Genetic Determinants of Skeletal Health and Disease
A. The personal and financial burden of skeletal disease
    Heritable and degenerative diseases of the skeleton have a profound personal impact on the affected individual and place a huge financial burden on our health care system. In most cases, the mental function of the individual is highly developed but the limitations of their skeletal disease is a constant frustration and challenge to realize their full intellectual potential. While diseases affecting the skeleton are not life threatening in most instances, they, more than any other diagnostic disease category, carry the highest burden to the health care system and to the economy overall due to their chronic disruptive impact on the quality of life (figure 1). As illustrated in the report United States Bone and Joint Initiative: The Burden of Musculoskeletal Diseases in the United States, (http://www.boneandjointburden.org), approximately 7.4% of the country’s total GNP (45% x 17%) is spent on treatments that are a direct consequence of a failing skeleton. Because these non-curative treatments only assist the affected individual to cope with his disability, these conditions impose a continuous lifelong economic and personal burden. If our national goal is to mitigate this economic burden, we must identify individuals at risk for a skeletal disease before it develops. Once identified, the individual can either modify his physical behavior so as not to overly strain the vulnerable tissue, or institute a behavioral change or pharmaceutical intervention that helps the tissue compensate for its vulnerability. An example of this principle is the study of bone mass in twins that shows the importance of genetically determined accumulated bone mass prior to the onset of bone loss as a predictor of osteoporosis in elderly subjects. Since accumulated bone mass is acquired during the adolescent stage of growth, osteoporosis should be regarded as a pediatric disease. It follows that identifying individuals with genetic predisposition for suboptimal bone accumulation needs in the early childhood years so steps can be taken to enhance bone mass throughout childhood and adolescence. Thus the challenge for the basic research community it to provide the genetic markers for the primary care giver that are predictive of an adult disease outcome.
Figure 1: Selected graphs from the Burden of Musculoskeletal Disease report.
A. The relative prevalence of major disease categories in the US population. B. The economic cost of disease categories in terms of lost days of employment. C. The inpatient financial cost for treatment of the major disease categories.
Figure 2: Heritability accounts for the 69% and 88% accumulated bone mass.
From: Moayyeri A, Hammond CJ, Hart DJ, Spector TD. Effects of age on genetic influence on bone loss over 17 years in women: the Healthy Ageing Twin Study (HATS). J Bone Miner Res. 2012 Oct;27(10):2170-8.
B. Identifying genetic loci that contribute to skeletal disease in humans
    The advances in DNAseq technology that allows for whole genome and exome sequencing has resulted in the discovery of many new and unappreciated coding genes that can result in disorders of skeletal health. Thus major advances in monogenic and complex genetic traits have identified all but the most rare Mendellian traits that have been observed in humans.

    1. Monogenic – There are over 200 heritable disorders of the skeleton (OMIM), each with a highly discordant phenotype but similar genetic mechanism. Most prominent are the dominantly inherited diseases of bone and cartilage such as osteogenesis imperfecta and spondyloepiphyseal dysplasia that result from a mutation in the major structural protein (type I and II collagen, respectively). These mutations act as a dominant negative because the presence of the misfolded collagen chain induces ER stress to the cell of synthesis and disrupts the hierarchical development of the extracellular matrix. However, other dominantly inherited disorders exert their effect by not sequestering cytokines or growth factors properly, leading to a continuous inflammatory and regenerative state. Another type of dominant mutation is an activation of a receptor that controls lineage expansion and differentiation and has been seen in most of the major signal transduction pathways (FGFs, IGFs, Wnts, TGFß and PTH). Usually disorders that are secondary to haploid insufficiency are less severe, although the phenotype can be significantly altered by the genetic background. Recessively inherited disorders usually have a severe skeletal disease because the genes affected by the mutations encode essential post-translational proteins necessary for processing the initial transcript to a mature extracellular matrix component.

    2. Complex genetic traits – OMIM lists over 250 genetic traits with skeletal abnormalities as part of a complex genetic phenotype. Most cases identify an underlying genetic unit as a gene that impacts skeletal development as well as other tissues essential for neural, cardiovascular or renal development. In addition, many metabolic disorders (mitochondrial, intermediary) are expressed in skeletal tissues and thus could serve as the in vivo surrogate for phenotyping.

    3. Polygenic – Just as genetic background has a major influence on the severity of dominantly inherited disease, it is the major contributor to adult diseases affecting the skeleton. GWAS studies have demonstrated multiple loci that have a statistical association with low bone mass, fracture, and degenerative joint disease. The ENCODE project, which uses the same sequencing technologies, has now demonstrated that the majority of GWAS identified loci are associated with molecular genetic variation in intergenic regions. Because many of the intergenic regions are sites of genetic regulation, it is likely that the inherited trait affects the expression of multiple genes that are on the same chromosome (cis) to the mapped site, or to the site of non-coding RNA or miRNAs that acts at multiple sites throughout the genome (trans). This reality poses a major challenge to the field because it is difficult from human studies alone to assess the impact of these genetic variant when the genetic background of the individual cannot be controlled.
C. Animal models of human skeletal disease.
     1. Monogenic disease: The ability to manipulate the mouse genome has provided a very useful tool for developing murine phenocopies of human disease to confirm that the gene locus is disease causing and to understand the molecular and physiological basis of the clinical phenotype. Thus germ line or tissue conditional gene knockouts have become the standard tool for the experimental geneticists or skeletal biologists. While major insights into genetic mechanism have been made, these studies can be difficult to interpret and inter-relate with similar studies done in other laboratories because the type of mutation, the genetic background of the animal, the phenotyping protocols employed and the reporting platform are all different. Furthermore only certain tissue are examine so that the total impact of the mutation is not appreciated. This lack of standardization can result in incomplete or misinformation that cannot be re-interpreted across different experimental platforms.
    To address this problem, the International Mouse Phenotyping Consortium (IMPC) initiated the KOMP to identify the consequence of a germ line inactivation of every coding gene in the mouse genome. The type of inactivating mutation is highly defined and the genetic background of the mice is held as a constant. A whole animal phenotyping program is utilized to assess the health and activity of the major organ systems and the results of the analysis is deposited into a web-based database that can be accessed the research community. As initially planned, the examination of the skeleton consists of a total body skeletal survey and DEXA study. The objective of this web site is to provide a more discriminating assessment of the bone and cartilage in these KOMP mice.
Figure 3: Organizational structure of the KOMP program. 
See website…..
     2. Quantitative trait loci (QTL) - Mouse geneticists have long appreciated significant differences in bone mass and architecture between different inbred strains of mice. Attempts to map the genetic identify of these differences by interbreeding strains with a low and high bone mass phenotype (recombinant inbred mice) did produce chromosomal intervals that could be linked to the phenotype, but the intervals were too large and contained too many gene loci to be specific. At the time these studies were being performed it was assumed that the difference would be in the coding region of the gene while we now believe these differences will be due to changes with the intergenic regions.
     The mouse geneticists retooled their interbreeding studies to include 8 divergent strains, all of which had been previously whole genome sequenced. This highly interbred mouse population is referred to as the diversity outbred (DO) mice. A dense SNP panel was developed for each strain that could be used for linkage studies. The mice are used to produce a wide spread for any trait and from that spread to identify the mice that reside in the tails of the spread. From the genotyping data, recurring SNPs that appear more frequently in either of the tails relative to the total population provide evidence of a segment affecting variation of the tested phenotype. The power of this genetic platform for identifying genes affecting the skeleton is illustrated in figure 4 which show the spectrum of trabecular bone pattern within this DO mouse population. Mapping these features in architecture as well as the underlying cellular basis for the changes is likely to identify new or unappreciated loci that can contribute to skeletal variation.
Figure 4: Variation in trabecular bone density within the DO mouse population.  .
The same sections are processed for AP and TRAP activity which we anticipate will show a similar spectrum of variation.  The measures will be related to the SNP genotype to identify segments associated with each observation.
D. Reinventing skeletal phenotyping for modern genetic platforms.
     The murine platform provides a plethora a examples of skeletal pathology and variation that needs to be categorized in term that are meaningful to bone biologists. However to be useful as a phenotyping resource to the skeletal biology community that can be related by to genomic databases, the following conditions need to be met.
     •The samples are collected and processed in a uniform manner
     •The mutations have to be produced on a uniform genetic background (with the exception of the DO mince in which variation is the phenotype being studies)
     •The genetic basis of the phenotype need to be easily discovered.
     •A sensitive screening method for segregating normal from the highly variant phenotype needs to be in place. We have chosen µCT.
     •Dynamic and cellular histomorphometry needs to be performed on animals with an abnormal/variant screening phenotype. Here we have developed a fluorescence cryohistological approach that generates digital images that are recognized by a computer image. The methods are relatively high through put and the imaging step are all computer generated.
     •An objective, consistent and observer independent measurement of the fluorescent images has to be implemented. A series of image analysis algorithms have been developed to provide this objective. •Access to the generated information through a public web portal.
     Figure 5 below compares the time line and data output of the work flow that has been developed for this skeletal phenotyping concept and contrasts it with that traditional plastic embedded and manually analyzed methods currently in use.
Figure 5: Schematic comparison of the analysis workflow designed from this web site (left column) relative to the traditional plastic embedded histology (right column).  
Our goal is to be able to get the results from the initial µCT screen to the web data site within 2 weeks of receiving the samples, and to perform and analyze the abnormal samples by histomorphology within another 2 weeks.