Clinical and Laboratory Investigations of Rare Genetic Skeletal Disorders

Although individually rare, the overall prevalence of skeletal dysplasias is approximately 1 in 3,000. In the past two decades, the study of rare skeletal disorders has led to the identification of new signaling pathways fundamental to bone physiology and has informed the development of new treatment approaches for common disorders such as osteoporosis.

We have the ability to draw participants from across the United States and international centers and to study them at the NIHCC in a uniform fashion. We employ a multidisciplinary approach that enables a careful assessment of this rare-disease cohort, expands our knowledge of rare diseases by evaluating patients with atypical presentations, and creates a unique resource of patients, clinical and genomic data, and expertise in the field. Our biorepository of unique samples will be used for studying disease mechanisms and facilitating collaborations. Our area of focus includes disorders of FGF23/phosphate biology, and skeletal dysplasias with an unknown molecular basis.

We routinely perform deep phenotyping via our omnibus clinical protocol (see above) to study the natural history of rare skeletal dysplasias, and we can evaluate the patients either in outpatient clinics or on inpatient wards. We will establish a repository of clinical, biochemical, and imaging data, and ascertain health-related quality of life and patient-related outcomes to inform endpoints in future clinical trials. We will also establish a large biorepository to identify biomarkers and facilitate translation between human pathology and experimental models.

For our patients with genetic skeletal disorders that lack a molecular diagnosis, we will perform exome sequencing (ES), genome sequencing (GS), long-read sequencing, optical genome mapping and RNA-Seq, as necessary. Implementation and prioritization of these approaches will be guided by the current clinical and basic-science understanding of the respective disorders. Our study of Trevor disease, a mosaic disorder characterized by tumoral overgrowth of the joints, is paradigmatic of this approach; we decided to pursue the identification of the molecular basis of this condition because it could inform new avenues for research of both typical and aberrant skeletal growth. Since the enrollment of over 40 families in our clinical protocol, we are focusing on two areas, one involved with disrupted FGF23 metabolism (fibroblast growth factor 23, a protein involved in phosphate metabolism) and the other on skeletal disorders of an unknown etiology.

Our clinical protocol is likely to lead to building additional cohorts in which to investigate the molecular etiology of other rare skeletal dysplasias. We expect to find novel molecular determinants of skeletal disorders using genomic analyses, and we have already successfully identified the etiology of one such rare skeletal dysplasia. This clinical and natural history study has the potential to become a model for the field and to allow us to define novel skeletal diseases, explore expanded phenotypes, and critically examine treatments.

The pathomechanisms of patients with skeletal dysplasias are poorly understood, and there is a paucity of targeted therapeutic options for such patients. An example is OGD, a skeletal disorder caused by gain-of-function variants in FGFR1, which encodes a receptor for fibroblast growth factor 1. I propose that OGD could serve as a model bone disorder because patients manifest cardinal features shared by many skeletal dysplasias, including impaired growth, severe bone deformities necessitating surgery, craniofacial abnormalities (e.g., craniosynostosis and midface retrusion), dysregulated FGF23 (fibroblast growth factor 23) levels, and overactive FGFR1 signaling. There is no treatment currently available for patients with OGD. The experience gained by studying OGD will provide expertise that will prove fruitful in studying novel skeletal dysplasias that we will identify.

There are no animal or cellular models of OGD, and the creation and characterization of new models will allow us to study pathomechanisms and assess the efficacy of potential therapeutic approaches. We also plan to re-purpose drugs to treat rare skeletal disorders, starting with small molecules that are FDA–approved. Given that OGD is caused by an overactive FGFR1 receptor, modulating the disrupted signaling via a tyrosine kinase inhibitor might be therapeutic. This type of drug re-purposing could serve as a paradigm for future treatments of skeletal disorders.

We used CRISPR to knock in the most common variant causative of OGD; we chose the variant Fgfr1 p.(N330I) not only based on its frequency but also on the fact that it has been associated with FGF23 excess and consequent hypophosphatemia in OGD patients. We performed skeletal phenotyping of the mutant Fgfr1^N330I/+ (OGD) mice via morphometrics, Faxitron imaging, microCT, and histological analysis. We will measure plasma phosphate and FGF23 levels via ELISA, and compare FGF23 mRNA in bone using RNA in situ hybridization technology. We have also undertaken transcriptomics with RNA obtained from bone to assess which pathways are dysregulated in the skeleton of OGD mice.

Fgfr1^N330I/+ mice display reduced body weight, shortened naso-anal length, shortened long bones, and craniofacial differences that closely recapitulate the human phenotype. Histology of femurs and tibiae revealed severe disorganization of the growth-plate cartilage. Plasma phosphorus was lower than in controls, driven by a marked elevation in circulating FGF23 levels.

As OGD is caused by an overactive FGFR1 receptor, I hypothesized that inhibiting the signaling pathway would lead to phenotypic improvement. We tested a small-molecule tyrosine kinase inhibitor with activity against FGFR (Infigratinib) for ERK (extracellular signal-regulated kinase) phosphorylation in cultured osteoblasts and for restoration of growth in our animal model of OGD. Infigratinib has been successfully used to treat advanced tumor-induced osteomalacia and was approved by the FDA for the treatment of cholangiocarcinoma resulting from FGFR2 rearrangements; it is currently being evaluated as a treatment for achondroplasia, a related skeletal dysplasia caused by overactivation of the FGFR3 receptor. Cultured osteoblasts from Fgfr1^N330I/+ mice incubated with Infigratinib for one hour exhibited a 76% reduction in ERK phosphorylation compared with vehicle control. We then performed a pilot project treating a small number of young Fgfr1^N330I/+ animals (P1-P16) with 2mg/kg/d of subcutaneous Infigratinib to assess its effect in vivo. The lengths and weights of the treated mutant animals were intermediate between those of the untreated mutant and the wild-type mice.

The underlying etiology of several skeletal dysplasias remains unknown. Finding novel molecular causes of congenital skeletal disorders will contribute to the understanding of skeletal physiology and pathology, which are crucial to generating ideas for targeted treatments. Trevor disease is characterized by skeletal tumors around the joints, leading to altered ambulation, deformity, and pain. It likely represents a mosaic developmental bone disorder and, given that involvement is usually unilateral, there has never been a credible report of familial recurrence, and one patient has been described with an unaffected monozygotic twin. We hypothesized that it involves somatic variants in a gene (or set of genes within a pathway) critical in development. An improved understanding of these pathways in bone and cartilage could lead to targeted therapeutic strategies.

Somatic mosaicism has long been implicated in cancer pathogenesis, but only recently have post-zygotic mosaic variants been studied in non-cancer disease and healthy individuals. The contribution of somatic variants to developmental skeletal disease is still poorly understood. The identification of the genetic etiology of Trevor disease, in particular, might have relevance not only to normal skeletal growth but also to dysregulated pathways in other mosaic disease states, such as cancer.

To test the hypothesis that Trevor disease represents a mosaic disorder, we investigated the clinical features of patients with this disease identified through my NIH cohort, obtained biospecimens (including biopsies from affected bone), and performed deep exome (500x) and genome sequencing (100x) to maximize the chance of identifying variants with low allele frequency. We analyzed the data using a combined ensemble and machine-learning method that has a higher overall somatic variant call accuracy than single variant calling approaches. In addition, to facilitate translation between human pathology and experimental models, we established primary cultures of bone marrow stem cells (BMSCs).

We identified the same somatic variant at an allele frequency of 4–37% in the affected osteo-cartilaginous lesions from five patients with Trevor disease. The specific variant was promising, as it has an existing association with certain cancers. We obtained a conditional mouse model of the variant that we identified in our patients, crossed it to a Prrx1-Cre (PRRX1 determines the fibroblast lineage with a myofibroblastic phenotype) and characterized the skeletal phenotype of this mouse model of Trevor disease. These mice recapitulate the human phenotype by exhibiting abnormal gait and epiphyseal overgrowth. Our current work is focusing on the mechanisms leading to skeletal overgrowth, and on repurposing a drug, panobinostat, to reduce the bone overgrowth caused by Trevor disease.