Episode Transcript
22. OI is the new AI. Have you heard about Organoid Intelligence? Feat. Johns Hopkins University prof. Thomas Hartung
===
[00:00:00] Professor Thomas Hartung: Imagine the following. I want to study the human brain. I'm interested in how the brain works and how diseases impact on this. How can I work on this? I can work to some extent with humans who are alive, but I cannot really look into their brain. I cannot intervene. I cannot manipulate a lot. I can use rats. Or mice. Or also monkeys. But this is very limited because first of all, they're very different. They learn differently. The access is also very limited. One macaque is at the moment increasing in price from about 5, 000 dollars before the pandemic, to 35,000 to 50,000 dollars.
[00:00:55] Mizter Rad: Hello, beautiful humans. I'm sitting down today with the one and only Thomas Hartung. He's a professor at Johns Hopkins Bloomberg School of Public Health, and an expert on OI - organoid intelligence. Listen to this. It's not ai, it is oi, and you heard it well. Organoid intelligence. Thomas, Come stai? Wie ghet's dir?
[00:01:20] Professor Thomas Hartung: Tutto bene, wundervoll!
[00:01:22] Mizter Rad: Well, Thomas speaks several languages, by the way, and he right now is in beautiful Italia. That's why I said, Come stai? Thomas, let's break it down right away because I, I think the title, it's already... A bit telling, but it is very unclear to me what is organoid intelligence. So maybe you can tell us what OI, organoid intelligence is.
[00:01:47] Professor Thomas Hartung: Okay, sure. Organ intelligence combines two of what we call disruptive technologies. So the things which are really changing the world as we know it, at least in biomedicine. Which is the possibility of creating organoids. And AI, and we call this organ intelligence because we combine brain organoids with the artificial intelligence capabilities.
And this is opening up for some quite incredible stuff of learning and memory taking place in a petri dish essentially.
[00:02:19] Mizter Rad: Taking place in what, sorry?
[00:02:20] Professor Thomas Hartung: This is opening up for learning and memory in a dish. We call them petri dishes originally. But it's it's just saying that cell cultures can suddenly do things which were before only seen in animals and humans.
[00:02:33] Mizter Rad: Okay, wait I need to go slowly here, slow, slower here because, uh, so, so basically a, a organoid intelligence is a combination between AI and organoids. I need to understand, let's leave AI on the side because I think by now a lot of people know or have an idea what AI is and what AI can do. But organoids, let's focus on organoids.
What is organoids?
[00:02:58] Professor Thomas Hartung: Yeah, this is actually a technology which has evolved mainly over the last 15 years or so. Which is very similar to AI disrupting the way we do research. Organoids have become possible mainly because of the advent of stem cells. Stem cells are cells coming from human donors, which are able to produce essentially every cell type and every tissue.
And this technology became possible in an ethically clean way only in 2006. Yamanaka received the Nobel prize for this technology already six years later, which is almost unheard of because it is it was so evident that this is changing by, by medical research. He found a way of taking whatever cells: skin originally, in the meantime, you take blood cells or cells from urine even, and we reprogram them to become something like an embryonic stem cell.
Technology has over the last one and a half decades enabled to produce brains, kidneys, livers, whatever you want. And this is really something which is extremely empowering because if you want to do research in medicine, in the life sciences, we need models. And these models were either animals in the past, which are very different to humans. I always say we are not 70 kilogram rats. But also the human cell models were terrible because I'm working on the brain. If I ask somebody for some cells of their brain, they will nicely decline. It is, it is really something which is, uh, where we don't get access. Because patients are brain dead by definition, but I want living brain to work on.
If I could possibly get them from after they deceased. So it's really about... Getting good cell culture materials from humans in a very reproducible way. And this is really an enabling technology which is allowing us in all fields of medicine at the moment to come up with new models.
[00:05:06] Mizter Rad: So, okay, let me see if I understand correctly, the, the organoids topic or concept is mainly used for basically improving the way we do research as humans.
Because you're using, you're using cells, let's say this technology helps researchers like you, specialists like you to use a cell that is human but is still alive basically, or? Explain me a bit more, because I think I'm not, I'm not getting it. Because you talked, you talked about organoid, sorry, stem cells.
[00:05:43] Professor Thomas Hartung: Okay. Imagine the following. I want to study the human brain. Um, because I'm, I'm interested in how the brain works and how diseases impact on this. How can I work on this? I can work to some extent with humans who are alive, but I cannot really look into their brain. I cannot intervene. I cannot manipulate a lot. I can use rats. Or mice. Or also monkeys. But this is very limited because first of all, they're very different. They learn differently. They, the access is also very limited. One macaque is at the moment increasing in price from about 5, 000 dollars before, before the pandemic, to 35,000 to 50,000 dollars to 50,000 dollars so it is very difficult to do research..
[00:06:29] Mizter Rad: Wow. Wait, wait, why did the price go up so much after the pandemic? Do you know?
[00:06:34] Professor Thomas Hartung: Mainly because China stopped exporting them. They did 60% of the supply in the US at least. And this is really putting companies at the moment under enormous pressures because for some registrations of drugs they need non human primates, as they call them monkeys, and they sometimes now have to wait for about a year to get their studies done they need for registrations.
[00:06:59] Mizter Rad: So there's a, there's a little bit of traffic jam in the research environment because of this...
[00:07:05] Professor Thomas Hartung: very much so. And you cannot imagine how valuable any day of delay is. Um, people have calculated that one day delay of registration is worth about 2 million dollars for a pharmaceutical company.
So you can imagine how they're all at the moment trying to solve this problem of not having access to this. But we're getting a bit sidetracked here. I'm just saying it is extremely difficult to get them and for researchers like me if I were to work with monkeys, which I don't it would be very difficult to get them even at the moment.
[00:07:40] Mizter Rad: Right.
[00:07:40] Professor Thomas Hartung: And this is why the possibility of getting something bioengineered from human cells is such a fantastic opportunity. And I mentioned already Yamanaka, who in 2006 described how to do this. Yeah. Because he found a way of taking an adult cell, for example, from the skin, and reprogram it to become something like an embryonic cell.
And as we know, embryonic cells produce all the tissues of the body. So it is only a matter of finding the right tricks to make, out of these embryonic cells, now the real tissues.
[00:08:20] Mizter Rad: Okay, wait, let's pause in there. So basically, instead of using animals, You say you don't do that anyway, but in general, instead of using monkeys, for example, for research, you could take a stem cell, an embryonic cell, and reprogram it so that it looks like a brain cell of a human, basically, and then do research on that.
[00:08:47] Professor Thomas Hartung: This is exactly what we did. So we started receiving funding for this type of research in 2012, from the NIH. And we started building brains. Unfortunately, others were faster than us, because already in 2013, when we just were experimenting with the first of these models, this was already published by somebody else.
We were only the fourth group actually to produce brain organoids in 2016. But we were the first to mass produce them. Thousands of identical brain organoids, tiny, tiny, tiny, we called them mini brains at the time. But because they were so standardized, they opened up for testing. Because if you can...
have many, many of these brain organoids, then you can test different substances on them. You can test different doses of a, of a substance on, on them, or you can manipulate them in very many different ways in order to do the research.
[00:09:47] Mizter Rad: Give me an example. Sorry, Thomas, give me an example. So I don't, I don't lose this track because I, I find this super interesting.
What do you mean with testing certain substances? Give me an example of what you test on this brain organoid. So now I know that this brain organoid sort of resembles The human brain, right? So instead of testing on a alive human brain, which would not be possible for obvious reasons, you can test on this mini, mini brain, mini organoid brain, mini brain organoid.
What kind of tests do you run on this for example?
[00:10:25] Professor Thomas Hartung: I mean, our work is, for example, motivated very much by the tremendous increase in autism, which we are observing you know, autism is a developmental problem of the brain. And it was something which was extremely rare in the seventies. About one in 10, 000 children was diagnosed in the United States to be autistic. The latest numbers from March this year say that in the U. S., one in 36 children is now diagnosed with autism before the age of eight. And this is an enormous increase. And this, this is no, can also no longer be explained by better diagnostics and awareness. It is really something which is a clear increase for which we need to find causes. And since the genes do not change in this short timeframe of a few decades, it must be something with our lifestyle, with the exposures we are seeing. And so we are asking what is the possible contribution of some chemicals to this type of, uh, of increasing disease. We have the chemicals, but we need a model. We need something which is showing us whether the development of the brain is derailed in the presence of these chemicals. And that's exactly the type of research we are carrying out with these brain organoids.
[00:11:40] Mizter Rad: So you, you, you sort of inject the chemicals into the, the mini brain, the mini organoid, and then you see some development, some results?
[00:11:48] Professor Thomas Hartung: Exactly. So we have fun.
[00:11:50] Mizter Rad: Is this, is this coupled to a computer, like to a software so you can like visualize it on a screen kind of?
[00:11:56] Professor Thomas Hartung: That's exactly where we are heading with the organs intelligence part. At this moment, we are mainly looking into how do these different cell populations form?
How do they connect to each other? Is there a regular development of the structures and architectures of the brain to the extent we can study this in these brain organoids?
And this is really fascinating. We can see, for example, that if we do these Brain organoids from cells, which originate from a patient which developed autism, they behave differently.
Or if we introduce...
[00:12:32] Mizter Rad: what do you mean?
[00:12:33] Professor Thomas Hartung: Yeah, they, um, um, we can see that they are, for example, more sensitive to the effects of certain chemicals.
[00:12:42] Mizter Rad: Okay. Wait, wait, give me a second. So let's say there is a patient with autism. Do you, do you, what would you do with this patient? Would you take a, would you, would you, okay.
Explain me from the beginning, you take a stem cell from this person from this individual and then you...
[00:12:58] Professor Thomas Hartung: yeah. So nowadays we would typically take a little bit of blood and um, the white blood cells are in the, in, in the blood can be reprogrammed to become stem cells. So it is nothing unusual. You need tiny amounts of nothing, which are taken in routine examinations anyway. But with the consent of the patient or and or the parents of the patient and you can start developing stem cell lines From them and can then work essentially in eternity with the material of this patient...
you can work endlessly with these with these with these cells. So we can freeze them. We can saw them. We can work with them. Whenever we need them We can make out of a few cells. We can make billions of cells. And for this reason we don't need continuous supply with this, this blood of patients we need once.
And then if everything is green lighted for our research, we can start working with this.
[00:14:01] Mizter Rad: Okay so this patient that comes to you, you take out some, some white blood cells. You've reprogrammed them into stem cells. With, few cells that you take from this person you can make millions of, of cells.
But and then when, when does it, when does the organoid come into play? Like you can now make an organoid based on, on that, right? And then you can test on this specific organoid that sort of relates to that specific patient, right?
[00:14:31] Professor Thomas Hartung: So the really cool thing is that we, that you have a patient and you know, what type of disease the patient developed, what symptoms are. And at the same time you have a functioning little representative of the brain of this person, which you can subject to tests and you can see is there any chemicals which make this worse or are there any treatments possibly, which could improve on this?
[00:15:00] Mizter Rad: Okay, super interesting. So you're basically testing sort of like real time, the brain of the patient, but in a non invasive way in a way.
[00:15:14] Professor Thomas Hartung: Exactly. So I have this, uh, in the laboratory available while at the same time know what the patient suffers from. And I can start to explain to which extent was it the genetic makeup of this patient.
And to which extent, for example, could this have been subject to effects of chemicals?
[00:15:36] Mizter Rad: What kind of chemicals or toxins do you inject into the organoids to test the different development?
[00:15:44] Professor Thomas Hartung: I mean, there's a couple of chemicals we know quite well who have these effects. For example, some heavy metals. Lead, is an example of it.
Heavy metals we call them. But also arsenic is among them or mercury, because we have seen that in some cases of intoxications, they have produced exactly these effects in children. But they don't explain the broad variety of of cases. Still we know that exposure to higher amounts of these metals is producing problems in brain development.
But there's also large groups of other substances like, uh, PEs, some pesticides. You know, we kill, we kill insects mainly be, uh, by, uh, interfering with the brain of the insects. And, uh, for this reason, these substances are always suspicious also to have some effects on the human brain.
[00:16:37] Mizter Rad: Ah, interesting. So, are these pesticides?
Um, maybe found in some foods?
[00:16:47] Professor Thomas Hartung: Yeah. Um, we have to say that mostly these pesticides nowadays are used in low doses and they're also decaying fast enough that we don't have relevant amounts in the food after harvest. But people, for example, who are working with them can be exposed to pretty high doses.
And if these are pregnant farm workers, children could be exposed. Or people living close to these farms could also be exposed to these pesticides. But it's not only pesticides. There's also a lot of other substances which are highly suspicious. For example, we are using, especially in the U. S., a lot of so called flame retardants.
[00:17:31] Mizter Rad: What is that?
[00:17:32] Professor Thomas Hartung: Flame retardants are substances which make your furniture burn not as fast. So it is meant to protect you against a fire in your home, but this is pretty nasty chemistry. A lot of this is actually looking very similar to some pesticides. And for quite a while already people have got worried that we are possibly trying to do good here, but we're actually doing really bad.
[00:18:00] Mizter Rad: You mean with the, with the flame retardant?
[00:18:03] Professor Thomas Hartung: Yes. The flame retardants. You have to imagine some furnitures sold in the U S are one third flame retardants. That's the amount of chemicals they have to put in to achieve a relatively minor delay in the possibility to set these things on fire.
[00:18:21] Mizter Rad: And is is this like transmissible or absorbable through the skin or how does it come into your, into the body?
[00:18:30] Professor Thomas Hartung: I mean, probably it comes into the body mainly as dust. Out of the furniture because these things are bound to the foam, to the materials in in the furniture. But after a while you get small particles like the microplastics, you know, from which are other discussions so much at the moment. And on in this way, we are inhaling this.
So imagine sleeping on the mattress, you're very close to possibly tons of chemicals. If this is not a biological material which is then making you inhale and by this in this way taking up particles which include these these chemicals.
[00:19:10] Mizter Rad: Wow.
[00:19:11] Professor Thomas Hartung: That should be very clear. This has not been proven completely.
We know about the risks and there is attempts to replace these substances because of these concerns, but nobody has shown it is the flame retardants, which are responsible, let's say for 10, 20, 30% of the increase in autism. This is something which is only possible to show if you have models. And this is why it is so important that our research in the life sciences, gets models with these organoids, which are human, which are relevant, which have the genetic set of an autistic patient, which means somebody who is vulnerable to develop the disease.
[00:19:51] Mizter Rad: Okay, so the objective is to get human relevant models that can be used to demonstrate certain, let's say, consequences of chemicals, heavy metals, pesticides, flame retardants, or any other sort of toxin, that is affecting our brain, sometimes with the effect of a condition like autism, but I guess there is more than that, is that correct?
[00:20:21] Professor Thomas Hartung: Exactly. So our personal interest was out of autism, so already in 2005. And we started to work on developing these models because we wanted to have something to test what chemicals do for autism. But now with these models becoming better and better. And we also work with people who are interested, for example, in Alzheimer's disease.
In the beginning I was a bit skeptical because Alzheimer's is something we develop if you're 50, 60, 70 years old. So I was wondering whether really cell cultures which are only reflecting early development of the brain could actually help us here. But astonishingly, again cells which come from Alzheimer's patients give us brain organoids which are somewhat different. Which shows some of the aspects of Alzheimer's disease.
And this is super cool because we can now ask, is something else contributing to this. For example Alzheimer is increasing dramatically, much more than the aging of the population would explain. So we forget to ask, is there possibly some chemicals in our lifestyle which are impacting on on this disease and are accelerating the development of of this devastating condition?
[00:21:42] Mizter Rad: That's super interesting. I have, I have, I have some, some things I wanna know without breaking the flow of the conversation because I think we have a good flow. Why the brain? First of all, because this could, as far as I understand, this could be applied to any other organ.
Is that correct?
[00:22:01] Professor Thomas Hartung: That's correct. And this is what people are doing. The field is at the moment really exploding. And, uh, we call these models, microphysiological system (MPS). Which means in a micro scale, very, very small our brain organics are just visible. Yeah. They have the size of an, the eye of a house fly.
And these...
[00:22:21] Mizter Rad: the size of what, sorry?
[00:22:23] Professor Thomas Hartung: The eye of a house fly. So it's really half a millimeter.
[00:22:27] Mizter Rad: Wow. So they're almost microscopic. You have to work always with sort of some, some sort of visor or microscope to, to, to manipulate them or?
[00:22:37] Professor Thomas Hartung: Exactly. But you can still grab them. Yeah. Place them from one vial into another, um, as, as you desire.
But, um, they are really small. Because we don't need more. We also have no concerns really what these organoids are possibly capable of, which is an issue if we come back to organoid intelligence later. But the, the point is we have these model system and you can do this with liver, with kidney, with heart, with muscle, name it.
I have had the fortune to be able to really become an opinion leader in this field. And we have just started a series of world summits for these microphysiological systems and also created the first international society in this field. And only two weeks ago, we had the second world summit in Berlin with 1, 300 researchers.
Which is really important as a community coming together and developing the models for the life sciences in the 21st century.
[00:23:34] Mizter Rad: Is this somehow solving the issue of not having enough monkeys for research and making, as a whole, research in these topics cheaper?
[00:23:47] Professor Thomas Hartung: Yes. From my point of view, these models make animal testing less important because human relevance is, is key. I always say we are not 70 kilogram rats. And this is a very important sentence because it is really showing you the quintessence of all of this.
We need something which is reflecting human. And this is only really possible in high quality since we have these micro physiological systems. It's a revolutionary technology.
[00:24:18] Mizter Rad: And is this... At the moment only thought of, let's say the main application of this technology is for research, right?
[00:24:27] Professor Thomas Hartung: Um, I would say, so the main application is certainly to understand how the organs are working really. But then it is very quickly coming to drug development. And it is coming to the area of of toxicity testing because there is an important contribution of chemicals and exposure to disease. And we have difficulties studying this because at the moment we are very much limited to animal tests, which are costly and not always relevant.
[00:24:57] Mizter Rad: Have you found any, some sort of ethical or legal barrier when it comes to testing on organoids or mini, mini brains, mini organoids?
[00:25:08] Professor Thomas Hartung: Sure there is. You have to see that whenever you work with these organoids, There is a human being, which is still running around, and it is very important that we have consent with this individual what we are doing.
[00:25:23] Mizter Rad: Let's stop right there. The organoid is related to a specific person, and it is unique.
Almost like a, it's non fungible, like an NFT, almost. Like it's, it's unique to that person that cannot be something that are going to cannot belong to someone else. Is that just to have that clear?
[00:25:43] Professor Thomas Hartung: Yeah, not belong. It is, it is different. So if I do one from you and one from me, the organ is somehow reflecting the two of us, which is beautiful because it allows us to do personalized medicine, for example.
You might benefit from a drug better than me. And we can could test this in such a system. Or looking from the perspective of a drug company. By using enough donors, I, but to find out how many of the patients will really benefit from a certain drug. Or how many will possibly show side effects, because I have them represented in my, in my test tubes. So it is really a game changer away from inbred rats, which are essentially identical twins. There have no variability to something which is really reflecting a human being is all of the different differences in genes and often also differences in the the experiences we had over life. The diseases we have, uh, we have had.
[00:26:50] Mizter Rad: The exposure with, uh, with the environment in general. Like if I grow up in, in the tropics I have a different sort of like bacteria composition than someone that grew up in the North Pole.
[00:27:04] Professor Thomas Hartung: Yeah. And for example, there's no such systems of the, of the gut of the gastrointestinal system where they even bring in the microbiome.
So the individual bacteria different people have, and they're studying what is different here? You have to imagine we have about 900 different types of bacteria in our gut. If we are, if we don't know each other, only about 300 are the same. If you live together with someone for many years, about 600 are the same.
But still, um, 300 roughly are different between me and my wife. Even though we are sharing our food and and we are living together. Yeah. So imagine how much influence this can have on the development of disease. And this is highly human specific. And this is really enabling technology to study things like this.
Another very prominent example is a virus infection. We just went through COVID and viruses are extremely human specific. Um, We, until now, don't have a reasonable animal model of COVID 19. But we have tons of models now for each and every organ which are based on these stem cells where we can study How is it impacting on the heart?
How is it impacting on the kidneys? And we have, and this is absolutely human specific and can only be studied in human systems.
[00:28:31] Mizter Rad: So you're sort of decentralizing, tailoring the way we study for example, the way toxins or chemicals in the environment affects each one of us, but not only that, each one of our organs.
And you can go maybe even on a deeper level, beyond organs. And this is fascinating. Am I getting the idea correctly?
[00:28:56] Professor Thomas Hartung: Absolutely. So to give you an example from COVID again., When we all were confronted with COVID in February of 2020 as suddenly becoming a major threat, we learned that at least 30% of the patients show some brain symptoms, neurological symptoms.
So I, I was, I asked instantly, I did a talk in early March: can the brain be infected with this? Because some of the Corona viruses, this is a family of 450 different viruses, some of them infect the brain and others don't. And it took me only until May to have the results of the respective experiments because we had these brain organoids ready and we were the first to show in May of 2020 that the brain is infected by coronavirus. And that they are multiplying in our nervous system. And this was really important and hasn't has been in the next year being reproduced by 10 other groups. We had a human brain available to very quickly test. And this was before any animal model could show something similar on this. Until today no brain infection of the brain in an animal model, which is not Of an, of an, yeah, of a naturally occurring animal, which we studied.
[00:30:10] Mizter Rad: So what effects, what effects did you find in the nervous system from the Corona virus?
[00:30:17] Professor Thomas Hartung: I mean, the good thing is the effects were relatively small which means the virus is multiplying.
We find about 500 times more after three days. But it is not a massive infection, which is erasing many brain cells. It seems to be a very slow infection, but it could very well be involved in long COVID. In some of these fatigue syndromes we are seeing in patients. This is something to be shown. Also could the fact that the brain can be infected and in some portion of the patients is infected as clinical studies have shown, is adding bad news to a pile of bad news. Because brain infection cannot be resolved by the immune system and also typically not by drugs.
So you have to live with this infection probably for your life. And we will only learn in the next decades what this means for autism again. Or what it means for Alzheimer. Because virus infections are also driving this type of diseases. I think it is really important that we have models to study this.
[00:31:24] Mizter Rad: Is there the, is the pharma industry or bigger organizations paying attention to this technology? And are they funding this? Or who is funding this? Who's behind this? Who is interested in this?
[00:31:38] Professor Thomas Hartung: I mean, our work has been funded first by the N I H. And then the FDA, the Food and Drug Administration. And the E P A, the Environmental Protection Agency, um, came in.
But increasingly now we see that the pharma industry is is interested. I have been presenting over the last six months to at least five pharma companies who are considering these type of models now for the, for their research. And this is really a very big trend that pharma, about 25% of all clinical trials are on brain diseases. So it is extremely important for them to have.
[00:32:15] Mizter Rad: Is this, sorry, is this after, is this after COVID? Cause brain health kind of became a big deal after COVID or, or in general, anyway,
[00:32:24] Professor Thomas Hartung: no, this is the data from before COVID already. So brain health has been one of the biggest things for pharma industry, because these are the big diseases where you have to treat patients, millions of patients for long periods of time.
Um, many, many years of neurodegenerative diseases, but you also have multiple sclerosis. You have Parkinson's. There's a, there's quite a few of these diseases, which are of high interest for pharma, and where we don't have adequate treatment at the moment.
[00:32:57] Mizter Rad: Do you fear sometimes that the connection between the food industry and the pharma industry, or the connection between the chemical industry and the pharma industry sort of hampers or limits the, the reaction time on, for example, the pharma testing through your technology or the technology we're talking about here? Certain effects of certain chemicals that are put into foods or, you told me the example of the furniture that I found incredible, and so: is there, is there sort of like conflict of interest here?
And so do you fear that the pharma is kind of like stopping or pausing or not funding fast enough this development so that we as humans can come up with fast diagnosis on what is actually causing autism?
What is actually causing Alzheimer's and all this mental health, but beyond mental, I guess, um, do you, do you sometimes think about this?
[00:34:04] Professor Thomas Hartung: Yeah, I'm not, I'm not really worried about this part. I mean, there's certainly a disconnect between them in the sense of the pharmaceutical industry has the big pockets and is investing a lot of money in relatively few substances.
There's about 20 new drugs which are coming to the market per year. 20. That's not a lot of chemicals. And they are studied extremely well because they absolutely don't want to poison their customers. They, they are tested not only in all of these animals studies, but they're also tested in patients.
We are observing enormous number of patients for possible side effects of these drugs. And this is in big contrast to the chemical industries. Whether it's now in the food sector or whatever. We have thousands, tens of thousands of chemicals which are, have never been probably tested because nobody can afford to test them.
They don't have the same profit margins as the pharmaceutical industry. So they are really very much there's, there's an enormous gap of knowledge. And this is one of the big reasons for why these technologies offer something also for the chemical side, because we can now ask to test certain things which were not tested before.
[00:35:22] Mizter Rad: Interesting. Why, why do you think, I mean, this is more maybe like a legal policy question, but why do you think the pharma has such a... Complex regulation and not the chemical? Because if the chemical industry or the food industry can just like launch products, thousands of products per year, without much pushback. And the pharma industry needs to go through like tons of clinical trials to launch one product.
You think that you see that as a problem as well, in some sense?
[00:35:59] Professor Thomas Hartung: Yeah, first of all, I have to say that's the fact. I mean, the, to bring a drug to the market is on average 2. 4 billion dollars of investment. 2. 4 billion. And it's a 12 year process. Any... Individual animal test or cell culture test is a tiny part, part of this. But it is for them very important to take the right decisions to put their money on the right horses in the process. To be better than the competition. Or to be faster in the development.
Drugs are completely different beast. We have thousands of chemicals which are entering the market. I recently found a study which showed that there's 350, 000 chemicals which are registered in the 19 most developed countries.
[00:36:53] Mizter Rad: 350?
[00:36:54] Professor Thomas Hartung: 350,000.
[00:36:55] Mizter Rad: 350,000?
[00:36:57] Professor Thomas Hartung: Okay. Yeah. So nobody could assess them reasonably. With with the same box of tools that the pharmaceutical industries employ.
They spend five to 10 million for a single drug which goes into into human trials.
[00:37:13] Mizter Rad: Why is it so expensive by the way, Thomas?
[00:37:17] Professor Thomas Hartung: Because animals are actually pretty expensive. And their care and everything around them, even more so if you want to have good data. The rat is 30 to 50 dollars. I already mentioned the monkeys, which used to be around 5, 000 dollars, if they're laboratory grade monkeys. They're now 35 to 50,000 dollars.
And this is why it becomes expensive. But more than the animals themselves, it is the facilities. It is the caretakers. It is all of the professional handling of these animals, which makes these experiments expensive. So to give you one example, most people would love to know whether a chemical is producing cancer.
That's, that's our biggest concern.
[00:38:03] Mizter Rad: Yeah, absolutely. I would love to know that. I mean, when I, when I literally changed my deodorant two weeks ago, because I didn't trust on the chemical composition of what I was using before. So, um, but like that, there's so many products that I would like to actually check real time.
If I go to the supermarket, it would be great if I could have an app. Okay. So that somehow tells me specifically to my body, is this going to be good or not? For example.
[00:38:31] Professor Thomas Hartung: Yeah. And, uh, what do we have at hand? There is an animal test, which is the accepted method to test whether chemical produces cancer.
This costs 850,000 dollars. And does take you about five years until you have a result.
But this does not match the needs of an industry, which is permanently exchanging chemicals, right? Because you have to treat these animals, 400 rats for two years, every day of their lifetime.
You need for this five to 10 kilogram of the chemical, which telechemists to produce. And then after these two years of treatment, it takes about two years to cut these animals into slices and look for tumors in any possible organ. And this is what makes it so costly. And this is why only a very, very tiny fraction of all chemicals has ever been tested, whether they produce cancer or not.
[00:39:30] Mizter Rad: So with OI, with this technology, you're kind, you're kind of solving that problem as far as I understand for the chemical industry as well. And that's even more in a way appealing to them because they're releasing, launching, dropping new products more so than the pharma industry.
[00:39:53] Professor Thomas Hartung: Yeah, it could become because if we now with the Organoid Intelligence do the next step, which is: we can, we are in the process of putting memory and learning into the systems.
So the real function of the brain. So this promises to be a much better, much more sensitive test system for the effects chemicals or drugs have on the brain. And at the same time, these are relatively inexpensive tests. These brain organoids are much less than a dollar a piece. It's not comparable. And they don't need a cage.
They don't need feeding. They don't need all of this infrastructure. You can put them in tiny, tiny devices and, and, and do your measurements. So it is really an enabling technology to understand what is possibly impacting on the function of the brain?
[00:40:47] Mizter Rad: You, you see like at some point having like OI banks, like organoid intelligence banks where I store my OI related to my heart, and then my OI related to my liver. And then somehow there, maybe this a bit too off in the future, but somehow you can have it... have like an app or something on your phone that is connected to that organoid that, and you can test sort of like the chemicals that are around you, whether it's your shampoo or deodorant or the food you're eating or whatever. Is that, is that somehow the, the way you see it, maybe far in the future?
[00:41:35] Professor Thomas Hartung: Uh, I would say it's not, this is all possible, whether you would do it on an individual level of let's say.
Having your organoids and test on them probably you would have need to have deep pockets to enable this, yeah. But to do it, to understand the variability of different humans, for sure. This is where we are already. We will...
[00:41:58] Mizter Rad: what do you mean with that?
[00:42:00] Professor Thomas Hartung: Yeah, we can already produce from essentially anybody brain organoids or other organoids and can compare them to see how different we humans are.
And we can look into gender differences. We can look into genetic differences because you have a certain disease. You can look into different ethnicities. So there's a lot enabled here. We also will be presenting this month at the at a conference for the first time that we can freeze these brain organoids.
So we can actually produce a bank of organoids. And then I can say, okay, in two weeks I want to do an experiment for which I need three patients... brains from three patients with Alzheimer's and two with Parkinson's. And then I take five controls. And I can just saw them and use them.
So you can imagine how much this enables. So you can bring the same organoid in any country. You can even bring it to the space station, if you like.
[00:42:57] Mizter Rad: Super, super, super interesting. What do you mean with memory? Injecting memory into the OI . Cause you, you, you mentioned that, but I didn't quite get it.
[00:43:08] Professor Thomas Hartung: Yeah, this is the, actually, this was the big message of, um, announcing these efforts and the programs on organs intelligence. We have for the first time now the machinery of learning and memory in these systems. So we have everything which is needed to make them learn. So for example, some of our partners from Australia, Cortical Labs, they had a nice paper in November on neural cultures playing pong, the computer game. So they could show that they can train these cell cultures to get better and better in playing a computer game. And they're now working with our brain organoids because we also have the machinery for long term learning. Because their cell cultures became better and better in the training session, but next day they had forgotten everything. Because they simply were not having all of the architecture of the brain, which is necessary.
And, and, um, in this way, also in our research projects at Hopkins, we are at the moment trying to demonstrate long term learning. We are trying to find ways of developing test systems which are allow on the one hand, the testing of chemicals and drugs, which I described so far, but also which allow us to better understand how long The brain is such a incredible computer outperforming in some aspects the best computers in the world.
[00:44:37] Mizter Rad: Help me understand. I would, I would like, because this is also on audio, I, we cannot show images or a video, but I would like to have you help us visualize. When you're standing in the lab, and you're telling me that an organoid can play some computer game and memorize some stuff. How do you visualize that?
How do you know that is happening in the lab? What? What do you see? Help me help me visualize this moment.
[00:45:08] Professor Thomas Hartung: Organoid Intelligence is actually the combination of three very quickly developing technologies. We call them disruptive technologies. The one we talked about so far is mainly the engineering of the brain organoid itself.
But in order to communicate with the brain organoid, I need sensors. And this is also something which is extremely rapidly developing. We call them microelectrode arrays, which means we have tons of small electrodes in a very compact design. So compact that essentially each and every of the cells of the brain organoid is in contact with at least one of these electrodes.
And this allows us to communicate, to record them like an EEG, an electroencephalogram, you know, when you record the brain activity of a human. But we can also feed information inside. We can tell the brain this is where the ball of the pong game is moving. And this is where the pedal is at the moment. And, um, so this is, this is the sensor technology side.
And the third technology is AI. Because we can really communicate with the organoid. So we can start understand how do we train them best? How do we get reproducible answers of the system in order to optimize their learning behavior?
[00:46:35] Mizter Rad: Would you say an organoid? Okay, so now I understand there's three technologies, the sensors, the actual physical engineered organoid, and the AI.
And when you combine these three, this new technology emerges. Would you say this is a live in itself? Does it, is it, would you consider this a living organism? You know what I mean?
[00:47:04] Professor Thomas Hartung: Yeah, I mean, it is first of all, it is living because I mean, these cells are alive. It is reacting to the outside signals, but this is a cell culture as well.
It is not in the sense of an autonomous system. It cannot walk around. It cannot decide which inputs it wants to get. It is also a very small system. So it has less neurons as a fly has, for example. But it is something which is clearly perceiving the signals and it is improving its behavior.
It is some type of synthetic biology.
[00:47:38] Mizter Rad: That's super, super interesting. And I guess for some people it's frightening.
[00:47:44] Professor Thomas Hartung: Yeah, I think the... this is why we have ethicists involved from the very beginning. I don't think that there's anything frightening at this stage. As I said, we are talking about very, very small systems.
We are talking about very limited information, experiences, they can make. So I'm absolutely not worried that I would produce a conscious system. Something which is developing emotions, sentiments in any way. But we should foresee that when we scale this, this could happen at some point. And this is why we work with eticists who design with us the experiments. Who come into our labs and when they see these tiny, tiny balls of cells, which are, uh, you normally need a microscope to see them well. This gets you much more humble and you understand that this is nothing to compete with even a handheld calculator in the foreseeable future. Yeah.
[00:48:44] Mizter Rad: Right.
[00:48:44] Professor Thomas Hartung: But it is at the same time, it is extremely important to have these discussions so that we are prepared to communicate this properly.
That we are also setting limits of what such a system can do and should do. And this is why this is such an important part, uh, to have these ethical discussions.
[00:49:01] Mizter Rad: No, I completely agree. I mean, I'm, I'm the reason why I invited you cause I think this is super, super exciting more than frightening. But this is my personal view.
Um, I understand that maybe some of my listeners would find it more frightening than exciting. But it's a good point that you have ethicists in your team to make sure that at least we think about this also from a moral, ethical standpoint. Even if we're still very early in this process. I have a question, actually, because when you were talking about the gut microbiome, I immediately thought about all those hundreds of bacterias that we have there.
What is the relationship or how can you actually use bacteria in the organoids in this structure between sensors, organoids and AI. Do they play any role? Could they play any role? Because I feel like bacteria is such a undervalued part of our bodies, and yet so important. So important. And so what's your take on this?
[00:50:15] Professor Thomas Hartung: First of all, you're right. We have 10 times more bacteria in and on our body than we have cells of our own. And they have gone through evolution with us. These bacteria have developed a symbiosis with us. We are caring for them. And they care for us in many aspects. And we only start to understand what the impact on diseases and also how much they make as different. If you take the twins, for example, they have, they start essentially with the same microbiome, their gut.
And then over their lifetime, they completely diverse and uh, in, in age, they only have about one third of the, of the bacteria in common. So there's a diversification which takes place and nobody knows how much this is impacting on you as a being. And probably there's quite a few substances which are coming from these bacteria which are permanently in our bloodstream. And we're only starting to understand how they impact on our metabolism and anything we do.
[00:51:25] Mizter Rad: Interesting. Interesting. Thomas, any last words? Anything you would like to share in this beautiful conversation that we just had?
[00:51:35] Professor Thomas Hartung: I think the big message from my side is really that we see these disruptive technologies which are changing the life sciences. Making models available. This has, first of all an instant impact on how we do medicine and medical research, drug development, or find health threats. But the big perspective is also by understanding how the brain works, we can learn how to build even better computers.
You know, the best supercomputer in the world, the frontier computer in Kentucky, only reached the performance of a human brain last June, one year ago. And this was a 680 square meter installation, which did cost 600 million. So, so far away are we from the computational power of a human brain. So there's an enormous prospect by understanding how the brain does the trick.
How we can do with a million times less energy on just 1. 4 kilograms of brain do these marvelous things.
So it is really very, very interesting how different the brain is in its performance. So for example if you want to teach a child, what is the difference between a horse and a donkey you need about 10 pictures, yeah?
And then they get it reasonably well. But you need several hundred pictures to train, um, a computer to distinguish them at the same error rate. And you need only a single picture to explain to a child what a unicorn is. We would need several hundred pictures again to make a computer distinguish this.
Or if you want to add knowledge, let's say, we've learned 10 words in Italian to stay with the, uh, with this from the, from the beginning. You would need to rerun the model. While we learn progressively. We can just add knowledge to what we already know. So we have some, some properties which are really unique and by learning how the brain does the trick, we can build better computers. So this is neuromorphic computing and and this is something where our better improved understanding can help a lot.
[00:53:50] Mizter Rad: That's that's fabulous. One last thing I want to know, and I want to pick your brain on this, Thomas, from your Perspective as a pharmacology, toxicology expert, professor off the Johns Hopkins Bloomberg School of Public Health and experienced, microbiological immunological, medical professor: what or how do you see the world from that lens developing in the next 50 years? From the lens, from the perspective that you are at the moment? In the following 50 years, 100 years, maybe go a bit further. You can get crazy with the answer. No one is gonna judge you.
[00:54:35] Professor Thomas Hartung: I mean, first of all, I'm a notorious optimist because I see how the availability of science and technology is improving our life, our longevity...
the last 40 years that Our original investment the human genome by sequencing human genes has really created enormous progress. A similar investment into the exposure side of human health. The exposome. Instead of the genome create something very similar. Because exposure has at least as much influence on our health as genetics has.
And if you understand both, we can do all those things in preventing diseases, but also in, um, healing and curing diseases. So I'm, I'm very optimistic that, uh, our increased understanding enabled by these type of technologies is really helping us to improve on a large scale,. Human health and hopefully not just for for the highest industrialized countries.
[00:55:41] Mizter Rad: That's beautiful. That's beautiful. Thomas. I have to thank you so much for your time. It was a fantastic conversation. I really had fun. I learned a lot of things. I hope our listeners also got their brains working and moving. And this is absolutely incredible what you're working on. Thank you so much once again for being here.
[00:56:05] Professor Thomas Hartung: My pleasure. Thanks for having me.
[00:56:07] Mizter Rad: Thank you so much. Buona serata.
[00:56:09] Professor Thomas Hartung: Buona serata. Alla prossima!
