Concerns regarding "TIPE2 specifies the functional polarization of myeloid-derived suppressor cells during tumorigenesis" (Yan et al., J. Exp. Med., 2020; DOI: 10.1084/jem.20182005)

Academic fraud report · Geng Detector

Summary

A text-based academic integrity review was conducted on the JEM 2020 paper by Yan et al. on TIPE2 and MDSC polarization. The overall verdict is highly suspicious, though confidence is limited by the absence of pixel-level image analysis. The most serious concern is a methodological impossibility: the authors reportedly used 250 mM H2O2 to stimulate bone marrow-derived MDSCs in vitro, a concentration roughly 500–25,000 times higher than standard oxidative-stress protocols (10–500 µM), which would be immediately cytotoxic and incompatible with downstream Western blotting. Additional concerns include an implausibly high B16F10 cell-seeding density of 1 × 10^7 cells/ml for overnight tumor supernatant preparation, where routine cultures use 0.1–2 × 10^6 cells/ml, and a statistical inconsistency between figure legends (unpaired Student's t test for Figures 1A, 1C, 6C, 6D, which depict tumor growth kinetics over time) and the methods section (two-way ANOVA for repeated tumor growth measurements). Figures 4B/C, 5A, and 6A/J, containing multiple Western blots of TIPE2, C/EBPβ isoforms, p-AKT, and p-STAT3, are flagged as high-risk for image reuse or splicing and require original-file verification. No findings are asserted as confirmed fraud.

Verdict

Highly suspicious (🟠). Based solely on textual analysis, the paper contains at least one near-fatal methodological error (250 mM H₂O₂ treatment of cultured MDSCs) and several secondary concerns. A definitive determination requires the original Western blot images, raw cell-viability records, and an institutional investigation.

Key findings

Implausible H₂O₂ concentration (Fig 6A / Methods): MDSCs were reportedly stimulated with 250 mM hydrogen peroxide. Standard in vitro oxidative-stress protocols use 10–500 µM. At 250 mM, immediate lipid peroxidation and cell death would preclude the 24-h survival and protein harvest required for Western blotting.
Implausible cell-seeding density (Methods – TES preparation): B16F10 cells were reportedly cultured overnight at 1 × 10⁷ cells/ml for tumor-derived supernatant. Routine mammalian cell culture uses 0.1–2 × 10⁶ cells/ml; the reported density would cause contact inhibition and nutrient exhaustion within hours.
Statistical inconsistency (Fig 1A, 1C, 6C, 6D vs. Methods): Figure legends indicate unpaired Student's t tests for tumor growth curves, while the methods section states that "repeated measurements of tumor growth were assessed by two-way ANOVA." t tests on endpoints of repeated time-course data are inappropriate.
Image-integrity risk flag (Fig 4B, 4C, 5A, 6A, 6J): Numerous Western blots probing TIPE2, C/EBPβ isoforms (LAP*, LAP, LIP), p-AKT, and p-STAT3 represent a known high-risk category for band reuse or splicing; original TIFF files are required.
Timeline check: The PDF header "Downloaded … on 27 May 2026" is consistent with the stated publication and review record.

Evidence highlights

DOI: 10.1084/jem.20182005
Reported H₂O₂ working concentration: 250 mM (vs. literature norm 10–500 µM; factor ≈ 500–25,000×).
Reported B16F10 seeding density: 1 × 10⁷ cells/ml (vs. norm 0.1–2 × 10⁶ cells/ml; factor ≈ 5–100×).
Statistical methods text vs. figure legend discrepancy for Figures 1A, 1C, 6C, and 6D.
Affiliations listed in the report: Chinese Academy of Sciences (Shenzhen Institute of Advanced Technology) and University of Pennsylvania.

Notes

All quantitative claims are taken directly from the report's text excerpts; values were not independently verified against the published PDF.
No pixel-level image analysis was performed; the image-duplication flag is precautionary.
The author's tone in the original Chinese review is informal and was neutralized here; underlying substantive concerns are retained.
Recommended actions: request raw data and uncropped blots from the corresponding author; post a structured concern on PubPeer; notify the JEM editorial office. Institutional notification has not yet been recommended by the reviewer.