Characterization of chromosome structure and centromeric activity of Robertsonian translocations

Abstract

Robertsonian translocation is the result of short arm fusion between acrocentric chromosomes. It is the most common structural chromosome aberration in humans. The majority of Robertsonian translocations are dicentric and segregate normally through meiosis and mitosis. Therefore, it is assumed that one centromere is inactivated or suppressed to ensure proper transmission of the translocation chromosome in cell divisions. A single study has specifically investigated the active and inactive centromeres of only two types of Robertsonian translocations, a task which has been complicated by the lack of criteria for either an active or inactive centromere. In this study, several approaches were applied in the identification of the active and inactive centromeres of dicentric Robertsonian translocations. Dual color fluorescence in situ hybridization (FISH) of {dollar}\alpha{dollar}-satellite DNA probes was used to assess centromeric activity based on DNA morphology in 54 dicentric, nonhomologous Robertsonian translocations. By in situ analysis, 90% (49/54) of the translocations demonstrated preferential activity of a particular centromere. The chromosome 14 centromere was most often active in its translocations, while the chromosome 15 centromere was least often active when involved in Robertsonian translocation. While the DNA components of active versus inactive centromeres have not been conclusively defined, essential centromere associated proteins (CENPs) have been localized in and around the primary constriction. Two of these centromeric antigens, CENPs-C and E, are thought to bind active centromeres only; thus application of antibodies to these two antigens is a more accepted and definitive means of differentiating between active and inactive centromeres of Robertsonian translocations. Nine of ten different translocations studied bound CENPs-C and E at one centromere only, confirming the active centromere as assigned by DNA in situ hybridization. The remaining translocation, interpreted as truly dicentric based on two constricted fluorescent hybridization signals in most cells, demonstrated CENP-C and CENP-E at both centromeres. To investigate if chromosome structure is similar among Robertsonian translocations, the residual short arm content of each translocation was studied with DNA probes for the acrocentric short arm region. The most distal repetitive DNA, {dollar}\beta{dollar}-satellite, was not retained in any translocation. However, most translocation chromosomes retained satellite III, a short arm, repetitive DNA located adjacent to the {dollar}\alpha{dollar}-satellite DNA of the primary constriction. All of the t(13q14q)s which were analyzed retained a proximal satellite III subfamily, pTRS-47 only, thus demonstrating grossly identical pericentromeric structure. The t(14q21q)s, however, varied in short arm DNA composition in that no, one or two satellite III DNA subfamilies were present on the translocation chromosomes. This study is the first to demonstrate preferential centromeric activity in almost every type of Robertsonian translocation and provides evidence for a functional centromeric hierarchy. Furthermore, this is the only study to show selective binding of centromere-associated proteins to Robertsonian translocations. The results additionally indicate that translocation structure can very depending on the chromosomes involved and may influence the activity of certain acrocentric centromeres.