Figure 4.

Secondary structure within the loop affects repair. Repair of large loops was tested in both an in vivo transformation assay (A) and in a biochemical assay using nuclear extracts (B). The transformation assay is performed as described, using the V₃₈, V₃₈ CTG and V₃₈ hairpin heteroduplexes diagrammed schematically in Figure 1B. The ratio of repair is determined by normalizing total repair of the structured loops to that of the unstructured V₃₈ loop. Total repair for V₃₈ was 92%. A schematic representation of the biochemical assay is found in Figure 5A. The in vitro heteroduplexes used in this structural study are diagrammed schematically in Figure 5B. The ratio of repair, as described above, is normalized to the V₃₀ substrate (an absolute value of 185 fmol/mg/h) in this biochemical assay. The error bars for both the in vivo and in vitro assays represent one standard deviation from three independent measurements.