Multilocus analysis of estrogen-related genes in Spanish postmenopausal women suggests an interactive role of ESR1, ESR2 and NRIP1 genes in the pathogenesis of osteoporosis

Abstract

Osteoporosis is a common disease with multiple environmental and genetic risk factors involved. Using a marker-by-marker approach, the role of different estrogen-related genes has been analyzed in different populations, but most of these studies ignore the complex multigenic nature of human osteoporosis. Looking for markers related to osteoporosis, we have analyzed five single nucleotide polymorphisms located in genes related to the estrogen pathway, Follicle Stimulating Hormone Receptor (FSHR) gene, the CYP19 aromatase (CYP19A1) gene, the Estrogen Receptor alpha (ESR1) gene, the Estrogen Receptor beta (ESR2) gene and the Nuclear Receptor Interacting Protein 1 (NRIP1) gene in 265 unrelated postmenopausal women. We have obtained nominal P values for the NRIP1 Gly75Gly and ESR2 *39A > G markers (P = 0.013 and P = 0.02 respectively), but no gene seems to be associated after multiple test corrections. Reanalysis of this study using 437 postmenopausal women confirmed our results and only detect marginal effects for ESR2 marker (P = 0.045).

By contrast, multilocus analysis predicted epistatic interactions between ESR1, ESR2 and NRIP1 loci and its involvement in postmenopausal osteoporosis (P = 0.003). We detected two digenic genotypes involving ESR2-NRIP1 and ESR2-ESR1 genes strongly associated with osteoporosis (P = 0.007). Replication of multilocus studies using 437 patients confirmed the detected interactions (P < 0.01). We proposed a non-additive non-multiplicative oligogenic model including ESR2 AG genotype modulated by NRIP1 A+ or ESR1 TT genotypes involved in osteoporosis. Our results reaffirm the polygenic nature and the genetic complexity of osteoporosis trait adding a new candidate gene (NRIP1) for association studies of bone-related traits.

Introduction

Osteoporosis fractures and bone mass loss are important health care problems affecting the majority of elderly individuals in developed countries [1]. This pathology is defined as the gradual reduction in bone strength with advancing age. It is also well established that age and sex are very important factors that influence this condition [2]. Moreover, a range of environmental, hormonal, and genetic factors is modifying the risk of osteoporosis and the variation of body mass index between individuals ([2], [3] and references therein).

The prevalence of osteoporosis and the presence of well-established risk factors for osteoporosis are being studied in Spain [4], [5], [6]. A recent population-based series of 5195 postmenopausal women has been analyzed looking for demographic, anthropometric, and clinical variables involved in osteoporosis aetiology. The most important risk factors detected in these epidemiological works included age, body mass index, early menopause (<45 years), benzodiazepine use, and previous falls [6].

In addition to previously mentioned environmental or clinical factors, there is enough previously reported evidence to support the existence of genetic factors influencing both body mass index and osteoporosis disease in humans ([2], [3], [7] and references therein). In fact, the current risk model of the pathology comprises the combination of both, genetic and environmental factors following a complex and interactive network of risk components that define – and is expected to predict – the exact risk of any individual [7]. For all these reasons, osteoporosis is now considered a complex and multifactorial disease. Unfortunately, the number of factors involved in each interaction, the link between them or a specific joint model integrating the available information has not been achieved to date.

Due to the vast amount of experimental, epidemiological, and genetic data available, it is well established that estrogenic steroids exert a broad range of effects on skeletal tissues and are considered calcicotropic hormones [3]. In fact, using a marker-by-marker approach, the role of different estrogen-related genes, including the Estrogen Receptor Alpha (ESR1) gene, the Estrogen Receptor Beta (ESR2) gene, and the CYP19 aromatase (CYP19A1) gene, has been analyzed in different populations [8], [9], [10]. For these reasons, we postulated that osteoporosis could be an estrogen-related disease in which genes involved in the production of estradiol (E2), genes involved in the mechanism of action via estrogen receptors, and other genes involved in E2 bio-availability or degradation could be related with osteoporosis following a complex multigenic and interactive model [11].

The existence of genetic interaction between unlinked loci (joint effects) in relation with osteoporosis or bone mineral density has been observed by several independent research groups. Most of these genetic studies combined Vitamin D Receptor (VDR) genotypes with two or more additional loci [12], [13], [14], [15], [16], [17], [18], [19]. Specifically, a two-locus genotype, including VDR Bsm I and ESR1 PvuII polymorphisms, was related to hormone replacement therapy (HRT) outcome in postmenopausal women [12]. This digenic VDR/ESR1 genetic interaction was also related to bone mineral density (BMD) variation in the Chinese postmenopausal women population [13] and BMD values in young children from the Iowa Bone Development Study [14].

Moreover, Interactive VDR and Collagen Type IaI genotypes have also been associated to susceptibility to fracture [15] and bone change and density [16]. More recently, a novel interaction between VDR and CYP19 genotypes in relation to bone mineral density in older African–American women has been detected [17]. Finally, an interesting interaction between ESR1 gene and AIB1 co-transcriptional activator has been associated with quantitative calcaneal ultrasound densitometry values [18], these data suggest the existence of modifier loci related to risk of osteoporosis via estrogen receptor signaling pathway.

Here, we present a multilocus analysis of the BMD trait in 437 unrelated postmenopausal women who underwent densitometric screening for osteoporosis or were diagnosed with osteoporotic fracture. We have analyzed five single nucleotide polymorphisms located at different candidate genes of the estrogen pathway. Association studies provided evidences of a digenic interaction between ESR2 and NRIP1 (Nuclear Receptor Interacting Protein 1) genes in relation to osteoporosis status. A trigenic model including ESR1, ESR2, and NRIP1 genes involved in postmenopausal osteoporosis and fracture risk is also proposed.