Characterization of 22q12 Microdeletions Causing Position Effect in Rare NF2 Patients with Complex Phenotypes

Abstract

Neurofibromatosis type 2 is an autosomal dominant tumor-prone disorder mainly caused by NF2 point mutations or intragenic deletions. Few individuals with a complex phenotype and 22q12 microdeletions have been described. The 22q12 microdeletions' pathogenic effects at the genetic and epigenetic levels are currently unknown. We here report on 22q12 microdeletions' characterization in three NF2 patients with different phenotype complexities. A possible effect of the position was investigated by in silico analysis of 22q12 topologically associated domains (TADs) and regulatory elements, and by expression analysis of 12 genes flanking patients' deletions. A 147 Kb microdeletion was identified in the patient with the mildest phenotype, while two large deletions of 561 Kb and 1.8 Mb were found in the other two patients, showing a more severe symptomatology. The last two patients displayed intellectual disability, possibly related to AP1B1 gene deletion. The microdeletions change from one to five TADs, and the 22q12 chromatin regulatory landscape, according to the altered expression levels of four deletion-flanking genes, including PIK3IP1, are likely associated with an early ischemic event occurring in the patient with the largest deletion. Our results suggest that the identification of the deletion extent can provide prognostic markers, predictive of NF2 phenotypes, and potential therapeutic targets, thus overall improving patient management.

Keywords: 22q12 microdeletion; NF2; genotype-phenotype correlation; haploinsufficiency; non-allelic homologous recombination; position effect.

Figure 1

Graphical representation of the 22q12 deleted region in patients 246, 366, and 160. The dotted line in light gray represents the deleted chromosomal region, while the dark line represents the centromeric (cen) and telomeric (tel) non-deleted regions. The filled rectangles represent the genes, colored light gray if deleted, dark gray if not deleted, and shaded if partially deleted, based on the results (shown in the empty rectangles) obtained from the qPCR-gDNA and MLPA assays for patient 246, and by aCGH for patients 366 and 160.

Figure 2

Genomic characterization of the deletion junction fragments. (A) Sequence of the deletion junction fragment of patient 246: the proximal BP falls into the L1ME3 repeated element in THOC5 intron 7, the distal BP falls into the MLT1A0 repeated motif in NF2 intron 14, and eight base pairs are of uncertain origin, being shared between the two regions. (B) Sequence of the deletion junction fragment of patient 366: the proximal BP is in an intergenic region, upstream of EMID1, belonging to the Tigger2b_Pri repeat, the distal BP falls into the AluSx repeated motif in the ZMAT5 intron 1, and thirty-three base pairs are of uncertain origin. (C) Sequence of the deletion junction fragment of patient 160: the proximal BP is in an intergenic region, within the L1MB7 repeated element upstream of ZNRF3, while the distal BP falls into the L2b repeated motif in OSBP2 intron 2, while fifteen base pairs are of uncertain origin.

Figure 3

Expression level of the AP1B1 gene. The AP1B1 quantitative expression level (2^−ΔCt) in the peripheral blood of patients 366 and 160 has been compared to the average value of ten healthy controls (WT) and three patients with NF2 intragenic mutations (NF2im). The AP1B1 mRNA was expressed at approximately half in our patients.

Figure 4

Gene expression analysis of the selected genes flanking the patients’ deletions. (A) The histogram shows quantitative expression levels (2^−ΔCt) in the peripheral blood of 4 centromeric and 8 telomeric genes to the patients’ deletions, in our 3 NF2 microdeletion patients compared to the mean of 10 wild-type subjects (WT) and to the mean of 3 NF2 patients with intragenic mutations (NF2im). For comparison, the mean WT expression level was set as equal to 1. Among the selected genes, AP1P1, located centromerically to patient 246’s deletion, was deleted in patients 366 and 160, and between telomeric genes, ZMAT5 was removed from the deletions of patients 366 and 160, while UQCR10, ASCC2, and MTMR3 were included in patient 160’s deletion. A trend towards dysregulation was found for LIMK2 in patient 246, MTMR3 in patients 246 and 366, LIMK2 in patients 246 and 160, and PIK3IP1 in patient 160 only, by comparison with both wild-type controls and NF2 patients with intragenic mutations. (B) The histogram shows the quantitative expression levels, in peripheral blood, of the previously identified dysregulated genes. For each patient, the value is mean ± standard deviation (SD) from three independent biological samples, while for the ten healthy controls and three NF2 intragenic mutation patients, the values are means ± SD. * BH-adjusted p < 0.05, Student’s t-test.

Figure 5

Visualization of topological domains mapping within the 22q12 genomic region. The heatmap, obtained from Hi-C data from blood-derived GM12878 lymphoblastoid cells, shows 22q12 DNA–DNA interactions and inherent TADs, whose positions along the chromosome are indicated by the yellow and cerulean blue bars. The barcode in brown represents the DNase I hypersensitive sites (DHSs). The UCSC genome browser screen shows the genes (in blue and black; those whose expression level has been evaluated by qPCR are highlighted in light blue), VISTA enhancers (in green), and CTCF binding sites (in red). The dashed lines correspond to the patients’ deletion breakpoints. Patient 246’s deletion, encompassing a single TAD, removed three genes, but no regulatory elements. Patient 366’s deletion partially involved two TADs, including 13 protein-coding genes and four CTCF-binding sites. Patient 160’s deletion, encompassing five TADs, included 38 protein-coding genes and 18 CTCF binding sites. Two enhancers were mapped outside the patients’ deletions.