Heterogeneous spectrum of EXT gene mutations in Chinese patients with hereditary multiple osteochondromas

Abstract

Hereditary multiple osteochondroma (HMO) is one of the most common genetic skeletal disorders. It is caused by mutations in either EXT1 or EXT2 resulting in abnormal skeletal growth and morphogenesis. However, the spectrum and frequency of EXT1 and EXT2 mutations in Chinese patients with HMO was not previously investigated.

Mutations were identified by performing Sanger sequencing analysis of the complete coding regions and flanking intronic sequences of EXT1 and EXT2, followed by multiplex ligation-dependent probe amplification (MLPA) analysis to detect gene deletions or duplications that could not be identified by the Sanger sequencing method.

The present study identified pathogenic mutations in 93% (68/73) of unrelated HMO probands from 73 pedigrees. Mutations in EXT1 and EXT2 were identified in 53% (39/73) and 40% (29/73) of families. We identified 58 distinct mutations in EXT1 and EXT2, including 20 frameshift mutations, 16 nonsense mutations, 7 missense mutations, 9 splice site mutations, 5 large deletions, and 1 in-frame deletion mutation. Twenty-six of these mutations were novel and 32 were previously reported. Most of the mutations in EXT1 were base deletions or insertions (21/33), whereas the majority of those in EXT2 were single base substitution (18/25).

Complete sequencing of both the EXT1 and EXT2 followed by MLPA analysis is recommended for genetic analysis of Chinese patients with HMO. This study provides a comprehensive characterization of the genetic aberrations found in Chinese patients with HMO and highlights the diagnostic value of molecular genetic analysis in this particular disease.

1 Introduction

Hereditary multiple osteochondroma (HMO), previously known as hereditary multiple exostoses (HMEs), is characterized by the growth of multiple osteochondromas (benign, cartilage-capped bone tumors) mainly in the metaphyses and diaphyses of long bones.^[1,2] Patients with HMO may present a variety of orthopedic deformities such as malformations of the forearm, inequality in limb length, varus or valgus angulation of the knee, deformity of the ankle, and disproportionate short stature.^[3] The risk of malignant degeneration to osteochondrosarcoma increases with age, although the lifetime risk of malignant degeneration is low (∼1%).^[1,2] The prevalence of HMO is estimated to be at least 1 in 50,000 in the general population^[4] and seems to be higher in males (male to female ratio 1.5:1).^[3]

The HMO is inherited in an autosomal dominant manner.^[1] Two genes, EXT1 (OMIM: 608177) and EXT2 (OMIM: 608210), which are located respectively at 8q24.11 and 11p11.2, have been identified to cause HMO.^[5–7] These genes encode homologous, Golgi-associated glycosyltransferases that are involved in the chain elongation step of heparan sulfate (HS) biosynthesis.^[8,9] Studies have indicated that HS influences various important processes in skeletogenesis, skeletal growth, and morphogenesis.^[8,9] Loss-of-function mutations in EXT1 and EXT2 lead to HS deficiency.^[8,9] Hundreds of mutations have been reported in the Human Gene Mutation Database (HGMD, http://www.hgmd.cf.ac.uk/ac/search.html) and Multiple Osteochondromas Mutation Database (MOdb, http://medgen.ua.ac.be/LOVD). Although most reported mutations involve one or several base changes in patients with HMO, the Sanger method of DNA sequencing is unable to detect all types of mutations.

In the present study, Sanger sequencing and multiplex ligation-dependent probe amplification (MLPA) were employed to determine the spectrum of mutations in EXT1 and EXT2 of 73 Chinese patients diagnosed with HMO.

2 Patients and methods

2.1 Patients

The study was approved by the ethics committee of Shanghai Children's Medical Center (SCMC), and written informed consent was obtained from the patients’ parents or guardians. Patients were initially diagnosed on the basis of radiology, and genetic analysis of the patients was performed subsequently. A personal medical history and physical examination were conducted. The diagnosis of HMO was confirmed by clinical and radiographic findings of multiple exostoses (arising from the area of the growth plate in the juxtaphyseal region of long bones or from the surface of flat bones), with the majority of cases have a positive family history.^[1–4]

2.2 Sanger method of DNA sequencing

Genomic DNA was extracted from peripheral blood samples of the probands and family members using a QIAamp Blood DNA mini kit (Qiagen GMBH, Hilden, Germany). Primers used in the amplification of EXT1 and EXT2 (GenBank accession numbers NM_000127.2 and NM_207122.1, respectively) were designed using the Primer3 online software (primer sequences are listed in Table 1). All exons and exon–intron boundaries of each gene were amplified by polymerase chain reaction (Takara Bio, Dalian, China). The amplified products were purified from an agarose gel using a QIAquick Gel extraction kit (Qiagen GMBH) and sequenced via an ABI3730XL sequencer (Applied Biosystems, Foster City, CA). An additional group of 105 Chinese patients without skeletal deformities were recruited as ethnicity-matched controls to examine the allele frequencies of various EXT1 and EXT2 sequence variants.

Table 1: Sanger sequencing primers for EXT1 and EXT2 genes.

2.3 Copy number variation analysis

The MLPA analysis was performed according to the manufacturer's protocol using the “SALSA MLPA probemix P215-B2 EXT” kit (MRC Holland, Amsterdam, the Netherland). The probe mix included in this kit contains 41 different probes with amplification products between 130 and 453 bp, including 13 specific probes for EXT1, 16 specific probes for EXT2, and 12 reference probes for the control. Data analysis and interpretation were performed using GeneMarker software (Softgenetics, State College, PA). This program identifies peaks as “normal” when the ratio of sample peak height or area to that of control samples falls in the range from 0.75 to 1.33; as “deleted” when the ratio is <0.75, and as “duplicated” when the ratio is >1.33.

2.4 In silico prediction of the novel missense mutations

In silico prediction of the identified variant was performed using prediction tools. Alamut Visual 2.7.1 (http://www.interactive-biosoftware.com/alamut-visual/, which included tools such as SIFT, PolyPhen-2, and MutationTaster) was used for the identification of novel missense mutations.

3 Results

3.1 Mutations identified

Between August 2008 and December 2015, 73 patients from different families with HMO were recruited at Shanghai Children's Medical Center (SCMC) China, including 50 boys and 23 girls; the mean age was 7.38 years (range 4.8–15.1 years). Sanger sequencing and MLPA were used for genetic analysis and revealed mutations in 93% of these probands (68/73; Table 2). We detected 33 different EXT1 mutations in 39 families and 25 different EXT2 mutations in 29 families. Among the 58 EXT1/EXT2 mutations, 26 were novel, whereas 32 had been previously reported. The mutation types included 20 frameshift (16 in EXT1 and 4 in EXT2), 16 nonsense (3 in EXT1 and 13 in EXT2), 7 missense (3 in EXT1 and 4 in EXT2), 9 splice site (5 in EXT1 and 4 in EXT2), 5 large deletions (4 in EXT1 and 1 in EXT2), and 1 small in-frame deletion (EXT1) (Fig. 1).

Table 2: Spectrum of EXT1 and EXT2 mutations in Chinese patients with hereditary multiple osteochondromas.

Figure 1: Example of MLPA detection results in EXT1/EXT2. Heterozygous deletion of the entire EXT1 in patient 30. (A) Distribution of the peak ratio of all probes. (B) The peak ratio values of control probes. (C) The peak ratio values of EXT1 probes. (D) The peak ratio values of EXT2 probes.

3.2 Pathologic prediction of novel missense mutations

Of the 26 novel mutations, 24 affected the length of the EXT protein by introducing a premature stop codon or changing the conserved splice site positions and thus were considered pathogenic. These mutations were not detected in 307 unrelated control individuals, nor were they reported in the Exome Aggregation Consortium (ExAC) database (http://exac.broadinstitute.org/).

In the case of the 2 novel amino acid substitutions (EXT1: c.1930A > T; p.Asn644Tyr; EXT2: c.1072T > C; p.Trp358Arg), the possibility that the changes represented nonpathologic polymorphisms was examined. The effect of amino acid changes on protein stability was examined using Alamut software. The prediction results are listed in Table 3. Both variants were predicted to be harmful to protein function with different in silico tools. In addition, the 2 patients’ parents were wild type indicating the de novo status of the 2 variants and offering strong pathogenic evidence for their causal relationship to the disease.

Table 3: In silico evaluation of novel missense mutations in the EXT1 and EXT2 genes.

4 Discussion

The HMO is an autosomal, dominant, benign tumor that is characterized by abnormal growth of long bone cartilaginous caps (osteochondroma). It is one of the most common skeletal dysplasias in adolescents. About 80% of patients with HMO develop clinical symptoms before 10 years of age and it can cause various complications, leading to skeletal deformities and short stature.^[1,10,11]

The EXT family of genes has been implicated in the pathogenesis of HMO. At least 5 members have been cloned and defined. EXT1 (NM_000127.2) consists of 11 exons that encode a 746-amino-acid protein, whereas EXT2 (NM_207122) comprises 14 exons that encode a 718-amino-acid protein. EXT1 and EXT2 are widely expressed genes and are highly homologous, particularly in terms of the carboxyl terminal sequences.^[12] HS, which is modified by EXT1 and EXT2 proteins, is generally distributed on the cell surface and the extracellular matrix. As a co-enzyme, HS is involved in cell adhesion, blood coagulation, and angiogenesis, as well as in the regulation of cell growth factors and various other biologic processes. In addition, based on its sulfate-modifying function, HS is involved in the distribution of various signal proteins in target cells, including a bone formation protein family (bone morphogenetic proteins), that are expressed in the bone growth plate. Research evidence suggests that mutations in either EXT1 or EXT2 may trigger the abnormal synthesis of the HS chain, which in turn leads to the development of HMO.^[8,13–16] Moreover, 3 EXT-like genes (EXTL1, EXTL2, and EXTL3) are also involved in HS synthesis and are believed to be correlated with HMO, although no mutations of these have been detected in patients with HMO.^[17,18]

Previous reports indicated that mutations in either EXT1 or EXT2 are responsible for nearly 90% of HMO cases, of which the majority involve EXT1 (60–70%). EXT1 mutations are more often found in the first 6 exons, whereas those of EXT2 involve the first 8 exons. Mutations in EXT1 are the most frequent cause of HMO in Europe and North America.^[19,20] The reported mutation frequencies of EXT1 and EXT2 are apparently similar (39/68 vs 29/68), yet differ from those observed in 36 Chinese families (14% in EXT1 and 33% in EXT2).^[21] These findings suggest that regional differences in EXT1 and EXT2 mutation frequencies occur in China. Moreover, the use of MLPA has dramatically increased the detection of mutations: it can identify large fragment deletions and duplications based on hybridization and ligation, followed by amplification of the ligation products.

To date, 463 mutations in EXT1 have been reported, including 154 missense/nonsense mutations, 223 small insertion/deletion mutations, 50 splice site mutations, and 36 other types, whereas mutations in EXT2 consist of 71 missense/nonsense mutations, 94 small insertion/deletion mutations, 24 splice site mutations, and 27 other types (data from HGMD). The percentage of small insertion/deletion mutations is higher than others in both EXT1 and EXT2. In the present study, most of the detected EXT1 mutations were base deletions or insertions (21/33). In contrast, the major mutation type in EXT2 involved single-base substitution (18/25). These findings add new insights into the spectrum of EXT mutations in Chinese patients with HMO. Furthermore, we detected 26 novel mutations in EXT1 and EXT2, which may be utilized in the clinical diagnosis of HMO.

Similar to most other single-gene inherited diseases, most of the mutations in EXT1 and EXT2 in HMO are caused by changes in a single or several bases, including missense/nonsense, splice site, and small insertion or deletion mutations.^[11] These small mutations are identified by using the Sanger sequencing technology; however, a few large gene fragments often go undetected and thus may result in an incorrect clinical diagnosis. Here, we introduced the MLPA assay to improve our detection specificity, as it can identify large fragment deletions and duplications (copy number aberrations) based on hybridization and ligation, followed by the amplification of the ligation products.^[22–24] Eight novel large deletions in EXT1 and EXT2 were detected by Sanger sequencing and verified by using the MLPA assay, with 93% of patients with HMO receiving verified diagnostic results. The limitation of this study is that no EXT1 or EXT2 mutations were detected in 5 patients with HMO, indicating that other genes that yet have to be identified may be responsible for this particular disorder. Next generation sequencing may be the best approach in identifying the candidate genes.^[25] In addition, this study only includes 73 samples that might result in the statistics bias of the variation spectrum. Moreover, functional studies of the missense mutations need be performed in the future studies, which can better understand the pathogenicity of these mutations.

In conclusion, we have successfully detected pathogenic mutations in 93% (68/73) of unrelated probands from 73 pedigrees. A total of 58 mutations were identified, which included 26 novel mutations. The range of EXT mutations slightly varied from those previously reported in Chinese and other ethnic groups around the world. Molecular genetic analysis is useful for the clinical diagnosis and genetic counseling of patients with HMO.

Author contributions

Formal analysis: Zhigang Wang.

Methodology: Bingqiang Han.

Resources: Niu Li.

Software: Tingting Yu, Yulin Chen.

Supervision: Yuchan Li, Yulin Chen, Qihua Fu.

Validation: Qihua Fu.

Visualization: Qihua Fu.

Writing – original draft: Jian Wang.

Writing – review & editing: Jingyan Tang, Yuchan Li.

Yuchan Li orcid: 0000-0002-2605-9727.