Homie blocks spreading of the <i>eve</i> Pc/H3K27me3 domain into the adjacent gene, <i>TER94</i>.

<p>ChIP assays were used to quantify the association of H3K27me3 with the endogenous <i>eve</i> and <i>TER94</i> coding regions, and with the transgenic <i>eve-lacZ</i> and <i>TER94-GFP</i> coding regions. <b>A, B:</b> ovaries were dissected and subjected to ChIP analysis, as described in <a href="http://www.plosgenetics.org/article/info:doi/10.1371/journal.pgen.1003883#s4" target="_blank">Materials and Methods</a>. In A is graphed the averages and standard deviations from triplicate PCR assays using primer pairs specific to each of the 4 coding region indicated along the bottom, and for the transgenic lines indicated (“intact t'gene” carries the entire transgene diagrammed in <a href="http://www.plosgenetics.org/article/info:doi/10.1371/journal.pgen.1003883#pgen-1003883-g001" target="_blank">Figure 1A</a>; “ΔHomie” is the same transgene with Homie deleted; “ΔHomie ΔPRE” has both Homie and the PRE deleted). Either H3K27me3-specific antibodies or non-specific IgG were used to precipitate cross-linked chromatin, as indicated in the inset key. In B is graphed the same data (“trial 2” in the inset key), and two other trials (“trial 1” and “trial 3”) using independent chromatin preparations (see <a href="http://www.plosgenetics.org/article/info:doi/10.1371/journal.pgen.1003883#s4" target="_blank">Materials and Methods</a>), normalized to the H3K27me3-specific signal from the <i>eve</i> coding region for each transgenic line. <b>C, D:</b> embryos (2–20 hr. after egg deposition) were collected and subjected to ChIP analysis, as in A and B above. Note that the specific signals from the transgenic <i>eve-lacZ</i> coding region do not change significantly when Homie is deleted, while the transgenic <i>TER94-GFP</i> coding region shows an increase in H3K27 trimethylation when Homie is deleted, correlating with its repression.