The two-domain hypothesis in Beckwith-Wiedemann syndrome

I was pleased to find that Maher and Reik, in their recent Perspective (1), agreed with our two-domain hypothesis, which we first proposed last year (2), for Beckwith-Wiedemann Syndrome (BWS), which causes prenatal overgrowth, midline birth defects, and cancer. According to this hypothesis, two separate domains of imprinted genes, on chromosomal band 11p15, are involved in BWS. (Imprinting involves epigenetic silencing of a specific parental allele, and loss of imprinting is a common cause of abnormal gene expression in cancer [reviewed in ref. 3].) One of these two domains, involving a novel antisense transcript within K_VLQT1 and termed LIT1, was the subject of our report. We found that approximately half of BWS patients undergo loss of imprinting (LOI) of LIT1, compared with 20% of BWS patients showing LOI of IGF2, which is located approximately 500 kb telomeric to LIT1. Furthermore, LOI of LIT1 was independent of LOI of IGF2 when examined in the same patients (2). Moreover, we and our collaborators have observed no evidence of LOI of LIT1 in Wilms tumors (4), despite frequent LOI of IGF2 in embryonal tumors (5). Finally, our two-domain model was consistent with other observations from our laboratory that a group of genes lying between these two domains, including TSSC4 and TSSC6, are imprinted minimally or not at all (6).

In Maher and Reik’s excellent review, they proposed a model of enhancer competition between LIT1 and p57^KIP2, similar to that known to exist between H19 and IGF2 in the more telomeric domain (1). We had also suggested in our article last year that p57^KIP2 might compete for a shared enhancer. However, I would like to point out an alternative to Maher and Reik’s and our own earlier speculation about the possible location of such an enhancer, which was hypothesized to lie on the telomeric side of a 450-kb cluster of germline balanced chromosomal rearrangement breakpoints in BWS patients, termed BWSCR1.

Alternatively, this enhancer might lie on the centromeric side of BWSCR1 and LIT1. In the absence of data, we cannot favor one possibility over the other at this point. However, if the alternative location were correct, one would have an elegant genomic structure of the two imprinted domains as mirror images of each other, with BWSCR1 in between them (Figure 1). Both the known H19/IGF2 enhancer and the hypothetical p57^KIP2/LIT1 enhancer would lie near one end of the entire 11p15 imprinted domain, and if one walked toward the center of the domain from either end, one would encounter, in this order: (a) the hypothetical enhancer for p57^KIP2/LIT, and the known enhancer for H19; (b) the p57^KIP2 gene, which is maternally expressed, and the H19 gene, which is maternally expressed; (c) an as yet unidentified insulator centromeric to LIT1, and the known insulator centromeric to H19; (d) an approximately 100-kb interval; and (e) the paternally expressed LIT1 gene, and the paternally expressed IGF2 gene. The disadvantage of this alternative hypothesis is that it would leave unexplained the mechanism of BWS in the BWSCR1 rearrangement patients, although we should not overlook the possibility of a more centromeric enhancer.

Figure 1

A model of a shared enhancer for p57^KIP2 and LIT1 centromeric to BWSCR1 and LIT1. If the enhancer were centromeric to LIT1, then it would activate p57^KIP2 when an insulator between p57^KIP2 and LIT1 is unmethylated, and it would not be affected by the BWSCR1 chromosomal rearrangements. The same enhancer would activate LIT1 when the insulator was methylated. Note that the precise location of the enhancer is immaterial as long as it is centromeric to LIT1. E, enhancer (green); I, insulator (red); maternally expressed p57^KIP2 and H19 (pink); paternally expressed LIT1 and IGF2 (blue); cen, toward the centromere; tel, toward the telomere. Arrows indicate alternate use of the enhancer.

Regardless of whether this alternative for the location of this theoretical enhancer, or a more telomeric location suggested earlier by us (2) and by Maher and Reik (1), is correct, the two-domain hypothesis that we first proposed (which is compatible with both) might also help to explain the apparent differences in the phenotype of BWS patients with altered imprinting of the two domains (M. deBaun et al., unpublished observations) or with p57^KIP2 mutations (7).

Note added in proof.

I appreciate Maher and Reik's description of the evolution of their independent thinking that led to the two-domain model in their JCI review. I emphasize my point that either location for a shared enhancer is possible with current data, but I did not suggest that KvDMR1 is itself the insulator. Indeed, I think that unlikely, as we find that the sequence is not conserved in the mouse.

References

1. Maher, ER, Reik, W. Beckwith-Wiedemann syndrome: imprinting in clusters revisited. J Clin Invest 2000. 105:247-252.
   View this article via: JCI PubMed CrossRef Google Scholar
2. Lee, MP, et al. Loss of imprinting of a paternally expressed transcript, with antisense orientation to KVLQT1, occurs frequently in Beckwith-Wiedemann syndrome and is independent of insulin-like growth factor II imprinting. Proc Natl Acad Sci USA 1999. 96:5203-5208.
   View this article via: PubMed CrossRef Google Scholar
3. Feinberg, A.P. 2000. Genomic imprinting and cancer. In The metabolic and molecular bases of inherited disease. 8th edition. C. Scriver et al., editors. McGraw-Hill. New York, New York, USA.
   View this article via: PubMed Google Scholar
4. Mitsuya, K, et al. LIT1, an imprinted antisense RNA in the human KvLQT1 locus identified by screening for differentially expressed transcripts using monochromosomal hybrids. Hum Mol Genet 1999. 8:1209-1217.
   View this article via: PubMed CrossRef Google Scholar
5. Rainier, S, et al. Relaxation of imprinted genes in human cancer. Nature 1993. 362:747-749.
   View this article via: PubMed CrossRef Google Scholar
6. Lee, MP, et al. Somatic mutation of TSSC5, a novel imprinted gene from human chromosome 11p15.5. Cancer Res 1998. 58:4155-4159.
   View this article via: PubMed Google Scholar
7. Lam, WW, et al. Analysis of germline CDKN1C (p57KIP2) mutations in familial and sporadic Beckwith-Wiedemann syndrome (BWS) provides a novel genotype-phenotype correlation. J Med Genet 1999. 36:518-523.
   View this article via: PubMed Google Scholar