A data-driven mapping of how 46 human chemokines and 23 GPCRs encode selective and promiscuous interactions. The team defines conserved, semi-conserved and variable determinants, identifies SLiMs in unstructured regions, and uses these rules to rewire a viral chemokine.
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 I want you to imagine for a 2nd.
0:11The sheer scale of the logistical nightmare that's happening inside your body right now. Oh, it's a lot of traffic. Yeah, exactly. You have, you know, 1000000000s of immune cells zooming through your bloodstream in your tissues.
0:22And to keep them from causing this massive biological pileup, your body relies on a GPS system made up of just 46 different molecular messengers, which are called chema kinds and 23 receptors. Right, which is remarkably small.
0:37Right. Right. That is a tiny amount of bandwidth to coordinate all that chaotic traffic. And when this system works, you know, it guides cells to heal wounds and it even helps your brain develop. Absolutely.
0:47But this very same communication network gets hijacked. It gets used by cancer cells to metastasize and by viruses like HIV and COVID-19 to, you know, invade your system. Yeah, it really is a double-edged sword.
1:00It makes you wonder, you know? What really happens when this communication network gets hacked and how could cracking its underlying code change how we treat disease entirely? Today, we celebrate the work of the team, led by Andrew B. Kleist, Brian F. Folkman, M. Medin Babu, and their collaborators at institutions, including the medical college of Wisconsin, St. Jude Children's Research Hospital, and the MRC Laboratory of Molecular Biology, who have advanced our understanding of the human Chemikine GPCR interaction network.
1:30Okay, let's unpack this because the core puzzle here is genuinely mind-bending when you think about it structurally. really is. We have these 46 chemicine ligands just floating around your tissues, right?
1:42Trying to find their match among the 23 GPCRs on the surfaces of your cells. And almost all of them share these highly conserved, practically identical structural folds. Yeah, if you look at them under a molecular microscope.
1:54They look incredibly similar. Exactly. Yet, somehow, they form this insanely complex network of over 100 validated interactions. I mean, some of them are strictly amogamous, like a highly selective one-to-one match.
2:07Right. And then others are wildly promiscuous, basically binding to almost anything they bump into. It's like, well, it's like having a giant bowl of keys and a massive wall of locks that all look identical to the naked eye.
2:20That's a great way to picture it. Yet somehow, the right keys usually find the right locks without jamming the entire cellular network. It is a remarkable feat of molecular engineering, but it's actually the paradox of those identical structures that makes this so fascinating.
2:36How so? Well, if we connect this to the bigger picture. This leg and receptor system is the fundamental mechanism for cellular migration. It's not just about, you know, guiding white blood cells to a paper cut.
2:48Right, it's bigger than that. much bigger. This system drives cerebellar development in the brain. It builds networks of blood vessels in your gut. It, uh, it regulates hematopoasis. The stakes for the organism are monumental.
3:01So the margin for error is basically zero. Exactly. If the wrong key slots into the wrong lock, the resulting misdirection of cellular traffic leads to severe autoimmune diseases, bleeding disorders, or, uh, it provides the exact chemical highway a tumor needs to spread to another organ.
3:18Wow. But mapping that out. I mean, with over 100 interactions and molecules that literally shift their physical shapes when they bind, how do you even begin to find the structural rules? It's definitely a challenge.
3:31Because you can't just take a snapshot of one protein and call it a day, right? The sheer volume of data you'd have to sift through to find tiny structural differences between identical looking proteins just sounds like an absolute nightmare.
3:43Oh, it is a nightmare, which is why previous attempts only gave us, you know, pieces of the puzzle. The scale of the data wrangling in this deep dive is really what sets it apart. Yeah. Yeah. They realize early on that looking at human proteins in a vacuum just wouldn't give them the answers. To understand what structural elements truly drive selectivity, they had to look at evolutionary history.
4:03Okay, so going back in time, essentially. Precisely. The team built a master alignment of over a 1000,58 to be exact chemachine sequences and 951 receptor sequences. And this was across more than 60 different species.
4:20Wait, so they're looking at a chemicine in a human and comparing it to the equivalent in, like, a mouse, a zebra fish, a chicken? Yes. Just trying to see what Mother Nature kept and what she discarded over time.
4:32Exactly that. If an amino acid at a specific position hasn't changed in 400 million years of evolution across 60 species. That tells you it is structurally load bearing. Ah, okay. Like a pillar in a house.
4:44Right. It's fundamental to the protein's core identity. But if a region is mutating rapidly between species, well, that is where the evolutionary arms race of selectivity is happening. Wait, I'm trying to picture how you even map that mathematically.
4:57You're comparing 60 species, but these proteins aren't all the exact same length, right? They are definitely not. Yeah, like a zebrafish protein might have an extra loop or a longer tail than a human one.
5:07How do you actually compare them without the alignment shifting completely out of phase? That's the $1000000 question. It sounds like trying to translate 60 different languages into a single Rosetta Stone when none of the sentences even had the same number of words.
5:20Yeah, that alignment problem was actually the very 1st major technical hurdle they had to solve, because the loops and tails vary in length, comparing them linearly is useless. So what do they do? So the researchers created a completely new, standardized topological framework.
5:38They called it the common chemachine numbering, or CCN, and the chemachine receptor numbering, or CRN system. Meaning they anchored into the 3D shape rather than just, you know, the linear string of letters.
5:51Yes, exactly. It allowed them to compare structural apples to apples, regardless of the sequence length. Well that's clever. It really is. If they wanted to investigate position 42 on one specific human chemicine, this system mathematically guaranteed they were looking at the exact structural equivalent position on every other chemicine across all 60 species.
6:12Wow. Yeah, and once they established that common language, they didn't just look at sequences. They analyzed 16 distinct, high resolution 3D structures of these chemicine DPCR complexes actively binding.
6:26So they basically caught them in the act of the handshake. Exactly. Using techniques like x-ray crystallography and cryo EM. They mapped out the physical geometry of the interactions, identifying 953 specific residue to residue contacts.
6:40953. Yes, down to the Angstrom level. And then uh, to bridge the gap between those abstract 3D models and real human biology. They integrated massive human genome variant databases, like Nomen D and genet TS.
6:54Okay, so looking at real people. Right. They wanted to see how natural everyday genetic variations in the human population affect these exact contact points. You know, here's where it gets really interesting for me.
7:04Looking at how they synthesize all this data, the whole system felt much less like a simple biological lock and key and much more like digital cryptography. Oh, I like that comparison. Yeah, like they were decoding a multifactor authentication system or something.
7:19I'd say that analogy holds up really well, provided we recognize that, you know, biological cryptography is physical, not mathematical. The classic lock and key metaphor kind of implies a rigid, solid state interaction.
7:34But what the researchers uncovered is this highly sophisticated layered structural encryption. The code that determines whether a chemicine will successfully bind to a receptor, is broken down into 3 distinct independent layers.
7:48Okay, let's break those down. The 1st layer is what they call the generalized determinants. So sticking with our encryption analogy, These would be the public keys. Right. The public keys are highly conserved, rigid, dysylside rich hotspots, specifically relying on Sistines.
8:03Sistines, okay. Yeah, these amino acids form strong codalent bonds that essentially hold the core 3D structure of the protein together. They don't tell the receptor the specific identity of the chemo kind.
8:15They just prove it's a chemakai. Exactly. They just provide a baseline molecular verification. They essentially say, yes, my architecture confirms I am a valid chema kind GPCR, and you are authorized to initiate a surface level handshake.
8:29So the public key gets you through the front door of the high security building, the guard recognizes you work there, but, you know, you still need to find your specific department. Which brings us to the 2nd layer.
8:40Subfamily specific determinants. We can think of these as the semi-private keys. The semi-private keys. Yes. Most human chema kinds fall into either the CC or CXC subfamilies, and the researchers identified distinct structural sensors that dictate which subfamily of protein belongs to.
8:57And it all comes down to the physical geometry of specific amino acids. For instance, CC chema kinds heavily rely on a proline amino acid at a specific structural junction. In the CXC ones. They rely on a glutemic acid.
9:10And what is swapping those 2 actually due to the physical shape? Like, why does it matter? It matters because of how they bend the protein backbone. Proline has a very rigid cyclical ring structure. When you put a prolene in a protein chain, it physically introduces a hard kink.
9:28Oh, so bends it. It forces the whole molecule to angle sharply in one direction. Glutamic acid on the other hand, is a longer, more flexible chain that allows the protein to swing the other way. This difference creates what the authors call an intrinsic handedness.
9:44It physically orients the molecules. Because of that proline kink, a CC chema kind perfectly slots into the binding pocket of a CC receptor. But if it tries the other one. Right. If it tries to bind a CXC receptor, the angle is completely wrong and severe molecular clashing happens.
10:00Okay, I love that. It's literally like a USB drive. You have the right port, but if the intrinsic handedness is flipped the wrong way no matter how hard you push, it just won't slide in. Exactly like a USB drive.
10:11So we've gotten in the building with the public key. The semi-private key, put us in the right hallway. Yeah. What is the final layer that dictates the exact specific match? The 3rd layer consists of network specific determinants or the private keys.
10:26This is where the true fine tuned selectivity happens. Real password. Yes. And the brilliance of this system is that these private keys aren't found in the rigid, structured core of the proteins, they are located in the highly unstructured regions.
10:40Yeah, the flexible, floppy tails called N termini, and the extracellular loops between the rigid protein domains. These regions are densely packed with slims or short linear motifs. Swims. That sounds like a 90s hacker handle.
10:54So these floppy slim tails are acting like the final private passcode. They are. And the reason they need to be floppy and unstructured is honestly evolutionary brilliance, because they aren't load bearing, you know, because they aren't holding the 3D shape, the protein together, these sequences can mutate and evolve rapidly without breaking the entire molecule.
11:12Oh that makes sense. Yeah, these slims are highly variable hot spots. The precise sequence of amino acids on these flexible tails dictates the exact selectivity of who binds to whom. Oh, it's a combination.
11:26Right. So it's the combinatorial code of the rigid public keys, the angled semi-private keys, and these rapidly evolving, floppy private keys that ultimately permits the cellular interaction. Okay, I want to push back on something here, or at least add a layer to it.
11:41What's really wild to me isn't just that these proteins are searching for a match. The analysis makes it clear they're also actively avoiding the wrong matches. Yes, exactly. It's not just a lack of attraction, right?
11:53is active repulsion. You're touching on the concept of negative design, which is arguably one of the most critical findings here. Nature doesn't just rely on thermodynamic attraction. It deliberately engineers molecular roadblocks.
12:05Roadblocks? Yeah. The researchers found multiple instances where receptors place massive bulky amino acids like tryptophan directly in the binding path to actively block incorrect pairings. So if the wrong chemicind floats by and tries to dock, it literally hits a physical wall.
12:21Exactly. It's called steric hindrance. If a non-cognate chemicine tries to bind, its own atoms clash with that bulky residue. Wow. It's a violent, energetic rejection, akin to trying to force 2 heavy duty magnets together by their positive polls.
12:37It ensures that even if a promiscuous chemicine has the right public and semi-private keys, the negative design actively vetoes the interaction. That is brilliant. It makes the traffic system practically foolproof.
12:49But uh, let's bring this down from the theoretical 3D models to actual human beings. We mentioned earlier that they integrated human genomic data like Noah May. We did. So how do these private keys and Flocky Slims actually play out in a living person's biology?
13:05This is where the structural biology directly intersects with human diversity. The researchers focused heavily on a specific receptor called ACKR1. ACKR one, okay? In the global human population, there's a very common genetic variant located at position 42, which sits right on this receptor's floppy end terminus right in the center of its private key.
13:24Okay, so right in the password zone. Exactly. Some people carry the amino acid glycine at this position. Well others have a spartic acid. Glycine to aspartic acid, a single amino acid change out of hundreds.
13:36It sounds totally negligible, right. But it radically rewrites the private key because of the chemistry involved. How so? Well, glycine is tiny and neutral. It's basically just a single hydrogen atom on its side chain.
13:49A Spartic acid, on the other hand, is bulky and carries a strong negative electrical charge. Oh, that's a big difference. Yeah. When you swap them, you fundamentally alter the electrostatic landscape of that floppy tail.
14:00This changes exactly how sticky the ACKR one receptor is. And what happens when the stickiness changes, like practically. ACR1 is heavily present on red blood cells. If that private key mutation makes the receptor super sticky, it pulls chemicines out of the bloodstream, essentially trapping them.
14:19But if it's less sticky, those chemicines flush freely into the surrounding tissue. This completely shifts the chemical gradient that tells white blood cells where to go. And crucially, this exact variant is the driver of the Duffin Alfino type.
14:32The Duffy Milfino type. Wait, that's the one historically known for providing natural resistance to certain types of malaria, right? That's the one. But what this deep dive highlights is that this altered binding directly drives the variations in white blood cell counts we see in populations with this phenotype.
14:49One single atomic edit to a private key slim alters the entire immune trafficking profile for 1000000s of people. So what does this all mean? We've essentially cracked the encryption. We know how the rigid public keys work.
15:04We know how the proline kinks, the semi-private keys. We understand the electrostatic charges of the private keys. And, you know, we see the negative design blocking the wrong matches. Does that mean we can stop just observing this system and start rewriting the code ourselves?
15:19That is exactly the proof of concept the researchers set out to demonstrate. They wanted to test if they could act as molecular hacker. Hack the system. To do it, they took a protein called VMIP2. This is a highly evolved viral chemachine produced by human herpes virus 8.
15:36In nature, VMIP 2 acts as a biological master key. It is wildly promiscuous and combined to a huge variety of human receptors to hijack our immune system, essentially allowing the virus to evade detection and spread.
15:49Wait, I need to pause you there. Herpes virus 8 has spent 1000000s of years evolving this perfect, promiscuous master key. It's the guy with the skeleton key who can break into any apartment in the building.
16:02Are you telling me researchers just walked into a lab, swapped a few amino acids and successfully reengineered it without the whole protein folding in on itself and falling apart? Because usually when we miss with viral proteins, we break them.
16:16We do, but that is exactly why understanding the layered cryptography was so essential because they knew the rigid core, the public keys, held the protein together. They knew to leave those alone. Oh, so they only touch the password.
16:28Exactly. They only targeted the unstructured floppy tails. They specifically mutated residues within the private key regions of VMIP 2 guided by their new rule book, and it worked beautifully. They successfully altered its selectivity.
16:40They did. They essentially filed down the teeth on the virus's skeleton key. They transformed this promiscuous master key into a highly selective tool that strongly binds to just a few specific receptors like CCR 5 or CCR 3, while entirely losing its ability to unlock any of the others.
16:59Wow, they preserve the structural scaffold, but completely rewrote the password. That is incredible. It's a monumental proof of concept. If we connect this to the bigger picture, it proves we are firmly entering the era of network editing.
17:14Think about what we could do with synthetic chemicines designed entirely from scratch. Like what? Take carte cell therapy, for instance, where we engineer a patient's immune cells to fight cancer. Currently, one of the biggest bottlenecks is getting those engineered T cells to actually traffic directly to the solid tumor.
17:31Right. Because right now we infuse them into the bloodstream and essentially cross our fingers, hoping they stumble upon the tumor before they get exhausted. Exactly. They get lost in the noise of the body's natural traffic system.
17:42But with network editing, we could use bespoke chemicines as programmable logic gates. Okay explain that. We could engineer a highly specific synthetic receptor onto the Sire T cell, and then design a synthetic chemicine that only binds to that specific receptor.
17:57We could essentially paint the tumor with this custom GPS beacon, guiding the engineered immune cells exactly where they need to go. Basically bypassing the body's natural traffic jams entirely. Exactly.
18:09That sounds like pure science fiction, but it's physically grounded in these structural alignments. I do have to ask, though, biology in a living organism is rarely as clean as a 3D computer model. Are there limitations to this rule book?
18:23There are significant complexities yet to be mapped, and the researchers are very transparent about that. This study decoded the primary structural sequences of the proteins themselves, but it didn't fully factor in tissue specific realities.
18:36Like what kind of realities? For instance, it doesn't account for gags, glycoseminoglycans? Yeah, it is. These are complex sugar chains coating the outside of your tissues. Chemo kinds actually stick to these shutters to form physical breadcrumb trails for cells to follow.
18:51Ah, so the environment, the key is operating in matters just as much as the key itself. Precisely. The structural map also doesn't fully account for receptor oligommerization. Oligomerization, meaning when they clump together.
19:04Yes, where multiple receptors clump together on the cell surface, which can radically change their binding affinity and how they process signals. Right, the dynamics change. Factoring in those 3D tissue realities represents the next great frontier of this research.
19:17But make no mistake. The fundamental baseline encryption rules, those have been conclusively cracked. Let's bring this all together for our listeners. Selectivity in the Chemo Kind GPCR network isn't just a simple, solid state lock and key mechanism.
19:31It is a brilliant, layered cryptographic system that relies on a precise mix of rigid structural scaffolds to get in the door, and highly flexible, rapidly evolving loops to find the exact match. By decoding these public and private keys, and understanding the physical roadblocks of negative design, science can now engineer bespoke molecular signposts to control cellular traffic.
19:54What does this mean for the future of custom built immune systems? It means we are crossing a vital threshold. We are no longer just passive observers trying to map the immune system. We are acquiring the root level access required to write its software.
20:08It really leaves you with a mind expanding thought as we wrap up this deep dive. If we can successfully edit this network to guide our own immune cells with pinpoint accuracy, could we eventually design synthetic decoy chemicines?
20:22Could we completely rewire a pathogen's GPS, sending viruses like HIV into cellular dead ends where they simply can't replicate, effectively using their own ancient trick of viral mimicry against them?
20:35That's the $10000 question for the future. This episode was based on an open access article under the CCBY 4.0 license. You can find a direct link to the paper and the license in our episode description.
20:47If you enjoy this, follow or subscribe in your podcast app and leave a 5 star rating. If 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 just heard about.
21:00Thanks for listening and join us next time as we explore more science base by base.
