Marfan syndrome (MFS) is a dominant disorder, mainly caused by mutations in the fibrillin-1 gene (FBN1) located on chromosome 15q21.1. The estimated prevalence of MFS is about 1 in 10000. Approximately 25% of MFS patients are sporadic cases due to new mutations [1, 2]. Different tissues and organs can be affected, especially the cardiovascular, skeletal, and ocular systems. Diagnostic criteria are well established and known as the Ghent criteria [3]. However, the inter- and intra-familial variability of the phenotype limits the establishment of genotype-phenotype correlations. To date, more than 1329 FBN1 mutations have been published (http://www.hgmd.cf.ac.uk/ac/gene.php?gene=FBN1), but only a few are recurring mutations. Missense mutations substituting or creating a cysteine molecule in one of the calcium-binding EGF domains are the most prevalent. The others are frameshift, splice-site, nonsense mutations and in-frame deletions and insertions. Heterozygous mutations in the genes coding for transforming growth factor beta receptors I (TGFBR1) and II (TGFBR2) have also been reported in patients with MFS and MFS-related disorders, indicating genetic heterogeneity [4–6]. Interstitial deletions involving the 15q21.1 band and the FBN1 gene are very rare. To our best knowledge, there are only six reports in the literature describing deletions of the whole FBN1 gene detected by molecular techniques, and only in five of them this gene deletion was associated with classical MFS [2, 7–11]. The patients described by Adès et al. [8] and Hutchinson et al. [7] had clinical features of the MFS spectrum and mental retardation, but the size of their deletions was not determined. The patient described by Faivre et al. [10] had a 2.97 Mb deletion and some features of MFS but no mental retardation. Hilhorst-Hofstee et al. [2] described 10 cases (five of the same family) with a complete FBN1 gene deletion, screened by Multiplex Ligation-dependent Probe Amplification (MLPA) analysis of 300 patients presenting from mild MFS features to the classical MFS or an MFS-related phenotype. Recently, Furtado et al. [11] studied 14 patients from 11 unrelated families with aortic aneurysm. Three patients of the one family who met clinical diagnostic criteria for Marfan syndrome had a 542 Kb deletion in chromosome 15 including the whole FBN1 gene; the study of the region was refined by MLPA and array analysis. Cases with chromosomal alterations including a genomic deletion of the whole FBN1 gene are rare, but their number has increased after the adoption of routine screening by molecular techniques (MLPA, array), especially in patients with MFS. To date, there is no description in the literature of a patient with MFS and a complex chromosome rearrangement (CCR). We report here on a 16-year-old female patient displaying features of the MFS spectrum and mental retardation, who was found to present a 1.9 Mb deletion at 15q21.1 (refined by array) including the FBN1 gene, and a novel CCR among chromosomes 6, 12 and 15.

We report here on a girl with clinical features of the MFS spectrum and a 15q21.1 deletion including the entire FBN1 gene. This is the seventh study in the literature in which the deletion of FBN1 is confirmed by molecular techniques. Thus, to this date 18 patients with deletions including the whole FBN1 gene were reported, five of them belonging to the same family. Interestingly, only 12 of these patients [2,7,11, present case] present the typical MFS phenotype according to the Ghent criteria. Clinical variability in patients with different point mutations has been described, with patients with a nonsense mutation presenting milder phenotypes than patients with a missense mutation [10]. Cases with a deletion can be compared to patients with a nonsense mutation, since the truncated mRNA in the latter cases is believed to be reduced, due to the nonsense-mediated decay (NMD) mechanism that prevents the expression of the truncated mRNA [2, 10, 13]. Thus, in cases of deletion and nonsense mutation, the MFS phenotype can result from FBN1 gene haploinsufficiency and, especially in cases of missense mutations, from a dominant negative effect. Hutchinson et al [7] suggested that the clinical variability in MFS could be also due to variable FBN1 expression of the normal allele. The size of the deletions reported varies from small (less than 300 kb in 5 members of the same family) [2], including only the FBN1 gene, to large (up to 17.7 Mb) [9]. As found in our patient, when the deletion involves other genes besides FBN1, other unusual features can be found, such as those described here (craniostenosis, hypothyroidism and intellectual deficit). The patient described by Hiraki et al [9] presented no sign of the MFS syndrome, probably due to her young age and severe clinical phenotype. Our patient showed most of the skeletal features of MFS and an aortic root dilatation, but no ectopia lentis. Concerning the ocular system, she presented myopia and astigmatism. Her deletion comprises 19 genes and predicted genes including FBN1, besides SEMA6D and COPS2 that may have contributed to the intellectual deficit and hypothyroidism, respectively. Of the 18 patients with a complete deletion of the FBN1 gene described so far, only seven were karyotyped: two presented normal karyotypes [2, 10], three had visible 15q deletions [7–9], and one had a de novo translocation between the long arms of chromosomes 12 and 15 and a 4.9 Mb interstitial deletion at the translocation breakpoint of the long arm of chromosome 15 between the bands q21.1 and q21.2 [2]. Hilhorst-Hofstee et al [2] performed karyotype analysis only in two patients, as part of the mental retardation screening. So, our patient is the first MFS case described presenting a complex chromosome rearrangement among chromosomes 6, 12 and 15. In the literature, up to 30-50% of the patients with a chromosomal rearrangement, both complex and reciprocal translocations, show an imbalance on the chromosomal or molecular level as an explanation for their phenotype [14]. The greater the number of breakpoints involved in a CCR, the greater the likelihood of genomic imbalances or position effect. Disruption of a gene could unmask a recessive mutation on the homologue allele, suggesting a greater chance for an abnormal phenotypic outcome [15].

We emphasize the importance of using a combination of different molecular cytogenetic techniques in cases of chromosomal and/or genomic rearrangements involving the FBN1 gene, in order to better understand the extent of the molecular etiology of the Marfan syndrome and also to elucidate the genetic constitution of CCRs associated with diseases.