Statistics from Altmetric.com
Editor—Carriers of de novo balanced reciprocal translocations and inversions have an increased risk of approximately 6% for developing multiple congenital abnormalities (MCA) and/or mental retardation (MR), compared to a 2-3% risk overall in newborn populations.1 2 Cytogenetically cryptic deletions or physical disruption or inactivation of a gene(s) in one or both breakpoint regions may account for the observed phenotypes.3 4 It seems plausible to assume that the risk for MCA/MR may be even higher in carriers of de novo complex chromosome rearrangements (CCRs), which involve at least three different chromosomes and breakpoint regions. Extreme cases involving up to seven chromosomes and 10 breakpoints have been described.5-7Indeed, most reported CCRs are associated with MCA/MR.8 9In addition, they have been found in infertile men10 and in women suffering from multiple miscarriages.11 12
The complex nature of CCRs renders karyotype interpretation by classical chromosome banding alone difficult. In many cases fluorescence in situ hybridisation (FISH) will be the best method to delineate the underlying chromosome rearrangements.7 13Here we have applied conventional FISH with chromosome painting probes and region specific large insert clones, comparative genomic hybridisation (CGH),14 15 and spectral karyotyping (SKY)16 17 to an apparently balanced and very complex rearrangement in a profoundly retarded patient with Moebius syndrome (MBS, MIM 157900).18 MBS is characterised by congenital paralysis of the seventh cranial nerve leading to facial diplegia.19 Other cranial nerves may also be affected. In addition, orofacial and limb malformations, defects of the musculoskeletal system, and MR may occur.
This patient, who has classical Moebius syndrome, has been reported previously.4 He was the third child born to a 35 year old father and a 32 year old mother. Because of paresis of the facial muscles as a newborn, he had feeding problems owing to inefficient sucking and swallowing. A more detailed examination at 17 years showed moderate to severe MR and deficient language development. At this time, his height was 1.56 m and his weight 39.5 kg (<3rd centile). Other clinical features included ptosis, blepharophimosis, atrophy of the optic nerve, and severe amblyopia. His mandible and facial muscles were hypoplastic. His metacarpals and metatarsals were short, limiting the mobility of his hands and feet. The pectoral muscles were hypoplastic.
Q band analysis20 showed an apparently balanced de novo rearrangement between chromosomes 7, 8, 11, and 13. Breakpoints were assigned to chromosome bands 7q21.1, 8q21.3, 11p14.3, and 13q21.2 (fig1). CGH did not detect any chromosomal imbalances (sensitivity: gain or loss of >10 Mb DNA) in the patient's karyotype (data not shown). Since an unambiguous interpretation of the rearranged karyotype was not possible by classical chromosome banding analysis, we applied chromosome specific and region specific FISH probes. A schematic representation of our molecular cytogenetic mapping results is shown in fig 2A.
Simultaneous visualisation of all 24 human chromosomes in different colours by SKY was very beneficial in dissecting the patient's karyotype (fig 2B). It confirmed the involvement of chromosomes 7, 8, and 13 in the CCR. The short derivative chromosome 7 was painted in its entirety by the chromosome 7 library. The material distal to 7q21 was translocated onto the der(8) and the der(13). The segment of the der(8) distal to 8q21 was translocated onto der(13). Inversely, the distal long arm 13q21-qter was moved to the der(8). Thus, the der(8) carried additional material from both chromosomes 7q and 13q, whereas the der(13) gained material from both chromosomes 7q and 8q. The chromosome 11 rearrangement, which had been suspected after Q banding, could not be characterised further by SKY. Some metaphases suggested a small insertion into 11p (data not shown); however, the chromosomal origin of the inserted material could not be identified. No other chromosome showed detectable abnormalities by SKY.
To overcome the decreased resolution of SKY compared with that of conventional FISH, we performed two colour FISH experiments with DNA libraries of chromosomes 7 and 13 (fig 2C) and of chromosomes 8 and 11 (fig 2D), respectively. In addition to the der(8) and the der(13) detected by SKY, we found an insertion of chromosome 8 material into chromosome 11p14.3 in all metaphases analysed (fig 2D). Sometimes the chromosome 11 paint produced a weak hybridisation signal on the long arm of the der(8) (fig 2D), indicating a reciprocal exchange event between the der(8) and the der(11) (fig 2A).
To FISH map this CCR in fine detail, we selected region specific non-chimeric clones from our standard set of, so far, >2000 cytogenetically and genetically anchored CEPH YACs. These YACs, on average located at 3 cM intervals spaced over the entire chromosome complement, can be adapted with considerable flexibility to the study of chromosome breakpoints.4 21 YACs 690h8, 808f3, and 929a10 mapped to the normal chromosome 7q21-31 and to der(13) (table1). This implies that the proximal part of 7q21-qter, which is missing on the der(7), was translocated onto der(13), whereas the distal part, 7q31-36, was presumably translocated onto der(8). Since the subtelomeric YAC 965c12 showed signals on both the normal and the derivative chromosome 7, but not on der(8) or on der(13), the chromosome 7q end is retained on the der(7). Interestingly, YAC 941d9 from chromosome 7q21 hybridised to the short arm of the der(11). This was not seen by SKY or chromosome painting probes. Taken together, our results indicate that material from the distal long arm of der(7) was relocated to three different derivative chromosomes, the der(8), der(11), and der(13), involving at least four breakpoints on chromosome 7 (fig 2A). The chromosome 8 material inserted into 11p is highlighted by YAC 662e12 from 8q21.3 (table 1). Thus, the small 11p insertion contains sequences from both chromosomes 7 and 8 (fig 2A). YACs 962g12 and 953h7 map proximal to the 8q21.3 breakpoint and, therefore, hybridise to both the normal and the der(8). YACs 940h5, 964f4, 942b11, and 765a11 gave signals on the normal chromosome 11p13-15 and on the der(11) (table 1). The breakpoint region on chromosome 13q21.2 has been studied previously.4 The breakpoint spanning YAC 882b4 from 13q21-22 produced a split signal on both the der(13) and der(8).
Autosomal dominant MBS has been linked to markers on 3q21 in a large Dutch family (MBS2, MIM 601471).18 22 To analyse whether rearrangements involving the MBS2 locus may be responsible for the phenotype in our patient, a microdissected DNA library specific for the long arm of chromosome 3,23 as well as YACs 819b5, 930a7, and 754c8 from the critical region on chromosome 3q21 (table 1) were hybridised to the patient's chromosomes. No structural abnormalities were detected on chromosome 3.
Collectively, molecular cytogenetics showed at least four breakpoints on chromosome 7, two each on chromosomes 8 and 11 and one on chromosome 13. To our knowledge, this case represents one of the most complex constitutional chromosome rearrangements reported: 46,XY,t(7;8;11;13) (7pter→7q21.1::7q36→7qter;8pter→8q21.3::11p14.3→ 11p14.3::7q31→7q36::13q21.2→13qter;11pter→11p14.3 ::8q21.3 → 8q21.3::7q21.1 → 7q21.1::11p14.3 → 11qter; 13pter→ 13q21.2::7q21.1→7q31::8q21.3→8qter).
Classical Q, G, or R banding is the screening method of choice for numerical and structural chromosomal abnormalities. However, in some instances, for example, subtelomeric24 and other cytogenetically cryptic rearrangements4 25 or CCRs,7 13 chromosome banding techniques alone may not allow an exact karyotype interpretation and correlation between genotype and phenotype. This is why FISH has become an important accessory technique in clinical cytogenetics. In this study, we have used state of the art molecular cytogenetic tools, most importantly SKY and a panel of region specific YACs, to elucidate a very complex chromosome rearrangement, which occurred as a de novo constitutional aberration in a MBS patient. Of course molecular cytogenetics cannot replace classical chromosome banding, but it certainly is a very useful complementary strategy.
Q banding identified four rearranged chromosomes. SKY was able to visualise this CCR in a single hybridisation experiment and confirmed the involvement of three chromosomes, 7, 8, and 13. However, it did not reliably detect the insertion on chromosome 11p, which contains sequences homologous to YAC 941d9 from chromosome 7 and YAC 662e12 from chromosome 8. Evidently, this insertion, which altogether extends over at least 2 Mb, is below the limit of DNA resolution by SKY. Under optimal experimental conditions, which may be hard to achieve in a routine laboratory and for all chromosomal regions, the resolution of SKY is estimated to be 1-2 Mb (from a single chromosome).16 17 We conclude that SKY is a very straightforward and fast screening technique for identifying (most) chromosomes involved in a given CCR. However, subsequent FISH experiments with chromosome painting probes and region specific YACs are crucial for exact karyotype interpretation. In particular, our standard set of cytogenetically and genetically anchored YAC probes allows the rapid and accurate characterisation of breakpoint regions of interest and of smaller chromosome segments involved in CCRs.
Identification of YAC clones spanning disease associated chromosome breakpoints will greatly facilitate positional cloning projects in cases where the phenotype is caused by disruption of a relevant disease gene in the breakpoint region. We have suggested previously that chromosome 13 may harbour a gene for MBS.4In addition to the patient described here, there is a t(1;13)(p34;q13) translocation segregating in a family with MBS26 and a de novo deletion of 13q12.2 in a girl with MBS.27 Although these chromosome rearrangements are located proximal to the 13q21 breakpoint region of our patient, it is still possible that the same gene is affected. The published cases have not been analysed by molecular cytogenetic techniques, and breakpoint localisation by chromosome banding alone may be inaccurate by several bands. A second locus for MBS has been linked to chromosome 3q21.22However, since this region did not show detectable rearrangements in the CCR studied, it is not very likely that theMBS2 gene on chromosome 3 is affected. MBS is a very heterogeneous clinical syndrome. In a family with autosomal dominant MBS, both the MBS1 locus on chromosome 13q12.2-q13 and the MBS2 locus on chromosome 3q21 have been excluded.22 Therefore, a cytogenetically cryptic deletion or disruption of a gene in a breakpoint region other than 13q21 may be responsible for the MCA/MR phenotype of our patient. In this context, it is interesting to note that an unrelated MBS patient with a breakpoint on chromosome 11p13 has been described.28 Therefore, the 11p insertion may be the chromosome rearrangement underlying the disease in our patient. Of course, we also cannot exclude the possibility that MBS/MR may be associated with the observed CCR by chance or be caused by a non-genetic mechanism.19
More than 100 CCRs have been described so far. With the advances in FISH and multicolour techniques, it has become possible for the first time to study even very complex cases in detail. Overall many CCRs may be even more complicated than suggested by classical chromosome banding. Interestingly, CCR breakpoints are not randomly distributed throughout the entire chromosomal complement, but appear to be clustered in certain chromosome regions. In particular, the 7q21, 8q21, and 11p14 regions, which are involved in our patient's CCR, seem to be prone to chromosome breakage. In a recent survey of 100 CCRs, nine breakpoints were found in 18q21; seven each in 3q13 and 7q22; six in 5p13, 7q11, and 8q22; five in 1p22, 1q42, 3p13, 7q21, 9p24, 11q14, and 21q21; four in 1p13, 2q13, 2q21, 2q33, 3p11, 3q26, 7q31, 8p23, 8q13, 8q24, 11p15, 13q32, 21q11, and 21q22; and three in 1q21, 1q32, 2q22, 2q23, 3p23, 5p14, 5q15, 6p21, 6q27, 7p14, 7q36, 8q21, 9q22, 9q32, 10p12, 10q11, 10q22, 11p14, 11p13, 11q13, 11q21, 12p13, 12q22, 12q24, 13q14, 15q22, and 21p11.8 This breakpoint clustering is probably not the result of ascertainment bias, but suggests the existence of chromosomal “hot spots” involved in complex rearrangements. The molecular mechanism(s) underlying frequent breakage remains to be elucidated.
A significant proportion of disease associated chromosome breakpoints involve submicroscopic deletions of several megabases of DNA.4 25 Considering the complex nature of the CCR described, excluding microdeletions in one or several of the breakpoint regions involved by high resolution FISH mapping alone will be a very time consuming or even impossible task. Only miniaturisation and automation of techniques will allow one to screen the entire genome in a reasonably short time for cytogenetically cryptic (unbalanced) rearrangements. However, with the development of CGH of patient and reference DNAs on microarray DNA chips,29 30 this goal seems to be within reach and a new era in the analysis of chromosome disorders has begun.
The authors thank S Freier, A Gerlach, C Menzel, and K Stout for excellent technical assistance and M A Ferguson-Smith and J M Trent for providing DNA libraries. This work was supported by the German Genome Programme (grant No 4763) and the EU Commission (BMH4-CT97-2268).
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.