Apyrases (NTP-diphosphohydrolases, EC 3.6.1.5) are a family of enzymes that catalyze the hydrolysis of phosphoanhydride bonds of nucleoside tri- and di- phosphates in the presence of divalent cations. In Arabidopsis, AtAPY1 and AtAPY2 function in part as ectoapyrases and have been shown to play important roles in controlling the concentration of extracellular nucleotides, which, in turn, can regulate pollen germination and growth, and cell expansion in diverse plant tissues. We used an NCBI nucleotide blast keyed to Apyrase Conserved Regions (ACRs) to identify five other AtAPYs (3-7). To assess the biological function of each of these five AtAPY genes, the phenotypes of their T-DNA insertion mutants were analyzed. We did not observe any obvious phenotypes from the T-DNA insertion single knockout of any of these genes. However, double knockout mutants of AtAPY6 and 7 exhibited late anther dehiscence and low male fertility. Transmission and scanning electron microscopy revealed that the pollen grains of double knockout mutant of AtAPY6 and 7 are largely deformed in shape and in most cases, the cell walls of the pollen grains are interconnected. Using an AtAPY6-YFP fusion protein in transgenic Arabidopsis, AtAPY6 was localized to intracellular vesicles. Quantitative reverse transcription polymerase chain reaction (qRT-PCR) assays and promoter:GUS fusion analysis demonstrated that the transcripts of both AtAPY6 and 7 are expressed in mature pollen grains, AtAPY6 is also expressed in the veins and hydathode regions of rosette leaves, and AtAPY7 is expressed in more diverse tissues such as roots, leaves, stems, pistils and sepals. The tissue specificity of AtAPYs 3, 4 and 5 expression was also determined using qRT-PCR assays and promoter:GUS fusion analysis. AtAPY3 and AtAPY4 were primarily expressed in roots but not in rosette leaves. AtAPY5 was expressed primarily in rosette leaves but only weakly in roots. AtAPY5 and AtAPY7 were upregulated when the rosette leaves are wounded or exposed to drought stress. RNA interference (RNAi) was performed to simultaneously suppress the gene expression of AtAPYs 3, 4, 5 to around 10% of that of the wild type. However, we did not observe any obviously altered phenotypes in the RNAi lines. The suppression of AtAPYs 3, 4, 5 by RNAi in the background of AtAPY6 single knockout did not cause any phenotype either. A possible explanation for this lack of phenotype in the RNAi lines is that functional redundancy exists between AtAPYs 3-5 and AtAPYs 1-2 and/or AtAPYs 6-7. / text