Introduction

Several genetic disorders, including tuberous sclerosis complex (TSC) and neurofibromatosis type 1 (NF1), show the effect of clinically unaffected parents who have multiple affected offspring. Germ line or somatic mosaicism in one of the parents has been described, resulting in a recurrence risk of up to 50%.1

TSC is an autosomal dominant disorder, caused by a mutation in either TSC1 or TSC2, inactivating the protein.2 About 70–80% of the TSC patients are sporadic cases.3 This observation does not exclude germ line mosaicism in one of the parents.

NF1 is an autosomal dominant neurocutaneous disorder,4 caused by a mutation in NF1, resulting in an inactive protein.

Point mutations as well as large rearrangements occur in TSC1, TSC2 and NF1. For the detection of large rearrangements, multiplex ligation-dependent probe amplification (MLPA) analysis is often used.5, 6 No data about the sensitivity of MLPA analysis are available. Here we present data that MLPA is not suitable for the detection of low-grade mosaicism. We recommend confirmation of an abnormal MLPA pattern by using either a mutation-specific PCR test after characterization of the breakpoint or, if possible, by fluorescent in situ hybridization (FISH) analysis. These methods can be used for analysis of other family members, especially the parents, for a better risk assessment.

Materials and methods

MLPA and FISH analysis

MLPA analysis was performed according to the manufacturer's instructions. Products were run on an automated sequencer (ABI 3730XL, Applied Biosystems, Foster City, CA, USA) and data were analyzed using Genemarker version 1.5 (Softgenetics, State College, PA, USA). In our routine diagnostic setting, the analysis is based on the MLPA ratio. Quantification is based on the peak height and normalization uses an internal control probe. For sequences present in two copies in the sample, the resulting value is 1.0. A reduction or increase in peak height values results in values <0.70 or >1.30, respectively (indicating a deletion or duplication). The kits used were: SALSA P046 for TSC2 and SALSA P081/P082 for NF1 (MRC-Holland BV, Amsterdam, The Netherlands).

Because of the possibility of somatic mosaicism, at least 50 metaphases derived from cultured PHA-stimulated peripheral blood lymphocytes were analyzed when performing FISH analysis for NF1 using the probe cP5.7

Characterization and confirmation of the breakpoint in TSC family

Using long-range PCR analysis, the breakpoints of an observed MLPA abnormality were characterised by sequencing the aberrant PCR product. A gene defect-specific PCR test was developed using three primers, one common primer, one primer specific for the gene defect and one primer more 3′ of the defect creating a wild-type fragment that is larger than the mutant fragment. In this way, detection of the mutant signal is favored (all primer sequences are available on request). The PCR products were run on an agarose gel.

Results and discussion

Previously, an abnormal pattern using Southern blot analysis with cDNA probes of TSC2 was obtained in the index patient of family 9703.1 DNA of the father showed a faint band at the same position of the abnormal fragment in DNA of the index patient, indicating that he was a somatic mosaic for his daughter's gene defect.

MLPA of TSC2 showed a pattern consistent with a duplication of exon 15–exon 26 (a probe for exon 27 is not present in the MLPA kit; Figure 1c). Using our routine MLPA protocol, the duplication was hardly detectable in DNA of the father (Figure 1b). Using a range of paternal input DNA of 25–250 ng DNA (Figure 1j–l), no clearly visible duplication could be shown. Using the internal control probe normalization as well as the population normalization using peak height as well as peak area calculation, also normalization and quantification of the peak results were tested. No differences between the results were seen (data not shown).

Figure 1
figure 1

MLPA analysis of TSC2. The graphics showed the MLPA results of peak ratio versus size of (a) a negative control sample, (b) father of the index patient, (c) index patient, (d–i) mixtures carrying the duplication in, respectively, 80 (d), 60 (e), 40 (f), 20 (g), 10 (h) and 0% of the cells (i); (j–l) father of the index patient with a range of input DNA of, respectively, 100 ng (normal amount) (j), 25 ng (k) and 250 ng (l). A blue dot refers to two copies of an internal control probe, a green dot to two copies of a gene specific probe and a red dot to respectively a deletion or duplication (<0.70 or >1.30) of a probe).

Full size image

Long-range PCR, followed by sequencing of the abnormal product resulted in the identification of pathogenic mutation. The duplication started in intron 14 and extended to the middle of exon 27 of TSC2, combined with an insertion of CACG (c.[3232insCACG;1600-75_3232dup13997]).

For confirmation, a duplication specific PCR was designed and used for testing the family members. The abnormal fragment of 627 bp and the wild-type fragment of 1315 bp were clearly visible in the index patient (Figure 2, lane 5). The duplication was not detected in control samples but was present in the father (Figure 2, lanes 1–4). However, the ratio of the mutant versus the wild-type fragment was lower compared with the index patient, indicating that the father was somatic mosaic for the mutation, confirming the result of the Southern blot analysis.

Figure 2
figure 2

Agarose gel electrophoresis of the TSC2 duplication specific PCR products. Lane 1, normal control; lane 2, father; lane 3 and lane 4, normal control; lane 5, index patient (100%); lane 6, 80% mosaic; lane 7, 60% mosaic; lane 8, 40% mosaic; lane 9, 20% mosaic; lane 10, 10% mosaic; lane 11, 0% mosaic; lane 12, no DNA.

Full size image

To estimate the percentage of affected white blood cells of the father, a titration curve was prepared by mixing DNA of the index patient with a negative control (Figure 2, lanes 5–11). The duplication-specific PCR showed that about 30% of the father's leukocytes carried the duplication (Figure 2, lane 2).

The detection rate of the MLPA was compared with that of the duplication-specific PCR using the previously mixed samples. Figure 2 showed that the duplication-specific PCR was able to detect a mosaic pattern of about 10%. A mosaic pattern of about 40% could be detected by MLPA analysis (Figure 1c–i), which explains why the 30% of the father could not be detected by MLPA (Figure 1b).

To investigate the sensitivity of the MLPA in detecting deletions, two patients with a total NF1 deletion previously identified using FISH analysis were tested. Patient 19578 carried the deletion in 85.4% of the cells and patient 26839 in 22%. MLPA analysis also showed the total gene deletions (Figures 3a and b). The sensitivity of the MLPA kit in detecting this NF1 total gene deletion was determined by a titration analysis using DNA of family 19578. Using the NF1-1 MLPA kit P081, this deletion may be detected when it is present in about 30% of the cells (Figure 3a). The NF1-2 MLPA kit P082 showed the same sensitivity (data not shown). These analyses showed that it is possible to detect a deletion present in 22–30% of the cells. These results indicate that, using a quantitative analysis, mosaic duplications are harder to detect than deletions. The size of the rearrangement and the probe density can influence the detection. It can be difficult to identify mosaic duplications or deletions of one exon, as the signal of the probe is also depending of DNA quality.8

Figure 3
figure 3

(a) MLPA NF1 results (only of the P081 kit). The graphics showed the MLPA results of peak ratio versus size. The percentage blood cells that carried the deletion varied from 85.4 to 0% as indicated in the graphics. (b) MLPA NF1 results (only of the P081 kit) of a patient with a 22% mosaic total gene deletion. A blue dot refers to two copies of an internal control probe, a green dot to two copies of a gene specific probe and a red dot to, respectively, a deletion or duplication (<0.70 or >1.30) of a probe).

Full size image

It has been shown that MLPA analysis can be used to detect high-grade mosaic NF1 patients.6 We investigated the sensitivity of the MLPA to detect a duplication of seven exons of TSC2 and total NF1 deletions. Our results indicated that a low percentage of somatic mosaicism could be missed. Somatic mosaicism of a single exon deletion or duplication will be even harder to identify. We recommend to confirm the abnormal MLPA patterns using an alternative technique; it is usually feasible to characterize the breakpoint and to develop a gene defect-specific PCR. Otherwise, for larger rearrangements, especially deletions, FISH analysis can be a reliable technique for the detection of low-grade mosaicism. Depending on probe availability and the identified abnormality, BACs, cosmids or fosmids can be used as a probe. Another possibility, depending on the probe number in the area of interest, is to use an oligonucleotide or SNP-based array. It is known that high-density SNP technology has a detection rate of up to ∼30%,9 which makes the sensitivity of this technique comparable with MLPA.

Detection of a low-grade mosaic parent has great implications for genetic counseling. In case the pathogenic mutation can not be detected in DNA of the parents, a low recurrence risk on affected offspring will be given. If somatic mosaicism is identified in one of the parents, the recurrence risk may be up to 50%, indicating the necessity to use the most sensitive technologies for the analysis of parental DNA.