COL1A1 Sp1 variation and bone phenotypes in an Italian population

Summary

Background

Osteoporosis is the most common metabolic bone disorder of the elderly, affecting the normal bone turnover with an increased bone resorption and subsequent higher risk of fragility fractures. Collagen type 1 is the most represented protein in bone matrix. A genetic variation (Sp1) in intron 1 of COL1A1 gene has been associated to modulation of expression of the alpha 1 chain of collagen type 1 and it is considered a candidate polymorphism for predisposition to osteoporosis status and fragility fractures. Association studies, in ethnically different populations, are needed to strongly confirm the role of this polymorphism in bone metabolism.

Materials and methods

We enrolled over 2,000 Italian individuals and studied their bone mineral density (BMD) and fractures in relation to age, sex and body mass index (BMI). Moreover, we analyzed the distribution of Sp1 polymorphism in these individuals and associated it to normal bone status, osteopenic condition or osteoporosis diagnosis, BMD and the presence of low-trauma fractures.

Results

The most rare ss genotype showed a trend for osteoporosis diagnosis with respect to both normal and osteopenic status. The same genotype resulted to be associated to lower values of BMD both at spine and femur sites. No association was found with fractures.

Discussion

In conclusion the presence of the homozygote ss genotype seemed to predispose to osteoporosis diagnosis and to be more frequent in subjects with lower spine and femur BMD values.

Introduction

One of the main abilities of bone tissue is to resist mechanical forces. Bone strength depends not only on bone mass and bone mineral density (BMD) but also by tissue quality, which is related to bone micro-architecture, bone turnover, and extracellular matrix protein composition, both quantity and quality. Osteoporosis is an elderly-related metabolic bone disease caused by a deregulation of normal homeostasis of bone turnover and an increased activity of osteoclasts. This disorder is characterized by the reduction of BMD and the alteration of bone micro-architecture and matrix protein composition, which combined are responsible for loss of bone strength and mechanical properties with the final endpoint represented by low trauma fractures.

Bone matrix is composed of a mineral phase which provides stiffness and a protein phase which is responsible for toughness and the ability to absorb energy. Collagen type 1 is the most represented protein of the bone matrix, accounting for about the 80% of the total proteins in bone tissue. Alteration of collagen properties and of its relative quantity in bone matrix can affect mechanical features of bone tissue and increase susceptibility to fragility fracture.

Collagen is composed by two chains, alpha 1 and alpha 2 (rate 2:1 in a collagen fiber), encoded by two different genes, COL1A1 and COL1A2. Mutations of these genes are associated to bone collagen abnormalities and severe clinical phenotype, such as osteogenesis imperfecta. A polymorphism in intron 1 of COL1A1 gene, Sp1 polymorphism (G>T substitution, rs1800012), is suspected to affect gene transcription, resulting in an altered chain alpha 1/chain alpha 2 ratio and an altered mature collagen protein. The s allele presents an increased affinity for the Sp1 transcription factor, resulting in a higher amount of chain alpha 1 with respect to chain alpha 2; the Ss genotype has been associated to a collagen chain ratio of 2.3 (respect to the normal 2, typical of the SS genotype) (1). An unbalanced ratio of collagen chains may be a potential risk factor for bone fragility and a predisposing factor to osteoporosis.

Some association studies have evaluated the role of this polymorphism in the susceptibility to osteoporosis and in the risk of fragility fractures, showing mild association to bone mineral density (BMD) and a stronger relationship to osteoporotic fractures, particularly vertebral ones (2–4). The multicentric Genetic Markers for Osteoporosis (GENOMOS) Consortium has performed a large-scale association study on over 20,000 individuals, from several centres all around Europe, between the Sp1 polymorphism, BMD and fracture risk (5). They evidenced that ss genotype was associated to a reduced BMD and to a non-significant trend for vertebral fracture in women, independent from BMD. However, since a significant ethnic difference in osteoporosis-related phenotypes and their potential underlying genetic determination are now well recognized, studies of genetic determinants, related to bone mass peak acquisition, bone architecture determination and/or to osteoporosis, should encompass ethnically different populations in order to improve the strength of association. Regarding the Italian population, COL1A1 Sp1 polymorphism distribution has been evaluated in two small cohorts of 100 healthy postmenopausal Sicilian women and 100 osteoporotic postmenopausal Sicilian women vs 360 healthy postmenopausal women from Burkina Faso to measure the influence of environmental factors in osteoporosis risk (6). The Sp1 allele frequency was also studied in 240 selected Italian osteoporotic subjects bearing a femoral neck fracture (7). No data, to date, are available regarding the general Italian population.

Thus, in the present study we have evaluated the allelic and genomic distribution of the Sp1 polymorphism in a large and well characterized population of over 2,000 male and female Italian subjects and we have studied its possible association to bone status, spine and femur BMD values and low-trauma fracture occurrence.

Materials and methods

Population

The recruited population consists of 2,299 Italian subjects (1,542 females and 757 males; mean age 61.8+14.6 years), afferent to the ambulatory of Metabolic Bone Unit, Department of Surgery and Translational Medicine, University of Florence. Individuals presenting a history of serious clinical conditions, affected by malignancies or metabolic bone diseases and/or receiving any long-term therapy known to interfere with bone metabolism have been excluded from the study. Data about fractures have been collected by a specific questionnaire, including information on time, localization, and cause of each fracture; suspected spinal fractures have been confirmed by X-ray screening. For the genetic association we considered data about fractures at wrist, hip, and spine. Fractures caused by high-trauma have been excluded from the association studies. Body mass index (BMI), expressed as weight/height² (kg/m²) has been determined for each individual. Informed consent to the genetic study and to the collection of clinical data has been obtained from each subject.

Bone measurements

Bone mineral density (BMD) has been evaluated at lumbar spine (L1–L4, LSBMD) in 2,215 subjects and at left femoral neck (FNBMD) in 1,043 subjects by dual-energy X-ray absortiometry (DXA) using a Hologic bone densitometer (Hologic, Bedford, Massachussets, USA).

For individuals over 70 years of age, the FNBMD has been used for bone clinical classification (normal, osteopenia or osteoporosis), since it is less susceptible to age-related osteoarthritis and bone degenerative lesions, which could influence the DXA measurement more than LSBMD.

For men under 55 years of age and for all pre-menopausal women, the Z-score has been used for bone clinical classification (normal, osteopenia or osteoporosis), and BMD has been evaluated only at lumbar spine.

For the remaining individuals (men over 55 years and under 70 years and post-menopausal women under 70 years) the bone clinical classification (normal, osteopenia or osteoporosis) has been made according to the lowest T-score between FNBMD and LSBMD has been used for bone.

Individuals have been classified as normal (T-score or Z-score >-1SD), osteopenic subjects (T-score or Z-score between −1 and −2.5 SD) or osteoporotic (T-score or Z-score < −2.5SD) and indicated, respectively, as phenotype 1, phenotype 2 or phenotype 3.

Genetic analysis

Genomic DNA was extracted from peripheral blood leukocytes using a microvolume extraction method, NucleoSpin Blood Quick Pure (Macherey-Nagel, Easton, PA, USA) according to the manufacturer’s instructions. The region containing the Sp1 polymorphism of COL1A1 gene has been amplified by PCR using 5′-AAAGTGACCTGGAGGCATTG-3′ as forward primer and 5′-CTTCCAACTCCAACCTCAGC-3′ as reverse primer. The PCR products were tested by 2% ethidium bromide-stained agarose gel electrophoresis and then purified by NucleoFast 96 PCR Plates for PCR product purification (Macherey-Nagel, Easton, PA, USA). One aliquot of each PCR product was sequenced, both with forward and reverse primers, using the BigDye Terminator Purification Kit (Applied Biosystems, Foster City, CA, USA) in a reaction consisting of twenty-five repeated cycles of denaturation for 10 sec at 96°C, annealing for 5 sec at 50°C and extension for 2 min at 60°C. The sequencing products were then purified with Montage SEQ96, Sequencing Reaction Cleanup Kit (Millipore, Bedford, MA, USA) and analyzed on the ABI Prism 3100 Genetic Analyzer (Applied Biosystems, Foster City, CA, USA). The obtained sequences were compared to the reference sequence of the Sp1 polymorphism (rs1800012). Subjects have been classified as genotype SS (GG; genotype 0), genotype Ss (GT; genotype 1) or genotype ss (TT; genotype 2), with T being the minor allele.

Statistical Analysis

Patients’ characteristics (BMI, LSBMD, and FNBMD) have been expressed as mean plus/minus standard deviation (SD). Correlation between Sp1 genotypes and diagnosis, Sp1 genotypes and mean value of FNBMD, Sp1 genotypes and mean value of FNBMD have been evaluated by using Scheffe’s test. Differences were considered significant if the p-value was less than 0.05.

All statistical analyses were carried out using SPSS, 10.1.0 SPSS Inc. Chicago, USA.

Results

Bone measurements and fracture incidence

Results from DXA measurements evidenced 908 individuals with normal BMD (39.5%), 714 individuals with osteopenia (31.1%), and 677 individuals with osteoporosis (29.4%). We found 382 men with normal BMD (50.5%), 258 men with osteopenia (34.1%), 117 men with osteoporosis (15.4%), 526 women with normal BMD (34.1%), 456 women with osteopenia (29.6%), and 560 women with osteoporosis (36.3%), confirming that women are more affected by this bone disease than men. Chi-square analysis demonstrated a strong correlation between sex and diagnosis. Sex has no correlation with BMI, but has been associated to LSBMD and FNBMD, both resulting significantly lower in women than in men (Table 1). No association has been found between sex and spine, hip or wrist fractures.

Table 1.

Associations between sex and BMI, LSBMD and FNBMD.

	Sex	Number of subjects	Mean	Standard deviation	Mean standard error
BMI	1	877	26.8569671	12.8043210	0.432371198
	2	874	26.3516101	11.5126465	0.389421076
LSBMD	1	872	1.000252	0.1739596	0.0058910
	2	853	0.873273	0.1586676	0.0054327
FNBMD	1	254	0.860114	0.1575428	0.0098851
	2	238	0.745013	0.1455119	0.0094321

Footnotes: sex: 1= males; 2 = females

We found that age, diagnosis, BMI and BMD values showed different correlations and trends when we consider different kinds of fractures. Regarding spinal fractures: the presence of one or more spine fracture showed a significant relation to a worse diagnosis, and also LSBMD and FNBMB were both significantly lower in these subjects. The mean age of subjects with a spine fracture resulted to be higher than non-fractured individuals. No association has been evidenced between spine fracture and BMI. Regarding hip fractures: the diagnosis is significantly worse in subjects with fracture, and BMI, LSBMD, and FNBMB were also all significantly lower in these subjects. The mean age of subjects with a hip fracture resulted to be higher than non-fractured individuals. Regarding wrist fractures: the diagnosis is worse in subjects with fracture, and BMI, LSBMD, and FNBMB were also all significantly lower in these subjects. No difference in mean age has been shown in wrist fractured vs non-fractured subjects.

Genetic analysis

The genotypes of Sp1 polymorphism have been successfully obtained in all the 2,299 subjects of the study. Genotype distribution showed the following frequencies: genotype SS (59.81%), genotype Ss (34.54%), and genotype ss (5.65%). No difference was found in Sp1 genotype distribution between male and female genders.

Association studies

Sp1 genotypes have been statistically analyzed regarding their correlation to BMI, diagnosis (normal BMD, osteopenia, or osteoporosis), BMD value at lumbar spine and femoral neck fractures.

BMI showed no association with sex, bone DXA diagnosis, BMD values at spine or femur. No association has been found between BMI and Sp1 genotypes.

Distribution of Sp1 genotypes with respect to diagnosis is shown in Table 2. Chi-square analysis showed no association between Sp1 genotypes and diagnosis when analyzed as non-linear variable (three groups of subjects: normal bone, osteopenia, and osteoporosis). Interestingly, if we considered diagnosis as a continuous variable, and we analyzed correlation with COL1A1 genotype by Scheffe’s method, we found that genotype ss (homozygote TT) was characterized by a worse diagnosis with respect to both the other two genotypes (Figure 1). The same clinically negative trend of genotype ss has been found also for continuous mean values of LSBMD (Figure 2A) and FNBMB (Figure 2B), for whom this genotype resulted to be characterized by lower values than both the other two genotypes.

Table 2.

Distribution of Sp1 genotypes with respect to diagnosis.

Diagnosis	COL1A1 Sp1 genotype			Total
	SS (GG)	Ss (GT)	ss (TT)
1	526	321	61	908
2	426	252	36	714
3	423	221	33	677
Total	1375	794	130	2299

Footnotes: Diagnosis: 1 = normal bone phenotype; 2 = osteopenia; 3 = osteoporosis

Figure 1.

Correlation between COL1A1 genotypes and mean value of diagnosis analyzed by Scheffe’s method.

Figure 2.

Correlation between COL1A1 genotypes and mean value of LSBMD and FNBMD by Scheffe’s method.

Distributions of Sp1 genotypes with respect to fractures at spine, hip, and wrist are shown, respectively, in Figure 3A, B, and C. Chi-square analyses showed no association between Sp1 genotypes and fragility fractures, neither at the individual evaluated sites, nor considering the total number of fractures.

Figure 3.

Distribution of Sp1 genotypes with respect to spine fractures (SPFX), hip fractures (HIPFX) and wrist fractures (WRISTFX).

Discussion

Osteoporosis is a multifactorial complex disease caused by interaction between predisposing genetic factors, environmental influences and lifestyle habits. Age and female gender are two fixed risk factors. Women enrolled in our study presented a percentage of 36.3% of osteoporosis vs only 15.4% in men. Both LSBM and FNBMD resulted to be significantly lower in women than in men. These data confirmed the female gender as one of the main osteoporosis risk factors. A higher mean age has been associated to subjects who have reported spine and/or hip fractures. Conversely, no difference in mean age has been shown in wrist fractured subjects vs non-fractured ones, perhaps due to the fact that fragility wrist fracture is usually the first clinical outcome of osteoporosis occurring earlier (mean age 50–60 years) than spine (usually over 65 years of age) and femur (usually over 75 years of age) fractures. Our data seem to suggest that age is not a risk factor for wrist fracture.

For years, low weight and low BMI have been considered well-documented risk factors for future osteoporosis fractures, while a high BMI has been indicated as protective (8). A meta-analysis (8) has studied the association between BMI and fracture risk in males and females from 12 prospective population based international studies, demonstrating that low BMI was associated with an increased risk of fractures both in men and in women, independently of the sex. Conversely, our data from the present study failed to find any association between BMI and fracture incidence in any of the considered sites. BMI also showed no association with diagnosis, LSBMD, or FNBMD. Data from our study seemed to exclude low BMI as a risk factor for low BMD values and/or for fracture risk, both in women and men, or when considered together.

Despite the role of environmental factors and lifestyle in the determination of individual bone mass peak, today, in genes regulating bone metabolism, genetic variations are strongly suspected to play a pivotal role in the determination of BMD and bone matrix composition, in the acquisition of personal bone mass peak and in the structuring of bone microarchitecture. However, since bone tissue metabolism is extremely complex and finely regulated by numerous pathways, it is very difficult to determine the genes involved in osteoporosis predisposition, and it is possible that their total number is very high, with some genes exerting a high degree of influence (major genes) and others playing only a marginal role (minor genes). Numerous genes and their polymorphisms, suspected to regulate bone metabolism, have been intensively studied during recent decades in various international association studies, but always with contradictory and inconclusive results. COL1A1 gene encodes the alpha chain of collagen type 1, the most represented protein in bone matrix, responsible for bone ductility and its resistance to traction. A polymorphism in intron 1 of COL1A1 gene alters a binding site for the transcription factor Sp1, modifying the production of alpha 1 chain and altering its 2:1 ratio with respect to alpha 2 chain with a subsequent reduction of mature collagen type 1 functionality in bone. This polymorphism has been largely investigated for its putative role in osteoporosis predisposition and fragility fracture incidence, and it has been slightly associated to low BMD and/or to increased risk of fractures in various studies (2–4, 9–18). Recently, the GENOMOS consortium performed a retrospective association analysis on over 20,000 European subjects, evidencing a faint non-significant association between the Sp1 polymorphism and vertebral fracture risk and a very modest association between ss genotype and reduced BMD (5). Our results confirmed a trend of the ss genotype toward worse diagnosis and lower BMD values, both at spine and femur, in the Italian population. However, we failed to find any association between this genotype and higher risk of fracture, perhaps due to the total low number of fractures in our population.

Predisposition to osteoporosis and fracture risk are due to numerous genetic and non-genetic factors interacting with each other during a lifetime. To date, no specific genetic tests are available for the sure prediction of osteoporosis occurrence and of fragility fracture risk and, given the high complexity of bone metabolism, they may not be available in the near future. Currently, the best approach to lower osteoporosis risk is prevention, beginning in childhood in the general population, such as correct intake of dietary calcium, physical activity, reduction of smoking and of alcohol intake, and minimal daily exposition to sunshine during spring and summer, or oral active vitamin D supplement during autumn and winter, etc. Moreover, additional prevention, consisting of routinely measurement of BMD and serum and urinary bone turnover markers, is strongly suggested in postmenopausal women, men over 65 years of age, and subjects with family history of fragility fractures and/or bearing osteoporosis risk alleles, such as the COL1A1 ss genotype.

In conclusions, given that it is estimated that by 2050 the incidence of osteoporotic hip fracture will increase by 310 percent in men and 240% in women (19), primary and secondary osteoporosis preventions should become an urgent priority to be adopted by the national health care systems worldwide, mostly in developed countries. At the same time, the application of high-throughput genome-wide techniques, such as genome-wide scan association and next generation sequencing techniques, will allow analysis of genetic bases of bone metabolism regulation through a whole-genome point of view, permitting the bypass of all the limitations of the traditional single candidate gene approaches, as well as the discovery of unsuspected risk candidate genes and/or loci.