This study used enhanced CRISPR base editor screens, structural biology, biochemical assays, and in vitro/in vivo selection to map MEN1 mutations that drive resistance to five clinical menin inhibitors and to explain their mechanisms.
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. Imagine you are a locksmith, right?
0:11But you're not designing a key for just a standard front door. You are engineering a master key for a highly complex microscopic lock. Right, a lock that is actively fighting back against you. Exactly.
0:24In this scenario, the lock is an aggressive form of leukemia. And you spend years researching the exact mechanism, and eventually, well, you craft the perfect targeted key. It slides right in, the lock turns, and the disease goes into complete remission.
0:37Which is an incredible achievement in oncology. Oh, absolutely. The therapy works exactly as intended, but months later, the cancer shifts the shape of the lock. And this alteration is almost imperceptibly small.
0:48You know, yet it's just enough that your perfect g just falls right out. Yeah, and it's um, it's one of the most persistent hurdles in modern medicine. The therapeutic key hasn't lost its edge at all, but the biological lock is dynamic.
1:01I mean, it evolves under pressure. So what really happens when cancer outsmarts our newest, most promising drugs, and more importantly, how could this change the way we predict and outmaneuver drug resistance before the cancer even makes its next move?
1:15Okay, let's unpack this. We are really looking at a fundamental shift here and how we approach targeted therapies. You know, instead of waiting for resistance to just occur in the clinic, researchers are now developing ways to force that evolution in the lab.
1:29They're basically mapping the cancer's escape routes before it even takes them. Exactly. So today we celebrate the work of Wallace Bourgeois, Hannah E. Rice, Daniela V. Wenge and colleagues, including co-corresponding authors, Giovanni Cutler and Scott A. Armstrong at the Dana Farber Cancer Institute, Boston Children's Hospital, and Harvard Medical School, who have advanced our understanding of how acute leukemias develop resistance to targeted therapies.
1:54Yes, and this deep dive is based on their newly published open access article in Nature Communications. It's an article in press for 2026, and the title is uh, CRISPR base editor screening identifies spectrum of men, one mutations, impacting men and inhibitors in clinical trials.
2:11Right, so to really understand the impact of this screening. I think we 1st need to establish what these specific leukemias are doing you know, at a cellular level. The paper focuses on acute leukemia is driven by KMT2A genery arrangements or MPM1 mutations.
2:27Right. Those 2 specific genetic drivers. Yeah, so what is the normal function of these genes and what happens when they go rogue? Well, in a healthy cell, KMT2A and NPM1 are, um, there are central regulators, KMT2A provides instructions for an enzyme that modifies his stones.
2:44And histones are those protein spools that DNA wraps around, right? Exactly. Yeah. So it essentially helps control which genes are turned on or off during normal blood cell development. And NPM one is a protein that helps manage cell growth and division.
2:57Okay, so they're basically the project managers of the cell. That's a good way to put it. But when these genes undergo mutations or, you know, structural rearrangements, they essentially get stuck in the on positions.
3:08I see. Yeah, they force the white blood cells to multiply just uncontrollably without ever maturing. And that is the absolute hallmark of acute leukemia. And this runaway process relies on a very specific physical partnership, doesn't it?
3:21It does, yeah. The mutating KMT2A protein, it can't drive this cancer program all by itself, it actually requires a physical connection with another protein called Menin. Menin. Right. Menon acts as sort of a structural scaffold.
3:34So when KMT2A binds to that men and scaffold, the whole leukemia machinery is fully assembled and operational. Which makes that connection an incredibly appealing target for drug developers, obviously.
3:45I mean, if you can break that physical link between Venend and Cam T2A, you disassemble the whole machinery. Precisely. And since the 1st small molecules designed to block this interaction were described back in 2012, the field has moved really rapidly.
3:59Wow, so over a decade of work. Yeah, and there are currently 8 different men and inhibitors in clinical trials globally right now. The furthest along is one called SNDX 5613, which is also known as Revumim.
4:10And how is that one doing? Really well, actually. In recent clinical trials, it achieved roughly a 23% complete remission rate in patients with advanced, heavily pretreated leukemia. Wow, 23% in heavily pre-treated patients is massive.
4:24But there is a snag. Patients are relapsing, right? Unfortunately, yes. The cancer is finding workaround. It's acquiring mutations in the MEN1 gene. That's the gene that holds the blueprint for the men in scaffold we talked about.
4:38Right. So leukemia introduces these slight alterations in the binding pocket of the men in protein itself. And because of that shape change, the drug just loses its grip. Exactly. But the terrifying part is that the interaction between Menin and KMT2A, which is the actual engine of the cancer, that remains completely intact.
4:59It's like, think of it like wedging a doorstop under a really heavy door. The doorstop is our drug, right? And it does a flawless job keeping the door from swinging shut. Right. But over time, the door frame itself morphs just a tiny bit.
5:11So the wedge just slips out. The doorstop hasn't failed mechanically. It's not broken, but it just has nothing to grip anymore. So the door swings freely again. That visualizes the mechanical failure really well.
5:21So to overcome this, the researchers basically realized they needed to map out the exact structural changes the door frame could undergo. But how do you even do that? How do you map 1000s of potential protein mutations systematically?
5:34Well, they utilized CRISPR base editor screening. Okay, base editing. Yeah, and this is a huge technological leap from traditional CRISPR Cast 9. You know, traditional cast 9 acts like a pair of molecular scissors that just cuts both strands of the DNA helix to knock out a gene entirely.
5:50Right, just totally breaks the gene so it stops working. Exactly. But here, the team used Spcast 9, Nicases. Nickel fees. Yeah, these editors only nick one strand of the DNA, and they carry an enzyme that chemically alters just a single DNA letter, so they can convert an A to a G or a C to a T without breaking the DNA backbone.
6:11Oh, wow. So instead of deleting the gene, they're literally rewriting it. One single letter at a time. That's exactly it. So they engineered a line of human leukemia cells. They're known as MV4, 11 cells to express these base editors.
6:23And then they introduced a massive library of guide RNAs. And these guides would tell the editors where to go, right? Right. They directed the base editors to meticulously mutate the MEN1 gene across 1000000s of cells, changing just single amino acids in the men in protein.
6:39So they basically created this incredibly diverse population of leukemia cells in a petri dish, each with a slightly different men in the scaffold. Yes. And then they applied the clinical inhibitors to see which cells would survive.
6:50They tested 5 specific clinical men and inhibitors. DS 1594, JNJ 6616, KO539, SNDX 5633, and DSP 5336. Okay, so 5 different drugs. Yeah. And they cultured these mutated cells in the presence of these drugs for 21 days.
7:08That's a long time for a cell culture. It is. And the cells were the drugs still effectively bound to the men in protein? Well, they died. The survivors were the ones harboring a mutation that granted them resistance.
7:18So by sequencing the DNA of the survivors at the end of the 21 days, they had a complete genetic catalog of resistance. Okay, but wait, taking cells in a Petri dish and forcing genetic mutations with a CRISPR editor is, I mean, it's highly controlled.
7:33How did they know a CRISPR screen in a Petri dish actually mimics the chaotic, spontaneous mutations that happen in a human patient taking these drugs. That's the perfect question. And the researchers actually anticipated that exact critique.
7:47So they validated their findings using Invivo systems. Specifically, patient derived xenographed, or PDX, mouse models. Okay, so how do the PDX models work in this context? Well, you take actual human leukemia cells derived directly from a patient and you and graph them into immunocompromise mice.
8:05Once a human leukemia establishes itself in the mouse's bone marrow, you treat the mouse with the men and inhibitors, and over several months, you just observe the natural evolution of the cancer under the pressure of the drug.
8:16Oh I see. So did the human leukemia and the mice develop the same mutations as the cells in the CRISPR screen? They absolutely did. The natural mutations that evolved spontaneously in the mice perfectly mirrored the genetic data from the base editor screen.
8:29Yeah, the lab setting accurately predicted the biological reality. That is incredible. But the team didn't just stop at genetic sequencing, did they? I know they combined this data with, um, TR FareFret, biochemical assays and x-ray crystallography to really understand the physical mechanics of this resistance.
8:46They did. So TR Fart T, which stands for Time Resolve fluorescence resonance energy transfer. Which is a mouthful. It really is. But it's just a technique used to measure the invisible molecular grip between 2 proteins, or, in this case, a protein and a drug, you basically attach a light emitting donor molecule to one side and an acceptor molecule to the other.
9:07And when the drug and the protein bind really tightly together, the energy transfers and emits a specific signal. If a mutation causes the drug to slip even a little, the distance increases, the energy transfer drops, and the signal just fades.
9:22So it quantifies the exact loss of grip strength. Precisely. And then the x-ray crystallography provides the visual blueprint of what's happening. So what did they find when they looked at the baseline data?
9:32Well, initially, against wild type leukemia cells, meaning cells with no mutations at all. All 5 inhibitors demonstrated similar loan anomaler potency. They all bound tightly and worked well. Okay, so the differences only emerge when they looked at the mutated cells from the screen.
9:47Right. And the screening revealed 2 distinct categories of resistance, which is fascinating. There's shared resistance and then inhibitor specific resistance. Let's start with the shared resistance. The study highlights a specific location on the minin protein called the M327 residue.
10:02Yes, the M 327 position. When that specific location mutates, it basically creates a universal shield. It causes pan class resistance, effectively blocking all 5 of the clinical inhibitors tested. All five.
10:14Wow. Yeah. To put the impact in perspective, the TRFret assay showed that a specific mutation at this site, the M327 I mutation, reduced the binding affinity of the drug S and DX 5613 by a staggering 102 fold.
10:28102fold. So the drugs grip is just completely neutralized. But the study also found inhibitor specific resistance, right? Where a mutation blocks one drug, but lisa cancer completely vulnerable to the others.
10:41Yes. Here's where it gets really interesting for you listening. Think of these 5 drugs as 5 different rock climbers, right? Scaling the exact same cliff face. The cliff face is the binding pocket on the men in protein, and the climbers are the 5 drugs.
10:56I like that analogy. Since these 5 drugs have distinct chemical structures. They really do represent climbers with different reaches, different body types, different gripping techniques. They are ascending the same general route, but they depend on slightly different microscopic handholds.
11:11Right. So if a major ledge gives way like that M 327 position we just discussed. All 5 climbers lose their footing in fall. But what happens if a very specific tiny handhold breaks? That is exactly where the specific resistance emerges.
11:25There is a location on the protein designated as E 368. The researchers found a mutation that changes the amino asset at that position. It's called the E 368 k mutation. When that specific handhold breaks.
11:38Only one climber falls. And which crug fails? The drug, DSP 5336. The resistance to that specific drug increased by more than 75 fold, basically rendering it largely ineffective. The others were fine. Yeah, the other 4 drugs were barely affected by this change at all.
11:55They just kept climbing. That is wild. What makes that single handhold so critical for DSP 5336 while the other drugs just ignore it entirely? Well, the x-ray principleography data answers that beautifully.
12:07The E in E 368 stands for glutamic acid, which is an amino acid that carries a negative electrical charge. Okay. And the drug DSP 5336 has a positively charged region designed to align perfectly with that specific spot.
12:21Oh, so they form an ion pair. They lock together like magnets, +and negative. Exactly. But the mutation changes the glutamic acid into Lacine, which is represented by the K, and Lacine carries a positive electrical charge.
12:34Positive meets positive. Right. So the magnetic attraction is broken, and the drug is literally physically repelled from the binding pocket. The other 4 inhibitors in the study. They're chemically structured differently.
12:45They don't rely on that specific magnetic bond to maintain their grip. So they are just indifferent to the charge flipping. And the thing is, the cancer isn't strategizing this. You know, it's just generating random mutations.
12:58And when it accidentally flips that magnetic charge, that one cell survives, it divides, and the resulting tumor is suddenly immune to DSP 5336. Exactly. But here's the brilliant part. If we know exactly which handholds are broken, we don't have to guess the next line of treatment.
13:16We can strategically deploy a drug that uses a different set of handhold. Right, because this maps out a completely clear clinical strategy. I mean, if a patient relapses, clinicians could eventually sequence the tumor, identify the specific structural shift like that E 368 k mutation, and immediately switch the patient to a different minute inhibitor that bypasses that broken handhold.
13:37Exactly. But is it merely a matter of finding an alternative drug that still works? Or do these structural changes actually weaken the cancer against the other options? Well, in some cases, the mutation induces a collateral vulnerability.
13:49Really? Yeah. When the researchers tested the E 368K mutant cells against 2 of the other inhibitors, JNJ 6627 and SNDX 5613. The cells were actually more sensitive to those drugs than the unmutated cancer cells were.
14:04Wow, so the mutation that built an impenetrable shield against one drug compromised the proteins defense against the others. Exactly. It fundamentally alters how we view drug resistance. It isn't always a dead end.
14:15Sometimes it literally exposes a new target. is so cool. Now, I know the study also uncovers some fascinating dynamics regarding dosage, right? In the mouse models, the researchers manipulated the concentration of the drugs to see how the evolutionary pathways responded.
14:29Did the volume of the drug change the type of vacations the cancer developed? It didn't just change the type. It changed whether the mutations could develop at all. When the researchers administered highly potent doses, such as providing feed containing .3% of the SNDX 5613 drug, or administering 100 milligrams per kilogram of JNJ 6617.
14:49They completely suppress the emergence of these resistance mutations. Wait, completely. Completely. The cancer was just overwhelmed before it could successfully mutate and survive. Okay, but if higher doses suppress the mutations entirely in mice, why not just start every human patient on the absolute maximum dose from day one?
15:06Right. It's a logical question. But while that holds true mechanically within the isolated environment of the tumor, clinical oncology involves the entire human system. Ah, right. toxicity. Exactly. Mice metabolized drugs very differently than humans.
15:22A highly concentrated dose that suppresses a tumor in a mouse over a short period could cause severe potentially fatal toxicity to a human patient's liver, kidneys, or their healthy bone marrow. So the ceiling isn't what the cancer can survive.
15:37The ceiling is what the patient can tolerate. Precisely. The maximum tolerated dose in a human is dictated by systemic toxicity, which these structural studies and isolated models just don't account for.
15:47We cannot always rely on sheer volume to outpace the cancer's evolution, which makes mapping these specific microscopic vulnerabilities in absolute necessity. That makes total sense. Now, looking at the methodology broadly, are there limitations to what this base editor screening can actually uncover?
16:05Do they map every single possible move the cancer could make? The screening is remarkably comprehensive, but the technology does have constraints. Crisper base editors require a specific neighboring DNA sequence, which is called a PAM site, to anchor themselves before they can edit a letter.
16:23If the ideal anchor site isn't present near the target amino acid, the editor might create a slightly different mutation than intended. So the molecular pencil can't perfectly trace every single line of the blueprint.
16:35That's a great way to say it. For instance, there is a highly aggressive mutation known as G331 D that has been observed in actual clinical relapses. But because of the editing constraints. The CRISPR screen predominantly produced a variation called G331N instead.
16:50Oh, I see Yeah. So the lab screening alone slightly underestimated the danger of that specific location. It was actually the integration of the PDX mouse models that clarified the true clinical threat of the G331 D mutation.
17:02It really highlights why combining lab editing with living models is so critical. Oh, it does. And furthermore, the scope of this study is entirely confined to MEN1 driven resistance. It catalogs how the specific lock changes shape.
17:16However, cancer is highly adaptive. It could potentially develop an entirely novel escape route that bypasses the Menin and KMT 2A pathway altogether, relying on completely different epigenetic drivers to maintain its growth.
17:29Right. could just build a new door. But still, this research provides a high resolution, atomic level map of how acute leukemia evolves to evade 5 major targeted therapies. I mean, it exposes their shared vulnerabilities, those universal shields, as well as their unique structural weaknesses like that flipped magnetic charge.
17:48It really does. By cataloguing these atomic handholds, science is shifting from merely reacting to drug resistance after a relapse to actually predicting it. What does this mean for the future of personalized oncology?
17:58Could we one day treat cancer like a chess match? always staying one move ahead of its next mutation before it even makes a move. This episode was based on an open access article under the CCBY 4.0 license.
18:12You can find a direct link to the paper and the license in our episode description. If you enjoy this, follow or subscribe in your podcast app and leave a five-star rating. If you'd like to support our work, use the donation link in the description.
18:24Now stay with us for an original track created especially for this episode and inspired by the article you've just heard about. Thanks for listening, and join us next time as we explore more science based by base.
