Hybrid Element Insertion

It has previously been shown that the combination of two deleted Pelements in trans, one containing the left functional end and the second element the right functional end, can lead to high levels of male recombination. This finding strongly suggests that P-element, ends from different chromosomes can become associated, followed by “pseudo-excision.” We show that two different processes are involved in resolving the pseudo-excision event: (1) the excised P-element ends continue to function as a single unit (Hybrid Element) and insert at a nearby site in the chromosome or into the element i self [Hybrid Element Insertion (HEI)] and (2) free ends that do not contain Pelements repair and rejoin [(Hybrid Excision and Repair (HER)]. Both types of resolution can lead to recombination, and this paper concentrates on the HE1 class. One type of HE1 event predicts the exact reverse complementary duplication of an 8-bp target site, and we have confirmed the existence of such a structure in six independently derived recombinant chromosomes. There is also a high tendency for insertion events to occur within a few bases of the original 8-bp target site, including six apparent cases of insertion into the exact site. H IRAIZUMI (1971) showed that male recombination in Drosophila melanogaster, which had previously been found only at very low levels, can occur at frequencies of up to 1% in crosses involving recently collected wild-type strains. This recombination has been shown to be associated with Pelements (BINGHAM et al. 1982; MCCARRON et al. 1989; SVED et al. 1990). Using a system in which a single P(CaSpeR] element (PIRROTTA 1988) is mobilized by a transposase source, P(A2-3](99B) (ROBERTSON et al. 1988), SVED et al. (1990) showed that levels of -0.5-1.0% recombination can be produced. SVED et al. (1991) found that the rate of recombination rises by an order of magnitude, to -2096, if two qCaSpeR} elements are present at homologous sites on the chromosome. SVOBODA et al. (1995) then showed that these two elements could function to produce even higher rates of recombination, 30% or more, if one had only a functional left end and the other only a functional right end. The high rate of recombination produced by enddeleted elements can be understood in terms of the “cut-and-paste” model of ENGELS et al. (1990). Under this model, it is postulated that the normal method of P-element propagation consists of excision of an element at the four-strand stage of division, followed by repair using the sister P element and insertion of the excised element at a new site on the chromosome. AlCorresponding author: John Sved, School of Biological Sciences A12, University of Sydney, NSW 2006, Australia. E-mail: [email protected] versity, Stanford, CA 94305. ’ Present address: Department of Biological Sciences, Stanford UniGenetics 144: 1601-1610 (December, 1996) though the details of how P elements excise are not clear, it seems likely that this is achieved by an association of left and right ends. The result of SVOBODA et al. (1995) strongly suggests that this association need not involve a left and right end from the same element. Association of opposite P-element ends at the four strand stage, followed by “excision,” can be pictured as in Figure 1 (I.$ SVOBODA et al. 1995, Figure 11). The diagram shows the two DNA strands of both sister chromatids of both chromosomes. ’ There are several possibilities for the resolution of the breaks shown in panel 3 of Figure 1. The first is that the two ends with P elements (el and +) remain associated as a “hybrid element.” This element could then insert elsewhere in the genome. This event is labeled as Hybrid Element Insertion (HEI) , and evidence is presented in this paper for its regular occurrence. The second possibility is that there is template repair followed by rejoining of two of the four ends [Hybrid Excision and Repair (HER)]. In cases where the P element-containing ends retain their association, the only ends available for such repair are nl and Q, and evidence is presented that these regularly rejoin. The possibility of rejoining events involving ends other than nl and Q is also discussed. MATERIALS AND METHODS Stocks and procedures used in this paper are as outlined in SVOBODA et al. (1995). Sequencing was carried out using the Circumvent Cycle Sequencing Kit (New England Biolabs). All results in the paper are from crosses of males that contain deleted elements derived from an initial insertion of flCaSpeR] in the 50C region of chromosome 2. Test males 1602 Y. H. M. Gray, M. M. Tanaka and J. A. Sved L +( > + . A C cn ~ l l l l l l l l l l l l l I ~ >