Custom oligonucleotide array-based CGH: a reliable diagnostic tool for detection of exonic copy-number changes in multiple targeted genes

Abstract

The frequency of disease-related large rearrangements (referred to as copy-number mutations, CNMs) varies among genes, and search for these mutations has an important place in diagnostic strategies. In recent years, CGH method using custom-designed high-density oligonucleotide-based arrays allowed the development of a powerful tool for detection of alterations at the level of exons and made it possible to provide flexibility through the possibility of modeling chips. The aim of our study was to test custom-designed oligonucleotide CGH array in a diagnostic laboratory setting that analyses several genes involved in various genetic diseases, and to compare it with conventional strategies. To this end, we designed a 12-plex CGH array (135k; 135 000 probes/subarray) (Roche Nimblegen) with exonic and intronic oligonucleotide probes covering 26 genes routinely analyzed in the laboratory. We tested control samples with known CNMs and patients for whom genetic causes underlying their disorders were unknown. The contribution of this technique is undeniable. Indeed, it appeared reproducible, reliable and sensitive enough to detect heterozygous single-exon deletions or duplications, complex rearrangements and somatic mosaicism. In addition, it improves reliability of CNM detection and allows determination of boundaries precisely enough to direct targeted sequencing of breakpoints. All of these points, associated with the possibility of a simultaneous analysis of several genes and scalability 'homemade' make it a valuable tool as a new diagnostic approach of CNMs.

Figure 1

Detection by genomic CGH array of small CNMs and CNVs in different genes. (a) Detection of a hemizygous duplication of 2.2 kb encompassing exon 7 of the DMD gene and a duplication of 1.4 kb in intron 2 (patient 14). (b) Detection of a heterozygous 1.4 kb deletion carrying away exons 25 and 26 of the CFTR gene. The horizontal axis shows the position along the genome (NCBI36; Hg18) and the vertical axis the Cy3:Cy5 log₂ ratios. Patient sample was fluorescently labeled using Cy3 and control sample using Cy5. Control was sex matched with patient. The arrows indicate the location of the copy-number change.

Figure 2

Heterozygous deletion of exon 1 of CDKL5 and neighboring genes (patient 27) is identified more reliably and accurately by CGH array compared with MLPA. (a) MLPA showing a reduction of the peak corresponding to CDKL5 exon 1 among CDKL5 exonic probe and control probe peaks. (b) Detection by genomic CGH array of a 294 kb deletion on chromosome X from intron 1 of the CDKL5 gene to the SCML2 and CXorf20 neighboring genes.

Figure 3

Detection of a somatic mosaicism corresponding to a duplication of exons 61 and 62 of the DMD gene in a carrier female (patient 16). (a) Real-time quantitative PCR of exons 61 and 62 performed in genomic DNA from blood of patient 16 (P), suggested that 50% of cells are heterozygous for the exons 61 and 62 duplication. DupC designates a control female heterozygous in all cells for duplication of exons 61 and 62 of the DMD gene. N indicates a normal control female. The vertical axis shows the relative quantification of the tested DNA compared with the normal control allele, in three different PCR experiments, the mean of the three quantifications is indicated above each peak. (b) Detection by genomic CGH array of the mosaic duplication. The arrow indicates the location of the duplication that extends on 2.3 kb encompassing exons 61 and 62. The log₂ ratio is at +0.3, in accordance with the quantification of 50% of heterozygous mutant:wild-type cells.

Figure 4

Detection of rearrangements involving whole genes and neighboring genomic regions. (a) Heterozygous deletion of 740 kb involving the entire LIS1 (PAFHB1) gene and neighboring MET1OD, KIAA0064 and GARNL4 genes identified in patient 31 with a Mieller Diecker syndrome. The MNT and OR3A2 genes are not deleted. (b) Hemizygous deletion of the entire KAL1 gene in patient 35 with Kallman syndrome associated with ocular albinism, corresponding to a neomutation event. A 2297-kb deletion is detected in the patient sample and involves not only the KAL1 gene, but also neighboring genomic region including the GPR143 (OA1) gene. Mutations in this gene are known to be associated with ocular albinism phenotypes. The deletion is not detected in DNA sample from the mother.

Figure 5

Identification of complex rearrangements. (A) Complex rearrangement in the DMD gene: duplication of exons 61–62, duplication of exons 65–67, triplication of exons 68–79. (a) CGH array. (b) Confirmation by real-time PCR of exons 63, 67 and 68. (c) Genealogical data. Tested women are indicated by an asterisk. The complex rearrangement was stably transmitted. (B) Triplication of the MECP2 gene embedded within a duplication in a male patient with severe encephalopathy. (a) CGH array. The log₂ ratios (indicated as log₂ R) from 0.822 to 1.393 indicated a hemizygous triplication of a 82-kb region that includes the entire MECP2 gene, embedded within a duplicated region of 329 kb involving several other genes. (b) Confirmation of the triplication by real-time PCR of exons 3 and 4 of the MECP2 gene. Dup C, duplicated control; N, normal control; P, patient.