I-R hybrid dysgenesis correlates with maternal piRNA inheritance. (A) Normalized piRNA counts for Repbase transposons are plotted for w¹¹¹⁸ inducer and w^K reactive ovaries. (B) (Left) Fold differences in piRNA counts comparing w¹¹¹⁸ and w^K mothers are shown (red line indicates a 1:1 ratio). (Right) Transposon piRNA ratios for mothers, embryos, and F1 progeny (SF: RSF ratio) are shown as a heat map. (C) Scatter plots indicating transposon piRNA correlations between w¹¹¹⁸ and w^K mothers (top) and their respective intercross progeny (bottom).

Suppression of P-M hybrid dysgenesis correlates with maternally deposited piRNAs. (A) Normalized number of ovarian (black) and early embryonic (orange) piRNAs from the indicated strains mapping to the F-, I-, and P-elements with up to three (F and I) or with one (P) mismatch(es). To the right (sense, upper; antisense, lower), densities of piRNAs are displayed over the P-element in parental ovaries of the Har, LK, and NA strains. The extent of the truncation of the telomeric P-insertion in the NA strain is indicated. The intron-exon structure of the P-element is shown below. (B) F1 ovarian morphology, ping-pong signature, and piRNA densities from nondysgenic (LK × Har and NA × Har) and dysgenic (w¹¹¹⁸ × Har) F1 progeny are shown.

The I-R hybrid dysgenesis system. (A) Crossing scheme to generate nondysgenic (left) and dysgenic (right) progeny. Small RNA libraries were made from flies indicated by asterisks. Bar diagrams indicate fertility analysis of parental and F1 females on the basis of egg hatching rates. n, number of counted eggs. (B) Immunofluorescence for I-element ORF1 (green) is shown for nondysgenic (RSF) and dysgenic (SF) daughters. Oo, oocyte; NC, nurse cells.

A piRNA amplification loop between active I-elements and ancestral fragments. (A) Density profile of piRNAs matching the I-element with up to five mismatches (left) and profiles for those species with zero (top right) or at least one mismatch (below) to the active element (sense/antisense fractions in red and green) are shown. Profiles below indicate species that have the potential (upper) or must (lower) derive from the 42AB piRNA cluster. I-element fragments contained within the 42AB cluster are indicated in red. (B) Shown are fractions of I-element piRNAs from w¹¹¹⁸, w^K mothers, and SF daughters (right) that match the active sequence with the indicated number of mismatches, split into sense (+) and antisense (−). (C) (Right) All I-element fragments in the Release 5 genome sequence [split into modern insertions (blue) and ancestral fragments (black) and sorted by chromosomal position] are shown (gray coloring indicates heterochromatic; HC, heterochromatin; p. HC, pericentromeric HC). Red fragments map to the 42AB piRNA cluster. (Left) Bar diagrams indicate the density of piRNAs mapping uniquely to the ancestral fragments shown at right (42AB fragments in red). As fragments with high piRNA density were not more divergent overall, clustering in 42AB is not an artifact of analysis (fig. S7).