This article is mentioned in:

Introduction

Hereditary multiple exostoses (HME), also known as multiple osteochondroma (MO), is an autosomal dominant bone disorder with an incidence of 1 in 50,000 live births in western populations (1). The morbidity rate is greater in males than females, with a ratio of 1.5:1, due to the mutations exerting a weaker phenotypic effect in females (2). HME is characterized by the presence of multiple benign cartilage-capped tumors, localized primarily in the long tubular bones, particularly in the humerus (10–50%), forearm (39–60%), knee (33%) and ankle (25%). The majority of patients with HME (~70%) have a family history of the condition (1,3,4).

HME is a genetically heterogeneous disorder with two disease-causing genes identified, exostosin glycosyl transferase-1 (EXT1) and exostosin glycosyl transferase-2 (EXT2) located at chromosomes 8q24 and 11p11-p12, respectively (5,6). The proteins encoded by human EXT1 and EXT2 are type II transmembrane glycoproteins, localized in the endoplasmic reticulum. The EXT1/EXT2 complex is involved in the biosynthesis of heparin sulphate (HS) proteoglycan (HSPG) (7). Prior to deacetylation, the EXT1/EXT2 complex catalyzes the elongation of the HS chain. EXT1 and EXT2 are ubiquitously expressed in developing limb buds, and in osteochondromas their expression was decreased in correlation with mutation status (8,9). To date, ~422 separate pathogenic mutations in EXT1 and ~221 mutations in EXT2 have been identified. Mutations in EXT1 account for 56–78% of cases in MO families, whereas EXT2 mutations have been identified in 21–44% of cases (4,10–15). However, in China, EXT2 mutations are identified more frequently than EXT1 mutations (16). The majority of patients have been identified in a single family or as sporadic cases. Nonsense, frameshift and splice-site mutations, which represent the majority of MO-causing mutations (80%), have been predicted to lead to the premature translational termination of the associated amino acids, and the subsequent production of a truncated protein (17). Mutations in EXT1 are dispersed along the gene, and may occur in various exons (14,18); however, EXT2 mutations do not appear to occur in the final third of the gene-coding region (17).

Other than EXT1 and EXT2, the EXT3 gene has been mapped to chromosome 19p (19). It appears to be a minor locus in HME families and no causative mutations in EXT3 have been identified (20). Three additional EXT-like genes, designated EXTL1, EXTL2 and EXTL3 have been identified and mapped to chromosomes 1 (1p36, 1p11-p12) and 8 (8p12) (21–23). Although the EXTL genes are considered strong candidate genes for MO, to date no HME family has been associated with these loci.

The present study investigated a rare large family with MO, and identified a novel splice-site mutation in EXT2.

Materials and methods

Subjects

Written consent was obtained from all study participants, and the present study was approved by the ethics committee of The Second Xiangya Hospital (Changsha, China). The proband was admitted to The Second Xiangya Hospital in 2012, and presented with a large osteochondroma. The five-generation Chinese family of the proband was subsequently investigated and a pedigree constructed based on clinical and radiographical evaluations of all family members (Fig. 1). Of the 33 family members, there were 13 affected individuals (11 males and 2 females) aged 8–80 years (average age, 52 years). The incidence rate was therefore 40% within the family, and MO occurred in each generation. The affected individuals had 6–16 exostoses, typically located at the juxtaepiphyseal regions of long bones; however, these were not as large as those present in the proband. There were no other lesions that were atypical of MO and no evidence of short stature (average height of adult male and female was 165 and 156 cm, respectively). All affected individuals had lesions detected prior to age 10, but had never received surgery, with the exception of the proband.

Figure 1. Pedigree analysis of a family with hereditary multiple exostoses. Affected individuals are present in every generation. Solid symbols, affected individuals; open symbols, unaffected individuals; squares, males; circles, females; arrow, proband; oblique lines, deceased.

The proband, family member IV-2, a 42-year-old male, presented with 41-year history of multiple osteochondroma. Physical examination revealed a large mass on the left side of the back (Fig. 2A), (39×33×19 cm), and >14 osseous nodules of varying sizes located on the prothorax wall, left scapula, bilateral forearms, knees and left ankle. The patient had had an operation at age 22 due to a rapidly enlarging lumbar spinal osteochondroma.

Figure 2. Clinical assessment of family member IV-2. (A) Osteochondroma present on back. (B) X-ray revealed multiple bony protrusions on limbs. (C) CT angiography did not reveal large artery imaging, indicating the potential presence of a chondroma. (D) CT revealed the left back mass had mixed density and multiple flecked calcifications, which broke into the left chest. (E) Histopathological analysis revealed an osteochondroma with focal malignant transformation. Scale bar=100 µm. CT, computed tomography.

X-ray analysis revealed a diffuse flocculent shadow with high density in the left lung field and multiple bony protrusions on limbs, as presented in Fig. 2B. Computed tomography (CT) angiography did not reveal any imaging of large arteries, indicating a potential chondroma (Fig. 2C). CT revealed the left back mass was of mixed density and multiple flecked calcifications, which spread into the left side of the chest (Fig. 2D). Histopathological analysis revealed an osteochondroma with focal malignant transformation, as presented in Fig. 2E. Surgery was conducted at The Second Xiangya hospital to remove the osteochondroma. The patient was subsequently followed up once every three months in the first year, and once every year thereafter. No recurrence has been detected as of October 2015.

DNA analysis

DNA of all affected family members was extracted from peripheral blood as previously described (24). The coding regions of the EXT1 and EXT2 genes were amplified by polymerase chain reaction (PCR) using primer sets, as previously described (25). The PCR products were purified by BigDye® Terminator version 1.1 (Applied Biosystems; Thermo Fisher Scientific, Inc., Waltham, MA, USA) followed by direct DNA sequencing with an ABI 3100 Automatic Sequencer (Applied Biosystems; Thermo Fisher Scientific, Inc.), using forward and reverse primers.

Reverse transcription-PCR (RT-PCR)

Total RNA was extracted from the tumor tissue of the proband and healthy lung tissue of an unrelated individual, using the QIAgen RNeasy Mini kit (Qiagen, Inc., Valencia, CA, USA), according to the manufacturer's protocol. cDNA synthesis was performed using RevertAid™ First Strand cDNA Synthesis kit (Fermentas; Thermo Fisher Scientific, Inc.), using the following primers: Forward, 5′-AACCAGAACACACTGCGCATCAAG and reverse, 5′-AGCTCCACGAAGAACCACACAGAA for exons 2–5. Amplification of cDNA was performed and products were purified by BigDye Terminator version 1.1 followed by direct DNA sequencing with an ABI 3100 Automatic Sequencer.

Results

DNA was extracted from family members with HME (Fig. 1A, solid squares), and the possible mutations residing in the EXT genes were scanned in the exon and intron junctions. DNA sequence analysis revealed a heterozygous mutation, c.939+1 G>T in EXT2 (Fig. 3) in all family members with HME; none of the unaffected family members carried this mutation. This indicated the association of this inheritable mutation with HME. No other mutations were reported in any of the analyzed samples.

Figure 3. Mutation in exostosin glycosyl transferase-2. (A) DNA sequencing of the polymerase chain reaction product. The position of the mutation in exon 5 is indicated by the red arrow. (B) Resulting alterations to the amino acid sequence in the mutated protein, due to the splicing out of exon 5. WT, wildtype.

The result of PCR gel electrophoresis is presented in Fig. 4. Following RT-PCR at exons 2–5, two bands were identified in the proband; one band of normal size (~500 bp) and one smaller band (~300 bp). The PCR products were purified and sequenced. The size of one band was revealed to be 429 bp, and the other, 296 bp. Compared with cDNA of EXT2, sequence analysis confirmed exon 5 skipping in the aberrant allele, resulting in an in-frame deletion of the EXT2 protein.

Figure 4. PCR analysis of proband. (A) PCR gel electrophoresis of EXT1 and EXT2 genes of the proband. (B) Reverse transcription-PCR of EXT2 mRNA from the tumor tissue of the proband (T1-T4) and healthy lung tissue from an unrelated individual (+C). There was no band present in the T1 column, indicating no mRNA was present in the first test tube, which may have occurred from RNA degradation due to length of storage. (C) DNA sequencing of the 196 bp fragment revealed the precise excision of exon 5 sequence from the transcript. The genomic DNA sequence of EXT2 between exons 4 and 6 is presented and aligned for reference. PCR, polymerase chain reaction; EXT1, exostosin glycosyl transferase 1; EXT2, exostosin glycosyl transferase 2.

Discussion

The MO family investigated in the present study exhibited a ratio of male and female patients at 5.5:1, which differs from the ratio of 1.5:1 reported previously (26). Therefore, the present study performed a detailed physical examination of all healthy females in the 3rd to 5th generation, to rule out misdiagnosis due to a weak phenotype. It was concluded that the difference may be due to a unique family structure.

Linkage analysis has confirmed that HME is genetically heterogeneous, and the genes that have indicated the greatest levels of association are EXT1, EXT2 and EXT3. It has been revealed that mutations in EXT1 or EXT2 are responsible for the majority of HME cases. The proteins EXT1 and EXT2 form a hetero-oligomeric complex that functions in HSPG biosynthesis. This complex has a substantially greater glycosyltransferase activity than homo-oligomers of EXT1 or EXT2 (27); therefore, a mutation present in EXT1 or EXT2 may result in a critical reduction in HSPG (9,28,29). This may subsequently alter the balance of fibroblast growth factor and Indian hedgehog homolog signals (30–32). Thus, the normal signaling pathway involved in bone development may be affected, leading to premature differentiation of cartilage, cartilage cell proliferation and abnormal bone growth in the adjacent areas (33), resulting in HME.

According to the MO Mutation Database (medgen.ua.ac.be/LOVDv.2.0/home.php), 713 mutations of the EXT1 gene and 386 mutations of the EXT2 gene have been identified; however, no mutations in EXT3 have been reported. Among the 386 EXT2 mutations, the majority are nonsense mutations, followed by frameshift and substitution mutations. Mutations occur primarily in exons 2–8, seldom occurring downstream. The present study revealed a novel splicing mutation (C.939+1 G>T) leading to deletion of 196 bp in exon 5 of EXT2, which may result in a truncated and subsequently pathogenic protein. The EXT2 gene encodes a protein 718 amino acids in length. This mutation resulted in deletion from codon 744 to 939 of exon 5 of the mRNA, causing a shift in the codon-reading frame, followed by the synthesis of 266 novel amino acids that terminate with a stop codon at position 994. Certain studies have suggested that no mutations exist downstream of exon 8; however, according to the Human Gene Mutation Database (www.hgmd.cf.ac.uk/ac/index.php), 3 patients with mutations in exon 10 and 4 patients with mutations in exon 11 have been identified. Therefore, mutations in the last 6 exons of EXT2 are very rare. The truncated protein that arises due to the C.939+1 G>T mutation in the family investigated in the present study did not contain the amino acids encoded for by exon 7 to l4. This alteration may cause disease; however, whether it is associated with the rare large osteochondroma that occurred on proband remains to be elucidated. The present study suggested that Knudson's two hit hypothesis or potential mutations in EXT3 or EXTL may explain the occurrence of the osteochondroma.

In conclusion, the present study demonstrated that the C.939+1 G>T (EXT2) mutation, present in a five-generation 33-member MO family, resulted in the splicing out of exon 5. These results have extended the mutational spectrum of EXT2.

References