Introduction Deoxycholic acid (DCA), a component of bile, has been associated with the development and progression of Barrett's oesophagus (BO). In addition, the “mutator-phenotype hypothesis” states benign cells with high rates of random mutations have an inherent predisposition for malignant conversion. Mitochondria are highly susceptible to random mutations due to inefficient DNA repair mechanisms. The role of mitochondrial mutations and dysfunction in Barrett's progression is unknown. The aim of this study was to investigate the levels of random mitochondrial mutations (RMMs) and dysfunction in BO and determine if DCA drives progression. In parallel, determine if a hydroxylase inhibitor, dimethyloxalylglycine (DMOG), could rescue these effects.

Methods We assessed mitochondrial dysfunction and mutagenesis using different models in vitro, ex vivo and in vivo. RMMs in BO extending to oesophageal adenocarcinoma (OAC) were examined. Mitochondrial function (reactive oxygen species (ROS), mitochondrial membrane potential (MMP), mitochondrial mass and cytochrome c) +/− DCA +/− DMOG were measured. A large scale gene array assessing oxidative stress and antioxidant defence was used to determine potential genes DMOG therapy may target.

Results In vitro; RMMs were higher in Qh (intestinal metaplasia [IM] cells) compared with GO (high grade dysplasia [HGD] cells) (p=0.06) and OE33 (OAC cells) (p=0.01). DCA induced mitochondrial dysfunction at all stages of disease progression (p<0.04). DCA increased mutagensis in Qh cell lines only. Using ex vivo explants, cytochrome c release was significantly increased in Barrett's compared to matched normal controls (p=0.0006). DCA significantly decreased cytochrome c secretion in Barrett's patients (p=0.015). In vivo; two distinct Barrett's IM groups were evident; those with low RMMs and a small cohort with significantly higher levels of RMMs (p<0.005). DMOG caused a significant reduction in mitochondrial dysfunction in Go and OE33 cells (p<0.05). DMOG down-regulated 3 genes, CYGB, FOXM1 and GLRX2 (13, 6 and 4 fold respectively), validations of these are being performed.

Conclusion Mitochondrial instability appears to be an early event in BO. DCA significantly alters mitochondrial function. Two distinct Barrett's patient groups are evident. Applying the “mutator-phenotype hypothesis”; random mitochondrial mutation may act as a potential biomarker in differentiating IM patients into low and high risk groups for malignant conversion. DMOG may have the potential to reduce this mitochondrial instability in Barrett's oesophagus.

Competing interests None declared.