Abstract
Several genes are implicated in the etiology of early onset osteogenesis imperfecta (OI). The various genes causing severe OI include WNT1 , SERPINF1 , P3H1 , CREB3L1 , and CRTAP , although glycine substitutions in COL1A1chains have also been predicted to cause perinatal lethal OI . Patients with early onset OI present decreased mobility, recurrent rib fractures, bony deformities, and chest infections that lead to an early death. We reported our experience in children with OI in Asian Indian families, which includes two patients with SERPINF1 pathogenic variants; and another two patients with severe OI and antenatal fractures caused by pathogenic variants in the CRTAP gene, identified by next generation sequencing (NGS). For one affected fetus, medical termination of pregnancy was done. The other baby was started on zoledronate therapy just after birth and is now 3 years old. Prenatal diagnosis was subsequently done on chorionic villus sample in the latter family.
Keywords: Brittle Bones, next generation sequencing, prolyl hydroxylation, recurrent fractures, zoledronate
Introduction
Osteogenesis imperfecta (OI) is more frequently inherited as autosomal dominant and is most often related to mutations in collagen genes. However, autosomal recessive gene mutations are being increasingly identified in severe cases of early onset OI. So far, around 15 genes have been identified in autosomal recessive or X-linked OI, being CCDC134 ––a transcriptional regulator—the latest implicated. The clinical features of OI include short stature, recurrent fractures, skeletal deformities, osteopenia, and scoliosis. Mutations in the CRTAP gene cause severe OI with prenatal and/or early postnatal fractures. 1 2 The CRTAP product is a scaffolding protein involved in the response to several cellular stimuli and importantly in bone metabolism. Bi-allelic disease-causing mutations in this gene cause severe bone fragility and reduced bone mass. Autosomal recessive gene mutations are expected to cause more severe forms of OI within affected families.
Methods
Thirty-seven children with OI registered in a genetic clinic of a tertiary care center in India were included. The patients had an age range from 1 month to 16 years. In postnatal cases, OI diagnosis was done in presence of recurrent fractures and significant osteoporosis on dual-energy X-ray absorptiometry (DXA) scan bone densitometry. Most OI patients were on zoledronate therapy and some older age patients with oral risedronate. Bisphosphonates were given to all children with OI having antenatal/neonatal fractures or recurrent fractures (>3/year). Zoledronate was given every 3 months at 2 mg/dose administered as a 30-minute to 1-hour intravenous infusion, using 4 mg/dose in older children.
Mutation analysis in some of these patients was performed by targeted next generation sequencing (NGS) technology. NGS testing was done mostly in such families who planned to have a child to facilitate prenatal diagnosis. On NGS raw data, the variant filtering and annotation were done by standard bioinformatic pipelines, and the probably causative or pathogenic variants were identified following the American College of Medical Genetics guidelines. 3 The mutations identified in these patients are described here, with focus on CRTAP gene variants.
Results
Gene variants were identified by NGS and confirmed by Sanger sequencing in 10 out of 37 OI patients, and in a fetus with OI suspected. There was consanguinity in only two out of 37 families. Two variants have been described elsewhere: 4 5 one COL1A2 gene variant and another WNT1 gene homozygous deletion. The variants identified in eight additional patients are listed in Table 1 . Two patients showed heterozygous variants in COL1A1 gene. Two patients showed homozygous SERPINF1 variant. History of consanguinity was present in one of the families with SERPINF1 variants. Another patient had compound heterozygous variants in the CRTAP gene; one inherited from their mother (delG) and another paternally inherited (dupG), which were confirmed by Sanger sequencing. One patient with distal contractures showed a FBN2 variant. Since patient's mother was also positive for this variant but did not have history of backache or fractures, it was considered as a variant of uncertain significance (VUS). Spontaneous fractures decreased significantly in the child with the CRTAP variants after zoledronate therapy, given as 3 monthly infusions. She had multiple fractures before therapy but only two fracture episodes in past 2 years. The X-rays done before the start of therapy confirmed multiple fractures in the limbs ( Fig. 1A ). The DXA scan done in the child showed decreased bone mineral density (BMD)––more pronounced in the ribs and limbs in comparison with the spine. However, DXA results in children cannot be interpreted unless there is an age-matched and size-matched control group, and serial DXA scan values may be needed in growing children, especially in OI patients. Prenatal diagnosis was done in this family using chorionic villus sample at 12 weeks for mutation analysis and maternal cell contamination (MCC) testing with appropriate consent. Mutation testing identified the same two variants in the fetus, suggestive of an affected fetus. Simultaneously, a level II fetal ultrasonography at 19 weeks of gestation revealed upper and lower limb shortening indicating 2 to 3 weeks difference in bi-parietal diameter and limb length measurements. The family opted for the termination of pregnancy. X-ray showed short limbs and generalized osteopenia ( Fig. 1B ).
Table 1. Gene variants in children with osteogenesis imperfecta.
| No. | Age/sex | Presentation | Gene variant | Condition |
|---|---|---|---|---|
| 1. | 3 years/male | Recurrent fractures from 1 year of age on zoledronate therapy |
SERPINF1
NM_002615.6: c.3 G > A: p.(Met1Ile), homozygous |
Osteogenesis imperfecta, type VI (AR) |
| 2. | 5 years/male | Recurrent fractures noted after 1 year of age |
COL1A1
NM_000088.3: c.1121G > C: p.(Gly374Ala), heterozygous |
Osteogenesis imperfecta, type IV (AD) |
| 3. | 3 years/male | Recurrent fractures on minor trauma, short stature |
COL1A1
, heterozygous
ENST00000225964: c.C1243T: p. (Arg 415 Ter) |
Osteogenesis imperfecta, type III (AD) |
| 4. | 3 years/female | Recurrent fractures from birth, antenatally detected short limbs |
CRTAP
, ENST00000320954:
c.22del: p.(Ala8ProfsTer5) and c.22dup: p.(Ala8GlyfsTer153) |
Osteogenesis imperfecta, type VII |
| 5. | 1.5 years/female | Recurrent fractures, antenatally detected short limbs, small eyes |
WNT1
NM_005430.3: c.506delG: p.(Gly169fs), homozygous |
Osteogenesis imperfecta type XV (AR) |
| 6. | 3 years /female | Recurrent fractures noted from infancy |
SERPINF1
NM_002615.6: c.72dupC: p.(Glu27GlyfsTer38) Homozygous exon 2 |
Osteogenesis imperfecta, type VI (AR) |
| 7. | 4 months/ female |
Multiple fractures in upper and lower limbs, finger fractures, low set crumpled ears |
FBN2,
heterozygous exon 39
ENST00000508053: c.4312G > A: p.(Glu1438Lys) |
Inconclusive, one parent with same variant (without fractures/osteopenia) |
| 8. | 5 years/male | Multiple fractures since birth, predominantly lower limbs |
BMP1,
homozygous, exon 12
NM_006129.5: c.1492 G > A: p.(Gly498Arg) |
Osteogenesis imperfecta, type XIII (AR) |
Abbreviations: AD, autosomal dominant; AR, autosomal recessive.
Fig. 1.

X-ray of child with cartilage-associated protein related osteogenesis imperfecta at birth ( A ), and X-ray (fetogram) of the aborted affected fetus with generalized osteopenia ( B ).
In another family, the couple came for antenatal counseling for short and bowed limbs and increased nuchal fold thickness on antenatal ultrasonography. Keeping a possibility of prenatal manifestation of OI, NGS testing was done on cord blood, identifying homozygosity for the variant c. 55–70 del (p.Cys19Glyfs*18) in the CRTAP gene. The X-ray done after termination of pregnancy confirmed the shortness and bowing of limbs, and thin beaded ribs, but autopsy was not done. Table 1 lists the different genes involved in causation of OI in the children in homozygous, heterozygous, or hemizygous state. The X-ray of one child showing fractures and bowed deformed femurs ( Fig. 2 ). Increased density at ends of bones in this child is effect of zoledronate therapy. X-rays of another child with deforming OI are shown in Fig. 3 .
Fig. 2.

X-ray of a child of type III osteogenesis imperfecta with recurrent fractures in newborn period showing hyperplastic callus ( A ) and bowed deformed femurs, thin cortices with zebra lines ( B ), showing increased density at ends due to zoledronate.
Fig. 3.

X-ray of a child with deforming osteogenesis imperfecta revealing the deformed forearm bones ( A ), osteopenia, and deformity of the leg bones including femur and tibia ( B, C ).
Discussion
Different mechanisms including abnormal bone mineralization ( COL1A1 , COL1A2 , SPARC ), disruption of osteoblast development or function ( WNT1 , SP7 ), and defective post-translational modification of collagen ( CRTAP , LEPRE1 , PPIB , PLOD2 , FKBP10 , IFITM5 ) or altered processing of collagen ( SERPINF1 , SERPINH1 , BMP1 ) can result in weak bones and increased tendency for fractures. 6 Autosomal dominant OI is commoner, and usually has later onset and is due to mutations in COL1A1 and COL1A2 genes. Perinatal lethal OI has also been reported due to specific variants especially glycine substitutions in COL1A1 gene. The gene implicated in X-linked OI is MBTPS2 on Xp22.12.
Early onset recessive osteogenesis imperfecta (OI) can be due to variants in genes like WNT1 , CRTAP , SPARC , and SP7 ( Fig. 4 ). Homozygous and compound heterozygous variants in the CRTAP gene are causative of OI type VII. 2 7 The manifestations in severe forms of OI include unprovoked fractures, deformity of limbs, limb bowing, short stature, compression fractures of vertebrae, joint hypermobility, scoliosis, and hearing loss. Bisphosphonates have given some hope of these patients, and the commonly used bisphosphonates include pamidronate and zoledronate. The latter is the more potent form and has shown clinical benefits in adults and children with osteoporosis. 8
Fig. 4.

Flow chart showing different patterns of inheritance in OI, the genes involved, and some additional features. * WNT1 mutations cause early onset osteoporosis in heterozygous state and severe early onset OI with neurological or ocular involvement in homozygous state. Since WNT1 is associated with central nervous system involvement/malformations, it may have contribution to prenatal or perinatal lethal OI. DD, differential diagnosis; OI, osteogenesis imperfecta.
Mutations in CRTAP gene cause severe osteoporosis, and hence, an increased risk of mortality in affected babies. CRTAP protein is expressed in both osteoblasts and osteoclasts, and it also interacts with prolyl-3-hydroxylase-1 and cyclophilin-B. There is co-expression of CRTAP , LEPRE1 ( P3H1 product), and PPIB . There is also possible significant interaction with SERPINF1 protein. Prolyl hydroxylation is an important post-translational modification, and the abnormal hydroxylation can affect the structure and the function of the protein. 3 Thus, bi-allelic mutations in P3H1 or CRTAP genes have similar phenotypes. A 7-year-old boy and a 12-year-old Pakistani girl have earlier been described with CRTAP -related OI. 7 9 In the girl described in present report, there was antenatal onset of limb shortening and bowing, and she was treated for fractures from birth. Interestingly, she also responded to zoledronate therapy. In an earlier report, there was progression to scoliosis even with bisphosphonate therapy at 4 years of age. 10 Starting bisphosphonate therapy at an earlier age probably may result in better outcomes. Pamidronate and zoledronate are the bisphosphonates used in therapy of OI patients with fractures. 8 These bisphosphonates are used to treat patients with deforming OI who have multiple, recurring fractures, especially if associated with pain, vertebral fractures, and requiring rodding of long bones. Additional therapies under trial are denosumab, teriparatide, and the antisclerostin antibodies, which include romosozumab and blosozumab. 11 The CRTAP variants found in this child were predicted to cause premature truncation of protein. These variants have not been described in 1,000 Genomes, ExAC, or ClinVar databases, and are predicted to be deleterious on Mutation Taster2. The 22del and 22dup CRTAP variants were described as pathogenic variants, since causing premature truncation of protein, also found damaging on in silico bioinformatic algorithms. A variant in the same region of the gene c.21_22dupGG has been described earlier in three sibs in one family. 12 The fetuses/babies in this family also had antenatal fractures and bowing. The baby, who was delivered at 39 weeks, had round face and short thorax. Early onset fractures and deformities are also seen in patients with homozygous WNT1 variants, but spine is more involved than limbs. 5 In CRTAP patients, vertebral compression fractures can be seen later in untreated patients. 1 Patient 5 in Table 1 showed the WNT1 variant, which was a likely pathogenic variant. The variant allele frequency in the gnomAD database was 0.000004. The SERPINF1 variant in patient 1 is an initiation codon VUS but found to be disease causing on Mutation Taster and damaging on scale invariant feature transform (SIFT) database. However, the SERPINF1 variant in patient 6 in Table 1 is a pathogenic variant, previously reported in ClinVar database as likely pathogenic. The BMP1 variant found in patient 8 in the Table 1 is a VUS. The variant has not been described in the 1,000-genome database, and the frequency was found to be 0.0004% in gnomAD database.
The del55–70 variant in CRTAP gene in the fetus with OI in present report has not been reported earlier, including on ClinVar database, is predicted to be damaging on bioinformatic algorithms, and hence was reported as pathogenic. Null variants are expected to be more deleterious than the missense nontruncating variants, which can lead to partly functional protein.
Thus, in our patient cohort of nine patients tested and having conclusive reports, only three showed COL1A1 or COL1A2 gene variants, but others ( n = 6) showed variants in autosomal recessive genes which included SERPINF1 , WNT1 , CRTAP , and BMP1 . Around 50 to 60% of Asian OI patients show COL1A1 and COL1A2 mutations, which include Taiwanese, Vietnamese, and Korean patients with OI in different published literature. 13 14 15 This frequency was reported to be higher (87%) in Estonian population. 16 A part of the contribution may be due to consanguinity or inbreeding within ethnic groups. In a previous report in 10 patients from Egypt, Lebanon, and Spain, three revealed FKBP10 variants, and another three showed SERPINF1 variants. 17 Two of 10 patients tested in present cohort also revealed homozygous SERPINF1 variants. Consanguinity is more common in South Indian population due to prevalence of uncle–niece marriages, but inbreeding or marriage within same caste or ethnic group is observed all over India. Few variants may be founder mutations, but this can only be clarified on further studies. Also, more studies are needed to identify role of CRTAP variants in lethal skeletal dysplasias, and additional genomic alterations modifying the phenotypes or response to zoledronate or denosumab therapy in these CRTAP -related OI patients.
Footnotes
Conflict of Interest None declared.
References
- 1.Barnes A M, Chang W, Morello R. Deficiency of cartilage-associated protein in recessive lethal osteogenesis imperfecta. N Engl J Med. 2006;355(26):2757–2764. doi: 10.1056/NEJMoa063804. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Morello R, Bertin T K, Chen Y. CRTAP is required for prolyl 3- hydroxylation and mutations cause recessive osteogenesis imperfecta. Cell. 2006;127(02):291–304. doi: 10.1016/j.cell.2006.08.039. [DOI] [PubMed] [Google Scholar]
- 3.ACMG Laboratory Quality Assurance Committee . Richards S, Aziz N, Bale S. Standards and guidelines for the interpretation of sequence variants: a joint consensus recommendation of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology. Genet Med. 2015;17(05):405–424. doi: 10.1038/gim.2015.30. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Kumar C, Panigrahi I, Somasekhara Aradhya A, Meena B L, Khandelwal N. Zoledronate for osteogenesis imperfecta: evaluation of safety profile in children. J Pediatr Endocrinol Metab. 2016;29(08):947–952. doi: 10.1515/jpem-2015-0351. [DOI] [PubMed] [Google Scholar]
- 5.Panigrahi I, Didel S, Kirpal H. Novel mutation in a family with WNT1-related osteoporosis. Eur J Med Genet. 2018;61(07):369–371. doi: 10.1016/j.ejmg.2018.01.017. [DOI] [PubMed] [Google Scholar]
- 6.Forlino A, Marini J C.Osteogenesis imperfecta Lancet 2016387(10028):1657–1671. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Valli M, Barnes A M, Gallanti A. Deficiency of CRTAP in non-lethal recessive osteogenesis imperfecta reduces collagen deposition into matrix. Clin Genet. 2012;82(05):453–459. doi: 10.1111/j.1399-0004.2011.01794.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Saraff V, Sahota J, Crabtree N, Sakka S, Shaw N J, Högler W. Efficacy and treatment costs of zoledronate versus pamidronate in paediatric osteoporosis. Arch Dis Child. 2018;103(01):92–94. doi: 10.1136/archdischild-2017-313234. [DOI] [PubMed] [Google Scholar]
- 9.Balasubramanian M, Pollitt R C, Chandler K E. CRTAP mutation in a patient with Cole-Carpenter syndrome. Am J Med Genet A. 2015;167A(03):587–591. doi: 10.1002/ajmg.a.36916. [DOI] [PubMed] [Google Scholar]
- 10.Amor I M, Rauch F, Gruenwald K. Severe osteogenesis imperfecta caused by a small in-frame deletion in CRTAP. Am J Med Genet A. 2011;155A(11):2865–2870. doi: 10.1002/ajmg.a.34269. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.McClung M R. Sclerostin antibodies in osteoporosis: latest evidence and therapeutic potential. Ther Adv Musculoskelet Dis. 2017;9(10):263–270. doi: 10.1177/1759720X17726744. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Van Dijk F S, Nesbitt I M, Nikkels P GJ. CRTAP mutations in lethal and severe osteogenesis imperfecta: the importance of combining biochemical and molecular genetic analysis. Eur J Hum Genet. 2009;17(12):1560–1569. doi: 10.1038/ejhg.2009.75. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Lin H-Y, Chuang C-K, Su Y-N. Genotype and phenotype analysis of Taiwanese patients with osteogenesis imperfecta. Orphanet J Rare Dis. 2015;10:152. doi: 10.1186/s13023-015-0370-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Ho Duy B, Zhytnik L, Maasalu K. Mutation analysis of the COL1A1 and COL1A2 genes in Vietnamese patients with osteogenesis imperfecta. Hum Genomics. 2016;10(01):27. doi: 10.1186/s40246-016-0083-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Lee K-S, Song H-R, Cho T-J. Mutational spectrum of type I collagen genes in Korean patients with osteogenesis imperfecta. Hum Mutat. 2006;27(06):599. doi: 10.1002/humu.9423. [DOI] [PubMed] [Google Scholar]
- 16.Zhytnik L, Maasalu K, Reimann E, Prans E, Kõks S, Märtson A. Mutational analysis of COL1A1 and COL1A2 genes among Estonian osteogenesis imperfecta patients. Hum Genomics. 2017;11(01):19. doi: 10.1186/s40246-017-0115-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Caparrós-Martin J A, Valencia M, Pulido V. Clinical and molecular analysis in families with autosomal recessive osteogenesis imperfecta identifies mutations in five genes and suggests genotype-phenotype correlations. Am J Med Genet A. 2013;161A(06):1354–1369. doi: 10.1002/ajmg.a.35938. [DOI] [PubMed] [Google Scholar]
