I really don’t like the portmanteau “cross-talk”
https://portlandpress.com/clinsci/article/140/6/1137/237626
chatGPT:
Paper
Emelyanov & Bulavin, “A cross-talk between p16High senescence and cellular reprogramming”, Clinical Science, 2026. This is a perspective/review, not a new primary experimental paper. It mainly synthesises recent work, especially the authors’ own 2023–2026 studies, on how p16High senescent cells regulate OSKM-mediated reprogramming and partial rejuvenation.
Summary
The paper argues that cellular senescence and reprogramming are not simply opposing processes. Instead, they interact bidirectionally. OSKM reprogramming can generate or encounter senescent states, while senescent cells can either block or, in some contexts, assist reprogramming.
The authors focus specifically on p16High senescence, rather than treating all senescent cells as equivalent. They argue that p16High cells are qualitatively different from p21High or transient injury-associated senescence. p16High fibroblasts are described as deeply entrenched, resistant to full OSKM reprogramming, and poor contributors to iPSC formation or teratoma formation. This resistance is not explained merely by cell-cycle arrest; rather, it reflects a more stable anti-plasticity state.
Mechanistically, the paper proposes three linked barriers:
-
Epigenetic entrenchment: p16High senescent fibroblasts have stress- and AP-1-dominated enhancer landscapes, with depletion of developmental and pluripotency-associated regulatory regions. This may sequester or misdirect reprogramming factors away from productive pluripotency sites.
-
Metabolic restriction: p16High cells show mitochondrial dysfunction, altered NAD+/NADH, chronic mTOR activation, and reduced availability of metabolites needed for chromatin remodelling, including acetyl-CoA and α-ketoglutarate.
-
NNMT–SAM axis: the authors emphasise elevated NNMT activity in p16High fibroblasts, reducing intracellular SAM. Since SAM is the universal methyl donor, this restricts methylation-dependent epigenetic remodelling. Importantly, this effect is proposed to act not only inside p16High cells but also non-cell-autonomously, via SASP/JAK–STAT3-mediated induction of NNMT in neighbouring p16Low cells.
Figure 1 summarises this model: in a senescence-proficient environment, p16High cells produce NNMT-inducing SASP signals, reduce SAM availability, and suppress iPSC formation and partial reprogramming. In a senescence-deficient environment, SAM balance is restored, iPSC formation improves, 2C-like/totipotent-like features emerge, and partial reprogramming in old tissues becomes more efficient.
The paper then discusses a subtle point: senescence can sometimes promote reprogramming. Injury-associated or transient inflammatory senescence, especially involving p21High cells or senescence-like immune cells, may provide IL-6/STAT3 and related signals that help plasticity. But chronic inflammation or persistent p16High fibroblast senescence tends to block successful reprogramming. Thus, the effect of senescence depends on cell type, timing, tissue context, and inflammatory duration.
A major section concerns totipotent-like states. Removal of p16High cells can allow reprogrammed cells to access states resembling 2-cell-like cells, with high Dux and MERVL, but also enhanced Line-1 and Stella/Dppa3 compared with classical 2C-like cells. The authors argue that this is not ordinary embryogenesis, because p16High cells are largely absent in embryonic stages; rather, it is a reprogramming-specific constraint imposed by adult senescence.
The final section links this to partial reprogramming for rejuvenation. Short pulses of OSK/OSKM can restore some youthful epigenetic and functional features without full dedifferentiation. The authors argue that this works best before cells cross into a deeply entrenched p16High state. However, they also note apparently contradictory evidence: targeted partial reprogramming of p16-positive cells in aged mice can improve wound healing, inflammatory gene expression, haematopoietic profiles, and lifespan without obvious tumour increase. They suggest this may reflect plasticity in some p16High cell types, especially immune cells, rather than fibroblasts.
The conclusion is that rejuvenation may require selectively dismantling senescence-associated transcriptional, epigenetic, and metabolic barriers, rather than simply erasing cell identity or indiscriminately killing all senescent cells.
Novelty
The main novelty is the paper’s integration of p16High senescence, reprogramming resistance, metabolic methyl-donor control, and totipotent-like plasticity into one model.
The strongest conceptual contribution is that p16High senescence is framed as an active regulator of plasticity, not merely a passive marker of old or damaged cells. The paper argues that p16High fibroblasts create a hostile reprogramming environment through AP-1/stress enhancers, SASP signalling, NNMT induction, and SAM depletion.
A second novel element is the emphasis on non-cell-autonomous metabolic epigenetic control. The idea is not just that senescent cells themselves cannot reprogram, but that they can reduce the plasticity of neighbouring, otherwise reprogrammable p16Low cells by altering NNMT/SAM metabolism.
A third notable point is the proposed relationship between p16High clearance and access to totipotent-like states. The paper suggests that removing p16High cells does not merely increase iPSC efficiency; it may alter the attainable cell fate space, allowing 2C-like or totipotent-like features that are normally suppressed.
A fourth useful contribution is the distinction between different senescent states. The paper resists the simplistic “senescence is bad” model: transient injury-associated senescence may help reprogramming, whereas persistent p16High mesenchymal/fibroblast senescence constrains it.
Critique
The paper is valuable, but it is a perspective with a strong interpretive stance, not a neutral systematic review. Much of the framework is built around the authors’ own recent work, especially on p16High cells, NNMT, SAM, and reprogramming. That does not make it wrong, but it means the model needs independent replication across tissues, species, ages, and reprogramming systems.
A central weakness is that p16High is treated as a major functional category, but p16 expression is not a single cell state. p16High fibroblasts, p16High immune cells, p16High epithelial cells, and p16High progenitor-like cells may behave very differently. The authors acknowledge this, especially in discussing immune cells, but the model still risks overgeneralising from p16High fibroblast biology.
The NNMT–SAM mechanism is interesting and directly relevant to methylation-dependent chromatin plasticity, but the paper may understate how many other metabolic pathways could be equally important. Acetyl-CoA, α-ketoglutarate, NAD+, mitochondrial redox, mTOR, autophagy, one-carbon metabolism, and methionine flux are all connected. SAM depletion may be a key bottleneck in some settings, but not necessarily the master control point for reprogramming resistance.
The link to totipotent-like states is intriguing but should be handled cautiously. “2C-like”, “totipotent-like”, and “expanded potential” states are difficult to define rigorously. Marker expression such as Dux, MERVL, Line-1, and Stella/Dppa3 is suggestive, but developmental potential is the harder test. The paper’s Table 1 is useful, but it also shows that p16High-deficient totipotent-like cells are not identical to classical 2C-like cells.
The paper’s therapeutic implications are also not straightforward. If p16High cells block partial reprogramming, senolytic or senomorphic pre-treatment might help. But the authors also point out that senolytics can impair early reprogramming independently of senescent-cell clearance. That makes the practical design problem much harder: timing, dose, target cell type, and drug mechanism may determine whether senolytics help or harm.
The oncogenic-risk discussion is appropriate but incomplete. The paper suggests that p16High cells are resistant to full reprogramming and may therefore reduce teratoma or tumour risk. But removing p16High barriers or increasing plasticity in old tissues could, in principle, allow mutated aged cells to enter dangerous partial states. The authors mention mutation burden, but the field still lacks a convincing safety framework for combining senolysis, metabolic rewiring, and OSK/OSKM in aged organisms.
Overall assessment
This is a strong and useful synthesis for the partial reprogramming field. Its best idea is that senescence is not merely a brake on proliferation, but a metabolic-epigenetic brake on cell fate plasticity. The p16High–NNMT–SAM model is especially interesting because it connects senescent-cell burden to the ability of neighbouring cells to remodel chromatin.
The main caveat is that the framework is still relatively early. It needs stronger causal testing outside fibroblast-heavy systems, better distinction among p16High cell types, and clearer evidence that manipulating NNMT/SAM improves rejuvenation safely in aged tissues. The paper is therefore best read as a hypothesis-shaping perspective rather than a settled mechanistic account.