Secondary and tertiary derivatives of a P-element insertion allele at the vestigial (vg) locus were induced by hybrid dysgenesis. The derivatives were characterized by Southern analyses and, in four cases, by DNA sequencing. The alterations found were P-element internal deletions, deletions of the insert and/or adjacent vg region DNA, or novel insertions of P-element sequences into existing P-element inserts. The relatively high frequency of secondary insertions into P-element sequences observed herein is unusual, since secondary insertions have seldom been recovered in other dysgenic screens. The effects of the alleles on vg expression were determined. The results are consistent with a model in which the insertions disrupt vg gene expression by transcriptional interference.Hybrid dysgenesis mediated by P elements is a powerful tool for studying gene expression in Drosophila melanogaster. It is induced in the progeny of crosses between P-cytotype males and M-cytotype females, but not in the reciprocal cross. P strains contain multiple copies of chromosomal P elements, while M strains lack functional P elements (27). Dysgenic crosses cause mutations due to P-element insertions into genes or imprecise excisions from genes with preexisting P-element sequences. The functional P element is a 2.9-kilobase (kb) transposon which encodes a transposase that is required for transposition (17,26). Most P elements in a P strain are smaller than the 2.9-kb element and are derived from this element by an internal deletion (24). P elements have 31-base-pair (bp) terminal repeats on each flank which are required for transposition (14), and 8 bp of chromosomal DNA is duplicated at the insertion site (24). The internally deleted P elements are unable to produce their own transposase, but they are mobilized when supplied with transposase produced by complete P elements (28, 32). Transpositional activity occurs only under dysgenic conditions, since P strains also encode a repressor which prevents transposition in P-cytotype flies. It is unknown what the structure of the repressor is or exactly how it works, but it appears to be encoded by P-element sequences (for a review, see reference 6).P-element mutagenesis can be used to produce primary or secondary mutants at a locus. Primary mutants are usually insertions of P-element sequences into a gene and are particularly useful in facilitating the initial cloning of genes. A variety of loci have been cloned in this manner (1, 18, 31). Secondary mutants arise when preexisting P-element alleles are induced to undergo further dysgenic activity (at rates as high as 10-2 to 10-3) (24, 34). The secondary mutants are often revertants, but they may also be more extreme derivatives of the original allele. They have been extensively studied at the rudimentary (34), RP11215 (30, 35), and yellow loci (3). Secondary mutants have been shown to be due to either precise or imprecise P-element excisions (3,24,30,34,35), internal deletions within the resident P element (4, 30, 34), deletion of DNA adjace...