This episode reviews a study that links in vivo insulin sensitivity phenotyping with proteome and signaling-pathway mapping to define molecular-phenotype associations across heterogeneous populations. The paper emphasizes population heterogeneity and maps proteomic signatures to functional signaling pathways associated with insulin sensitivity.
0:00Welcome to Base by Base, the papercast that brings genomics to you, wherever you are. Thanks for listening, and don't forget to follow and rate us in your podcast app. So imagine 2 people sitting in a clinic waiting room.
0:11Okay, just a standard everyday doctor's visit. Right. Patient A is officially diagnosed with type 2 diabetes. Patient B is declared perfectly healthy with completely normal glucose tolerance. Which seems like a very clear binary medical situation.
0:27You would naturally think so. I mean, we expect medical diagnoses to be precise like that, right? Like, you break your arm, the x-ray shows a fracture, and you're putting a cast. It's either broken or not broken.
0:37Exactly. Black and white. But if you were to look deep inside their muscle tissue and measure exactly how their cells respond to insulin, you might find something completely baffling. Patient A, the one with the official diabetes diagnosis, is actually more responsive to insulin than patient B.
0:53Wait, really? The diabetic patient is responding better. Yes. Which begs the question, how is that even medically possible? When you zoom into the metabolic machinery of the human body, That neat categorization we rely on just completely falls apart.
1:08It really does. So what actually happens inside our muscles when insulin knocks on the door. And why do some bodies answer while others, you know, just keep the deadbolt locked? And more importantly, how could unlocking this exact cellular mechanism change the way we treat metabolic diseases forever?
1:25Well, to answer that, today we celebrate the work of the researchers at the Novo Nordis Foundation Center for basic metabolic research at the University of Copenhagen and the Karolinska Institute. They have significantly advanced our understanding of the molecular mechanisms behind insulin resistance.
1:40It's a huge step forward. It really is. This deep dive into the underlying architecture of metabolic disease is based on a truly staggering piece of research. The paper is titled Personalized Molecular Signatures of Insulin Resistance and Type 2 Diabetes.
1:55It was published in the journal Cell, volume 188 on July 24, 2025. Hot off the presses. Exactly. And it was authored by Geppie Kergaard, Ben Stocks, a Tuls Deschmuck, Anna Crook, Julian Arzirath, and their very extensive team of collaborators.
2:12And the scale of what they are tackling here. I mean, it really cannot be overstated. We are looking at a disease that affects over 500000000 people worldwide. It's massive global health issue. Yeah. And as that paradox with patient A and patient B hints at, tied to diabetes is just incredibly heterogeneous, like it looks acts and damages, completely differently depending on the person.
2:34Right, it's not a one size fits all condition. Definitely not. But out of all the organs involved in this, you know, the liver, the pancreas, the fat tissue. Why did this massive study, 0 in specifically on skeletal muscle?
2:47That's a great question. It's because skeletal muscle is quantitatively speaking. The main tissue involved in insulin stimulated glucose uptake. Oh, really? Not the liver. Nope. When you eat a meal and your body releases insulin, your skeletal muscle acts as the body's primary glucose sink.
3:03It handles the vast majority of the load. Wow. So because it's the primary sync, It is naturally the primary site of insulin resistance. If the muscle stops responding to insulin, well the whole system just backs up.
3:15Okay, let's unpack this. If the muscle cell is a factory, and insulin is the delivery truck bringing the glucose, we've actually known for a while that the trucks arrive at the loading dock just fine. Exactly.
3:27Yeah. The body produces the insulin, the cells have the insulin receptors on their surface, and they have the glucose transporters ready to go. So if the loading dock is clear, and the trucks are arriving.
3:39Why isn't the delivery being processed? That right there is the multi-billion dollar question. For decades, the scientific community has recognized this as a post-receptor defect. Meaning the problem happens after the insulin binds.
3:52Right. The issue isn't a lack of receptors on the outside of the cell. It's a profound communication breakdown somewhere on the factory floor inside the cell long after the insulin has already knocked on the door.
4:03But mapping that exact failure across a large, diverse population has been technologically out of reach until now, right? Yeah, totally. You couldn't just look at the whole factory floor at once. Historically, we could only interrogate maybe a few workers or a couple of machines at a time.
4:20Which is why the methodology of this study is just so fascinating. They didn't just take a quick blood draw. They managed to look at the entire factory in real time. Walk us through how they actually pulled this off.
4:32It was a massive undertaking. They recruited over 120 individuals. Wow. Yeah, and they split them into a discovery cohort of 77 people and a validation cohort of 46 people to, you know, ensure the data was robust.
4:46Okay, that makes sense. And this group included both men and women, comprising individuals categorized with normal glucose tolerance, and those officially diagnosed with type 2 diabetes. See how the factory handles a delivery.
4:58They used a procedure called a hyper insulinemic euglycemic clamp. I've heard of this, but it sounds incredibly intense to actually perform on a human being. Oh, it is. It's the gold standard for measuring insulin resistance, but it's very complex to run.
5:15Hyper insulinemic. Basically means they infuse the participant with a constant high amount of insulin. Right. Flooding the system. Exactly. And euglycemic means they simultaneously infuse glucose to keep the patient's blood sugar perfectly stable and normal.
5:31So they're artificially driving the delivery trucks while making sure the blood sugar doesn't crash. Precisely. By measuring exactly how much glucose they have to pump in to keep things stable, they calculate a highly precise score called the M value.
5:44So a high M value means the factory is working perfectly, like you are highly sensitive to insulin and chewing through that glucose. And a low M value means you are highly resistant. Correct. But here is the really critical part.
5:57While that clamp procedure was running, They took physical muscle biopsies from the participants' thighs, specifically the vastest lateralis muscle. Yeah, not the most comfortable procedure. They took one biopsy before the infusion started.
6:11When the patient was in a fasted resting state, and then they took a 2nd biopsy 30 minutes in when insulin was actively flooding the system. Taking muscle biopsies while someone is hooked up to dual IV infusions is real dedication from those volunteers.
6:26But the technology they use to analyze those tiny pieces of muscle is where I get a bit lost. The paper mentions state of the art mass spectrometry. Uh, specifically DIA PaceF on a Timbstone Pro 2. What does that actually mean for the data?
6:42Like, why couldn't we do this 10 years ago? It really comes down to resolution scale. Older mass bitrometry was like, well, trying to take a picture of a crowded room where everyone is moving around. The faces just blur together.
6:52Too much noise. Exactly. This specific technology uses something called ion mobility separation. Think of it like sorting a massive pile of mail, not just by zip code, but by the exact weight, the shape, and the electrical charge of every single envelope simultaneously.
7:08Oh, wow, that is precise. It separates the molecules so cleanly that they don't overlap. Because of this, they were able to map over 3000 distinct proteins in the muscle. Which, going back to our analogy would be the physical hardware of our factory.
7:23The machines, the conveyor bolts, the actual workers. Yes, but even more impressively, they mapped over 29,000 phosphorelation sites. Okay, let's pause there and make sure we really understand the biochemistry.
7:36If the proteins are the hardware, How should we think about these phosphor relation sites? Think of phosphorelation sites as the software or like the physical on and off switches on those machines? When a protein is host related, meaning a phosphate group is chemically attached to it, The protein physically changes shape.
7:55And that shape change does what? It alters its function. It turns it on, turns it off or flags it to move to a completely different part of the cell. By mapping 29,000 of these specific sites, the researchers got a real-time snapshot of the entire communication network flashing across the factory floor.
8:12That is mind blowing. So we have this unprecedented map of the hardware and the software. When they laid it all out, what happened to those standard medical categories we talked about in the hook? The diabetic versus healthy buckets?
8:23They completely dissolved. It gone. Totally gone. When the researchers plotted the complete molecular map, the signatures did not cluster neatly into a healthy group and a diabetic group, the molecular realities of the patients mapped to a vast, continuous spectrum.
8:39Wow. Yeah. And that continuum aligned perfectly with the M value, the actual physiological insulin sensitivity we measured with the clamp. Wait, how wide was the spectrum of, like, between the most and least sensitive?
8:51Within just the discovery cohort, there was a staggering 36 fold difference in insulin sensitivity from the most resistant, the most sensitive person. A 36 fold difference. That is massive. And I guess that completely explains the paradox from earlier.
9:08You could have a patient technically labeled diabetic based on, like, a fasting glucose test from their doctor, but their actual cellular factory floor, their M value is operating healthier than someone labeled normal, who was secretly spiraling towards severe resistance.
9:25Exactly. The clinical labels are just too blunt. In fact, the researchers discovered that the fasting proteom. So, just looking at the baseline state of the muscle hardware before insulin even enters the picture, strongly predicts whole body insulin sensitivity.
9:40Wait, before the insulin even arrived? Yes, it actually outperformed standard clinical tests like HBA1c in predicting how the muscle would eventually handle the insulin delivery. If the baseline hardware is that predictive.
9:51It really makes me wonder about the power grid of the factory. And that leads to their next finding, which addresses a massive debate in the metabolism world, right? Oh, the mitochondria debate, yes. Because I've always heard that type 2 diabetes is caused by having fewer or damaged mitochondria.
10:08Did they find that the diabetic group just had fewer power plants? Well, this is where that continuous spectrum becomes so vital. If they had just used the old binary categories and compared the diabetic box to the healthy box, they would have found no significant difference in total mitochondrial protein.
10:24The debate would have just continued. But because they use the continuous M value scale? Exactly. When mapped against the M value, mitochondrial abundance correlated perfectly with insulin sensitivity.
10:36It is strictly about how sensitive the tissue is, regardless of the patient's official clinical label, but they found something even more granular by looking at the ratio of 2 specific enzymes, LDHA and LDHB.
10:49Okay, I am curious about the mechanism here. How do these 2 enzymes change the power grid? It's all about how the cell chooses to make its energy? LDHA is an enzyme that favors glycolysis, which is creating lactate.
11:00It is a very fast but somewhat dirty way to generate energy without using oxygen. Like a backup generator. Right. LDHB, on the other hand, favors taking those energy precursors and pushing them into the mitochondria, where they are burned cleanly, using oxygen.
11:17That's the oxidative pathway. So it's literally the difference between running your factory on a clean, steady electrical grid versus firing up a bunch of loud, exhaust heavy backup generators. That's a great way to visualize it.
11:30What the data revealed is that a higher ratio of glycolytic proteins, meaning an overabundance of LDHA favoring that quick lactate energy, strongly correlates with insulin resistance. The muscle cell fundamentally rewires its power grid away from the clean burning mitochondria, even if the mitochondria are physically still sitting right there.
11:49That's incredible. So that's the hardware. But here's where it gets really interesting. Let's talk about the software, though, phos relations sites. What did the communication network look like in the fasted state, you know, before the insulin delivery truck even pulled up?
12:02You would naturally assume a resting muscle in a fasted state would be completely quiet. Like the intercom should be off. Right. There's no delivery to process yet. But they found massive chaotic phosphorelation activity.
12:13The resting muscle and insulin resistant individuals was already highly stressed out. Specifically, they saw hyperactivity in a stress response network called the JNK P38 pathway. What is stressing the factory out so much if there's no insulin and no food delivery to process?
12:29Well, the JNKP 38 pathway acts like an alarm bell for cellular stress. In metabolic disease, it's often responding to lipotoxicity, which is high levels of circulating fatty acids, or chronic low grade inflammation.
12:42So this hyperactive alarm system creates a really hostile environment on the factory floor, effectively predetermining the muscle's inability to handle insulin long before the hormone actually arrives.
12:53And deep within this blaring alarm system, they found one specific software switch that seems to control the entire cascade, didn't they? They did. It's a phosphoreation site known as S 65, and it's located on a protein called AMPK Gamma 3.
13:07Now, AMPK is a pretty famous protein. This is the master energy sensor for the whole body, right? It is, but this specific subunit AMPK Gamma 3 is unique. It is only expressed in skeletal muscle, and the phosphoration of the single site, S 65, was the strongest predictor of insulin resistance in the entire study.
13:26I found the evolutionary biology around this specific switch completely mind blowing. The S 65 site is unique to homo sapiens. It literally does not exist in chimpanzees. Our closest evolutionary relatives.
13:38Yeah, it is a distinctly human piece of biological software. But the paper highlighted something equally fascinating about another animal, pigs, or susscropha. In pigs, this exact site on the protein has naturally mutated over evolutionary time.
13:53Really, mutated how? Instead of a serene amino acid at position 65, pigs have an aspartic acid. Wait, I need a biochemistry, explain it like I'm 5 here. How does swapping one amino acid for another, mimic a software switch?
14:05It's all about the electrical charge. A phosphate group carries a heavy negative charge. When it attaches to a serene amino acid, That negative charge physically forces the protein to change its folded shape.
14:15Okay, following so far. Aspartic acid, by its very chemical nature, already carries a strong negative charge. So, by mutating that spot to an asbartic acid, the pig's DNA naturally mimics the exact physical shape and electrical charge of a permanently phosphor related state.
14:32So pigs are walking around with this S 65 metabolic switch permanently glued in the on position. Why would evolution do that? Well, the paper doesn't dive super deeply into porceline evolution, logically, an animal like a pig is evolutionarily adapted for constant foraging and rapid energy storage.
14:50Right, they eat a lot and grow fast. Exactly. Having a metabolic rheostat locked in a state that alters glucose and lipid handling might be really beneficial for that rapid growth. But humans are built for endurance, fasting, and metabolic flexibility.
15:03So we need the switch to be able to turn off. Yes. In a human muscle, when this S 65 switch gets stuck in the on position due to chronic stress and overphosphorelation, it is deeply tied to profound insulin resistance.
15:15And the researchers actually proved this at a petri dish, didn't they? They did. They took human muscle cells. applied a targeted drug to block the JNK stress pathway, and successfully turned down the phosphorelation of the S 65 switch.
15:28The alarm bells finally went quiet. Okay, so we have a stressed out resting state with a stuck human specific switch. But what actually happens when the insulin truck finally hits the loading dock. Does the entire factory just ignore the delivery, or are just specific highways inside the cell shut down?
15:48The data strongly supports the selective shutdown idea. This was their finding regarding preserved versus dysregulated signaling. You see, we typically visualize insulin resistance as a brick wall, like the signal hits the outside of the cell and just dies.
16:00Right, a total blackout. But that is completely inaccurate. Because the mass spec data showed that even in people with the lowest M values. The most severe insulin resistance, some of the classic primary insulin pathways still fired perfectly.
16:13Yes, exactly. When insulin hit, a famous signaling Kines, called AKT, and its immediate downstream target TBC one D4, they lit up exactly the way they do in a perfectly healthy person. The proximal signaling.
16:26The steps closest to the receptor works fine, but the distal signaling, the steps further down the chain of command, were deeply broken. Right. Specifically, substrates of the MTR pathway, like a protein called TNS 2 and another AMPK subunit, AMTK alpha 2, The expected phosphorelation of these targets completely fail.
16:45So to bring back to our factory, the delivery truck arrives, the dock worker radios the floor manager over the intercom, and the floor manager yells the orders perfectly. The AKT pathway works. But the conveyor belt motors further down the assembly line, the MTOR and AMPK Alfred 2 substrates, have frozen software.
17:03They just don't turn on. That is precisely what is happening. Insulin resistance is a highly selective breakdown of specific molecular nodes, not a uniform failure. Whenever we look at large population maps like this.
17:15I have to ask, does this map look the same for everyone? We know there are significant baseline metabolic differences between men and women. Does biological sex change the nature of the factory floor? It does, and providing a comprehensive sex resolved atlas of the muscle was actually a major achievement of this paper.
17:32The base muscle prodium, the baseline hardware configuration, is vastly different between men and women. The data showed females had higher levels of proteins dedicated to lip it uptake and fat storage, right?
17:45Which perfectly aligns with the physiological reality of having higher circulating free fatty acid. Exactly. And males, conversely, showed a much higher abundance of proteins dedicated to glucose metabolism.
17:57And importantly, the researchers noted that very few of these massive hardware differences were driven by genes on the X chromosome. Yeah, that's a key point. They are mostly autosomal meaning driven by non-sex chromosomes and are instead governed by the distinct whole body hormonal environments of men and women.
18:14So men and women have fundamentally different hardware setups. But if the baseline factory is entirely different, does insulin resistance look different? This is perhaps the most elegant finding in the whole study.
18:24Despite these massive differences in the baseline machinery, the molecular transducers of insulin resistance were remarkably identical. So the exact same conveyor belt motors freeze in both sexes. Yes.
18:36The specific signaling pathway is breaking down those selective highway closures we talked about, they occur the exact same way. The underlying mechanism of the disease does not care about the biological sex of the host.
18:47Wow. If we connect this to the bigger picture, it completely reframes our clinical approach. For decades, our primary strategy for type to diabetes has been to use blunt instruments, you know. Give the patient more insulin to force the signal through, or use broad drugs to force blood sugar down systemically.
19:05But if the AKT pathways working fine, and only the MTOR and AMPK gamma 3 pathways are broken, flooding the system with excess insulin isn't fixing the root problem. Exactly. It might even be overstimulating the pathways that are still working perfectly well.
19:20It's like turning up the master volume on a stereo because the base speaker is blown. The base is still dead, but now the trouble is deafening and you're damaging your hearing. That is a perfect analogy.
19:30This data set is the foundation for true precision medicine. Imagine a future where a doctor doesn't just diagnose you with type 2 diabetes. Instead, they know that the specific AMPK gamma 3 S 65 switch is malfunctioning in your muscle.
19:45And because AMPK gamma 3 is uniquely expressed in skeletal muscle, pharmaceutical companies could theoretically design a targeted molecular drug that binds only to that switch, right? Exactly. You could flip that switch and restore insulin sensitivity without causing dangerous off target side effects in the liver, the brain, or the heart.
20:04That is the ultimate promise. Of course, we must acknowledge the limitations of the current study. I mean, this is incredibly complex data set, but it is primarily associational. Yes, that's true. It maps the terrain with incredible resolution, but it doesn't prove every single cause and effect.
20:19Right. And while analyzing 120 people at this depth is an astonishing feat. They are still a specific cohort. It doesn't capture every global phenotype of metabolic disease. You have confounding variables like lifelong dietary patterns, exercise histories, and previous medications that add really dense layers of complexity to these protein signatures.
20:40Furthermore, regarding that tantalizing S 65 switch, robust functional testing is the required next step. We need to perfectly understand the mechanical physics of how that exact phospholation event controls glucose uptake inside a living human muscle before a drug can be safely synthesized.
20:56So what does this all mean? If we synthesize this massive molecular atlas, insulin resistance is not a simple binary disease state where your body either works or it doesn't. It is a highly personalized continuous spectrum.
21:08Absolutely. And while our baseline muscle metabolism differs wildly between sexes and populations, specific targetable signaling nodes, like the unique human-specific AMPK Gamma 3S65, switch hold the key to selectively repairing the cellular factory.
21:23We are finally moving from treating a broad descriptive label to targeting the exact mechanical failure. What does this mean for the future of treating complex diseases? Will we eventually stop diagnosing broad conditions like type 2 diabetes entirely, and instead diagnose the precise molecular traffic jam happening inside your own muscles?
21:43It's a truly fascinating possibility. This episode was based on an open access article under the CCBY4.0 license. You can find a direct link to the paper and the license in our episode description. If you enjoyed this, follow or subscribe in your podcast app and leave a 5 star rating.
22:00If you'd like to support our work, use the donation link in the description. Now stay with us for an original track created especially for this episode and inspired by the article you've just heard about.
22:09Thanks for listening and join us next time as we explore more science, base by base.
