(a) A schematic overview of the production and amplification of a single piRNA from a transposon transcript. The transposon mRNA is processed by the primary piRNA biogenesis machinery to generate primary sense piRNAs that have a 1U bias. Indicated in yellow is a representative piRNA that will be processed from this transcript. Once bound to MILI, the piRNA guides endonucleolytic cleavage of complementary transcripts between 10th and 11th nucleotide counting from a 5′ end of bound piRNA. The 3′ cleavage product will be processed to form a secondary antisense piRNA, the location of which is indicated in blue. The nascent secondary piRNA is loaded into MILI or MIWI2. This licences MIWI2 to enter the nucleus, where it is proposed to tether to a nascent transposon transcript and direct de novo DNA methylation by an unknown mechanism. The binding of the secondary antisense piRNA to MILI results in the piRNA-guided cleavage of transposon transcript and the biogenesis of the original piRNA. This MILI:MILI homotypic ping-pong cleavage cycle amplifies transposon-derived piRNAs during fetal reprogramming. (b) Features of amplified piRNA pairs. Due to the fact that most primary piRNAs exhibit 1U bias, newly generated secondary piRNAs will have a bias for A at their 10th position (10A). Complementarity between the first 10 nucleotides of amplified piRNA pairs is a key signature of the ping-pong cycle.