Article Text
Abstract
We have performed an extensive analysis of TP53 in 474 French families suggestive of Li–Fraumeni syndrome (LFS), including 232 families fulfilling the Chompret criteria. We identified a germline alteration of TP53 in 82 families (17%), in 67/232 of the families fulfilling the Chompret criteria (29%) and in 15/242 which did not fulfil these criteria (6%). Most of the alterations corresponded to missense mutations (67%), and we identified in four families genomic deletions removing the entire TP53 locus, the promoter and the non-coding exon 1, or exons 2–10. These results represent a definitive argument demonstrating that LFS results from TP53 haplodeficiency. The mean ages of tumour onset were significantly different between patients harbouring TP53 missense mutations and other types of alterations, missense mutations being associated with a 9 year earlier tumour onset. These results confirm that missense mutations not only inactivate p53 but also have an additional oncogenic effect. Germline alterations of TP53 that lead exclusively to loss of function are therefore associated with a later age of tumour onset and the presence of such mutations should be considered in atypical LFS families with tumours diagnosed after 40 years.
Statistics from Altmetric.com
The Li-Fraumeni syndrome (LFS, MIM 151623), which represents one of the most dramatic inherited forms of cancer, is characterised by an early age of tumour onset and a wide tumour spectrum.1–3 The characteristic LFS tumour spectrum includes soft tissue sarcomas, osteosarcomas, pre-menopausal breast cancers, brain tumours and adrenocortical tumours. Patients also have an increased risk for the development of other malignancies, such as leukaemias and lymphomas, gonadal cell tumours, gastric, lung and colorectal cancers.4 5 This wide spectrum of malignancies, the range of the age of tumour onset, the high risk of developing multiple primary tumours,6 and the abnormal sensitivity of the patients to ionising radiations7 8 complicate the clinical management of the families. Seventeen years ago, it was shown that LFS results from germline heterozygous alterations of the TP53 tumour suppressor gene,9 10 located on chromosome 17p13. Although some studies have suggested that LFS could also result from CHEK2 mutations,11–13 or be linked to chromosome 1q23 in some families,14 there is, at the present time, no definitive data supporting the involvement of a gene other than TP53 in LFS.15 The dramatic clinical presentation of the syndrome is probably explained by the key cellular role of the p53 protein, a potent transcription factor able, in the presence of DNA damage, to activate numerous target genes involved in cell cycle arrest, DNA repair and apoptosis.16
Identification of a germline TP53 mutation in a patient will make it possible to: (1) confirm on a molecular basis the diagnosis of LFS, whose clinical recognition can be complicated, considering the remarkable heterogeneity of the presentation; (2) ensure a regular clinical review by an informed clinician in order to avoid a delay in the diagnosis of a second tumour; (3) offer to women breast imaging screening programme; (4) avoid radiation, whenever possible, and (5) offer prenatal diagnosis to the families. In order to facilitate the recognition of the syndrome, the French LFS working group has elaborated practical criteria,17 which are looser than the original criteria, defined by Frederick Li and Joseph Fraumeni.3 These “Chompret” criteria integrate the different clinical situations suggestive of LFS: (1) a proband with a tumour belonging to the narrow LFS tumour spectrum (soft tissue sarcoma, osteosarcoma, brain tumour, pre-menopausal breast cancer, adrenocortical carcinoma) before 36 years and at least one first or second degree relative with a narrow LFS tumour (except breast cancer if the proband is affected by breast cancer) before 46 years or with multiple tumours; or (2) a proband with multiple tumours, two of which belong to the narrow LFS tumour spectrum and the first of which occurred before 36 years; or (3) a child with adrenocortical carcinoma irrespective of the family history.
We have performed an extensive analysis of TP53, based on complete sequencing of the 11 exons and on QMPSF to detect genomic rearrangements,18 in 474 families suggestive of LFS, and our results provide new insights into the molecular basis of the syndrome.
METHODS
Patients
A total of 474 unrelated French families were recruited by the French LFS network. Approximately half of these families (232, 49%) fulfilled the Chompret criteria.17 The 242 other families were selected because they fulfilled either the age or tumour spectrum Chompret criteria (197) or because they corresponded to sporadic early onset breast cancers (diagnosed before 33 years of age) without detectable germline BRCA mutation (45). Among the 474 families, only 34 fulfilled the original criteria for LFS.3
Molecular analyses
The 11 exons of TP53, including the first non-coding exon, and the intron exon boundaries were polymerase chain reaction (PCR) amplified from genomic DNA (primers sequences are supplied in supplemental tables), and analysed using an automated sequencer (PE Applied Biosystems, Foster City, California, USA). The QMPSF analysis of the 11 exons of TP53 has previously been described.18 Identification of a TP53 alteration in an index case was systematically confirmed on two analyses performed on two independent blood samples. The MDM2 SNP309 genotyping was performed, as previously described.19
Statistical analysis
The Mann–Whitney U test was used to compare the mean age of onset between the different groups.
RESULTS AND DISCUSSION
We detected a germline alteration of TP53 in 82/474 families (17%). Among the 232 families fulfilling the Chompret criteria, 67 were found to have a TP53 mutation (29%), and among the 242 other families, 15 carried a mutation (6%). These 15 families, which did not fulfil the Chompret criteria when they were analysed for TP53, included 12/197 (6%) families fulfilling only the tumour spectrum or age Chompret criteria: four families were characterised by a tumour onset after 46 years of age in the relative or after 36 years of age in the proband in case of multiple primary cancers; in eight families, the relative had developed a tumour not included in the narrow spectrum: leukaemia, lung bronchoalveolar carcinoma or neuroblastoma. Finally, we found a TP53 mutation in 3/45 (7%) sporadic cases of early onset breast cancers (<33 years) without detectable BRCA mutation.
In this series of 82 families with TP53 mutation, the median age of tumour onset was 25 years. Tumour frequencies are presented in table 1. The germline TP53 alterations corresponded, in most of the cases (78/82, 95%), to point mutations or small deletions or insertions, widely distributed between exons 3–11 (table 2 and supplemental tables). In four families, QMPSF analysis revealed a genomic rearrangement of the TP53 locus: a complete deletion extending on 44.6 kb, that we had previously reported20; two deletions removing only the promoter and the non-coding exon 1 of TP53 and extending to intron 3 of the centromeric FLJ10385 gene, as shown by subsequent quantitative multiple PCR of short fluorescent fragments (QMPSF) scanning; and a deletion of exons 2–10. The families harbouring these genomic rearrangements presented with typical LFS syndrome fulfilling the original LFS criteria. The identification of germline TP53 deletions in LFS families is of interest for our general understanding of the basis of LFS since it provides the final argument that LFS results from TP53 haplodeficiency. In our series, missense mutations represented 67% of the germline alterations (table 2). All the missense TP53 mutations that we tested in the FASAY (functional assay on separated alleles in yeast)20 were found to inactivate the transcriptional activity of the protein, in agreement with the fact that TP53 haplodeficiency causes LFS. Nevertheless, this does not explain the remarkable mutation spectrum characterised by the predominance of missense mutations.
Interestingly, TP53 wt/mt mice harbouring a heterozygous germline missense mutation have been shown to develop additional tumours, compared to TP53 wt/- mice,21 22 suggesting that germline TP53 missense mutations not only inactivate the transcriptional activity of the protein, but may have an additional oncogenic effect. This observation led us to compare in this series the mean ages of tumour onset between patients harbouring TP53 missense mutations and patients with other types of alterations. As shown in table 3, the mean ages of tumour onset between the two groups were significantly different (p = 0.004), missense mutations being associated with a 9 year earlier age of tumour onset. Our results confirm one previous study on smaller series reporting that TP53 missense mutations, compared to inactivating mutations, are associated with an earlier age of tumour onset.23 This is in agreement with the mouse models showing that missense TP53 mutations have an additional oncogenic effect. This oncogenic effect could be related, as previously suggested, by the ability of the missense mutants either to deregulate gene expression,24 25 or to induce genetic instability by disrupting ataxia telangiectasia mutated (ATM) pathway.26 The association of a later age of tumour onset with TP53 null mutations is illustrated by some families presenting a tumour spectrum suggestive of LFS but with a first tumour being diagnosed in the index case after 40 years of age. Figure 1 shows three LFS pedigrees characterised by a late tumour onset in the proband and associated with non-sense, splicing mutations or in frame deletions.
The other genetic factor, which has recently been shown by three independent studies19 27 28 to modulate the age of tumour onset in TP53 mutation carriers, is the SNP309 G allele of the MDM2 gene encoding the main negative regulator of p53. We analysed the effect of the MDM2 SNP309 G allele in this series for 88 TP53 mutation carriers when genomic DNA was available. The age of tumour onset in the G carriers (corresponding to the T/G+G/G genotypes) was clearly different from that observed in the T/T patients (p = 0.01), the presence of the G allele being associated with an 8 year earlier age of tumour onset (table 4). In our series, we observed a cumulative effect of the type of TP53 alteration and the MDM2 SNP309 (table 5).
In conclusion, this study which was performed on 474 patients shows that: (1) germline TP53 alterations can be detected in 29% of the LFS families selected on the Chompret criteria, and only in 6% of the families which do not fulfil these criteria. This confirms the validity of the Chompret criteria and argues for their use in clinical practice; (2) genomic rearrangements represent 5% of the germline TP53 alterations and their identification in LFS families demonstrates that LFS results from loss of function at the TP53 locus; (3) missense mutations, which represent 67% of the germline TP53 alterations, are associated with a 9 year earlier age of tumour onset, and this is probably explained by an additional oncogenic effect; (4) the MDM2 SNP309 G allele is associated with an 8 year earlier age of tumour onset; and (5) the effects of missense mutations and MDM2 SNP309 G allele may be cumulative. Germline alterations of TP53 that lead exclusively to loss of function are therefore associated with a later age of tumour onset and the presence of such mutations should be considered in atypical LFS families with tumours diagnosed after 40 years.
This study led us to propose that the Chompret criteria can be updated by extending the age of tumour onset and LFS tumour spectrum: (1) a proband with a tumour belonging to the LFS tumour spectrum (soft tissue sarcoma, osteosarcoma, brain tumour, pre-menopausal breast cancer, adrenocortical carcinoma, leukaemia, lung bronchoalveolar cancer) before 46 years and at least one first or second degree relative with an LFS tumour (except breast cancer if the proband is affected by breast cancer) before 56 years or with multiple tumours; or (2) a proband with multiple tumours, two of which belong to the narrow LFS tumour spectrum and the first of which occurred before 46 years; or (3) a patient with adrenocortical carcinoma or a patient with breast cancer before 36 years of age without BRCA mutation, irrespective of the family history.
Acknowledgments
This manuscript is dedicated to our colleague Agnès Chompret. We are grateful to Mario Tosi for critical review of the manuscript. This work was supported by the INCa, the French National Cancer Institute.
REFERENCES
Supplementary materials
web only appendix 45/8/535
Files in this Data Supplement:
Footnotes
▸ Supplemental tables are published online only at http://jmg.bmj.com/content/vol45/issue8
Competing interests: None declared.