A former researcher at the University of Pittsburgh inflated the number of mice used in his experiments, and faked data in a number of images in a paper reporting the results, according to the Office of Research Integrity (ORI).
Dong Xiao admitting to having
intentionally fabricated data contained in a paper entitled ‘Guggulsterone inhibits prostate cancer growth via inactivation of Akt regulated by ATP citrate signaling,’ specifically Figure 6G,
the ORI reports. The paper was published in in July 2014 in Oncotarget. Here’s Figure 6:
Figure 6: Anticancer effect of Z-Gug is mediated by mitochondrial ROS generation. Flow cytometric measurement of MitoSOX Red fluorescence in PC-3 (A) and LNCaP (B) cells treated with DMSO (control) or 20 or 40 µM Z-Gug. Results shown are mean ± S.E. total sample size is n = 4 per group. *significantly different (P<0.05) between the indicated groups by one-way ANOVA followed by Dunnett’s test. Effect of NAC on Z-Gug (40 µM 4 h for flow cytometric measurement of MitoSOX Red fluorescence, and 24 h for cytoplasmic histone-associated DNA fragmentation in PC-3 (C and E) and LNCaP (D and F) cells treated with or without 2 h pretreatment of 4 mM NAC. The detailed treatment information was described in the section of Methods. For C to F, Columns, mean of three determinations; bars, SE. *Significantly different (aP < 0.05) compared with DMSO-treated control and (bP < 0.05) compared with Z-Gug-treated group by one-way ANOVA followed by Dunnett’s test. Similar results were observed in two independent experiments. Representative data from a single experiment are shown. Immunoblotting analysis of S473phospho-Akt was performed using lysates from the PC-3 (G) and LNCaP (H) cells treated with or without 2 h pretreatment of 4 mM NAC. The blot was stripped and reprobed with-anti-β-tubulin antibody to ensure equal protein loading. The numbers on top of the immunoreactive bands represent change in levels relative to DMSO-treated control. Immunoblotting for each protein was performed at least twice using independently prepared lysates.
According to the ORI, Xiao’s misdeeds went beyond Figure 6G. He also reported that 10 mice per group were used in experiments, when he only had data for four. And he faked figures 1D, 4C, 4D, 5C, 5D, 6H, S2B, and S3B.
Xiao agreed to have his research supervised for three years, and to not serve on any NIH peer review committees for that period of time.
Update, 8:30 a.m. Eastern, 1/27/15: A Pitt spokesperson tells us:
…the University conducted an inquiry, self-reported the matter, provided the information to the ORI and cooperated with their inquiry.