More AI Fun With Hallmarks of Aging
Can small molecules hit all 12 Hallmarks?
I can’t resist revisiting the idea of asking some AIs what drugs or treatments hit several of the Hallmarks of Aging. Can you target all 12 Hallmarks without expensive gene therapies? Can the effects of multiple treatments be additive? [2] [3] (spoiler: yes)
AI Query 1:
What combination of drugs (like rapamycin,metformin, spermindine,berberine,cycloastragenol, acarbose, trehalose, phenformin, senolytics like Azithromycin /fisetin/quercetin, Urolithin A, senolytics), or methods (like 'heterochronic parabiosis', 'caloric restriction', 'senolytic treatment', 'stem cell therapy', 'fecal matter transplant/gut microbiome from a healthy donor), or genes (assuming a gene therapy like telomerase, existed that could inhibit or enhance that gene), or other lifespan extending treatments, might be able to have an additive effect on lifespan by targeting the 12 hallmarks of aging (Genomic Instability, Telomere Attrition, Epigenetic Alterations, Loss of Proteostasis, Deregulated Nutrient Sensing, Mitochondrial Dysfunction, Cellular Senescence, Stem Cell Exhaustion, Altered Intercellular Communication, Chronic Inflammation, Dysbiosis, Macronutrient Sensing) . Can you map some of these combinations to the Hallmarks they might affect in a chart with combinations in rows, and Hallmarks as columns.
Make combinations that do not duplicate effects, but impact as many Hallmarks of Aging as possible, to a medium to high impact. Can you map some of these combinations to the Hallmarks they might affect in a chart with combinations in rows, and hallmarks as columns. Focus on drugs shown to have additive effects in research. Assume the experimental animal does normal basic health techniques and lifestyle for longevity like exercise, avoiding obesity, balanced nutritional diet, and no poisons like cigarettes are used,so do not list any of those lifestyle effects in the combinations. Label each combination as being expected to be low med or high in impact. Only list those with an overall medium to high expected effect.That gives a table like this, and you can see that many hallmarks are targeted at least a little:
But, you may be saying, “Who can afford a gene therapy? They run a million plus for don’t they?” and of course, “Can we get coverage of ALL 12 Hallmarks?”.
Well, lets try this and leave out the expensive or hard to do treatments, and always include easy to obtain treatments:
AI Query 2:
What combination of drugs (like rapamycin,metformin, spermindine,berberine,cycloastragenol, alphaketogluterate (AKG), acarbose, trehalose, phenformin, or other small molecule drugs or chemicals from anti-aging literature, senolytics like Azithromycin /fisetin/quercetin, Urolithin A, senolytics), or methods (like 'caloric restriction', 'senolytic treatment', 'stem cell therapy', 'fecal matter transplant/gut microbiome from a healthy donor), or other lifespan extending treatments, might be able to have an additive effect on lifespan by targeting the 12 hallmarks of aging (Genomic Instability, Telomere Attrition, Epigenetic Alterations, Loss of Proteostasis, Deregulated Nutrient Sensing, Mitochondrial Dysfunction, Cellular Senescence, Stem Cell Exhaustion, Altered Intercellular Communication, Chronic Inflammation, Dysbiosis, Macronutrient Sensing) . Can you map some of these combinations to the Hallmarks they might affect in a chart with combinations in rows, and Hallmarks as columns.
Make combinations of 7 or more, and try to impact as many Hallmarks of Aging as possible, and prefer a medium to high impact. Always include rapamycin and metformin and trehalose and fisetin since these are readily available. Can you map some of these combinations to the Hallmarks they might affect in a chart with combinations in rows, and hallmarks as columns. Focus on drugs shown to have additive effects in research. Assume the experimental animal does normal basic health techniques and lifestyle for longevity like exercise, avoiding obesity, balanced nutritional diet, and no poisons like cigarettes are used,so do not list any of those lifestyle effects in the combinations. Label each combination as being expected to be low med or high in impact. Only list those with an overall medium to high expected effect.
That query seems better, and targets all of the hallmarks at a medium or high level, if you can believe an AI. But it does give you a direction to research this more perhaps.
Another AI query:
What combination of drugs (like rapamycin,metformin, spermindine,berberine,cycloastragenol, alphaketogluterate (AKG), acarbose, trehalose, phenformin, or other small molecule drugs or chemicals from anti-aging literature, senolytics like Azithromycin /fisetin/quercetin, Urolithin A), or methods (like 'caloric restriction', 'intermittent fasting', 'senolytic treatment', 'stem cell therapy', 'fecal matter transplant/gut microbiome from a healthy donor), or other lifespan extending treatments, might be able to have an additive effect on lifespan by targeting the 12 hallmarks of aging (Genomic Instability, Telomere Attrition, Epigenetic Alterations, Loss of Proteostasis, Deregulated Nutrient Sensing, Mitochondrial Dysfunction, Cellular Senescence, Stem Cell Exhaustion, Altered Intercellular Communication, Chronic Inflammation, Dysbiosis, Macronutrient Sensing). Can you map some of these combinations to the Hallmarks they might affect in a chart with combinations in rows, and Hallmarks as columns.
Make combinations of 7 or 8 or more, and try to impact as many Hallmarks of Aging as possible, and prefer a medium to high impact. Always include rapamycin and metformin and trehalose and fisetin since these are readily available. Include some things that reduce inflammation, Enhances autophagy. Can you map some of these combinations to the Hallmarks they might affect in a chart with combinations in rows, and hallmarks as columns. Focus on drugs shown to have additive effects in research. Assume the experimental animal does normal basic health techniques and lifestyle for longevity like exercise, avoiding obesity, balanced nutritional diet, and no poisons like cigarettes are used,so do not list any of those lifestyle effects in the combinations. Label each combination as being expected to be low med or high in impact. Only list those with an overall medium to high expected effect.
In this one, I asked to include more treatments, to get a higher impact to more Hallmarks. The AI was happy to comply, but that is of course one of the problems with an AI. They “want” to give you the answer that you seem to want. So you have to be skeptical of the answers that you get. And of course, this is biased. It is based on published Internet material, and that is often by people/researchers/health enthusiasts who really want to believe the good results. And the AI doesn’t do lab experiments; it just believes what it reads.
Dosing?
With anything like this, you understand that dosages matter. A small amount of foxglove extract has been used to treat a heart problem, but too much can kill you. Dose matters a lot, but the dose for mice (like the experiments) doesn’t cleanly scale to humans sometimes. In fact, many phase I clinical trials in humans are to establish a safe dosage. So I asked AIs to guess, and they gave me some ranges from the literature.
Human Dosing (mg/day)
Rapamycin:
Dosage: 1-5 mg/day, often used in specific clinical trials.
Reference: "Rapamycin in Human Trials"
Metformin:
Dosage: 500-2000 mg/day, often used in diabetes treatment.
Reference: "Metformin and Cardiovascular Health"
Trehalose:
Dosage: 2-10 grams/day, often taken as a supplement.
Reference: "Trehalose and Neuroprotection in Humans"
Fisetin:
Dosage: 100-200 mg/day, often taken as a supplement.
Reference: "Fisetin and Cognitive Function"
Spermidine:
Dosage: 1-5 mg/day, often taken as a supplement.
Reference: "Spermidine and Longevity in Humans"
Berberine:
Dosage: 500-1500 mg/day, often used in metabolic health.
Reference: "Berberine and Metabolic Syndrome"
Acarbose:
Dosage: 25-100 mg/day, often used in diabetes treatment.
Reference: "Acarbose and Glycemic Control"
Quercetin:
Dosage: 500-1000 mg/day, often taken as a supplement.
Reference: "Quercetin and Inflammation"
Urolithin A:
Dosage: 500-1000 mg/day, often taken as a supplement.
Reference: "Urolithin A and Muscle Health"
Phenformin:
Dosage: 50-150 mg/day, historically used in diabetes treatment but
less commonly today due to safety concerns.
Cycloastrogenol:
Dosage: 20 mg/kg of mice for 7 days (20 mg/kg bodyweight; equivalent
to a dose of 6.5 mg/m2 in patients)
https://www.sciencedirect.com/science/article/pii/S0047637422001385
Astragalus membranaceus root:
Dosage: 1-60g/day.
https://www.healthline.com/nutrition/astragalus#dosage
Do we believe it?
So, that is the opinion of an AI. As a human, I find this unconvincing for many reasons.
There were many warnings, by the AI about species differences, individual differences, the lack of clinical trials, and of course the ubiquitous, “Always consult with a healthcare professional before starting any new treatment regimen.” And of course, I am talking to AIs without safeguards, so that is something to remember, as well.
An AI with guardrails, like Gemini, gave me this response, in many cases, “Consult with a qualified professional for appropriate dosing.” And I agree, that makes sense, but I did want a ballpark just for a thought experiment. I wanted to know that it wasn’t some unreasonably large amount, like a kilogram. And there is a limit to how many 1 gram pills I might want to take every morning. The other side of this is the low dose that you might get in a supplement - many spermidine supplements are just wheat germ, which won’t hurt you, but may not help you either.
But even without that, there are additional issues, such as the variable dosage found in herbs, and the other unresearched compounds that will be found in plants such as “Astragalus membranaceus root”.
Interactions:
There are specific interactions, such as rapamycin and metformin. Rapamycin can induce insulin resistance, while metformin improves insulin sensitivity.
Berberine and Warfarin: There have been reports suggesting that berberine may interact with warfarin, potentially altering its anticoagulant effects and increasing the risk of bleeding.
Berberine and Metformin: While both are used for glucose control, their combined use in humans has not been thoroughly studied, and caution is advised due to potential additive effects on blood sugar levels and risk of hypoglycemia.
Metformin and Statins: There's concern about potential interaction between metformin and statins, which could lead to increased risk of myopathy (muscle weakness or pain).
Metformin and ACE Inhibitors: Potentially increases the risk of lactic acidosis when used together, particularly in individuals with renal impairment.
Metformin Interaction with Chemotherapy: Metformin has been reported to potentially enhance the efficacy of some chemotherapy drugs, but it may also increase the risk of lactic acidosis, a serious condition, in patients with compromised kidney function.
Rapamycin and Berberine can affect the mTOR pathway, potentially leading to over-suppression.
Rapamycin and Acarbose: Acarbose can cause gastrointestinal side effects, and combining it with other drugs might exacerbate these symptoms.
Rapamycin and Calcineurin Inhibitors / Immunosuppressants (e.g., Cyclosporine): Both drugs suppress the immune system, and their combination could lead to increased risk of infections or other immunosuppression-related complications.
Fisetin and Other Senolytics (Quercetin, Urolithin A): Combining multiple senolytics could lead to excessive cellular clearance, potentially harming healthy cells.
Trehalose and Other Osmolytes: High doses of osmolytes like trehalose might lead to osmotic stress.
Trehalose and Diuretics: Trehalose might potentially interfere with the action of diuretics, impacting fluid balance and electrolyte levels.
Fisetin and Anticoagulants: Fisetin's potential antiplatelet effect needs to be carefully managed in patients already on anticoagulants or antiplatelet drugs to avoid bleeding risks.
Fisetin and Chemotherapy Drugs: While fisetin has shown promising anti-cancer properties, its interaction with chemotherapy drugs is an area of concern that requires further research.
Fisetin and Anticoagulants: Fisetin's potential antiplatelet effect needs to be carefully managed in patients already on anticoagulants or antiplatelet drugs to avoid bleeding risks.
So I am not sure how useful any of this is, besides being an interesting thought experiment, not a Biology experiment.
Caveats:
Just to be clear, this is not a recommendation, because there are many unknowns. Some of these chemicals are prescription drugs. It is just a thought experiment. It is interesting to speculate that there are small-molecule drugs and chemicals that can affect the various Hallmarks of Aging, without expensive million-dollar gene therapies. And we do know that the oral rapamycin+metformin combination has been published, with positive results, so this type of speculation may be justified. Some real experiments like this have been done and shown additive effects.
There are questions of course, about whether this would actually extend the lifespan of a lab animal, or whether some of these would impact treatments in a negative way. For example, some drugs, like quercetin, do not absorb from the gut very well. Others are ‘detoxified’/ruined by enzymes in the intestine or liver. So this thought experiment is just the beginning of a much longer set of experiments that would need to be researched, and then completed, to get anything actionable. At best this is a first pass screening test, and there are better ones out there. [1]
So please don’t do something dumb and then try to sue me. Caveat Emptor. (It’s on you). Disclaimers follow from legal.
Postscript:
I just saw David Sinclair, a noted Harvard researcher, talking in ARDD [4], in the Q&A section say what he thought about rejuvenation by small molecules. His small molecules were targeting epigenetic rejuvenation, not just “do they affect the Hallmarks”, although they likely do that also. And his lab was testing molecules like C1, valproic acid, CHIR-99021, E-616452, tranylcypromine, forskolin, C6NYSA, C4, CHIR-99021, Y-27632, E-616452, TTNPB, and ABT-869. So these are chemicals that would be hard to obtain and may not be safe. He also gave a safety warning. You can see more of that in his ARDD presentation [4] [5] or the paper. [6]
I will take that as encouragement for this idea. And now I am asking myself if perhaps I got this idea from something that he published, as I do try to read everything that he publishes. I should point out something that may not be obvious to those new to aging research. Targeting the Hallmarks might make a person age slower, probably only a few years at best.
But doing epigenetic rejuvenation is the holy grail of aging research because you can conceivably rejuvenate whole cells, tissues, organs, or organisms. Turn back the clock. If that were successful, you might be left with only the damage cleanup from years of being unhealthy/old. And these damages, like occluded arteries, porous bone, buildups of damaged proteins, etc might be partially fixable, by the newly rejuvenated cells! If that sounds exciting to you, then you may want to watch the ARDD video.
immortality unlocked
Great prompts.