GEN BiotechnologyVol. 1, No. 1 Asked & AnsweredFree AccessA Matter of Timing: Rudy Tanzi on Alzheimer's DiseaseRudolph E. Tanzi and Malorye Allison BrancaRudolph E. TanziRudolph E. Tanzi, PhDis a co-founder, advisor and/or equity holder in several biotech companies mentioned in this interview, including ProMIS Neurosciences, Amylyx, and REACT Neuro. E-mail Address: MQuintanillaDieck@mgh.harvard.eduSearch for more papers by this author and Malorye Allison BrancaMalorye Allison Brancais a freelance science writer based in Acton, MA. E-mail Address: maloryea@gmail.comSearch for more papers by this authorPublished Online:16 Feb 2022https://doi.org/10.1089/genbio.2021.29004.intAboutSectionsPDF/EPUB ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions Back To Publication ShareShare onFacebookTwitterLinked InRedditEmail Rudolph E. TanziFor the better part of four decades, Rudy Tanzi, PhD, has been one of the most prominent and influential researchers in the field of Alzheimer's disease (AD) and other neurodegenerative disorders. A pioneer in the genetics of Alzheimer's, a founder and/or advisor to multiple biotech companies, co-director of the McCance Center for Brain Health at Mass General Hospital in Boston, and a successful book author, Tanzi has strong views about the lack of translational success in turning genetics insights into meaningful therapies.In this wide-ranging interview, conducted by contributing editor Malorye Allison Branca, Tanzi doesn't hold back. Gathering evidence from whole-genome sequencing studies and 3D Alzheimer's-in-a-dish models, Tanzi argues where the pharma industry has gone wrong. He lays out the key steps in disease pathogenesis and the early promise in clinical trials of new therapies that are as much about when to target the disease as what or where.This interview, originally conducted for GEN Edge, has been edited for length and clarity.Branca: It seems like Alzheimer's disease, after many years of hard news, has become a positive area again. What are some things that are making people optimistic about Alzheimer's research now?Tanzi: This optimism is also controversial regarding which hypothesis about the disease is correct. But if you follow the genetics, you won't go wrong. The genetics doesn't just tell you what needs to be addressed, it also tells you when it must be addressed. And it's that second part of when to address each pathology that a many people miss—they want to just be binary and say, “this is right, that's wrong,” and throw out the baby with the bath water. I'm first referring to β-amyloid, the gene for which I discovered back as a student at Harvard (amyloid precursor protein, APP).1 Later I co-discovered the presenilin genes.2,3 These are all early-onset familial AD genes.“The genetics doesn't just tell you what needs to be addressed, it also tells you when it must be addressed.”All these findings said amyloid was the culprit. Whether it was APP or the presenilins, which cut at the gamma secretase site to make β-amyloid, amyloid was it. Then APOE (apolipoprotein E) came around as the major late-onset risk factor. APOE controls the release of amyloid from the brain. So we can't deny that amyloid triggers this disease.What I learned early on in genetics is that genes that cause early-onset forms of disease usually tell you about events that come early—you have to treat preferably with early detection, early intervention. Take cholesterol—Mike Brown and Joe Goldstein first found a mutation in the LDL receptor gene that led to high cholesterol in a family with a rare mutation that had very early-onset hyperlipidemia. You have to treat cholesterol decades before you get heart disease. You don't wait until you get a bypass or have congestive heart failure to pop your first statin.This is exactly analogous in Alzheimer's. The amyloid genes are early-onset, even [homozygous] APOE4 onset is relatively early mid-60s. So the genes tell us you have to hit amyloid a decade or two before symptoms, just like we lower cholesterol decade before heart disease may occur. All the brain imaging shows that the amyloid starts rising, even in patients after 40 years old, up to 2–3 decades before symptoms arise. By the time you have even the mildest cognitive impairment, amyloid's peaked and is already plateauing. So amyloid is something you have to hit early.Does that explain why there has been so much failure in Alzheimer's disease?Absolutely! In some cases, the amyloid drugs didn't work well, but there have been therapies that clear amyloid just fine. Take aducanumab (Aduhelm) from Biogen. It was actually Rob Moir, from my lab who passed away in 2019, who inspired that drug: he discovered auto-antibodies to Aβ oligomers that protect against Alzheimer's.4 Roger Nitsch in Switzerland started the company Neurimmune and they found one such [auto-antibody], reverse translated that into aducanumab and licensed it to Biogen. So scientifically it's great to be an early part of that story. But if you're going to use an antibody to hit amyloid, using it in a patient who is already symptomatic is difficult.One [Biogen] trial had marginal improvement in cognition, the other had none. Previous amyloid trials had no cognitive improvement. So the FDA very carefully approved it for removing amyloid, not for treating cognition and said, “let's see if removing amyloid helps.” Well, if it's only removing amyloid in folks who are symptomatic, good luck. But if you use it for early detection, early prevention, then you'll see a result. The problem is, it's easy to give somebody Lipitor if they have high cholesterol. It's not as easy to give 38 million asymptomatic Americans, who have amyloid in their brain, a drug that costs $56,000 a year. That's the problem.The silver lining is that this decision has opened a door for trials on cheaper, safer drugs that can remove amyloid just as well [as aducanumab], that are exponentially less expensive and safer. If we get one of those approved, then we can argue it's best to use this early. Let's do early detection, a blood test. If you get a result, perhaps verify with imaging and then take this little drug.People who have extraordinarily high amyloid levels, when you do early detection, may be candidates for immunotherapy because they have so much amyloid you have to hit with the antibody first. Then in a serial combination, you bring them to some other drug—I'm working on γ-secretase modulators, which would keep beta-amyloid down—and check again after a year.Like Jerry Garcia once said, with every silver lining, there's a touch of gray. The touch of gray is that meanwhile, people may try to mortgage their homes to get dad aducanumab because he's got advanced Alzheimer's and it is unlikely to improve his memory.Your group published a study in 2021 using whole-genome sequencing to identify 13 new AD candidate genes (Figure 1).5 What was the difference between previous genome-wide association studies (GWAS) and this study?FIG. 1. The 2021 GWAS study from Tanzi and co-workers.The earlier GWAS told us that amyloid initiates the disease and that's still the reigning paradigm. The only controversy is when we treat amyloid in full-blown patients, we don't make them better. To me, that's like saying, “I treated this patient who has congestive heart failure with Lipitor. He didn't get better.” No! You had to do it before symptoms.The early genetics are going to tell us what to do for early detection, early intervention. If you have amyloid brewing in your brain—amyloid quickly causes tangles, look for the tangles. These are initiating pathologies, plaques and tangles. Those tangles have to spread like brush fires for decades, setting off neuroinflammation. These are like forest fires—that's what causes the symptoms. In a patient who's symptomatic, forest fires of neuroinflammation are already blazing. And you're hoping that you'll make them better by stomping out the brush fire of the tangle or blowing out the match of the amyloid that started it decades ago.“In a patient who's symptomatic, forest fires of neuroinflammation are already blazing.”Most of the GWAS hits are bringing us to neuroinflammation. The early-onset genes tell us the earliest events you have to hit, preferably pre-symptomatically, like amyloid and tangles. The late-onset AD genes from GWAS are telling us what you need to hit later in patients: neuroinflammation, microglial activation, astrogliosis. We found the first neuroinflammation-related AD gene CD33 in 2008.6 I remember Time magazine called it a top 10 medical breakthrough. We laughed because we don't even know what this gene did at the time!We later figured out that CD33 is the “on” switch for the microglial cells to be activated. Normally the microglial cells are housekeeping—while you're sleeping, they're eating amyloid, cleaning up your brain. If those cells instead start eating by chance some pieces of neurons that died, that gives them a signal that neurons are dying and the microglial cells are still evolutionarily programmed the way they were 30–40,000 years ago. If neurons are dying, the microglial cells interpret that as an infection and wipe that part of the brain out—that's neuroinflammation.When microglial cells get activated, they also turn on the astrocytes and [you] get astrogliosis, which causes more neuroinflammation and neuronal cell death. This is all part of an evolutionarily conserved program that is meant to protect the brain. These microglial cells didn't get the memo that now we live till 80, and neurons can die for other reasons. When they see neurons dying, they just wipe that part of the brain out, assuming its neuroinflammation, which is meant to protect you.I call it auto-innate immunity—when your innate immune cells in your brain, your only immune protection in your brain, turn against you.So a natural process has gone awry just because people are living too long?It's similar to COVID-19—COVID doesn't kill you. It causes a cytokine storm and inflammation in lungs and vessels. That's what kills you! Same thing in Alzheimer's: the plaques and tangles don't cause the disease, they initiate the original cell death, which together with amyloid deposits activate the microglia to cause a cytokine storm in the brain. Those cytokines turn on the astrocytes. Now you have active glia everywhere that is causing neuroinflammation. That process kills 10 times more neurons than the original plaques and tangles that initiated this process decades before.It's analogous to [American] football. I work with the New England Patriots as a brain health advisor. Football players and boxers get tangles, not from amyloid but from bangs to the head—in their teens, 20s, or 30s. But it takes decades of those tangles spreading like brush fires. The match in these sports are bangs to the head. In AD, the match is amyloid. In either case, the tangles are the brush fires ignited by the match, which then spread for decades. Then finally you get enough neuroinflammation—forest fires—to get chronic traumatic encephalopathy (CTE).Both CTE and Alzheimer's are tauopathies. Tangles spreading cause a disease but in Alzheimer's, amyloid induces it—that's the most common form of dementia. In CTE, a rare form of dementia, it's head bangs. But in both cases, neuroinflammation is needed to kill enough neurons to get symptoms.I'm optimistic. We can see the timeline because we created mini human brain organoid model of AD—“Alzheimer's-in-a-dish,” funded by the Cure Alzheimer's Fund. Mice have led us astray in this field. Mice are not humans! It's pretty obvious you're better off learning the chronic order of events and molecular mechanism in brain organoid systems, like the ones we first published in 2014.7What's new about the genetics of Alzheimer's that you've discovered in your recent GWAS?We've discovered a number of different genes. The bulk of the new genes we and others discovered over the past decade or two brought us away from amyloid to neuroinflammation. We found CD33; deCODE Genetics found TREM2 in 2013. Way back when I was a kid, in 1980 with Jim Gusella, we found the first five SNPs [single nucleotide polymorphisms] ever in the genome. Back when no-one was doing this—I was 20, he was 25—we mapped the Huntington's disease gene with the first five SNPs miraculously.8I've been working on SNPs since they were first discovered. GWAS takes the most common SNPs, equally spaced, and you scan the whole genome for hits. That's where we see these neuroinflammatory genes popping up. More recently, we asked, what if we use whole genome sequencing? We're looking at 50–70 million variants in each person. Now you can ask, what if there are rare variants and why is this so important? What if there are rare variants that are only present in 1% of the population that will cause the disease?For the first time, we could do a scan where the input was the whole genome. We were wondering if a lot of the missing part of the genetics was in the rare variants, where each variant is 1% of the population or less.To our surprise, what we found was that the 13 genes that came out with the highest significance brought us back from microglia and neuroinflammation to neuronal genes involved with synaptic function and maintenance. But then, how could Alzheimer's, a disease of the synapses, not involve synaptic genes? Finally, here they are. But it looks like it's rare variants in those synaptic genes whereas it's common variants that predispose to neuroinflammation. That makes sense because neuroinflammation is the end game of all neurodegenerative disease.Other misfolded proteins that initiate the disease, like alpha synuclein causing Lewy bodies in Parkinson's and Lewy body disease, TDP43 in ALS, analogous to amyloid and tangles in Alzheimer's, all kill neurons. But that relatively small amount of neuronal cell death can trigger the neuroinflammation, which kills 10 times more neurons. Without neuroinflammation, you don't get the symptoms. You'll have a resilient brain.Once in a while, people die in their 80s with no dementia but are then found at autopsy to have abundant levels of plaques and tangles in their brains. How did they not have Alzheimer's? It's always the same answer: They were spared of neuroinflammation. Their microglial cells stayed calm and carried on housekeeping and cleaning at night, rather than becoming reactive and saying, “neurons are dying, wipe out this area of the brain!” Resilience comes when the microglial cells never took off the housekeeping apron to put on the SWAT team outfit.Pharma companies have been sticking with Alzheimer's for a while but have taken a few body blows. Recently Lilly said it is refocusing. Is there going to be a new generation of drugs or are they just going to keep trying the same things?Most of the big pharmas still follow each other. Biogen got aducanumab approved and an approval is an approval. There's a chance to make money. Lilly has an antibody that's just as good, maybe even better than aducanumab. There are smaller companies too. For example, ProMIS Neurosciences (in which I also hold equity) has an immunotherapy against Aβ oligomers that is much more relevant for stopping neurodegeneration. Amyloid is still a big target because the FDA just set a precedent: They approved a drug that removes a certain amount of amyloid based on PET scans. This opens the doors, right? Anybody can come in and say, “Here's my drug. I can remove amyloid just as fast over the same period, using the same PET scan. Approve me too!” What is the FDA supposed to say? They already set the precedent. Companies would be foolish not to follow that example.My warning is that hitting amyloid is a crapshoot as to whether you're going to make somebody better. Most would say it's not even a crapshoot, it's a guaranteed loss. You have to hit the right form of amyloid. That's why I like to work with this smaller company, ProMIS, because they're targeting the oligomeric Aβ that I think is the most dangerous to synapses versus the plaque Aβ, which I think we should let lie. The plaques are more or less inert and neutral. Maybe then you might see some greater effects on cognition?Amyloid comes early and amyloid does its job early. By the time you have even the earliest signs of dementia or mild cognitive impairment, amyloid has already peaked, plateaued, and, if anything, starts to go down a little bit. You have to hit it on the way up, 1–3 decades before the onset of symptoms. That's where these antibodies that Lilly and Biogen and others are making are going to be useful. The question is, can you make them cheap and safe enough to use it for secondary prevention in millions of people who have amyloid but do not yet have symptoms?In Alzheimer's, are we talking about a range of diseases or one condition?The disease is heterogeneous in many ways. How does it start? Some folks are making normal amounts of Aβ and tangles, but they have genes that ramp up their neuroinflammatory response. Others have genes that just ramp up the amyloid early on. Others we just found have genes that make synapses more vulnerable to amyloid and neuroinflammation. There are many different routes to get there, but in the end it's neuroinflammation that takes you out. If you want to treat a patient who is symptomatic, no matter how you got there, you have to stop neuroinflammation or protect against it.At the pre-symptomatic, early-intervention stage, that's where heterogeneity comes in. In some folks, it may be more important to stop amyloid. In others it may be more important to stop the amyloid or the tau pathology from spreading. My guess is that if you have drugs that safely promote amyloid clearance by microglia—a little white powder that does what aducanumab does for a 100 × less money—or a drug that stops the tangles from spreading, any drugs that stop initiating pathologies despite the heterogeneity will be useful in everyone, early on, despite the early heterogeneity.“The match in these sports are bangs to the head. In AD, the match is amyloid. In either case, the tangles are the brush fires ignited by the match, which then spread for decades.”So you see this as similar to an anti-cholesterol drug? People deemed “at risk” will take a preventative medicine and that will reduce their risk?Exactly like checking your cholesterol when you're young and taking care of your cholesterol levels decades before you might get heart disease. That's how we put a big hit on heart disease. Remember, not many believed Brown and Goldstein when they first showed this one family with early-onset hyperlipidemia, and they said, it's the LDL receptor mutation, maybe cholesterol has something to do with heart disease. They got the same criticism that we got when we discovered the familial Alzheimer's gene and said, it's amyloid that causes Alzheimer's!This controversy—the media loves it! They love to whip up the controversy and make scientists look like stupid kids fighting with each other. The smart scientists don't do that. They don't look just at whether something's involved. They look at when it's involved. That's the mindset change we need—not whether but when to hit that pathology. You have to look at the temporal order of things and then determine when to hit. This is something that pharma has not done well.Your GWAS paper was a revelation in understanding Alzheimer's disease. What are the next things that we need to solve?I think our γ-secretase modulators that don't block γ-secretase, but they're allosteric modulators of the γ-secretase docking site. We don't just hit γ-secretase with a sledgehammer (like Lilly did) or β-secretase. You can't—these enzymes are too important. Instead, you have to tweak them with a jeweler's screwdriver. Then you have the other drugs that can promote microglial clearance of amyloid. That's what aducanumab does. We have many drugs from our drug screens, from our Alzheimer's in a dish model, where we could do what aducanumab does, with cheap and safe repurposed drugs and nutraceuticals, just as well. We're testing in mice now to see if those could become a replacement. They would be cheap enough to someday use for secondary prevention in 38 million people. That's how we're going to prevent the disease. Early detection—early intervention.But in terms of treating the disease, I like what Denali and Alector are doing, they're targeting the main neuroinflammation genes, like CD33, trying to turn it off. (For disclosure. I'm the inventor on the issued patents for CD33 gene therapy and immunotherapy). I like the idea of activating TREM2. I think that's the way forward for treating patients—you have to learn from the neuroinflammation genes.Finally, these new synaptic genes we found from whole-genome sequencing GWAS tell us how we might provide resiliency to synapses. At the end of the day, the degree of dementia correlates with the degree of synapse loss. That's why I tell everybody, always strive to learn new things and build up your synaptic reserve with the same rigor you build up your financial reserve, because the more synapses you make, the more you can lose before you lose it. This is one of the main messages in my most recent lay book, The Healing Self.“With Alzheimer's, we wait until the brain degenerates to the point of dysfunction before we treat it.”Are people going to volunteer for these trials?I'm a co-director of the McCance Center for Brain Health at Mass General. We're doing trials right now on nutraceuticals that work in our Alzheimer's in a dish model—our 3D human neural culture and brain organoid models.We put the Alzheimer's genes in these neurons and get amyloid plaques in four weeks. If you wait another week, we get full-blown tangles. We showed for the first time what mice never showed us—that amyloid directly induces tangles. The most referenced paper in our field is the amyloid cascade hypothesis.9 It was a handwaving argument for why mice didn't get tangles. We now know mice don't get tangles after you put amyloid genes in their brain, because they can't make tangles! They have the wrong isoforms of tau. You need a 50:50 ratio of four- and three-repeat tau, like in the adult human brain, to get a tangle. Mice don't have this and, thus, cannot make tangles.While we now know [mice] can't make tangles, we can see tangles made in our dish models—β-amyloid induces tangles directly. Then if you add in microglial cells on the side, you can see they get recruited once neurons start dying with plaques and tangles. The microglial cells rush in and start eating synapses and axons. They also activate the astrocytes, which also start eating synapses. The whole thing is modeled in a dish in five weeks. This has made drug screening, supplement screening, nutraceuticals, and natural product screening 100 times faster, 100 times cheaper.Do you have volunteers who are taking these drugs?We spent five years and many millions of dollars, thanks to the Cure Alzheimer's Fund and the NIH. Now we have identified over 100 approved drugs and natural products that either stop amyloid production safely and induce microglial clearance of amyloid, like aducanumab does, or stop amyloid from inducing tangles (40 different drugs or nutraceuticals do that). Then we have another 40 drugs and nutraceuticals that prevent the reactive astrocytes and microglia to curb neuroinflammation. These were all identified in our Alzheimer's-in-a-dish models. We've now put together a taskforce to pick out which of the drugs and natural products are safest and most efficacious to start platform clinical trials in Alzheimer's patients with Howard Feldman of UCSD and the ADCS.With have already started one of the trials at the McCance Center. We can screen for approved drugs and natural products using these dish models so quickly, we can target exactly what pathology we want and pick out combinations. That's what we originally did with Amylyx (a company in which I hold equity as co-founder). We picked two drugs, one that is a regular drug and that is a one natural product. Together, this combination was successful in an ALS clinical trial10 and has been submitted to the FDA as part of an NDA. The combination, called AMX035, protected neurons against oxidative stress and neuroinflammation.Now we're trying to repeat that, let's see if we can hit amyloid and tangles that way as well as microglial activation and provide neuroprotection. But now it's 100 times faster, 100 times cheaper, because we have these dish models and we don't have to depend on unreliable mouse models.What is standing in your way between now and taking these to the finish line?Money! It's hard to get companies to fund drug repurposing trials. If you're going to repurpose a drug (or a combination of a repurposed drug and a natural product), you have to start a small company that generates intellectual property around that combination, like we did with Amylyx. Big pharma won't come in until you do that combination.Luckily, we have donors, who help to fund these early platform trials. If you get the right combination, then you launch a biotech company around it, you get the IP, and then you get pharma interest. But it has to start organically.Are you doing any more genetics studies?We're always doing genetics. I've never stopped doing genetics since I helped find the first five human SNPs in 1980. We're doing more whole genome sequencing. We just analyzed 30,000 genome sequences and identified five or six more genes that reach genome-wide significance. It's important to keep that genetics pipeline coming in, because those are the targets you have to hit. But pharma has to [understand] that a target tells you not just what to hit, but when to hit it…The elephant in the room is that with Alzheimer's, we wait until the brain degenerates to the point of dysfunction before we treat it. We're reactive—imagine if we did that with diabetes! We'd wait until you lose half of your islet cells in your pancreas before we treat insulin and glucose levels. Or with heart disease, we'd wait to see when your heart starts to get clogged up and doesn't work well, then we're going to try to bring your cholesterol down. That sounds absurd, but that's what we do in Alzheimer's with amyloid. That's why started the McCance Center, to be proactive about early detection. What are the indicators of brain health?Another company I co-founded called React Neuro has a modified virtual reality device that does a neuropsych exam, eye scanning, and voice analysis. It will hopefully predict Alzheimer's or brain health issues decades before problems. We're hoping that this device that takes two minutes to do the brain health exam will be in every single doctor's office someday.When you go for your physical every year, you'll get your mental, as well. Like [former Congressman] Patrick Kennedy said, “There's no checkup from the neck up.” They get from the neck up and look at the holes in your head, mouth nostrils, eyes, ears and say “OK. you're fine.” How about that three pounds of gel under my skull?“When you go for your physical every year, you'll get your mental, as well.”We want to change this. We want to get to the point where you see early changes in the brain with early detection, just like we do for the heart pancreas, early lungs—early detection, early intervention, and make sure to use the right drug at the right time in the right person. That's the goal. We just hope pharma will adapt this mentality and stop just being reactive and chasing the next big drug that eventually fails and, perhaps, works with us to do promote brain and prevent brain disease in a more effective way. That's my hope.