First pig lung transplant, and the origins of dark energy

0:28This episode is sponsored by Aerie. and not hold you back. Visit Aerie.com to find the right bra and your perfect fit.

0:54Hello, welcome to this week's Naked Scientist podcast, the show that brings you the biggest breakthroughs and talks to the major movers and shakers in the worlds of science technology and medicine i'm chris smith and coming up a pig lung is transplanted into a man in china so what was the outcome also scientists have a theory for the origins of dark energy and it makes sense that a black hole might be the source we'll hear from one of the team and why the dawn chorus is starting earlier and finishing later these days what is getting into wildlife

1:35First this week, scientists in China have just reported the first successful transplant of a pig lung into a person. A 39-year-old man who was brain dead at the time of the transplant was the recipient because it was an experiment. It ran for nine days. Xenotransplantation, transplanting organs between species, is being explored as a solution to the ongoing organ shortage crisis. Pigs are a favourite because their organs are very similar in size to our own. Previously, hearts and kidneys have been tested in human transplants, but not pig lungs, which are judged to be more challenging owing to the fact that the tissue has to handle blood as well as airflow.

2:15So what did the experiment reveal and is this likely to be a feasible approach in the future. My name is Jas Palm. I'm a consultant physician at Royal Patworth Hospital. My specialist interest is lung transplantation and I've been involved in the field for 25 years. So this is the first time a donor lung from a pig has been implanted into a human recipient and this builds on background work in renal and cardiac transplantation. This was performed in a hospital in China. They have a laboratory where they generated these genetically modified pigs and implanted the donor organ into a recipient in a hospital.

2:55So this is going into a human being so it gives us a chance to test what happens when you put an animal lung into a person. What are the constraints of doing that first and foremost? Yes so the fact that the recipient was deceased obviously has some implication into the way this organ might behave. But what it did demonstrate was that firstly it was technically feasible to do it and the organ was able to function for a period of time. How long did they keep the person going for with the pig lung in there? The organ survived for nine days in total. They had a number of challenges which included using some fairly industrial immunosuppressant medication to lower the immune system.

3:41Did it go in the pig organ into the chest where a lung would normally go? So it behaved to all intents and purposes like the person's own lung? Yes, I mean the paper's a little light on detail around that but that's the assumption is that this was placed in the normal anatomical position with the normal connections. And how did they actually monitor what was going on? So they were using a number of parameters to look at the lungs They were including x-rays, blood tests and function of the lungs in terms of arterial blood gases. And with these they had a composite picture of how the organ was performing.

4:20And how was it performing? So it appeared to be doing reasonably well up until about day two when it developed a severe immunological based reaction which then stopped it working well. And thereafter? Again the paper's a little light on the detail around that but it seems that the experiment was stopped at nine days and the assumption is that was when the organ actually completely failed. You mentioned that the pigs that were the donors for these organs have been genetically modified. In what way? Clearly there are a species difference in the way in which we handle foreign proteins and so if a pig organ is transplanted into a human there will be a number of protective mechanisms which would attack the organ.

5:06So of the processes they use, they use genetic deletion to try and make the pig organ look a bit more like a human organ. And does this look like, therefore, on the basis of what they present, it is a direction to go in? Does this look promising? Because it's been done for kidneys. It's been done for hearts. OK, only in the short term at the moment. This was the first lung. But does this give us hope? Yes, I think as a proof of concept, it demonstrates a number of things there is still a huge amount of work that needs to go in to make this a durable and sustainable enterprise which can enter the clinical arena but yes it does it does demonstrate that this is technically feasible.

5:49Is this the sort of nuclear fusion of the transplant world where you know it's 10 years away and always will be is it the same sort of thing for that? Yes I remember conversations around this we had a very active xenotransplant program in Cambridge and many of the old sagely people around at that time said this is tomorrow's medicine and will remain tomorrow's medicine. But more seriously does this look like it's going to be practical or do you think that someone bright will come along and be 3D printing organs and we'll be having human-based tissues well before we ever get this problem solved?

6:24Is this just too big a nut to crack? The immune system is too hard to defeat when we're doing this? Yeah I mean I think the biology of the protection against you know foreign proteins is an immensely basic phenomenon and very strong and I think if you can get to a point where you can develop organs from stem cells you obviate the need for immunosuppression the risk of infection goes down and the whole thing is much more attractive so if I were to put some money on something it would be on the developing of 3D printed organs from stem cells because I think that's a way more attractive option. But there's no doubt there is a big clinical need for this kind of thing isn't there?

7:08Absolutely so there's a desperate need in the UK and we have a very large waiting list with a number of patients dying before they receive an organ. So each of these steps are welcome and if they enter the clinical arena they will have a place in treating patients. Royal Papworth Hospital lung transplant physician, Jas Palmer there, who took a look for us at the paper in the journal Nature documenting that recent pig to human lung transplant over in China. And if you want to hear more about these recent steps in xenotransplantation, the Naked Scientist actually interviewed the surgeons in the US who did the pig heart transplant into recipient David Bennett back in 2022.

7:46If you want to give that a listen, look up the man with a pig's heart, Naked Scientist, and you can hear what he had to say. Well now let's turn to space and one of the most enduring mysteries of the last century or so and that's the expanding universe. And it isn't just growing, as the universe expands the rate at which it grows expands too. To account for this scientists invoke an amorphous entity called dark energy which is the impetus behind the accelerating universe. But therein lies a question. Where is the dark energy coming from? If the universe is accelerating at an accelerating rate, there must be an increasing amount of it.

8:26Well, to be perfectly blunt, no one had the foggiest. But recently, an initiative called DESI, the Dark Energy Spectroscopic Instrument, which is tracking millions of galaxies, some of which date back 11 billion years has spawned an intriguing hypothesis that black holes at the centres of galaxies are the source of dark energy which they spew out as they consume matter. Colas Frank who's part of the DESI team is a cosmologist at Durham University. The program is called DESI where DESI stands for dark energy survey instrument so we're trying to answer one of the most fundamental questions in physics what is the dark energy?

9:08Now, dark energy is something we know exists because it's causing havoc with the expansion of the universe. We have known for almost 100 years that the universe is expanding. Hubble discovered that. But what was discovered at the end of the last millennium and came as a huge surprise and a shock is that the universe is not only expanding, but the expansion is getting faster and faster and faster. The expansion is accelerating. implying that there must be a force out there in the middle of nowhere that is pushing galaxies, making them expand away from each other at an ever-increasing rate. And the instrumentation you're using to try to study this, tell us what you're actually recording and measuring and how that will shed some light on dark energy.

9:59This is where the real breakthrough has come about. DESI is a unique instrument. The instrument is intended to measure and analyze the life from galaxies. Traditionally, you point your telescope at one galaxy and you get information about that galaxy. With DESI, we can target thousands of galaxies at the same time using fiber optics that are positioned with robotic arms. So that instead of only being able to get one galaxy at a time, each shot gives us tens of thousands of galaxies. We have measured now 40 million galaxies, which is 10 times more than before DESI. And have we now got, thanks to DESI's observations, clearer ideas as to what dark energy is and what it's doing and where it's coming from?

10:52Well, we now know exactly what it does. And people have come up with suggestions. is one that has become very popular in the last few weeks because of a very provocative paper where they posit that, in fact, the source of the dark energy is another type of fascinating object in the universe, a black hole. So according to this idea, when a black hole forms and sucks in matter, as black holes tend to do, then the very intense gravitational field around the black hole turns the infalling matter into dark energy. Particular types of black holes can cause this conversion of ordinary matter as it falls into dark energy.

11:40How does the dark energy get out of the black hole? Is it in the same way as we see radiation from things getting very, very hot when they orbit a black hole and it's radiated away? Or does the black hole have an arsehole, for want of a better way of putting it. And sort of the black hole is crapping out dark energy. How do you see it working? Right. So one of the great things that Stephen Hawking did was to demonstrate that actually black holes can produce particles, they can evaporate. But this is a different phenomenon from what we call Hawking radiation. So even though black holes are, apart from Hawking radiation, completely invisible, material as it falls in gets heated to such high temperatures that it can emit radiation, usually in the x-ray.

12:25So we see x-rays coming from black holes, that's how we know where the black holes are. Well, the dark energy is analogous to that. So it's not coming from any part of the anatomy of the black hole. It's simply that the material, just before it gets swallowed, gets converted into the dark energy, and then the dark energy spreads out in the universe, creating Hubble with the way galaxies expand. And is that dark energy everywhere all at once, or does it aggregate around a black hole for a while? So could we sort of do a survey looking for hot spots of acceleration because there's lots of dark energy there, or does it spread out so quickly that that experiment would be implausible and impractical?

13:06Well, yes, the dark energy, it's in fact affecting the geometry of space-time, and it does that essentially at the speed of light. Black holes are more or less uniformly distributed around the universe on the largest scales and so the dark energy very soon pervades the whole of space distorting the geometry of space-time and so there are no hotspots unfortunately. That would be very nice. If we knew, look, there's a black hole forming here let's go and see if we can detect an excess of dark energy that would never work because it is a global phenomenon that affects the universe as a whole. So how do you think we could test this then?

13:47It's an intriguing idea, but what experiments can we put in place to find out whether black holes are genuinely the origin of dark energy in this way? Well, that's an extremely good question. And the onus is on the people who put forward this theory to make concrete testable predictions. There's no answer that I know, but that clearly is the focus of intense research. That is very new. It was just published a few weeks ago. So not surprisingly, people are still working very hard on exactly that. How do we test it? How do we know if this is the right answer? And we'll just have to watch this space in the meantime.

14:27That was Durham University's Carlos Frank. This episode is sponsored by Aerie. If you know me, you know that my philosophy is all about looking good and feeling good every single day. And that honestly comes down to the first layer. I'm talking about bras. If your bra is digging into your shoulder or maybe it's riding up in the back or you are dying to take it off at the end of the day, then that bra is not the right fit. And what I love about Aerie is that they have 59 different sizes in really soft fabrics and really good fits. Because at the end of the day, your bra should hold you in and not hold you back.

15:03Visit Aerie.com to find the right bra and your perfect fit. The Naked Scientist podcast is produced in association with Spitfire, cost-effective voice, internet and IP engineering services for UK businesses. Find out how Spitfire can empower your company at spitfire.co.uk. Music in the programme is sponsored by Epidemic Sound, perfect music for audio and video productions. This is the Naked Scientist podcast with me, Chris Smith. Still to come, how light pollution is keeping birds up at night. But is that a good thing or a bad thing? We will find out. Before that, though, to the world of tech.

15:47And US technology stocks have been suffering something of a rough ride recently after a report from the Massachusetts Institute of Technology, MIT, found that billions of dollars' worth of investment in AI tools has so far offered no return to the businesses that have bet big on new models. The market turmoil was compounded by comments from Sam Altman, who's the head of ChatGPT's parent company OpenAI, who claimed that investors were, in his opinion, overexcited. Tech speak possibly for a bubble on the way to bursting. Shares in NVIDIA, which make the chips fuelling the AI boom, have continued to slide, wiping hundreds of billions off the company's market valuation, despite NVIDIA posting annual revenues above expectations.

16:32But Raghu Raup, who's Professor of Finance at the Cambridge Judge Business School, is more bullish about the situation. Here he is speaking to James Titko. What the MIT study did is something similar to what Robert Solow, who was a Nobel Prize winning economist, asked as a question back in the early 2000s. He said computers are showing up everywhere except in the productivity statistics. So his basic conclusion was we are not actually having much of an impact on organizations from computers. Of course, that was completely wrong because they were having a big impact, but companies needed to retool themselves to properly take advantage of these new methods of working.

17:16And the same thing is happening with AI right now. What's happening essentially is you may invest a lot in AI, but nobody really knows how to use it very well yet. The market spooking was really not that big a deal. 3 % to 5 % is normal on a normal day. So I wouldn't really call it spooking the market in the sense of more a mild market correction in a way. The report isn't saying that generative AI is a dud then. It's just turning out not to be quite so easy to integrate into our workflows as some people maybe had originally envisaged. Absolutely right. I think that's a very good way of putting it.

17:56I myself find it incredibly productive. I use it almost every day. But how do I actually measure the improvement in their productivity? That's really difficult to tell. Nevertheless, Sam Altman, who, as we've mentioned, runs OpenAI, the company that operates ChatGPT. He's had something of a reality check for investors when he spoke last weekend, didn't he? Well, yes, that's correct. We just saw the release of GPT-5, and honestly, it's been a little underwhelming. One of the questions we have is, are the large language models really keeping up the same rate of progress as they had from going from GPT-3 to GPT-4, or is that rate going down?

18:39At the moment, the idea consists of cleaning large language models on more and more data. But as we saw from DeepSync in China, there are other ways of doing this, and we don't know yet which system will take off. Investors have to struggle with how do we detect reality from fantasy. The hope is that these will be enormously transformative industries, like the airlines industry, like the internet, like blockchain. These will be transformative in some way or the other. So we want to buy anything which is vaguely associated with these industries, but we don't know which ones are the ones going to succeed.

19:19All Hartman said was, don't jump ahead of the gun by investing blindly just because something sounds like AI. But investors will say, if I don't do this, I'd be missing out on opportunities. So it's a sort of fear of missing out. And I don't think Aartman's comments are going to change any of that either. And some people have been predicting that the value of language models might come when they're designed for specific purposes, when they're trained on curated, heavily filtered data to complete bespoke specific tasks. Yet this MIT report seems to kind of fly in the face of that hypothesis. At the end of the day, we simply do not know how these large language models work.

20:03They have, think about GPT-4, it had over one trillion parameters. That's unimaginable for a human mind. So what you have is a system which is so complex that it is almost impossible for anyone to explain why the AI system is coming up with a particular type of answer. None of these are intelligent. They may sound human, but at the end, they're just taking the data from all over the web and putting that into a statistical model of the English language or any language you need that you may have. The OpenAI method has been to throw more and more data to get more and more parameters. And other countries like DeepSeek and Mistral in France are trying to compete by not going so much in the direction of a super large language model.

20:53The interesting part, of course, is if you curate the data, if you come down to a tiny fraction of the data, it's not a large language model anymore. It is a specialized data set. And in fact, a couple of companies, in fact, like McKinsey, tried doing things like this. They used only their own internal documents to create their version of a GPT. But it was simply the number of connections you can get from a smaller data set is much weaker than the number of connections you can get from a large language model. That's the essential issue. People are trying new innovative ways to try to figure out how to get the same kind of jumps in understanding with smaller models.

21:36Ragu Rao, and you can hear more about the AI space and some aspects of this story on Tuesday when we'll be taking a look at how the widespread uptake of AI is potentially harming users. That's next week. Before then, though, as human civilisations sprawl across the planet, so too does artificial light, often throughout the night. the consequences for human health of light pollution as we really ought to be thinking of it are beginning to be understood they upset our body's circadian rhythms our biological clocks which affect when we sleep and how we metabolize food but it's not just we humans that are affected the natural world is feeling it too and a new study published in science has found that birds are also waking up earlier and staying awake for longer researchers recorded calls from locations around the world and found, of the species that are usually awake in the day, those with the biggest eyes and the most open nests are also the most affected.

22:33Neil Gilbert is at Oklahoma State University. So we were interested in understanding how light pollution, or light created by humans at night time, is affecting bird singing and calling behaviour. Light pollution now affects about a quarter of Earth's surface. And previous work has indicated that it can mess up the timings that organisms have evolved in response to these light-dark cycles. How did you quantify and measure it then? What did you actually do? So we were using a big data set of audio recordings of birds. These came from a citizen science program called Bird Weather. Anyone can go and buy one of these listening devices for$250 and put these out in their backyards.

23:18They take them with them on hikes. They're embedded with a machine learning algorithm that automatically identifies bird songs and calls. And so these volunteers, these users, sort of get the experience of learning about their birds in their garden or backyard. But then these data are also stashed in this open source database. Millions and millions of bird detections. We have latitude and longitude coming from those listening devices. And then we related that to satellite remote sensing data on nighttime lights. And so we linked those up and did a bunch of fancy statistical modeling to get at the relationship between the first vocalization of birds in the morning and the last vocalization in the evening.

24:03And where are they across the Earth's surface? Is this just a U.S. project? Is it international? Where have you looked? It is an international study, but it does follow this pattern. It's quite common in ecological studies where there's a lot of data from the global north, places like US, Canada, Western Europe, and less data from the global south. So, you know, there's some gaps in the sampling, but still it's a quite impressive data set. We had data from almost 8 ,000 individual locations. And what did you see? So our major finding is that these birds have prolonged days in these light polluted areas.

24:37They start singing and calling about 20 minutes earlier in the morning and about 30 minutes later in the evening. So that's 50 minutes, almost an hour longer, across 583 species, across a full year of data, so multiple seasons, and across all of these locations. So of course there's a lot of variability. Some species show a lot stronger responses. Some species show less responses. And we found that species that have large eyes tend to be more sensitive to light pollution. The other was nest structure. So species that have open nests, cup nests, things like your American robins, Eurasian blackbirds, those showed stronger responses than species that nest in tree cavities.

25:21So things like woodpeckers, nuthatches, those sorts of species. And we suspect that's because these cavities are basically acting as shutters or blinds. It does sound like a compelling story for light, but how can you exclude other things? Because where there's light, there are people. Where there are people, there are noises and other disturbances. It could be sound as well, presumably. Right. Chris, were you a reviewer of the paper? There's a whole lot of correlated variables, things that go along with each other. So light pollution, of course, is prevalent in cities. Sound pollution is also prevalent in cities.

25:55And there has been quite a bit of work showing that sound pollution affects birds as well. So we did some extra analyses to try to check our results. There aren't really global data sets on sound the way there are with light pollution, but we used a distance to road as sort of a proxy for noise pollution. So we looked at the latitude and longitude of each one of these listening devices and measured how close it was to a road. Here the assumption is that most noise pollution is going to come from traffic noise. And really our results stayed the same. We found this effect of light even when we accounted for sound.

26:33The study dwells very heavily on light and therefore dayactive animals. What about nocturnal species like owls, for example, because you might expect to see the flip effect for them? Yeah, that's a great question. And we did do an extra analysis looking at nocturnal species, owls or night jars. And we did find evidence of what you're suggesting. And it seemed like on average they are vocalizing less and they are vocalizing over a shorter period of time. And is it a bad thing if the birds get up earlier, get to bed a bit later? Is there evidence that this is impacting them negatively? Or is that just a function of them fitting in their lives around our lives?

27:13You know, we initially assumed that, oh, this must be a bad thing for these birds. I was thinking of, you know, my own experience. If I lose an hour of sleep at night pretty quickly, I'll get pretty grumpy. And, you know, there's a lot of medical research on humans showing that sleep debt is associated with really adverse health outcomes. But bird sleep and human sleep is pretty fundamentally different. And so there's a possibility that prolonged activity might not be linked to sleep loss if birds can sleep during the day or perhaps sleep with half a brain hemisphere at a time. There's some work showing that birds can sleep mid-flight by shutting brain hemispheres on and off.

27:53So, you know, it's not necessarily a negative effect. There could potentially be a positive benefit for these birds of prolonged activity if, for example, they could feed their young for a longer period of time. There's been some evidence of that here and there of American robins, for example, feeding their young under streetlights at 1, 2 in the morning when it's dead of night. The one caveat that I'll give is that there are other mechanisms by which light pollution does have clear negative effects on birds. So one is migration. A lot of bird species migrate at night, and they are attracted to these artificial lights on skyscrapers and other human structures, often leading to fatal collisions.