A new twist on using ‘personalized’ stem cells for studying dementia | Hacker News Reader

A new twist on using ‘personalized’ stem cells for studying dementia

singularityhub.com

87 pointselorant43 comments

As someone who does research in biology and therapeutics, it seems to me that Alzheimer's research is such a dismal field. We don't have a good idea of what causes it, not only do we not have good animal models, we have misleading animal models which reflect some of the traits of Alzheimer's but nothing translates to humans. There was this mouse model they made which developed these plaques like in Alzheimers and they found they could clear them with brain stimulation pretty well - but trying it in humans there was no effect.

> There was this mouse model they made which developed these plaques like in Alzheimers and they found they could clear them with brain stimulation pretty well - but trying it in humans there was no effect.

Rinse and repeat many times over, I think. This dynamic is what pushed me away from the field in general, the knowledge that I could spend my whole career working on a mouse model that there's reason for only tepid confidence in.

There’s a Twitter feed dedicated to this caveat

https://mobile.twitter.com/justsaysinmice?lang=en

For diseases where the causal mechanism is known very well, mouse models aren't that bad: Rett syndrome, hemophilia, spinal muscular atrophy.

It gets to be a problem where the disease is poorly defined and the cause is unknown: Alzheimers, schizophrenia, ADHD, etc.

It is very important that people have this discrimination power and not jump to "All of this is bad" or "All of this is good".

Also in a lot of cases it’s the best we’ve got. You might be able to use monkeys but even that doesn’t change a lot of the problems. If you move from animal models you aren’t going to suddenly be able to start experimenting on humans.

Very good point. I always considered that Twitter profile as tongue-in-cheek, rather than being outright dismissive.

Do you know if there are new ideas or techniques to go around that issue ? With printed tissue and new computational means .. I was hoping new roads would emerge.

Hopefully :) it's been 10 years and I haven't kept up on the topic.

Some of this has to do with the funding incentives in academic AD research. NIH continues to dump money into mouse and even more ridiculously Drosophila (fruit fly) models of Alzheimer’s. I am all for basic research but for this disease, if we have learnt one thing, it is that the success of a therapeutic in animal models is nearly 100% predictive of its failure in humans. The funding incentives need to change for meaningful change in the clinic.

Do you think that the "organ on a chip" method could help?

Having tiny human brains (or brainlets) would be a closer model than using any animal, or not?

Missing a lot of signals from the in vivo context - for example taking a naive approach to organoids typically means you’re going to be missing vascularization, which is essential for the architecture of the brain in vivo.

Also typically organoids present as very immature brains, whereas Alzheimer's patients sort of have "maximally mature" brains.

I’m not an expert, but suppose Alzheimer’s onset and progression is affected by the gut-brain axis. There’s no way for an organoid to capture that.

Yes, also hormones controlling e.g. sleep, which is essential for brain function.

> Do you think that the "organ on a chip" method could help?

Can we even reliably diagnose Alzheimer's based on brain tissue alone? If so, what do we look for?

I think tau protein is a correlate?

As far as I know, it can only be tested post-mortem but there is some research related to detecting it with blood tests.

Is there a paper on that? I'm so interested.

is Alzheimers akin to Depression in the sense its a umbrella term/abstraction?

There’s a remarkable amount of hyperbole in this article (but not the original publication). In plain language, this is a project to generate, characterize, and openly share the data for iPSC models (and presumably neurons from them) bearing 100 APOE disease-associated mutations.

It must be remembered that one reason many AD drugs have failed is because models don’t capture the disease accurately and that may well turn out to be the case here.

Do AD drugs attempt to reduce the concentration of APOE in the brain? If so, is it known if these drugs either failed to reduce the concentration of APOE, or succeeded in reducing concentration but still didn't work?

“Reduce APOE” is an oversimplified view. The mechanisms by which the AD-associated E4 genotype (arg112, arg158) results in disease are still poorly understood. The main therapeutic hypothesis for AD is to reduce the levels of pathogenic plaques and aggregates using antibodies.

Interesting. And these studies have shown that medicine which reduces these plaques doesn't actually help with AD symptoms?

I've pinched the title from https://www.ninds.nih.gov/News-Events/News-and-Press-Release..., which the OP points to, since it seems less hyperbolic and more precise.

Not sure if the article itself should be changed also...

The problem with AD is by the time the symptoms appear, the brain is already irreparable ravaged. The damage is actually macroscopically visible. Treatment would have to begin long before symptoms show, which would mean identifying people who are at risk.

I think any person newly diagnosed with alzheimers would jump at the chance to stop its progression.

Once you get an alzheimer's diagnosis your brain is pretty ravaged. The diseases process probably starts decades before most diagnoses.

And I think anyone newly diagnosed would love to have progression stopped at diagnosis.

What’s your point?

It’s not a total cure and restoration, but it’s better than now- where your brain is ravaged at diagnosis and gets worse until death.

The problem is what would halt progression at diagnosis isn't necessarily the same thing that would halt progression 20 years earlier. And there is evidence to suggest this is the case.

Right, but these are separate things that would both be beneficial. They should be exclusive pursuits.

Currently we have neither. I would prefer some preventative therapy that makes it never happen. I would also like a therapy that halts progression once diagnosed.

Luckily https://www.nia.nih.gov/news/blood-tests-show-promise-early-... points at a test that looks like it can figure out who will develop Alzheimer's years before there are any symptoms. (In their dataset, up to 20 years before.)

What can you do if you find it early?

prepare your will

If you have MRI scans of your brain, you can upload them to brainkey.ai to track differences over time. They even offer a ≈400USD MRI scan out of pocket if you live close to one of their participating scanning centers (e.g. in the Bay Area).

There are many studies underway to find biomarkers eg changes in the blood proteome, sleep, etc for the pre-disease prodromal state of AD.

There is a fascinating episode on Alzheimer treatment using some sort of frequency exposure [0], I don’t know much about this but radiolab made me really excited that there could be an end to Alzheimers one day soon. Studies like this and the hopeful potential benefits unlocked with crispr also!

https://www.wnycstudios.org/podcasts/radiolab/articles/bring...

Interesting. Hopefully it doesn’t suffer from the mouse model flaw.

I think the overall goal assumes that genetics causes Alzheimer's, which is (in my opinion) a questionable assumption. That's not to say that the production of model systems for Alzheimer's will be useful, but will it be worth the tens or hundreds of millions of dollars that this will cost? ¯\_(ツ)_/¯

The book "The End of Alzheimers" makes a case that there are 36 causes of Alzheimers. It's the combination of those causes that results in the disease. The more of those causes a person has, the more likely they'll develop Alzheimers.

This is why research and treatments that focus on only one cause have all been stymied and went nowhere.

The book's treatment program is to address as many of the causes as possible.

As long as it's not diverting funds from anything more promising then spend all the money you want I say. Try everything. You may fail more often but you'll find the right answer faster.

With a limited NIH budget this will necessarily divert funds from other endeavors. In addition, I think such programs entrench the large lab bias prevalent in academia right now.

Gene therapy can treat diseases that aren't specifically caused by genetics.

Yes, but that would require knowing how to rescue the phenotype, in this case healthy neurons and connections. Alzheimer's is not like Parkinson's, which you can (most likely, see the recent work from BlueRock) treat by replacing dopaminergic neurons or rescuing the dopamine production.

_Crush_ Alzheimer's.

Unrelated: Tofisopam has been shown to have groundbreaking efficacy on reverting cognitive impairment: https://pubmed.ncbi.nlm.nih.gov/31981560/

btw I believe that Tifasopam is the best long term anxyolitic medication out there: No tolerance No cognitive impairment No sedation No side effects Very potent

Reply on news.ycombinator.com