“…However, hRad51 hydrolyzes ATP and promotes strand exchange at least 10-fold less e ciently than RecA, so that the length of heteroduplex DNA formed is limited to 1.3 kb (Baumann and West, 1999). Rad51 fails to generate stable joint molecules in the absence of ATP, although it still binds cooperatively and with high a nity to DNA and can renature complementary DNA strands (Baumann et al, 1996;Gupta et al, 1997;Kelley De Zutter and Knight, 1999). Opposite to the polarity of the bacterial recombinase, hRad51 seems to preferentially start transfer at the 5' end of the linear duplex (Sung and Robberson, 1995;Baumann et al, 1996;West, 1997, 1999).…”