The two proteins in question are PARK7, also known as DJ-1, and PINK1. Both are best known to the public through their connection to familial Parkinson’s disease, where loss-of-function mutations sensitize neurons to oxidative damage and mitochondrial dysfunction. In the new work, the Tsukuba team asked whether these same proteins do something entirely different in a very different cellular context: the dramatic transformation that occurs when mouse embryonic fibroblasts are pushed toward pluripotency by reprogramming factors. PARK7 is a multifunctional, redox-sensitive protein long implicated in buffering oxidative stress and modulating transcription, and earlier reports had suggested it could associate with the promoter of Nanog, a master gene of pluripotency, and even suppress reprogramming efficiency. That hint of an inhibitory role made PARK7 an intriguing suspect.

To test the idea, the researchers depleted Park7 in mouse embryonic fibroblasts undergoing reprogramming and monitored the emergence of Nanog-positive colonies, a standard readout of successful pluripotency induction. The result was striking: knocking down Park7 enhanced colony formation, with the effect most pronounced during the early phases of reprogramming. In other words, removing PARK7 made it easier for somatic cells to cross the initial barriers on the road to pluripotency. This positioned PARK7 not as a passive bystander but as an active regulator of the process, and it raised the obvious question of how a protein best known for antioxidant defense could, when removed, actually improve the efficiency of cellular rejuvenation.

The answer, it turns out, lies in reactive oxygen species. Reprogramming is metabolically traumatic: a cell that has spent its life burning glucose efficiently through mitochondrial oxidative phosphorylation must tear down and rebuild its entire energy economy, ultimately adopting the glycolytic, lactate-producing metabolism characteristic of embryonic stem cells. That transition is accompanied by bursts of reactive oxygen species, which at controlled levels act as signaling molecules that help drive the switch. Using paused iPSCs, a valuable experimental system in which cells are held in a partially reprogrammed state and can be interrogated in detail, the team showed that depleting Park7 increased both mitochondrial and cytosolic ROS, stabilized HIF-1α, the hypoxia-inducible transcription factor that commands the glycolytic program, promoted the expression of glycolytic genes, and upregulated pluripotency markers.

The causal chain became clear when the researchers added antioxidants to the culture. Antioxidant treatment substantially blunted all of these effects, indicating that the enhanced reprogramming phenotype seen upon Park7 loss was driven by ROS-mediated metabolic remodeling rather than by some unrelated function of the protein. In essence, PARK7 normally keeps a brake on ROS levels, and that brake, paradoxically, restrains the very oxidative signals that help push a cell toward pluripotency. When the brake is released, ROS rise, HIF-1α accumulates, glycolysis genes switch on, and the cell slides more readily into its new pluripotent identity. It is a vivid demonstration that in reprogramming, a little oxidative stress is not the enemy but an ally.

The study also uncovered a second, compensatory layer of redox regulation. The elevated ROS triggered by Park7 depletion activated NRF2, the master transcription factor of antioxidant defense, switching on NRF2-dependent antioxidant transcriptional programs. This reveals a built-in feedback loop: as ROS climb and drive metabolic remodeling, the cell simultaneously mounts a protective antioxidant response, presumably to keep the oxidative surge within tolerable limits. The interplay between these pro-signaling and protective arms of the redox system highlights how finely balanced the reprogramming environment must be, and it suggests that manipulating this balance could be a way to tune reprogramming efficiency in the laboratory.

Perhaps the most consequential discovery, however, concerns PINK1. The team found that Park7 depletion reduces the expression of Pink1, and that knocking down Pink1 phenocopies the loss of Park7, producing the same ROS elevation, metabolic shift, and pluripotency marker upregulation. Most tellingly, restoring wild-type PINK1 reversed the increase in pluripotency markers caused by Park7 loss. Together, these experiments establish PINK1 as the essential downstream mediator of PARK7 in this cascade. The two Parkinson’s-linked proteins, it appears, operate as a functional axis: PARK7 upstream, sustaining PINK1 expression, and PINK1 downstream, executing the ROS-suppressing and metabolic functions that keep reprogramming in check.

This PARK7–PINK1 axis is more than a curiosity of stem cell biology. It suggests that the same molecular machinery whose failure contributes to neurodegeneration also serves as a gatekeeper during the acquisition of pluripotency, coordinating three intertwined processes: suppression of reactive oxygen species, glycolytic metabolic remodeling, and the expression of pluripotency genes. For researchers trying to generate iPSCs more efficiently and more safely, the findings point to concrete levers. Modulating PARK7 or PINK1 activity, or carefully titrating ROS with antioxidants, could in principle shift cells along the reprogramming trajectory. At the same time, the work cautions that redox manipulation is a double-edged sword, since the NRF2 response shows the cell actively defending itself against the very signals that promote its transformation.

There are also broader implications for understanding disease. PARK7 and PINK1 are central figures in the mitochondrial quality control literature, and this study ties them to HIF-1α-driven metabolic switching and NRF2-mediated antioxidant transcription, linking several of the most important stress-response pathways in cell biology into a single regulatory circuit. Whether the axis operates similarly in human cells, and whether it influences the quality and genomic stability of the resulting iPSCs, are questions the field will now want to pursue. The experiments were conducted in mouse embryonic fibroblasts and paused iPSCs, so translating the mechanism to human reprogramming systems will be an essential next step.

What the study delivers today is a mechanistic framework with immediate experimental value. By showing that PARK7 restrains reprogramming primarily through PINK1-dependent control of ROS, the Tsukuba team has converted a long-standing puzzle about DJ-1’s role at the Nanog promoter into a coherent signaling pathway that connects redox balance, metabolism, and gene expression during one of biology’s most dramatic cellular transformations. As regenerative medicine moves closer to the clinic, understanding such gatekeepers, and learning when to release them and when to respect them, may prove as important as the reprogramming factors themselves. The full open-access study is available in Cellular and Molecular Life Sciences for readers who wish to explore the experimental details.

Subject of Research: The role of the PARK7–PINK1 axis in redox balance and metabolic remodeling during induced pluripotent stem cell reprogramming

Article Title: A PARK7–PINK1 axis coordinates redox balance and metabolic remodeling during somatic cell reprogramming

Article References: Mensah, E. O., Burramsetty, A. K., Kamata, H., Jiang, Z., Fukuda, A., Hisatake, K., & Nishimura, K. (2026). A PARK7–PINK1 axis coordinates redox balance and metabolic remodeling during somatic cell reprogramming. Cellular and Molecular Life Sciences. https://doi.org/10.1007/s00018-026-06468-8

Image Credits: AI Generated

DOI: 10.1007/s00018-026-06468-8

Keywords: PARK7, DJ-1, PINK1, iPSCs, reprogramming, pluripotency, reactive oxygen species, HIF-1α, NRF2, metabolic remodeling, redox biology, mitochondria

The intellectual foundations of the field reach back to classic experiments showing that the differentiated state of a cell is not permanently fixed. Nuclear transfer and cell fusion studies demonstrated that somatic cell identity could be reset, and the discovery that the transcription factor MyoD could convert mouse fibroblasts directly into myoblasts established the concept of direct reprogramming. The landmark identification by Takahashi and Yamanaka of four transcription factors capable of generating induced pluripotent stem cells, or iPSCs, then opened a path to generating nearly any cell type in the laboratory. Several iPSC-derived therapies have since received regulatory approval or entered clinical trials. Yet the review emphasizes that iPSC-based approaches face persistent challenges, including tumorigenesis risk, inefficient differentiation, and poor survival and integration of transplanted cells within host tissues.

Direct reprogramming, also called transdifferentiation, offers an alternative by converting one somatic cell type into another without passing through a pluripotent intermediate. Even this strategy, however, has struggled when performed in vitro. Cells reprogrammed in culture often remain incompletely or immaturely converted, show poor tissue homing after transplantation, and suffer low long-term survival. In vivo direct reprogramming flips the logic entirely: rather than making cells in a dish and delivering them, researchers deliver reprogramming factors directly to the injury site, converting tissue-resident cells in their native environment. The review attributes the superior maturity of in vivo reprogrammed cells to pro-developmental and regenerative cues within the tissue microenvironment, which help drive cells toward fully functional states and promote integration with surrounding tissue.

At the molecular level, direct reprogramming works by rewriting gene regulatory networks, the hierarchical webs of transcription factors, signaling pathways, non-coding RNAs, and chromatin remodelers that define cellular identity. Central to this process are pioneer factors, a class of transcription factors capable of binding compacted heterochromatin. During neuronal reprogramming of fibroblasts, for example, the pioneer factor Ascl1 binds neural-lineage targets regardless of chromatin accessibility, locally opens the chromatin, and recruits the master regulators Brn2 and Myt1l to establish a neuron-specific regulatory circuit. Pioneer factors begin by scanning chromatin through non-specific electrostatic interactions, then distort and unwrap DNA at specific binding sites, recruiting co-regulators that remodel histones and create a transcriptionally permissive state. Successful reprogramming culminates in activation of peripheral network components that govern terminal fate and maturity, such as the factors Mafa and Errγ, which drive glucose-responsive insulin secretion in reprogrammed pancreatic beta cells.

The reprogramming toolkit extends well beyond transcription factors. Small molecules can activate developmental signaling pathways and modify the epigenetic landscape; a cocktail of CHIR99021, SB431542, DAPT, and LDN193189, for instance, converted human fetal astrocytes into cortical neurons by suppressing glial programs and activating neural master regulators. MicroRNAs offer another route: a combination of miRNAs 1, 133, 208, and 499 reprogrammed cardiac fibroblasts into cardiomyocytes by reinforcing cardiac lineage programs while silencing fibroblast signatures. Short hairpin RNAs and antisense oligonucleotides against the RNA-binding protein PTB have converted astrocytes into neurons, with stereotactically injected PTB antisense generating dopaminergic neurons that alleviated Parkinsonian symptoms in mouse models. Even exosomes, tiny membrane vesicles released by cells, can carry reprogramming cargo; insulinoma-derived exosomes enriched in miR-127 and miR-709 converted exocrine cells into pancreatic beta-like cells in vitro.

Delivering these factors into living tissue remains one of the field’s central engineering problems, and the review systematically compares the options. Viral vectors dominate current practice. Adenoviruses carry large payloads and achieve high transduction of dividing and non-dividing cells, as demonstrated when delivery of the transcription factors Ngn3, Pdx1, and Mafa converted more than 20 percent of infected pancreatic exocrine cells into insulin-positive cells within ten days. Adeno-associated viruses persist episomally for months to years and have reprogrammed retinal Müller glia into rod photoreceptors and astrocytes into neurons at efficiencies near 20 percent. Lentiviruses integrate into the host genome, enabling sustained expression but raising insertional mutagenesis concerns. Non-viral alternatives include plasmids, which can carry very large cargoes with minimal immunogenicity but express transgenes only transiently, and nanoparticles, particularly lipid nanoparticles, which can be surface-engineered for tissue-specific targeting. In one striking example, neutrophil-membrane-coated mesoporous silicon nanoparticles carrying cardiac reprogramming miRNAs homed to injured heart tissue via fibrinogen-derived targeting peptides, improving cardiac function and reducing fibrosis after myocardial infarction in mice.

The therapeutic results assembled in the review span nearly every major organ system. In the brain, delivery of the transcription factors Ascl1, Lmx1a, and Nurr1 converted striatal glial cells into functional neurons that integrated into host circuits within three months. AAV-mediated suppression of PTB converted astrocytes into neurons with efficiencies rising from roughly 20 percent at three weeks to 80 percent at twelve weeks, replenishing lost dopaminergic neurons and restoring striatal dopamine in Parkinson’s disease models. In the heart, retroviral delivery of Gata4, Mef2c, and Tbx5 generated induced cardiomyocytes that developed mature sarcomeres, fired action potentials, and electrically coupled with endogenous tissue, improving cardiac function after infarction. A seven-molecule chemical cocktail administered orally and intraperitoneally achieved lower but measurable conversion with functional recovery. In the pancreas, reprogrammed beta cells secreted insulin and improved glucose tolerance in diabetic mice, while in adipose tissue, exosome-driven conversion of inflammatory macrophages and browning of white fat improved insulin sensitivity and energy expenditure in obese mice.

Despite these preclinical successes, the review is unflinching about why in vivo direct reprogramming has yet to enter a single clinical trial. Reprogramming kinetics and efficiency in living tissue lag far behind culture conditions: a small-molecule cocktail that converted 94 percent of neonatal astrocytes into neurons in vitro achieved only about 11 percent in adult mouse brain, and in vitro-optimized doses proved insufficient in vivo, sometimes requiring tripled concentrations. The complex three-dimensional tissue microenvironment, with its biophysical cues such as matrix stiffness and shear stress, and its biochemical milieu of injury-induced inflammatory signals, reactive oxygen species, and cellular senescence, can either promote or antagonize reprogramming in ways that two-dimensional cultures cannot predict. Safety concerns compound the problem: reprogramming factors such as c-Myc carry oncogenic potential, integrating vectors risk insertional mutagenesis, transient progenitor-like states may mimic cancer phenotypes, and non-target cells sharing lineage history with targets may be inadvertently converted. Partially reprogrammed cells that revert or drift into undefined states represent a further unresolved risk.

The path forward, the authors argue, will require a convergence of emerging technologies. CRISPR-based transcriptional activators offer programmable activation of lineage-specific gene networks, as shown when AAV-delivered CRISPR repression of Nrl converted rod photoreceptors into cone-like cells in a retinitis pigmentosa model. Artificial intelligence and machine learning models trained on single-cell multi-omics data can predict gene perturbation effects, identify bifurcation points where reprogramming trajectories succeed or fail, and screen for novel factor combinations. Organ-on-chip platforms that replicate tissue-specific stiffness, fluid flow, and electrical properties may finally allow preclinical optimization of dose and exposure time under realistic three-dimensional conditions. Stimuli-responsive delivery systems, including microneedle patches, hydrogels, and electromagnetic-field-activated gene switches, promise controlled and targeted release. Before first-in-human trials, the review concludes, the field must validate outcomes in large animal models, establish GMP-compliant manufacturing, define patient stratification criteria, and develop regulatory frameworks that evaluate the integrated platform of factors and delivery vehicles as a single therapeutic entity. If those challenges can be met, in vivo direct reprogramming could move regenerative medicine beyond repair toward genuine rejuvenation of organs currently considered untreatable.

Subject of Research: In vivo direct reprogramming of tissue-resident cells for regenerative therapy

Article Title: In Vivo Direct Reprogramming: Current Progress and Future Prospects from Mechanisms to Therapeutic Application

Article References: Singh, R. D., Calvoli, M., & Kim, K. K. (2026). In Vivo Direct Reprogramming: Current Progress and Future Prospects from Mechanisms to Therapeutic Application. Advanced Science, Article e77370. https://doi.org/10.1002/advs.77370

Image Credits: AI Generated

DOI: 10.1002/advs.77370

Keywords: direct reprogramming, in vivo reprogramming, regenerative medicine, transcription factors, pioneer factors, gene regulatory networks, iPSCs, cardiac regeneration, neuronal reprogramming, viral vectors, lipid nanoparticles, CRISPR

The technical foundation of the field rests on two reprogramming strategies. Integrative methods, which use retroviral or lentiviral vectors to deliver the classic quartet of factors—OCT4, SOX2, KLF4 and c-MYC—are efficient but insert themselves permanently into the genome, raising the specter of insertional mutagenesis and tumor formation. Non-integrative alternatives, including Sendai virus, episomal plasmids, synthetic mRNA and protein-based approaches, leave no genetic scars and are now favored for clinical translation, even though they tend to be less efficient and more expensive. The payoff is the ability to manufacture transgene-free, GMP-compliant cell lines suitable for personalized medicine and regenerative therapies, a goal that has come steadily closer over the review period.

What makes iPSCs so potent for drug discovery is their capacity to reproduce human disease in a dish. Patient-derived iPSCs can be coaxed into neurons, cardiomyocytes, hepatocytes or pancreatic beta cells that carry the exact genetic variants underlying a person’s illness, allowing researchers to study molecular mechanisms that animal models frequently misrepresent. This has proven particularly valuable for neurodegenerative disorders, cancer and metabolic conditions, where interspecies differences between mice and humans have derailed countless promising compounds. High-throughput screening campaigns can now run thousands of candidate drugs against disease-relevant human cells, identifying molecules that correct cellular pathologies rather than merely surviving toxicity filters designed around animal physiology.

The review highlights striking successes in drug repositioning, the rediscovery of new uses for existing medicines. Using patient-specific iPSC models, researchers identified statins as potential treatments for achondroplasia and the anticonvulsant ezogabine as a candidate therapy for amyotrophic lateral sclerosis. Screens against familial dysautonomia and ALS have similarly surfaced compounds that would have been difficult to find through conventional pipelines. These platforms are increasingly woven into clinical trials themselves, serving as pre-screening tools that predict how individual patients will respond to therapy, cutting down on trial-and-error prescribing and reducing adverse effects by matching the right drug to the right genetic background before the first dose is ever administered.

Nowhere is the approach more consequential than in Alzheimer’s disease, a condition that has devoured decades of research funding and an unbroken series of failed clinical trials. iPSC-derived neurons and glial cells now replicate the disease’s signature pathologies in vitro, including amyloid-beta accumulation and tau phosphorylation. High-content screens using these models flagged the anti-inflammatory drug cromolyn and antiparasitic avermectins as potential amyloid-lowering agents, while statins emerged as modifiers of phosphorylated tau. Tri-culture systems combining neurons, astrocytes and microglia have illuminated the glial dimension of neurodegeneration, and single-cell RNA sequencing has mapped gene expression changes across disease progression, surfacing new biomarkers. CRISPR/Cas9 editing of iPSC lines derived from patients with PSEN1 and PSEN2 mutations has allowed precise validation of therapeutic targets such as BACE1 and MAPT, and clarified unexpected links between cholesterol biosynthesis and amyloid precursor protein dysfunction.

Diabetes research has followed a parallel trajectory. Differentiation protocols guided by the transcription factors PDX1, NKX6.1 and Ngn3 now generate glucose-responsive, insulin-secreting beta-like cells from iPSCs, and recent studies have shown these cells can reverse hyperglycemia in streptozotocin-induced diabetic mice. Three-dimensional organoid systems have sharpened the maturity and glucose responsiveness of the derived cells, better approximating native pancreatic islets. Beyond replacement therapy, patient-derived iPSCs have been differentiated into endothelial cells that model diabetic vascular dysfunction, a screening platform that identified angiotensin receptor blockers as candidates for restoring endothelial function. Encapsulation of iPSC grafts in elastin-like recombinamers has shown promise for immunoprotection, addressing one of the central obstacles to durable cell replacement in type 1 diabetes.

In arthritis, the technology has produced some of the most inventive therapeutic engineering to date. CRISPR-engineered iPSC-derived macrophages have been designed to sense tumor necrosis factor-alpha and respond by autonomously secreting anti-inflammatory biologics such as soluble TNF receptor 1, quelling joint inflammation and bone erosion in murine models. iPSC-derived extracellular vesicles have reduced IL-1beta-driven chondrocyte death in osteoarthritis models while suppressing the cartilage-degrading enzymes ADAMTS5 and MMP13. Even more strikingly, vesicles from anti-inflammatory M2 macrophages can reprogram pro-inflammatory M1 macrophages within inflamed synovial tissue with efficiencies exceeding ninety percent, restoring a critical immune balance and boosting production of IL-10 and other tissue-protective cytokines. The iPSC secretome itself—exosomes, vesicles and growth factors—is emerging as a drug class in its own right, offering regenerative and immunomodulatory effects without the risks of transplanting live cells.

Cancer modeling has been equally transformed. Reprogramming tumor cells back to a pluripotent state can erase epigenetic damage, such as the hypermethylation of tumor suppressor genes in glioblastoma multiforme, creating a window into how malignancies originate. iPSC-based chimeras have modeled osteosarcoma and pancreatic ductal adenocarcinoma, while transient expression of the oncogene MLL-AF4 in iPSCs has reproduced B-acute lymphoblastic leukemia, and iPSCs carrying germline TP53 mutations have illuminated osteosarcoma formation in Li-Fraumeni syndrome. Organoids derived from iPSCs recreate the three-dimensional architecture of tumors, offering a far more faithful stand-in for the tumor microenvironment than flat cultures, and when combined with patient-derived xenografts they support biomarker discovery and pharmacological testing in vivo.

None of this has been easy. The review is candid about the field’s technical debts: genetic instability accumulates during reprogramming and long-term expansion, and even minor mutations introduced during cell generation can confer tumorigenic potential. Differentiation protocols vary between laboratories, producing inconsistent disease models and irreproducible drug response data. The costs of growth factors and specialized media under GLP and GMP standards are punishing, and residual undifferentiated cells raise unresolved immunogenicity concerns. Ethical questions compound the technical ones—donors are often inadequately informed about whether their cells may be used in human-animal chimera studies or germline-related research, incidental genetic findings are disclosed inconsistently, and regulatory oversight of cell therapies remains fragmented across jurisdictions, leaving openings for unproven treatments to reach vulnerable patients.

The authors’ prescriptions are pragmatic. They call for standardized differentiation protocols, shared reference cell lines and validation criteria to make results comparable across the field; deeper collaboration among academia, biotechnology and pharmaceutical companies and regulators to build scalable, compliant products; expanded investment in automation and machine learning to predict and control differentiation outcomes; transparent consent frameworks that spell out the full scope of possible research uses; long-term safety monitoring with centralized tracking of patients receiving iPSC-derived therapies; and sustained public engagement to build the trust on which the entire enterprise depends. If those foundations hold, the five years from 2020 to 2024 may be remembered as the period when drug development finally began to run on human biology rather than approximation—and the diseases that have resisted medicine longest may at last face models that truly resemble them.

Subject of Research: Applications of induced pluripotent stem cells in novel drug development from 2020 to 2024

Article Title: Recent progress in the use of induced pluripotent stem cells in novel drug development over the past 5 years (2020–2024)

Article References: Moseki, M. C., Makhzoum, A., Masisi, K., & Gaobotse, G. (2025). Recent progress in the use of induced pluripotent stem cells in novel drug development over the past 5 years (2020–2024). Discover Biotechnology, 2(1), Article 33. https://doi.org/10.1007/s44340-025-00042-x

Image Credits: AI Generated

DOI: 10.1007/s44340-025-00042-x

Keywords: induced pluripotent stem cells, iPSCs, drug discovery, disease modeling, Alzheimer's disease, diabetes, arthritis, cancer, personalized medicine, CRISPR, organoids, high-throughput screening