47.279 --> 00:08:47.839
So you do it.

00:08:48.560 --> 00:08:52.980
They have to devote some portion of their surface to making these things, so they can't get too large.

00:08:53.220 --> 00:08:58.540
I was just thinking how long I made millimeter long things, which we would have to put in memory somewhere.

00:08:59.340 --> 00:08:59.460
Exactly.

00:09:00.580 --> 00:09:01.920
So that's one example for how to do it.

00:09:02.280 --> 00:09:06.080
I would say that's maybe actually relatively rare, these membrane tubules style of...

00:09:06.000 --> 00:09:10.760
building these wires. A more common set of wires are ones that are entirely out of proteins.

00:09:11.300 --> 00:09:16.620
And so this is what geobacter does. And so it creates basically a polymeric protein. So

00:09:16.960 --> 00:09:20.520
you can think about this like, you know, chains that are all linked together, and then they have

00:09:20.640 --> 00:09:26.380
sort of a heme at the interface of that chain. And these can also grow to be several microns in length.

00:09:26.900 --> 00:09:27.340
[LAUGH]

00:09:28.019 --> 00:09:33.080
But they're only, I think they're tens of nanometers in diameter. So very thin hairs that are coming

00:09:33.220 --> 00:09:38.000
off the cells that are entirely made out of protein can transport electrons across those hairs.

00:09:39.140 --> 00:09:44.640
And so geobacter makes upwards of, I think, five different cytochrome polymers at the same time

00:09:44.839 --> 00:09:47.560
on the surface of the cell. So they have different genes that get

00:09:47.600 --> 00:09:50.380
expressed, then they all make these similar wires.

00:09:51.759 --> 00:09:53.540
They're examples of convergent evolution.

00:09:53.740 --> 00:09:55.779
You get this very similar heme stacking structure,

00:09:56.000 --> 00:09:57.420
even though they didn't necessarily

00:09:57.779 --> 00:10:00.139
evolve from the same common ancestor.

00:10:02.139 --> 00:10:04.240
And I'm just coming back to this post-to-session.

00:10:06.260 --> 00:10:12.779
Did you decide to say, "Oh, let's build a new sensor, let's build some living electronics."

00:10:13.920 --> 00:10:14.980
That's basically what happened.

00:10:15.149 --> 00:10:19.180
I talked to Moshe at the poster session, we chatted for a few hours that night.

00:10:19.370 --> 00:10:22.399
I went back up into my hotel room and I sat down looking like, "Okay."

00:10:22.400 --> 00:10:24.360
The system that I have makes sulfide.

00:10:24.730 --> 00:10:27.420
The system that you make requires a quinone.

00:10:27.680 --> 00:10:30.400
So how do we get the electron between the sulfide molecule

00:10:30.740 --> 00:10:31.340
into a quinone?

00:10:31.370 --> 00:10:33.960
And so quinones are these intermembrane molecules

00:10:34.170 --> 00:10:36.840
that contain electrons, all of the electrons for sort

00:10:36.920 --> 00:10:38.600
of your electron transport chain and your mitochondria

00:10:38.720 --> 00:10:39.400
also uses a quinone.

00:10:40.000 --> 00:10:44.220
And so I started digging through the literature what enzymes actually oxidized sulfide and reduced quinones.

00:10:44.340 --> 00:10:46.960
And there's a whole suite of bacteria called sulfide oxidizing bacteria

00:10:47.520 --> 00:10:51.860
that very commonly use an enzyme called a sulfide quinone reductase to do exactly this purpose.

00:10:52.660 --> 00:10:57.840
And so I walked back into the meeting the next day and found Mochae in the crowd and was like,

00:10:57.840 --> 00:11:01.600
"Hey, Mochae, we just need to express this one enzyme with both of our pathways and this auto work."

00:11:02.100 --> 00:11:05.600
So let's schedule a conversation with our two PIs. Let's make this collaboration happen.

00:11:06.680 --> 00:11:07.940
And it went from there.

00:11:08.820 --> 00:11:10.480
We had a follow up meeting the next week.

00:11:11.100 --> 00:11:14.520
I began ordering genes for especially these SQRs,

00:11:14.580 --> 00:11:15.820
as we like to call them in E. coli,

00:11:16.040 --> 00:11:17.300
just to see if they would function at all.

00:11:18.460 --> 00:11:21.560
Also decided Caroline's group was-

00:11:21.640 --> 00:11:26.060
The only group really doing a heterologous expression of this pathway in E. coli, so I was like,

00:11:26.120 --> 00:11:29.440
"I need to go work with these people to learn how to do this because it's a non-trivial pathway

00:11:29.720 --> 00:11:34.960
to get expressed inside the cells." So I started writing some funding applications for a graduate

00:11:35.140 --> 00:11:39.200
fellowship. I was a graduate student at the time, and so the Department of Energy has very nice

00:11:39.400 --> 00:11:44.520
graduate fellowships or funded labs from the DOE to go work in the national lab system.

00:11:44.710 --> 00:11:48.980
And so the National Labs are, I think, probably, you know, America's greatest contribution to

00:11:49.040 --> 00:11:54.780
science as far as their scope and what they do. They think, you know, very big at the scale that

00:11:54.940 --> 00:11:59.260
you can't think as an individual PI. And so it was a really great opportunity to go and learn how

00:11:59.260 --> 00:12:03.040
the National Labs work and go work with, you know, the best person doing that type of work.

00:12:03.160 --> 00:12:03.400
Yeah.

00:12:04.480 --> 00:12:08.560
And then you became proof of, so if something needs to have happened because you became

00:12:08.700 --> 00:12:08.820
Professor.

00:12:09.440 --> 00:12:13.780
Oh yeah, yeah, so it turns out you could express all these pieces together and you could couple

00:12:14.060 --> 00:12:18.180
sort of ferredoxin and metabolism to production of an electrode as the output.

00:12:18.680 --> 00:12:23.600
And so we were able to basically have electrons come from NADPH go through our synthetic electron

00:12:23.840 --> 00:12:24.100
transport chain.

00:12:24.000 --> 00:12:25.940
and then find themselves out at an electrode.

00:12:26.610 --> 00:12:29.680
And what we ended up doing is using that to build a biosensor

00:12:30.000 --> 00:12:31.580
for endocrine disrupting chemicals.

00:12:31.910 --> 00:12:35.260
And so I had built a bunch of these sensor ferredoxins

00:12:35.310 --> 00:12:37.840
that would respond to endocrine disrupting molecules.

00:12:38.060 --> 00:12:39.360
So things like bisphenol A,

00:12:39.520 --> 00:12:41.360
as well as a bunch of breast cancer drugs

00:12:41.470 --> 00:12:43.000
and other sort of cancer treatments as well.

00:12:43.500 --> 00:12:48.080
And we use that as our model test system to show that you could put in an engineered protein

00:12:48.240 --> 00:12:50.800
to actuate electron transfer from the cell to the electrode.

00:12:51.780 --> 00:12:53.240
And this happens at really fast time scales.

00:12:53.640 --> 00:12:59.320
So typically a biosensor that uses a gene circuit takes maybe a half hour at the fastest,

00:12:59.900 --> 00:13:02.040
usually typically hours to really get your sensor response.

00:13:02.240 --> 00:13:04.600
And so this we're able to see our molecules within minutes.

00:13:05.500 --> 00:13:08.980
So it takes diffusion into the cell and then it can generate enough current that we can

00:13:09.040 --> 00:13:09.840
detect it at the electrode.

00:13:10.540 --> 00:13:10.780
Okay.

00:13:11.860 --> 00:13:17.860
So it's fast, it's not as fast as maybe like neuronal signaling that's maybe pushing it a little faster than what we can do, but maybe we'll get there.

00:13:18.079 --> 00:13:18.139
Yeah.

00:13:18.720 --> 00:13:26.740
I mean, it also depends on the type of signal because sometimes you just don't want to have the quick responses and rather, I mean, if you, I recall somewhere.

00:13:27.660 --> 00:13:31.960
I guess in the talk, I listened to, you mentioned wastewater things.

00:13:32.620 --> 00:13:35.140
These things don't change within milliseconds.

00:13:36.160 --> 00:13:40.440
Yeah, you don't necessarily need the fastest timescale, but faster than hours is 10 to

00:13:40.540 --> 00:13:40.720
10.

00:13:40.880 --> 00:13:40.980
Yeah.

00:13:42.240 --> 00:13:43.420
And just help me out.

00:13:44.380 --> 00:13:44.580
The

00:13:46.100 --> 00:13:52.040
the mechanism behind this is then we quickly touched the topic of quantum

00:13:52.240 --> 00:13:56.100
electrochemistry, electron density distribution. Can you walk me a bit

00:13:56.200 --> 00:14:01.740
through what the hypothesis or the published state is or what we are going

00:14:02.040 --> 00:14:07.200
to do? Yeah, if we look broadly at electron transfer extracellular and

00:14:07.280 --> 00:14:12.080
bacteria, the primary mechanism by which this occurs is through redox hopping

00:14:12.160 --> 00:14:15.360
through cytochromes. And so what you end up having there is

00:14:15.400 --> 00:14:20.560
you have multiple heme molecules that are decorated in these proteins, an electron hops

00:14:20.760 --> 00:14:24.200
from one heme molecule to the next heme molecule, and it just continues to hop up the chain.

00:14:24.660 --> 00:14:28.640
And so this ends up looking like a thermally activated process. It looks a lot like redox

00:14:28.820 --> 00:14:32.080
polymers also look the same way. So, you know, as we increase the

00:14:32.000 --> 00:14:33.820
the temperature, this conduction increases.

00:14:34.430 --> 00:14:36.560
And so we can measure this in sort of our Shewanella

00:14:36.700 --> 00:14:37.260
that we work with.

00:14:37.850 --> 00:14:40.800
We can deposit them onto an electrode in sort of an ordered

00:14:41.000 --> 00:14:41.120
way.

00:14:41.280 --> 00:14:42.360
We can begin to vary the temperature,

00:14:42.650 --> 00:14:44.660
measure the conductance at each of those different temperatures,

00:14:45.180 --> 00:14:46.620
derive out conductivities from this.

00:14:46.670 --> 00:14:48.340
This is a lot of what we spend our time doing,

00:14:48.450 --> 00:14:50.600
is to try to get these material properties measured

00:14:50.960 --> 00:14:53.000
for these living biofilms and living materials

00:14:53.250 --> 00:14:53.740
that we're making.

00:14:55.260 --> 00:15:00.500
With the idea that we can use this to build more complex things than just a resistor, which is kind of what we currently are.

00:15:01.040 --> 00:15:10.600
And I'm obsessed, not obsessed, but this is my private show. So I'm obsessed with quantum effects in biology.

00:15:12.600 --> 00:15:17.280
And I mean, there's a lot of things. I mean, you mentioned photosynthesis. Nobody really understands how this works.

00:15:18.240 --> 00:15:22.420
We understand but we can't re-engineer maybe and I was just thinking when you said

00:15:23.480 --> 00:15:26.660
building the bridge between bio and electronics.

00:15:29.910 --> 00:15:33.440
There are a lot of things, variables, what have you,

00:15:33.500 --> 00:15:35.080
but there are also much more interesting things,

00:15:35.300 --> 00:15:36.880
which is sensing quantum properties.

00:15:38.700 --> 00:15:40.080
Have you thought about this?

00:15:40.970 --> 00:15:42.340
Just to give you one thing before you answer,

00:15:42.960 --> 00:15:44.760
I had Keeper Sharkey, for example,

00:15:44.980 --> 00:15:48.140
who detected that mentioned in the talk,

00:15:48.660 --> 00:15:50.900
that, for example, one isotope of lithium

00:15:52.440 --> 00:15:57.360
is active for in the brain and the other not.

00:15:58.120 --> 00:16:00.480
Which is a spin effect, which is a nuclear spin effect.

00:16:01.580 --> 00:16:05.220
Do you, have you absorbed or have your plans to extend?

00:16:06.080 --> 00:16:08.200
Yeah, so I haven't myself worked too much

00:16:08.230 --> 00:16:10.660
on sort of using these to measure quantum effects, right?

00:16:11.240 --> 00:16:12.699
but some colleagues of mine have.

00:16:12.819 --> 00:16:15.319
And so I did have my postdoc in Moel Nijars lab

00:16:15.360 --> 00:16:16.699
at the University of Southern California.

00:16:17.560 --> 00:16:21.100
I was like a physicist by training, right?

00:16:21.540 --> 00:16:24.440
And so he does a lot of solid state physics background,

00:16:24.600 --> 00:16:25.680
but he does it on biology nowadays.

00:16:26.420 --> 00:16:26.940
And so anyways,

00:16:28.040 --> 00:16:31.900
They're very interested in this property called chiral-induced spin cell activity.

00:16:33.000 --> 00:16:35.800
It's a property that's generic to organic molecules in general,

00:16:36.100 --> 00:16:42.300
along the sort of, they have chiral centers in them. They get bias, the spin of the electrons

00:16:42.520 --> 00:16:47.540
that are flowing through them. And so they've done some nice work looking at these same multiheme

00:16:47.620 --> 00:16:52.300
side-acromes that we work on from the perspective of them effectively being spin filters. And so,

00:16:52.700 --> 00:16:58.160
Basically, you have a bias towards spin-up electrons flowing through these cytochrome

00:16:58.280 --> 00:16:58.400
wires.

00:16:58.730 --> 00:17:01.080
And so they've done some really nice measurements on purified proteins.

00:17:02.370 --> 00:17:06.860
Basically, they'll use a whole devices to do these measurements in an AFM.

00:17:08.360 --> 00:17:13.780
They've now done some very nice work in living biofilms, showing that if you expose them

00:17:13.790 --> 00:17:18.020
to a magnetic field, depending on the orientation of that field, you have increases in the rates

00:17:18.180 --> 00:17:21.020
of electron transfer coming out of the biofilms themselves onto the electrodes.

00:17:22.600 --> 00:17:31.340
And so this is sort of indicating that, you know, this might be something that's real from the biology perspective of influencing how efficiencies processes occur.

00:17:32.140 --> 00:17:36.580
You know, maybe they're not being exposed to the same type of magnetic fields that we're exposing them to in the lab.

00:17:36.740 --> 00:17:38.780
They're much higher than you would find sort of out in nature.

00:17:39.539 --> 00:17:43.720
But what you might end up finding is minerals that have sort of some bias towards the spins that they accept.

00:17:44.600 --> 00:17:46.820
while thinking about the directionality of how electrons

00:17:47.200 --> 00:17:51.820
blow through these wires. If you can basically build up some spin resistances, you can sort of

00:17:51.960 --> 00:17:56.500
biases toward moving electrons out of the cell versus taking electrons into the cell. So what I

00:17:56.500 --> 00:18:00.380
did mention is this property is bidirectional. And so some cells use it to dispense electrons,

00:18:00.740 --> 00:18:08.679
other cells use it to take electrons. Because before we started, we mentioned the connection to

00:18:08.720 --> 00:18:11.899
life. The other thing is...

00:18:12.320 --> 00:18:14.080
Clarice Aiello, I'm not sure if you know of her.

00:18:14.860 --> 00:18:16.760
Clarice Aiello, she was at the UCLA,

00:18:17.230 --> 00:18:19.800
and now she's doing, she's my go-to person

00:18:19.940 --> 00:18:20.940
for quantum biotopics.

00:18:22.800 --> 00:18:24.400
In the episode I had with her,

00:18:24.500 --> 00:18:26.200
she was of course mentioning the birds navigation,

00:18:26.760 --> 00:18:28.080
birds eye navigation, so this.

00:18:28.220 --> 00:18:29.620
Yeah, right, yep, that's a sort of one.

00:18:30.100 --> 00:18:33.220
And I was really wondering because it was the horse.

00:18:34.820 --> 00:18:42.880
for making noise. Do we have a magnetic sense and these things? And regarding the

00:18:44.720 --> 00:18:50.660
importance of magnetic components compared to the electric components, are you

00:18:51.000 --> 00:18:54.960
Do you need these or can you work with electric circuits?

00:18:55.620 --> 00:18:57.060
I don't need for magnetic.

00:18:57.940 --> 00:18:58.980
They're one in the same, right?

00:18:59.440 --> 00:18:59.920
It's electromagnetic.

00:19:00.960 --> 00:19:01.680
So they're always sort of...

00:19:01.680 --> 00:19:02.540
Hands on frequency, yes.

00:19:03.180 --> 00:19:06.460
So I say they are a unified subject here, right?

00:19:06.500 --> 00:19:12.900
But we've become more interested in sort of magnetic fields and how you can use these from a bioengineering perspective.

00:19:13.140 --> 00:19:22.940
And so there are other folks who have been working in this space quite heavily of looking at basically some of the same type of redox active proteins that bind flavon molecules.

00:19:23.700 --> 00:19:28.320
The difference being flavons can be doubly reduced so you can have two electron reductions rather than one electron reduction.

00:19:29.300 --> 00:19:30.500
They have a...

00:19:31.580 --> 00:19:39.220
basically magnetic field sensitivities that can be used for actuating some phenotypes in different proteins.

00:19:39.400 --> 00:19:44.980
So some people have used this for looking at basically increasing magnitudes of fluorescence in the presence of magnetic fields.

00:19:45.600 --> 00:19:48.480
We've been interested in starting to think about them as basically little action.

00:19:48.500 --> 00:20:03.600
to actuate

00:20:03.679 --> 00:20:08.940
Go on. So primarily to govern turnover of enzymes is really what we're thinking. But

00:20:09.600 --> 00:20:14.140
it as a platform rather than any one particular application. So basically, can you use one of

00:20:14.220 --> 00:20:18.760
these domains just as a magnetic field sensor so that we can, you know, within a bioreactor,

00:20:19.020 --> 00:20:25.419
maybe turn on or off one of these metal pathways just by applying a magnetic field as sort of a

00:20:25.440 --> 00:20:29.679
simple stimulus. You could think about this like optogenetics, but using magneto-tintics more

00:20:29.679 --> 00:20:31.780
Yeah, is somebody else working on that or just

00:20:32.620 --> 00:20:34.420
tons of people who are currently working in this space.

00:20:34.720 --> 00:20:34.780
Yeah.

00:20:35.250 --> 00:20:37.320
You know, it gives me my ignorance.

00:20:37.430 --> 00:20:39.820
But there's been some very nice papers

00:20:40.040 --> 00:20:41.940
who've kind of come out in the last three to four years

00:20:42.260 --> 00:20:44.480
where people have sort of started to look at these proteins

00:20:44.770 --> 00:20:47.280
and thinking about this-- typically, this domain

00:20:47.330 --> 00:20:48.840
that I'm talking about, the love domain,

00:20:49.100 --> 00:20:50.420
has been used as a blue light,

00:20:50.800 --> 00:20:52.060
activatable domain in optogenetics.

00:20:53.080 --> 00:20:55.200
You can make a point mutation that makes it no longer really

00:20:55.760 --> 00:20:57.300
reply to light in the same way that it does,

00:20:57.370 --> 00:20:59.140
and now it sort of replies to the magnetic fields.

00:21:00.900 --> 00:21:03.900
So it has this sort of, it's even in the name of the domain, right?

00:21:04.540 --> 00:21:08.720
It's a light oxygen voltage sensitive domain and turns out it's also slightly sensitive

00:21:08.770 --> 00:21:09.740
to magnetic fields, right?

00:21:13.260 --> 00:21:14.820
And in your work?

00:21:16.480 --> 00:21:16.580
Oh, okay.

00:21:17.020 --> 00:21:17.460
Actual work.

00:21:17.650 --> 00:21:22.100
Do you need the classical, the quantum description

00:21:22.600 --> 00:21:24.820
or can you work with a classical description?

00:21:24.980 --> 00:21:26.540
Because I'm just thinking building electronics

00:21:27.070 --> 00:21:29.520
and then if we say, let's build electronics

00:21:29.650 --> 00:21:32.680
or circuits with classical things to be easier

00:21:32.840 --> 00:21:36.120
then using quantum, I mean.

00:21:38.300 --> 00:21:40.020
Even chips work with quantum, of course.

00:21:40.480 --> 00:21:44.760
Yeah, so a lot of our work happens at sort of a higher level

00:21:44.900 --> 00:21:46.600
than individual molecules themselves.

00:21:46.760 --> 00:21:49.620
And so classical descriptions are sufficient for a lot

00:21:49.660 --> 00:21:51.900
of what we're trying to measure from sort of like

00:21:51.900 --> 00:21:56.200
whole cell biofilm populations of cells and transport through those biofilms.

00:21:56.960 --> 00:22:02.500
As you push into the actual molecules themselves and the direct interfaces that they have with

00:22:02.540 --> 00:22:05.120
the electrodes, some quantum descriptions are actually quite useful.

00:22:05.540 --> 00:22:10.220
I'm trying to describe sort of how those couplings work, but it's not something that

00:22:10.260 --> 00:22:13.020
we've had to use in sort of our day in and day out of trying to build a biotechnology

00:22:13.560 --> 00:22:14.400
using some of these features.

00:22:16.920 --> 00:22:17.120
Okay.

00:22:23.060 --> 00:22:27.100
And let me come to this sensing part.

00:22:29.140 --> 00:22:35.640
Or engineer microbes, which detect whatever, something,

00:22:35.730 --> 00:22:37.500
a signal, a component of water.

00:22:38.340 --> 00:22:44.840
Typically, and then you have an electrical signal, which we can kind of read in kind of read time or whatever.

00:22:45.360 --> 00:22:45.440
Yep.

00:22:50.740 --> 00:22:59.500
Is this, I'm going a bit in the, not in the area of speculation, is this something you

00:22:59.720 --> 00:23:01.840
think makes them aware?

00:23:04.120 --> 00:23:07.799
Yeah, yeah, so maybe thinking more of a consciousness type thing.

00:23:07.860 --> 00:23:13.720
Yeah, yeah, I'm just thinking because people are starting to talk about self-replic, there

00:23:13.720 --> 00:23:18.059
was just news about these parts of the whole, how they were called and then you you build.

00:23:18.700 --> 00:23:22.120
you build artificial cells that replicate themselves,

00:23:22.310 --> 00:23:23.340
I think you might have read it.

00:23:23.400 --> 00:23:25.220
And now comes the other thing and talks about

00:23:26.020 --> 00:23:27.820
cognitions, are we generating God or whatever?

00:23:28.300 --> 00:23:32.440
Just boiling it down to what really works or not.

00:23:33.200 --> 00:23:36.740
Right, so the cells are deeply aware

00:23:37.760 --> 00:23:40.500
from a biochemical perspective of the electrode.

00:23:41.980 --> 00:23:46.580
So the electrode is serving as the main thing they are breathing off of.

00:23:46.660 --> 00:23:50.380
And so the entire rate of their metabolism is dictated by the potential we apply at that

00:23:50.480 --> 00:23:50.620
electric.

00:23:51.200 --> 00:23:56.940
If we want to talk about awareness from the responsiveness of that, we're certainly aware

00:23:57.220 --> 00:23:57.440
of that.

00:23:57.980 --> 00:24:00.679
Is that consciousness, that's maybe a different philosophical?

00:24:01.000 --> 00:24:01.320
It's a awful discussion.

00:24:01.840 --> 00:24:02.640
Yeah, yeah, yeah.

00:24:03.520 --> 00:24:04.760
But they are responding to it.

00:24:05.320 --> 00:24:09.200
And we have very nice videos of basically looking at the

00:24:09.720 --> 00:24:12.860
voltages they generated, their membrane in response to these

00:24:13.760 --> 00:24:14.920
voltages we applied at electrode.

00:24:15.540 --> 00:24:17.980
And you get these very strong biochemical responses that

00:24:17.980 --> 00:24:18.320
are happening.

00:24:20.200 --> 00:24:25.180
they don't persist beyond like if you shift the electrode away. This signal disappears. It has

00:24:25.260 --> 00:24:29.780
some decay time that maybe is something that lingers a little bit about there like yes, you

00:24:29.900 --> 00:24:35.220
know here. The other thing that Chiuenella in particular likes to do is it likes to

00:24:35.600 --> 00:24:38.360
try to find a terminal electronic receptor to breathe off of.

00:24:38.940 --> 00:24:43.760
And so it has to make decisions again from a biochemical network perspective on where it

00:24:43.900 --> 00:24:46.960
should physically be. And so these are motile organisms they can swim around.

00:24:48.020 --> 00:24:51.740
There's a property called electrotaxis. And if you've been investigating with these organisms,

00:24:51.830 --> 00:24:55.440
we basically can. They sense the electric signature from one of these electrodes.

00:24:56.039 --> 00:24:59.039
Is this a field? Is this a concentration gradient of a redox molecule?

00:24:59.700 --> 00:25:01.919
That's sort of an area of mechanism perspective.

00:25:02.480 --> 00:25:03.580
Could it be magnetic? No.

00:25:04.800 --> 00:25:06.940
Maybe they have magnetic field sensors.

00:25:07.200 --> 00:25:07.400
Okay.

00:25:07.820 --> 00:25:09.260
Yeah, if you have an AC current,

00:25:09.500 --> 00:25:09.820
yeah sure.

00:25:09.960 --> 00:25:10.880
AC current could be, yeah.

00:25:12.480 --> 00:25:15.480
But it's, you know, my gut says this is likely

00:25:15.720 --> 00:25:18.380
a concentration gradient of reduced or oxidized molecules

00:25:18.580 --> 00:25:20.060
that they are introducing up in town

00:25:20.140 --> 00:25:21.980
that just like all the other chemotaxis

00:25:22.100 --> 00:25:22.980
is typically work.

00:25:24.460 --> 00:25:28.659
Now, something that's different than a colleague of mine here at Princeton works on is electrotaxis

00:25:28.940 --> 00:25:30.580
in mammalian cells.

00:25:31.500 --> 00:25:33.960
So Daniel Cohen is my colleague's name.

00:25:34.120 --> 00:25:38.820
And so he's been looking at basically the response of human cells to electrical fields.

00:25:38.960 --> 00:25:42.500
And so he'll look at kidney cells, he'll look at white blood cells, all sorts of different

00:25:42.580 --> 00:25:42.679
types.

00:25:43.360 --> 00:25:47.760
And there's a protein that's in the membrane of the cells that's now has a fun named called

00:25:47.760 --> 00:25:51.000
Galvanon, the basically in response to an electrical

00:25:51.279 --> 00:25:52.120
sealed, it, you know.

00:25:52.580 --> 00:25:57.280
migrates on the cell surface to one pole or the other pole depending on the charge and the field that's being applied

00:25:57.900 --> 00:25:59.720
And then it causes the cell to drive that way

00:26:00.660 --> 00:26:01.660
So this is actually how

00:26:02.840 --> 00:26:05.960
These cells are traversing wounds in the human body

00:26:06.080 --> 00:26:10.740
So you get a small electrical field that gets generated when you have a wound just from sort of buildup of ion

00:26:10.920 --> 00:26:18.600
Concentration and then cells specifically immune cells can actually traverse fields of that strength and find themselves where they need to be

00:26:18.520 --> 00:26:19.700
Um...

00:26:20.100 --> 00:26:27.679
which opens the door to electro-seuticals in a very straightforward thing to just for whom to care.

00:26:27.860 --> 00:26:32.700
Exactly. Yeah, yeah. So he has some projects that he's thinking about sort of, you know, electronic bandages, him and others.

00:26:32.840 --> 00:26:34.980
I mean, there are other great folks in that space as well.

00:26:38.260 --> 00:26:41.179
I just mentioned human things.

00:26:42.330 --> 00:26:44.720
You know, I was in the lab 25 years ago, so I did my PhD.

00:26:46.120 --> 00:26:49.500
During PhD, you are sometimes in the lab at night.

00:26:50.779 --> 00:26:54.220
You get lost completely in the topic.

00:26:55.159 --> 00:26:57.399
I'm not sure maybe you entangled with your experiment.

00:27:00.590 --> 00:27:07.179
I recall one thing when I was in the lab.

00:27:02.419 --> 00:27:06.299
when you spend too much in the lab, you get strange ideas.

00:27:07.100 --> 00:27:11.139
And we have been working with single cells from interoscee,

00:27:11.190 --> 00:27:14.299
from rats, so we were taking the feed from rats

00:27:14.539 --> 00:27:20.179
and then taking the small muscles, I think it's really my micrometer long.

00:27:21.220 --> 00:27:24.179
And when you look at these, a colleague of mine had a look at it

00:27:24.179 --> 00:27:29.159
and they were shimmering, like, I'm not sure, I don't know the word for...

00:27:31.200 --> 00:27:34.600
Shinken, so if you slice muscles and then you look at it, you see a shimmer.

00:27:35.280 --> 00:27:41.300
And he said, it also fuses and then he said, "Wow, maybe the seat of the soul is in the

00:27:42.160 --> 00:27:42.600
muscle."

00:27:43.190 --> 00:27:47.460
And this still sits in a very deep sense.

00:27:47.890 --> 00:27:49.840
But the...

00:27:50.100 --> 00:27:58.120
So we didn't put this in a paper, but when you said so these things sense somehow or

00:27:58.360 --> 00:28:06.100
the cells sense something and they have the tendency to move to a longer gradient, I mean

00:28:06.160 --> 00:28:14.600
it's a very small gradient. It's very vulnerable because if it doesn't

00:28:15.539 --> 00:28:17.120
go there, it will die, etc.

00:28:17.679 --> 00:28:22.019
And this, for me, now looking at all the AI topics,

00:28:22.340 --> 00:28:27.679
so we're jumping, and people claim they are intelligent, etc.

00:28:27.820 --> 00:28:28.600
But they don't have a body.

00:28:29.279 --> 00:28:30.940
So what I want to go is, do you think,

00:28:31.279 --> 00:28:33.820
regarding coming back to this poor two consciousness or whatever,

00:28:34.400 --> 00:28:38.380
And if it's too woo, we can cut short.

00:28:38.470 --> 00:28:40.900
But not the idea.

00:28:41.030 --> 00:28:44.960
I was just thinking, do you think with these behavior

00:28:45.130 --> 00:28:49.900
you observe and building electronics, living electronics?

00:28:51.520 --> 00:28:55.540
We have found a way to say these things work by itself,

00:28:55.880 --> 00:28:57.200
and they are embodied.

00:28:57.960 --> 00:28:59.620
And by this, they might have a consciousness.

00:29:01.600 --> 00:29:05.400
compared to the other guys that just are digital signals and don't have a body.

00:29:06.280 --> 00:29:08.380
Do you think there's a fundamental difference or is it?

00:29:08.780 --> 00:29:09.320
I'm just asking.

00:29:09.760 --> 00:29:14.980
Yeah, so this is sort of a philosophical question of basically, can you be conscious

00:29:15.280 --> 00:29:18.960
or an organ, or something with a soul if you are in absence of a body and then if you

00:29:19.060 --> 00:29:19.280
were to.

00:29:19.400 --> 00:29:20.800
If you cannot die, it's a bad luck.

00:29:22.680 --> 00:29:24.680
Yeah, it's a very interesting question, right?

00:29:26.240 --> 00:29:29.720
So these bacteria, they certainly have, they can die, right?

00:29:32.710 --> 00:29:41.260
They can't, on their own, aside from deciding whether or not they are there or whether or not their metabolism is active, decide much more than that, right?

00:29:41.430 --> 00:29:47.640
Like they don't have to roll over the system, right? They're sort of being dictated to by the electrode that they happen to have found themselves on.

00:29:47.690 --> 00:29:50.340
And so, you're like with a single neuron, I mean, no.

00:29:50.760 --> 00:29:56.760
Yeah, so I guess they don't have like, you know, this is maybe a different philosophical question, but control over their own fate, right?

00:29:58.100 --> 00:30:02.320
Their fate is being driven by sort of the way that we set them up in our experiments.

00:30:03.480 --> 00:30:08.980
And so, you know, maybe if they were able to freely communicate with us, they could tell us, you know, leave that electorate on.

00:30:09.140 --> 00:30:20.480
Please stop changing the role. Please stop changing the role.

00:30:14.140 --> 00:30:18.420
Yeah, which we do see in their their metabolism, but they're relatively robust so that you know you can have this around

00:30:19.720 --> 00:30:22.920
Yeah, and if you scale these things, I mean

00:30:24.520 --> 00:30:29.780
Yeah, so scaling is one of the challenges in these systems in general and so

00:30:30.360 --> 00:30:35.320
Like all of the early work on engineering these electrowactive organisms was done at sort of bulk scale.

00:30:36.020 --> 00:30:40.500
And so you would be, you know, maybe making a modification in a genome, you'd have a whole strain in a reactor,

00:30:40.820 --> 00:30:45.360
you'd have one electrode, those cells would, you know, maybe they can form a biofilm, they'll come and attach to it on their own.

00:30:45.560 --> 00:30:48.400
But you're just sort of letting the biology do what biology does, right?

00:30:48.670 --> 00:30:52.400
And so it's whatever is already pre-encoded, it's genome, it's going to respond in that particular way.

00:30:53.180 --> 00:30:55.580
That's challenging if you want to go and try to build devices out of these things

00:30:55.730 --> 00:30:59.960
And so that's sort of where a lot of synthetic biology work has been focused to date is how do you

00:31:00.520 --> 00:31:03.820
Construct devices from these and so you could you can do this from you know

00:31:03.980 --> 00:31:05.560
Just more of a materials engineering perspective

00:31:05.670 --> 00:31:09.480
We can build hydrogels where we trap the cells to the other way

00:31:22.420 --> 00:31:22.780
Yeah

00:31:13.080 --> 00:31:14.940
make a porous electrode, load it with cells,

00:31:15.559 --> 00:31:17.639
put some skin on it and the cells will stay there.

00:31:18.419 --> 00:31:20.399
We've been taking a slightly different approach more recently

00:31:20.600 --> 00:31:22.880
which is to engineer the biology to have programs

00:31:23.130 --> 00:31:24.940
that allow us to have itself assemble

00:31:25.139 --> 00:31:26.000
where we want it to assemble.

00:31:26.190 --> 00:31:28.700
And so we can basically use light to guide

00:31:28.980 --> 00:31:30.120
the formation of biofilm.

00:31:30.340 --> 00:31:32.600
So we have sort of a lithographic system for shining light

00:31:32.649 --> 00:31:33.940
and getting the biofilm to grow

00:31:34.080 --> 00:31:35.100
in the way that we want it to grow.

00:31:35.710 --> 00:31:37.399
It helps make that, you know,

00:31:37.460 --> 00:31:39.200
manufacturing process slightly simpler

00:31:39.679 --> 00:31:40.760
because now our cells are sort of,

00:31:41.020 --> 00:31:46.920
predisposed to doing what we want. This is also the backbone of the entire silicon industry

00:31:47.060 --> 00:31:52.640
and all of our traditional electronics is the ability to do lithography and make our chips smaller

00:31:52.840 --> 00:31:55.840
and smaller right now. We are limited in so far as like

00:31:56.200 --> 00:31:58.260
we have a cell that's approximately three microns long

00:31:58.400 --> 00:31:59.299
by one micron wide.

00:31:59.940 --> 00:32:01.539
And so we can't get any smaller than that.

00:32:02.480 --> 00:32:04.559
That's a fundamental limitation of these types of things.

00:32:05.179 --> 00:32:07.200
But then to start to scale, if you want to build devices

00:32:07.440 --> 00:32:09.400
that you go and deploy in a bunch of different environments,

00:32:09.760 --> 00:32:11.740
you can do this in parallel and just build a bunch of small

00:32:12.700 --> 00:32:12.980
devices.

00:32:13.980 --> 00:32:15.620
That's perfectly feasible with the technology

00:32:15.880 --> 00:32:16.860
we have off the shelf right now.

00:32:17.520 --> 00:32:20.440
But something that a lot of folks have been spending their time

00:32:20.539 --> 00:32:23.780
thinking about in this field is how do you use-

00:32:24.640 --> 00:32:28.600
The biology as a chemical catalyst and the electrode as a source of electrons to make

00:32:28.760 --> 00:32:32.520
molecules right and how can you do this at rates and scales that are actually feasible

00:32:32.800 --> 00:32:37.120
to compete with traditional chemical synthesis processes and this is where scale really is

00:32:37.120 --> 00:32:40.760
a killer for these types of technologies and we don't have a solution to this problem currently.

00:32:41.400 --> 00:32:45.980
So the thickness of the biofilm that you would need to reach the rates that are competitive

00:32:46.190 --> 00:32:49.960
with a lot of chemical processes, depending on obviously whatever chemistry you're precisely

00:32:50.220 --> 00:32:56.560
trying to do, is one of the limiting factors for making this actually a tractable technology

00:32:56.750 --> 00:32:57.820
that can be economically feasible.

00:32:57.980 --> 00:32:58.720
Currently, it's not there.

00:32:59.570 --> 00:33:01.520
And so one way is porous electrodes.

00:33:01.920 --> 00:33:03.660
It's one strategy people have been going after here.

00:33:03.820 --> 00:33:07.040
Other ways are trying to enhance biofilm formation to make them thicker.

00:33:08.180 --> 00:33:10.420
if you can pump up the charge transport.

00:33:11.000 --> 00:33:14.220
pathways that you have from the electrode into the cell, use multiple in parallel.

00:33:14.920 --> 00:33:21.600
One very useful way of building this electrical to biochemical interfaces to just use an electrolyzer

00:33:21.600 --> 00:33:26.320
to make a molecule like hydrogen or formate, which are very good electrolyzers these days,

00:33:26.580 --> 00:33:31.480
make these molecules and then scale up and just sort of traditional stirred batch bioreactors

00:33:31.510 --> 00:33:34.680
that we've been using for several decades.

00:33:34.800 --> 00:33:35.240
and.

00:33:38.980 --> 00:33:44.020
The biofilms you mentioned, are they forming themselves?

00:33:44.180 --> 00:33:47.980
I mean, biofilms, once you have a biofilm, it's quite tough to get rid of it.

00:33:49.900 --> 00:33:53.320
So are they forming naturally or would you have...

00:33:53.700 --> 00:33:55.220
to build them or

00:33:55.919 --> 00:33:59.220
Yeah, no, no, so that's a good question. So Shewanella will naturally form biofilms

00:33:59.360 --> 00:34:04.820
They're relatively weak biofilms, so they're actually easier to get off than some biofilms like a pseudomonas aruginosa

00:34:05.039 --> 00:34:08.899
It's a very strong biofilm. They have tons of extracellular matrix that

00:34:09.000 --> 00:34:13.580
keeps them adhered. She would not like use this one large protein to do this and some polysaccharides

00:34:14.000 --> 00:34:18.340
and DNA on the outside of the cell. On their own, they don't form the most robust biofilms,

00:34:18.639 --> 00:34:23.820
and they just form them wherever the cells land and begin to divide and begin to populate.

00:34:24.600 --> 00:34:26.419
And so what we've been doing is,

00:34:26.440 --> 00:34:26.860
side.

00:34:27.760 --> 00:34:31.060
basically putting in a genetic circuit to express particular

00:34:32.160 --> 00:34:37.280
extracellular matrix proteins, and then using light to control the expression of those proteins,

00:34:37.920 --> 00:34:40.660
which then causes the cells to self-generate a biofilm,

00:34:40.770 --> 00:34:43.340
but it's using proteins they don't normally use to make biofilms.

00:34:44.200 --> 00:34:47.640
And so they make a slightly more robust biofilm than the native Schuenella one.

00:34:48.960 --> 00:34:54.099
Albeit, it's still weaker than some of your natural multicellular biofilm or multi species biofilm.

00:34:55.070 --> 00:34:58.300
Do we understand this if you want to build an electric circuit or whatever you would,

00:34:59.050 --> 00:35:02.060
then do also like they do in classical chip production.

00:35:02.420 --> 00:35:05.320
So you say, I have my whatever.

00:35:09.599 --> 00:35:12.760
surface and then I put a mask on it and then I activated

00:35:13.160 --> 00:35:16.500
wear light and then there where the shadow is, there is nothing

00:35:16.740 --> 00:35:18.980
happening where the light is coming, they express this bio film

00:35:19.119 --> 00:35:22.000
so I can build really is this, are you doing it?

00:35:22.160 --> 00:35:22.400
Press yes.

00:35:22.960 --> 00:35:24.000
>> That's exactly how we're doing it.

00:35:24.400 --> 00:35:25.820
We're using the same sort of photo mask.

00:35:26.500 --> 00:35:28.480
Our stuff looks a little bit more simplistic.

00:35:29.089 --> 00:35:29.180
>> Yeah.

00:35:29.570 --> 00:35:30.920
>> And of three, none of me does like this.

00:35:31.490 --> 00:35:33.619
>> Yeah. No, we make,

00:35:34.200 --> 00:35:36.579
we go to the nano fab to make some of

00:35:36.589 --> 00:35:38.619
the electrode arrays that we use to do these measurements.

00:35:38.980 --> 00:35:42.520
So we use those types of lithography processes.

00:35:43.099 --> 00:35:44.680
So ours is just layered on top of that.

00:35:44.849 --> 00:35:45.720
>> Then, did.

00:35:49.060 --> 00:35:53.540
Are these acting things as one big circuit or are they showing emergent?

00:35:55.920 --> 00:36:01.800
So the way that we interrogate them currently, they act as sort of one big conductor is the

00:36:01.860 --> 00:36:02.560
way I would think about them.

00:36:02.560 --> 00:36:03.660
You can kind of think about that.

00:36:03.880 --> 00:36:05.960
What we're doing here is we have a semiconducting channel

00:36:06.060 --> 00:36:08.160
and we're controlling how much dopant is in there.

00:36:08.200 --> 00:36:10.880
And so we have this biofilm that in and of itself

00:36:11.420 --> 00:36:13.960
without the cytochromes is not intrinsically conductive.

00:36:14.260 --> 00:36:16.340
And then we control how many cytochromes

00:36:16.420 --> 00:36:17.600
those cells have on their surface.

00:36:17.700 --> 00:36:18.900
And so that's sort of our dopant is-

00:36:18.900 --> 00:36:20.860
And it just has to be along the thing.

00:36:21.779 --> 00:36:26.460
Yeah, exactly. And so electrons are top from cytochrome to cytochrome and then from cell to cell through that

00:36:26.900 --> 00:36:28.740
Bifilm. So we think about these as sort of conductive channels.

00:36:29.839 --> 00:36:35.119
Now if you step away from this and you start to think about if we just kept our electrode poised how that

00:36:35.460 --> 00:36:38.779
population of cells is going to be channeling electrons to that electrode,

00:36:39.240 --> 00:36:42.220
that's where you could probably see some emergent properties that are starting to emerge

00:36:43.339 --> 00:36:46.440
from their collective needs as a metabolic entity.

00:36:46.599 --> 00:36:46.880
Thanks.

00:36:48.000 --> 00:36:53.880
Yeah, we haven't ever left them on an electrode long enough and looked at the membrane voltages

00:36:53.930 --> 00:36:57.599
are changing in heterogeneous ways throughout the population.

00:36:58.630 --> 00:37:01.640
I had so much things to pop up.

00:37:02.130 --> 00:37:05.240
I had a recent episode with Neil Thiese.

00:37:06.300 --> 00:37:07.760
I'm not sure if you have heard of him.

00:37:08.000 --> 00:37:10.200
He's a pathologist.

00:37:12.270 --> 00:37:15.800
And a few years back he discovered he was working with Fastia.

00:37:18.380 --> 00:37:20.300
And fascia, so these things.

00:37:21.880 --> 00:37:27.640
It's a bit of, it was at the edge of speculative, it's now clinically published.

00:37:28.610 --> 00:37:35.040
And he has something which you call interstitium, which is the space between

00:37:38.359 --> 00:37:45.059
other cells. And apparently this is a body wide network with connective tissue and the

00:37:45.099 --> 00:37:50.460
fluid system. In principle, just like the thing you built. But just internally, I was

00:37:50.520 --> 00:37:54.319
thinking if if if

00:37:57.200 --> 00:37:58.600
We can learn from each other.

00:37:58.800 --> 00:38:07.460
Or if, for example, those guys say maybe this network is a second organ or independent

00:38:07.640 --> 00:38:12.400
of our nerve system and our endocrine-occurial system.

00:38:12.680 --> 00:38:14.140
There is this interstitium-4.

00:38:15.820 --> 00:38:19.420
message transport, I'm trying to come to the word coherence.

00:38:19.940 --> 00:38:20.340
(laughs)

00:38:21.660 --> 00:38:24.580
To be the way to coherence, not to orchestrate,

00:38:25.210 --> 00:38:28.480
to orchestrate the whole organism

00:38:29.910 --> 00:38:32.420
so that just

00:38:34.160 --> 00:38:38.900
you vibe in the more classical terminology.

00:38:40.080 --> 00:38:45.360
With your network or the biofilm network,

00:38:45.770 --> 00:38:51.180
would this be possible to do experiments that if

00:38:51.180 --> 00:38:52.380
those guys come with hypothesis?

00:38:53.480 --> 00:39:02.520
So we assume X and it's okay, let's build a model of what you assume and we just measure it because it seems to be behaving the same.

00:39:02.700 --> 00:39:11.460
So we have on the one hand we have just a system with electrical conductance and you just have the other thing but you can build it yourself and it apparently naturally emerged.

00:39:14.160 --> 00:39:15.359
I'm just want to make a stink.

00:39:16.140 --> 00:39:19.940
What would you be reaction on this?

00:39:20.640 --> 00:39:23.799
So the interstitial thing that they discovered

00:39:23.980 --> 00:39:25.260
has electron transport properties,

00:39:25.369 --> 00:39:27.539
is that the idea or ionic transport properties?

00:39:28.150 --> 00:39:28.339
Yeah.

00:39:29.840 --> 00:39:34.640
The biofilms that we're building, the matrix protein that we're having, aside from as much

00:39:34.780 --> 00:39:38.700
as a protein as a conductor, they're not particularly ionic or electronically conductive.

00:39:39.020 --> 00:39:43.500
The particular molecules we're using, geobacter, a different organism, like a metric for,

00:39:44.080 --> 00:39:49.120
uses the electronically conductive molecules as sort of their ECM.

00:39:49.600 --> 00:39:55.000
some degree. And so I would say yes, in that sense, you sort of have this, you know, extracellular

00:39:55.960 --> 00:39:59.640
network of molecules that have transport properties and allow them to sort of

00:40:00.240 --> 00:40:03.980
transported distances. They're not necessarily using it for communication, but maybe they are.

00:40:05.740 --> 00:40:08.220
Yeah, so I think there's a good analogy there to be had.

00:40:10.070 --> 00:40:14.420
What sort of, I guess, questions they were to ask and whether or not those track is sort

00:40:14.430 --> 00:40:16.680
of maybe down to the physiology of the organism itself.

00:40:17.640 --> 00:40:18.680
It's sufficiently different.

00:40:20.090 --> 00:40:23.940
But I mean, people sort of the way that we think about sculpting a biofilm in this case,

00:40:24.210 --> 00:40:35.280
people have been thinking about this and tried to doing artificial tissues for

00:40:28.300 --> 00:40:33.700
many decades, right? This is sort of fundamental to that. And people use the same sort of optogenetic

00:40:33.840 --> 00:40:39.800
control to sort of build, you know, larger scale tissues and organs out of mammalian cells.

00:40:40.760 --> 00:40:44.940
Again, that's not that's more structural scaffoldy than it is communication and coherence.

00:40:45.280 --> 00:40:50.240
Yeah, yeah, yeah. If somebody would come like DARPA or the big funding group and say, we have

00:40:50.580 --> 00:40:56.440
five billion, can you design an experiment for us? What would you do?

00:40:56.840 --> 00:40:57.560
Doo doo.

00:40:58.960 --> 00:41:00.579
to compare across this two system.

00:41:02.660 --> 00:41:03.299
Yeah, that's a good one.

00:41:04.779 --> 00:41:06.799
I would want something where basically we can,

00:41:07.339 --> 00:41:08.859
in a localized population of those cells,

00:41:09.220 --> 00:41:10.960
stimulate them in a way so then we can have

00:41:11.059 --> 00:41:12.980
the communication networks basically transfer

00:41:13.279 --> 00:41:14.099
to cells at a distance.

00:41:14.279 --> 00:41:16.599
I think distance dependence in these types of things

00:41:16.619 --> 00:41:18.020
is what I find to be very interesting.

00:41:18.220 --> 00:41:18.559
And so,

00:41:19.000 --> 00:41:26.520
You can ask these from the very simplistic question of how far away can one of these bacteria be from an electrode and still sense and feel and use it.

00:41:27.119 --> 00:41:36.020
And so that's a very simple question, but I don't think we have really great answers in sort of what are the mechanistic limitations used by the different mechanisms cells use,

00:41:37.440 --> 00:41:39.660
what dictates sort of what the farthest distances can be.

00:41:40.140 --> 00:41:43.400
The same could go for communication, right? It doesn't have to just be for energy for metabolism.

00:41:43.900 --> 00:41:45.900
And I think that's where

00:41:46.579 --> 00:41:53.579
like fidelity of signal transfer becomes an interesting question like as we begin to walk away if you have this network that's penetrating all of these cells

00:41:54.520 --> 00:41:58.700
Does this cell population over here feel the same thing with the cell population over here is there's sort of some

00:41:59.560 --> 00:42:05.640
network topology that changes how individual populations of the cells experience the signal that's coming from a single sender cell

00:42:05.839 --> 00:42:08.280
I think those are the sort of questions I find to be interesting

00:42:08.480 --> 00:42:08.900
(chuckles)

00:42:10.120 --> 00:42:12.380
Yeah. Are there dead zones where you can't transport?

00:42:13.100 --> 00:42:14.940
And then, you know, how do you overcome those dead zones?

00:42:15.100 --> 00:42:16.000
Because maybe that's, you know,

00:42:16.000 --> 00:42:16.720
Yeah, I'd rather go.

00:42:17.260 --> 00:42:18.300
Yeah. Yeah.

00:42:19.220 --> 00:42:20.420
And talking about limits.

00:42:23.640 --> 00:42:24.320
What's the current?

00:42:26.140 --> 00:42:31.680
Practical or engineering failure you hit when you, when you don't don't you have limits?

00:42:33.280 --> 00:42:37.920
Yeah, we certainly have limits and I would say probably the practical one is the amount of

00:42:37.980 --> 00:42:43.900
signal that we generate. So I say these organisms are conducted right and they are but the conductivity

00:42:43.930 --> 00:42:49.900
of a Shiwenella biofilm from our measurements is nanoseconds per centimeter and so it's not that

00:42:50.100 --> 00:42:54.860
conductive right if you put it on a spectrum maybe it's closer to an insulator than it is a metallic

00:42:55.300 --> 00:43:00.040
Right. And so we are practically limited basically on the amount of signal that we can measure.

00:43:00.200 --> 00:43:04.760
We have to build large devices with mini channels for these cells to transport electrons across

00:43:05.180 --> 00:43:09.400
for us to measure these types of conductivities. And you know we were talking about scaling. If we

00:43:09.460 --> 00:43:13.860
want a downscaler device and use fewer and fewer cells to actually do the same amount of sensing,

00:43:14.640 --> 00:43:16.440
this is sort of one of our key limitations and so.

00:43:16.920 --> 00:43:22.740
That's something we've run up against in sort of we have a device fabrication set up that works

00:43:22.960 --> 00:43:26.540
reasonably well for our instrumentation that we use to do these measurements. We use these

00:43:26.960 --> 00:43:31.460
research grade potentia stats that could measure picoamps of current without issue.

00:43:32.599 --> 00:43:39.079
We are generating more on the order of microamps for our device so that it's more in a measurement

00:43:40.440 --> 00:43:45.020
Limit where we don't worry about Faraday cages. I'm thinking back to the patch clump

00:43:47.220 --> 00:43:54.300
Yeah, and then a lot of signal signals and I was just thinking as I mentioned patch clumps so

00:43:55.860 --> 00:44:01.960
the difference again between, I was having these glass vials, so you have a glass

00:44:02.260 --> 00:44:07.420
vial and then you make a very small thing to patch them on the muscle and you put

00:44:07.660 --> 00:44:15.540
a silver whatever into it into ringer solution or whatever, and then the

00:44:15.620 --> 00:44:22.200
difference between the wires you built.

00:44:23.860 --> 00:44:25.820
and the just salt.

00:44:40.880 --> 00:44:42.280
(laughs)

00:44:42.860 --> 00:44:46.680
Instead of this using these biofilms or the nanowires,

00:44:46.870 --> 00:44:48.120
would this make sense as an,

00:44:48.280 --> 00:44:50.020
as an, from an engineering perspective?

00:44:50.800 --> 00:44:52.360
Yeah, yes, sorry, you're saying using the,

00:44:52.470 --> 00:44:54.460
the, the, the Persian wires that these cells are a,

00:44:54.830 --> 00:44:58.400
as, uh, helpful in with, uh, sort of patch clamp type

00:44:58.500 --> 00:44:58.620
experience.

00:44:58.920 --> 00:45:00.400
Yeah, I think that is something that's, uh,

00:45:00.680 --> 00:45:02.460
very much a viable, uh,

00:45:03.600 --> 00:45:08.520
device to try to make these types of measurements. The challenge there is generating the protein

00:45:08.700 --> 00:45:15.960
wires at scale to build devices in a way that's not so sort of unstable over time. And so some

00:45:16.140 --> 00:45:19.460
folks who've been spending a little bit of time on how do you manufacture more of these wires that

00:45:19.900 --> 00:45:23.540
just tends to be grow more of the biomass of the one cell. Some people will try to put these

00:45:23.560 --> 00:45:28.120
in E. coli to synthesize them in an easier manner. There hasn't been great successes in that space.

00:45:29.579 --> 00:45:35.440
But I certainly think you can have them function as sort of, you know, sensors for voltages that you might experience.

00:45:37.460 --> 00:45:38.480
We have not tried this now.

00:45:39.020 --> 00:45:42.740
Yeah, a colleague of mine, she has a company who's working on it.

00:45:43.380 --> 00:45:51.500
Yeah, I was just at a conference, this forum on new materials and I'm blanking our name right now.

00:45:52.180 --> 00:45:57.200
But someone really gave a wonderful talk on using organic, electrochemical transistor devices for measuring

00:45:57.900 --> 00:46:00.440
I think I forget what cell type she was using.

00:46:00.720 --> 00:46:01.859
But as an alternative, I would say.

00:46:02.579 --> 00:46:03.740
>> It really could be a sense.

00:46:03.859 --> 00:46:05.140
>> Just a capture.

00:46:05.299 --> 00:46:08.359
>> Yeah. What those devices are a little easier to

00:46:08.359 --> 00:46:10.020
maybe manufacture at scale, I would say.

00:46:10.079 --> 00:46:10.339
>> Yes.

00:46:11.859 --> 00:46:13.660
>> Again, a follower rather than biology.

00:46:13.819 --> 00:46:14.819
>> You just mentioned Pico,

00:46:15.400 --> 00:46:17.220
what have you, milliamps or whatever, my crumbs?

00:46:17.920 --> 00:46:18.440
>> My crumbs.

00:46:18.779 --> 00:46:20.819
>> The voltage is what before the thing burns?

00:46:22.740 --> 00:46:28.700
Oh, so we operate all, so most of biology operates within plus and minus one volt sort

00:46:28.700 --> 00:46:31.040
of from the chemistry of biology that it natively does.

00:46:31.140 --> 00:46:32.860
And so we're operating within that regime as well.

00:46:32.940 --> 00:46:34.080
So it's very up to minus voltage.

00:46:35.480 --> 00:46:42.360
So we typically, what we're measuring is electrons flowing from the cells to the electrode, and

00:46:42.460 --> 00:46:46.080
we'll have the electrode poised at a positive potential of about 400 millihertz.

00:46:46.540 --> 00:46:46.740
I'm sorry.

00:46:46.980 --> 00:46:51.420
relatively little. And then the chemistry of them oxidizing lactate, it's

00:46:51.460 --> 00:46:54.240
sufficiently favorable to drive electrons from the cell to the electrode.

00:46:54.700 --> 00:46:58.740
If we're trying to push electrons from an electrode into a biofilm and then

00:46:58.880 --> 00:47:01.940
back into an electrode, what we need to do is actually keep the potential within

00:47:02.240 --> 00:47:02.480
sort of

00:47:03.100 --> 00:47:06.920
a favorable potential near the redox potential of the multi heme cytochromes

00:47:06.960 --> 00:47:09.480
themselves. So this is about minus 100 millivolts.

00:47:10.080 --> 00:47:11.100
For example, for example. - For example,

00:47:11.140 --> 00:47:12.420
an electrophysiology in principle.

00:47:13.140 --> 00:47:16.380
Yeah, exactly. You just need to be near to sort of, you know, keep those Fermi

00:47:16.700 --> 00:47:20.680
levels near the actual electronic potentials of the heme's themselves.

00:47:20.860 --> 00:47:20.920
Yeah.

00:47:23.600 --> 00:47:26.880
And just thinking 10, 20 years in the future.

00:47:28.800 --> 00:47:29.160
(laughs)

00:47:29.260 --> 00:47:31.160
What have you built in 20 years?

00:47:32.300 --> 00:47:33.140
Yeah, there's a good question.

00:47:33.260 --> 00:47:35.200
So actually coming off your question of potentials

00:47:35.200 --> 00:47:36.200
that you operate at, right?

00:47:36.240 --> 00:47:38.040
Like I'm talking about what we do with Schuenella.

00:47:38.410 --> 00:47:40.380
There are a bunch of diversity of molecules

00:47:40.550 --> 00:47:41.580
that interact with electrodes,

00:47:41.730 --> 00:47:42.600
the biology is using.

00:47:42.790 --> 00:47:44.660
They all have different formal redox potentials.

00:47:44.970 --> 00:47:45.260
And so.

00:47:45.680 --> 00:47:55.560
One thing that we currently are sort of missing at a larger scale is the ability to do sensors with multiple channels of communication at different potentials.

00:47:55.630 --> 00:48:00.580
So thinking about this like we think about fluorescent proteins, we have different spectra of light.

00:48:00.900 --> 00:48:04.520
like we can have different excitation emissions spectra for each of these different proteins.

00:48:04.720 --> 00:48:08.820
We have the same ability with redox proteins. We just haven't been employing them sort of as

00:48:09.940 --> 00:48:16.500
multi-potential elements. So in Prince William would say, I have three different sensors and

00:48:16.500 --> 00:48:20.520
the one is for whatever millivolts and the other one is for whatever or

00:48:21.380 --> 00:48:24.160
Yeah, yeah, so like I have a 200 millivolt sensor. I have a 400 millivolt sensor

00:48:24.360 --> 00:48:27.940
I have a 600 millivolt sensor each of these are hooked up to individual, you know

00:48:28.400 --> 00:48:34.780
populations of cells so that we can watch whatever they're designed to sense so the the goal would be to sense more things and

00:48:34.920 --> 00:48:37.840
Things that we would be sensing is what we would build protein sensors

00:48:41.680 --> 00:48:43.880
Finitely, you know, or maybe we want to measure a magnetic?

00:48:44.000 --> 00:48:47.360
field alongside of some organic molecules if we had that sensor build, right?

00:48:48.340 --> 00:48:53.000
I think the main point being is basically a multiplexing of sensing in devices.

00:48:54.090 --> 00:48:58.480
You can do this from the potential perspective by having different potentials over the redox

00:48:58.560 --> 00:48:58.880
molecules.

00:48:59.410 --> 00:49:03.420
You can do this through device fabrication by having smaller and smaller electrodes with

00:49:03.580 --> 00:49:04.380
the same.

00:49:03.500 --> 00:49:05.460
with more populations of cells that are all individually

00:49:05.700 --> 00:49:07.780
engineered, but we don't currently have the ability

00:49:07.920 --> 00:49:08.600
to sort of do that.

00:49:08.650 --> 00:49:11.100
And so that's where I think this is heading over the next,

00:49:11.240 --> 00:49:13.460
probably in the next decade, more than the next two decades,

00:49:13.840 --> 00:49:18.440
but basically trying to build out more of this.

00:49:18.900 --> 00:49:20.280
What could go wrong?

00:49:20.540 --> 00:49:23.340
So I think what could go wrong is that people are bored

00:49:23.480 --> 00:49:24.200
and it doesn't work or?

00:49:32.790 --> 00:49:32.900
(laughs)

00:49:28.060 --> 00:49:33.740
Yeah, I guess finding the use case where this is the most useful for society will help me

00:49:33.780 --> 00:49:36.099
survive the work in this space for longer.

00:49:36.460 --> 00:49:40.380
I think right now we have really wonderful tools for measuring organic molecules.

00:49:40.859 --> 00:49:43.420
This is what analytical chemistry has brought us.

00:49:44.100 --> 00:49:50.420
the ability to have insight into chemical molecules. But it is throughput limited. And you have

00:49:51.080 --> 00:49:57.720
labs that are built out to do this. They can do this as so many samples a day. And we are currently

00:49:57.850 --> 00:50:01.060
in need of more chemical information than we currently have. So

00:50:01.579 --> 00:50:06.220
We're seeing this a lot in response to these forever chemicals, all of these polyfluorol

00:50:06.579 --> 00:50:07.099
alcohol substances.

00:50:08.720 --> 00:50:12.940
This is a place where we now have rules in the United States, at least from an EPA perspective

00:50:13.089 --> 00:50:18.020
on drinking water quality, but we do not have the capacity of LCMS needed to measure it.

00:50:18.100 --> 00:50:21.120
to actually work for all of our wastewater treatment facilities.

00:50:21.340 --> 00:50:27.160
So the idea would then be to just deploy the kind of life sensors somewhere and then have them collect

00:50:27.320 --> 00:50:31.060
something and then you do just magic upload it in the internet. I mean that's...

00:50:31.400 --> 00:50:34.640
Yeah, more or less. So to have a sensor that's deployed in the environment where we're interested

00:50:34.760 --> 00:50:37.540
in sensing this, maybe this is a drinking water system, maybe this is a...

00:50:37.820 --> 00:50:42.339
a river effluent from that treatment plant and then have that just continuously be monitoring.

00:50:43.060 --> 00:50:46.700
Maybe you have to come and replace the sensor with some frequency, right? Maybe they'll

00:50:46.839 --> 00:50:51.920
foul and they'll go bad over a week or a month or a year. But what you end up getting there is

00:50:52.000 --> 00:50:55.880
continuous information rather than discrete information that comes from sampling campaigns

00:50:56.050 --> 00:51:01.720
that are very infrequent and also very costly for sort of the facility to actually operate, right?

00:51:01.720 --> 00:51:02.040
I am.

00:51:03.320 --> 00:51:08.600
And so that's been one of the regulatory burdens associated with the new past regulations is simply

00:51:08.900 --> 00:51:13.060
that there's a cost associated with measuring these molecules and there's a limited capacity

00:51:13.520 --> 00:51:17.720
for measurement within the globe. And so how do we address those two concerns?

00:51:18.000 --> 00:51:21.560
It's been principle ethical to stay let's just go pandemic prediction.

00:51:22.950 --> 00:51:23.560
Yeah, yeah, yeah.

00:51:23.780 --> 00:51:28.980
In wastewater and then you just know know if you analyze the wastewater that people are whatever.

00:51:29.300 --> 00:51:31.580
So you can detect it in the wastewater first.

00:51:32.980 --> 00:51:37.140
I think this wastewater epidemiology or wastewater surveillance, as people will call this,

00:51:38.020 --> 00:51:42.340
is probably one of the neatest things that emerged out of the COVID.

00:51:43.400 --> 00:51:48.200
And I think it was that we now have capacity to do this. But there are ethical concerns about

00:51:48.340 --> 00:51:52.480
understanding disease spread within populations, especially if they're not well-mixed populations.

00:51:53.680 --> 00:51:57.620
So you end up having treatment facilities that are, you know, very localized to one core population,

00:51:57.850 --> 00:52:03.040
and now you have some idea that they have some disease that maybe has a stigma associated with it.

00:52:04.000 --> 00:52:08.200
I do think there are sort of ethical and privacy concerns related to that that need to be addressed.

00:52:09.600 --> 00:52:11.300
I don't know what the right answer there is.

00:52:12.570 --> 00:52:14.260
You're going to have some, I don't know,

00:52:14.690 --> 00:52:16.560
nowadays we'll have some all-seeing AI

00:52:16.860 --> 00:52:18.180
that's going to be taking in all this information

00:52:18.330 --> 00:52:20.800
and making maybe some decisions about what to do.

00:52:21.060 --> 00:52:23.700
And then it's a fast-changing world.

00:52:23.740 --> 00:52:26.040
Not too fast-changing, but I think the ethics

00:52:26.170 --> 00:52:27.540
is really important discussion because

00:52:29.010 --> 00:52:29.680
when you talk about

00:52:30.680 --> 00:52:36.260
Ethics, we have an external ethics advisory board on a lot of topics also on AI.

00:52:36.380 --> 00:52:40.420
And the question is always, would you do it even without AI or even without

00:52:40.640 --> 00:52:46.859
Europe's thing? And then also is it ethical not to deploy it?

00:52:48.079 --> 00:52:51.640
Right, yes, yeah, I mean this where you could potentially have interventions to overcome

00:52:52.359 --> 00:52:57.800
Specifically diseases that you know can spread readily from human human like should we not be addressing these?

00:52:58.800 --> 00:53:04.460
If we have the ability to monitor when they have upticks in the population and these days always we have to think about cyber

00:53:08.640 --> 00:53:14.380
If I would be interested in protecting the data these things produce

00:53:16.220 --> 00:53:22.600
Can you build maybe with these different sensing proteins?

00:53:22.750 --> 00:53:23.840
Did you say the one is sensing?

00:53:23.930 --> 00:53:25.200
This is what kind of voltage the other one?

00:53:25.280 --> 00:53:28.580
Can you build fault hack proof, bio hack proof,

00:53:29.280 --> 00:53:33.560
signal that just assume in the case we live in a, well,

00:53:34.300 --> 00:53:42.740
Not hopefully far and maybe not ever coming future, but that people are locked down because in the area

00:53:43.880 --> 00:53:46.080
There is a certain signal based on whatever

00:53:47.540 --> 00:53:50.380
And then yes, so maybe we just want to think that you just want to hack it

00:53:51.740 --> 00:53:51.840
Yeah

00:53:54.060 --> 00:54:00.060
That's a good question. We've not put much effort into trying to build safeguards or

00:54:00.300 --> 00:54:03.760
basically a lack of signal. Mostly a lot of our current challenges are just getting the

00:54:03.820 --> 00:54:07.940
signals to happen at all in response to the analytes that we're interested in and not

00:54:08.110 --> 00:54:11.620
other analytes, right? So specificity is like one of the challenges we're addressing.

00:54:11.700 --> 00:54:12.960
Um, I

00:54:14.460 --> 00:54:18.700
The way that I would say maybe to think about that is having multiple discrete mechanism

00:54:19.099 --> 00:54:22.640
by which sensing of our, you know, we think about organic contaminants and so I guess

00:54:22.820 --> 00:54:26.680
the worst biohack here is you think there's a contaminant when there's not a contaminant

00:54:27.040 --> 00:54:30.640
and maybe your population then would act in a way that they don't use drinking water

00:54:30.780 --> 00:54:33.000
or something and this would be a...

00:54:33.300 --> 00:54:34.660
negative for society, right?

00:54:35.040 --> 00:54:37.760
I think what I would like our tools to be used for

00:54:37.940 --> 00:54:40.560
is sort of the initial indication that there's something wrong

00:54:41.560 --> 00:54:44.860
followed by the wonderful analytical chemistry tools that we have.

00:54:44.980 --> 00:54:46.340
This is basically telling, you know,

00:54:46.680 --> 00:54:47.900
one of your operators of the facility,

00:54:48.140 --> 00:54:48.940
we have a problem over here.

00:54:49.020 --> 00:54:51.040
You should go do a sample of chemistry right now

00:54:51.160 --> 00:54:52.680
and so that you have a secondary mechanism

00:54:52.920 --> 00:54:54.020
for doing that sensing.

00:54:54.600 --> 00:54:56.720
So I think that would overcome that challenge in a lot of ways.

00:54:57.940 --> 00:55:02.560
I never relying on that information on its own to make operational decisions per se.

00:55:07.270 --> 00:55:14.320
I'm just looking at the clock and I have a tradition, a kind of tradition in my podcast,

00:55:14.690 --> 00:55:18.820
that I have a quantum computer picking a random number.

00:55:19.859 --> 00:55:20.000
>> Yes.

00:55:20.440 --> 00:55:23.200
>> I have already run it because sometimes you don't know.

00:55:24.780 --> 00:55:26.880
>> How long the queue is greetings to IBM at this?

00:55:27.640 --> 00:55:27.880
>> Yes.

00:55:27.880 --> 00:55:30.980
>> I did run it. You want the question?

00:55:32.230 --> 00:55:32.500
>> Sure.

00:55:32.710 --> 00:55:38.800
>> It did really run on IBM machine and it worked.

00:55:39.280 --> 00:55:40.380
It was not too long in queue.

00:55:41.780 --> 00:55:42.760
The question it picked,

00:55:43.100 --> 00:55:48.640
was a one coming from Ere Gienzen in Copenhagen.

00:55:49.280 --> 00:55:51.600
So Copenhagen, I mean, I'm not sure what it is going on, actually.

00:55:53.300 --> 00:56:00.100
And it is called, "What is the most beautiful pattern you have seen in nature?"

00:56:01.520 --> 00:56:03.360
Most beautiful pattern that I've seen in nature.

00:56:05.940 --> 00:56:11.140
I'm going to answer this with something that I find very interesting that maybe is not the most beautiful pattern in the world,

00:56:11.690 --> 00:56:14.580
but it's the thing that I'm obsessed with, so we find beauty in that way.

00:56:15.140 --> 00:56:18.320
So we were talking about electron transport limitations in JUANELA.

00:56:19.060 --> 00:56:20.800
Another organism I didn't talk about that we mentioned,

00:56:20.960 --> 00:56:22.160
actually it was discovered in Denmark,

00:56:22.380 --> 00:56:24.240
so this is a app to come from someone from Copenhagen.

00:56:25.830 --> 00:56:27.400
Are these organisms called cable bacteria?

00:56:28.340 --> 00:56:32.360
They can form these really lovely cage-like structures

00:56:32.820 --> 00:56:33.820
on their outside of the cell,

00:56:34.520 --> 00:56:35.620
and they use them to connect

00:56:36.000 --> 00:56:37.900
thousands of cells together in a line,

00:56:38.620 --> 00:56:40.860
and then to transport electrons of that line.

00:56:41.520 --> 00:56:44.560
So I find beauty in the simplicity of

00:56:44.680 --> 00:56:47.580
what this organism looks like and then the function that it's actually performing.

00:56:48.320 --> 00:56:48.920
It's just a wire,

00:56:49.560 --> 00:56:52.340
but it's nature's most conductive wire that we've discovered today.

00:56:52.660 --> 00:56:56.160
So this is where I think we're going to be able to use this to address some of our,

00:56:57.100 --> 00:57:00.800
charge transport limitation questions, but also we're just fascinated in how these

00:57:00.940 --> 00:57:04.800
organisms operate. I don't know, it's maybe not the best picture. On my wall over here we have some

00:57:04.940 --> 00:57:08.740
SEM images of these cable bacteria. Okay, tell me, tell me, because then I think one of these

00:57:09.020 --> 00:57:13.660
actually pretty, this one is not maybe our most pretty image. And this is so cool, because I think

00:57:13.730 --> 00:57:19.420
that that's just maybe just going to the closing vision or whatever. So if you could

00:57:20.319 --> 00:57:22.839
Are you planning to use these bacteria or also product?

00:57:23.380 --> 00:57:24.319
How can you produce them?

00:57:24.819 --> 00:57:28.760
So right now, the challenge is understanding how these organisms tick.

00:57:28.980 --> 00:57:32.319
We have a very good mechanistic understanding how she and LNG have to do their business.

00:57:32.559 --> 00:57:35.680
We don't have as good of a mechanistic understanding of what these organisms are doing.

00:57:35.859 --> 00:57:38.059
They're not using the cytochromes everyone else is using.

00:57:38.559 --> 00:57:41.039
They're using some nickel sulfur polymer that they're generating.

00:57:41.260 --> 00:57:43.299
You can see it on Raman spectroscopy very well.

00:57:44.400 --> 00:57:48.060
They also seem to have a charge transport that's not acting in the same

00:57:48.560 --> 00:57:51.820
thermally activated way at all temperatures. And so this is where

00:57:52.580 --> 00:57:55.260
they might be using a fundamentally different mechanism for electron

00:57:55.380 --> 00:57:57.980
transport. And that's why they're interesting. And what I was mentioning

00:57:58.140 --> 00:58:00.660
earlier, the nano Siemens per centimeter conductivity, you get two

00:58:00.760 --> 00:58:03.600
anella. These guys are, you know, Siemens per centimeter,

00:58:03.800 --> 00:58:14.280
that tends of Siemens per centimeter conductivity. So

00:58:06.779 --> 00:58:10.940
Many orders of magnitude more conductive and we don't understand how they work.

00:58:11.109 --> 00:58:13.960
And so we've been trying to get an idea on that other picture didn't come up.

00:58:15.039 --> 00:58:17.740
Trying to get an idea of what is the molecule that they're making,

00:58:18.039 --> 00:58:20.359
how can we make it and then use that to make devices.

00:58:21.000 --> 00:58:25.740
That's sort of our, you know, arc for the next five or so years, I think, on this work.

00:58:27.160 --> 00:58:28.900
Basically, can we synthesize these polymers?

00:58:29.280 --> 00:58:32.859
This is something that biology has learned to extract trace amounts of nickel from their

00:58:32.960 --> 00:58:38.540
environment, turn it into a polymer that's a core part of this skeleton of this organism

00:58:38.940 --> 00:58:39.280
more or less.

00:58:40.280 --> 00:58:41.760
How do we learn from what they're doing there?

00:58:42.260 --> 00:58:44.660
Maybe this will help us inform how we recover nickel from the environment.

00:58:45.280 --> 00:58:49.780
Maybe this will help us make more conductive inorganic materials using enzymatic processes

00:58:50.099 --> 00:58:55.640
that are operating at sort of mesophilic temperatures and pressures and maybe a little

00:58:55.839 --> 00:59:00.040
more green chemistry to get some of the molecules we want at the end of the day.

00:59:00.900 --> 00:59:01.839
And can we source this?

00:59:02.100 --> 00:59:04.660
Or how would this work?

00:59:04.750 --> 00:59:06.700
So do we need to go to the sea and dig it out?

00:59:06.770 --> 00:59:07.940
Or can we produce these things?

00:59:08.270 --> 00:59:08.940
I just want to-

00:59:09.180 --> 00:59:12.000
Yeah, so to date, everyone had been going to the sea

00:59:12.260 --> 00:59:13.680
and basically bringing home some sediment

00:59:13.850 --> 00:59:14.480
or going to a lake.

00:59:14.640 --> 00:59:15.440
They also live in fresh waters.

00:59:16.300 --> 00:59:18.160
Bringing home sediments and cultivating these organisms,

00:59:18.920 --> 00:59:20.680
something that we've sort of one of our innovations here

00:59:20.780 --> 00:59:23.460
is basically coming up with-

00:59:22.100 --> 00:59:28.060
artificial matrices to groom them in. So we basically build our own mud, if you will,

00:59:28.480 --> 00:59:33.820
to build a barrier. We take sand, we take silica, we take organic matter, we mix all of these together

00:59:34.100 --> 00:59:38.880
into a tube, and we can cultivate these organisms on that substrate. And so we can begin to now

00:59:39.400 --> 00:59:44.000
cultivate them indefinitely on something that's chemically defined. Now the question again is like

00:59:44.120 --> 00:59:52.080
how do you make

00:59:45.180 --> 00:59:47.840
a lot of this material and so scaling them is a challenge.

00:59:47.960 --> 00:59:49.860
We are doing these in small tombs in Aquaria

00:59:50.280 --> 00:59:51.880
and so that's something we're working on right now.

00:59:52.300 --> 00:59:53.120
But you can grow them.

00:59:53.440 --> 00:59:55.760
You don't have to go out into the environment necessarily to grab them.

00:59:55.900 --> 01:00:00.220
You can keep them in perpetuity as something you passage from sediment to sediment.

01:00:01.420 --> 01:00:05.220
We're beginning the stages of cultivating them like we do other organisms.

01:00:14.080 --> 01:00:14.180
[NOISE]

01:00:08.260 --> 01:00:11.760
but not in the clean, external cultures that we have for E. coli and Shewanella oneidensis

01:00:11.869 --> 01:00:12.380
Yeah, nothing.

01:00:12.869 --> 01:00:12.980
Nothing.

01:00:13.240 --> 01:00:19.880
There's work done on the biology perspective there, as well as just a better understanding

01:00:20.070 --> 01:00:23.440
of what is the molecule that we're looking at and what are the enzymes that produce those

01:00:23.520 --> 01:00:26.540
molecules because we have no notion of right now other than that.

01:00:26.900 --> 01:00:27.859
There's just nickel sulfur polymer.

01:00:29.210 --> 01:00:34.000
I sensed a bit of all when you showed the picture.

01:00:34.369 --> 01:00:36.040
So you spent your days engineering.

01:00:37.680 --> 01:00:53.700
Still, no

01:00:53.700 --> 01:00:59.619
Why? Absolutely. I tend to actually find field work to be very fun. So, you know, I came up as a,

01:01:00.380 --> 01:01:05.280
you know, there are two tracks of microbiology. There's coax microbiology of doing things on

01:01:05.480 --> 01:01:09.420
augur plates and, you know, having one organism that causes that disease. And then there's one

01:01:09.500 --> 01:01:13.460
gratsky's microbiology where you go out in the environment, you get that organism, you isolate it,

01:01:13.460 --> 01:01:16.440
and you study it as close to the isolation as possible. And so,

01:01:16.839 --> 01:01:20.460
Most of my career I spent doing coke style microbiology and I've gotten to do more of

01:01:20.540 --> 01:01:24.980
Winogratzky's recently and I found that to be very enriching and rewarding. It's hard.

01:01:25.619 --> 01:01:29.700
Often times we just use omics techniques to deal with these types of samples. We extract

01:01:29.920 --> 01:01:34.040
DNA, RNA, proteins, we look at what's there, we make observations about them. The active

01:01:34.540 --> 01:01:37.740
interrogating communities that are in the wild is something that is

01:01:38.860 --> 01:01:42.480
becoming an engineering space in the microbiology world.

01:01:42.560 --> 01:01:45.900
And so there are people who do microbiome engineering in the human gut.

01:01:46.000 --> 01:01:50.420
And we're using a lot of those same tools to work in sediments and wastewater facilities as well.

01:01:51.080 --> 01:01:57.600
So how do you basically deploy a genetic circuit that you've engineered into one of these very messy, very environmental systems?

01:01:58.200 --> 01:02:02.060
And there's some challenges associated with deployment and booting circuits up in these

01:02:02.599 --> 01:02:04.560
hosts that we can't grow conventionally.

01:02:05.180 --> 01:02:09.920
And that's where I'm finding the most interesting new engineering work to be done, right?

01:02:10.640 --> 01:02:12.940
We have these awesome molecular biology tools now.

01:02:13.040 --> 01:02:17.140
How do we go and use them out in our agricultural field or in our sediments to try to sculpt

01:02:18.020 --> 01:02:20.000
the biogeochemistry that's happening in those environments?

01:02:20.440 --> 01:02:21.180
So that keeps you awake.

01:02:22.400 --> 01:02:26.880
It does, it keeps me awake and it's different than building electrochemical devices, right?

01:02:27.580 --> 01:02:32.340
It ends up flexing an ecology muscle that had been long dormant in my brain that's

01:02:32.340 --> 01:02:33.320
now coming back, which is funny.

01:02:34.100 --> 01:02:34.200
Cool.

01:02:37.340 --> 01:02:38.580
If you would have a...

01:02:39.680 --> 01:02:42.540
I'm limited resources and what would you do?

01:02:46.400 --> 01:02:46.800
That's a good question.

01:02:47.080 --> 01:02:51.120
So lately, obviously interrogating the cable bacteria

01:02:51.620 --> 01:02:54.060
conducted mechanisms is the number one thing.

01:02:54.580 --> 01:02:58.800
And then throwing a lot of energy at this question of magnetic field actuation

01:02:59.300 --> 01:03:00.140
and then basically

01:03:00.740 --> 01:03:02.680
Building out the design rules to actuate diverse,

01:03:02.940 --> 01:03:03.340
pretty functionality.

01:03:03.860 --> 01:03:03.940
Yeah.

01:03:05.000 --> 01:03:06.840
A lot of people have spent wonderful time building sensors

01:03:07.160 --> 01:03:09.000
and these engineered proteins for fluorescent

01:03:09.600 --> 01:03:10.000
form borders.

01:03:10.400 --> 01:03:11.800
And it was really great for biosensing.

01:03:12.220 --> 01:03:14.500
But we can use this to control the behavior of biology.

01:03:15.040 --> 01:03:18.140
You see this in sort of optogenetics in the neuroscience

01:03:18.500 --> 01:03:18.680
systems.

01:03:19.000 --> 01:03:24.440
I'm interested in extending that beyond into metabolic

01:03:24.539 --> 01:03:29.319
pathways that happen at ecology scales, how can you re-sculpt the entirety of a microbial community?

01:03:30.220 --> 01:03:34.420
Starting to think of things that look like terraforming, right? I find this to be very interesting.

01:03:35.920 --> 01:03:39.900
And you know, maybe these are useful technologies if we ever want to have an off-earth presence of,

01:03:40.190 --> 01:03:43.099
like, how do we get the chemistry to look like the chemistry that's compatible with us?

01:03:43.670 --> 01:03:48.200
The way to do this is with microbiology. And I mean, the resources you would need to do this on a

01:03:48.260 --> 01:03:50.880
planet are the ones where you would need sort of unlimited resources.

01:03:53.080 --> 01:04:02.420
I'm happy to do it in an aquarium in one of our artificial sediments, but you know, eventually maybe this would be something someone 50, 100 years is going to use to try to do that.

01:04:02.600 --> 01:04:05.740
So, if the usual suspects listen to this.

01:04:09.120 --> 01:04:19.820
Cool. So I very much like the discussion. I would just like to also have my guests to have the last words, so take home messages or wishes or whatever you want to tell our.

01:04:21.260 --> 01:04:25.740
audience before, but before that, before I forget where do people find you if they want

01:04:25.820 --> 01:04:25.880
to?

01:04:26.290 --> 01:04:29.040
Yeah, so I'm based in Princeton, New Jersey.

01:04:29.490 --> 01:04:30.920
So if you're ever in Princeton, drop me a line.

01:04:31.160 --> 01:04:32.140
I'm happy to chat with people.

01:04:32.570 --> 01:04:35.740
But also you can find me on the web on the sort of the usual suspects.

01:04:36.660 --> 01:04:37.560
I'm President on LinkedIn.

01:04:37.880 --> 01:04:39.440
I'm President on Blue Sky.

01:04:39.980 --> 01:04:42.120
My Twitter presence is sort of faded as of late.

01:04:43.290 --> 01:04:48.060
And then, you know, my email, you can find me at jaf4509@princeton.edu.

01:04:48.800 --> 01:04:53.960
people want to send an email that's also great. So good. Do we have that? And now your final wish

01:04:54.240 --> 01:05:02.480
or take home message? Yeah, I guess my take home message is, you know, if you pull up a gram of soil

01:05:02.780 --> 01:05:07.680
or a gram of sediment, it may not look like anything interesting. But what you have that's

01:05:07.740 --> 01:05:09.180
happening inside of that is this,

01:05:09.800 --> 01:05:17.280
dense microbial ecology that's entirely driven like the key organizing principle is the flow of

01:05:17.340 --> 01:05:22.360
electrons from one molecule to another catalyzed by these cells. And so appreciate that. You can't

01:05:22.400 --> 01:05:24.840
see this, but there are, you know,

01:05:25.160 --> 01:05:30.880
Billions of electrons that are flowing around and this is really what makes biology biology right each of your cells are doing it

01:05:30.960 --> 01:05:32.100
They're doing one type of chemistry

01:05:33.260 --> 01:05:39.100
Microbes do oh so many more types of chemistry and so it's good to appreciate the you know the unseen majority of our planet

01:05:39.500 --> 01:05:39.840
I'm

01:05:40.300 --> 01:05:41.760
And for those who are not my microbiologists.

01:06:02.720 --> 01:06:03.280
[laughs]

01:06:03.540 --> 01:06:07.480
Stay curious, until next time and keep exploring the unseen.