Notes on Direct Conversion of FBs into iENPs (hESC-ENP-TF–enriched) and Disease Modeling

Overview

• Directly convert human fibroblasts (FBs) into expandable neural progenitor-like cells (iENPs) by overexpressing transcription factors (TFs) enriched in human ESC-derived ENPs (hESC-ENPs).

• Two TF combinations were identified that efficiently generate iENPs from FBs: a 6-TF set (iENP-6F) and a 7-TF set (iENP-7F).

• iENPs resemble hESC-ENPs in morphology, gene expression, proliferation patterns, and differentiation potential to CNS and PNS lineages.

• iENPs can differentiate into neurons (multiple subtypes), astrocytes, and oligodendrocytes in vitro; they also survive and differentiate after in vivo transplantation in rat brains.

• Disease modeling: iENPs can be derived from FBs of patients with Alzheimer’s disease (AD) and Huntington’s disease (HD); disease-relevant phenotypes (Aβ, pTAU, DNA damage) recapitulate key features in their neuronal derivatives and respond to disease-relevant treatments.

• The approach avoids iPSC reprogramming and associated tumorigenicity risks from pluripotent states, offering a scalable platform for disease modeling and potential autologous therapies.

Background and rationale

• Neurodegenerative diseases such as Huntington’s and Alzheimer’s are not curable with current therapies; patient-specific disease models are needed for pathogenesis studies and drug screening.

• iPSCs can model disease but carry tumorigenic risks due to pluripotent states and unpredictable differentiation.

• Induced neurons (iNs) directly converted from FBs can yield specific neuronal subtypes quickly but often lack expandable progenitor populations with broad differentiation potential.

• hESC-ENPs display broad differentiation capacity to CNS and PNS lineages, and identifying neural TFs enriched in hESC-ENPs offers a path to convert FBs into ENP-like cells that are expandable.

Key concepts and definitions

• ENP: Embryonic neural progenitor population capable of giving rise to diverse neural lineages.

• iNP / iENP: Induced neural progenitor from somatic cells; iENP here denotes induced ENP-like cells expandable in culture.

• hESC-ENP: ENPs derived from human embryonic stem cells.

• TF (transcription factor): Protein that binds DNA and regulates gene expression; selecting TFs enriched in hESC-ENPs aims to bias FBs toward ENP identity.

• Reporter systems: PAX6:EGFP and SOX1:EGFP used to monitor neural fate progression and ENP induction efficiency.

• iENP-25F, iENP-15F, iENP-13F, iENP-6F, iENP-7F: iENP populations generated with 25, 15, 13, 6, and 7 TF combinations, respectively.

• AT8: Antibody recognizing phosphorylated tau (pTAU), used to assess AD-related tau pathology.

• Ab40/Ab42: Secreted amyloid-β peptides; Ab42/Ab40 ratio is a disease-relevant readout in AD models.

• gH2AX: A marker of DNA double-strand breaks/DNA damage.

• CGS21680: A selective A2A adenosine receptor (A2AR) agonist.

• SB415286: A GSK-3β inhibitor used to modulate tau phosphorylation.

• 1-Aza: 5-aza-2′-deoxycytidine (DNA methylation inhibitor) used to probe epigenetic effects.

TF discovery strategy and experimental workflow

• Compare global gene expression between FBs and hESC-ENPs to identify hESC-ENP-enriched neural TFs (25 TFs identified; NR2F2 included due to known neural differentiation role).

• Two neural reporters (PAX6:EGFP and SOX1:EGFP) were validated in hESC-ENPs to monitor neural fate progression.

• FBs were transduced with lentiviruses encoding the 25 TFs plus a neural reporter. By ~6 days post-infection, PAX6:EGFP+ or SOX1:EGFP+ cells emerged with rounded morphology, unlike UbC:EGFP controls.

• After FACS purification of GFP+ cells, they formed neural sphere-like structures about 2 days post-replating.

• Characterization of iENP-25F showed expression of common ENP markers such as NESTIN, OTX2, and ZO1 and neural genes (assessed by ICC and RT-PCR).

• A two-step TF reduction identified TFs essential for ENP induction by removing one TF at a time and assessing the effect on GFP+ cell generation (via flow cytometry).

• TF reductions yielded two potent combinations: 15 TFs and 13 TFs with specific removals that significantly decreased PAX6:EGFP+ or SOX1:EGFP+ cells, respectively.

• The selected 15TF and 13TF sets were placed under a doxycycline-inducible system; after purification of GFP+ cells and induction, iENP-15F and iENP-13F could form neural spheres and express ENP markers, with endogenous ENP gene activation after DOX withdrawal.

• Global gene expression (microarray) showed iENP-15F/ iENP-13F were more similar to hESC-ENPs than to FBs.

• Both iENP-15F and iENP-13F differentiated to TUJ1+ neurons, GFAP+ astrocytes, and GALC+ oligodendrocytes in vitro, indicating multipotency.

• A refined second TF screen (6TF and 7TF) further pared down to minimal essential sets that still produce iENPs: iENP-6F and iENP-7F.

• After selecting 6TFs and 7TFs, the fractions of PAX6:EGFP+ and SOX1:EGFP+ cells were: 10.54 ext{%} \, ext{±} \, 0.47 ext{%} ext{ for PAX6:EGFP+; } 11.22 ext{%} \, ext{±} \, 0.44 ext{%} ext{ for SOX1:EGFP+} following the final TF sets and purification.

• iENP-6F and iENP-7F also formed neural spheres after purification, with exogenous transgene expression silenced upon DOX withdrawal and endogenous ENP gene expression activated, demonstrating reprogramming stability.

Minimal TF sets and validation

• 6TF combination (iENP-6F): removal of any single TF significantly reduced PAX6:EGFP+ cell generation; 10.54% ± 0.47% of PAX6:EGFP+ cells after purification.

• 7TF combination (iENP-7F): removal of any single TF significantly reduced SOX1:EGFP+ cell generation; 11.22% ± 0.44% of SOX1:EGFP+ cells after purification.

• After infection with the 6TF or 7TF sets under doxycycline control and subsequent purification, iENP-6F and iENP-7F formed neural spheres and expressed ENP markers and genes.

• Integration of exogenous transgenes confirmed by PCR; endogenous ENP gene activation confirmed after doxycycline withdrawal (RT-PCR).

• Global gene expression clustering showed iENP-6F and iENP-7F profiles closer to hESC-ENPs than FBs.

• iENP-6F and iENP-7F were capable of prolonged expansion (>20 passages) with normal karyotype and maintained NP characteristics; they could be cryopreserved and recovered.

Molecular and cellular characterization of iENPs

• Morphology: neural sphere-like structures after purification; NP-like colonies.

• Marker expression (ICC/RT-PCR): ENP markers such as NESTIN, OTX2, ZO1, plus neural genes; endogenous ENP genes activated after DOX withdrawal.

• Genomic status: exogenous TFs integrated; exogenous expression silenced after withdrawal; endogenous TFs expressed.

• Global expression: iENP-6F and iENP-7F resemble hESC-ENPs more than FBs; iENP-6F and iENP-7F express ENP genes and NP markers (see Figures in the study).

• Proliferation and survival:

  • iENP-6F proliferates similar to hESC-ENPs; iENP-7F proliferates slower.

  • BrdU+ higher in iENP-6F; TUNEL+ lower in iENP-6F vs iENP-7F, indicating less apoptosis in 6F.

• Long-term expansion and stability: iENP-6F and iENP-7F can be passaged >20 times with a stable karyotype; scalable population for downstream applications.

In vitro multipotency and neuronal subtypes

• Differentiation into three major neural lineages:

  • Astrocytes: GFAP+。

  • Oligodendrocytes: GALC+.

  • Neurons: TUJ1+ with mature neuronal markers MAP2, NEUN, TUJ1 co-expression with synaptic marker SYP.

• Relative neuronal differentiation propensity:

  • iENP-6F: neuronal differentiation robust and comparable to hESC-ENPs; higher neuronal yield than iENP-15F.

  • iENP-15F: reduced neuronal generation relative to iENP-6F; oligodendrocyte generation also lower than hESC-ENPs.

  • iENP-7F and iENP-13F: neuronal differentiation comparable to or slightly lower than hESC-ENPs.

• Neuronal subtype diversification (iENP-6F and iENP-7F): capable of generating

  • GABAergic neurons (GABA+).

  • Cortical neurons (TBR1+).

  • Dopaminergic neurons (TH+).

  • Motor neurons (HB9+/ISL1+).

  • Peripheral neurons (BRN3A+, PRPH+, NAV1.7+).

• Subtype-specific differentiation cues:

  • Cortical differentiation conditions enhance TBR1+ TUJ1+ neurons.

  • Dopaminergic differentiation conditions enhance TH+ TUJ1+ neurons.

  • Peripheral neuron differentiation conditions enhance PRPH+ or NAV1.7+ populations.

• Electrophysiological maturity (iENP-6F and iENP-7F derived neurons):

  • Resting membrane potentials: $V<em>{ ext{rest}} = -35.25 ext{ mV} \pm 0.64 ext{ mV}$ for iENP-6F-derived neurons; $V</em>{ ext{rest}} = -64.3 ext{ mV} \pm 17.96 ext{ mV}$ for iENP-7F-derived neurons.

  • Action potentials elicited by depolarizing current steps in current-clamp mode.

  • Spontaneous action potentials observed in some iENP-derived neurons.

  • Inward Na+ currents blocked by tetrodotoxin (TTX), confirming voltage-gated Na+ channel activity.

In vivo differentiation and integration

• Transplantation into rat brains: undifferentiated iENP-6F and iENP-7F transplanted into the corpus callosum; analyzed at 12 weeks post-transplantation.

• Tumor safety: no tumor formation; negative RT-PCR/IHC for tumor markers; confirmed by H&E staining.

• Migration and differentiation in vivo:

  • Some transplanted cells migrated toward ventricular zones (neurogenic niche) and expressed GFAP (radial glia progenitor marker).

  • Differentiation into GFAP+ astrocytes, NG2+ oligodendrocytes, and TUJ1+/MAP2+ neurons observed in vivo.

• Conclusion: iENPs can survive, migrate, and differentiate into major neural lineages in adult brain, consistent with hESC-ENP behavior.

The two iENP populations differ in developmental propensity

• Global gene expression similarity:

  • iENP-6F and iENP-7F have similar global profiles, but ~170 genes differ significantly (R2-fold change cutoff used).

• Pathway and functional differences (via IPA and GO analysis):

  • iENP-7F shows relatively lower expression of cell-cycle/mitosis genes; cell-death pathways more active in iENP-7F.

• Growth and survival:

  • iENP-6F proliferates more rapidly, akin to hESC-ENPs; iENP-7F proliferates more slowly.

  • BrdU incorporation higher in iENP-6F; TUNEL+ (apoptosis) higher in iENP-7F.

• Regional identity and differentiation propensity:

  • iENP-6F/iENP-Ns enriched for forebrain, midbrain, and spinal cord markers.

  • iENP-7F/iENP-Ns enriched for hindbrain and peripheral nervous system (PNS) markers.

  • In undifferentiated iENPs and iENP-derived neurons, iENP-6F favored rostral identity; iENP-7F favored caudal identity.

• A proposed model (Figure 6G): differential neural identity outcomes can be steered by using different hESC-ENP TF panels and neural reporters to define iENP subpopulations with distinct differentiation propensities.

Disease modeling with AD and HD iENPs

• Patient-derived FBs used to generate AD- and HD-iENPs using the 6TF (or 7TF) combos; iENPs could be induced from AD and HD FBs with the same reporters and TFs as WT controls.

• Alzheimer’s disease (AD) features recapitulated in iENP-derived neurons:

  • Conditioned media Ab40 and Ab42 levels were elevated in neurons differentiated from PSEN1-mutant AD iENPs (AD2 and AD3) relative to control neurons; Ab42/Ab40 ratio increased in AD2-derived neurons; AD3 showed variable Ab42/Ab40 changes.

  • Phospho-tau (pTAU) detected in AD-iENP-derived neurons (via AT8) in processes and in cell bodies; pTAU aggregates observed.

  • Pharmacological reduction of pTAU with GSK3β inhibitors SB415286 and other agents reduced AT8 signal in AD-iENP neurons, indicating a reversible tau pathology component in this model.

• Huntington’s disease (HD) features recapitulated in iENP-derived neurons:

  • HD iENPs and their neuronal derivatives exhibit increased DNA damage markers (gH2AX) relative to control iENPs.

  • Activation of A2A adenosine receptor (A2AR) with CGS21680 reduced gH2AX (DNA damage) in HD-iENP-derived neurons and HD iENP populations, suggesting potential therapeutic modulation of DNA damage in HD models.

• Overall significance: diseased iENPs recapitulate core pathological features of AD and HD, supporting their use for disease mechanism studies and drug screening in a human cellular context.

Ethical, practical, and translational implications

• A direct conversion approach reduces tumorigenic risk associated with iPSC reprogramming, as iENPs are generated without passing through a pluripotent state.

• The ability to derive iENPs from patient FBs offers a path to autologous disease models and personalized drug discovery.

• In vivo transplantation in rats demonstrates potential for regenerative strategies, but long-term tumorigenicity, integration, and functional consequences need thorough evaluation.

• The reliance on TF combinations to define iENP identity implies that selecting TF panels and neural reporters can tune the fate and regional identity of iENPs, enabling disease-relevant regional modeling.

• Limitations to consider: long-term stability, safety, scalability, and reproducibility across donors and disease states; potential epigenetic memory from somatic origin; and translational gaps between rodent models and human clinical contexts.

Experimental methods (highlights)

• Human materials: FBs from healthy donors and patients; appropriate consent and IRB approvals.

• TF constructs: 25 neural TFs cloned into FUW or FUW-tetO vectors; reporter constructs PAX6:EGFP and SOX1:EGFP.

• iENP generation workflow:

  • Infect FBs with lentiviruses carrying candidate TFs + neural reporters.

  • At ~6 days post-infection, identify GFP+ cells (PAX6:EGFP+, SOX1:EGFP+).

  • Purify GFP+ cells by FACS; plate on Matrigel in iENP media; neural spheres appear within ~2 days.

  • Characterize iENPs by ICC and RT-PCR for ENP markers; confirm exogenous transgene integration by genomic PCR; withdraw doxycycline to assess endogenous ENP gene activation.

• TF reduction strategy:

  • Remove one TF at a time from 25TF pool; measure impact on GFP+ cell generation via flow cytometry to identify essential TFs.

  • Repeat with 15TF and 13TF sets to identify potent reductions; finalize with 6TF and 7TF sets.

• Microarray and gene expression analyses: compare FBs, hESC-ENPs, and iENP populations (GEO accession GSE81554; GSE27280; E-MEXP-2668).

• In vitro differentiation assays: differentiation into GFAP+ astrocytes, GALC+ oligodendrocytes, TUJ1+ neurons; neuronal subtype markers (GABA, TBR1, TH, HB9/ISL1, BRN3A, PRPH, NAV1.7).

• Electrophysiology: whole-cell patch-clamp on iENP-derived neurons after maturation (2 weeks in maturation medium).

  • Resting potentials and action potentials measured; Na+ currents blocked by TTX.

• In vivo transplantation: inject undifferentiated iENP-6F or iENP-7F into rat corpus callosum; analyze after 12 weeks; assess tumorigenicity; evaluate differentiation into GFAP+ astrocytes, NG2+ oligodendrocytes, TUJ1+/MAP2+ neurons.

• Disease assays:

  • AD: assess Ab40/Ab42 secretion; Ab42/Ab40 ratio; pTAU by AT8 staining; GSK3β inhibitor effects.

  • HD: assess gH2AX DNA damage and response to CGS21680; compare HD-iENPs to control iENPs.

Key quantitative findings (selected figures and data)

• iENP induction efficiency (PAX6:EGFP+): 5.31 ext{%} \pm 0.38 ext{%}; SOX1:EGFP+ 6.31 ext{%} \pm 0.45 ext{%} after initial 25TF infection.

• Post-purification, iENP-25F spheres formed by day 2 post re-plating; neural markers expressed by ICC/RT-PCR.

• After TF reduction to 15TF and 13TF, significant reductions in GFP+ cell production were observed for specific TF removals (statistical significance denoted in figures; p < 0.05).

• Final 6TF (iENP-6F) yield: 10.54 ext{%} \pm 0.47\% of PAX6:EGFP+ cells after purification; 7TF (iENP-7F) yield: $11.22\% \pm 0.44\%$ of SOX1:EGFP+ cells after purification.

• iENP-6F and iENP-7F could be passaged >20 times with normal karyotype; endogenous ENP gene activation observed after DOX withdrawal.

• Functional electrophysiology:

  • iENP-6F neurons: $V<em>{rest} = -35.25\pm0.64\ \text{mV}$; iENP-7F neurons: $V</em>{rest} = -64.3\pm17.96\ \text{mV}$.

  • Action potentials elicited by depolarizing current steps; spontaneous APs observed; Na+ currents blocked by TTX.

• In vivo results at 12 weeks:

  • No tumor formation observed (RT-PCR/IHC for tumor markers negative).

  • GFAP+ astrocytes, NG2+ oligodendrocytes, and TUJ1+/MAP2+ neurons detected from transplanted iENPs in rat brains.

• Disease-relevant molecular phenotypes (AD iENPs):

  • AD2/AD3 iENP-derived neurons show elevated Ab40 and Ab42 in conditioned medium; Ab42/Ab40 ratio increased in AD2 lineage.

  • pTAU detected in AD-iENP-derived neurons; pTAU aggregates observed in cell bodies.

  • GSK3β inhibitors reduced pTAU levels.

• Disease-relevant molecular phenotypes (HD iENPs):

  • HD-iENPs and their neuronal derivatives show higher gH2AX than controls; CGS21680 (A2AR agonist) reduces gH2AX in HD-iENPs and HD-derived neurons.

Implications and take-home messages

• iENP generation from FBs using hESC-ENP TFs yields expandable, multipotent progenitors capable of CNS and PNS differentiation—closer to embryonic NPCs than adult brain NPCs.

• The TF combination and neural reporter used during selection can dictate iENP properties, including proliferation, regional identity, and neuronal versus glial propensity.

• The approach provides a relatively rapid and scalable platform to model neurodegenerative diseases with patient-specific cells and to screen therapeutic compounds in a human cellular context.

• The disease models recapitulate hallmark AD and HD features and respond to targeted agents, supporting their utility for mechanism studies and drug discovery.

• Cautions: long-term safety, tumorigenicity after extended periods, and translational relevance to humans require further investigation; ethical considerations for patient-derived cells and in vivo experiments remain essential.

Data access and supplementary materials

• Microarray and expression data are available at:

  • GEO: GSE81554 (NP2, FB1, hESC-ENP, and iENPs)

  • GEO: GSE27280 (FB2 and FB3)

  • ArrayExpress: E-MEXP-2668 (NP1)

• Supplemental materials, including detailed experimental procedures and additional figures, are available with the article online.

Methods in brief (for reference)

• Constructs: 25 neural TFs cloned into FUW or FUW-tetO; reporter plasmids for PAX6:EGFP and SOX1:EGFP.

• iENP induction: FB infection with TFs + reporters; GFP+ cells purified by FACS; cells form neural spheres; DOX withdrawal activates endogenous ENP genes.

• Differentiation assays: standard neural differentiation media; cortical, dopaminergic, and PNS differentiation protocols to assess lineage potential.

• Electrophysiology: patch-clamp recordings on iENP-derived neurons after maturation.

• In vivo: rat models with intracranial transplantation of iENPs; 12-week analysis for integration and differentiation.

• Disease experiments: AD iENPs and HD iENPs generated from patient FBs; Ab and pTAU assays; DNA damage assays; pharmacological interventions with SB415286, 1-Aza, CGS21680.

Glossary of abbreviations

• FBs: Fibroblasts

• iENP: induced embryonic neural progenitor

• NP: neural progenitor

• ENP: embryonic neural progenitor

• hESC: human embryonic stem cells

• iPSC: induced pluripotent stem cell

• IIS: immunocytochemistry

• IHC: immunohistochemistry

• MA: microarray

• IPA: Ingenuity Pathway Analysis

• gH2AX: phosphorylated histone H2AX

• AT8: antibody for pTAU

• Ab42/Ab40: amyloid-β peptide species

• DH: Huntington’s disease; AD: Alzheimer’s disease

• A2AR: adenosine A2A receptor

• CGS21680: A2AR agonist

• SB415286: GSK-3β inhibitor