Given the interaction of human ATRX with heterochromatin protein 1, we asked whether the Drosophila protein could also interact biochemically with dHP-1 using a GST-pulldown assay with dHP-1a. Bacterially expressed Succinylsulfathiazole dHP-1a protein bound strongly to full length in vitro translated dATRX. Analysis of deletion constructs mapped the interaction to amino acids 233�C332. This region contains a CxVxL motif that is divergent from a consensus PxVxL motif, but which has been shown to mediate interactions between SP100 and the chromoshadow domain of HP-1. Disruption of this motif in dATRX by Costunlide mutation of the central valine to a glutamic acid abolished binding to dHP-1a in vitro. We then sought to analyse interaction of dATRX with dHP1a in vivo. Co-transfection of tagged dATRX and dHP-1a constructs into S2 cells showed a strong interaction between the two proteins when immunoprecipitated with anti-FLAG agarose. As expected, this was specific for the long isoform of dATRX, since dHP-1a-FLAG specifically enriched for the long isoform of dATRX when immunoprecipitated from cell extracts. Mutation of the CxVxL motif in the long isoform of dATRX also abolished the interaction with dHP-1a. These results are consistent with the ability of the dATRX3 allele to strongly suppress PEV, since it removes specifically the long isoform of the protein, which contains the dHP-1a interaction domain. Since dATRX is involved in formation or maintenance of pericentric heterochromatin, and interacts with dHP-1a, we expressed tagged constructs of dATRX and dHP-1a in 3T3 cells to analyse colocalisation of these proteins at heterochromatin during interphase. Expression of the long isoform of dATRX showed localisation to DAPI dense regions of the nucleus, and upon expression with dHP-1a, showed colocalisation with this protein. As expected, the short isoform did not localise to the heterochromatin or colocalise with dHP-1a. Indeed, this protein appeared to be excluded from the heterochromatin, and its localisation was non-overlapping with the long isoform. Upon mutation of the CxVxL motif, the long isoform of dATRX failed to localise to heterochromatin and showed a similar staining pattern to the short isoform, showing that the interaction with dHP-1a, or its mouse homologue, is necessary for localisation of the long isoform of dATRX to heterochromatin.