Type-1 a-1 Collagen and Osteoporosis. The gene COL1A1, which encodes type-1 a-1 collagen, is the one for which there is convincing evidence of a role in the determination of bone density and fracture risk. A polymorphism in the binding motif of the nuclear transcription factor Sp1 in the promoter region of the gene in 22% of Caucasians, decreased the binding affinity for the ligand protein, and has been associated with bone density, bone fracture risk and bone turnover in several studies [10-13]. The effects of the Sp1 allele are age-dependent and probably play a role in the rate of bone loss in menopausal women. This is the first genetic marker associated with fracture risk with an LOD score (a measure of linkage) (RR = 1.5-3.2). The COL1A1 gene locus has been linked with bone density in twins and families [14,15]. In one longitudinal study of the Sp1 COL1A1 polymorphism, it was found in association with the at risk “s” allele, with bone loss in the lumbar spine in postmenopausal women with low calcium intake, and no effect of this allele on the femoral neck [16]. The opposite effect was found in women with high calcium intake, suggesting that this marker may be useful in predicting the effect of calcium supplementation. In vitro studies suggested that the Sp1 COL1A1 polymorphism affects the production of a-1 procollagen and formation of a-1 (I) homotrimer and bone strength [17]. Further prospective studies, and in more populations, are needed to define the mechanisms of association of the Sp1 COL1A1 polymorphism with osteoporosis, and the clinical value of screening this marker. There is an agreement that this is the most promising genetic marker for prediction of postmenopausal bone mass and fracture risk [18,19].

      Vitamin D-Receptor (VDR) Alleles and Osteoporosis. The active metabolite of vitamin D plays an important role in the regulation of bone cell function and homeostasis of serum calcium. Also 1,25 (OH)2D3 has antiprolifer­ative and immunomodulating effects. It exerts its biological effects when binding to its receptor/VDR. There are normal variations in the population/polymorphism of a VDR gene. Morrison et al. [20], reported an association between the restriction fragment length polymorphism (RFLP) of the VDR gene and serum osteocalcin concentration. The same authors reported the association between VDR polymorphism and BMD in twins and the adult population of Australia, and came to the conclusion that 75% of the genetic effect on BMD is due to variations in the VDR gene [21].

      Three polymorphisms in VDR genes are recognized: alleles ‘b’ or ‘B’, ‘a’ or ‘A’, ‘t’ or ‘T’. These polymor­phisms determine the serum osteocalcin as a marker of bone synthesis and bone density [18,20]. Many association studies with contradictory results were published. One large linkage study of UK families found a weak linkage between COL2A1-VDR locus and bone density with maximum LOD score for the lumbar spine 1.7 [14]. The VDR polymorphism may affect the absorption of calcium in the intestine [22,23], but no differences in the expression of VDR in the intestine were found between different genotypes [23]. The coding polymorphism in exon 2 of the VDR gene has been reported to produce different isoforms of the VDR protein [24,25]. Women with longer VDRs have low bone density, which was not confirmed by some authors [27-30]. The VDR gene polymorphism is correlated with the effect of therapy. The effect of Raloxifene seems to be controlled by BsmI VDR gene polymorphisms [31]. The steroid hormones and vitamin D exert their biological effects binding with nuclear receptors and specific binding proteins associated with the plasma membrane [membrane associated rapid response steroid (MARRS) binding proteins]. Osteoblasts respond to therapy with 1,25(OH)2D3 through both nuclear VDR and MARRS, and both mechanisms are necessary for bone formation. Osteoclasts do not have the nuclear VDR but bear 1,25D3-MARRS. The intestines express 1,25D3-MARRS which decrease with age, and can explain the reduced calcium absorption and bone loss [31]. It is generally accepted that there are great geographic and demographic variations in the association strength between the VDR allele polymorphism and BMD [25,28,29,31]. The available data show that the allele variations in the VDR gene locus have a role in the genetic regulation of bone mass, probably modified by calcium intake, vitamin D and intestinal calcium absorption, but the molecular mechanisms are not clear. A synthetic VDR ligand (ED-71) possesses selective effect on bone, and is under clinical testing. Its effect on lumbar spine is dose-dependent without hyper­calcemia and hypercalciuria [33]. In experimental ovari­ectomized rats, ED-71 increases BMD of the spine more than Alfacalcidol, while increasing calcium absorption and decreasing parathyroid hormone level [32,33].

      Estrogen, the Estrogen Receptor (ER) Gene and Osteoporosis. The rapid bone loss which occurs after menopause and the effect of hormone replacement therapy (HRT) prove the importance of estrogen and its receptor for bone. Knockout mice with null allele for ERa have reduced bone mass [34]. In men, a mutation in the ER gene also leads to osteoporosis [35]. Studies of three poly­morphisms of the ER a gene have shown that they may play a role in determining bone density [36-38], but these have not been confirmed by other authors [39]. Ushiro­yama et al. [40] studied the gene polymorphisms of two fragments, Pvu and Xba I RFLPs, and found an association with low vertebral BMD.

      Genetic association between VDR and ER with an effect on bone mass, growth and peak bone mass, was found in children [33,41]. No such correlation was found with bone mass and rate of bone loss in postmenopausal women [41]. It is accepted that this gene interaction is age dependent.

      Interleukin 6 and Osteoporosis. Interleukin 6 (IL-6) is an important cytokine in the mediation of bone loss that is associated with estrogen deficiency. A gene polymorphism is associated with reduced bone mass of spine, femoral neck and radius in postmenopausal women [43,44]. One study which looked at a mini-satellite in the 3’ flanking region of IL-6, was found to be associated with low bone density in homozygotes [45]. Further linkage disequilibrium mapping of this region is required to identify the disease-causing variants.

      Transforming Growth Factor b (TGFb) and Osteoporosis. The super family TGFb includes more than 30 growth factors. TGFb1 plays an important role in cell growth, morphogenesis, inflammation and bone growth. It is suspected to take part in the regulation of bone turnover through the interaction of osteoblast with osteoclast. In vitro studies show that estrogens stimulate its synthesis from osteoblasts, and TGFb1 induces apoptosis in osteoclasts [46]. In some clinical studies, one of the gene alleles was associated with low bone mass and high bone fraction of alkaline phosphatase in osteoporotic women [45,46]. TGFb is thought to be an important regulator of bone formation and bone resorption. TGFb knockout mice have low total bone mineral content and a reduced rate of growth [47]. The lack of biglycan, an extracellular matrix proteoglycan that binds TGFb and negatively regulates its function, also caused osteoporosis and reduced growth rate [48]. The experimental overexpression of TGFb by osteoblasts caused osteopenia due to increased bone remodeling [49]. In a study of Danish women, an association was found between TGFb gene polymorphism and low bone density and increased bone turnover [46].

      Candidate Genes and Osteoporosis. Several candidate “osteoporosis genes” have been associated with bone density: apolipoproteon E [50], the calcitonin receptor [51], the IL-1 receptor antagonist [52], osteocalcin [53], parathyroid hormone [54], and MHC genes [55]. Chen et al. [56], studied the Taq I IL-1 b exon 5 polymorphism and found an association with low BMD. Confirmation studies are required to determine the general relevance to osteoporosis.

      Osteoporosis is one of the few remaining polygenic diseases for which large whole genome screening has not been reported. A candidate gene and whole genome screen reported by Devoto et al. [57], in a heterogenous family collection suggested a linkage to loci in chromosomes 1p, 2p, 4q, and the highest individual marker in 11q (LOD 2.08 for CD3D). A study of 115 Caucasian families in the UK (about 600 individuals), reported linkage between bone density and the parathyroid hormone receptor type 1 (LOD score at the femoral neck: 2.7-3.5), suggestive evidence of linkage with COL1A1, IL-6, the ER, COL2A1 and VDR, IL-1, IL-4 and epidermal growth factor (LOD score >1.0 [14]. A genome screen of a family with 22 members with high bone density showed a linkage with chromosome 11q12-13 with a maximal LOD score of 5.74 at the marker d11s987 [58], a locus thought to be involved in the determination of peak bone mass. A linkage study of Caucasian and African-American sister pairs confirmed this locus as a determinant of peak bone mass [59]. Also, autosomal-recessive osteopetrosis and osteoporosis-pseu- doglioma syndrome, monogenic diseases with abnormal bone density, map to the same region [60,61].

      The best model for progress in the field of osteoporosis genetics comes from the field of diabetes mellitus genetics and the pioneering work of John A. Todd [62], where a combination of whole genome, candidate gene linkage and linkage disequilibrium studies, dissection of the features of animal models of the disease and research into the pathophysiology of the disease, have resulted in major advances in the understanding of the genetics of type 1 diabetes mellitus. No single method can produce answers to all questions in complex genetic diseases, but progress in this field will overcome the mapping of polygenic diseases.