BGPT: Paper Review: Perfusable 3D models of ureteric bud and collecting duct tubules

Explore by Goal

Quick Explanation Copied

Perfusable UB→CD tubules: what’s new (and what’s still uncertain)

Key advance: a perfusable 3D ureteric bud (UB) tubule platform from hiPSC-derived UB cells that budded into ECM, supported lumen-to-lumen fusion with adjacent bioprinted UB networks, and then enabled UB-to-collecting-duct (CD) differentiation under luminal flow—a combination the authors argue is missing in prior UB/CD-on-chip models.
Skeptical takeaway: marker-level evidence suggests flow-associated maturation, but several functional endpoints (e.g., transepithelial transport/transport kinetics) are not demonstrated in the provided text, and gene↔protein discrepancies (notably AQP2) remain unresolved.

Want deeper critique on model validity, controls, and what would falsify the “flow drives CD maturation” claim? Use the agent buttons below.

Long Explanation

Paper Review

Perfusable 3D models of ureteric bud and collecting duct tubules

Preprint DOI: 10.1101/2025.06.19.659798
Paper date (from provided metadata): June 24, 2025.

Visual model map (what the system is)

Input: hiPSC differentiation into UB organoids, dissociation into UB cells.

Core engineering: perfusable cylindrical channel embedded in fragmented basement membrane matrix + collagen I (BsM(frag.)–Col I), tuned for monolayer formation, budding, and embedded bioprinting.

Network creation: embedded bioprinting of UB network filaments adjacent to the central tubule, aiming for lumen-to-lumen fusion.

Phenotype conversion: perfused differentiation from UB tubules into CD tubules, comparing low vs high flow and static controls; markers (RET, WNT9B, AQP2, CALB1, etc.) are used to assess UB vs CD states.

Fluidic cue: low vs high perfusion (reported shear stress)

Shear stresses are stated in the paper as 0.006 dyn/cm² (LF; 5 µL/min) and 0.08 dyn/cm² (HF; 66 µL/min), calculated from an initial ~1 mm tubule diameter.

Important caveat: experienced shear can be ~order-of-magnitude higher

The authors estimate maximum fluidic shear stresses of ~0.031±0.015 dyn/cm² (LF) and ~0.34±0.14 dyn/cm² (HF), deviating up to an order of magnitude from values computed using the initial 1 mm diameter.

Morphogenesis readout: budding length scale

The maximum bud length is reported to range ~200–600 µm, with some variation in budding number/density across batches but not in maximum length.

Claim-by-claim evaluation (support, uncertainty, falsifiability)

1) ECM enables UB monolayer formation + epithelial lumen budding

Supported by: the paper reports that a hybrid BsM–Col I matrix yields nearly 100% confluency after 2 days in 2D adhesion, and that only BsM–Col I supports both monolayer formation and budding with lumens expressing SOX9/CDH1/KRT8 over 21 days in a 3D budding assay.
Uncertainties / missing: the provided text does not show whether luminal formation is functionally polarized (e.g., transporter polarity) at this ECM optimization stage—marker/cell-morphology readouts may not guarantee epithelial barrier function in 3D budding assays.

2) Perfusable 3D UB tubules maintain monolayer integrity under perfusion

Supported by: the paper states that 3D UB tubules can be perfused at 5–66 µL/min for >1 week without monolayer disruption and that cross-sections show a confluent monolayer connected to buds via central lumen.
Important caveat: the authors explicitly note that tubule geometry changes with reservoir pressure and that contraction during monolayer formation introduces variability in diameter and thus shear stress experienced.

3) UB bioprinted networks fuse luminally to the central perfusable tubule

Supported by: the paper reports GFP/mCherry lineage-tagged printed UPC features forming tubules that fuse and generate contiguous lumens, and then shows lumen-to-lumen fusion between bioprinted UPC-derived features and the central UB channel after long co-culture/differentiation.
Uncertainty: the provided text excerpt does not include quantitative metrics of fusion efficiency or continuity (e.g., fraction of constructs with fully connected lumens) beyond qualitative imaging.

4) Flow promotes UB→CD differentiation/maturation (marker-level evidence)

Supported by:

The paper reports CD marker upregulation (e.g., WNT9B) and UB marker downregulation (e.g., RET/WNT11/ETV5/GFRA1) when comparing differentiated CD-like states under flow vs UB-like controls across multiple conditions.
The paper states that high-flow conditions reduce RET/WNT11/ETV5 expression while increasing AQP2 expression at the gene level compared with static CD controls, with UB markers largely unaffected aside from UPK2/AQP2 changes.
Protein imaging reports qualitatively less membranelocalized RET in CD vs UB, particularly under flow, and increased CALB1 protein in flow-exposed cells.

Key counterpoint (important): AQP2 gene and protein do not align: the paper says AQP2 upregulation under flow is not observed at the protein level and polarization/trafficking trends are unclear.
What would falsify the “flow drives CD maturation” interpretation: flow conditions could be changing shear-stress-mediated epithelial mechanics without truly maturing specific functional CD transport phenotypes. A falsification would require demonstrating that CD-like marker shifts disappear when shear is controlled or neutralized independently from differentiation media composition—control layers for this are not fully shown in the excerpt.

Extracted numeric parameters (from provided full text excerpt)

Stage	Parameter	Value(s)	Unit	Context / comparison
Perfusion	Low flow rate	5	µL/min	Defines LF condition
Perfusion	Shear stress (applied calc.)	0.006	dyn/cm²	Calculated from initial ~1 mm diameter
Perfusion	High flow rate	66	µL/min	Defines HF condition
Perfusion	Shear stress (applied calc.)	0.08	dyn/cm²	Calculated from initial ~1 mm diameter
Perfusion (re-estimated)	Max shear (LF)	0.031	dyn/cm²	Average max after contraction (±0.015)
Perfusion (re-estimated)	Max shear (HF)	0.34	dyn/cm²	Average max after contraction (±0.14)
Morphogenesis	Max bud length	200–600	µm	Maximum bud protrusion range reported
Microfabrication	Channel diameter target	1.0	mm	Operationally maintained ~1 mm by reservoir height strategy
Matrix printing	BsM fragment area range	~100–10,000	µm²	Cross-sectional areas reported for fragmented particles

Values are transcribed from the provided paper text excerpt.

What’s strong vs what remains unproven

Strengths (mechanistic alignment + engineering rigor)

Integrated bidirectional engineering (ECM rheology/fragmentation + perfusable geometry + bioprinting adjacency + luminal fusion) rather than a single-variable experiment.
Shear-stress caveat acknowledged (experienced shear may differ from applied estimates due to contraction). This is a key skeptical feature because it prevents overclaiming that “shear = exactly X.”

Limitations / blind spots (as suggested by the provided excerpt)

Marker-level differentiation without full function: the excerpt mentions barrier function evidence using FITC-inulin retention, but does not show core CD functional readouts such as Na+/water transport rates or vasopressin-responsive functional assays in the provided text.
Gene–protein discrepancies weaken “maturation” claims if not resolved: AQP2 mRNA vs protein do not show clear alignment under flow.
Attribution of flow effects is mechanistically unresolved: the paper states RET flow effects remain unknown and calls for deeper pathway/subpopulation analysis (e.g., single-cell RNA-seq or proteomics).
Reproducibility signals not fully visible in the excerpt: the paper mentions N=3 independent experiments and 33/36 tubules reaching endpoint in perfused culture, but the provided text does not list detailed sharing of chips CAD files or deposition of raw image data in public repositories (it says “upon request” for some items).

Known vs inferred vs uncertain (skeptical epistemic ledger)

Known (directly stated in the paper text excerpt):

Flow regimes are defined (LF 5 µL/min; HF 66 µL/min) and shear stresses are stated, with contraction-driven re-estimation.
UB tubules show lumen connectivity to buds and are described as having barrier function by FITC-inulin retention relative to non-epithelialized controls.

Inferred (interpretation):

The authors infer flow may promote CD maturation because flow-associated trends include RET ↓ and CD marker shifts.

Uncertain / not established in the excerpt:

Mechanism of flow effects on RET (the paper states it remains unknown).
Functional CD transport capacity (beyond gene/protein markers) under LF/HF.

Author reviews (click to open BGPT page)

Feedback:

Updated: April 25, 2026