Nonetheless, we wished to observe clear CTCF nucleolar localization, and so used a stage of the cell cycle when binding to nucleolar DNA is cytologically distinct. We therefore detected CTCF in the secondary constrictions on neuroblast sex chromosomes. rDNA localization is the only heterochromatic binding that we could detect, however the strong signal from the euchromatic compartments of the genome limited our ability to detect CTCF in distal heterochromatic blocks that juxtapose euchromatin. This localization at a time when nucleoli are disassembled and transcriptionally silent suggests localization is not due solely to protein-protein or protein-rRNA interactions in a mature nucleolus, but instead is due to direct DNA binding by CTCF. CTCF is a sequence-specific Zinc-Finger DNA binding protein, and we thought it was likely to bind the rDNA directly. To confirm this, we predicted potential CTCF binding sites in the entire rDNA sequence including the non-transcribed spacer that WY 14643 purchase separates the 35S primary transcription units, and the 28Sinterrupting R1 and R2 arthropod transposable elements, using the Patser algorithm informed by two different published Drosophila CTCF consensus sequences. We identified six potential sites, and manually scanned the R1 and R2 sequences for similar potential consensus sites that differed in only one nucleotide of the core conserved consensus, which identified an additional 15 sites which served as an expected ����negative���� out-group. In addition to potential binding sites identified by Patser, we designed primers to amplify sequences approximately every 350 base pairs to test potential non-consensus binding across the entire 35S rDNA transcription unit. We did not identify any sequences in the NTS that were similar to the CTCF consensus, although sites have been reported in the human rDNA NTS, so we designed primers for chromatin immunoprecipitation of the NTS regardless of lack of obvious consensus. Of the 46 tested sites, only one Nutlin-3 showed robust binding using chromatin immunoprecipitation with antibodies raised against CTCF. This site was in the DNA that corresponds to the R1 element, near the beginning of this transposable element. Although one other site showed moderate but statistically significant immunoprecipitation, its close linkage to site 29 makes ancillary immunoprecipitation from incomplete DNA shearing a likely explanation. Even for site 28, the relative enrichment by chromatin immunoprecipitation was modest relative to the positive control of the Fab-8 element, but the biology of the rDNA locus makes chromatin immunoprecipitation potentially insensitive to occupancy at this locus, since we must consider that of the hundreds of copies of the cistron, only a fraction are silent and thus possess corresponding histone modifications or regulatory protein binding.