Rpp20 and Rpp25 are subunits of the human RNase MRP and RNase P endoribonucleases belonging to the Alba superfamily of nucleic acid binding proteins. These proteins, which bind very strongly to each other, transiently associate with RNase MRP. Here, we show that the Rpp20-Rpp25 heterodimer is resistant to both high concentrations of salt and a nonionic detergent. The interaction of Rpp20 and Rpp25 with the P3 domain of the RNase MRP RNA appeared to be strongly enhanced by their heterodimerization. Coimmunoprecipitation experiments demonstrated that only a single copy of each of these proteins is associated with the RNase MRP and RNase P particles in HEp-2 cells. Both proteins accumulate in the nucleoli, which in case of Rpp20 is strongly dependent on its interaction with Rpp25. Finally, the results of overexpression and knock-down experiments indicate that their expression levels are codependent. Taken together, these data indicate that the Rpp20-Rpp25 heterodimerization regulates their RNA-binding activity, subcellular localization, and expression, which suggests that their interaction is also crucial for their role in RNase MRP/P function.

RNase MRP is a eukaryotic endoribonuclease involved in nucleolar and mitochondrial RNA processing events. RNase MRP is a ribonucleoprotein particle, which is structurally related to RNase P, an endoribonuclease involved in pre-tRNA processing. Most of the protein components of RNase MRP have been reported to be associated with RNase P as well. In this study we determined the association of these protein subunits with the human RNase MRP and RNase P particles by glycerol gradient sedimentation and coimmunoprecipitation. In agreement with previous studies, RNase MRP sedimented at 12S and 60-80S. In contrast, only a single major peak was observed for RNase P at 12S. The analysis of individual protein subunits revealed that hPop4 (also known as Rpp29), Rpp21, Rpp20, and Rpp25 only sedimented in 12S fractions, whereas hPop1, Rpp40, Rpp38, and Rpp30 were also found in 60-80S fractions. In agreement with their cosedimentation with RNase P RNA in the 12S peak, coimmunoprecipitation with VSV-epitope-tagged protein subunits revealed that hPop4, Rpp21, and in addition Rpp14 preferentially associate with RNase P. These data show that hPop4, Rpp21, and Rpp14 may not be associated with RNase MRP. Furthermore, Rpp20 and Rpp25 appear to be associated with only a subset of RNase MRP particles, in contrast to hPop1, Rpp40, Rpp38, and Rpp30 (and possibly also hPop5), which are probably associated with all RNase MRP complexes. Our data are consistent with a transient association of Rpp20 and Rpp25 with RNase MRP, which may be inversely correlated to its involvement in pre-rRNA processing.