Feeder-Free Directed Differentiation of Human iPSCs to Definitive Endoderm via Dual GSK3/Nodal Signaling in mTeSR-Adapted Cultures

Three independent iPSC lines (or three biological replicates of one line at different passages, p25-p45) are differentiated in parallel, each in technical triplicate wells of a 12-well plate. Cells are seeded at 2.0-2.5 x 10^5 cells/cm^2 on day -1 to reach ~80-90% confluence at induction. Fixed time points: day 1 (primitive streak, brachyury/T), day 2, day 3, and day 4 (DE endpoint). A pluripotent (day 0) reference and an Activin-omitted negative arm are run alongside. CHIR concentration is the key titrated variable (1, 3, 5 µM) in optimization runs; 3 µM is used for production.

BSL-1; human iPSCs handled as potentially infectious primary-derived material under BSL-2 practices in many institutions. Wear nitrile gloves, lab coat, and eye protection in a Class II biosafety cabinet. CHIR99021 and DMSO are handled as irritants. PFA is a fixative and suspected carcinogen — use in a fume hood, collect in halogenated/aldehyde waste. Aspirated medium and dissociated cells are decontaminated with 10% bleach (>=20 min) before disposal. Sharps to puncture-proof containers.

Positive control: a previously validated iPSC line known to yield >90% SOX17+/FOXA2+ DE. Negative/differentiation control: Activin A omitted (cells remain OCT4+, do not gain SOX17). Vehicle control for CHIR: equal-volume DMSO (0.03%) confirms the GSK3 effect is compound-specific, not solvent. Isotype-matched IgG controls (goat IgG, rabbit IgG) set flow gates. Day 0 undifferentiated sample anchors the OCT4-high / SOX17-negative baseline. Secondary-only IF wells exclude nonspecific binding.

Day 1 shows a transient brachyury/T peak (40-70% positive). By day 4, >=85% of cells co-express SOX17 and FOXA2 by flow cytometry, with residual OCT4+ <5%. Viability >85% by LIVE/DEAD. IF shows nuclear SOX17/FOXA2 in confluent sheets with characteristic cobblestone DE morphology. Failed runs cluster below 60% SOX17+ and retain OCT4 islands.

• Human iPSC line (e.g., WTC-11 or a karyotype-normal in-house line), passage 25-45
• mTeSR Plus or Essential 8 maintenance medium
• Vitronectin (VTN-N, 0.5 µg/cm^2) or hESC-qualified Matrigel (1:100 in DMEM/F12)
• RPMI 1640 with GlutaMAX basal medium
• B-27 Supplement minus insulin (50x)
• Recombinant human Activin A (100 ng/mL working)
• CHIR99021 GSK3 inhibitor (3 µM working, 10 mM DMSO stock)
• Y-27632 ROCK inhibitor (10 µM, for seeding only)
• Accutase cell detachment solution
• DPBS without Ca2+/Mg2+
• Antibodies: anti-SOX17 (goat), anti-FOXA2 (rabbit), anti-OCT4 (mouse), anti-Brachyury/T (goat); AF488/AF594/AF647 secondaries
• LIVE/DEAD Fixable stain; 4% paraformaldehyde; 0.1% Triton X-100; 1% BSA block

To convert undifferentiated human iPSCs into definitive endoderm (DE) using a chemically defined, serum-free protocol driven by Activin A and a transient Wnt/GSK3 pulse, achieving a reproducible SOX17+/FOXA2+ population exceeding 85% by day 4 with minimal residual OCT4+ pluripotent cells. The DE produced serves as the validated starting material for gut-tube, hepatic, and pancreatic organoid lineages.

1. Day -1: Dissociate iPSC colonies with Accutase (5 min, 37 C), quench, count, and seed 2.0-2.5 x 10^5 cells/cm^2 onto vitronectin-coated 12-well plates in mTeSR Plus + 10 µM Y-27632 (1 mL/well).
2. Day 0: Confirm 80-90% confluence; aspirate and wash once with DPBS.
3. Day 1: Add DE induction medium A: RPMI 1640 + 100 ng/mL Activin A + 3 µM CHIR99021 (no B-27). 1.5 mL/well, 24 h.
4. Day 2: Replace with medium B: RPMI 1640 + 0.2x B-27 minus insulin + 100 ng/mL Activin A (omit CHIR). 1.5 mL/well.
5. Day 3: Replace with medium C: RPMI 1640 + 0.5x B-27 minus insulin + 100 ng/mL Activin A. 1.5 mL/well.
6. Day 4: Replace with medium C (1x B-27 optional); harvest at the start of day 4 for endpoint analysis.
7. For flow cytometry: dissociate with Accutase (8-10 min), fix in 4% PFA (15 min, RT), permeabilize 0.1% Triton, block 1% BSA, stain SOX17/FOXA2/OCT4 (1 h, RT), acquire >=10,000 events.
8. For immunofluorescence: fix wells in 4% PFA, stain in situ, image on confocal at 20x.

Independent variables: CHIR99021 concentration (1/3/5 µM), Activin A presence, differentiation day (0-4). Dependent variables: % SOX17+/FOXA2+ cells, % residual OCT4+ cells, brachyury/T peak on day 1, viability. Controlled variables: seeding density (2.0-2.5 x 10^5 cells/cm^2), passage range, medium volume (1.5 mL/well), incubation 37 C/5% CO2, lot of Activin A and B-27, plate coating matrix.

A 24-hour pulse of high Activin A (100 ng/mL) combined with CHIR99021 (3 µM) on day 1, followed by Activin A alone in progressively richer basal medium (days 2-4), will posteriorize the epiblast through a transient brachyury+ primitive-streak intermediate and commit cells to anterior definitive endoderm. We expect SOX17/FOXA2 co-expression to rise above 85% while OCT4 falls below 5%, confirming exit from pluripotency.

Flow cytometry .fcs files are gated in FlowJo v10: FSC/SSC singlets -> live (LIVE/DEAD-negative) -> SOX17+/FOXA2+ quadrant set on isotype controls. Percent-positive and median fluorescence intensity (MFI) are exported per replicate. IF images are quantified in ImageJ/Fiji: nuclei segmented (DAPI), and SOX17+/FOXA2+ counted via thresholded nuclear overlap, expressed as fraction of total DAPI+ nuclei. Data normalized to day-0 baseline.

1. Low SOX17 yield: cells were over- or under-confluent at induction — re-seed to hit 80-90% on day 0; verify Activin A lot activity. 2. Massive cell death day 1-2: omit Y-27632 too early or CHIR too high — drop CHIR to 1.5 µM and ensure ROCKi only at seeding. 3. Persistent OCT4 islands: incomplete streak induction — extend the CHIR pulse to 36 h or raise Activin to 100 ng/mL fresh. 4. Detachment of the sheet during media changes: add medium gently down the well wall and pre-warm media. 5. Variable run-to-run results: standardize passage number, aliquot Activin A, and avoid freeze-thaw cycles of growth factors.

Percent-positive values across CHIR concentrations are compared by one-way ANOVA with Tukey HSD post hoc (alpha = 0.05) using GraphPad Prism. n = 3 independent differentiations (biological replicates), each the mean of 3 technical wells. A power analysis (effect size d = 1.5, power 0.8, alpha 0.05) indicates n = 3 detects a 15-percentage-point difference in SOX17+ fraction. Replicate variability reported as mean +/- SD; CV across runs should be <15%.