Saccharomyces cerevisiae prion [PSI ] is a self‐propagating isoform of the eukaryotic release factor eRF3 (Sup35p). Sup35p consists of the evolutionary conserved release factor domain (Sup35C) and two evolutionary variable regions – Sup35N, which serves as a prion‐forming domain in S. cerevisiae, and Sup35M. Here, we demonstrate that the prion form of Sup35p is not observed among industrial and natural strains of yeast. Moreover, the prion ([PSI + ]) state of the endogenous S. cerevisiae Sup35p cannot be transmitted to the next generations via heterologous Sup35p or Sup35NM, originating from the distantly related yeast species Pichia methanolica. This suggests the existence of a ‘species barrier’ in yeast prion conversion. However, the chimeric Sup35p, containing the Sup35NM region of Pichia, can be turned into a prion in S. cerevisiae by overproduction of the identical Pichia Sup35NM. Therefore, the prion‐forming potential of Sup35NM is conserved in evolution. In the heterologous system, overproduction of Pichia Sup35p or Sup35NM induced formation of the prion form of S. cerevisiae Sup35p, albeit less efficiently than overproduction of the endogenous Sup35p. This implies that prion induction by protein overproduction does not require strict correspondence of the ‘inducer’ and ‘inducee’ sequences, and can overcome the ‘species barrier’.

To test the hypothesis that inaccurate DNA synthesis by mammalian DNA polymerase (pol ) contributes to somatic hypermutation (SHM) of Ig genes, we measured the error specificity of mouse pol during synthesis of each strand of a mouse Ig light chain transgene. We then compared the results to the base substitution specificity of SHM of this same gene in the mouse. The in vitro and in vivo base substitution spectra shared a number of common features. A highly significant correlation was observed for overall substitutions at A-T pairs but not for substitutions at G-C pairs. Sixteen mutational hotspots at A-T pairs observed in vivo were also found in spectra generated by mouse pol in vitro. The correlation was strongest for errors made by pol during synthesis of the non-transcribed strand, but it was also observed for synthesis of the transcribed strand. These facts, and the distribution of substitutions generated in vivo, support the hypothesis that pol contributes to SHM of Ig genes at A-T pairs via short patches of low fidelity DNA synthesis of both strands, but with a preference for the non-transcribed strand.H igh affinity antibodies result from somatic hypermutation (SHM) of Ig genes followed by selection. The SHM process introduces base substitutions at a very high rate into DNA encoding the variable regions of immunoglobulins (1-4). Although the mechanism for introducing these sequence changes is currently unknown, several features of SHM specificity offer clues to the DNA transactions that might be involved. For example, SHM primarily occurs in two highly mutable DNA sequence motifs (5-9). One is the RGYW sequence (the underlined G is mutated, R ϭ A or G, Y ϭ T or C, and W ϭ A or T), which is found in SHM substitution spectra in equal proportions in both DNA strands. The other is the WA motif, where substitutions are more likely in one strand than the other (5, 7-10). A clue to the origins of the substitutions in the WA motif is the observation that their type and location correlates with the base substitution error specificity of human DNA polymerase when copying a bacterial gene sequence in a model system in vitro (8). Based on that correlation and additional considerations of the two mutable motifs (11), we suggested that errors at A-T base pairs by pol may contribute to as much as one third of somatic mutations in Ig genes, preferentially during synthesis of the non-transcribed strand. This hypothesis is supported by the observation that XP-V patients lacking active polymerase have a lower proportion of somatic substitutions at A-T base pairs in Ig genes (12). Because pol error specificity does not correlate with substitutions in the RGYW sequence motif, and because those substitutions are distributed equally on both strands, we further suggested that SHM may involve more than one DNA transaction and more than one DNA polymerase (8). This finding is consistent with the two-phase model of SHM proposed earlier (9, 13). Other DNA polymerases suggested to participate in SHM include pol (14-16), pol (17, 18) ...