A Hypothetical Project: the Fast Track to Partial Reprogramming in Human Volunteers

In a recent post, I suggested that is practical and useful for small organizations to run low-cost clinical trials in large numbers in order to build physician support for treatments for aging that should, by rights, already be in the clinic. The senolytic treatment of dasatinib and quercetin is the most obvious candidate, given its low cost, availability for off-label use, broad, large, and reliable benefits in animal models of aging and age-related disease, and human evidence for efficacy in clearing senescent cells to a similar degree as it does in mice.

Today I'll propose a different angle on early, small trials. In this case the goal is to fast-track access for human volunteers to whole-body partial reprogramming. In partial reprogramming, cells are exposed to Yamanaka factors for a limited time, long enough to reset epigenetic marks to a youthful configuration, but (hopefully!) not long enough for any significant number of cells to lose their differentiated state and become induced pluripotent stem cells capable of forming tumors. In mice, a variety of gene therapy approaches have been used to introduce expression of reprogramming factors, and in the short term the benefits appear interesting enough to follow.

As long-term readers might recall, I've long been dismissive of attempts to adjust epigenetic changes characteristic of aging, as (a) these changes were, in my eyes, a long way downstream from root causes, and (b) the research community was likely to try to make these changes one at a time, with limited individual benefit resulting from any given intervention. What changed my mind on this was the discovery that cycles of DNA damage and repair cause characteristic age-related epigenetic changes. That work needs expansion and replication, but it places some sizable fraction of epigenetic change very much closer to the root causes of aging than previously thought, and makes reversal of those changes a good point of intervention if there is a cost-effective way of doing it. Which there is, in principle, in the form of partial reprogramming.

A great deal of funding is now devoted to the matter of developing partial reprogramming into therapies. NewLimit will be much more nimble than the behemoth that Altos Labs has become, and Turn.bio nimbler still, but I'd still expect a decade to pass between where we are now and the first partial reprogramming therapies becoming available in the clinic in any meaningful sense. These entities will conduct a significant amount of preclinical research, and will be following the standard regulatory playbook thereafter. That takes a long time. Even then, there is a strong chance that the first therapies will be very cautious implementations, such as by being limited to the treatment of retinal diseases and only introduced into the eye.

As an alternative, I believe it would be feasible for a smaller, more agile, directed group to put together a gene therapy for most-of-the-body expression of reprogramming factors and administer it in a small trial of volunteers outside the US, accomplishing that goal in two years or so. The important challenges in reaching that milestone in just a few short years, likely consuming most of that time, are people matters rather than technical matters.

A good approach for a gene therapy capable of only short-term expression appropriate to partial reprogramming would be lipid nanoparticles (LNPs) carrying mRNA encoding the Yamanaka factors, to be injected intravenously in initially low and then ascending doses in human volunteers. The LNPs would be one of the later generation of low immunogenicity variants, while the mRNA would be optimized to reduce immunogenicity in the ways that are presently standard practice in the industry. These are existing technologies, a known sequence for expression of the reprogramming factors, and a matter of running a simple but multi-step manufacturing process that involves two distinct companies and some shipping back and forth.

This gene therapy really doesn't have to be produced using highly expensive, slow Good Manufacturing Practices (GMP) methods in order to be reasonably safe. While some medical technologies do require great care in their manufacture, in this case low-cost research grade materials will do just fine. To ensure correct manufacture at reasonable cost, one runs a quality control study for each batch in cell cultures and in mice, looking for expression of proteins, LNP size, correct sequence of mRNA, and a few other items. That data should be enough to convince anyone that the result is as expected. When injecting into humans, doses should start very low in order to assuage concerns about unexpected immunogenicity.

From a technical perspective, good options for manufacturing of the LNPs are Entos Pharmaceuticals and Acuitas Therapeutics, given what is known of the biodistribution (e.g. not passing the blood-brain barrier, so excluding brain tissue from reprogramming) and safety profiles of their products. For the mRNA there are more companies on the table, but TriLink Biotech is the leading manufacturer, owning some important process patents. The first people matter is to convince the LNP and mRNA companies to act as hands-off manufacturers for a group intending to perform human trials with research grade materials, likely outside the US. There will probably be reputational concerns amongst the leadership of companies that must work closely with the FDA.

All of the other people matters revolve around regulatory approval to perform these trials: which jurisdictions, how the regulatory bodies work in those regions, finding willing clinic owners, and so forth. The Bahamas is a favorable location for a number of groups that are presently setting up clinics for potential anti-aging therapies and would likely be interested in enabling a fast track to partial reprogramming trials. That said, given the good relationship between Bahamas regulators and the FDA I suspect they would require some form of GMP or GMP-like manufacture, significantly increasing costs.

Healthy volunteers in middle age would be a better choice at the outset of this project than those who are very old or very ill, as they will be more resilient in the case of, for example, unexpected immunogenicity. When looking for efficacy, outcomes to measure include epigenetic age, all the omics data that is shown to be rejuvenated by partial reprogramming in mice, and physical function: kidney and liver function, immune function, blood pressure, aerobic capacity, and so forth. The most important question is that of cancer risk, and regardless of how much is spent on clinical trials, or whether they are conducted by large or small organizations, that data will only emerge many years later.

Conducting this project seems to me largely an exercise in organization and finding the funding, with no major technical roadblocks. The big unknown, cancer risk, will remain a big unknown for a long time yet.



First let me say that David's insight concerning the epigenetic consequences of cycles of DNA repair is very smart and almost certainly correct... qualitatively. However, when it comes to the magnitude of the effect, I am hesitant about the extrapolation to humans from David's findings in yeast. I know he is assiduously working to answer that question.

But that's not actually what I want to focus on here, because (as is my wont) I am more interested in how to fix damage than in how it arises.

The main thing that makes me skeptical about partial reprogramming as a useful modality for human rejuvenation is its risk of tumorigenesis. It is well known that tumour cells typically have a more "undifferentiated" gene expression profile than the cells from which they arose, which says that dedifferentiation can in some circumstances be tumorigenic. But of course the Yamanaka factors are precisely what naturally turn gametes into a zygote and beyond, i.e. not into tumours, so there must be ways around that risk. But... can indiscriminate application of those factors to an adult get around it? My main concern here is telomerase, which:

- is turned on by the Yamanaka factors
- is (as was shown decades ago by Shay and Wright) typically the LAST thing to change during tumorigenesis (simply because basically everything else that needs to change entails INactivating a gene, which mutationally is far easier)
- is necessary (unless ALT happens) for a tumour to become big enough to be clinically detectable, let alone metastatic
- is typically turned on in cancers by an accumulation of epigenetic changes to its promoter region, rather than by a single mutation, hence is statistically sure to be nearly turned on in a great many not-yet-full-blown-tumour cells before it gets properly turned on in any such cell.

Thus, until such time as we have cancer totally licked (and here, as usual, I will highlight the amazing drug 6-thiodG, which was also discovered by Shay and is being taken forward by MAIA - look it/them up), I strongly believe that any partial reprogramming approach that can potentially activate telomerase must be very painstakingly accompanied by the absolute best possible technology for early detection of cancer, such as high-sensitivity detection of circulating tumour cells. Even then there is a big risk, because shedding of cells by the primary tumour may often NOT precede telomerase activation (the Shay/Wright study only looked at the primary tumour).

So I'm strongly in favour of the exploration of dedifferentiation factors that (even when expressed constitutively) take cells less far back: that make them more regenerative but do not activate telomerase. Full disclosure - as is well known, I worked for some time at AgeX, which is looking at exactly that - but I have no financial stake there.

Posted by: Aubrey de Grey at February 14th, 2022 4:24 PM

@Aubrey de Grey

Good you see you here!! You have been missed!!

Thanks for all this great insight, but of course what everyone really wants to ask you is -

When will you be back running things at SENS???

Without you the organization has no soul

Posted by: James Petrov at February 14th, 2022 4:59 PM

Welcome back, Aubrey!!

Posted by: mcmp at February 14th, 2022 5:33 PM

@ Aubrey de Grey,

Regardless of whether you return to SENS or decide to start something new, I suspect there is an army of people willing to back the man who started it all. Be encouraged!

Posted by: Morpheus at February 14th, 2022 6:26 PM

@Aubrey & everyone:
Sebastiano lab (as you I'm sure remember) found that OSKMLN reversed 8 hallmarks but not telomeres.
Jacob Kimmel's lab at Calico and Daniel Ives of Shift Bio have explicitly already begun work on reversing epigenetic age with minimal de-differentiation. I expect separating these 2 effects will also be a large focus for Altos, Retro, NewLimit, and probably others.

Given this flurry of new work that I think is likely to produce increased understanding of cocktails that better reduce dedifferentiation, probably within 3-5 years, I suspect for those without urgent need that a wait a little bit approach (maybe while laying some of this human/organization/bureaucratic groundwork in parallel) is probably a good idea.

But the main point of the post that there aren't insurmountable unknown technical hurdles is an important one. Aubrey's point can be generalized to not only monitoring for cancer as important but also more broadly the problem is detecting when there has been a meaningful beneficial effect at the lowest dosage without having to wait a long time, including monitoring for different degree of effect in different tissues/organs. We need less invasive biopsies and innovations like Morgan Levine's systems epigenetic clocks from blood samples (which should be coming soon I think).

Posted by: Karl Pfleger at February 14th, 2022 10:33 PM

Always with the telomerase, Aubrey.

My take is somewhat different. The cycles of epigenetic maintenance that you and David describe following DNA damage most likely also affect replication, i.e. new DNA strand produced, DNMTs add appropriate methylation, TETs then remove the excess, is a highly imperfect process (compared to DNA repair). We know that with age some regions gets undermethylated (relative to youth) and some promoters are over methylated. But what Horvath showed looking at patterns across many species, is that it is largely the over methylation that is correlated with age. It is especially interesting that the promoters affected are highly enriched for development/differentiation genes. Given their long life the process I describe will especially affect stem cells. So with age we should expect to see a kind of block on differentiation. Stem cells get stuck as stem cells, or if they differentiate do it only partially. This is probably the real reason cancer incidence rises with age, and shows how cancer and aging are intertwined. From this perspective de-differentiation isn't at all what you want to do - you want to force MORE differentiation. Of course you will run into a lack of proliferative potential without telomerase; for some reason people forget that there can be no reversal of aging without maintaining telomeres.

Posted by: Mark at February 15th, 2022 6:12 AM

Reason, the Acuitas LNP biodistribution has been well characterized in mice and non human primates. These LNPs are largely taken up by hepatocytes, with relatively minor transfection of liver endothelial cells and splenic cells. So your proposal could be rephrased as "a fast track to reprogramming hepatocytes." If limited to hepatocytes, would this reprogramming result in a functional phenotype that would be easily measured and desirable enough to warrant the risk? Especially outside of the context of an existing chronic liver disease (where the cancer risk would be even more pronounced)?

Posted by: Skeptical at February 15th, 2022 11:05 AM

@Skeptical: Yes, their approved, most prominent LNPs are liver-only, and very useful for people who are targeting the liver. They have a whole bunch of others at various stages of development and data.

Posted by: Reason at February 15th, 2022 11:17 AM

Awesome idea! Have you thought of approaching Vita DAO?

Could potentially be a great way of getting the initial funding

Posted by: Daniil Gorbatenko at February 15th, 2022 12:15 PM

I can't imagine that a healthy middle-aged person would want to risk cancer. Apes, on the other hand... It's not very ethical but perhaps to speed things up without killing humans in the process, someone could start a small trial of partial reprogramming in our closest relatives and see how they fare with cancer. Given a large enough number (perhaps a dozen?) of test subjects we could have solid answers in a few years. In any case, no later than if we were using humans as guinea pigs and without the risk - not only to the guinea pigs but to the field as a whole. Remember what happened in the late 90s with gene therapy...

Posted by: Barbara T. at February 15th, 2022 5:06 PM

@Brabara T - it could be tested topically (which would maybe be less expensive) on the skin of Monkeys to see if it had a cosmetic effect. If it did, it could spawn an underground industry similar to those for other cosmetic drugs.

What effects you would look for in aging Monkey skin that could translate to humans I don't know? Increased dermal thickness? Decreased shininess?

Posted by: jimofoz at February 16th, 2022 12:21 AM

For those worried about "too much age reversal" and cancer - it appears that the "clock" is actually made up of many loosely independent modules - the most sensitive to reprogramming are involved with disease (all-cause mortality is the most sensitive, cancer is second most sensitive) - while the modules involved with cellular differentiation are considerably less sensitive or not at all sensitive to reprogramming.

How cool is this! From Morgan Levine's Lab - someone inform Aubrey! https://www.biorxiv.org/content/10.1101/2022.02.13.480245v1

Posted by: Max Darling at February 16th, 2022 10:18 AM

@Aubrey - I thought the old model of cancer being the process of randomly and over time acquiring several mutations in a cell, hence the rising rates of cancer in later life power law, had been thrown out in favour of a theory of decline in immune system surveliance?

If this is the case, more cancerous cells in a young animal/human with a robust immune system shouldn't lead to rising rates of cancer unless the epigentic reset allows more cancer cells to achieve "escape velocity"? Although we don't have a good model of the probability distribution curve of this happen, the data in mice seems to suggest that it is an extreme tail end risk at best.

Posted by: jimofoz at February 16th, 2022 12:10 PM

I know from my work and research with hematopoietic stem cells that genetic defects not only can trigger apoptosis but also lineage differentiation. Interestingly in a way that differentiation happens in directions where the defect genes aren't functionally of necessity and are deactivated. Therefore it's once more safe to say that differentiation is associated with accumulation of genetic trash. Asymmetric divisions in the stem cell pool can be expected to preferably keep the high quality genetic material upstream in the stem cell pools and send "trash" downstream. I propose to start rejuvenating the immune system including the thymus, addressing tumor response and senolytic capacity for example of T-cells, then the bone marrow and other stem cell pools/niches and to avoid "empowering the trash" with significant risk of tumorigenesis as Aubrey pointed out above. In other words, a specific sequence and targeting of niches should be considered when using epigenetic reprogramming in vivo. BTW the tissues just mentioned lend themselves to manipulation outside of the body and applications can be manifold in terms of therapeutic approaches and end of life interventions. Good for acquiring funding.

Posted by: Wolfgang Möcklin at February 16th, 2022 5:54 PM

@ Reason - TBH the science being discussed in the comments above is well beyond my level of understanding. Reason, it seems to me, you have recently started commenting more on possibilities / the desirability to, a bit more boldly than the norm, push forward with clinical trials on humans. This could very well be an area where a new innovative, but still safe, approach is needed. I guess it will need to be driven by someone with the appropriate qualifications, foresight & drive (perhaps yourself), paired with someone with the appropriate organisational skills & funding. That someone might be Michael Greve. After all he did say " My ambition is clear: I want to be able to use what we find out for myself, for my life." & he is 57 years old so I guess he's not interested in waiting 30 years for treatments to come along.

Posted by: TB at February 20th, 2022 10:02 PM

Wow, I missed a lively discussion last week!
@Aubrey: On Sinclair's model of double strand break-driven epigenetic decay: it seems that this would have much greater clarity if we understood what DNAm clocks were telling us about biological aging. To the extent that they reflect epigenetic forces that are drivers of hallmarks of aging, then I'd think the Sinclair hypothesis gains in relevance.

You mention yeast, but what do you think of the work with the ICE mice in their BioRxiv manuscript?
These are the ones with an inducible PpoI restriction enzyme to produce double-strand breaks in noncoding regions of DNA, hence Inducible Changes to the Epigenome. I've been watching to see if they ever get the damned thing published, because I'm concerned I've missed a significant flaw that the reviewers caught. If you trust the data, though, they claim that three weeks of accelerated epigenetic decay produced a broad progeria:

Alopecia, hair graying, weight loss, reduced physical activity, frailty, kyphosis, optic nerve degeneration, skin thinning, sarcopenia, memory impairment, and CNS inflammation.

Obviously there are thousands of ways to break things that produce a progeria phenotype, but this seems unique in that
1. Methylation clocks can be constructed that predict biological age with high accuracy
2. Double strand breaks occur during normal living
3. DSB repair appears to produce epigenetic changes including changes in methylation that may drive DNAm clocks
4. Epigenetic decay similar to that produced by DSB repair is among the recognized hallmarks of aging and is a proposed cause of aging
5. Transient, physiologically relevant increases in the level of DSB's in vivo drives epigenetic decay, increases DNAm age, and produces a progeria-like phenotype
6. Yamanaka factors can repair epigenetic decay, producing an induced pluripotent state
7. In data from multiple labs, transient expression of Yamanaka factors in vivo appears to reverse some aging phenotypes

Although it doesn't recapitulate every facet of aging exactly, DSB-driven progeria seems to capture most of them. I agree with the authors' conclusion that the immediate value of this work might be the creation of a much better model of accelerated life testing. All models are bad, but some models are useful. This model, or models like it, should be evaluated in several labs ASAP if it isn't already happening.

(Does anyone know if there are other labs using the ICE mice, or similar models? I've written Sinclair a couple times, but haven't received a reply. )

I agree with cancer concerns following reprogramming, but in the same way that it may be possible to separate the epigenetic age reversal from de-differentiation, hopefully it'll be possible to separate the telomerase induction, too.

@Reason @Barbara: Regarding informal tests of reprogramming, I'm wondering if a compromise between non-human primates and healthy adults might be to turn to our canine and feline friends. I think some people would take the risk on their pets if there is a reasonable likelihood that they could contribute to science and possibly extend the healthspan of their pets. Then there's Sus domesticus - pigs. I assume if people are willing to raise them just for food, then it's acceptable to try to extend their healthspan without their consent. Jean Hébert has also pointed out that we could produce non-human primates lacking a neocortex with only one or two gene knockouts. They would capture almost everything except potential cognitive effects without the ethical dilemma. Lastly, back to humans: there's the suggestion that reprogramming may be effective in Werner's syndrome, which is a pretty awful genetic disease.

@Wolfgang: I'll second this call for the use of ex vivo manipulation of tissues. We have thousands of unusable donor organs that are wasted every day. Let's put these to good use. I'll go one further and mention the concept of pseudo-clinical study. The first kidney from a genetically modified pig was tested last year by tying it to the circulatory system of a brain dead person who had given prior consent to have their body used in this way. This type of creative use of existing human tissues could really improve our hit rates for emerging therapies. We can keep legally deceased bodies alive for days, even weeks. Could we do it long enough to gather evidence of efficacy and safety of candidate reprogramming strategies?

Posted by: Robert Cargill at February 21st, 2022 7:24 PM

To optimize the outcome of epigenetic reprogramming approaches, we could try first if we can find approaches for increasing genetic and cellular quality. For example in fast replicating tissues can yield fast results even in vivo. To understand better, gain a headstart for subsequent dedifferentiation interventions and later apply both together. Increasing immune surveillance, senolytics, reduction of inflammation, simulating "fasting cycles", NAD, nutrient/vitamin optimization, mitochondrial rejuventation (another approach:rejuvenate mitos first), Histone Deacetylase Inhibitors, helicase activation (opposite of Werner's syndrome), oxidative eustress, shielding from oxidative and radiation distress, stem cell recruitment, induction/restoration of a juvenescence-friendly extracellular matrix (ECM). An optimzed environment might be necessary for mitigating damaging effects on reprogrammed cells during sensitive phases where the risk of triggering malignancies is biggest. Development of a quality control matrix for this "headstart protocol for epigenetic reprogramming" would be another issue. Epigenetic clocks of course, sings of accumulated genetic damage, quantitative and qualitative distribution of the cellular compartments (stem cell markers, mitotic compartment, senescent cells, tissue organisation versus pathology), markers of inflammation, ECM quality, mitochondrial age and density, healthy transmembrane potentials (?), biophotons (?) eventually. Epigenetic reprogramming embedded into a comprehensive approach for body regeneration, mitigating the risks and doing more good than bad. Still additionally there need to be developed better means - the fire brigade so to speak - for addressing catastrophic events of tumor formation. Maybe in the future we'll be able, instead of acting in the dark in terms of tracing cellular events in vivo - to see everything happening in realtime. A big goal needs bright minds and many approaches. Therefore we need advanced analytics and imaging methods. @Altos

Posted by: Wolfgang Möcklin at February 23rd, 2022 6:54 AM
Comment Submission

Post a comment; thoughtful, considered opinions are valued. New comments can be edited for a few minutes following submission. Comments incorporating ad hominem attacks, advertising, and other forms of inappropriate behavior are likely to be deleted.

Note that there is a comment feed for those who like to keep up with conversations.