Simultaneous and independent detection of C9ORF72 alleles with low and high number of GGGGCC repeats using an optimised protocol of Southern blot hybridisation

Abstract

Background: Sizing of GGGGCC hexanucleotide repeat expansions within the C9ORF72 locus, which account for approximately 10% of all amyotrophic lateral sclerosis (ALS) cases, is urgently required to answer fundamental questions about mechanisms of pathogenesis in this important genetic variant. Currently employed PCR protocols are limited to discrimination between the presence and absence of a modified allele with more than 30 copies of the repeat, while Southern hybridisation-based methods are confounded by the somatic heterogeneity commonly present in blood samples, which might cause false-negative or ambiguous results.

Results: We describe an optimised Southern hybridisation-based protocol that allows confident detection of the presence of a C9ORF72 repeat expansion alongside independent assessment of its heterogeneity and the number of repeat units. The protocol can be used with either a radiolabeled or non-radiolabeled probe. Using this method we have successfully sized the C9ORF72 repeat expansion in lymphoblastoid cells, peripheral blood, and post-mortem central nervous system (CNS) tissue from ALS patients. It was also possible to confidently demonstrate the presence of repeat expansion, although of different magnitude, in both C9ORF72 alleles of the genome of one patient.

Conclusions: The suggested protocol has sufficient advantages to warrant adoption as a standard for Southern blot hybridisation analysis of GGGGCC repeat expansions in the C9ORF72 locus.

Figure 1

Southern blot analysis of GGGGCC repeat expansion in the C9ORF72 locus. (A) Scheme of a region around exons 1a and 1b of human C9ORF72 gene with a single copy of GGGGCC sequence (top) or expansion of this repeat (bottom). The black bar in the middle denotes a DNA probe used for Southern blot hybridisation. The sizes of fragments produced by double EcoRI and XbaI digestion and detected by hybridisation with this probe are shown. (B) Hybridisation of EcoRI and XbaI digested DNA extracted from the whole blood of patients positive (+) or negative (-) for GGGGCC repeat expansion according to results of the repeat-PCR analysis. The ratios of band intensities obtained by scanning the X-ray film are shown below the image. Note that for sample 7 the intensity used for calculation is combined intensities of two close ~1.33 kb size bands. (C) Hybridisation of EcoRI and XbaI digested DNA extracted from the cultured lymphoblastoid cell lines of patients positive (+) or negative (-) for GGGGCC repeat expansion according to results of the repeat-PCR analysis.

Figure 2

Identification of a patient with GGGGCC repeat expansion in both alleles of the C9ORF72 locus. Hybridisation of EcoRI and XbaI digested DNA extracted from the cultured lymphoblastoid cell lines of patients positive (+) or negative (-) for GGGGCC repeat expansion according to results of the repeat-PCR analysis. Note the presence of ~1.6 kb band representing an allele with relatively small, although what is considered pathological, i.e. >30, number of repeats in one of the C9ORF72 loci in the genome of three patients (lanes 1 – 3). In contrast to patients carrying the second allele without pathological repeat expansion (lanes 2–6), one patient (lane 1) has the second allele with very high number of repeats. Where relevant, the ratios of the 1.33 kb to 1.05 kb and/or ~1.6 kb to 1.05 kb bands intensities are shown below the image. Note that for sample 6 the intensity used for calculation is combined intensities of two close ~1.33 kb size bands.

Figure 3

Somatic cell genomes of ALS patients have variable number of GGGGCC repeats in the C9ORF72 locus. Two sections of the same Southern blot (see raw image in Additional file 2: Figure S1) show hybridisation of EcoRI and XbaI digested DNA extracted from peripheral blood, cerebellum and cerebral cortex of two patients positive for GGGGCC repeat expansion. The ratios of the 1.33 kb to 1.05 kb bands intensities are shown below the image. Note that in this case these ratios are unconventionally high, which is due to the gel has been run for too long and therefore the 1.05 kb bands appeared too close to the edge of the gel, resulting in less efficient transfer and hybridisation of this band. However, the ratio for normal DNA sample is still approximately twice as high as the ratio for all DNA samples with GGGGCC repeat expansion.