Characterization of a Novel Ribonucleotide-polymerizing Enzyme from a Fungus, Histoplasma capsulatum*

Yeast-like cells of the pathogenic dimorphic fungus Histoplasma eapsulatum have been found to contain a novel ribonucleotide-incorporating enzyme, which we have termed ribonucleotide polymerase. The enzyme, not detectable in the mycelial form of the fungus, is capable of incorporating any of the ribonucleoside triphosphates into short oligonucleotide chains (5 to 6 residues/average chain). Ribonucleotide polymerase has an absolute requirement for Mn’+ as the divalent cation; Mg”+ cannot replace MI?+ at any concentration tested (up to 9.6 mM). DNA template is not required. Activity is inhibited by pyrophosphate but not by phosphate ions. Heparin and rifamycin AF/013 are strongly inhibitory, while actinomycin D and cu-amanitin have no effect on the enzyme. Ribonucleotide polymerase shows a rapid incorporation of ribonucleoside triphosphates into oligonucleotide product. The reaction reaches a plateau within 20 to 30 min, with little additional incorporation upon extended incubation. The cessation of incorporation is due to inhibition by the product which remains tightly bound to the enzyme. Ribonucleotide polymerase has been purified by two glycerol gradient centrifugations and treatment with ribonuclease Tl. The purified enzyme (estimated molecular weight of 900,000) yields a single band upon electrophoresis on polyacrylamide gels under nondenaturing conditions. Electrophoresis under denaturing conditions reveals two bands of protein with estimated molecular weights of 75,000 and 69,000 present in approximately 1:l molar ratio.

Yeast-like cells of the pathogenic dimorphic fungus Histoplasma eapsulatum have been found to contain a novel ribonucleotide-incorporating enzyme, which we have termed ribonucleotide polymerase. The enzyme, not detectable in the mycelial form of the fungus, is capable of incorporating any of the ribonucleoside triphosphates into short oligonucleotide chains (5 to 6 residues/average chain). Ribonucleotide polymerase has an absolute requirement for Mn'+ as the divalent cation; Mg"+ cannot replace MI?+ at any concentration tested (up to 9.6 mM). DNA template is not required. Activity is inhibited by pyrophosphate but not by phosphate ions. Heparin and rifamycin AF/013 are strongly inhibitory, while actinomycin D and cu-amanitin have no effect on the enzyme.
Ribonucleotide polymerase shows a rapid incorporation of ribonucleoside triphosphates into oligonucleotide product. The reaction reaches a plateau within 20 to 30 min, with little additional incorporation upon extended incubation. The cessation of incorporation is due to inhibition by the product which remains tightly bound to the enzyme. Ribonucleotide polymerase has been purified by two glycerol gradient centrifugations and treatment with ribonuclease Tl. The purified enzyme (estimated molecular weight of 900,000) yields a single band upon electrophoresis on polyacrylamide gels under nondenaturing conditions. Electrophoresis under denaturing conditions reveals two bands of protein with estimated molecular weights of 75,000 and 69,000 present in approximately 1:l molar ratio.  (12,14). Whether this also accounts for the absence of ribonucleotide polymerase activity is at present unknown.
As Fig. 1 shows, the yeast ribonucleotide polymerase activity can be detected in two regions of the gradient. A obtained by glycerol gradient centrifugation (Fig. 1)      7. Polvacrvlamide gel electrophoresis of ribonucleotide polymerase A. &ei A, ribonucleotide bolymerase (6.0 pg) obtained from the second glycerol gradient (Table I) was subjected to electrophoresis under n&denaturing conditions as described under "Experimental Procedures." After electrophoresis one gel was stained for product (Table V), as if it were tightly bound to the polymerase and thus protected from digestion. Binding of Oligonucleotides to Ribonucleotide Polymeruse -Further evidence for the interaction of ribonucleotide polymerase product with the enzyme was obtained when the ribonucleotide polymerase reaction mixture was centrifuged on a 15 to 30% glycerol gradient, and the distribution of radioactivity was studied (Fig. 6). When the ribonucleotide polymerase was incubated under standard reaction conditions the bulk of the label sedimented at exactly the same rate as the light enzyme species, ribonucleotide polymerase C ( Fig. 11 and a consistently smaller, but significant, amount sedimented at the ribonucleotide polymerase A position. However, when incubation was carried out in the presence of RNase Tl, the radioactivity was redistributed so that more product was found in the rapidly sedimenting fractions (note a one-fraction shift in sedimentation).
Finally, when the material was treated with pronase, essentially all radioactivity remained at the top of the gradient. Therefore, it is likely that oligouridylate formed in the reaction binds tightly to ribonucleotide polymerase proteins unless some contaminating proteins exhibit the same sedimentation velocities as the two ribonucleotide species (also see "Discussion").
In addition, dialysis of ribonucleotide polymerase reaction mixture resulted in the removal of unreacted UTP but there was no loss of radioactivity in the oligonucleotide material. Had this material not been bound, it would have been removed by dialysis.
Polyacrylamide Gel Electrophoresis -When samples of ribonucleotide polymerase A were subjected to electrophoresis, under nondenaturing conditions, a single band appeared on staining with Coomassie blue (Fig. 7A). The band migrated relatively slowly in 3.85% acrylamide gel indicating, as expected, a high molecular weight protein. As shown in the figure, the position of the protein band coincided with the pattern of ribonucleotide polymerase activity in the parallel unstained gel, although the enzyme recovery was quite poor (less than 5% of the amount applied to the gel).
protein and scanned (----), a parallel gel was sliced, and ribonucleotide polymerase activity in the slices were measured (--4. Panel B, ribonucleotide polymerase (5.5 pg) was treated with sodium dodecylsulfate and electrophoresed as described under "Experimental Procedures." In both panels, the anode end of the gels is indicated by +.

Electrophoresis
under denaturing conditions revealed the presence of two prominent bands of protein (Fig. 7Bl. The molecular weights of these proteins were estimated to be 75,000 (slower band) and 69,000 (faster band). On the basis of densitometer scanning we conclude that the two components are present at about 1:l molar ratio. However, these data are not sufficient to provide firm clues as to the possible subunit structure of ribonucleotide polymerase.

Relation to Known
Ribonucleotide Polymerizing Enzymes -There have recently been a number of reports describing the occurrence of enzymes catalyzing the synthesis of different polynucleotides in a variety of organisms (2'7-351. Our results indicate that the ribonucleotide polymerase activity isolated from yeast phase extracts of H. capsulatum differs from these enzymes in several respects. In contrast to the majority of polynucleotide polymerases, most of ribonucleotide polymerase activity is associated with a very rapidly sedimenting fraction of the yeast extract (HSE). Niessing and Sekeris (27) have shown that rat liver nuclei contain 30 S ribonucleoprotein particles capable of polymerizing UTP, CTP, GTP, and ATP to the corresponding ribohomopolymers. However, the polymers formed in their preparations, and in other systems as well (28,31,321, were of greater length than those made by the ribonucleotide polymerase, and the incorporation was linear for long periods of time as compared to only about 20 min for ribonucleotide polymerase. In addition, the incorporation was stimulated by the presence of RNA or other polynucleotides, indicating a primer requirement for the activity of those enzymes. We could not demonstrate polynucleotide stimulation in ribonucleotide polymerase system even after an extensive treatment of ribonucleotide polymerase with RNase A or Tl, although the enzyme does seem to use an RNA primer (tightly bound). Moreover, the ribonucleolytic treatment of ribonucleotide polymerase did not seem to affect the integrity of the enzyme as evidenced by its sedimentation behavior and catalytic properties. The small size of the oligonucleotide product, the utilization of any of the individual nucleoside triphosphates (or combination thereof), and the lack of dependence on DNA template (Table II, lines one and two) make it unlikely that ribonucleotide polymerase is involved in the process of transcription. Since the enzyme produces homooligonucleotides with an average chain length of 5 to 6 residues, a DNA or RNA template of 20 to 24 nucleotides (or 10 to 12 nucleotide pairs, if doublestranded) would be needed to account for incorporation of each of the four nucleoside triphosphates.
A template of such length should be at least partially sensitive to RNase or DNase attack because even the short ribonucleotide polymerase product itself is about 80% degraded by RNase (Table  V). Instead, a stimulation of incorporation was observed after RNase treatment (Table IV), and even a 60-min incubation with DNase I prior to assay had no effect on ribonucleotide polymerase activity (