Introduction

Fragile X syndrome (FXS), caused by loss-of-function of the FMR1 gene, has been recognized as the most frequent cause of X-linked mental retardation and autism spectrum disorders.1, 2, 3, 4, 5 FXS is caused by a full mutation, an expansion of CGG trinucleotide repeats (>200 CGG repeats) in the 5'UTR of the FMR1 gene. Individuals harboring an allele with 55–200 CGG repeats are premutation carriers. The FMR1 full mutation causes a broad spectrum of symptoms, including intellectual disabilities, autism spectrum disorders, social anxiety, hyperarousal, inattention, impulsivity and hyperactivity.6, 7 Few cases presenting a clinical phenotype typical of fragile X have been reported to be caused by relative gross duplication in the FMR1 gene.8 Microduplications involving the FMR1 gene alone have not been reported. A prior report in the literature described a family with a 5.1 Mb duplication including both the FMR1 locus and FMR2 and 26 other genes, leading to a heritable syndrome of intellectual disability and short stature with hypogonadism8. Members of this family had a phenotype consisting of mild developmental delay with relative preservation of verbal skills, microcephaly and small hands, feet and testicles. These authors proposed that a functional disomy of FMR1 may contribute to the observed clinical phenotype.

Here, we report the clinical and molecular characterization of a case of a 4-year-old boy with a distinctive phenotype, who has an 86 kb de novo microduplication on Xq27.3 including only the FMR1 gene, the FMR1 antisense RNA gene (AS-FMR1), as well as two paternally inherited genomic alterations: a 363-kb duplication on 1q44 and a 168-kb deletion on 4p15.31.

Clinical description

The patient first came to medical attention at around 18 months of age when he was diagnosed with myoclonic seizures, although in retrospect he had fine motor and speech delay since at least 9 months of age. The myoclonic seizures were characterized by episodes of muscle twitching that led to falls and focal activity in both frontal lobes on video EEG. The patient’s clinical symptoms improved after treatment with levetiracetam. A metabolic workup, including measurement of blood lactate, pyruvate and plasma amino acids levels, was normal. An MRI scan of the brain was also normal.

At around 2 years of age, the patient developed behavioral issues, characterized by aggressiveness toward others. He also had breath-holding spells. His neurologist noticed language delay, more expressive than receptive, suggestive of possible oral apraxia. Speech therapy was initiated, and the patient showed considerable improvement. The patient had a follow-up EEG at 3½ years of age, which showed persistent focal seizure activity despite the absence of overt clinical seizure episodes. His speech delay, although milder, persisted. It also became apparent that he was having difficulty attaining new fine motor skills, such as drawing age-appropriate geometric figures.

At 4 years of age, the patient developed absence seizures. These were described as ‘staring spells’ that were increasing in frequency over 2 months. A new video EEG confirmed the presence of absence seizures, with classical generalized 3 Hz spike-and-wave appearance. He had persistent speech and fine motor delay. He also had signs and symptoms suggestive of hyperactivity. A clinical geneticist was consulted at that time.

On physical examination, the patient’s growth parameters were within the normal range. The patient was hyperactive, running around constantly in his hospital room, and was unable to focus on a single task for more than a few seconds. He had bilateral fifth-finger clinodactyly. His strength, sensation and coordination were normal. The patient spoke very few words, and these were difficult to understand, but he had good speech comprehension. The patient was unable to draw a stick figure or a circle.

Molecular findings

Molecular testing for fragile X syndrome and an oligonucleotide CGH microarray were carried out to investigate possible underlying genetic mechanisms. Southern blot and PCR analysis for fragile X syndrome were carried out as previously described.9, 10 The Southern blot displayed a normal male pattern and PCR analysis showed the presence of a 29 CGG repeat allele. Oligonucleotide array CGH (aCGH) was performed using a custom-designed NimbleGen 135K array developed by Signature Genomics (Spokane, WA, USA) (SignatureChipOS, version 2). The probes on the array have an average spacing of one probe every 35 kb throughout the genome and one probe every 10 kb in regions with known clinical significance. The aCGH revealed an approximately 86-kb copy number gain in chromosome Xq27.3 between nucleotides 146 783 933 and 146 868 568, which included 11 consecutive markers that were duplicated, with adjacent probe gaps of 9.89 and 2.27 kb (nucleotide coordinates were based on the UCSC Genome Browser (hg18, March 2006, NCBI build 36.1) (Figure 1).

Figure 1
figure 1

Microduplicated region in Xq27.3, according to the USCS Genome Broswer (hg18, March 2006, NCBI build 36.). FMR1 and ASFMR1 are the only genes in this region.

The duplication only encompasses the FMR1 and the FMR1 antisense RNA gene (AS-FMR1). In addition to the above duplicated region, an approximately 363.16-kb copy number gain in the 1q44 region (nucleotides 245 377 586–245 556 179) and an approximately 168.86-kb copy number loss in the 4p15.31 region (nucleotides 22 897 572–23 066 434) were also detected.

To further characterize the copy number changes found in this patient, parental studies were carried out. The copy number changes in the 1q44 and 4p15 regions were detected in the patient’s father, who does not have mental retardation, seizures or speech delay. The copy number gain in Xq27.3 was determined to be a de novo event, as it was not detected in the mother.

FMR1 mRNA expression levels were measured in peripheral blood leukocytes by quantitative RT-PCR as previously described,11 and were found to be within the normal range (1.09± 0.16).

Discussion

Mutations in the FMR1 gene have been associated with various phenotypes. These include fragile X syndrome, in which loss-of-function in FMR112 leads to intellectual disability.13 The most common mechanism leading to fragile X syndrome is a CGG trinucleotide repeat expansion in the promoter region of the FMR1 gene leading to methylation of the promoter of the gene, with subsequent silencing of transcription and the absence of the encoded product FMRP,14 which is important for synaptic plasticity and synaptic protein synthesis.15, 16 However, other loss-of-function mutations, including deletions, missense mutations and splice-site mutations, have been reported.17, 18, 19, 20

The case described here has an 86-kb microduplication that only encompasses the FMR1 and ASFMR1 genes. The ASFMR1 gene promoter is located within intron 2 of the FMR1 gene.21 Interestingly, the expression of ASFMR1, similar to that of FMR1, is elevated in fragile X premutation carriers and decreased in individuals with full mutation alleles.21 To date, translation of the ASFMR1 gene has not been reported, and whether it has a pathogenic role is not known.

Parental microarray analysis indicates that the microduplication in the FMR1 gene region in our patient is a de novo event. The gain in the 1q44 and the loss in the 4p15 regions were both inherited from a healthy father. Both copy number gains and losses of a larger chromosome 1q44 region encompassing genomic coordinates 245 141 053 to 246 395 458 had been reported in normal controls.22 The 1q44 region contains three genes (ZNF124, ZNF496 and VN1R5). ZNF124 is a transcription factor that inhibits apoptotic death in hematopoietic and leukemia cell lines23 and has been identified as part of large pathogenic deletions in 1q44,24 although patients who have a normal copy number of ZNF124 may have the same phenotype as those with a deletion. ZNF496 is a transcription cofactor of NSD1. Children with Sotos syndrome have mutations in NSD1, some of which prevent it from interacting with ZNF496.25 A third gene in our patient’s 1q44 gain, VN1R5, is thought to be a pseudogene.26 None of these genes have been reported as causative of human disease. The 4p15 loss has been reported as a normal copy number variant in the Database of Genomic Variants (http://projects.tcag.ca/variation). There are no genes in this region.

A large 5.1-Mb interstitial duplication in the Xq27.3q28 region containing 28 genes, including FMR1 and FMR2, has been reported in a French family.8 Three adult men in this family had a syndrome consisting of short stature, microcephaly, small hands and feet, low testicular volume, as well as facial features including deep-set eyes, bulbous nasal tip and thin lips. The individuals had a history of intrauterine growth restriction, failure to thrive in infancy, undescended testes requiring orchidopexy, walking after the age of 16 months and mild learning difficulties, with preserved verbal skills. The expression level of FMR1 was not measured in the French family; it is unclear whether the phenotype described was due to increased FMR1 transcript levels. X-inactivation studies suggested that female carriers preferentially inactivated the X chromosome with the interstitial duplication.

Our patient’s phenotype differs from the one described in the French family. The patient does not have any distinctive facial features; he did not have cryptorchidism, small testicular volume, nor small hands or feet. The larger duplication observed in these cases suggests that these distinctive traits identified in the French family may be caused by duplication of genes other than FMR1. Our patient has a relative milder phenotype, but he has additional features such as speech delay, fine motor delay and hyperactivity, which were not documented in the French family. However, it is difficult to compare these traits in a 4-year-old with those of the adults in the French case series.

Around 13–16% of individuals with fragile X develop seizures during their lifetime,27 which suggests a role for the FMR1 gene in epilepsy. Decreased inhibitory signaling from GABA-A receptors28 and overactivity of the metabotropic glutamate receptor (mGluR) with resulting altered AMPA receptor activity29 have been reported as pathogenic mechanisms for increased risk of seizures in fragile X syndrome. Animal models of fragile X syndrome with heterozygous knockout of the mGluR5 receptor suggest phenotypic rescue of many of the features of fragile X, including increased risk for seizures.30 These studies demonstrate a clear role for the FMR1 gene in seizures. However, whether the observed increased gene dosage of the FMR1 could also lead to the seizure activity seen in our patient is not clear.

Although the lack of FMRP causes fragile X syndrome, an increase in FMR1 mRNA, leading to RNA toxicity, results in disease phenotypes such as fragile X-associated tremor/ataxia syndrome (FXTAS)31 and premature ovarian failure.32 This may indicate the dosage-sensitive character of the FMR1 gene. However, we cannot state that excess mRNA is the cause of our patient’s phenotype, as the levels of FMR1 expression were found to be normal. This does not rule out a pathogenic microduplication, especially if it is a de novo event in which the only gene involved is known to cause disease. In addition, blood mRNA expression levels do not necessarily reflect those in the brain. CGG-binding protein sequestration has been reported to depend on the number of CGG repeats and may occur despite normal mRNA levels.33 The level of sequestration may be insufficient to cause a disease but may modulate the penetrance of a phenotype due to another genetic alteration.

It is possible that the microduplication represents a second hit that triggers the phenotype in the presence of another genomic alteration, such as the duplication on chromosome 1 or the deletion on chromosome 4 detected in this case. Indeed, copy number variation burden has been reported to positively correlate with the severity of childhood disability.34

Finally, the patient we report may have other genetic or environmental exposures to neurotoxicants that could contribute to his phenotype. The patient in the present report has a phenotype with features such as speech delay or hyperactivity, which change with time, may be age dependent and are difficult to accurately measure. Despite these limitations, we believe that it is important to share this case report with the medical genetics community.

Conclusion

A microduplication in Xq27.3 encompassing exclusively the FMR1 and ASFMR1 genes may be causative of motor and speech delay, focal seizures, absence seizures and hyperactivity, which are clinical traits frequently observed in individuals with an altered function of the FMR1 gene. Other genetic mechanisms may underlie the clinical phenotype presented here.