Schaff et al. use multi-treatment clonal tracing combined with single-cell RNA-seq to show that rare, pre-existing transcriptional states in melanoma predict resistance to diverse therapies and that high CD44 marks cells with multi-treatment tolerance.
0:00Welcome to Base My 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 you are playing a high stakes game of whack-a-mole.
0:12Okay, I can picture that. Right. You're standing there, mallet in hand, just waiting for the mole to pop up. But uh, this isn't your standard arcade game. In this version, the mole somehow knows exactly which mallet you're going to use next.
0:26Like it's a reading your mind or something. Exactly. Maybe you're going to use a standard wooden hammer or, you know, a metal baseball bat. Or maybe you just suddenly switch it up and throw a water balloon.
0:38Right. totally different types of attacks. Yeah. And in a normal game, the mole reacts after being hit. But what if, instead of adapting after the fact, some rare moles are already wearing like a tiny rain jacket or a metal helmet, they're perfectly prepared to survive strikes they have never even seen before.
0:54Oh, wow. Okay, so in the world of oncology, those moles are rare cancer cells. You got it. What really happens when a cancer cell is inherently equipped to dodge an entire arsenal of distinct therapies.
1:07Okay, let's unpack this. Today we celebrate the work of Dylan L. Schaff, Sidney M. Schaffer, and their collaborating teams at the University of Pennsylvania, Johns Hopkins, and other leading institutions, who have advanced our understanding of multitreatment cancer resistance.
1:23Which is such a crucial area of research right now. It really is. They published these findings in the journal Cell genomics in 2026, offering us a totally new lens on how we view tumor evasion. We are moving away from looking at resistance as, you know, a purely reactive process.
1:40Right, and instead seeing it as a pre-existing condition within certain rare cells. Exactly, which represents a massive shift in how we approach treating the disease. Because, I mean, clinically speaking, tumors are just famously stubborn.
1:52Oh, absolutely. If you know anyone who has navigated cancer treatment, you know that patients are often given complex sequences of therapies or, you know, combinations of different toxic drugs to try and cover all the bases.
2:03Throwing everything at the wall to see what sticks. Right. But tragically, some patients still develop resistance to absolutely all of them. The cancer simply keeps, well, it just keeps finding a way to survive.
2:14And historically, the scientific paradigm has been to assume this resistance comes primarily from genetic mutations. Like a physical change in the DNA code. Right. The prevailing thought was that a cell acquires a random typo in its DNA that miraculously allows it to survive the drug.
2:33Or, alternatively, we thought cells dynamically adapted their intermolecular workings, you know, under the immense stress of the treatment itself. So they changed during the attack. Exactly. But recently, a nongenetic angle has emerged.
2:46The scientific community is realizing that differences in gene expression, meaning which specific genes are turned on are often a cell at any given moment can actually dictate resistance. Totally independent of any underlying change to the DNA code itself.
3:00Yes. But trying to study that initial gene expression creates a massive methodological paradox for researchers. Oh, totally, because if you want to know what a cell looked like before it became resistant, you have to examine it before you treat it.
3:13But the very act of dropping toxic chemotherapy drugs onto a cell drastically alters its gene expression. It changes the very thing you're trying to observe. Yeah, it's like it's like trying to investigate a crime scene to figure out how a fire started while the building is actively burning down around you.
3:31That's a great way to put it. The evidence you need is literally going up in smoke. Exactly. This raises an important question. How can we look back in time to see what a surviving cell look like before the treatment? Because standard molecular analysis destroys that initial pre-treatment baseline.
3:48Exactly. But the researchers came up with a brilliant workaround for this deep dive. They combined a technique called clonal tracing with single cell RNA sequencing. Or CRNA sec for short. Right. And we can think of this like giving a unique microscopic tracking number to 1000s of identical twins.
4:07Okay, I like this analogy. You tag them, you send them off to different extreme survival camps, and then you just wait to see who makes it out alive. And when you check the tracking numbers of the survivors, You know exactly which original family they came from.
4:20Precisely. And to actually get those tracking numbers into the cells. They use the lentivirus. Ah, right. A lentiviral approach is crucial here. It is. Alentivirus is essentially a virus that has been hollowed out and turned into, well, a microscopic delivery truck.
4:37So instead of delivering a disease payload, They packed it with unique DNA barcodes. Exactly. When they infect the cells, the virus inserts that unique barcode directly into the cell's genome. And what makes this so powerful is that every time that cell divides, it copies the barcode.
4:53So it passes the tracking number down to its descendants. Right. And the researchers introduce this barcode library into a specific type of melanona cell line, the WM989 V600 EBRAF line. And they did this on an absolutely massive scale, didn't they?
5:08Oh, yeah. huge. I was looking at the methodology, and they used fluorescence activated cell sorting to isolate exactly 350,000 uniquely bar coded cells. It's incredible volume. Right. And from there, they let those cells multiply, expanding them for about 6 doublings.
5:25So one barcode itself becomes 2 to become four. Until you have a whole microscopic colony sharing the exact same tracking number. Yeah, and that expansion yielded a population of about 23000000 cells. Right.
5:38And because they share a barcode, we know for a fact they share a lineage. They are clonal copies of each other. So what did they do with those 23000000 cells? Well, they took this massive population and divided it.
5:49They kept an untreated control group, and then split the rest across 12 different treatment arms? 12 Yeah. This consisted of 2 replicates for 6 highly diverse, extremely harsh treatments. These were the survival camps you mentioned.
6:02Man, and to really test these cells, they didn't just use one type of chemical, they threw the absolute worst case biological scenarios at them. It really did. First, they used targeted clinical inhibitors.
6:14Right. They use Dabrafenib, which specifically targets a mutation in the BRAF protein. And Trementinib, which inhibits the MEK protein. And both of these drugs are designed to physically block a specific cellular signaling highway, right?
6:29The one that melanoma relies on to grow uncontrollably. Exactly. But they didn't stop there. Following those targeted inhibitors, they introduced biological selective stressors. To kind of chemically mimic the suffocating toxic environment found deep inside a solid tumor in the human body.
6:44Right, because as tumors grow, their blood supply often just can't keep up. So to simulate this, the team used cobalt chloride, or coCL2, to chemically induce hypoxia. Essentially starving the cells of oxygen.
6:57Yes, and they also use highly acidic media to simulate extracellular acidosis, pushing the fell's delicate pH balance right to the brink. And finally, they brought in the heavy chemotherapeutics. The really toxic stuff.
7:10Yeah, they used cisplatin, which literally binds to and causes physical breaks in the cell's DNA strands. Right. And Doxarubicin, which inhibits an enzyme called poisomerase. And if toosomerase is blocked, the cell cannot unwind its DNA to replicate.
7:26Meaning it physically cannot divide. So that is 6 totally different mechanisms of attack. You have targeted signaling blocks, oxygen starvation, acid baths, and DNA shredders. Which brings us to the RNA time machine.
7:40Oh, this is the best part. Right. So before exposing the cells to these 6 harsh conditions. The researchers took a sample of the untreated cells and ran single cell RNA sequencing. And this technique captures all the Messenger RNA cells currently reading.
7:54Exactly. It gives the researchers a complete snapshot of the molecular blueprint showing exactly which genes are turned on and off for every single barcode family before any stress was ever applied. And then, what, a whole month later, they sequence the barcodes of the rare cells that actually survived those survival camps.
8:10You got it. By matching the barcodes of the final hardened survivors back to that pre-treatment RNA snapshot. They could see exactly what those specific cells were doing differently before the chemical warfare even started.
8:23So cool. So what did the data actually show? Well, the 1st major observation from mapping those barcodes was that resistance is highly heritable over those 6 doublings. Meaning if a clone survived in one replicate of a drug, its identical cousins almost always survived in the second replica.
8:40Right. But looking at how they survived across the different drugs, revealed a really striking divide between what we can call specialists and generalists. Okay, tell me about the specialists first. So about 20 to 40% of the top resistant clones were specialists.
8:56This means they possessed a pre-existing gene expression state that allowed them to survive one and only one specific type of treatment. Which is exactly what we have traditionally expected in cancer biology, right?
9:08Right. But the generalists completely flipped the script on how we view tumor resilience. Out of 100s of 1000s of original clones, the team found 11 extremely rare clones that were captured in the top 10% of survivors across all 6 treatments.
9:23Wait, all six? All six. These specific cells were inherently multitreatment resistant. So, they possessed an internal state that made them virtually invincible to targeted therapy, chemotherapy, oxygen deprivation, and acid all at once.
9:37Exactly. And tracing those 11 generalist barcodes back to the pretreatment data revealed heavily overlapping gene expression signatures. So they shared a specific profile. Yes. Two genes in particular, CD 44 and FN1, were highly expressed in the clones that later went on to resist Daberfinib, Trematinib, and the hypoxy mimic CoCL2.
10:00Okay, and FN one stands for fiber nectin one, right? Right. It is a protein that sells used to interact with the extracellular matrix around them. It essentially helps them anchor to their environment and communicate with neighboring cells.
10:11So the untreated cells destined to survive had already dialed up the volume on these specific genes. They had. Wait, so it's having high CD 44, a magical impenetrable shield against cancer drugs, or is it just a red flag indicating the cell happens to be tough?
10:25That is the $1000000 question. Because there must be some kind of biological cost to maintaining that super survivor state, right? Or else evolution would just ensure every single cancer cell look like that.
10:36Exactly. The researchers ask that exact question about causation versus correlation. They actually ran a validation experiment. where they physically sorted the melanoma cells into 2 distinct groups. Based on the CD 44 express.
10:51Yes, those naturally high-end CD 44 and those low in CD 44. And as the clonal tracing predicted, the CD 44 high cells were vastly more resistant to the subsequent treatments. But they needed to test if CD 44 was the shield itself, right?
11:07Right. So they used a peptide inhibitor called Angstrom 6. This inhibitor is designed to block CD 44 from interacting with its normal binding partners on the outside of the cell. Okay, so if CD 44 was the functional shield.
11:18Blocking it should have immediately sensitized the cells to the drug, de Brefanib. Precisely. Let me guess. Taking away the CB44 interaction didn't stop them from surviving. You guessed it. The Engstrom 6 inhibitor did not consistently sensitize them to the treatment, which strongly indicates that CD 44 is acting more as a prominent marker of a resistant state, rather than being the direct physical mechanism of the shield.
11:41It is the red flag on the tough cell. Exactly. And the biological cost you mentioned earlier, likely comes from the massive energy requirements of the actual mechanism keeping them alive, which turned out to be hyperactive lysosomes.
11:54Lysosomes? Oh, I remember from biology class that those are essentially the cellular garbage disposals. That's a perfect description. They're like small compartments inside the cell filled with highly acidic enzymes that just break down waste and cellular debris.
12:07Right. And the CD 44 high cells had highly elevated expression of lysosomal pathway genes, specifically genes like SRGN and VMP1. Okay, what do those do? Well, as RGN produces a protein that helps package materials inside these acidic compartments.
12:23And VMP one is heavily involved in forming the cellular vesicles that transport waste to the lysosomes. So they basically built a more robust trash transport system. Yeah, and to prove it, they use a special dye called Lysotracker that literally lights up under a microscope in the presence of active Lysosomes.
12:42Oh, that's clever. Right. And they proved the CD 44 high cells had hyperactive garbage disposals even before any treatment was applied. That is wild. So the theory is that the cells are surviving because their disposal systems are just in overdrive.
12:58Exactly. They are actively sequestering the toxic cancer drugs. Pulling them into these highly acidic lysosome compartments and then degrading the chemical structure of the drug before it can ever reach its target inside the cell.
13:11They're essentially swallowing the poison and digesting it into harmless waste. That perfectly explains the generalist super state, but, uh, does a cell facing dabrafendim have to use the specific garbage disposal state to live?
13:23Here's where it gets really interesting. Because biology always finds multiple paths to survival. Always. The researchers wanted to see if totally different starting states could lead to the exact same destination of survival.
13:34So how did they test that? To analyze this, they utilized a complex computational technique called consensus non-negative matrix factorization, or CNMF, clustering. Okay, let's break that down because CNMF sounds incredibly dense.
13:50It is a bit of a mouthful. Is it accurate to think of CNMF like looking at the listening habits of 1000000s of Spotify users? How do you mean? Like instead of just looking at one single song a user plays, the algorithm finds hidden playlists?
14:05It identifies coordinated programs of gene expression, 100s of genes that always seem to activate together in the background? That is a brilliant way to conceptualize it. Yes. By finding those hidden genetic playlists, the algorithm revealed that cells starting in totally different initial states could become resistant to the exact same drug, Deborah Finnib by taking entirely different molecular paths.
14:25Wow, okay. What were the path? They found 2 main starting states. First, there were what they called differentiated clones. Okay. These cells had high expression of typical Milan acidic markers, like M-LA-N, which is a protein involved in Milanum production, and MITF, which acts as a master regulator of Milanocyte development.
14:43So before treatment, these just look like standard run of the mill melanoma cells. Exactly. And how did those standard looking cells managed to survive a targeted drug like Debrethinib, which is literally designed to block their main BRAF growth pathway?
14:58They survived by upregulating a totally different signaling pathway known as KRAS. Ah, okay. By activating the KRAS pathway, the cell essentially sends an alternative keep growing signal to the nucleus.
15:11This completely bypasses the blocked BRAS proteins. It is a biological detour around the drugs roadblock. Perfect way to describe it. But then you have the 2nd group, right? The Messenciml clones. Yes. These are the cells that started out high in our red flag markers, CD 44 and FN1.
15:27They obviously didn't take the KRS detour. No, they survived by turning up the volume on entirely different programs. Specifically, they upregulated epithelial mesenchymel transition or EMT pathways. Which is a fascinating process where a cell fundamentally changes its shape, right?
15:42It loses its rigid adhesion to neighboring cells and becomes much more mobile and resilient to external stress. Exactly. Yeah, and alongside EMT, they also upregulated oxidative phosphoration. Okay. And that's a highly efficient way for cells to generate large amounts of energy using oxygen inside their mitochondria.
16:00So you have 2 different types of cells facing the exact same targeted threat, and they both survive, but by pulling entirely different molecular levers inside their cellular machinery. One uses a signaling detour and the other sheep shifts and cranks of its energy production.
16:16But did I remember reading the population data in the deep dive sources? Wait, a 33 fold expansion compared to a one. 6 fold expansion. Yes. That means the Mesenchymal clones aren't just surviving the targeted therapy.
16:29They are aggressively thriving in it. The disparity in their survival efficiency is really stark. During the prolonged treatment period. The population of the mess and chimo clones expanded by 33.2 times.
16:40That's huge. Meanwhile, the differentiated clones, you know, the ones relying on the KRAS signaling detour. They only managed to increase their population by one. 6 times. Wow, so one group is grudgingly holding on, barrely surviving the chemical assault, while the other group is multiplying exponentially while bathed in toxic drugs.
17:00Exactly. If we connect this to the bigger picture. Yeah, let's do that. The clinical implications of this are profound. If we know that tumors naturally harbor these rare pre-existing generalists that are capable of surviving almost anything we throw at them.
17:15And if we know they use distinct transcriptional pathways, like hyperactive lysosomes or EMT, to achieve that survival. Then we can completely change how we treat the disease, instead of just reacting to resistance after a tumor starts growing again, we can extract these targetable gene expression states to eliminate multi-treatment resistance before it even starts.
17:37Because if you sequence a patient's tumor and see it has a high population of these CD 44 high cells with hyperactive garbage disposals, you wouldn't just give them a standard targeted drug that is inevitably going to get shoot up and digested.
17:49Right, it would be pointless. You would need to proactively target the garbage disposal system itself, or, you know, target the specific oxidative phosphorelation energy pathways they rely on, alongside the traditional therapy, take away their shield first.
18:04Yes. And the methodology itself, using Leniviral clonal tracing combined with single cell RNA sequencing, could be deployed far beyond just these 6 drugs. Oh, totally. We could use this exact framework to study resistance against cutting edge immunotherapies, like checkpoint inhibitors.
18:21Or cancer vaccines or even card T cell therapy. We could look at how well different pre-existing cellular states absorb targeted drug delivery systems, like nano carriers. The applications are really limited only by the treatments we wish to test.
18:34It feels like we are finally getting a look at the opponent's playbook before the game even begins, but, um, we absolutely need to responsibly outline the boundaries of this specific study. Of course. This is incredibly exciting paradigm shifting data, but it was all done in vitro, meaning in a plastic petri dish in a laboratory.
18:52And it relied on a single established melanoma cell line. The WM 9889 line. A vital point to make. Well, the researchers went to great lengths to use things like cobalt chloride in acidic media to mimic the severe stress of a tumor, a plastic Petri dish severely lacks the complexity of a true in vivo tumor micro environment.
19:13Right, because inside a human body, a solid tumor is a chaotic three-dimensional ecosystem. Exactly. There are constant dynamic interactions with the patients attacking immune cells. There are structural stromal components, like colligan and blood vessels.
19:28And there are true shifting metabolic gradients where oxygen and asset levels change millimeter by millimeter. Right. So a cell that looks and acts like a super survivor, when it is sitting flat in a plastic dish, might behave very differently when it is surrounded by a swarm of attacking key cells, fluctuating blood supply and physical tissue barriers.
19:46The plastic dish is a highly controlled environment, which is strictly necessary to isolate these specific genetic variables and prove causation. But it isn't the whole picture of human disease. Future studies must use patient derived tumor models and complex and vivo systems to validate these markers.
20:03We need to see if CD 44 access the same reliable red flag in a living breathing organism as it does in the lab. Exactly. But what's fascinating here is that despite those in vitral limitations, the underlying ability to look back in time and definitively link a cell's initial molecular state to its final fate across multiple treatments is a massive leap forward for oncology.
20:26It really is. To summarize the findings, rare cancer cells possess pre-existing gene expression states, marked by genes like CD 44 that grant them generalist resistance to multiple diverse treatments simultaneously.
20:39Furthermore, by mapping a cell's initial transcriptional state, we can predict the divergent molecular pathways it will take to survive, proving that tumors possess multiple blueprints for evasion. What does this mean for the future of personalized medicine when we can read a tumor's generalist playbook before administering the very 1st dose of therapy?
20:58That is the big question. 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. If you enjoyed this, follow, or subscribe in your podcast app, and leave a 5 star rating.
21:13If 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.
21:23Thanks for listening and join us next time as we explore more science base by base.
