Pseudoautosomal linkage of familial hodgkin's lymphoma: molecular analysis of a unique family with leri–weill dyschondrosteosis and hodgkins lymphoma

Hodgkin’s lymphoma (HL) is generally a sporadic disease that may be caused by an infectious agent, possibly Epstein‐ Barr virus (EBV). However, there is consistent evidence of familial clustering in HL (Ferraris et al, 1997), suggesting a role for heritable predisposition in a proportion of patients. Although progress with the identification of an HL-predisposing gene has been hindered by the rarity of multiplepatient HL families, our description of a family in which two sisters with the skeletal dysplasia Leri‐Weill dyschondrosteosis (LWD) both developed HL in adolescence led us to suggest that an HL-predisposing gene might be linked to the LWD gene (Gokhale et al, 1995). Subsequent identification of the LWD gene as the pseudoautosomal short stature homeobox gene SHOX (Shears et al, 1998) led to the hypothesis, based on observations in our HL-LWD family, that an HL-predisposing gene might be located in the major pseudoautosomal region (PAR1) of the X and Y chromosomes (Horowitz & Wiernik, 1999). We have studied further the PAR1 in the HL-LWD family using microsatellite marker analysis and fluorescence in-situ hybridization (FISH). Fluorescent polymerase chain reaction (PCR) primers were used to amplify the pseudoautosomal microsatellite markers DXYS233, DXS6814, DXYS228, DXYS230 and the SHOX CA repeat [located approximately 15 kilobases (kb) upstream of SHOX]. PCR products were analysed using an ABI Prism TM 377 Genetic Analyser and allele sizes calculated using GENOTYPER TM software (Applied Biosystems, Warrington, UK). The microsatellite marker DXYS232 and the 4 basepair (bp) insertion ⁄ deletion polymorphism at the MIC2 locus were amplified using nonfluorescent primers, and PCR products resolved using polyacrylamide gel electrophoresis and silver staining. FISH was performed on metaphase spreads of lymphocytes using a panel of cosmid probes from the Xp ⁄ Yp telomere spanning PAR1. One of these, LLNOYCO3¢M¢34F5, encompasses the SHOX gene. Cosmids were labelled using the Bio-Nick labelling system (Life Technologies). The results of the microsatellite and FISH analysis are summarized in Fig 1A and B. The individuals tested included the probands P1 and P2, both affected with LWD and Hodgkin’s lymphoma, their mother (M), who only had LWD, their unaffected father (F) and their maternal grandmother (GM). Our results confirmed that the siblings with HL-LWD each harboured a maternally inherited microdeletion within PAR1 encompassing SHOX, estimated to be 900 kb, and located between 200 kb and 1100 kb from the Xp telomere. Three genes of potential importance in HL are located within the pseudoautosomal region, including the interleukin 3 receptor a chain (IL3RA), the granulocyte‐ macrophage colony stimulating factor receptor a chain (CSF2RA) and MIC2, which encodes the CD99 cell adhesion protein. These genes were not, however, within the deleted region in the two siblings (Fig 1C). While it is not yet possible to exclude haploinsufficiency due to the deletion of some unknown tumour suppressor gene, an intriguing possibility is that the PAR1 deletion caused a long-range position effect by downregulating expression of one of the above genes. MIC2 is a particularly strong candidate as decreased expression is associated with the development of Hodgkin’s Reed‐Sternberg cells (H-RS cells; Kim et al, 1998) and is downregulated by EBV latent membrane protein 1, which is highly expressed in H-RS cells (Lee et al, 2001). Against this hypothesis is the fact that PAR1 microdeletions of a similar size are the most common defect in LWD, but there is, as yet, no report of an association between LWD and HL. The existence of an HL-predisposing gene in PAR1 (Horwitz & Wiernik, 1999) thus remains speculative, but further studies of genes in this region in familial Hodgkin’s lymphoma should help to resolve this issue.