ANALYSIS OF CLONALITY USING X-LINKED POLYMORPHISMS

Abstract

Analysis of clonality can be used to determine whether proliferative disorders are clonal in origin and may also be applied to assess disease progression. A clonal population of cells is defined as the cells arising from mitotic division of a single somatic cell. A variety of different methods for clonality analysis may be applied e.g. characteristic karyotypes, gene rearrangements, deletions or point mutations. However, in many disorders, specific genetic abnormalities have not been defined and it is necessary to use an indirect approach based on X-chromosome inactivation. Clonality analysis using X-Iined polymorphic loci are based on the hypothesis that only one allele in the X-chromosomes is inactivated by methylation of cytosine residues and that polymorphisms existing in the general female population at different loci on the X-chromosome allows the distinction of the paternal from the maternal X-chromosome. The informativeness of any X-linked polymorphic site is therefore related to the frequency of polymorphism at that locus in the female population. Because the X-inactivation process that occurs during embryogenesis is random, some cells from a normal female may carry an active maternal X-chromosome and others an active paternal X-chromosome. Changes in the methylation pattern of the X-chromosome can be distinguished by cleavage with the methylation sensitive restriction endonucleases, Hpall or Hhal, which cleave only the unmethylated (active) DNA. Differential methylation patterns between the active and inactive X-chromosome have been documented at several loci. PCR-based clonality assays have been developed for the analysis of the phosphoglycerate kinase (PGK), human androgen receptor (HUMARA), monoamine oxidase A (MAOA) and fragile X (FMR1) genes. Clonality analysis using the HUMARA locus is particularly advantageous because it has a high incidence of heterozygosity (> 85 %) in the general female population and because the methylation of cytosines in the sequences recognized by Hpall and Hhal are correlated with X-inactivation. We have devised nested primers to enhance the sensitivity of the HUMARA clonality assay. We then established, by clonality analysis of peripheral blood samples from healthy Chinese females, that the HUMARA locus is more informative (91%) than the PGK locus for clonality assays in this population. Subsequently, we carried out HUMARA clonality analysis in isolated eosinophils in the idiopathic hypereosinophilic syndrome (IHES), in formalin-fixed, paraffin-embedded tissues in Erdheim-Chester disease (ECD) and Langerhans cell histiocytosis (LCH), and in isolated Hodgkin and Reed-Sternberg (H/RS) cells from the nodular sclerosing subtype of Hodgkin's disease (NSHD). Eosinophils purified by flow cytometric sorting from a female patient presenting with IHES were found to show a clonal pattern of X-inactivation by HUMARA gene analysis. In contrast, eosinophils purified from a patient with Churg-Strauss syndrome and from six patients with reactive eosinophilias attributed to allergy, parasitic infection or drug reaction, showed a polyclonal pattern of X-inactivation. In the analysis of paraffin-embedded tissue blocks from three female patients with LCH and a female patient with ECD, the ECD tissues were found to be polyclonal in origin, whereas LCH tissues were clonal. This provides evidence that the two disorders are distinct disease entities that arise from the macrophage/dendritic cell lineage. In studies of X-chromosome inactivation patterns of isolated H/RS cells from six cases of NSHD, the same high- or low-molecular weight allele in about 83% (range 73-91%) of single H/RS cells was found in tissues from all six cases. In contrast, HUMARA analysis of whole NSHD tissues from the six cases showed balanced allelic patterns. The results provide evidence that a clonal population of H/RS cells was likely to be present in each case. Our results also indicate that, although the HUMARA loci are useful for clonality analysis, results must always be interpreted with caution since several factors can potentially bias PCR amplification.