The mechanism describing the guanine specific hydrolysis by the different microbial endoribonucleases which has been proposed cannot account in barnase (B.amyloliquefaciens) in the case of GpN dinucleotide hydrolysis, the observed preference for N being A≳G≳C≳U. Similarly the much higher activity toward long RNA molecules as compared to dinucleotides is not understood. A possible explanation for these observations is the existence in barnase of secondary substrate binding sites.