Abstract
Spondyloepimetaphyseal dysplasias (SEMDs) comprise a heterogeneous group of autosomal-dominant and autosomal-recessive disorders. An apparent X-linked recessive (XLR) form of SEMD in a single Italian family was previously reported. We have been able to restudy this family together with a second family from Korea by segregating a severe SEMD in an X-linked pattern. Exome sequencing showed missense mutations in BGN c.439A>G (p.Lys147Glu) in the Korean family and c.776G>T (p.Gly259Val) in the Italian family; the c.439A>G (p.Lys147Glu) mutation was also identified in a further simplex SEMD case from India. Biglycan is an extracellular matrix proteoglycan that can bind transforming growth factor beta (TGF-β) and thus regulate its free concentration. In 3-dimensional simulation, both altered residues localized to the concave arc of leucine-rich repeat domains of biglycan that interact with TGF-β. The observation of recurrent BGN mutations in XLR SEMD individuals from different ethnic backgrounds allows us to define “XLR SEMD, BGN type” as a nosologic entity.
Main Text
Spondyloepimetaphyseal dysplasias (SEMDs) are characterized by anomalies of the spine and the epiphyses and metaphyses of the long bones, resulting in short stature and osteoarthritic changes of the joints. In 1994, X-linked SEMD (MIM: 300106) was described by Camera et al.1 in eight male individuals from six generations of an Italian family. Recently, we encountered a Korean family in which male siblings and their grandfather exhibited radiological features of SEMD. The older brother of the male siblings was suspected to have pseudoachondroplasia (MIM: 177170); however, no mutation was found in COMP (MIM: 600310). An Indian person subsequently referred for radiological diagnosis showed the phenotype similar to those of Italian and Korean individuals.
We analyzed Korean (Figure 1A) and Italian (Figures 1B) families with multiple affected males and one simplex Indian individual (Figure 1C) (see case reports in Supplemental Data). All affected individuals were born at term from healthy non-consanguineous parents. The birth weights and birth lengths were within the normal range. All persons came to medical attention between 12 and 24 months of age because of growth retardation with body disproportion (Table 1). They showed resomelic shortening of the limbs and short limbs-to-trunk ratio. The affected individuals showed significant bowing of the legs, a waddling gait with lumbar lordosis, and brachydactyly. Their facial appearance, teeth, and sclerae were unremarkable, as were their cognitive development and intelligence. The data from biological tests (i.e., routine blood cell count, blood and urinary levels of calcium, phosphate, creatinine, serum alkaline phosphatase, 25-hydroxyvitamin D, parathyroid hormone, and urine analysis) were within the normal range. The originally described1 Italian individuals (V-8, V-9, and VI-6) have been regularly visited since 1994. Individual VI-6 underwent 15 orthopedic procedures to lengthen both the femora and the tibiae since adolescence, when his height was 122 cm, and his final height after successful orthopedic procedures was 160 cm. His gait was almost normal, and he could swim. Obligate carriers V-6 and V-10 showed no specific health problems except hypertension in later adulthood. A picture of the Italian family taken in 2015 is shown in Figure S1. Adult Korean and Italian individuals were relatively healthy except for hypertension after the age of 60 years. They had life-sustaining occupations, enjoyed hiking and bicycling, walked without support, and had no symptoms or signs of early arthritis or neurological issues. The Indian individual (III-4) was 10 years old at the time of our evaluation. His height was 103 cm (−5.58 SDS). He had no family members showing severe short stature.
Figure 1.
Pedigree of Three Families Showing the Segregation of BGN Mutant Alleles
Shown are the pedigrees of the Korean family (A), the Italian family (B), and the Indian individual (C). Open boxes represent healthy males, and open circles represent healthy females. Filled boxes represent affected males. Boxes or circles with a line across indicate that the person has died. All circles with a dot in the middle indicate the status of carrier. Alleles with the wild-type genotype are indicated by a plus sign. Samples were not available for individuals lacking a genotype designation.
Table 1.
Clinical Findings of the Affected Individuals with XLR SEMD BGN Type
|
Family |
F1 (Korean) |
F2 (Italian) |
F3 (Indian) |
||||||
|---|---|---|---|---|---|---|---|---|---|
| Subject | IV-1 | IV-2 | II-1 | II-2 | II-6 | V-8 | V-9 | VI-2 | III-4 |
| Age at onset of symptoma | 12 months | 15 months | <2 years | <2 years | <2 years | 18 months | 18 months | 18 months | around 2 years |
| Age at exam | 5 years 1 month | 2 years 9 months | 72 years | 70 years | 60 years | 52 years (73 yearsb) | 50 years (51 yearsb) | 25 years (46 yearsb) | 10 years |
| Height at exam (cm) (SD) | 90 (−4.17) | 79.6 (−3.56) | 115 (−9.67) | 113 (−10) | 115 (−9.67) | 119 (−8.98) | 125 (−7.95) | 131 (−6.91)c | 103 (−5.58) |
| Weight at exam (kg) (SD) | 17.1 (−0.78) | 13.5 (−0.83) | ND | ND | ND | 68 (0.51) | 58 (−0.55) | ND | 30 (−1.15) |
| Mesomelic short limbs | + | + | + | + | + | + | + | + | + |
| Genu varum | + | + | + | + | + | + | + | + | + |
| Lumbar lodosis | + | + | + | + | + | + | + | ND | − |
| Brachydactyly | + | + | + | + | + | + | + | + | + |
| Joint laxity | + | + | + | + | + | − | − | − | − |
| Waddling gait | + | + | + | + | + | + | + | + | + |
| Episode of fractures | − | − | − | − | − | − | +d | +e | − |
| Current medication | − | − | AHD | AHD | AHD | AHD | AHD | − | − |
| Intelligence | N | N | N | N | N | N | N | N | N |
| BGN mutation | c.[439A>G];[0] | c.[439A>G];[0] | ND | c.[439A>G];[0] | c.[439A>G];[0] | c.[776G>T];[0] | c.[776G>T];[0] | c.[776G>T];[0] | c.[439A>G];[0] |
All affected individuals had no hypotonia, limitation of movement, joint pain during childhood, early degenerative osteoarthritis, dentinogenesis imperfecta, hyperextensibility, visual difficulty, hearing difficulty, or neurologic abnormalities. Abbreviations are as follows: N, normal; ND, not described; SDS, standard deviation score; DI, dentinogenesis imperfecta; AHD, antihypertensive drug.
Disproportionately short stature with markedly bowed legs.
Follow up examination in 2015.
160 (−1.91) cm after surgery.
Fracture of femur secondary to accidental trauma at 60 years.
Various fractures after lengthening of the lower limbs.
Radiographs of the Korean and Indian boys showed platyspondyly with central protrusion of the anterior vertebral bodies, kyphotic angulation, and increased lumbar lordosis. Radiographs showed flared ilia with horizontal and irregular acetabular roofs and long and constricted femoral necks (Figures 2A and 2C). The long bones were very short with dysplastic epiphyses and flared, irregular, and cupped metaphyses. Uniform shortness of the metacarpals and phalanges with coning at the metacarpals was seen in affected children and adults (Figure 2B). These findings were similar to those in the previous report.1
Figure 2.
Radiologic Characteristics of Individuals with XLR SEMD BGN Type
(A) Korean family member IV-1’s radiographs at age 5 years. Spine lateral shows platyspondyly with beaking appearance of its anterocentral portion and accentuated lordotic spinal curve. Note brachymetacarpals and phalanges and delayed carpal bone ages. Metaphyseal flaring and cupping are noted at the distal radii and ulnae. Pelvis shows flared iliac wings with flat acetabuli. The femoral necks are constricted and elongated. Small and dysplastic capital femoral epiphyses are seen. Note marked metaphyseal flaring and cupping of the lower extremity.
(B) Korean family member II-2, grandfather of IV-1, radiographs at age 70 years. Lateral spine shows platyspondyly and anterior wedging at L1 and L2 vertebral bodies, with kyphosis at the thoracolumbar spine. Hands show uniform brachydactyly, especially brachymetacarpals. Pelvis reveals flared iliac wings and constricted femoral necks. Hip joints are preserved for his age. Marked shortening of the femora, tibiae, and fibulae with metaphyseal widening are noted. Knee joints are intact.
(C) Indian family 3, a simplex case, radiographs at 10 years of age. Lateral spine shows platyspondyly with residue of anterior beaking. Wedging is noted at L1 vertebral body. Pelvis shows flared iliac wings with flat and irregular acetabuli. Femoral necks are constricted and elongated, and dysplastic capital femoral epiphyses with lateral subluxation are seen. Metaphyseal flaring and cupping are noted at the distal femoral and proximal tibiae with impinging epiphyses within cupping metaphyses. Hands show uniformly short metacarpals and phalanges and markedly delayed bone age.
To identify the pathogenic mutation, we screened the germline genetic variations of the three families with array comparative genome hybridization (aCGH). Using aCGH, we examined copy-number alteration in the affected individuals and selected unaffected family members (IV-1, III-1, and III-2 of the Korean family; V-8, V-9, and V-10 of the Italian family; and III-4, II-3, and II-4 of the Indian family), but no candidate gross aberration was detected. Next, whole-exome sequencing (WES) was conducted to identify the causal variants in nine members of two different families (Korean and Italian). Written informed consent was obtained from the individuals or their parents, and the Institutional Review Board approved this study (IRB file number: 2014-09-095-002).
The detailed methods of WES and the quality of WES data are summarized in Tables S1 and S2. Based on the hypothesis that this disease is inherited in an XLR fashion, we eliminated nonpathogenic variants with our own script according to the following conditions (Table S3): (1) variants showing allele frequency more than 1% in ESP6500 or 1000 Genomes Project; (2) variants found in our in-house controls (n = 452); (3) synonymous amino acid changes; and (4) low quality of reads (read number < 20, QS < 30, or minor allele frequency < 20%).
Considering the clinical and radiological findings of two families, skeletal dysplasia with severe short stature such as SEMD aggrecan type (MIM: 612813), Dyggve-Melchior-Clausen disease (MIM: 223800), acromesomelic dysplasia, Maroteaux type (MIM: 602875), pseudoachondroplasia, and spondyloepiphyseal dysplasia, Maroteaux type (MIM: 184095) should be checked. No significant variation was found in genes associated with these disorders (ACAN, DMY, NPR2, COMP, and TRPV4) in affected individuals; moreover, mutations in these genes are transmitted by autosomal-dominant or -recessive mode.
In the Korean family, among the six genes found in all affected individuals, one variant, c.439A>G (p.Lys147Glu) in BGN (RefSeq: NM_001711.4), was identified as a possible XLR candidate after segregation analysis using other Korean family members’ blood samples. In the Italian family, the two genes carried variants in all affected members, but one variant, c.776G>T (p.Gly259Val) in BGN, was identified as a possible candidate after a segregation test. After combining the variants selected by the XLR mutation model, the gene shared by individuals from two families was found to be BGN (Table S3). Interestingly, the same BGN variant as was found in Korean individuals, c.439A>G (p.Lys147Glu), was identified de novo in the Indian person. Neither of these two variants in BGN was found in 904 control chromosomes from individuals of Korean descent in addition to 800 control chromosomes from individuals of Japanese descent. These variants were not observed in the ExAC database. The two variants in BGN have been validated by Sanger sequencing of three different ethnic family members (Figure S2).
Biglycan, a member of the class I family of small leucine-rich proteoglycans (SLRPs) in the extracellular matrix, is highly expressed in bone.2, 3 It is an important structural component of articular cartilage and participates in the assembly of the chondrocyte extracellular matrix through the formation of protein interactions with type VI collagen and large proteoglycan aggregates. Biglycan also plays a role in cell signaling, and it is thought to play a critical role in the regulation of chondrocyte and pericellular matrix homeostasis.4 Its disruption affects postnatal skeletal growth and bone formation5 by disturbing key signaling pathways and the activity of the multifunctional cytokines TGF-β,6 bone morphogenic protein 4 (BMP4),7, 8 and WNT.9 In addition, BGN regulates osteoblast differentiation mediated by the extracellular signal-regulated kinase (Erk)-activated Runx2 pathway and activation of SMAD1/5/8 pathway.10 On this basis, abnormalities of biglycan are expected to cause a skeletal dysplasia. This has now been delineated. In accordance with these findings, Bgn-deficient mice (bgn−/0) show a reduced growth rate and low bone mass that is not expressed at birth but becomes obvious with age.5 X-ray images show short tubular bones with decreased radiodensity compared with age-matched controls. Histological analysis reveals reduced trabeculation and cortical thickness.5 The comparison of phenotypes between affected humans with BGN mutation and Bgn knockout mice is shown in Table S4.
Biglycan has two unique structural motifs: cystein repeated sequence and leucine-rich repeat (LRR) defined as a consensus sequence with leucines in conserved positions.11 The identified variant residues (Lys147 and Gly259) are located in LRR and conserved in decorin, which is in the same class I SLRP family as biglycan (Figure 3A). The two residues are conserved in LRR motif in different species as well as in different SLRP proteins. To determine the structural properties of biglycan and its mutated protein, we created a 3-dimensional (3D) model of human biglycan using the crystal structure of bovine biglycan as a template. We assumed that there are two possible scenarios that could explain the pathogenicity of two substitutions, p.Lys147Glu and p.Gly259Val. One scenario consists of the decrease of biglycan stability caused by physical impact of the point mutation on the overall biglycan structure. The other predicts the interference of the substitution with the binding interaction between biglycan and TGF-β, which is known to bind with biglycan.12, 13
Figure 3.
Pathogenic Variant in BGN and Molecular Interactions between Biglycan and TGF-β
(A) The schematic diagram of BGN and the position of two identified variants. The two BGN variants (c.439A>G [p.Lys147Glu] and c.776G>T [p.Gly259Val]) are located in the leucine-rich repeat (LRR) domain. The asterisk (∗) marks conserved amino acid residue in biglycan (BGN) and decorin (DCN). Boxed amino acid residues indicate conserved functional leucine in the LRR domain.
(B) Structural change brought about two substitutions (p.Lys147Glu and p.Gly259Val) predicted by molecular dynamics simulation. Arrows at merged structures indicate the place where beta sheets at the biglycan (BGN) concave region are significantly perturbed by mutations.
(C) Structural model of BGN (green) and TGF-β (blue) dimer. Red functional residues indicate the substituted sites (p.Lys147Glu).
(D) Structure simulation models for BGN p.Lys147Glu with TGF-β in overall and interface close-up view. Yellow dotted lines in close-up view show polar interactions between functional groups around Glu147. Abbreviations are as follows: WT, wild-type biglycan; SUB, substituted biglycan.
To investigate the decrease of protein stability, each substituted monomer model was analyzed with Gromacs Molecular Dynamics (MD) simulation. The visualization of the altered residues on 3D structures is shown for the two substitutions. To evaluate the probability of interference with binding interaction, we performed a global docking simulation between biglycan and TGF-β to build a complex model. The complex models were compared between wild-type and the two variants. If the position of the substitution did not appear to be located at the binding interface between biglycan and TGF-β (as in the case of p.Gly259Val), we also performed MD simulation for the complex model.
According to the result of the MD simulation, both p.Lys147Glu and p.Gly259Val substitutions have impacts on biglycan structure (Figure 3B). Several side chains around the mutation become translocated in MD simulation. In addition, a few beta sheets on the concave region are not maintained. The complex model suggests that the residues in which substitutions (p.Lys147Glu and p.Gly259Val) occur are located in the concave region of biglycan and that biglycan binds to TGF-β through its concave region (Figure 3C). In particular, biglycan interacts with TGF-β through the Lys147 residue of biglycan through several polar interactions (Figure 3D). The p.Lys147Glu substitution causes a considerable polarity change because lysine has a positive-charged side chain and glutamic acid has a negative-charged side chain. We found that polar interactions between biglycan and TGF-β are removed when p.Lys147Glu substitution has occurred (Figure 3D). These results indicate that p.Lys147Glu substitution might reduce the binding affinity between biglycan and TGF-β by changing the interface polarity of biglycan and interfering with polar interactions with TGF-β. The position of the p.Gly259Val substitution does not seem to participate directly in the binding interaction (Figure S3). However, it might change the overall biglycan structure, disturbing binding interaction with TGF-β. The MD simulation performed on the complex model showed that p.Gly259Val does not interact with TGF-β (Figure S3). The binding between biglycan and TGF-β is specific for TGF-β out of several other growth factors, which suggests an important role for these small proteoglycans in regulating the activity of TGF-β in the synthesis of matrix components.13 In addition, there might be some matrix disruption caused by biglycan dysfunction because biglycan is an important constituent of the matrix and organizes the assembly of extracellular matrix by binding to collagen fibrils.7, 10
To investigate the functional consequences of the BGN mutation, dermal fibroblasts were obtained from the Korean family. Variant biglycan (p.Lys147Glu) was expressed in lower levels in an affected person (IV-1) than in an unaffected person’s dermal fibroblasts (III-2) (Figure S4A). The transcription of variant BGN mRNA did not show the differences between two fibroblasts. On the other hand, biglycan was degraded more rapidly in IV-1’s fibroblasts than in III-2’s fibroblasts (Figure S4B). Missense mutation can lead to the misfolding of variant proteins and reduction of the protein stability.14 The decreased protein level of variant biglycan in affected person’s cells might alter the extracellular signaling of TGF-β, which induced the phenotype of affected individuals.
To conclude, we describe XLR SEMD and define its biglyan-related molecular basis in individuals from three different ethnic backgrounds. Structural simulation studies of the two biglycan substitution indicate functional alterations in both cases. Further clinical, genetic, and biological studies are necessary to understand the pathological mechanism of the disease. A more detailed understanding of the physicochemical and structural properties of the binding sites through which biglycan interacts with its various binding partners will be a critical component of this effort.
Acknowledgments
This study was supported by a grant from Samsung Medical Center (#GFO1130061 to D.-K.J.) and the Korean Health Technology R&D Project, Ministry of Health and Welfare, Republic of Korea (Hl12C0014 to W.-Y.P.). The Galliera Genetic Bank member of “Network Telethon of Genetic Biobanks” (project no. GTB12001A) and of EuroBioBank Network provided us with specimens. The authors express gratitude to the family members in the study, especially the Italian family who allowed their group photo to be included.
Published: May 26, 2016
Footnotes
Supplemental Data include four figures, four tables, and case reports of individuals from three different ethnic backgrounds and can be found with this article online at http://dx.doi.org/10.1016/j.ajhg.2016.04.004.
Contributor Information
Woong-Yang Park, Email: woongyang.park@samsung.com.
Dong-Kyu Jin, Email: jindk@skku.edu.
Web Resources
Ensembl Genome Browser, http://www.ensembl.org/index.html
GraphPad, http://graphpad.com/
NHLBI Exome Sequencing Project (ESP) Exome Variant Server, http://evs.gs.washington.edu/EVS/
OMIM, http://www.omim.org/
PolyPhen-2, http://genetics.bwh.harvard.edu/pph2/
RCSB Protein Data Bank, http://www.rcsb.org/pdb/home/home.do
UCSC Genome Browser, http://genome.ucsc.edu
Supplemental Data
References
- 1.Camera G., Stella G., Camera A. New X linked spondyloepimetaphyseal dysplasia: report on eight affected males in the same family. J. Med. Genet. 1994;31:371–376. doi: 10.1136/jmg.31.5.371. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Bianco P., Fisher L.W., Young M.F., Termine J.D., Robey P.G. Expression and localization of the two small proteoglycans biglycan and decorin in developing human skeletal and non-skeletal tissues. J. Histochem. Cytochem. 1990;38:1549–1563. doi: 10.1177/38.11.2212616. [DOI] [PubMed] [Google Scholar]
- 3.Ameye L., Young M.F. Mice deficient in small leucine-rich proteoglycans: novel in vivo models for osteoporosis, osteoarthritis, Ehlers-Danlos syndrome, muscular dystrophy, and corneal diseases. Glycobiology. 2002;12:107R–116R. doi: 10.1093/glycob/cwf065. [DOI] [PubMed] [Google Scholar]
- 4.Iacob S., Cs-Szabo G. Biglycan regulates the expression of EGF receptors through EGF signaling pathways in human articular chondrocytes. Connect. Tissue Res. 2010;51:347–358. doi: 10.3109/03008200903427695. [DOI] [PubMed] [Google Scholar]
- 5.Xu T., Bianco P., Fisher L.W., Longenecker G., Smith E., Goldstein S., Bonadio J., Boskey A., Heegaard A.M., Sommer B. Targeted disruption of the biglycan gene leads to an osteoporosis-like phenotype in mice. Nat. Genet. 1998;20:78–82. doi: 10.1038/1746. [DOI] [PubMed] [Google Scholar]
- 6.Bi Y., Stuelten C.H., Kilts T., Wadhwa S., Iozzo R.V., Robey P.G., Chen X.D., Young M.F. Extracellular matrix proteoglycans control the fate of bone marrow stromal cells. J. Biol. Chem. 2005;280:30481–30489. doi: 10.1074/jbc.M500573200. [DOI] [PubMed] [Google Scholar]
- 7.Chen X.D., Fisher L.W., Robey P.G., Young M.F. The small leucine-rich proteoglycan biglycan modulates BMP-4-induced osteoblast differentiation. FASEB J. 2004;18:948–958. doi: 10.1096/fj.03-0899com. [DOI] [PubMed] [Google Scholar]
- 8.Young M.F., Bi Y., Ameye L., Chen X.D. Biglycan knockout mice: new models for musculoskeletal diseases. Glycoconj. J. 2002;19:257–262. doi: 10.1023/A:1025336114352. [DOI] [PubMed] [Google Scholar]
- 9.Berendsen A.D., Fisher L.W., Kilts T.M., Owens R.T., Robey P.G., Gutkind J.S., Young M.F. Modulation of canonical Wnt signaling by the extracellular matrix component biglycan. Proc. Natl. Acad. Sci. USA. 2011;108:17022–17027. doi: 10.1073/pnas.1110629108. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Wang X., Harimoto K., Xie S., Cheng H., Liu J., Wang Z. Matrix protein biglycan induces osteoblast differentiation through extracellular signal-regulated kinase and Smad pathways. Biol. Pharm. Bull. 2010;33:1891–1897. doi: 10.1248/bpb.33.1891. [DOI] [PubMed] [Google Scholar]
- 11.Hocking A.M., Shinomura T., McQuillan D.J. Leucine-rich repeat glycoproteins of the extracellular matrix. Matrix Biol. 1998;17:1–19. doi: 10.1016/s0945-053x(98)90121-4. [DOI] [PubMed] [Google Scholar]
- 12.Chen G., Deng C., Li Y.P. TGF-β and BMP signaling in osteoblast differentiation and bone formation. Int. J. Biol. Sci. 2012;8:272–288. doi: 10.7150/ijbs.2929. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Hildebrand A., Romarís M., Rasmussen L.M., Heinegård D., Twardzik D.R., Border W.A., Ruoslahti E. Interaction of the small interstitial proteoglycans biglycan, decorin and fibromodulin with transforming growth factor beta. Biochem. J. 1994;302:527–534. doi: 10.1042/bj3020527. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Bross P., Corydon T.J., Andresen B.S., Jørgensen M.M., Bolund L., Gregersen N. Protein misfolding and degradation in genetic diseases. Hum. Mutat. 1999;14:186–198. doi: 10.1002/(SICI)1098-1004(1999)14:3<186::AID-HUMU2>3.0.CO;2-J. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.