Sulfate assimilation provides reduced sulfur for synthesis of the amino acids cysteine and methionine and for a range of other metabolites. Sulfate has to be activated prior to reduction by adenylation to adenosine 5'-phosphosulfate (APS). In plants, algae, and many bacteria, this compound is reduced to sulfite by APS reductase (APR); in fungi and some cyanobacteria and gamma-proteobacteria, a second activation step, phosphorylation to 3'-phosphoadenosine 5'-phosphosulfate (PAPS), is necessary before reduction to sulfite by PAPS reductase (PAPR). We found previously that the moss Physcomitrella patens is unique among these organisms in possessing orthologs of both APR and PAPR genes (Koprivova, A., Meyer, A. J., Schween, G., Herschbach, C., Reski, R., and Kopriva, S. (2002) J. Biol. Chem. 277, 32195-32201). To assess the function of the two enzymes, we compared their biochemical properties by analysis of purified recombinant proteins. APR from Physcomitrella is very similar to the well characterized APRs from seed plants. On the other hand, we found that the putative PAPR preferentially reduces APS. Sequence analysis, analysis of UV-visible spectra, and determination of iron revealed that this new APR, named PpAPR-B, does not contain the FeS cluster, which was previously believed to determine the substrate specificity of the otherwise relatively similar enzymes. The lack of the FeS cluster in PpAPR-B catalysis is connected with a lower turnover rate but higher stability of the protein. These findings show that APS reduction without the FeS cluster is possible and that plant sulfate assimilation is predominantly dependent on reduction of APS.