Statistics from Altmetric.com
Editor—Germline mutations in the dual specificity phosphatase genePTEN (also known as MMAC1 orTEP1) have been associated with susceptibility to two related hamartomatous disorders, Cowden syndrome (CS, MIM 158350) and Bannayan-Riley-Ruvalcaba syndrome (BRR, MIM 153480).1 2It has recently been established that PTEN functions as a 3-phosphatase towards phospholipid substrates in the phosphatidylinositol 3-kinase (PI-3 kinase) pathway.3 Lack of PTEN results in the accumulation of phosphatidylinositol-(3,4,5)-P3, which is required for activation of protein kinase B (PKB)/Akt, a downstream target of PI3-kinase and a known cell survival factor.4-8
While up to 81% of CS and approximately 60% of BRR cases have detectable PTEN germline mutations, no mutations in the coding region or exon-intron boundaries of PTEN have been found in the remaining affected subjects.2 9-13Informative PTEN mutation negative families have been shown to be linked to the 10q23 region, where PTENlies,2 14 although recently there has been a report of two CS families in which linkage to 10q23 has been excluded.9 This has raised the possibility that either a regulatory region of the PTEN gene not included in previous studies, such as the promoter region, or another, closely located gene might be responsible for the CS and BRR cases in which noPTEN mutation has been found. The first alternative is unlikely to represent the majority of such cases, as no evidence ofPTEN transcriptional silencing has been detected in the tissue of affected CS subjects in which no PTEN mutation was identified (Dahia and Eng, unpublished observations). Transcription levels of PTEN were found to be similar in affected and unaffected tissues of at least three unrelated CS patients and were equivalent to those of normal subjects. This suggests that methylation of the promoter or mutation within the promoter affecting transcription ofPTEN does not occur in at least a subset of thesePTEN mutation negative CS and BRR cases. To investigate the possibility that a closely mapped gene was the target of such mutations, we examined the coding region of a recently identified gene mapping to 10q23, next to D10S579, a marker estimated to lie no more than 1 Mb centromeric of PTEN.15 The multiple inositol polyphosphate phosphatase, known as MINPP1 orMIPP, has been cloned and shown to encode a conserved domain common to histidine phosphatases.15 16 MINPP1 codes for an approximately 52 kDa enzyme with the ability to remove the 3-phosphate from inositol phosphate substrates, such as Ins (1,3, 4,5)P4, as well as other inositol moieties. It has been shown that human MINPP1 has a wide tissue distribution pattern and its subcellular localisation appears to be targeted to the endoplasmic reticulum (ER).15 16 While little is known about the human MINPP1 function, its most well studied homologue, chick HiPER1, has a more restricted tissue distribution and appears to be critical to regulate the transition of growth plate chondrocytes from proliferation to hypertrophy.17 It is presumed that humanMINPP1 plays a role in differentiation and apoptosis, although details on the pathways involved in such signalling are as yet unknown. Thus, owing to its chromosomal location and to the fact that, like PTEN, it encodes a phosphatase with activity towards lipid substrates, we sought to investigate whether mutations inMINPP1 would account for cases of CS and BRR without detectable PTEN mutations.
We obtained DNA from 36 subjects who met stringent criteria for the diagnosis of CS (n=14) and BRR (n=22) and in whom noPTEN mutation had been detected.12 13 In at least one of the families, linkage data were compatible with linkage of the CS phenotype with the 10q23 region.14 The rest of the cases were isolated or belonged to small families where linkage analysis was impossible. Informed consent was obtained from all subjects enrolled in this study, according to institutional Human Subjects Protection Committee protocols. All samples were screened for mutations in the coding region of MINPP1 and most intron-exon boundaries of the gene by PCR based (primer sequences and PCR conditions in table 1) direct sequence analysis, as previously described.18 No MINPP1 mutations were found in germline DNA from any of the subjects examined in the present study. In particular, no mutations were found at the highly conserved histidine phosphatase motif, RHGxRxP, which defines members of the histidine acid phosphatase family. In addition, a second highly conserved site in this group of phosphatases comprising a histidine residue located at position 370 was found to be intact in all samples examined. This represents a proton donor site at the carboxy-terminal region of the protein which appears to be critical for full catalytic activity of this group of enzymes.15 16 We identified five variations from the reference MINPP1 sequence from the database in all samples, as well as in three normal controls (GenBank accession number AF046914). All of these sequence variants were identical to the reference MINPP2 sequence (GenBank accession number AF084943). A sixth variant, c.444A→G, was noted in all our sequences which is in agreement with the MINPP1reference sequence, but at odds with that of MINPP2. It is likely, therefore, that these variations might represent errors in sequence entry on the database, rather than being associated with any particular phenotype, as they were identical in all samples, including the normal controls.
While described as independent hamartoma syndromes with shared clinical features until recently, it has been generally accepted that only CS bears a higher susceptibility to malignancies.19 20 A broad analysis of genotype-phenotype data in the largest series of both CS and BRR recently undertaken in our laboratory has suggested that they might in fact represent distinct spectra of the same primary disorder.13 These findings have clear implications for the follow up of affected subjects, in which systematic cancer surveillance is now recommended for both disorders, and not only for patients with CS.
In several human malignancies, such as breast, prostate, and thyroid cancer with loss of heterozygosity of 10q and in which noPTEN mutations have been found, it has been suggested that a region proximal to PTEN might be the main target in the tumorigenesis pathway.18 21-24 It remains to be determined whether somatic abnormalities of MINPP1 might be related to any of these sporadic tumours.
In conclusion, we have excluded an important candidate gene as the primary genetic abnormality underlying CS and BRR in subjects without identifiable PTEN mutation. It is possible that some degree of genetic heterogeneity exists, as suggested by a study that has excluded linkage to 10q23 in two PTEN mutation negative CS families.9 The major genetic defect responsible for CS and BRR in cases without detectable PTENmutation still remains to be established.
This work was partially supported by the US Army Breast Cancer Research Program (to CE) and P30CA16058 from the National Cancer Institute, Bethesda, MD (Ohio State University Comprehensive Cancer Center). PLMD is a Susan G Komen Breast Cancer Research Foundation Postdoctoral Research Fellow (to CE) and OG a Fellow of the DFG.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.