Volume 4, Issue 5, Pages 705-714 (November 1999) Rapid Exchange of A:T Base Pairs Is Essential for Recognition of DNA Homology by Human Rad51 Recombination Protein  Ravindra C. Gupta, Ewa Folta-Stogniew, Shawn O'Malley, Masayuki Takahashi, Charles M. Radding  Molecular Cell  Volume 4, Issue 5, Pages 705-714 (November 1999) DOI: 10.1016/S1097-2765(00)80381-0

Figure 1 Preferential Switching of A:T Base Pairs in Homologous Recognition This model is based on the observation that increasing GC content preferentially inhibited the rate of initiation of strand exchange, while the rate of homologous pairing remained unchanged (Gupta, et al. 1999). In 37% GC DNA, the switching of A:T base pairs suffices to mediate homologous recognition by HsRad51, but not to initiate significant strand exchange. The lines denoted on the right by minus and plus signs represent complementary strands of DNA, with A:T and G:C base pairs symbolized as indicated. For simplicity, the bases in the single-stranded substrate and product have not been identified by a symbol. The diagram in the middle represents synaptic intermediate. At the left of the figure, U denotes unswitched base pairs, and S denotes switched base pairs. Molecular Cell 1999 4, 705-714DOI: (10.1016/S1097-2765(00)80381-0)

Figure 2 Effects on Strand Exchange of Substituting Inosine for Guanine The oligonucleotides G13.I24, G37.I13, and G49.I22 were derived by substituting inosine for guanine in G37, G50, and G71 oligonucleotides, respectively. For reactions with oligonucleotides that contained inosine, the minus strand in the filament and the minus strand in the duplex substrate both contained the same inosine substitutions for guanine so that exchanges would be isoenergetic. (A) The yield of products after 45 min, as measured by gel electrophoresis (see Experimental Procedures). (B) Homologous pairing as measured by the fluorometric assay. G37(−) and G13.I24(−) single-stranded oligonucleotides labeled with fluorescein at their 3′ ends were used to form filaments with HsRad51. Both duplex substrates, G37(−)/G37(+) and G13.I24(−)/G37(+) contained rhodamine at the 5′ ends of the G37(+) strand. Quenching of the emission from fluorescein was recorded, after the addition of duplex to preformed HsRad51 filament on ssDNA. (C) Strand exchange as measured by the fluorometric assay: the substrates were the same as in (B), except that the reporter dyes were both on the duplex substrate, fluorescein at the 3′ end of the minus strands and rhodamine at 5′ end of the plus strands. Enhancement in the intensity of fluorescein emission monitors separation of (−) and (+) strands. Molecular Cell 1999 4, 705-714DOI: (10.1016/S1097-2765(00)80381-0)

Figure 3 The Preferential Effect of Mismatches Opposite A:T Base Pairs To make substrates with mismatches opposite either A:T or G:C base pairs, we made A:T or G:C transversions in the duplex substrates. The HsRad51 nucleoprotein filament on G37(−) ssDNA was incubated with duplexes derived from G37(−)/G37(+), but containing 0, 2, 3, 4, or 6 evenly spaced transversions, all of which were either A:T transversions or G:C transversions in any given substrate (Table 3, and Experimental Procedures). The use of substrates derived from G37, which is 37% GC, strongly reduces strand exchange (see Figure 2). Effects of mismatches opposite G:C base pairs are shown in (A) and (C), effects of mismatches opposite A:T base pairs are shown in (B) and (D). (A and B) Assay of reaction products by gel electrophoresis. Percent products refers to the amount of displaced radiolabeled ssDNA from labeled duplex DNA. The number on right side of each curve indicates the number of mismatches. (C and D) Fluorescence assay for homologous pairing. HsRad51 nucleoprotein filament was formed on fluorescein-labeled G37(−) oligonucleotide followed by the addition of rhodamine-labeled duplex oligonucleotides containing 0, 2, 3, or 4 transversions. Decrease in fluorescein emission due to homologous pairing was monitored. The heterologous control, GC6(−)/GC6(+) was a duplex oligonucleotide closely related to G16 and GC10 (see Experimental Procedures). In (D), three tracings overlapped, those for the heterologous control, and for substrates with 3 and 4 mismatches. Molecular Cell 1999 4, 705-714DOI: (10.1016/S1097-2765(00)80381-0)

Figure 4 Effects of Mismatches on Homology-Dependent Exchanges (A) Graphic representation of the preferential effect of mismatches opposite A:T base pairs. The yields of products at 45 min, as presented in Figure 3A and Figure 3B, were plotted versus the number of mismatches. Relative yield of products is calculated with respect to the perfectly homologous substrates. (B) Effects of mismatches opposite A:T base pairs in AT-rich versus GC-rich DNA. A:T transversions were made in G16(−)/G16(+) and G37(−)/G37(+) duplexes at the locations specified in Table 2, creating mismatches relative to HsRad51 filaments formed on G16(−) and G37(−). Reactions promoted by HsRad51 were stopped at 45 min and monitored by gel electrophoresis. The yield of products was normalized with respect to the yield from substrates lacking mismatches. The result shown in this graph is the average of two separate experiments. (C) Effects of mismatches opposite 2-AP:T base pairs: evidence that switching of base pairs is an early event. As a fluorescent reporter, 2-aminopurine was substituted for 3 adenine residues in the minus strand of GC10(−)/GC10(+) duplex DNA. The latter has 10 base pairs at each end that are heterologous to oligonucleotide G16(−), which was used to form the HsRad51 filament. Homologous pairing of these substrates produces paranemic joints that can not complete strand exchange because of their heterologous ends. We monitored at 370 nm the enhancement in fluorescence emission that occurs when 2-aminopurine is displaced from its position in duplex DNA. (Circles) Enhancement of emission from 2-AP in paranemic joints; (×) Lack of enhancement of emission from 2-AP when mismatches in the single strand were placed directly opposite the residues of 2-AP; (inverted triangles) control, enhancement of emission when mismatches were placed 4–9 nucleotide residues away from each residue of 2-AP. Molecular Cell 1999 4, 705-714DOI: (10.1016/S1097-2765(00)80381-0)

Figure 5 Coincident Kinetics of Homologous Pairing and Strand Exchange in AT-Rich DNA Pairing and strand exchange were monitored by the fluorescence assays for substrates containing (A) 63% AT, (B) 74% AT and (C) 84% AT. Tracings of fluorescence emission were converted into concentrations of the intermediate produced by homologous pairing or the product of strand exchange, and theoretical curves were fit to the data, both as described previously (Bazemore et al. 1997b). (Circles) homologous pairing; (×) strand exchange. Molecular Cell 1999 4, 705-714DOI: (10.1016/S1097-2765(00)80381-0)