Investigation Report: Butyrate Suppresses Glucose Metabolism of Colorectal Cancer Cells via GPR109a-AKT Signaling Pathway and Enhances Chemotherapy

Academic fraud report · Geng Detector

Summary

This report assesses a 2021 Frontiers in Molecular Biosciences paper (DOI: 10.3389/fmolb.2021.634874) by Geng et al. The overall verdict is highly suspicious. Four main concerns are identified: (1) Results text contains an implausibly high frequency of round-numbered percentages and fold-changes (e.g., 60%, 70%, 1.5x, 2.5x, 5x), inconsistent with the natural variance expected from Western blot densitometry and metabolomics quantification; (2) The Methods section describes CCK-8 (WST-8) assays read at 590 nm, which contradicts the manufacturer's specification of ~450 nm—the 590 nm wavelength is characteristic of legacy MTT protocols, strongly suggesting copy-paste from a prior manuscript; (3) The submission-to-acceptance window of only 74 days for a multi-omics, multi-assay study involving CE-TOFMS metabolomics, membrane fractionation, flow cytometry, qPCR, and siRNA work is implausibly short for a five-author team; (4) Triplicate experiments allegedly produced overly concordant trends without visible standard deviations in textual descriptions. Limitations: image-based reuse and splicing checks could not be performed without high-resolution figure files. Confidence is high on the wavelength error and suspicious round-numbering, moderate on production-volume claims.

Verdict

Highly Suspicious (高度可疑). Multiple independent textual and methodological anomalies converge. The wavelength error in the CCK-8 protocol is, on its own, a near-fatal internal inconsistency. Combined with the suspiciously round data values and an implausibly short publication timeline, the paper warrants formal institutional investigation and journal-level inquiry.

Key findings

Implausibly round numeric values across Results. The text reports effects as perfectly rounded percentages and fold-changes (60%, 70%, 1.5x, 2.5x, 5x, 3.5x, 80% knockdown), atypical for genuine densitometry or LC-MS/CE-TOFMS quantification.
Critical methodological error: CCK-8 read at 590 nm. The Methods specify absorbance at 590 nm for the Dojindo Cell Counting Kit-8 (WST-8). The correct wavelength is ~450 nm; 590 nm corresponds to legacy MTT protocols. This strongly indicates copied Methods text.
Impossibly short research-to-publication cycle. Received 29 Nov 2020, accepted 11 Feb 2021 (~74 days), during which a five-author team purportedly completed extensive metabolomics, multiple Western blots (including membrane fractionation), siRNA optimization, flow cytometry, qPCR, CCK-8, and BrdU assays.
Idealized triplicate reproducibility. "Independent triplicates" are claimed to yield highly concordant trends for inherently variable assays (membrane protein CBB staining, metabolomics), with no visible standard-deviation reporting in the Results narrative.
Image forensics not performed. Pixel-level reuse and splicing checks (Geng methods 1 and 3) were not conducted because only text-extracted PDF content was available.

Evidence highlights

Wavelength quote: "Absorbance at 590 nm was recorded using a microplate reader... Cell Counting Kit-8 (Dojindo...)"—contradicts Dojindo CCK-8 specification (~450 nm).
Round-number exemplars (Figure 1–3 textual descriptions):
Glucose uptake "dropped from 38 to 27% and from 36 to 24%"
"around 70% decrease" in downstream signaling
Membrane fraction reduction "by 60 and 40%"
GPR109a "knockdown by around 80%"
Increases "by about 1.5 and 2.5 times" and "five times and 3.5 times"
Lactate "reduced by 60%"; R-5P-related intermediates "reduced... by 70, 58, and 60%"
Timeline: Received 29 Nov 2020 → Accepted 11 Feb 2021 (≈74 days).
DOI: 10.3389/fmolb.2021.634874

Notes

Confidence is high for the CCK-8 wavelength mismatch and the suspiciously round numeric reporting; moderate for the production-rate inference; low/none for image manipulation, which remains untested.
Recommended follow-ups: request raw microplate-reader exports for viability assays, request raw immunoblot images and CE-TOFMS chromatograms, file a PubPeer comment highlighting the wavelength error and the round-number pattern, and audit prior publications from the corresponding author (Yun Yang) for systematic reuse of Methods boilerplate.
This report was AI-assisted and is intended for academic-discussion purposes only; final determination of misconduct requires institutional investigation.