Article Text

Download PDFPDF

Pseudoxanthoma elasticum: evidence for the existence of a pseudogene highly homologous to the ABCC6gene
  1. Dominique P Germain
  1. Clinical Genetics Unit, Department of Genetics, Hôpital Européen Georges Pompidou, 20 rue Leblanc, 75015 Paris, France
  1. Dr Germain,dominique.germain{at}

Statistics from

Editor—Pseudoxanthoma elasticum (PXE, MIM 264800) is an inherited disorder of connective tissue in which the elastic fibres of the skin, eyes, and cardiovascular system slowly become calcified, causing a spectrum of disease involving these three organ systems, with highly variable phenotypic expression.1 2Mutations in the ABCC6 gene (previously known as MRP6), encoding a 1503 amino acid membrane transporter, have recently been identified by our group and others3-7 as the genetic defect responsible for PXE. We subsequently designed a strategy for a complete mutational analysis of the ABCC6 gene, in order to provide accurate molecular and prenatal diagnosis of PXE. During this mutational screening, we have found evidence for the existence of at least one pseudogene highly homologous to the 5′ end ofABCC6. Sequence variants in thisABCC6-like pseudogene could be mistaken for mutations in the ABCC6 gene and consequently lead to erroneous genotyping results in pedigrees affected with pseudoxanthoma elasticum.

Material and methods

Seven unrelated patients presenting with PXE were evaluated for mutational analysis of the ABCC6 gene. For each proband, diagnosis of PXE was consistent with previously reported consensus criteria,8 which include a positive von Kossa stain of a skin biopsy, indicating calcification of elastic fibres, in combination with specific cutaneous and ocular manifestations (angioid streaks).

Whole blood samples were obtained after participants had provided written consent using a form that was approved by the Institutional Review Board of our academic institution. High molecular weight DNA was isolated from peripheral blood leucocytes, using a standard salting out procedure. Primers for amplification of theABCC6 gene were designed from the published sequence of human chromosome 16 bacterial artificial chromosome (BAC) clone A-962B4 (GenBank accession number U91318). PCR amplifications ofABCC6 exon 2 and exon 9 were done in 20 μl volumes with 100 nmol/l of each of the respective PCR primers (table1), 100 ng of genomic DNA, 100 μmol/l of each dNTP, 1.0 U Amplitaq Gold DNA polymerase (PE Biosystems), 10 mmol/l pH 8.3 Tris-HCl, 50 mmol/l KCl, and 1.5 mmol/l MgCl2. The thermal cycling profile used was 95°C for 10 minutes, followed by 35 cycles at 95°C for 30 seconds, 58°C for 30 seconds, and 72°C for one minute, followed by one cycle at 72°C for 10 minutes and a soak at 6°C.

Table 1

Primers for amplification of the ABCC6 gene and cDNA

RNA was isolated from lymphoblastoid cell lines or skin fibroblasts using Rneasy™ (Qiagen) and was reverse transcribed using random hexamers and Superscript-RT (Gibco BRL). Following reverse transcription, RT-PCR amplifications encompassing exon 2 in the published ABCC6 cDNA sequence9(GenBank accession number AF076622) were done in 20 μl volumes with 4 μl of RT reaction. An aliquot of the amplified product was analysed by ethidium bromide visualisation on a 1.5% agarose gel. A 1/50 dilution was submitted to a nested PCR, using internal specific primers (table 1) and 2.0 U Amplitaq Gold DNA polymerase (PE Biosystems), under the following conditions: 95°C for 10 minutes, followed by 20 cycles at 95°C for 30 seconds, annealing temperature gradient ranging from 48°C to 70°C for 30 seconds, and 72°C for one minute 30 seconds, followed by one cycle at 72°C for 10 minutes, and a soak at 10°C, on a Robocycler gradient 96 (Stratagene).

All PCR and RT-PCR fragments were purified using QIAquick Spin PCR Purification Kit (Qiagen) according to the manufacturer's protocol, and 4 μl of the purified PCR products were sequenced using the Big Dye Terminator AmpliTaq FS Cycle Sequencing Kit on an automated ABI 310 DNA sequencer (PE Biosystems). DNA sequences were handled with Navigator 2.0 software.


During our mutational screening of theABCC6 gene, we disclosed sequence changes which, although predicted to be truncating mutations, were unexpectedly detected not only in PXE patients but also in all tested controls. We first identified a single nucleotide insertion (c196insT) in the heterozygous state in a sporadic 14 year old female PXE patient. This mutation causes a shift in the reading frame, predicting a premature stop at codon 100 of the ABCC6 protein. Since we found this single nucleotide insertion in six other PXE patients, we initially interpreted this sequence change as a mutational hotspot. However, sequencing of the sporadic case's parents' DNA showed that, although unrelated to each other and phenotypically normal, they were both heterozygous for this frameshift mutation. These results were puzzling; if autosomal dominant inheritance with a de novo mutation2had occurred, neither of the unaffected parents should be a carrier, and, conversely, if autosomal recessive transmission had occurred, the proband should be a compound heterozygote on the basis of our results, and, consequently, only one of the parents would be expected to be a carrier of the c196insT mutation. These odd results prompted us to investigate 58 controls, all of whom showed a heterozygous profile for the frameshift mutation in what was thought to be exon 2 of theABCC6 gene (fig1).

Figure 1

Detection of a frameshift mutation (c196insT) in what was initially thought to be exon 2 of the ABCC6 gene. Genomic PCR products were sequenced using an antisense primer. Chromatograms show the insertion of an adenine (A) on the antisense strand (arrow), corresponding to a thymine (T) insertion on the sense strand. This single nucleotide insertion is responsible for a frameshift with consequent premature appearance of a stop codon. Heterozygosity for this frameshift mutation is shown here in two PXE patients (lanes 1 and 2) and two controls (lanes 3 and 4), but was also found in all 58 tested controls without exception. Sequencing of the other strand yielded the same result (not shown).

Similarly, a heterozygous C to T transition was found at cDNA position 1132 in exon 9 of the ABCC6 gene. This nucleotide substitution alters the codon (CAG) for glutamine to a stop codon (TAG), predicting termination of translation at position 378 of the ABCC6 protein (Q378X). This nonsense mutation was detected in the heterozygous state in all seven PXE patients and in one healthy volunteer. Mutation Q378X predicted the loss of aPstI restriction site. To test for the presence or absence of this nucleotide change, we usedPstI to digest PCR amplified genomic DNA of 79 additional controls and found all of them to be heterozygotes for the Q378X nonsense mutation (fig 2).

Figure 2

Detection of a nonsense mutation (Q378X) in what was initially thought to be exon 9 of the ABCC6 gene. (A) Upper panel: identification of a heterozygous nonsense mutation in four patients affected with PXE by direct automated sequencing of exon 9 of the ABCC6 gene. The heterozygous C to T transition alters the codon (CAG) for glutamine to a stop codon (TAG) at position 378 of the ABCC6 protein. The position of the mutation is shown by the letter Y (Y=C and T). (B) Lower panel: mutation c1132C>T (Q378X) predicted the loss of a PstI restriction site. Restriction digests using PstI were performed on PCR amplified exon 9 of the ABCC6 gene. Five healthy volunteers, who although unaffected with PXE display heterozygozity for the Q378X nonsense mutation, are shown. The study of 75 additional white controls yielded the same result. This indicates that rather than being amplified from two genomic copies, the PCR products were being amplified from four genomic copies.


Both results are interesting although surprising, since they identify two mutations, one nonsense and the other one inducing a frameshift, expected to cause truncation of the protein and thereby compromise its function. However, these mutations have been shown to be non-pathogenic since they are consistently found in healthy subjects. Among possible explanations for these results, we initially thought of the existence of mutational hotspots, but this hypothesis was ruled out through the discovery of the same mutations in controls. A technical artefact of direct automated sequencing was also considered, but was eliminated through the use of other experimental techniques, including restriction digest experiments. We then checked for possible homologies within the ATP binding cassette (ABC) superfamily.10 ABC genes are divided into seven distinct subfamilies (ABC1,MDR/TAP,MRP, ALD,OABP, GCN20, andWhite). However, if homologies do exist within the MRP subfamily (subfamily C) to whichABCC6 belongs, they are not important enough to explain our results, according to our database searches.

Finally, one likely explanation is the existence of a pseudogene with high homology with the 5′ end of the ABCC6gene, the PCR products being amplified from four rather than two genomic copies. Since we have used intronic primers to amplifyABCC6 genomic sequences, anABCC6-like pseudogene with introns is expected, as has been, for instance, reported for the gene encoding acid β-glucosidase.11 Indeed, as is found in Gaucher disease, the existence of a highly homologousABCC6 pseudogene hampers the accuracy of molecular diagnosis of PXE, since sequence variants in the pseudogene might be mistaken for pathogenic mutations in the activeABCC6 gene.

Pseudogenes are thought to arise from tandem gene duplication events caused by chromosome misalignment and unequal crossing over during meiosis. This mechanism would explain the high homology observed in both exonic and intronic sequences of theABCC6 gene and pseudogene.

The pseudogene could be located on a different chromosome or could be close to ABCC6 on chromosome 16. In favour of the later hypothesis is the fact that the short arm of chromosome 16 has been shown to be a site where complex rearrangements have taken place.12 Further evidence also comes from preliminary FISH experiments which detected double signals at 16p13.1, when fragments of BAC containing the ABCC6 gene were used as probes.13

In order to determine whether the pseudogene is expressed or not, RNA was isolated from skin fibroblasts and lymphoblastoid cell lines and RT-PCR experiments amplifying exon 2 inABCC6 cDNA were performed. We foundABCC6 mRNA to be expressed at low level in cultured skin fibroblasts and lymphoblastoid cell lines from both PXE patients and controls, in agreement with previous data indicating thatABCC6 is mainly expressed in liver and kidney.9 14 This prompted us to develop a nested PCR strategy, which proved efficient for the molecular analysis ofABCC6 mRNA. No frameshift was shown when nested RT-PCR fragments, encompassing the region corresponding toABCC6 exon 2, were sequenced (fig 3). This indicates that the pseudogene that we describe belongs to the unprocessed category.

Figure 3

RT-PCR chromatograms of the region corresponding to exon 2 in the ABCC6 cDNA in the four subjects shown in lanes 1-4 in fig1. ABCC6 mRNA was reverse transcribed and amplified through a nested PCR procedure. Direct automated sequencing of the region corresponding to the mutation detected in exon 2 at the genomic level does not show a thymine insertion in the mRNA and consequently no frameshift is seen. The same nucleotides as in fig 1 are shown, but sequencing was performed with a sense primer. These data indicate that the pseudogene is not expressed.

In conclusion, we have found nonsense and frameshift sequence variations in the ABCC6 gene, both of which appear to be non-pathogenic, thereby indicating the existence of at least one highly homologous pseudogene, which greatly complicates genotyping in families affected by pseudoxanthoma elasticum. Further studies are needed to map and fully characterise the sequence of the pseudogene(s). However, our results already emphasise the importance of not confusing variants in the pseudogene with pathogenic mutations in the ABCC6 gene, especially in genetic counselling or prenatal diagnosis.

  • Pseudoxanthoma elasticum (PXE) is an inherited systemic disorder of connective tissue with highly variable phenotypic expression.

  • Mutations in the ABCC6 gene were recently identified as the genetic defect responsible for PXE.

  • We have characterised two truncating mutations (c196insT and Q378X) in the ABCC6 gene, always found in the heterozygous state, not only in PXE patients but also in all controls.

  • This indicates the existence of a highly homologous pseudogene.

  • Sequence variants in the pseudogene should not be confused with mutations in the ABCC6 gene, especially in genetic counselling or prenatal diagnosis.


The author thanks the patients and their families for their help during this project and gratefully acknowledges F Letourneur, J Perdu, and I Roncelin for excellent technical assistance and Dr D Recan for providing lymphoblastoid cell lines.


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.