Deletion of exon 8 from the EXT1 gene causes multiple osteochondromas (MO) in a family with three affected members
© Zhuang et al. 2016
Received: 4 November 2015
Accepted: 8 January 2016
Published: 22 January 2016
Multiple osteochondromas (also called hereditary multiple exostoses) is an autosomal dominant disorder characterized by multiple cartilaginous tumors, which are caused by mutations in the genes for exostosin-1 (EXT1) and exostosin-2 (EXT2). The goal of this study was to elucidate the genetic alterations in a family with three affected members. Isolation of RNA from the patients’ blood followed by reverse transcription and PCR amplification of selected fragments showed that the three patients lack a specific region of 90 bp from their EXT1 mRNA. This region corresponds to the sequence of exon 8 from the EXT1 gene. No splice site mutation was found around exon 8. However, long-range PCR amplification of the region from intron 7 to intron 8 indicated that the three patients contain a deletion of 4318 bp, which includes exon 8 and part of the flanking introns. There is evidence that the deletion was caused by non-homologous end joining because the breakpoints are not located within a repetitive element, but contain multiple copies of the deletion hotspot sequence TGRRKM. Exon 8 encodes part of the active site of the EXT1 enzyme, including the DXD signature of all UDP-sugar glycosyltransferases. It is conceivable that the mutant protein exerts a dominant negative effect on the activity of the EXT glycosyltransferase since it might interact with normal copies of the enzyme to form an inactive hetero-oligomeric complex. We suggest that sequencing of RNA might be superior to exome sequencing to detect short deletions of a single exon.
KeywordsMultiple osteochondromas (MO) Hereditary multiple exostoses (HME) Exostosin-1 (EXT1) Glycosyltransferase Genomic deletion
Multiple osteochondromas (MO), also called hereditary multiple exostoses (HME), is a genetic, autosomal dominant disorder characterized by multiple benign cartilaginous tumors. These tumors (or exostoses) form next to the growth plate of long bones, ribs and pelvis and protrude into the adjacent perichondrium and neighboring tissue (Huegel et al. 2013). Patients with MO are frequently of short stature and show shortened forearms, changes in the angle of fingers and ankle and unequal length of limbs. Most of the patients have to undergo repetitive surgery to remove some of the exostoses because they cause pain or functional impairment (Pedrini et al. 2011). The most serious complication of MO is a malignant transformation into a secondary chondrosarcoma, which occurs in 1–5 % of the patients.
MO is caused by heterozygous inactivating mutations of the genes for exostosin-1 (EXT1) or exostosin-2 (EXT2). The majority of these mutations are nonsense, frame shift and splice-site mutations, but there are also missense mutations and deletions (Jennes et al. 2009; Pedrini et al. 2011). About 65 % of the MO mutations occur in EXT1 and 25 % in EXT2. The residual 10 % of the patients do not display any overt mutations in EXT1 or EXT2. Recent results suggest, however, that some of these cases might be explained by somatic, mosaic mutations in the EXT genes (loss of heterozygosity) and/or by mutations in some of the large introns (Szuhai et al. 2011; Waaijer et al. 2013).
EXT1 and EXT2 code for glycosyltransferases in the Golgi apparatus that are involved in the formation of heparan sulfate chains for proteoglycans (Esko and Selleck 2002; Nadanaka and Kitagawa 2008). Both enzymes are typical type II transmembrane proteins with the C-terminal part of the protein outside of the membrane. The growth plate of long bones is known to contain large amounts of heparan sulfate proteoglycans, mainly syndecan, glypican and perlecan (Farach-Carson et al. 2005). EXT1 and EXT2 belong to a small family of related proteins, which also includes the three exostosin-like proteins EXTL1, EXTL2 and EXTL3. With the exception of EXTL2, all EXT proteins contain a transmembrane domain at the N-terminal end, an Exostosin interaction domain in the center and a catalytic domain at the C-terminal end (Additional file 1: Fig. S1). EXTL2 in contrast contains only the transmembrane domain and the catalytic domain. EXT1 and EXT2 form a hetero-oligomeric complex (Kobayashi et al. 2000; McCormick et al. 2000; Senay et al. 2000) that adds repeating units of glucuronic acid (GlcA) and N-acetyl-glucosamine (GlcNAc) to the nascent chain of heparan sulfate. EXTL1, EXTL2 and EXTL3 on the other hand, appear to be involved in the addition of the first carbohydrate residue to the linker tetrasachararide of the proteoglycan core protein (Nadanaka and Kitagawa 2008).
Loss-of-function mutations in one allele of the EXT genes cause a significant decrease of glycosyltransferase activity. Why the loss of a single copy of EXT1 or EXT2 leads to a nearly complete loss of enzymatic activity is not yet clear (Hall et al. 2002). It is possible that the mutated proteins interact with normal proteins to form an inactive hetero-oligomeric complex. Furthermore, studies with isolated chondrocytes from patients showed that cells with an EXT mutation have a significantly decreased survival rate (Hecht et al. 2005). At any rate, the decrease in glycosyltransferase activity leads to a general deficiency of heparan sulfate in the growth plate. This deficiency causes an increase in the distribution range of signaling proteins, including Ihh, BMP, FGF and Wnt. Diffusion of these signaling proteins is usually restricted by binding to heparan sulfate proteoglycans found on cell surfaces and in the extracellular matrix. Finally, the increased distribution range of the growth factors might initiate ectopic chondrogenesis and exostosis formation.
In the present work we have studied a patient with MO from a family containing three affected members. The patient revealed a localized deletion of a single exon of EXT1. Deletions, especially deletions encompassing only a single exon have rarely been found with MO patients, probably because such deletions are more difficult to detect by conventional methods.
Description of patients
The currently 29-year-old younger brother (#14) of the index patient was diagnosed at the age of 12. Like his older brother, he had multiple exostoses, some of which required surgical resection. He shows shortened legs and forearms and has a mild walking impairment.
The 61-year-old mother (#8) of the index patient was diagnosed at age 10. Most of her limbs are shorter than normal. She underwent multiple surgical interventions before the age of 18 in order to remove exostoses. Due to a length difference of her legs she has severe walking impairment.
Based on these clinical findings it is likely that mother #8 harbors a de novo mutation, which was passed on to her sons #13 and #14. This mutation might have occurred early during development or alternatively, it was already present in the germline of one of the grandparents #1 or #2.
Identification of the mutation
Mechanism of deletion
In this study, we have investigated the molecular cause of MO syndrome in a family with three affected members. By PCR amplification and DNA sequencing, we were able to identify a discrete deletion of 4318 bp that includes exon 8 of the EXT1 gene and parts of the flanking introns.
In contrast to many other studies, which rely on exon sequencing of genomic DNA, we have started from purified RNA of the index patient and analyzed the sequences of the EXT1 and EXT2 mRNAs. We reasoned that leukocytes from the patient’s blood must contain heparan sulfate proteoglycans on their cell surfaces and consequently contain detectable levels of mRNA for these glycosyltransferases. It is likely that we would have missed the relatively short deletion of exon 8 if we had used exome sequencing. In fact, a primer pair flanking this exon amplified a fragment with the expected size from the intact allele, whereas no product was obtained from the mutant allele.
Large genomic deletions spanning several kb of genomic DNA are usually detected by fluorescent in situ hybridization (FISH). However, short deletions spanning only a few hundred bp are difficult to identify by this method due to its limited sensitivity. In this case, quantitative comparative hybridization to a gene array containing genomic DNA fragments (array-CGH) is the method of choice. But again, short deletions as in our case are difficult to detect by gene arrays, especially when these arrays are based on common SNPs with poor resolution. In order to detect short deletions, some researchers have developed special arrays with “tiling resolution” that contained synthetic DNA fragments based on the genomic sequence of interest (Szuhai et al. 2011). However, such arrays are not commercially available and have to be prepared individually for each project.
The most common mutations identified in MO patients are frameshift mutations (44 %), nonsense mutations (24 %), and splice site mutations (11 %). In most cases, these mutations result in premature termination of translation and polypeptide truncation (Jennes et al. 2009). Furthermore, it is likely that the affected mRNAs are degraded by nonsense-mediated decay NMD (Miller and Pearce 2014). Deletions, especially deletions of a single exon as in our case, are rather rare. Jennes et al. (2008) analyzed a cohort of 63 MO patients and found only 3 patients with EXT1 deletions. Furthermore, we are aware of only one additional family with a deletion of exon 8 alone, which has been published in the literature (Jennes et al. 2008, 2011). It is of interest to note that the exact deletion breakpoints differ between our family (117,814,822-117,810,504) and the one reported in the literature (117,815,648-117,809,763).
Exon 8 is 90 bp long and codes for 30 amino acids. Since it is flanked by splice sites of phase 0, deletion will not change the reading frame. Thus, the transcribed mRNA will contain the normal stop codon and the normal 3′UTR; consequently it will not be recognized by NMD for degradation. In fact, we observed roughly equimolar amounts of the cDNAs from the wildtype allele and from the mutated allele. It is possible that this fact makes the mutation particularly devastating as discussed in the following.
In a pairwise alignment (Fig. 8) the portion of the human EXT1 sequence, which is encoded by exon 8 but missing in our patients, aligns with residues 131-160 of mouse ExtL2. This is exactly the region that would interact with the UDP-sugar and the essential manganese ion. The DXD motif, which represents the signature of all UDP-sugar dependent glycosyltransferases (Breton and Imberty 1999), is found in mouse ExtL2 at residues 151-153. The two aspartate residues have been shown to interact with Mn2+ and the UDP-sugar and to position them in the proper orientation for catalysis (Pedersen et al. 2003). This motif is conserved as 565-DED-567 in human EXT1. Furthermore, residue R135 of mouse ExtL2, which forms a hydrogen bond with the donor sugar, is conserved as R549 in human EXT1 (Fig. 8). Based on this alignment, it is obvious that the EXT1 protein of our patients must be enzymatically inactive.
Quantitative analysis of the glycosyltransferase activity of mutated EXT1 proteins from several MO patients has shown that the residual enzyme activity was much lower than 50 % as would be expected in the heterozygous state (Hecht et al. 2005; Cheung et al. 2001). Theoretically there are two ways to explain why mutation of a single EXT1 copy might result in MO in a dominant fashion. Either, the residual glycosyltransferase activity is not sufficient to fulfill the function of the enzyme (haploinsufficiency), or alternatively, the mutated protein exerts a dominant negative effect on the activity of the wildtype protein. Work with knockout mice has provided evidence against the first possibility. Homozygous Ext1 knockout mice (expected Ext enzyme levels 0 %) are lethal and die at embryonic day E8.5, indicating the importance of heparan sulfate for embryonic development (Lin et al. 2000). However, heterozygous knockout mice (expected Ext enzyme levels 50 %) are largely normal and lack an obvious skeletal phenotype, thus ruling out haploinsufficiency in mice (Hilton et al. 2005). Likewise, homozygous Ext2 knockout mice are embryonically lethal. However, heterozygous Ext2 knockout mice develop, although with very low frequency, exostosis-like protrusions along the ribs, suggesting that a partial loss of the Ext glycosyltransferase activity may be sufficient for exostosis formation (Stickens et al. 2005). Moreover, compound heterozygous mice lacking one allele of Ext1 plus one allele of Ext2 frequently exhibit exostoses on the long bones (Zak et al. 2011). These mice should possess approximately 25 % of the Ext enzyme levels when compared to wildtype littermates. Thus, a significant, but not necessarily a complete loss of Ext protein expression is sufficient to explain exostoses formation.
Our MO patients contain a heterozygous deletion in EXT1 and show only a relatively small reduction of the normal EXT1 mRNA levels (Fig. 3a). Mutation of the second allele (loss of heterozygosity) in somatic cells would explain a further decrease of EXT enzyme activity, but loss of heterozygosity is very rare and has been ruled out in the majority of the MO patients (Hall et al. 2002). It is therefore likely that the mutant protein encoded by the affected allele exerts a dominant negative effect on the intact protein encoded by the wildtype allele.
How can we explain a dominant negative effect of the mutated EXT1 protein? EXT1 is known to form an oligomeric complex with EXT2 (Kobayashi et al. 2000; McCormick et al. 2000; Senay et al. 2000). It is likely that this complex contains several copies of EXT1 and EXT2 (e.g. stoichiometry 2:2 or 4:4) because it has to interact with the nascent chain of heparan sulfate that consists of at least four sugar residues, the starting tetrasaccharide (Nadanaka and Kitagawa 2008). It is therefore conceivable that incorporation of a mutated copy of EXT1 into such an assembly will render the entire complex inactive. One false copy could therefore compromise the activity of several normal copies of the polypeptide. Since the N-terminal domain of EXT1, which is involved in protein–protein interactions, remains intact in our case, it is likely that the mutated enzyme will still be incorporated into the final complex. Faithful incorporation of a mutant protein into an oligomeric complex would thus explain a dominant negative effect. However, it should be emphasized that there might be additional ways to explain a dominant negative effect.
We have demonstrated that sequencing of the mRNA, rather than the genomic DNA, might offer an advantage over exome sequencing if one is searching for short deletions spanning only a few hundred bp. It remains to be determined whether deletions affecting a single exon are more common than initially thought because they were missed by exome sequencing in some reports.
Total RNA was isolated from the blood of the patients using the QIAamp RNA Blood Mini kit from Qiagen (Hilden, Germany). Written consent was obtained to use the samples for genetic testing and to publish the results in accordance with the guidelines of the Swiss Society of Medical Genetics (www.sgmg.ch).
The blood (2 ml) was collected in the presence of the anti-coagulant EDTA and immediately processed. Erythrocytes were lysed in hypotonic buffer. Leukocytes were collected by centrifugation, lysed with a QIAshredder and dissolved in RLT buffer. Total RNA was adsorbed to a silica-based membrane, washed and selectively eluted with TE buffer. The purified RNA was transcribed into first-strand cDNA using hexamer primers and Improm-II Reverse Transcriptase (Promega, Madison WI, USA).
Isolation of genomic DNA
DNA was isolated from freshly drawn blood using the Wizard genomic DNA purification kit from Promega. Red blood cells were lysed in Cell Lysis Solution. White blood cells were collected by centrifugation and lysed in Nuclei Lysis Solution. RNA was digested with RNase and cellular proteins were removed by salt precipitation. Finally, the genomic DNA was concentrated by precipitation with isopropanol and washed with ethanol.
PCR amplification and DNA sequencing
Fragments of interest were amplified from the cDNA by PCR through 35 cycles (20″ at 95 °C, 20″ at individual annealing temperature (Additional file 1: Table S1), 15″ at 72 °C) with KAPA high fidelity DNA polymerase (KAPA Biosystems, Boston, USA). The fragments were resolved on 1 % agarose gels. Single bands were recovered from the agarose gel with the GeneElute extraction kit (Sigma, St. Louis MO, USA). For long-range PCR, the cycle number was reduced to 33 and the extension time at 72 °C was increased to 2′ 40″. In this case, the amplified fragments were resolved on a 0.5 % agarose gel. DNA sequences were determined by Sanger cycle sequencing on an ABI 3730 platform.
All sequences were analysed with the program package of MacVector Inc. For sequence alignments, the program ClustalW was used with the following parameters: Scoring matrix Gonnet, open gap penalty 10, extend gap penalty 0.1. Transmembrane sequences and PFAM protein domains were identified with the simple modular architecture research tool (Schultz et al. 1998).
LZ and BT carried out the experiments. SDG and SK provided X-rays and interpreted them. BT designed the study and drafted the manuscript. All authors read and approved the final manuscript.
This study was supported by a Grant from the Swiss National Science Foundation (31003A_143350).
The authors declare that they have no competing interests.
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.
- Abeysinghe SS, Chuzhanova N, Krawczak M, Ball EV, Cooper DN (2003) Translocation and gross deletion breakpoints in human inherited disease and cancer I: nucleotide composition and recombination-associated motifs. Hum Mutat 22:229–244View ArticleGoogle Scholar
- Breton C, Imberty A (1999) Structure/function studies of glycosyltransferases. Curr Opin Struct Biol 9:563–571View ArticleGoogle Scholar
- Chen JM, Cooper DN, Férec C, Kehrer-Sawatzki H, Patrinos GP (2010) Genomic rearrangements in inherited disease and cancer. Semin Cancer Biol 20:222–233View ArticleGoogle Scholar
- Cheung PK, McCormick C, Crawford BE, Esko JD, Tufaro F, Duncan G (2001) Etiological point mutations in the hereditary multiple exostoses gene EXT1: a functional analysis of heparan sulfate polymerase activity. Am J Hum Genet 69:55–66View ArticleGoogle Scholar
- Esko JD, Selleck SB (2002) Order out of chaos: assembly of ligand binding sites in heparan sulfate. Annu Rev Biochem 71:435–471View ArticleGoogle Scholar
- Farach-Carson MC, Hecht JT, Carson DD (2005) Heparan sulfate proteoglycans: key players in cartilage biology. Crit Rev Eukaryot Gene Expr 15:29–48View ArticleGoogle Scholar
- Hall CR, Cole WG, Haynes R, Hecht JT (2002) Reevaluation of a genetic model for the development of exostosis in hereditary multiple exostosis. Am J Med Genet 112:1–5View ArticleGoogle Scholar
- Hecht JT, Hayes E, Haynes R, Cole WG, Long RJ, Farach-Carson MC, Carson DD (2005) Differentiation-induced loss of heparan sulfate in human exostosis derived chondrocytes. Differentiation 73:212–221View ArticleGoogle Scholar
- Hilton MJ, Gutiérrez L, Martinez DA, Wells DE (2005) EXT1 regulates chondrocyte proliferation and differentiation during endochondral bone development. Bone 36:379–386View ArticleGoogle Scholar
- Huegel J, Sgariglia F, Enomoto-Iwamoto M, Koyama E, Dormans JP, Pacifici M (2013) Heparan sulfate in skeletal development, growth, and pathology: the case of hereditary multiple exostoses. Dev Dyn 242:1021–1032View ArticleGoogle Scholar
- Jennes I, Entius MM, Van Hul E, Parra A, Sangiorgi L, Wuyts W (2008) Mutation screening of EXT1 and EXT2 by denaturing high-performance liquid chromatography, direct sequencing analysis, fluorescence in situ hybridization, and a new multiplex ligation-dependent probe amplification probe set in patients with multiple osteochondromas. J Mol Diagn 10:85–92View ArticleGoogle Scholar
- Jennes I, Pedrini E, Zuntini M, Mordenti M, Balkassmi S, Asteggiano CG, Casey B, Bakker B, Sangiorgi L, Wuyts W (2009) Multiple osteochondromas: mutation update and description of the multiple osteochondromas mutation database (MOdb). Hum Mutat 30:1620–1627View ArticleGoogle Scholar
- Jennes I, de Jong D, Mees K, Hogendoorn PC, Szuhai K, Wuyts W (2011) Breakpoint characterization of large deletions in EXT1 or EXT2 in 10 multiple osteochondromas families. BMC Med Genet 12:85View ArticleGoogle Scholar
- Kobayashi S, Morimoto K, Shimizu T, Takahashi M, Kurosawa H, Shirasawa T (2000) Association of EXT1 and EXT2, hereditary multiple exostoses gene products, in Golgi apparatus. Biochem Biophys Res Commun 268:860–867View ArticleGoogle Scholar
- Krawczak M, Cooper DN (1991) Gene deletions causing human genetic disease: mechanisms of mutagenesis and the role of the local DNA sequence environment. Hum Genet 86:425–441View ArticleGoogle Scholar
- Lin X, Wei G, Shi Z, Dryer L, Esko JD, Wells DE, Matzuk MM (2000) Disruption of gastrulation and heparan sulfate biosynthesis in Ext1-deficient mice. Dev Biol 224:299–311View ArticleGoogle Scholar
- McCormick C, Duncan G, Goutsos KT, Tufaro F (2000) The putative tumor suppressors EXT1 and EXT2 form a stable complex that accumulates in the Golgi apparatus and catalyzes the synthesis of heparan sulfate. Proc Natl Acad Sci USA 97:668–673View ArticleGoogle Scholar
- Miller JN, Pearce DA (2014) Nonsense-mediated decay in genetic disease: friend or foe? Mutat Res, Rev Mutat Res 762:52–64View ArticleGoogle Scholar
- Nadanaka S, Kitagawa H (2008) Heparan sulphate biosynthesis and disease. J Biochem 144:7–14View ArticleGoogle Scholar
- Pedersen LC, Dong J, Taniguchi F, Kitagawa H, Krahn JM, Pedersen LG, Sugahara K, Negishi M (2003) Crystal structure of an alpha 1,4-N-acetylhexosaminyltransferase (EXTL2), a member of the exostosin gene family involved in heparan sulfate biosynthesis. J Biol Chem 278:14420–14428View ArticleGoogle Scholar
- Pedrini E, Jennes I, Tremosini M, Milanesi A, Mordenti M, Parra A, Sgariglia F, Zuntini M, Campanacci L, Fabbri N, Pignotti E, Wuyts W, Sangiorgi L (2011) Genotype-phenotype correlation study in 529 patients with multiple hereditary exostoses: identification of “protective” and “risk” factors. J Bone Joint Surg Am 93:2294–2302View ArticleGoogle Scholar
- Schultz J, Milpetz F, Bork P, Ponting CP (1998) SMART, a simple modular architecture research tool: identification of signaling domains. Proc Natl Acad Sci USA 95:5857–5864View ArticleGoogle Scholar
- Senay C, Lind T, Muguruma K, Tone Y, Kitagawa H, Sugahara K, Lidholt K, Lindahl U, Kusche-Gullberg M (2000) The EXT1/EXT2 tumor suppressors: catalytic activities and role in heparan sulfate biosynthesis. EMBO Rep 1:282–286View ArticleGoogle Scholar
- Stickens D, Zak BM, Rougier N, Esko JD, Werb Z (2005) Mice deficient in Ext2 lack heparan sulfate and develop exostoses. Development 132:5055–5068View ArticleGoogle Scholar
- Szuhai K, Jennes I, de Jong D, Bovée JV, Wiweger M, Wuyts W, Hogendoorn PC (2011) Tiling resolution array-CGH shows that somatic mosaic deletion of the EXT gene is causative in EXT gene mutation negative multiple osteochondromas patients. Hum Mutat 32:E2036–E2049View ArticleGoogle Scholar
- Waaijer CJ, Winter MG, Reijnders CM, de Jong D, John Ham S, Bovée JV, Szuhai K (2013) Intronic deletion and duplication proximal of the EXT1 gene: a novel causative mechanism for multiple osteochondromas. Genes Chromosom Cancer 52:431–436View ArticleGoogle Scholar
- Zak BM, Schuksz M, Koyama E, Mundy C, Wells DE, Yamaguchi Y, Pacifici M, Esko JD (2011) Compound heterozygous loss of Ext1 and Ext2 is sufficient for formation of multiple exostoses in mouse ribs and long bones. Bone 8:979–987View ArticleGoogle Scholar