English:Aging – Partial reprogramming with preserved identity

aiMOOC

Aging – Partial reprogramming with preserved identity

QR-Code

Introduction

Partial reprogramming is the deliberate, time-limited activation of cell-reprogramming programs with the aim of reversing selected age-associated states without allowing cells to become fully pluripotent. The central scientific claim is stronger than "a molecular clock becomes younger": a useful rejuvenation intervention would have to improve or preserve function while maintaining the correct cellular lineage, avoiding selective expansion of hazardous clones, and producing no dysplasia or organ toxicity. This aiMOOC treats those claims as separate hypotheses rather than as interchangeable meanings of "rejuvenation".

You should read the field adversarially. A result can be internally valid and still leave a different safety question unanswered. For example, a DNA-methylation clock can move toward a younger reference state while a subset of cells temporarily suppresses somatic identity programs. A tissue can show improved regeneration without proving that the same cells were rejuvenated rather than replaced by a fitter subpopulation. A mouse can tolerate one schedule without establishing a general dose, tissue, vector, genotype, age, or species safety window.

The course therefore separates four evidence planes throughout:

1. Molecular age markers: DNA methylation, transcriptomic age, chromatin state, proteomic and metabolomic features.
2. Cell identity: lineage-specific transcription, chromatin, protein expression, spatial architecture, differentiation competence, and transient versus persistent identity changes.
3. Selective advantage: whether reprogramming changes which clones survive, proliferate, dominate, or acquire a competitive advantage.
4. Organ function: physiological performance, injury recovery, behavior, frailty, survival, and tissue-specific function.

The intended audience is an expert colloquium in aging biology, regenerative medicine, epigenetics, stem-cell biology, pathology, quantitative biology, or translational safety science. The course is educational and describes preclinical research design; it is not a clinical protocol.

Foundational distinction: rejuvenation is not pluripotency

The Yamanaka factors OCT4, SOX2, KLF4, and MYC, commonly abbreviated OSKM, can reprogram differentiated cells toward induced pluripotent stem cells. Full reprogramming deliberately erases somatic identity. Partial reprogramming instead attempts to stop or reverse the trajectory before stable pluripotency is established. The difficulty is that aging-associated state and somatic identity are encoded in overlapping regulatory systems, so a clean separation cannot be assumed.

A useful vocabulary distinction is:

1. Continuous identity preservation: a cell remains within its original lineage state throughout exposure and recovery.
2. Identity recovery: a cell transiently leaves or weakens its lineage program, then returns after the factors are withdrawn.
3. Population-level identity preservation: a bulk sample retains the expected average markers, even though a minority subpopulation may deviate.
4. Functional identity: the cell performs the mature function expected of its lineage, not merely expressing a few markers.

These are not equivalent. Gill and colleagues' maturation-phase transient reprogramming in human fibroblasts produced substantial molecular rejuvenation, but fibroblast identity was temporarily lost and then reacquired. Roux and colleagues, using single-cell transcriptomics in mouse adipogenic and mesenchymal cells, likewise observed youthful shifts together with transient suppression of somatic identity programs. These findings make "preserved identity" a measurement problem, not a label that can be inferred from a short marker panel.

Load video

YouTube

YouTube might collect personal data. Privacy Policy

Continue Dismiss

Evidence Framework

Four claims that must be tested separately

Claim	Strong evidence	Weak substitute	Main failure mode
Molecular age is shifted	Orthogonal clocks plus locus-level and multi-omic changes replicated across cohorts	One clock or one biomarker	Clock hacking, distribution shift, batch effects, cell-composition change
Identity is preserved	Single-cell transcriptomic and chromatin identity, protein markers, spatial architecture, lineage competence, and longitudinal recovery	Bulk expression of a few markers	Rare dedifferentiated states hidden in averages
No selective advantage is created	Longitudinal lineage tracing, clone-size distributions, driver-mutation surveillance, and competitive-fitness analysis	No visible tumor during short follow-up	Expansion of a hazardous clone below histological detection
Organ function improves	Pre-specified physiological endpoints with blinded assessment and durable benefit after washout	Younger molecular markers or histology alone	Biomarker improvement without physiological benefit

The most important adversarial rule is: evidence in one column cannot silently substitute for another. If a study measures methylation age but not clone dynamics, the correct conclusion is "molecular age changed; clonal safety was not tested", not "the intervention was safe".

Molecular age markers: what they can and cannot prove

DNA-methylation clocks summarize age-associated CpG patterns. Transcriptomic clocks summarize age-associated gene-expression structure. Other studies measure histone modifications, inflammatory pathways, mitochondrial parameters, serum metabolites, or proteomic features. Agreement across orthogonal layers is more informative than any single clock, but no current clock is a validated surrogate for long-term clinical benefit from partial reprogramming.

Three adversarial questions should accompany every clock result:

1. Reference distribution: Was the clock trained on untreated aging, and is reprogramming pushing samples outside the training distribution?
2. Cell composition: Did a younger predicted age arise because the same cells changed, or because the mixture of cell types or clones changed?
3. Causal relevance: Does changing the measured clock component improve the function that matters, or is the marker merely correlated with age?

Chondronasiou and colleagues showed that one transient OSKM cycle in naturally aged mice shifted DNA methylation, transcription, and serum metabolites toward younger patterns. Importantly, some methylation changes disappeared during recovery while others emerged after factor withdrawal, showing that the post-induction phase is biologically active rather than a neutral waiting period.

Identity: preserved, suppressed, or recovered?

Identity should be treated as a vector rather than a single marker. A strong identity panel includes lineage-defining transcription factors, accessible chromatin at lineage enhancers, mature effector proteins, morphology, tissue position, electrophysiology or secretory function when relevant, and the ability to respond appropriately to physiological cues.

A particularly important challenge comes from single-cell studies. Roux and colleagues found that transient reprogramming could restore youthful gene expression while temporarily suppressing somatic identity programs; some factor combinations separated these two effects better than others. Gill and colleagues found a related pattern in human fibroblasts: substantial rejuvenation was compatible with temporary identity loss followed by reacquisition. Therefore, the phrase "identity preserved" should be reserved for data that actually observe identity during exposure, not only after washout.

For an identity-preserving intervention, pre-specify two tests:

1. Continuous-preservation test: no biologically meaningful excursion from the lineage reference during treatment.
2. Recovery test: after washout, cells return to the lineage reference without residual pluripotency, fetalization, metaplasia, or altered lineage competence.

Passing only the second test supports identity recovery, not continuous preservation.

Selective advantage: the under-measured axis

Partial reprogramming may change survival, proliferation, stress resistance, immune visibility, or niche competition. Any of these can alter clone frequencies. If a rare clone expands because it tolerates reprogramming better, a later sample can look "younger" even if the intervention did not rejuvenate the original population. Conversely, selective loss of damaged clones could be beneficial while still representing population replacement rather than within-cell rejuvenation.

This creates a causal ambiguity:

Observed post-treatment state = within-cell state change + differential survival + differential proliferation + migration + sampling.

Most published partial-reprogramming studies were not designed for high-resolution, long-term clonal surveillance. Short-term absence of teratoma does not exclude a selective advantage that becomes visible months later. This is a major uncertainty that should remain explicit.

Organ function: the highest bar

Organ function requires tissue-specific endpoints. Lu and colleagues used the retina and optic nerve as a central-nervous-system model and reported OSK-associated restoration of youthful methylation and transcription together with axon regeneration and improved visual function in old and glaucoma-model mice. Xu and colleagues later showed that partial reprogramming in the aged mouse subventricular-zone niche increased neuroblast and neural-progenitor proportions and improved production of new neurons, while the effects varied by cell type and pathway.

In 2026, Berdugo-Vega and colleagues reported OSK-mediated partial reprogramming of engram neurons in aged and Alzheimer-model mice, with changes in epigenetic-transcriptional state, excitability, and learning-memory performance. This is a strong functional result in a defined neural population, but it is not a substitute for long-term clonal or whole-organ safety data.

Adversarial Review of Primary Studies

Study-by-study evidence matrix

The table below deliberately records both what each study supports and what it leaves unresolved.

Study	Model and exposure	Molecular age evidence	Identity evidence	Selective-advantage evidence	Functional evidence	Adversarial reading
Ocampo et al. 2016	Progeroid and wild-type mice; cyclic OSKM	Multiple aging hallmarks improved, including epigenetic and damage-associated features	No gross loss of identity in the cyclic low-exposure regime	No long-term clone-resolved analysis	Lifespan extension in a progeroid model; improved injury and metabolic recovery in older wild-type mice	Seminal proof of principle, but progeroid longevity does not establish normal-aging efficacy or clonal safety
Sarkar et al. 2020	Naturally aged human fibroblasts, chondrocytes, endothelial cells and muscle stem cells; transient non-integrating mRNA reprogramming factors	Epigenetic age and multiple age-associated features shifted younger	Cell-type markers were retained in bulk measurements	No long-term clone-resolved follow-up	Improved inflammatory and regenerative cellular phenotypes	Human-cell relevance is high, but in vitro cellular function is not organ function
Lu et al. 2020	Mouse retinal ganglion cells; AAV-delivered inducible OSK	Younger DNA-methylation and transcriptomic patterns	Targeted mature retinal ganglion cells remained non-proliferative, but identity was not exhaustively tracked over years	Post-mitotic targeting reduces but does not eliminate selection questions; no broad clonal surveillance	Axon regeneration and visual-function improvement	Strong coupling of marker and function in one tissue, but not a general systemic safety demonstration
Chondronasiou et al. 2022	Naturally aged mice; one week of low-dose OSKM followed by recovery	DNA methylome, transcriptome and serum metabolome shifted toward younger states	Pancreas showed transient histological alteration that normalized after recovery	No clone-resolved follow-up	Mainly molecular and histological readouts	Demonstrates that recovery is active and that transient tissue disturbance can coexist with later molecular rejuvenation
Gill et al. 2022	Middle-aged human dermal fibroblasts; maturation-phase transient reprogramming	Transcriptomic and epigenetic age reduced substantially	Fibroblast identity was temporarily lost and then reacquired	No long-term clonal test	Collagen expression and migration improved	Rejuvenation and continuous identity preservation are not synonymous
Roux et al. 2022	Aged mouse adipogenic and mesenchymal cells; transient factor combinations with single-cell RNA sequencing	Youthful gene-expression programs restored	Somatic identity programs were transiently suppressed in subpopulations	Single-cell state resolution, but not long-term in vivo clone tracking	Mainly cell-state and myogenic assays	Bulk marker panels can miss rare identity excursions
Browder et al. 2022	Physiologically aged wild-type mice; cyclic OSKM for months	Epigenetic-clock, transcriptomic, metabolomic and inflammatory changes in selected tissues	No gross pathological identity loss reported under the tested single-copy schedule	No barcode-based long-term clonal surveillance	Skin wound-healing and organismal health measures improved; effects were tissue dependent	Long exposure can be tolerated in this genetic system, but "safe" remains model- and schedule-specific
Parras et al. 2023	Multiple reprogrammable mouse strains; continuous OSKM	Biological-age measures could decrease in a modified strain	Continuous exposure drove dedifferentiation-associated pathology in susceptible tissues	Not designed as a clonal evolution study	Liver and intestinal dysfunction, weight loss and premature death within about one week in susceptible models	A direct counterexample to any assumption that more exposure yields more benefit
Cano Macip et al. 2024	Extremely old male wild-type mice; systemic AAV9 inducible OSK with cyclic activation	Epigenetic age changes in tissues and human keratinocytes	No reported gross dedifferentiation in the study	No high-resolution clonal lineage surveillance	Median remaining lifespan and frailty improved	Remaining lifespan approximately doubled from a very late treatment start, but whole lifespan did not double; the highly selected old-mouse design and company-affiliated study require independent replication
Xu et al. 2024	Old mice; whole-body and subventricular-zone-targeted OSKM	Cell-type-specific transcriptomic aging signatures changed	Major SVZ cell types remained identifiable, but pathway responses were heterogeneous	No long-term clone-resolved follow-up	Increased neuroblast/progenitor representation and neurogenesis	Rejuvenation was cell-type specific and not uniformly aligned across all pathways
Berdugo-Vega et al. 2026	Aged and Alzheimer-model mice; OSK targeted to engram neurons	Epigenetic-transcriptional age-associated features improved	Targeted mature neuronal population retained recognizable neuronal function	No long-term clonal oncogenicity study	Learning and memory improved across tested paradigms	Important functional extension, but long-term safety and generalization beyond targeted neurons remain open

Primary-study links: Ocampo et al.; Sarkar et al.; Browder et al.; Parras et al.; Cano Macip et al..

A critical note on effect-size language

The 2024 systemic OSK study in very old male mice reported a 109 percent increase in median remaining lifespan, from about 8.9 weeks in controls to about 18.5 weeks after treatment, with median ages at death of roughly 133 and 142.5 weeks, respectively. Describing this as "doubling lifespan" would be incorrect. The denominator was remaining life after treatment began at 124 weeks, not total lifespan. This distinction matters whenever relative effects are calculated late in life.

The strongest negative-control lesson

Parras and colleagues showed that continuous OSKM can produce rapid liver and intestinal failure and premature death in mice. Excluding reprogramming from liver and intestine attenuated the toxicity, demonstrating that tissue susceptibility is a decisive variable rather than an implementation detail.

A defensible safety claim must therefore be indexed to at least: factor set, expression level, pulse duration, recovery duration, tissue distribution, delivery system, age, sex, genotype, disease state, and observation period.

Designing a Clonal Long-Term Follow-Up

Core hypothesis

The clonal study should distinguish two competing explanations:

H1: within-cell rejuvenation. The same long-lived clones retain lineage identity, shift molecular age markers, and maintain or improve function without gaining abnormal competitive fitness.

H2: population selection. The intervention changes survival or proliferation so that a subset of cells disproportionately contributes to the post-treatment tissue, creating an apparently younger population without demonstrating stable rejuvenation of the original cells.

A study that cannot discriminate H1 from H2 cannot claim clone-neutral rejuvenation.

Experimental architecture

Use an aged, immunocompetent mouse model with lineage-restricted, inducible labeling of the target tissue before reprogramming. The lineage label should be neutral and sufficiently complex to identify thousands of independent clones. Prefer a design in which clone identity can be read together with single-cell RNA and chromatin state; add targeted DNA sequencing for cancer-driver mutations and copy-number abnormalities.

A minimal longitudinal design includes matched untreated aged controls, treatment controls that receive the delivery system without reprogramming factors, a positive-control condition known to produce excessive reprogramming, and at least two independently replicated treatment cohorts. Both sexes should be represented unless the biology of the target tissue makes that impossible.

Sample the same conceptual endpoints at baseline, early exposure, end of exposure, early recovery, and long-term recovery. Terminal tissue collections can be staggered across cohorts. For a mouse study, a defensible long-term horizon should extend far beyond the transient molecular response and cover a substantial fraction of remaining lifespan.

What to measure in each clone

Domain	Clone-resolved readout	Safety interpretation
Identity	Lineage classifier score, lineage-enhancer accessibility, mature effector genes, spatial localization	Detects transient or persistent drift away from the intended lineage
Molecular age	Epigenetic or transcriptomic age score where technically valid, plus orthogonal age-associated pathways	Tests whether age shifts occur within the same clones
Fitness	Clone frequency, growth rate, persistence, death rate, contribution to regeneration	Detects selective advantage or disadvantage
Genomic integrity	Driver mutations, copy-number change, structural variation, vector-integration events where relevant	Detects emerging neoplastic risk
Phenotype	Proliferation, stress response, differentiation output, tissue position	Connects molecular state to biological behavior

Clone-size distributions should be analyzed with more than a single "largest clone" metric. Report the fraction of tissue contributed by the top clones, diversity indices, clone-size Gini coefficient, growth trajectories, and whether expansions are reproducible across independent animals. A clone that is large but stable is biologically different from one that is accelerating.

Long-term falsification criteria

The clone-neutral hypothesis is falsified if any of the following reproducible patterns occurs relative to matched controls:

1. A clone with a new oncogenic driver or copy-number abnormality shows sustained expansion after intervention.
2. Clone-size inequality increases beyond the pre-specified equivalence margin and is driven by treatment-responsive lineages.
3. Molecular rejuvenation is concentrated in expanding clones rather than distributed across pre-existing clones.
4. Identity drift predicts subsequent clone expansion.
5. A histologically abnormal focus maps to one or a few expanding labeled clones.

These criteria are intentionally stricter than "no tumor observed". Tumors are late endpoints; selection can be detected earlier.

Time-Limited Intervention With Stopping Criteria

Candidate preclinical intervention logic

A falsifiable safety-window experiment should not begin by declaring a schedule safe. It should test a bounded exposure matrix. One practical mechanistic design is to compare short, cyclic inductions with full recovery intervals, because this is the regime with the strongest historical mouse evidence. The exact exposure must remain model-specific; a schedule tolerated in a single-copy transgenic mouse is not automatically transferable to a viral vector or a human tissue.

For a mechanistic mouse study, define several exposure levels around the literature-supported short-pulse regime and include a recovery phase long enough to observe whether identity and tissue architecture return to baseline. Escalation to another pulse occurs only after a pre-specified gate is passed.

Gate A: molecular response. At least two orthogonal age-associated measures move in the youthful direction beyond assay noise.

Gate B: identity. No unacceptable lineage excursion is detected by single-cell, protein, and spatial criteria.

Gate C: pathology. No dysplasia, metaplasia, ectopic proliferative focus, or tissue-architecture disruption beyond the pre-specified pathology threshold.

Gate D: function. Organ function is non-inferior to baseline and matched controls.

Gate E: clonal behavior. No treatment-linked expanding clone crosses the pre-specified surveillance threshold.

If any safety gate fails, additional exposure stops even if molecular age markers improve.

Stopping criteria

The following thresholds are proposed as research-operational criteria, not established clinical safety limits. Their purpose is to make the experiment falsifiable.

Domain	Immediate stop	Confirmatory rule
Dysplasia	Any blinded pathologist-confirmed dysplastic focus, teratoma-like structure, or new metaplastic lesion in a treated tissue	Independent pathology review and spatial molecular mapping of the lesion
Functional loss	At least 20 percent deterioration from the pre-specified organ-function baseline or a decline beyond the 95 percent control interval, whichever is more conservative	Repeat measurement when technically possible; stop persists if confirmed
Identity change	More than 1 percent of sampled target cells enter a pre-defined off-lineage or endogenous pluripotency state, or the median lineage score breaches the equivalence margin	Confirmation by a second modality such as protein, chromatin, or spatial transcriptomics
Clonal selection	Any clone exceeds 5 percent of sampled target cells and is still increasing, or doubles in representation across two consecutive post-treatment time points	Deep sequencing plus driver-mutation and copy-number analysis
Systemic toxicity	Pre-specified severe weight loss, organ injury biomarkers, or mortality imbalance relative to controls	Independent safety review; no further induction

The numerical thresholds above are deliberately conservative starting hypotheses for a preclinical experiment. They are not validated universal cutoffs. A central goal of the experiment is to learn whether these margins are too permissive, too strict, or biologically inappropriate.

A falsifiable safety window

Define the safety window as the set of exposures for which all of the following hold after a complete washout and during long-term follow-up:

1. Molecular age: at least two orthogonal age-associated measures improve beyond the pre-specified minimal effect.
2. Identity: the treatment remains within the continuous-preservation margin, or, if the project allows recovery rather than continuous preservation, returns within the recovery margin without residual off-lineage states.
3. Clonality: clone diversity and clone-growth distributions remain equivalent to controls within pre-specified margins.
4. Pathology: no treatment-associated dysplasia, teratoma, metaplasia, or ectopic proliferative lesion is detected.
5. Function: target-organ function is non-inferior during exposure and improves or remains stable after recovery.
6. Durability: the above conditions persist at late follow-up, not only during the first weeks after withdrawal.

The candidate window is falsified if no tested exposure simultaneously satisfies all six conditions. This outcome would be scientifically informative: it would mean that, in the tested model, molecular rejuvenation could not be separated from identity, selection, pathology, or functional risk.

Why washout is part of the intervention

Chondronasiou and colleagues found that some molecular changes disappeared after OSKM withdrawal while others emerged during the recovery period. Gill and colleagues found identity reacquisition after transient reprogramming. These studies imply that washout is not merely an observation period; it is a distinct biological phase that can determine whether the final state is rejuvenated, destabilized, or simply returned to baseline.

Therefore, every safety-window study should specify both an exposure window and a recovery window.

Uncertainty Ledger

The field still has important missing evidence. Preserving these uncertainties is part of scientific rigor.

Question	Current status	What would reduce uncertainty
Can molecular age be reset without any transient identity suppression?	Not established generally; some studies show temporary suppression or loss with later reacquisition	Dense single-cell time courses across tissues and factor combinations
Does partial reprogramming create long-term clonal selective advantage?	Insufficiently tested in most published studies	Prospective lineage barcoding and long-term competitive-fitness surveillance
Are mouse safety windows transferable between tissues?	No; susceptibility differs strongly by tissue and model	Tissue-specific comparative dosing and pathology
Are OSK and OSKM safety profiles equivalent?	No; MYC removal changes risk but does not make all other risks disappear	Head-to-head, matched-delivery, matched-expression studies
Do younger methylation clocks predict durable organ benefit?	Not established as a validated surrogate	Studies linking clock change to independent function and late outcomes
Can systemic delivery be safe for years?	Long-term human data are absent	Multi-year human follow-up and preclinical clonal surveillance
Are benefits maintained after factor withdrawal?	Variable by model and endpoint	Long washout studies with repeated molecular, clonal and functional measures
Is there a human efficacy signal?	As of September 2026, a first-in-human Phase 1 OSK trial for optic neuropathies is recruiting, but no efficacy or long-term safety results are posted	Completed controlled trials and multi-year follow-up

The current Phase 1 study, NCT07290244, is a small first-in-human safety and tolerability study of an investigational OSK therapy for open-angle glaucoma and non-arteritic anterior ischemic optic neuropathy, with planned long-term follow-up. Its existence does not establish efficacy or a general anti-aging indication. ClinicalTrials.gov: NCT07290244

Practical Reading Rules for an Expert Colloquium

When reviewing a new partial-reprogramming paper, ask these questions in order:

1. What exactly was perturbed: OSKM, OSK, another factor set, chemicals, or a downstream pathway?
2. Where was the perturbation expressed, and which cells actually received it?
3. Was the measured "younger" state within cells, between cell types, or between clones?
4. Was identity measured during exposure or only after recovery?
5. Was a rare off-lineage state detectable with the assay used?
6. Was function measured independently of the molecular age readout?
7. Was selective advantage tested rather than inferred from absence of tumors?
8. Was the follow-up long enough for the relevant tissue turnover and cancer latency?
9. Were effect sizes reported in absolute as well as relative terms?
10. Which claims remain unsupported by the experiment?

A strong paper may answer only some of these. The correct response is not to dismiss it, but to restrict the conclusion to the evidence plane it actually measures.

Media for Conceptual Orientation

The epigenetic diagram above helps you distinguish DNA methylation and histone modification from cell identity itself. Reprogramming affects regulatory architecture, but a change in one epigenetic layer does not uniquely determine a safe cell state.

Interactive Tasks

Quiz: Test Your Knowledge

Which result most directly demonstrates molecular age change rather than organ rejuvenation? (A decrease in a validated DNA methylation age estimate) (!Improved visual acuity after treatment) (!Absence of dysplasia at necropsy) (!Stable clone size over one year)

What is the strongest evidence that cell identity was continuously preserved? (Single-cell lineage measures remain within a pre-specified identity margin throughout exposure and recovery) (!Bulk lineage markers are normal after washout) (!The tissue contains no visible tumor) (!A methylation clock becomes younger)

Why can a younger bulk transcriptomic profile be misleading? (Cell composition or clone frequencies may have changed) (!Transcriptomes cannot measure gene expression) (!Young cells never express lineage markers) (!Bulk RNA sequencing measures only DNA)

Which observation most directly tests selective advantage? (Longitudinal expansion of individually labeled clones) (!A younger methylation clock) (!Improved wound closure) (!Higher collagen expression)

What did continuous in vivo OSKM highlight as a major safety concern in susceptible mouse models? (Hepatic and intestinal dysfunction) (!Universal preservation of cell identity) (!Guaranteed lifespan extension) (!Complete resistance to dysplasia)

Why is the recovery phase scientifically important? (Some molecular and identity changes can reverse or emerge after factor withdrawal) (!All reprogramming stops instantly at the molecular level) (!Recovery eliminates the need for pathology) (!Recovery makes clonal tracking unnecessary)

Which statement best describes the 2024 very-old-mouse OSK lifespan result? (Median remaining lifespan increased substantially after treatment began late in life) (!Total lifespan doubled in all treated mice) (!The study proved a human lifespan effect) (!The study eliminated the need for replication)

What would falsify a clone-neutral rejuvenation hypothesis? (Treatment-linked expansion of a clone carrying a new driver mutation) (!A stable clone-size distribution matching controls) (!Improved function without identity drift) (!Concordant molecular age improvement across many pre-existing clones)

Which finding supports identity recovery rather than continuous identity preservation? (Lineage markers are transiently suppressed and later return) (!Lineage markers never leave the equivalence margin) (!Clone sizes remain stable) (!Organ function improves)

What is required for a defensible preclinical safety window? (Molecular benefit and simultaneous passage of identity clonal pathology and function gates) (!A single younger epigenetic clock) (!No visible tumor during treatment) (!A positive result in one tissue)

Memory Game

Epigenetic clock	Statistical age estimate derived from age-associated molecular patterns
Identity recovery	Return to the original lineage state after a transient excursion
Clonal expansion	Increase in the relative contribution of one labeled lineage over time
Dysplasia	Abnormal tissue growth with disturbed cellular architecture
Washout phase	Post-exposure period in which reprogramming factors are withdrawn
Organ function	Physiological performance of the treated tissue or organ

Drag and Drop

Match the correct terms.	Topic
DNA methylation clock	Molecular age
Lineage classifier	Cell identity
Barcode trajectory	Selective advantage
Blinded histopathology	Dysplasia surveillance
Visual acuity	Organ function

...

Crossword Puzzle

Methylation	Which DNA modification is commonly used in epigenetic aging clocks?
Identity	What cellular property must remain lineage appropriate during a preservation claim?
Dysplasia	What term describes abnormal growth with disturbed tissue architecture?
Barcode	What one-word label can uniquely mark a clone for lineage tracking?
Washout	What phase follows withdrawal of the reprogramming stimulus?
Rejuvenation	What process aims to shift age-associated states toward younger function?

Cloze Text

Complete the text.

Partial reprogramming aims to reverse selected age-associated features without reaching stable . A younger molecular clock does not by itself prove preserved . Long-term lineage tracing is needed to detect treatment-related . A safety window must exclude dysplasia while preserving organ . Because some changes appear or disappear after factor withdrawal, the phase is part of the biological experiment. When long-term evidence is absent, the correct scientific conclusion is continued .

Open-Ended Tasks

Easy

1. Evidence-plane annotation: Choose one primary partial-reprogramming paper and label each result as molecular age, identity, selective advantage, organ function, or unmeasured.
2. Clock versus function diagram: Create a one-page visual explaining why an epigenetic clock shift is not equivalent to physiological rejuvenation.
3. Identity vocabulary audit: Find every use of preserved, maintained, recovered, or retained identity in one paper and decide which experimental measurement supports each word.
4. Absolute effect calculation: Re-express a reported relative lifespan effect using absolute weeks or months and explain how the denominator changes interpretation.

Standard

1. Single-cell critique: Reanalyze a published figure conceptually and identify which rare cell states could be hidden by bulk averaging.
2. Pathology stopping rule: Draft a blinded pathology rubric for dysplasia, metaplasia, teratoma-like change, and normal regenerative hyperplasia.
3. Clonal surveillance dashboard: Design a dashboard showing top-clone frequency, diversity, driver mutations, identity score, and molecular age over time.
4. Organ-function preregistration: Choose one organ and write a preregistered primary functional endpoint, measurement schedule, equivalence margin, and failure rule.

Advanced

1. Adversarial replication plan: Design an independent replication of one positive study using a different laboratory, randomized allocation, blinded outcomes, and pre-specified negative controls.
2. Causal decomposition project: Build a statistical model that separates within-cell rejuvenation, cell-composition shift, clone selection, and sampling effects.
3. Long-term lineage experiment: Design a barcoded 12-to-18-month mouse follow-up that tests whether rejuvenated molecular states are distributed across pre-existing clones or concentrated in expanding clones.
4. Falsifiable safety-window protocol: Specify exposure levels, recovery phases, molecular benefit thresholds, identity margins, clonal stopping criteria, pathology rules, and functional non-inferiority conditions, then state the exact result that would falsify the existence of a safe window.

Learning Assessment

1. Evidence separation assessment: Given a new paper abstract, classify every claim into molecular age, identity, selection, pathology, or function and identify any category error.
2. Safety-window inference: Compare three exposure schedules and determine which observations are sufficient to include or exclude each schedule from a candidate safety window.
3. Clone dynamics assessment: Interpret longitudinal barcode-frequency plots and distinguish stable dominance, stochastic drift, and treatment-associated selective expansion.
4. Translational reasoning: Explain why a result from a transgenic mouse cannot be transferred directly to an AAV-based human intervention even when the same factors are expressed.
5. Uncertainty defense: Write a short review conclusion that states what the current data support, what they do not support, and which missing experiment would most reduce uncertainty.
6. Function-surrogate challenge: Propose a case in which molecular age improves while organ function worsens, then design measurements that would detect the discrepancy.

Evidence of Learning

Evidence of learning should include both knowledge and scientific judgment. You should be able to explain the mechanistic difference between full and partial reprogramming; distinguish continuous identity preservation from identity recovery; interpret molecular clocks as biomarkers rather than automatic surrogates of benefit; and explain why cell-composition and clone-selection effects can mimic rejuvenation.

Skill evidence includes the ability to construct an evidence matrix from primary studies, identify hidden denominators in relative effect claims, design longitudinal lineage tracing, pre-specify stopping rules, interpret single-cell heterogeneity, and separate statistical significance from safety relevance.

Product evidence can include an adversarial review, preregistered experimental protocol, clone-tracking dashboard, blinded pathology rubric, organ-function analysis plan, or an uncertainty ledger.

Transfer evidence is demonstrated when you can apply the same four-plane framework to a new reprogramming factor set, another tissue, a chemical reprogramming study, or an early clinical trial without assuming that findings from one evidence plane validate the others.

OERs on the Topic

Useful related articles include Induced pluripotent stem cell, Epigenetics, DNA methylation, Cellular senescence, Stem cell, Lineage tracing, Dysplasia, Oncogene, Neurogenesis, and Aging.

For primary evidence, consult the linked peer-reviewed studies in the evidence matrix rather than relying on secondary summaries alone.

Primary Sources and Current Research Status

The following primary studies form the core evidence base for this aiMOOC:

1. Ocampo et al. 2016: In vivo amelioration of age-associated hallmarks by partial reprogramming.
2. Sarkar et al. 2020: Transient non-integrative reprogramming in naturally aged human cells.
3. Lu et al. 2020: OSK-mediated retinal and optic-nerve rejuvenation in mice.
4. Chondronasiou et al. 2022: Multi-omic rejuvenation after a single transient OSKM cycle in naturally aged mice.
5. Gill et al. 2022: Maturation-phase transient reprogramming in human fibroblasts.
6. Roux et al. 2022: Single-cell evidence for youthful expression with transient identity suppression.
7. Browder et al. 2022: Long-term cyclic partial reprogramming during physiological mouse aging.
8. Parras et al. 2023: Liver and intestinal toxicity during continuous in vivo reprogramming.
9. Cano Macip et al. 2024: Systemic AAV-OSK, frailty, and remaining lifespan in extremely old male mice.
10. Xu et al. 2024: Partial reprogramming of the aged neurogenic niche.
11. Berdugo-Vega et al. 2026: Targeted partial reprogramming of engram neurons and cognitive outcomes.

As of September 2026, the strongest human evidence is still cellular and early translational. A first-in-human Phase 1 OSK study for optic neuropathies is recruiting, with long-term follow-up planned, but no posted efficacy or multi-year safety results are yet available. NCT07290244

Linked Learning Areas

Aging – Partial reprogramming with preserved identity Cellular reprogramming Epigenetic clock Cell identity Lineage tracing Clonal selection Dysplasia Organ function Single-cell sequencing Regenerative medicine Translational safety