Unit on Genetics and Endocrinology (UGEN) Developmental Endocrinology Branch National Institute of Child Health and Human Development National Institutes of Health Bethesda, Maryland 20892-1862
To the editor:
We read with great interest the article by Gadelha et al. (1) in the February issue of JCEM. The authors had previously described two families with apparent isolated familial somatotropinomas (IFS) and reported loss of heterozygosity (LOH) of their tumor tissue at 11q13 (2). They then examined their families for possible linkage to the multiple endocrine neoplasia type 1 (MEN 1) locus at 11q13 and the Carney complex (CNC) loci at 2p16 and 17q22-24. To perform this analysis, Gadelha et al. (1) used a program by Rohde et al. (3), which adjusts logarithm of odds (LOD) scores for LOH data. The modified maximum LOD score determined in this way was 4.0 for 11q13 (family B). Although it is likely that another gene with tumor suppression function in endocrine tissues is located on 11q13 (4), several questions are raised regarding the actual linkage of the families reported by Gadelha et al. (1) to 11q13 vs. 2p16.
We have recalculated the LOD scores for 11q and 2p using the germ-line
alleles reported by Gadelha et al. (1); this
analysis yielded two-point LOD scores that are strongly positive for
both loci, but not conclusive (Table 1).
We did not include LOH data in this analysis because we believe it is a
finding common in CNC patients (see below). Furthermore, the increase
in power expected from the use of LOH in combination with germ-line
allelotyping for LOD score analysis, "depends on the certainty with
which we can infer the phase between the cancer-predisposing allele and
the marker allele retained in the tumor" (5). In the
case of family A, Gadelha et al. (1) seem to
have assigned the "disease-causing allele" for chromosome 2 to the
unaffected father, with no apparent explanation. For chromosome 11, it
was the maternal allele that was lost in the tumors, suggesting that,
for this chromosome, the father was an unaffected carrier. However, the
authors used their knowledge of genotyping information for 11 on their
linkage analysis for chromosome 2, by assuming that in both cases it
was the father who transmitted the disease-causing allele. Even if we
accept the inclusion of LOH data for 11q13 calculations, the above
assumption cannot be made for the 2p16 analysis. If the father in
family A is considered as "unknown disease status" (UDS), which
should be the case in the absence of any other information, the use of
this kindred does not exclude 2p16, and the LOD scores become positive
for the CNC locus (data not shown).
|
Is there another somatotropinoma-causing tumor suppressor gene on 11q13? We have recently shown that LOH at 11q13 is present in CNC tumors, including a somatotropinoma (6). Other investigators, too, have shown that 11q13 LOH may occur in other tumors (4) and that most GH-producing adenomas that demonstrate LOH at 11q13 do not harbor somatic (tumor) mutations of menin (7).
In summary, although we cannot exclude entirely digenic inheritance or
linkage to 11q13 (1), the two-point LOD scores and the
presence of a possible recombination for 11q13 in family B lead to
either inconclusive linkage or support that at least one of the
families maps with greater likelihood to the CNC locus on
2p16 (Table 1). LOH at 11q13 is likely to be a tertiary hit at the
tumor tissue level. In agreement with Gadelha et al.
(1), there is much evidence to suggest that a tumor
suppressor gene other than MEN 1 is located in 11q13, but we
do not think that germ-line (vs. somatic) mutations of that
gene are present in their families with IFS.
Received February 24, 2000.
References