In bacteria, a single RNA polymerase core enzyme is known to be responsible for synthesizing all RNAs, including messenger RNAs (mRNAs), ribosomal RNAs (rRNAs) and transfer RNAs (tRNAs). However, until the early 1970s, it was unclear whether the same was true for eukaryotic cells.
At the time, several lines of evidence implied that a distinct polymerase transcribed rRNAs. The (G+C)-rich base composition of rRNAs was distinct from other RNA classes, their genes were heavily redundant and localized in the nucleolus, and rRNA synthesis was regulated independently from the other RNA types.
In late 1969, Roeder and Rutter reported the discovery of three chromatographically separable forms of eukaryotic RNA polymerase from sea urchin embryos (I, II and III) and two species from rat liver (I and II). Further characterization of the polymerases showed differing salt requirements for maximal activity, and indicated that sea urchin polymerase I and II were similar to their respective enzymes in rat. The authors proposed that polymerase I and II localized primarily in the nucleolus and the nucleoplasm, respectively. Furthermore, they proposed that increases in polymerase I levels were responsible for observed increases in rRNA levels during gastrulation. However, it was still not known whether the three forms of polymerase were the products of three distinct genes or simply the result of the differential regulation of a single gene product.
Work published by the Chambon laboratory a few months later began to clarify things. They examined the ability of the toadstool toxin 
-amanitin to inhibit two RNA polymerase activities, which they called A and B, isolated from calf thymus. Eukaryotic RNA polymerase activity was known to be inhibited by the toxin, in contrast to RNA polymerase from Escherichia coli. Chambon and colleagues showed that polymerase A activity was insensitive to -amanitin, whereas polymerase B activity was inhibited. In addition, their study showed that -amanitin affected the elongation stage of RNA transcription. This work supported previous studies indicating that the bacterial RNA polymerase differed from eukaryotic RNA polymerases, and allowed the researchers to speculate that there were structural differences between polymerases A (I) and B (II).
Shortly thereafter, several laboratories showed that distinct forms of RNA polymerases could be distinguished based on their sensitivities to -amanitin — with polymerase I being insensitive, polymerase II being inhibited and polymerase III being only moderately affected. By correlating the -amanitin sensitivities of the isolated enzymes with the concentrations of -amanitin required to inhibit synthesis of different classes of RNA in cells, these studies supported the idea that RNA polymerases I, II and III, respectively, are responsible for the synthesis of the major rRNAs, mRNA and small RNAs (including 5S RNA, tRNA and a subset of the small-nuclear RNAs). Despite these differences in gene targets, the later discovery that the TATA-binding protein was a transcription factor used by all three polymerases enforced the evolutionary relatedness of these enzymes.
