The amyloid-beta precursor protein (APP) and its pathogenic byproduct beta-amyloid peptide (Abeta) play central roles in the pathogenesis of Alzheimer's disease (AD). Reduction in levels of the potentially toxic Abeta is one of the most important therapeutic goals in AD. Recent studies have shown that bivalent metals such as iron, copper, and zinc are involved in APP expression, Abeta deposition, and senile plaque formation in the AD brain. However, the underlying mechanisms involved in abnormal homeostasis of bivalent metals in AD brain remain unclear. In the present study, we found that two isoforms of the divalent metal transporter 1 (DMT1), DMT1-IRE, and DMT1-nonIRE, were colocalized with Abeta in the plaques of postmortem AD brain. Using the APP/PS1 transgenic mouse model, we found that the levels of both DMT1-IRE and DMT1-nonIRE were significantly increased in the cortex and hippocampus compared with wild type-control. We further verified the proposed mechanisms by which DMT1 might be involved in APP processing and Abeta secretion by using the SH-SY5Y cell line stably overexpressing human APP Swedish mutation (APPsw) as a cell model. We found that overexpression of APPsw resulted in increased expression levels of both DMT1-IRE and DMT1-nonIRE in SH-SY5Y cells. Interestingly, silencing of endogenous DMT1 by RNA interference, which reduced bivalent ion influx, led to reductions of APP expression and Abeta secretion. These findings suggest both that DMT1 plays a critical role in ion-mediated neuropathogenesis in AD and that pharmacological blockage of DMT1 may provide novel therapeutic strategies against AD.