Differential evolution (DE) is an efficient and robust evolutionary algorithm, which has been widely and successfully applied to solve global optimization problems. Although many methods have been developed based on the population topology to improve the performance of DE, the effects of population topology interacted with the functions being optimized are not considered in most of the algorithm designs. Moreover, the synergy among multiple population topologies has not been systematically exploited in DE. In this paper, a novel DE algorithm with neighborhood-dependent mutation operator, named neighborhood-dependent DE (NDE), is proposed. In NDE, a pool of population topologies is used to define multiple neighborhood relationships for each individual and then the neighborhood relationships are adaptively selected for the functions being solved during the evolutionary process. The experimental results on the benchmark functions from CEC2013 demonstrate that NDE is able to enhance the performance of the studied DE algorithms. Additionally, the synergy among different neighborhood topologies in NDE is also revealed.