A multi-cohort neuroimaging and genetics study (n=1,354) used PET and causal path modelling to test how sex, APOE-ε4 dosage and TREM2 rare variants influence stages of the canonical amyloid→tau cascade, focusing on entorhinal (EC) and neocortical (MetaTemp) tau.
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 2 patients.
0:11They like the exact same age. They have identical lifestyles, and their brain scans show the exact same buildup of amyloid plaques. You know, those sticky protein fragments we've always been taught are the primary hallmark of Alzheimer's disease.
0:24Yeah, the classic warning sign. Exactly. So based on the scans alone, you would totally expect their clinical journeys to look identical, but they don't. They really don't. Patient A experiences a rapid, just devastating cognitive decline over a few short years.
0:39A patient B, they remain remarkably stable, like still doing Sunday crosswords and living independently a whole decade later. Which is wild when you think about it. It is. And for decades, we've treated these 2 patients with the exact same broad brush, even though their underlying biology is obviously telling a completely different story.
0:57Yeah, and now that we have these new breakthrough anti-amyloid treatments, This variance isn't just some, you know, academic puzzle anymore. It's an urgent clinical problem. Right, because these drugs work remarkably well for some people, but surprisingly fail for others. Specifically, they fail in certain genetic groups, and the results even differ between men and women.
1:18Okay, let's unpack this. Imagine you hire a specialized cleanup crew to remove toxic spills in a house. But for some houses, the foundation keeps rotting anyway, because the cleanup crew is targeting the entirely wrong stage of the decay.
1:32That's a great way to put it, which forces us to look past the amylite itself. I mean, the plaques are really just the trigger. The question we're exploring today is what happens when our unique genetic blueprints interact with that initial pathology.
1:44If we think of Alzheimer's as a biological fire, the initial MLA buildup is the match. But your genes dictate, well how dry the kindling is, how strong the wind is blowing, and whether your internal fire engines even have brakes.
1:55Oh, wow. So today we celebrate the work of Joseph Giorgio and his colleagues, along with the Alzheimer's disease neuroimaging initiative, or ADNI, who have advanced our understanding of the genetic risk factors in Alzheimer's progression.
2:09And their work finally maps exactly where these individual genetic differences hijack the disease process. Right. So the insights for a deep dive today are drawn from their open access article titled, Variable and interactive effects of sex.
2:22APOE, Epsilon 44, and trim 2 on the deposition of Tao, in entortal, and neocortical regions. Published in nature communications on July 1st, 2025. And what's fascinating here is what they accomplish biologically.
2:37It's going to fundamentally shift how we approach personalized medicine. So we all know the canonical Alzheimer's Cascade. I mean, it's practically neurobiology 101 at this point. Amyloid beta accumulates in the brain and acts as a catalyst.
2:49Right, it triggers the buildup of toxic towel proteins, 1st in the internal cortex, which is sort of the brain's memory hub. Yeah, and from there, the tao spreads like a wildfire into the neocortex. And that governs like higher order executive functions, language, spatial reasoning.
3:04Exactly. And when Tao hits the neocortex, that's when cognitive decline becomes highly visible. But we've treated that sequence like a straightforward, inevitable chain of falling dominoes. Like amyloid tips over Tao, which tips over cognitive decline.
3:18But do the dominoes really fall at the exact same speed in the exact same way for everyone? Well, no, that linear model works perfectly for the rare genetically dominant forms of Alzheimer's, where a single mutation basically guarantees the disease, but for sporadic, laid on set Alzheimer's.
3:36Which is the vast majority of cases for you. Yeah, the vast majority. For those, that linear model is proving entirely inadequate. The cascade simply does not flow the same way across the general population.
3:46And we're seeing the fallout of that inadequate model right now in the recent clinical trials for drugs like Lecanamab and Don Anamab. Which is a huge issue. These drugs clear amyloid from the brain, and they're incredibly effective at that specific job.
3:58But their benefits very wildly. In the trials, the actual preservation of memory was all over the map, based on a patient's sex, and whether they carry the APOE epsilon war abream. For instance, patients carrying 2 copies of the APOA epsilon 4 brain gene showed no significant treatment effect on primary clinical outcomes.
4:17Yeah, the clinical frustration there is immense. You administer a drug. It does exactly what it was engineered to do biologically, but the patient still loses their memory. Wow. It strongly suggests that for certain genetic profiles, clearing the 1st domino just doesn't matter because the rest of the dominoes are already falling under their own momentum.
4:35So if the dominoes don't fall the same way for everyone, how do we actually prove where the divergence happens, we can't just slice open living brains to watch protein spread in real time, moving from the theoretical problem to the actual mechanics of the study, how did the researchers actually visualize this variability in living humans?
4:53They engineered a massive, highly coordinated, multi-cohort design. They pulled 1354 participants across the ADNI, A4, and ABS HD studies. Okay, so a huge data set. Yeah, and integrating A4 and HBS HD is critical here, because those specific studies track older adults who are still cognitively normal.
5:16If the goal is to understand Alzheimer's dementia, why is the study heavily weighted toward people who are cognitively normal? Aren't we just guessing they might get sick? It sounds counterintuitive, I know, but it's really the only way to solve the timeline problem.
5:31By the time someone presents with clinical dementia in a neurologist's office, the fire has already burned through the forest. Oh I see. Yeah, the Tao pathology has spread so widely that tracing it back to its origin is like trying to figure out which room a house fire started in after the roof has already collapsed.
5:47Right. That makes total sense. You have to catch the very 1st spark in the preclinical stages. Exactly. Okay, let's unpack the tech they use to see that spark. Because they aren't just giving cognitive tests, right?
5:58They're looking at the molecular level. Right. They used multimodal PET imaging. They inject mildly radioactive tracers that bind specifically to amyloid and tow proteins in the living brain. So they can literally light up and see the exact burdens of these proteins.
6:14Yes, and they paired those brain maps with whole genome sequencing to track specific gene variants, focusing on APOE epsilon 404 and trim too. But the real innovation here is what they did with the data, right?
6:28Causal path modeling. Instead of just looking at who gets sick, they mapped the cascade step by step. What's fascinating here is that the causal path model acts like a way to reverse engineer the biological supply chain.
6:40It isolates direct effects for mediated effects. So, like, does a gene directly increase amyloid, or does it just make Towspread faster once amyloid is there? Exactly. It pinpoints exactly where in the biological chain of events, a specific gene jumps in to make things worse.
6:55Okay, let's get into the key findings. The heart of the deep dive. They broke down 3 major genetic modifiers, right? starting with sex. Yeah, we've known female prevalence in Alzheimer's is higher, but the model showed the vulnerability occurs at the very beginning of the sequence.
7:09For a given level of amyloid, females develop significantly more tao in the interhinal cortex than males. Wow. So the match, the amyloid strikes with the same intensity, but the kindling in the internal cortex just catches fire much more easily in female brains.
7:25That is the exact mechanical takeaway. Okay, so what about APOE? Epsilon 404? That's the most notorious genetic risk factor? Right. So everyone inherits 2 copies of the APOE gene. If you have 2 copies of the Epsilon 4 variant, meaning your home is I goat, you get hit with a massive double multiplier.
7:43Double multiplier. Yeah. First, they have more primary tau in the intorinal cortex for a given level of amyloid, but second, and this is the most profound finding, they show a profoundly greater spread of that tau into the neocortex for a given level of entorinal tau.
8:00Wait, really? So they get hit on both ends. But hold on. If the drugs clear amyloid, shouldn't that eventually starve the towel of its catalyst, even in these homicide goats? You would think so, but the causal path model shows why that fails.
8:11In Homazy goats, once the tower reaches a certain threshold, its proliferation into the neocortex becomes uncoupled from the amyloid. It just dries itself. Exactly. It's essentially driven by a separate self-sustaining engine.
8:22Interestingly, heterozygos people with just one copy, do not show this severe Tao specific proliferation independent of amyloid. That is wild. Okay, here's where it gets really interesting. What happens to the brain's natural defenses?
8:37Shouldn't the immune system step in? That brings us to the trem 2 gene, right? Yes. Trim 2 controls microglia, which are the brain's resident immune cells, the cleanup crew. Normally, microglia surround amyloid plaques, creating a physical barrier to limit the seating of tau, but rare risk variants in trem 2 cause a failure in that immune response.
8:56The microglia become paralyzed. So the cleanup crew is completely asleep at the wheel. What is the model show happening then? A massive unregulated spread of tau from the enterhinal cortex straight into the neocortex.
9:08Nothing contains it. So, if APOEs, Epsilon 44 drives the spread. And trem 2 variants cause a failure to clean it up. What happens if someone has both? Do they interact? They absolutely do. The researchers found that individuals who have both a trim 2 risk variant and an APOE epsilon 4 her allele suffer an interactive effect.
9:27It results in significantly higher neocortical tau than just adding the 2 risks together. So to synthesize this, if amyloid is the spark, APOE, epsilons 40 quarter longosity, acts as a heavy gust of wind blowing the fire into the forest, and a trim 2 variant means the immune system's fire engines have broken brakes, letting the fire spread uncontrollably into the city.
9:48It's a vivid analogy and perfectly accurate. The immune cells fail to isolate the toxic proteins, allowing the disease to just race forward. So what does this whole mean? Connecting this back to the patient in the clinic in those Lacanamab and Don MMab trial results?
10:03It perfectly contextualizes why the drugs failed for APOE, Epsilon 4 or 4 homozygotes. If their tau proliferates almost independently once amyloid is present, clearing amyloid after the cascade has started simply won't stop their cognitive decline.
10:16Because the forest is already blazing. Right. This demands highly personalized medicine. Homozygotes might need anti-amyloid treatments much earlier at much lower abidale levels, or they might strictly require anti-tow therapies.
10:30And females might also need earlier treatment compared to males to prevent that primary towel buildup, right? Absolutely. The universal amyloid threshold using clinical trials right now is fundamentally flawed.
10:41But you know, we should note the limitations here. This is a cross-sectional study. Meaning it's just a snapshot in time. Right. Longitudinal tracking over time is needed to fully confirm these dynamic pathways.
10:52Also, trim 2 variants are really rare, so the sample sizes for those specific interactions were quite small. Sure, but the central insight is still huge. Genetic variations like sex, APOE, epsilon 44, and trim 2 don't just increase a generalized risk of Alzheimer's.
11:08No, at all. They act as specific localized multipliers at distinct anacomical stages of the amyloid tau cascade. Understanding exactly where a patient's genetics accelerate the disease is clearly the key to unlocking personalized effective treatments.
11:23I couldn't agree more Which leaves us with a lot to think about. What does this mean for the future of clinical trials? As we map the genome with greater precision, Will we soon see an era where Alzheimer's is no longer treated as a single disease, but is a uniquely personalized genetic cascade, requiring its own biscope cocktail of interventions?
11:42This 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.
11:56If 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.
12:06Thanks for listening and join us next time as we explore more science base by base.
