Article Text

Download PDFPDF

The contribution of germline rearrangements to the spectrum of BRCA2 mutations
  1. F Casilli1,
  2. I Tournier1,
  3. O M Sinilnikova2,
  4. F Coulet3,
  5. F Soubrier3,
  6. C Houdayer4,
  7. A Hardouin5,
  8. P Berthet5,
  9. H Sobol6,
  10. V Bourdon6,
  11. D Muller7,
  12. J P Fricker7,
  13. C Capoulade-Metay8,
  14. A Chompret8,
  15. C Nogues9,
  16. S Mazoyer10,
  17. P Chappuis11,
  18. P Maillet11,
  19. C Philippe12,
  20. A Lortholary13,
  21. P Gesta14,
  22. S Bézieau15,
  23. C Toulas16,
  24. L Gladieff16,
  25. C M Maugard17,
  26. D M Provencher17,
  27. C Dugast18,
  28. C Delvincourt19,
  29. T D Nguyen19,
  30. L Faivre20,
  31. V Bonadona21,
  32. T Frébourg1,
  33. R Lidereau9,
  34. D Stoppa-Lyonnet4,
  35. M Tosi1
  1. 1Inserm U614, IFRMP, Faculty of Medicine, Rouen, France
  2. 2Plate-forme Mixte de Génétique Constitutionnelle des Cancers Fréquents, Hospices Civils de Lyon/Centre Léon Bérard, Lyon, France
  3. 3Unité d’Oncogénétique et Angiogénétique Moléculaire, Pitié Salpétrière, Paris, France
  4. 4Service de Génétique Oncologique, Institut Curie, Paris, France
  5. 5Laboratoire de Biologie Clinique et Oncologique, Centre François Baclesse, Caen, France
  6. 6Inserm EPI 9939, Institut Paoli-Calmettes, Marseille, France
  7. 7Unité d’Oncogénétique, Centre de Lutte Contre le Cancer Paul Strauss, Strasbourg, France
  8. 8Service de Génétique, Institut Gustave Roussy, Villejuif, France
  9. 9Service d’Oncogénétique, Inserm U735, Centre René Huguenin, Saint-Cloud, France
  10. 10Laboratoire de Génétique, UMR5201 CNRS, Université Claude Bernard Lyon I, Lyon, France
  11. 11Service d’Oncologie, Hôpital Universitaire de Genève, Geneva, Switzerland
  12. 12Laboratoire de Génétique, CHU, Nancy, France
  13. 13Centre Catherine de Sienne, Nantes, France
  14. 14Service d’Oncologie, Centre Hospitalier, Niort, France
  15. 15Laboratoire de Génétique Moléculaire, CHU, Nantes, France
  16. 16Institut Claudius Regaud, Toulouse, France
  17. 17CR-CHUM, Service de Médecine Génique, Département de Médecine, Université de Montréal, Montreal, Canada
  18. 18Centre Régional de Lutte Contre le Cancer Eugene Marquis, Rennes, France
  19. 19Centre Régional de Lutte Contre le Cancer Institut Jean Godinot, Reims, France
  20. 20Oncogénétique, Centre Georges Francois Leclerc et Hôpital Le Bocage, Dijon, France
  21. 21Unité de Prévention et d’Epidémiologie Génétique, Centre Léon Bérard, Lyon, France
  1. Correspondence to:
 Mario Tosi
 Inserm U614, Faculty of Medicine, 22 boulevard Gambetta, 76183 Rouen, France; mario.tosi{at}


Background: Few germline BRCA2 rearrangements have been described compared with the large number of germline rearrangements reported in the BRCA1 gene. However, some BRCA2 rearrangements have been reported in families that included at least one case of male breast cancer.

Objective: To estimate the contribution of large genomic rearrangements to the spectrum of BRCA2 defects.

Methods: Quantitative multiplex PCR of short fluorescent fragments (QMPSF) was used to screen the BRCA2 gene for germline rearrangements in highly selected families. QMPSF was previously used to detect heterozygous deletions/duplications in many genes including BRCA1 and BRCA2.

Results: We selected a subgroup of 194 high risk families with four or more breast cancers with an average age at diagnosis of ⩽50 years, who were recruited through 14 genetic counselling centres in France and one centre in Switzerland. BRCA2 mutations were detected in 18.6% (36 index cases) and BRCA1 mutations in 12.4% (24 index cases) of these families. Of the 134 BRCA1/2 negative index cases in this subgroup, 120 were screened for large rearrangements of BRCA2 using QMPSF. Novel and distinct BRCA2 deletions were detected in three families and their boundaries were determined. We found that genomic rearrangements represent 7.7% (95% confidence interval 0% to 16%) of the BRCA2 mutation spectrum.

Conclusion: The molecular diagnosis of breast cancer predisposition should include screening for BRCA2 rearrangements, at least in families with a high probability of BRCA2 defects.

  • familial breast cancer
  • molecular diagnosis
  • quantitative multiplex PCR of short fluorescent fragments

Statistics from

Screening for BRCA1 rearrangements has become part of the routine molecular diagnosis of predisposition to breast/ovarian cancer, but relatively few BRCA2 germline rearrangements have been reported.1–6 Germline rearrangements of BRCA1 and BRCA2 have been recently reviewed.7 We have previously shown that BRCA2 germline rearrangements are present in a significant number of male breast cancer families3 and this finding suggested that they might represent a sizable proportion of the BRCA2 mutation spectrum. To test this hypothesis in more depth, we screened for BRCA2 rearrangements in 120 high risk breast cancer families negative for BRCA1 and BRCA2 mutations, by using quantitative multiplex PCR of short fluorescent fragments (QMPSF), a rapid and sensitive method able to detect heterozygous genomic deletions and duplications.3,8,9



Inclusion criteria for families were at least four cases of female breast cancer in the same parental lineage and average age at diagnosis of 50 years or less. Families also presenting ovarian or male breast cancer cases were included. All index cases had been found negative for BRCA1 and BRCA2 mutations in previous analyses, including screening for large deletions or duplications of BRCA1. Screening of all exons and exon-intron boundaries for point mutations and microdeletions/insertions was carried out by direct sequencing or by denaturing high performance liquid chromatography (dHPLC). All DNA samples studied were from centres with institutional and national approval to carry out molecular diagnosis of BRCA1 and BRCA2 defects. All patients had signed the appropriate informed consent for participation in mutation analysis.

Screening for BRCA2 rearrangements

QMPSF was carried out as described recently.3 DNA concentrations were measured using a picogreen assay (PicoGreen dsDNA Quantitation Kit, Invitrogen, Cergy-Pontoise, France) and 100 ng samples were used in each QMPSF assay. The deletion of exons 12 and 13 found in one family was confirmed by using a multiplex ligation-dependent probe amplification (MLPA) BRCA2 kit (MRC Holland, Amsterdam, the Netherlands). All rearrangements found were confirmed by long range PCR as previously described8 and the boundaries of each deletion were determined by DNA sequencing. Conditions used for long range PCR and the sequences of primers used in QMPSF assays are available upon request.


We have estimated the prevalence of BRCA1 and BRCA2 mutations by family type, considering the number of cases as well as cancer type and age at diagnosis, in families studied in 14 genetic counselling centres in France and one centre in Switzerland and tested for mutations over a 1 year period. From subgroups with relatively large numbers of families and a high proportion of BRCA2 mutations, we selected a subgroup consisting of 194 families with four or more breast cancers with an average age at diagnosis of 50 years or less.

BRCA2 mutations were detected in 36 index cases (18.6%) and BRCA1 mutations, including one germline BRCA1 rearrangement, were detected in 24 index cases (12.4%) from these families. Among the 134 BRCA1/2 negative index cases from this subgroup, 120 were screened for large rearrangements of BRCA2 using QMPSF.3,8 The QMPSF assay covers all 27 exons of BRCA2, except for exon 24, which is separated from exon 23 by a very short intron. The large exon 11 is targeted by two distinct amplicons located near its 5′ and 3′ ends. All QMPSF assays were carried out in the same laboratory using as controls DNA from a chromosome 13 trisomy case and from a patient with a complete BRCA2 deletion.3 Three families revealed distinct BRCA2 rearrangements (fig 1 and table 1). The deletion of exons 14–18 (family IC3808) and the deletion of exons 15 and 16 (family PS1188) had not been previously described, whereas deletions of exons 12 and 13 had previously been reported in other families.2,3

Table 1

BRCA2 germline rearrangements found in this study

Figure 1

 Detection by QMPSF of a novel deletion of BRCA2 exons 14–18. Fluorescence profiles obtained from patient DNA and control DNA (red and blue, respectively) were superimposed and adjusted with the GeneScan (Applied Biosystems, Foster City, CA) program to match the peak heights of the control amplicon (C), consisting of a short segment of the MLH1 gene. The deletion of five consecutive exons is shown by a ∼50% reduction in the heights of the corresponding peaks.

Long range PCR assays followed by sequencing analysis were used to determine the boundaries of each of the three deletions. In particular, it was found that the boundaries of the exon 12-exon 13 deletion differ at the sequence level from the other two reported deletions involving the same exons (table 2). While most BRCA1 and BRCA2 rearrangements, whose boundaries have been characterised, involve at least one Alu repeat, a recent report indicates that this is not always the case for BRCA2 rearrangements, suggesting the presence of other recombinogenic elements.6

Table 2

 Summary of BRCA2 germline rearrangements

Two of the three novel BRCA2 deletions found in the present study are mediated by unequal homologous recombination involving two Alu repeats, whereas the deletion of exons 15 and 16 reveals the presence of an Alu repeat only at the breakpoint within intron 16, while the other breakpoint is at the 3′ boundary of intron 14 (table 2). This table shows that the boundaries of BRCA2 rearrangements are more often within introns, and especially within Alu sequences. However, large deletions with boundaries in exons as well as insertions of Alu elements into exons of BRCA1 or BRCA2 have also been reported.1,7,10,11 This type of rearrangement may be missed in some cases by gene dosage methods, such as QMPSF or MLPA12 which target only portions of exons or exon-intron boundaries. In this regard, it is worth noting that QMPSF (range of amplicon sizes from 152 to 280 bp) provides larger coverage of BRCA2 exons than MLPA (target size ∼60 nucleotides). While large double blinded studies are still lacking, our present experience indicates that both methods provide similar sensitivity of detection, if one compares profiles obtained from good quality DNA samples, preferably extracted with the same method. With both techniques, confirmation by a second method is necessary when deletions or duplications are indicated by a change affecting a single peak. MLPA is a multi-step assay, in which probe hybridisation, probe ligation, and PCR amplification are performed sequentially, whereas QMPSF involves only one step before capillary electrophoresis. This feature of QMPSF minimises the risk of sample cross-contamination and facilitates full automation of the assay. Conversely, MLPA allows analysis of the entire BRCA2 gene in a single run of a capillary electrophoresis procedure, whereas, using QMPSF, three QMPSF reactions followed by three parallel capillary runs are currently needed for each DNA sample (fig 1).

The three BRCA2 deletions identified in the present screening can be added to 36 BRCA2 mutations previously detected in the subgroup of 194 families selected for this study (four or more breast cancers with an average age at diagnosis of ⩽50 years). Our finding suggests that 7.7% (95% confidence interval 0% to 16%) of BRCA2 mutations are germline rearrangements. This estimate appears to be rather similar to that recently reported in a study of 121 selected breast and breast-ovarian cancer families from Italy, where three deletions were found and their contribution to the BRCA2 spectrum was estimated to be 11% (95% confidence interval 3% to 23%).6 Considering the large confidence intervals, the conclusions of both studies are in good agreement and suggest that the contribution of rearrangements to the mutation spectrum of BRCA2 is comparable to that of rearrangements in the mutation spectrum of BRCA1, currently estimated at about 15%.7

The present study extends the conclusions of our previous screening of male breast cancer families for BRCA2 rearrangements3 and concurs with two recent reports5,6 in suggesting that searching for large BRCA2 deletions/duplications should be added to the procedure for the molecular diagnosis of breast cancer predisposition, in particular in breast cancer families having a high probability that their disease is due to a BRCA2 defect, as in families with multiple cases of female breast cancer and families with cases of male breast cancer.


F Casilli wishes to thank E Ricevuto, A Pavan, P Marchetti, and AISCuP (Associazione Italiana Studio e Cura del Paziente Oncologico) for help and support during her visit to Rouen. We wish to thank M Belotti for assistance.



  • This work has been supported by grants from Association pour la Recherche sur le Cancer (ARC) and Ligue contre le Cancer, Comité de l’Eure. I Tournier was supported by a fellowship from Ministère de l’Education Nationale, de la Recherche et de la Technologie (MENRT) and is currently supported by a fellowship from ARC.

  • Competing interests: none declared

Request Permissions

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.