Concerns regarding 'Mitophagy activation for jawbone radiation injury therapy via angiogenesis/osteogenesis-active nanozymes' (DOI: 10.1007/s11427-025-3104-4)

Academic fraud report · Geng Detector

Summary

This report evaluates a 2026 publication in SCIENCE CHINA Life Sciences by Tianxiao Wang and colleagues concerning jawbone radiation injury therapy. The verdict is severe: multiple findings strongly suggest academic misconduct. First, Figure 4A (EC cells) and Figure 4E (HBMSCs) AM/PI live/dead staining images display near-identical background textures and cell distribution patterns across the 0h, 24h, and 48h time points and across Control, NBP, and CeSNs groups, indicating image reuse inconsistent with independent biological replicates. Second, a glaring methodological inconsistency: the Results state RNA-seq was performed on human bone marrow mesenchymal stem cells (HBMSCs) treated with 6 Gy and 6 Gy+NBP@CeSNs, yet the Methods section states RNA was extracted from ECs and aligned to the mouse reference genome (GRCm38), implying a direct copy-paste error that also questions data authenticity. Third, Figure 5C HSP60/LC3 immunofluorescence images across treatment groups show suspiciously similar patterns. Fourth, Figure 6A Transwell migration images appear unnaturally uniform. Overall confidence is high for findings 1 and 2; findings 3 and 4 are suggestive but require raw image verification.

Verdict

🔴 Confirmed concerns — The report identifies multiple serious issues, most notably apparent image duplication in Figure 4A/4E and a severe cross-species/cross-cell-type methodological contradiction in the RNA-seq description. These findings collectively raise strong doubts about data integrity and warrant formal institutional investigation.

Key findings

Image reuse in Figure 4A and 4E: AM/PI staining images for EC cells (Figure 4A) and HBMSCs (Figure 4E) at 0h, 24h, and 48h across Control, NBP, and CeSNs groups display highly similar or near-identical background textures and cell distribution patterns, inconsistent with independent experimental replicates.
Methods/Results contradiction in RNA-seq: The Results section states RNA-seq was performed on human HBMSCs (human bone marrow mesenchymal stem cells) treated with 6 Gy and 6 Gy+NBP@CeSNs (Page 2367), while the Methods "RNA sequencing" subsection (Page 2372) states: "Total RNA was extracted from ECs... Cleaned reads were aligned to the mouse reference genome (GRCm38) using HISAT2." This contradiction between cell type (HBMSCs vs ECs) and species (human vs mouse) strongly suggests an unedited copy-paste from another paper's Methods.
Possible over-manipulation of Figure 5C: HSP60/LC3 immunofluorescence images in EC cells across Control, 6 Gy, and 6 Gy+NBP groups show highly similar noise patterns and cellular outlines, suggestive of repeated use with brightness adjustment.
Statistical chart visual anomaly in Figure 6A: Transwell migration crystal violet staining images show unnaturally uniform and regular cell cluster morphology, inconsistent with the expected randomness of migration assays.

Evidence highlights

DOI: 10.1007/s11427-025-3104-4
Figure 4A (EC cells) AM/PI staining at 0, 24, 48 h under Control, NBP, CeSNs groups — visual duplication with Figure 4E
Figure 4E (HBMSC cells) AM/PI staining at 0, 24, 48 h — background textures and cell positions closely match those of Figure 4A
Page 2367 (Results): RNA-seq performed on HBMSCs treated with 6 Gy and 6 Gy+NBP@CeSNs
Page 2372 (Methods): "Total RNA was extracted from ECs... aligned to the mouse reference genome (GRCm38) using HISAT2"
Figure 5C: HSP60/LC3 dual immunofluorescence in ECs under Control, 6 Gy, 6 Gy+NBP — similar noise pattern across groups
Figure 6A: Transwell migration assay images — uniform, textbook-like cell distribution

Notes

Findings 1 and 2 are based on direct textual and figure-caption evidence in the published article and are considered reliable.
Findings 3 and 4 are based on visual assessment; raw, unprocessed images would be required for conclusive determination.
The cross-species genome alignment contradiction (human HBMSCs aligned to mouse GRCm38) is particularly damaging: it not only indicates a copy-paste error but also raises fundamental questions about the validity of any RNA-seq results.
Recommended actions: request raw FASTQ files for RNA-seq and unprocessed original images for Figures 4, 5, and 6 from the authors; post concerns on PubPeer; report to the journal editorial office and to the institutional research integrity committee (Nanjing Medical University affiliation noted).