Response to the Letter to the Editor by Dunning Hotopp and Klasson

In Leung et al. (2017), we provide evidence that Drosophila transposons have been major contributors to the expansion of the Drosophila ananassae fourth chromosome (Muller F element) arms. Dunning Hotopp and Klasson agree with this finding, but they take exception to our second conclusion, that lateral gene transfer (LGT) from theWolbachia endosymbiont of D. ananassae (wAna) had only a minor role in the expansion of the D. ananassae fourth chromosome arms. This disagreement might have arisen because of a misunderstanding of the scope of our report. As stated in the paper, our analysis is restricted to the “assembled portions” of the D. ananassae F element ( 18.7 Mb, all within the chromosome arms, not including the telomeric and pericentromeric regions). Wedid not generalize ourfindings to thewhole F element.Hence, we agree withDunningHotopp andKlasson that the results reported in Leung et al. (2017) do not preclude the possibility that LGT fromwAna is a contributor to the expansion of the D. ananassae F element in its entirety; specifically, LGT could be contributing to an expansion of the pericentric heterochromatin regions of the F element, as reported in Klasson et al. (2014). Because we only manually improved 1.4 Mb of the D. ananassae F element, Wolbachia sequences that are integrated into the D. ananassae F element could be located within gaps or unassembled portions. However, the presence of large-scale LGT from wAna in the manually improved regions of the D. ananassae F element is very unlikely, as the gap sizes in the improved regions were estimated by multiple restriction digests, or by long reads produced by the Pacific Biosciences (PacBio) sequencer. We did find several scaffolds within the D. ananassae CAF1 assembly that show sequence similarity to the Wolbachia endosymbiont of D. melanogaster (wMel) along the entire length of the scaffold (e.g., scaffold_12940). Hence, if sequencing reads that exhibit sequence similarity to wMel were filtered as part of the sequencing or assembly pipeline, as suggested by Dunning Hotopp and Klasson, then the parameters that were used allow at least some of thewMel reads to be assembled in the CAF1 assembly. We note also that most of the telomeric and pericentric heterochromatin regions are missing from the CAF1 assembly. Thus, our findings do not contradict the work by Klasson and colleagues, who used fluorescence in situ hybridization of Wolbachia to conclude that “[h]ybridization to the fourth chromosome is consistent with the LGT being largely heterochromatic (Figure 9)” (Klasson et al. 2014), describing hybridization to the pericentric region. Both sources of expansion could well have occurred in different regions of the F element, with retrotransposons contributing to the expansion of the chromosome arms and LGT from wAna contributing to an expansion of the heterochromatic regions. We certainly agree with Dunning Hotopp and Klasson that higher quality genome assemblies forD. ananassae and forwAna are needed in order to ascertain the extent to which LGT fromWolbachia contributes to the expansion of the D. ananassae F element overall. For example, assemblies based on long reads produced by PacBio or Nanopore sequencers might enable the clear identification ofWolbachia sequences that are integrated into the D. ananassae genome. Nonetheless, even with long-read technologies, it will remain difficult to generate a highquality assembly for the D. ananassae pericentric heterochromatin region and, as such, to get a reliable estimate of the level of LGT from Wolbachia in these regions. However, within the assembled portions of the chromosome arms, which contain 64 complete D. ananassae F element genes, the available data argue that retrotransposon expansion was the major driver of chromosome expansion. Note added in proof: See Dunning Hotopp and Klasson in this issue for a related work.