Statistics from Altmetric.com
The existence of parent of origin differences in the expression of some genes, a process known as genomic imprinting, has been recognised and documented over the past several years. This epigenetic marking process results in the differential expression of normal genes, depending on whether they are of maternal or paternal origin. A number of human disorders have been identified as resulting from alterations in genomic imprinting.1 One type of genetic abnormality which can unmask genomic imprinting is uniparental disomy (UPD), in which both chromosomes of one pair are inherited from one parent, with no contribution from the other.2 Distinct phenotypes exhibited by patients with maternal and paternal UPD(14) strongly suggest that at least some genes on human chromosome 14 are subject to imprinting effects.3-6 This finding is supported by studies in the mouse indicating that the distal portion of chromosome 12, recognised as a candidate imprinted region, is syntenic with human chromosome 14q.7 Nevertheless, no imprinted genes have yet been identified on chromosome 14, and the location of the imprinted region on human chromosome 14 remains unclear.
Analysis of parental origin effects in human trisomy for chromosome 14q8 or monosomy for the same chromosome9makes the region 14q23-q32 a candidate region for containing imprinted genes.10 Moreover, the observation of a partial duplication of 14q in a developmentally delayed girl with minor abnormalities and her phenotypically normal father led Robinet al 11 to propose that the 14q24.3-q31 region may be imprinted. As genotype-phenotype correlations in patients with aneusomy for this region may help the identification of imprinted genes on human chromosome 14, we report here a direct 14q31 duplication observed in a child with mild developmental delay and his phenotypically normal mother. This duplication was also detected in five relatives with a normal phenotype and maternal inheritance through three generations.
Our patient, a 2 month old boy, was referred following a cyanotic episode. He was the first child of healthy, unrelated parents. The family history was uneventful and he was born by caesarean section. Birth weight, length, and head circumference were 3515 g, 54 cm, and 33 cm, respectively. At the age of 2 months, clinical examination showed dysmorphic features (macrosomia, metopic suture, neck hyperextension without opisthotonos, large nose, prominent philtrum, gingival hyperplasia, and bilateral palmar simian crease) and axial hypotonia with moderate distal hypertonia. EEG and MRI scan of the brain, medullary ultrasound, abdominal ultrasound, pulmonary and skeletalx ray, ECG and retinal examination were all normal. The patient walked at 20 months. By the age of 27 months, clinical evaluation of the patient showed very slight dysmorphism, slight overgrowth (+2 SD), moderate psychomotor delay with specific speech delay, hypotonia, and behavioural and sleep disturbance.
Routine R and G banding techniques, performed on metaphases from PHA stimulated lymphocytes of the proband, showed an elongated chromosome 14 with an extra G positive band on the long arm (fig 1). High resolution banding suggested that the patient had a duplication of band 14q31. The same 14q duplication was found in the proband's mother and four other members of his family, all of whom had a normal phenotype (fig 2A).
Fluorescence in situ hybridisation (FISH) was performed with various probes, in order to characterise the rearrangement. A chromosome 14 specific paint, used as recommended by the manufacturer (ONCOR), confirmed that the extra material originated from chromosome 14 (data not shown). FISH experiments using YAC probes estimated to cover the 14q31 band were performed, in order precisely to define the duplicated chromosomal region in the proband. The YAC clones 813e6, 930c12, 749g4, 815d5, 872e3, 756a4, 945d2, 856g8, and 965b9, obtained from the CEPH (Centre d'Etude du Polymorphisme Humain), were labelled with biotin-14dCTP or digoxigenin-11dUTP by random priming (Bioprime DNA labelling system, Life Technologies and Dig high prime, Roche Diagnostics), then hybridised on metaphase chromosomes or interphase nuclei. The hybridised biotin signals were made visible with fluorescein labelled avidin, and the digoxigenin signals were visualised with rhodamine labelled anti-digoxigenin antibody following standard protocols.12 The results are summarised in fig2B. The more proximal YACs (813e6 and 930c12), as well as the more distal ones (945d2, 856g8, and 965b9), gave signals of equal intensity on the normal and the duplicated chromosomes 14, indicating the presence of only two copies in the patient's genome. In contrast, YACs 749g4, 815d5, 872e3, and 756a4 showed two signals of different sizes, the larger one being located on the duplicated chromosome 14. In order to confirm these results, two YACs with different FISH patterns were cohybridised on the same metaphase preparations: YAC 815d5 (green signal) presumed to be included in the duplication and YAC 965b9 (red signal) presumed to map outside the duplication. As expected, the green signal (815d5) on the duplicated chromosome 14 was larger than the red signal (965b9) on the same chromosome (fig 3A, B). The same experiment was performed on interphase nuclei, where the undercondensation of the chromatin allows increased resolution. This allowed the detection of two red signals (YAC 965b9) and three green signals (YAC 815d5), clearly showing a duplication of the region covered by YAC 815d5 (fig3C). Different YAC probes were hybridised with metaphase spreads from subjects III.2 and III.3, confirming the presence of the same 14q duplication in the proband and his relatives.
In order to determine whether the duplication is direct or inverted, the proximal and distal duplicated YACs (749g4 and 756a4 respectively) were labelled differently, hybridised to interphase nuclei, and visualised with specific fluorochromes. Alternate green and red signals observed in nuclei show that the two copies of the duplication are organised in direct orientation on chromosome 14 (fig4).
In order to gain insight into the mechanism that gave rise to this duplication, we used microsatellite analysis to determine whether the duplication originated from one or two homologous chromosomes 14. Genomic DNA from the proband and other family members was extracted from whole blood samples using standard methods. DNA polymorphisms were analysed, as previously described,13 using polymerase chain reaction (PCR) amplification of short sequence repeats from the following markers: D14S68, D14S67, D14S1052, and D14S1033 (Genethon). For three informative markers located in the duplicated region (D14S68, D14S67, and D14S1052), only two different alleles were observed in the patient (IV.1), his mother (III.2), and his uncle (III.3) (fig 5). One of these alleles is shared by these three members of the family and probably represents the abnormal chromosome 14. This result suggests a monochromosomic origin of the duplicated segment, probably resulting from unequal sister chromatid exchange.
Finally, on the basis of the genetic map of chromosome 14,14 we were able to estimate the approximate length of the duplicated segment to be between 3 and 9 cM. These data are in agreement with our cytogenetic observations, as the duplication corresponds approximately to cytogenetic band 14q31, whose physical size is estimated to be 8.8 megabases.15 16
In this study, we describe a phenotypically abnormal patient carrying a duplication of the 14q31 band, which occurs in the potentially imprinted region on this chromosome.8 9 10 11 The duplication, which was also found in five other family members, was maternally inherited through three generations. Apart from the proband (IV.1), who has minor dysmorphic features and psychomotor delay associated with behavioural and sleep disturbance, all family members carrying the abnormal chromosome are phenotypically normal. This suggests that the abnormal phenotype of the proband is probably unrelated to the duplicated segment. Moreover, duplication of the 14q31 band, when maternally inherited, does not seem to be associated with an abnormal phenotype.
It has been suggested by Robin et al 11 that a similarly duplicated chromosome 14 (duplication 14q24.3-q31), if paternally transmitted, does induce an abnormal phenotype. This parent of origin dependent phenotypic disparity associated with similar duplications suggests a possible imprinting effect. The fact, however, that the duplications are not identical allows three different hypotheses to be formulated with respect to imprinting effects.
Firstly, the imprinted genes could be located outside the 14q24.3-q31 interval. This region may contain genes that do not have a deleterious effect if overexpressed, and duplication will therefore produce no phenotypic effect. In that case, the association of the 14q duplication with an abnormal phenotype, both in the proband described here and in that described by Robin et al,11 would be merely coincidental.
Secondly, the imprinted genes could be located in band 14q24.3 and not in band 14q31. In that case they would be duplicated in the patient described by Robin et al 11 but not in the family described here. In such a situation, no parent of origin phenotypic effect would be associated with a 14q31 duplication. Then, a 14q31 duplication of paternal origin would be associated with a normal phenotype, as we show to be the case here for a 14q31 duplication of maternal origin. This hypothesis is supported by the fact that G positive bands, such as the 14q31 band, contain very few genes.17
Thirdly, imprinted genes could be located in band 14q31 and duplicated both in the family reported here and in that of Robinet al. 11 If this were the case, both observations are in agreement and suggest that imprinted genes from this region are paternally expressed. However, the presence of paternally expressed genes only is not consistent with the observation that the maternal UPD(14) phenotype is moderate compared to the paternal UPD(14) phenotype. This would suggest that maternally expressed genes are located elsewhere in the 14q23-q32 region. Unfortunately, it is not possible to choose between these three hypotheses because no case of paternally inherited 14q31 duplication has yet been described.
So far, only a few genes have been mapped to the human 14q31 band (NCBIhttp://www.ncbi.nlm.nih.gov), and none of them has been shown to be imprinted. Those that have been mapped include the genes encoding galactosylceramidase (GALC), the thyroid stimulating hormone receptor (TSHR), the suppressor of lin-12 (C elegans)-like (SEL1L), neurexin III (NRXN3), and a member of the G protein coupled receptor superfamily (GPR68/OGR1).
Further studies of chromosome 14 aneusomy are needed in order to define the imprinted region better and identify the corresponding genes.
We would like to thank the family for its participation, Dr A Nivelon-Chevalier and Dr C Robinet for their kind collaboration in the clinical study of the patient, and Dr M Mitchell for reviewing the English language. This work was supported by INSERM and the Association pour la Recherche contre le Cancer (ARC).