Segmental Maternal UPD6 with Prenatal Growth Restriction

Abstract

We report a child with segmental maternal uniparental isodisomy of chromosome 6, involving most of the long arm distal to 6q16, detected by SNP microarray. Clinical features include prenatal growth restriction, global developmental delay, and severe gastro-esophageal reflux disease. Maternal uniparental disomy (UPD) of chromosome 6 has previously been reported to cause intrauterine growth restriction. Paternal UPD of this chromosome is well known to cause transient neonatal diabetes mellitus. We discuss reported cases of maternal UPD of chromosome 6 and consider whether our patient's features may be due to disordered imprinting or unmasking of an autosomal recessive condition.

Uniparental disomy (UPD) was first recognized by Engel [1980], and in the era of SNP microarrays its detection rate has increased [Papenhausen et al., 2011; Schaaf et al., 2011; Schwartz, 2011].

In uniparental heterodisomy (heteroUPD), both alleles are inherited from 1 parent. This can result in disordered imprinting, and indeed paternal UPD of chromosome 6 is well-recognized to cause transient neonatal diabetes mellitus and macroglossia, due to aberrant imprinting of the PLAGL1 and HYMAI genes [Milenkovic et al., 2006; Mackay and Temple, 2010]. The significance of maternal UPD on imprinting is less certain.

In uniparental isodisomy (isoUPD), a single parental allele is duplicated. As well as disordered imprinting, this may result in a child inheriting a homozygous mutation from a heterozygous parent, and thus being affected by a recessive condition of which only 1 parent is a carrier. At least 4 recessive conditions have been reported as a result of isoUPD of chromosome 6: 3M syndrome [Sasaki et al., 2011], C4 deficiency [Welch et al., 1990], molybdenum cofactor deficiency [Gümüş et al., 2010] and congenital adrenal hyperplasia [Spiro et al., 1999; Parker et al., 2006].

isoUPD may be segmental or it may involve the whole chromosome. Segmental UPD means that part of the chromosome is inherited biparentally, and another part is inherited uniparentally. It is the result of a postzygotic, mitotic event [Kotzot, 2008]. Possible mechanisms include homologous chromatid exchange or repair of a partial monosomy [Kearney et al., 2011]. In either case, the homozygosity usually includes the telomere [Papenhausen et al., 2011]. Being a postzygotic event, it results in isoUPD and may cause a recessive condition to be unmasked [Fernández-Rebollo et al., 2010; Pérez et al., 2012].

Materials and Methods

Clinical Report

The proband was the 3rd child of 4 to healthy, unrelated Caucasian parents. Her older brother had pervasive developmental disorder not otherwise specified; her other siblings were healthy. Her mother reported reduced fetal movements during the pregnancy. Though the midtrimester morphology scan was normal, later in the pregnancy the fetus was found to be small for gestational age. She was delivered at 35 weeks of gestation by elective cesarean section with a birth weight of 1.20 kg (0.7 kg < 10th percentile). Apgar scores were 9 and 9 at 1 and 5 min.

She had severe dysphagia and required nasogastric and then gastrostomy feeding. She was treated for gastro-esophageal reflux disease and had a fundoplication at the age of 1 year, after which proton pump inhibitor therapy was successfully ceased. There was also intolerance to cow's milk protein, with an alternative formula used.

Global developmental delay became apparent though there was continued developmental progress. At the age of 18 months she could roll over and sit up, and she began walking at the age of 23 months. At the age of 22 months she had a vocabulary of 4 words; by the age of 35 months this had increased to 10–15 words.

She remained small for the first 2 years of life, but her growth improved thereafter. At 22 months of age her height was 75 cm (2 cm < 1st percentile), weight 9.0 kg (0.7 kg < 3rd percentile) and head circumference 46 cm (25th percentile). By the age of 35 months her height and weight measured on the 10th percentile and head circumference remained on the 25th percentile.

She was not facially dysmorphic (fig. 1). She had not had any disturbance of blood glucose control, and there was no macroglossia. Clinical examination of her cardiovascular, respiratory, gastrointestinal, and neurological systems was normal.

Fig. 1.

The proband at the age of 17 (left) and 22 (center, right) months.

Placental Histopathology

The placenta was examined macro- and microscopically.

Cytogenetic Studies

A standard G-banded karyotype at a resolution of 550 bands was performed, as well as FISH testing for deletion/duplication of the SHOX locus at Xp22.

Microarray Analysis

An Affymetrix whole genome 2.7M array was performed using the standard Affymetrix protocol on DNA extracted from peripheral blood. Array results were visualized using the Affymetrix Chromosome Analysis Suite (ChAS v 1.01) software and reference sequence NCBIv36/hg18.

UPD testing was performed using 13 microsatellite markers spanning chromosome 6 in the proband, her mother and her father (trio analysis).

Results

Placental Histopathology

The placenta was small, weighing 225 g and measuring 20 mm in thickness. Its size was on the 3rd percentile for gestational age. It contained patchy areas of abnormal villous maturation.

Cytogenetic Studies

The cytogenetics showed a female karyotype (46,XX) with no abnormality detected at the 550 band-level. SHOX FISH showed 2 normal signals.

Microarray Analysis

On SNP array testing, a long contiguous stretch of homozygosity (LCSH) was identified on the long arm of chromosome 6, from 6q16.1 to 6qter (fig. 2). No significant copy number changes (deletions or duplications) were found across the genome at a resolution of 100 kb, including at the crossover point of the biparental and uniparental regions on chromosome 6.

Fig. 2.

SNP data displayed in ChAS showing the long contiguous stretch of homozygosity (LCSH) from 6q16.1 to 6qter.

On UPD testing, markers D6S1610 (6p21.1) and D6S257 (6p12.1) indicated biparental inheritance at these loci. Markers D6S308 (6q24.1), D6S1581 (6p25.3) and D6S446 (6q27) indicated maternal isoUPD. Markers D6S1574 (6p25.1), D6S289 (6p22.3) and D6S460 (6q14.1) were uninformative.

These results confirmed maternal isoUPD of at least 6q24.1q27 and biparental inheritance of at least 6p21.1p12.1. No markers were available for testing in the region between 6q16.1 and 6q24.1, or distal to 6q27. The SNP array result showed homozygosity distal to 6q16.1. These results suggest that biparental inheritance spans 6pterq16.1, with maternal isoUPD from 6q16.1 to 6qter. This is consistent with a mitotic recombination event with homologous chromatid exchange at 6q16.1 during somatic development. It is most likely, therefore, that the LCSH detected by the array corresponds with the region of maternal isoUPD (i.e. 6q16.1qter).

Discussion

This case provides further evidence that intrauterine growth restriction (IUGR) is a feature of maternal UPD of chromosome 6. There are relatively few reported cases of maternal UPD6, and it is still unclear whether growth restriction is due to disordered imprinting or to an unmasked recessive condition.

Of the 8 cases of maternal UPD6 previously reported [Sasaki et al., 2011], 5 had maternal isoUPD6, 2 had maternal heteroUPD6 and 1 had areas of both. Interestingly, only the patients with maternal isoUPD6 had IUGR (with 5 of 6 affected). This led the reviewing authors to conclude that IUGR in these patients may be due to homozygosity, rather than disordered imprinting. However, there is only partial overlap between the areas of homozygosity in the reported cases of maternal isoUPD6. Four of the above 6 cases had maternal isoUPD6 of the entire chromosome. One [Sasaki et al., 2011] had maternal isoUPD6 in 2 regions (1 on each arm of chromosome 6, including 6q25.1q27). The other [Gümüş et al., 2010] had maternal isoUPD6 affecting part of the short arm, as well as UPD6 of 6q13q22.31 (the parent of origin of this region was not investigated by the authors). Notably, this patient did not have IUGR. Our patient has segmental maternal isoUPD6 of 6q16.1qter. If an autosomal recessive condition is responsible for IUGR, it is presumably in an isodisomic region shared by all growth-restricted individuals. One patient [Sasaki et al., 2011] had an identified recessive condition known to cause IUGR (3M syndrome). Disregarding this patient, the common isodisomic region shared by growth-restricted individuals is the region affected in our patient (6q16.1qter).

However, it is not necessarily true that homozygosity leads to recessive disease. Stretches of homozygosity with biparental inheritance are common in the normal population (without IUGR). This is due to identical by descent inheritance. Thus, 9.5% of the outbred population may carry a >5-Mb LCSH [Simon-Sanchez et al., 2007], and an increased rate of IUGR has not been reported.

Alternatively, growth restriction in maternal UPD6 may be due to disordered imprinting. Paternal UPD of chromosome 6 is known to cause IUGR by overexpression of the maternally imprinted PLAGL1 and HYMAI genes [Mackay and Temple, 2010]. These genes are at 6q24.2. A dosage-sensitive effect is also possible; perhaps loss of expression of PLAGL1 in maternal UPD6 may also have negative effects on intrauterine growth. This hypothesis is supported by IUGR in 5 reported patients with maternal UPD including the PLAGL1/HYMAI locus, and the absence of UPD beyond 6q22.31 in 1 patient with normal birth weight [Gümüş et al., 2010]. It is also supported by an animal study which showed that mice with a knockout of the paternally inherited Zac1 allele had a birth weight 23% lower than wild-type mice (PLAGL1 is the human ortholog to Zac1) [Varrault et al., 2006]. However, normal growth has been reported in at least 1 post-term human individual with maternal UPD of the whole of chromosome 6 [Salahshourifar et al., 2010].

There is also growing evidence that other loci on chromosome 6 may be imprinted. In 1 study, methylation tests were performed on 79 patients with IUGR or unexplained short stature [Turner et al., 2010]. Gain of methylation was found at the IGF2R locus (at 6q25.3) in 7/79 patients. This was significantly more common than in the control, non-growth-restricted group. Whether these patients had UPD of chromosome 6 is not reported.

As well as IUGR, our patient had significant developmental delay. Though she has not been identified as having an autosomal recessive condition, her isodisomic DNA contains 95 OMIM-listed genes resulting in a known phenotype. Mutations in at least 3 of these (GRIK2, MED23 and MRT28) have been reported to cause autosomal recessive mental retardation. It is interesting to note that another patient with maternal UPD6 also has unexplained intellectual impairment [Sasaki et al., 2011]. Further analysis of these genes remains an option in the future.