Rapid genetic analysis of x-linked chronic granulomatous disease by high-resolution melting

Abstract

High-resolution melting analysis was applied to X-linked chronic granulomatous disease, a rare disorder resulting from mutations in CYBB. Melting curves of the 13 PCR products bracketing CYBB exons were predicted by Poland's algorithm and compared with observed curves from 96 normal individuals. Primer plates were prepared robotically in batches and dried, greatly simplifying the 3- to 6-hour workflow that included DNA isolation, PCR, melting, and cycle sequencing of any positive products. Small point mutations or insertions/deletions were detected by mixing the hemizygous male DNA with normal male DNA to produce artificial heterozygotes, whereas detection of gross deletions was performed on unmixed samples. Eighteen validation samples and 22 clinical kindreds were analyzed for CYBB mutations. All blinded validation samples were correctly identified. The clinical probands were identified after screening for neutrophil oxidase activity. Nineteen different mutations were found, including seven near intron-exon boundaries predicting splicing defects, five substitutions within exons, three small deletions predicting premature termination, and four gross deletions of multiple exons. Ten novel mutations were found, including (c.) two missense (730T>A, 134T>G), one nonsense (90C>A), four splice site defects (45 + 1G>T, 674 + 4A>G, 1461 + 2delT, and 1462-2A>C), two small deletions (636delT, 1661_1662delCT), and one gross deletion of exons 6 to 8. High-resolution melting can provide timely diagnosis at low cost for effective clinical management of rare, genetic primary immunodeficiency disorders.

Figure 1

Predicted and observed normalized melting curves for PCR products bracketing the 13 exons of CYBB. A: Predicted melting curves were calculated using Poland's recursive algorithm, as described in Material and Methods. B: The observed melting curves were obtained by normalization using linear background extrapolation after PCR and high resolution melting of a normal DNA sample.

Figure 2

High-resolution melting analysis of a variant X⁻ CGD family with a mutation in a three-domain melting curve. In the pedigree, black squares are affected males, gray squares are normal males, and half-black circles are heterozygous carrier females. Melting analysis of all family members and the wild-type control subjects clustered together except for exon 9. A: Normalized exon 9 melting curves after overlay reveals two clusters. Unaffected males, normal controls, and affected males without mixing are shown in gray. Heterozygous females and affected males mixed with wild-type DNA are shown in black. B: The differences between clusters are easier to see in “difference plots,” where all curves are subtracted from the wild-type average. Exon 9 sequencing of the affected males revealed a c.1016C>A mutation, also present in the carrier females in heterozygous form.

Figure 3

High-resolution melting analysis of a variant X⁻ CGD family with a mutation isolated to a minor low temperature domain. In the pedigree, black squares are affected males, gray squares are unaffected males, gray circles are wild-type females, and half-black circles are heterozygous carrier females. Melting analysis revealed variation within exon 3. A: Two melting curve clusters are shown and a low-temperature domain is present that is easy to miss unless melting curve prediction is used (Figure 1). B and C: When the two domains are analyzed separately, the variation is isolated to the low-melting domain. Targeted sequencing revealed a c.252 + 5G splicing mutation in the patients and carriers.

Figure 4

Deletion of exons 6 to 8 detected by PCR and high-resolution melting. A: Normalized melting curves for exons 6 to 8 are shown for both wild-type (WT) controls and a 1:1 mixture of unknown patient and wild-type DNA (gray traces). However, when the unknown sample is tested without mixing, no specific melting curves are observed (black traces). B: Fluorescent image of the plate. Primers for exons 1 to 7 are in columns 1, 3, 5, 7, 9, and 11 and primers for exons 8 to 13 are in columns 2, 4, 6, 8, 10, and 12. Mixed DNA is present in columns 1 to 4, the unknown sample without mixing is present in columns 5 to 8 (exons 6 to 8 are boxed), and wild-type male DNA is present in columns 9 to 12.