Review Papers Claim by Claim

Key visual takeaways

• Platform engineers uniaxial axon bridges up to 2000 µm with high orientation (mean 90.21°, Gaussian fit R²=0.98) and progressive GALC/MBP increases consistent with oligodendrocyte maturation and myelination (IF quantification) Jin et al. 2026: orientation & myelination data
• Cup Cocktail (cuprizone 1000 µM + TNF-α 50 ng/mL + IFN-γ 50 ng/mL) caused widespread loss of MOG and disrupted β3-tubulin networks (axon degeneration), while LPC (30 µg/mL, 24 h) produced more selective myelin loss in the axonal gap with axons largely intact — enabling separation of demyelination vs combined demyelination+axonopathy phenotypes Jin et al. 2026: demyelination regimes
• MEA shows marked functional impairment after both treatments: amplitude dropped from ≈1000 µV → ≈200 µV and velocity ≈1.2 → ≈0.4 m/s, consistent with loss of saltatory conduction after myelin loss Jin et al. 2026: MEA results

Concise critical evaluation

1. Strengths
   • Human-cell origin: hiPSC-derived hNSCs enable human-specific cell interactions and oligodendrocyte differentiation rather than relying on rodent cells (Jin et al. 2026: human cell origin).
   • Architectural realism: microwell-guided colonies + aligned piezoelectric scaffolds produce long (up to 2 mm), uniaxial, myelinated axons—important because spinal-cord lesions extend over long distances and conduction deficits amplify with axon length (Jin et al. 2026: architecture).
   • Functional validation: MEA electrophysiology links morphology to physiology and distinguishes demyelination-only (LPC) from demyelination+axonopathy (Cup Cocktail), strengthening face validity.
2. Limitations and blindspots
   • Single hiPSC line and limited donor diversity — raises questions about inter-line variability and generalizability (authors acknowledge single-line usage and n≥3 biological replicates) Jin et al. 2026: limitations.
   • Acute chemical exposures (24 h) followed by 1-week recovery capture early demyelination but not chronic lesion biology, inflammation dynamics, or remyelination kinetics; many MS processes unfold over weeks–months and involve adaptive immunity and microglia interactions (the model lacks microglia/macrophages as a separate immune compartment) Jin et al. 2026: exposure regime.
   • Reagents & concentrations: cuprizone is a systemic toxin in vivo; in vitro dosing (1000 µM) plus exogenous TNF/IFN is pragmatic but may not recapitulate microglia‑driven cytokine timing and gradients; LPC is a focal detergent—both are valid tools, but neither fully models autoimmune MS pathology (authors discuss these caveats) Jin et al. 2026: toxin model caveat.
   • Quantitative details: while marker ratios and MEA summary metrics are reported, the paper does not (within provided text) include raw spike trains, electrode-by-electrode conduction latencies, or EM ultrastructure quantification (g-ratios) to confirm ultrastructural myelin loss vs protein loss.
3. Reproducibility & methods transparency
   • Methods are well-detailed (electrospinning parameters, scaffold annealing, microwell fabrication, MES parameters, antibody lists, MEA stimulation protocol), which supports reproducibility for labs experienced in electrospinning and MEA work Jin et al. 2026: methods detail.
   • Data availability: authors state data are available from corresponding author upon reasonable request; public deposition of raw images, electrophysiology recordings, and fabrication CAD/print files would materially improve reproducibility and community adoption.

Practical recommendations to strengthen the platform

• Introduce i) microglia/macrophage co-culture (iPSC-derived microglia) or ii) conditioned microglial secretomes to model debris clearance and cytokine timing; compare Cup Cocktail with microglia-mediated cuprizone pathology (Jin et al. 2026: cuprizone + cytokine note).
• Provide ultrastructural confirmation (TEM g-ratio analysis) to demonstrate compact myelin loss vs MBP/MOG protein loss, and deposit representative raw MEA spike data (to allow conduction latency maps & propagation analyses).
• Test multiple hiPSC lines (diverse donors, sexes, genetics) to quantify inter-line variability in myelination, demyelination susceptibility, and remyelination potential.
• Longitudinal experiments: extend demyelination durations and recovery windows and assay remyelination-promoting interventions (e.g., TREM2 modulation, CXCR2 blockade, mitochondrial Ca2+ support) to demonstrate platform utility for therapeutic screening (see relevant mechanisms in literature) TREM2 agonist accelerates debris clearance (preclinical)).

Conclusion and what would change it

Conclusion: The paper presents a technically rigorous, human-cell in vitro spinal-cord–like platform that meaningfully reproduces uniaxial axonal myelination and produces two experimentally separable demyelination phenotypes with correlated electrophysiological deficits — a useful screening and mechanistic platform, though limited by single-line use, short toxin exposures, and absence of microglial immune dynamics (Jin et al. 2026: authors).

Data that would change this assessment: public release of raw MEA traces and TEM ultrastructure showing preserved axon ultrastructure after LPC but frank axonal degeneration after Cup Cocktail; demonstration across multiple hiPSC lines; and inclusion of microglia that change the cuprizone response would materially strengthen or, if absent, weaken claimed physiological relevance.