Article Text

PDF

Incidence and molecular mechanism of aberrant splicing owing to a G→C splice acceptor site mutation causing Smith-Lemli-Opitz syndrome
  1. H R WATERHAM*,
  2. W OOSTHEIM*,
  3. G J ROMEIJN*,
  4. R J A WANDERS*,
  5. R C M HENNEKAM
  1. * Laboratory for Genetic Metabolic Diseases (F0-226), Departments of Clinical Chemistry and Paediatrics, Emma Children's Hospital, Academic Medical Centre, University of Amsterdam, PO Box 22700, 1100 DE Amsterdam, The Netherlands
  2. Department of Paediatrics, Emma Children's Hospital, Academic Medical Centre, University of Amsterdam, PO Box 22700, 1100 DE Amsterdam, The Netherlands
  1. Dr Waterham, h.r.waterham{at}amc.uva.nl

Statistics from Altmetric.com

Editor—Smith-Lemli-Opitz (SLO) syndrome (MIM 270400) is an autosomal recessive disorder characterised by a wide spectrum of developmental abnormalities including craniofacial malformations, growth and mental retardation, and multiple congenital anomalies.1-3 The disorder is caused by deficient activity of 7-dehydrocholesterol reductase (7-DHCR), the enzyme that catalyses the reduction of the C7-C8 (Δ7) double bond of 7-dehydrocholesterol to produce cholesterol.4 5 As a consequence, patients with SLO syndrome have low plasma cholesterol and raised plasma 7-dehydrocholesterol concentrations, thus constituting the biochemical hallmark used to confirm the clinical diagnosis of the syndrome.

Recently, we and others have identified the cDNA coding for human 7-DHCR and confirmed that SLO syndrome is caused by mutations in the corresponding 7-DHCR gene.6-8 So far, 19 different disease causing mutations have been reported, which were identified after analysis of only 33 patient alleles. In our previous report, we described an aberrantly spliced mutant 7-DHCR cDNA with an insertion of 134 bp which we identified in two of three Dutch patients analysed with SLO syndrome, one of whom was a homozygote and the other a compound heterozygote.6 Upon translation, this insertion not only introduces additional amino acids but also causes a frameshift, which leads to an inactive, truncated protein with a changed C-terminus.6 Since the same insertion was also reported by two other groups,7 8 and since the affected patients were not related to one another, this finding suggested that the insertion is the result of a frequently occurring mutation. To evaluate this, we analysed 17 additional patients with SLO syndrome for the occurrence of the 134 bp insertion using allele specific RT-PCR. Furthermore, we studied the underlying mechanism that gives rise to the aberrant splicing, which results in the 134 bp insertion at the cDNA level.

First strand cDNA was synthesised from total RNA isolated from leucocytes or primary skin fibroblasts9 obtained from 17 patients who were diagnosed with SLO syndrome on the basis of clinical features and raised serum levels of 7-dehydrocholesterol. In the patients for whom fibroblasts were available, the biochemical defect was confirmed by the detection of reduced enzyme activity of 7-DHCR using a previously described ergosterol to brassicasterol conversion assay10 followed by sterol analysis using gas chromatography-mass spectrometry.11 To determine the presence of the 134 bp insertion, the first strand cDNA was used as template for the amplification by PCR of the second half of the coding region of 7-DHCR cDNA using primer set DHCR684-703 and DHCR1563-1544 (table 1). This primer set generates a single fragment of 915 bp in control subjects and patients with SLO syndrome, who do not harbour the 134 bp insertion (fig 1, lanes 2 and 3). In the previously described patient who was a homozygote for the insertion,6 a single fragment of 1049 bp is amplified (fig1, lane 4), while in a patient who is a compound heterozygote for the insertion, both fragments of 915 and 1049 bp are amplified. In compound heterozygotes, an additional intermediate sized fragment is frequently observed, which is a heteroduplex form composed of one strand of both fragments, as could be shown by mixing the 915 and 1049 bp fragments followed by one cycle of melting and annealing (not shown). Of the 20 patients thus analysed, including the three patients previously reported6 and three pairs of sibs, one patient was a homozygote and 13 patients heterozygotes for the insertion. Of these 14 patients, eight patients (including the homozygote and two pairs of sibs) were of Dutch origin, four patients of German origin (including one pair of sibs), one patient of Belgian and one patient of Spanish origin.

Table 1

PCR primers used in this study

Figure 1

Identification of the 134 bp insertion by RT-PCR. The second half of 7-DHCR cDNA was amplified by RT-PCR from a control subject (lane 2) and three genotypically different patients with SLO syndrome using primer set DHCR684-703 and DHCR1563-1544 (see table 1). One patient contained two missense mutations (lane 3) while the other two patients were either homozygous (lane 4) or heterozygous for the 134 bp insertion (lane 5). Lane 1 contains a 100 bp DNA molecular weight marker.

In their recent paper, Fitzky et al 7 reported that two of their 13 patients with SLO syndrome were compound heterozygotes for the 134 bp insertion. They also indicated that the insertion was caused by a G→C transversion in the splice acceptor site of intron 8 of the7-DHCR gene, which is included in the 134 bp insertion of the corresponding cDNA, but they did not study the molecular background in further detail. Therefore, we amplified the corresponding DNA region by PCR from genomic DNA isolated from two control subjects and two patients with SLO syndrome to confirm that the insertion is caused by a mutation in a splice acceptor site and to determine the underlying mechanism that causes the specific splicing defect. To this end two separate primer sets were used. The first primer set consisted of a forward primer designed on the cDNA sequence 5′ of the site of insertion (DHCR930-949) and a reverse primer designed on the 134 bp insertion sequence (DHCR-ins113-95), which amplified a genomic DNA fragment of approximately 2 kb. The second set consisted of a forward primer designed on the 134 bp insertion sequence (DHCR-ins21-39) and a reverse primer designed on the cDNA sequence 3′ of the site of insertion (DHCR996-978), which amplified a genomic DNA fragment of approximately 120 bp. Taken together, these results indicated that the 134 bp insertion resulted from a partial retention of the 3′ end of an intron with an approximate size of 2 kb. Indeed, sequence analysis of the two overlapping genomic DNA fragments which were amplified from the two control subjects identified consensus sequences for both a splice donor site and a splice acceptor site, which confirmed that the complete intron sequence, which comprises 1965 bp, was isolated (fig 2). Sequence analysis of the same overlapping genomic DNA fragments amplified from the two patients with SLO syndrome confirmed that the retention was caused by a G→C transversion at the −1 position of the splice acceptor site of the intron. This mutation disrupts the consensus sequence of the splice acceptor site and leads to alternative splicing of the transcript at a site located 134 bp 5′ of the original splice site. The intron DNA sequence adjacent to this alternative site of splicing shows strong homology to the consensus sequence of a splice acceptor site and is preceded by a DNA sequence which has strong homology to the branch point consensus sequence, thus constituting quite a strong alternative splice acceptor site (fig2).

Figure 2

Molecular basis of the 134 bp insertion. (A) Sequence analysis of the genomic DNA region from which the 134 bp insertion originates and amplified by PCR from control subjects identified an intron of 1965 bp (small lettering) with consensus splice donor, branch point, and splice acceptor site sequences (italics and underlined). The complete nucleotide sequence of the intron can be obtained from GenBank under accession number AF132981. (B) In the patients, a G→C mutation at the –1 position of the authentic splice acceptor site (double underline) results in alternative splicing at a cryptic splice acceptor site located 134 bp 5′ of the authentic splice acceptor site and preceded by a consensus branch point sequence (italics and underlined). The 134 bp intron sequence, which is retained as a result of the alternative splicing, is indicated in bold. Original exon sequences are indicated in capitals. Consensus sequences are defined as follows: splice donor site, [exon]..(C/A)AG ↓ gt(a/g)agt..[intron]; branch point, (t/c)n(t/c)t(g/a)ac (18-40 bp 5′ of splice acceptor site); splice acceptor site, [intron]..(t/c)11n(t/c)ag ↓ G/A..[exon].

The (compound) heterozygous presence of the 134 bp insertion enabled allele specific amplification by PCR of the two cDNAs, which could then be directly sequenced to identify the syndrome causing mutations on both alleles. To amplify the cDNA without insertion, primers DHCR979-960 and DHCR949-967, which both span the site of insertion (after nucleotide 963), were used in combination with, respectively, a forward primer in the 5′ (DHCR-58–38) and a reverse primer in the 3′ region (DHCR1563-1544). This amplification produced two overlapping fragments of 1073 and 650 bp, respectively (fig 3). The cDNA containing the 134 bp insertion was specifically amplified using the same 5′ and 3′ region primers in combination with the DHCR-ins113-95 and DHCR-ins21-39 primers designed on the insertion DNA sequence. This amplification produced two overlapping fragments of 1170 and 748 bp, respectively (fig 3). Using the latter two primer sets, the cDNAs containing the 134 bp insertion which were identified in the patients with SLO syndrome (see above) were amplified and analysed by sequencing. In all cases, the same G→C mutation at the −1 position of the splice acceptor site was identified.

Figure 3

Allele specific amplification of cDNAs with and without the 134 bp insertion. The 7-DHCR cDNA without the 134 bp insertion was specifically amplified in two overlapping fragments from a patient heterozygous for the insertion (lanes 2 and 3) and a patient with two missense mutations (lanes 6 and 7) using primer sets DHCR-58–38 and DHCR979-967 (lanes 2 and 6), and DHCR949-967 and DHCR1563-1544 (lanes 3 and 7), respectively. The cDNA containing the 134 bp insertion was specifically amplified in two overlapping fragments from the heterozygous patient using primer sets DHCR-58–38 and DHCR-ins113-95 (lane 4), and DHCR-ins21-39 and DHCR1563-1544 (lane 5). Using the latter two primer sets, no fragments were amplified from the patient who had two missense mutations and lacked the splice site mutation causing the insertion (lanes 8 and 9).

To conclude, using allele specific 7-DHCR cDNA amplification, we analysed 20 patients with SLO syndrome and identified one patient as a homozygote and 13 patients, including three pairs of sibs, as heterozygotes for the G→C transversion which results in the partial retention of intron 8 owing to alternative splicing.7After excluding the three sibs, this amounts to an incidence of ∼35% (12/34 alleles) for the mutant allele in our group of patients. When patients previously reported by others are also included,7 8 however, the incidence of this allele approximates ∼25% (15/59 alleles). This makes this allele the most frequently occurring among SLO patients, since all other causative mutations identified to date have been recurrent among a limited number of patients6-8 (unpublished results). The RT-PCR based methods presented in this paper provide an easy and rapid screening at the cDNA level for this frequently occurring mutant allele. The allele specific amplification of cDNAs in case of heterozygosity for the 134 bp insertion allows the establishment of compound heterozygosity without subcloning of the cDNAs, which is valuable in cases where no parental material is available.

The relatively high incidence of the splice acceptor site mutation among patients with SLO syndrome remains unexplained at present, but could be the result of a founder effect. This is supported by the complete absence of any of the quite common polymorphisms,7 189A>G, 207C>T, 231C>T, 438C>T, 1158C>T, and 1272T>C, in all 15 mutant alleles identified in our patients. On the other hand, this seems less likely as the allele was identified in unrelated patients of Dutch, German, Belgian, Spanish, and North American origin6-8 (this paper).

References

View Abstract

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.