Polynucleotide Phosphorylase:
     In 1955, Marianne Grunberg-Manago and Severo Ochoa reported the isolation of an enzyme that catalyzed the synthesis of RNA. Their work built upon the earlier work of Jerard Hurwitz & J.J. Furth who performed experiments to see if isolated E. coli protein fractions could polymerize radioactively labeled nucleotides. Later Grunberg-Manago & Ochoa tested a protein fraction that could make RNA. For this work, Ochoa shared the 1959 Nobel Prize (but not Grunberg-Manago) in Medicine with Arthur Kornberg (who received the Prize for his work on DNA polymerase I).
Their enzyme could convert ribonucleoside diphosphates into RNA:
            (RNA)n + NDP   --►   (RNA)n+1 + Pi   
However, the enzyme had a number of unsettling properties. It did not need a template; and, it could use as little as 1 NDP or as many as 4 NDPs as substrate. In fact, the sequence of the product RNA depended entirely on the number and concentration of substrate NDPs.  These are not the properties of an enzyme that must faithfully copy the genetic material for expression!   We now know that Grunberg-Manago and Ochoa had isolated the enzyme polynucleotide phosphorylase which usually catalyzes the breakdown of RNA - not its synthesis! i.e., its a ribonuclease.
               (RNA)n + NDP   ◄--    (RNA)n+1 + Pi

                      

Polynucleotide  Phosphorylase

Authors:  M. F. Symmons, G. H. Jones & B. F. Luisi
Reference: Structure, 8:1215-1226, 2000
Description: PNPase is a disk-like trimeric exoribonuclease.
   This side view of the trimer shows the interface between two subunits. The cores of subunits are constructed from two homologous domains shown in shades of red and blue. The lower and upper accessory domains are in green and grey respectively.

                          The channel through the PNPase has been suggested to function in the entrapment of the enzyme's single-stranded RNA substrate. The cores of each of the trimer subunits are formed from pairs of homologous domains colored here as the pairs violet and blue, blue-green and green, and finally orange and red respectively.
                           This view of the channel through the center of the trimer channel shows sidechains of the conserved 'FFRR loops' that may act as a mechanism to entrap RNA substrate and select against base-paired segments. This provides a possible structural explanation for both the high processivity and the regulation of the PNPase activity by RNA secondary structure.