To the Editor:
Osteoporosis and obesity are two inheritable diseases indexed by low bone mineral density (BMD) and high body mass index (BMI), and are correlated. Osteoblast cells and adipocyte cells share the same progenitor, and can trans-differentiate into each other. Thus, pleiotropic genes may exist to influence the risks of both diseases. In order to identify such pleiotropic genomic loci, we performed a large scale bivariate genome-wide association (GWA) meta-analysis in 14 489 participants from seven samples. Details of the seven samples were described previously.1
Raw femoral-neck BMD and BMI values were adjusted by age, age squared, gender, height (in case of BMD) and the first five principle components derived from genome-wide genotype data. The residuals were normalized by inverse quantiles of standard normal distribution. Genotype imputation based on 1000 genomes project was performed in each GWA sample, resulting in 6 879 267 qualified bi-allelic variants. Both univariate and bivariate association tests were performed in each individual sample, followed by an inverse variance weighted fixed-effects meta-analysis. The in-house meta-analysis software is available upon reasonable request to the corresponding author.
Univariate BMD GWA meta-analysis identified seven distinct loci at the genome-wide significance (GWS, 5.0 × 10−8) level: 1p31.3, 1p36.12, 4p16.3, 5q14.3, 7q21.3, 7q31.31 and 11p15.2. Univariate BMI meta-analysis identified three distinct loci at the GWS level: 10p14, 16q12.2 and 18q21.32. All the 10 (7 + 3) loci remain GWS significant in the bivariate analysis. Four loci were nominally significant (P < .05) in both univariate analyses, and were declared as pleiotropic loci. The bivariate analysis identified additional three loci at the GWS level: 2p23.2, 14q32.33 and 18p11.21, all of which are nominally significant in both univariate analyses. The seven pleiotropic loci (4 + 3) were subjected to replication in the UK Biobank (UKB) cohort sample; three were successfully replicated: 2p23.2, 16q12.2 and 18q21.32 (Table 1).
TABLE 1.
Main association results of the three pleiotropic loci
| Discovery | Replication | ||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| EAF | BMD | BMI | Bivariate | BMD | BMI | Bivariate | |||||||||||
| SNP | Band | EA/OA | AFR | AMR | ASN | EUR | Gene | Beta (SE) | P | Beta (SE) | P | P | Beta (SE) | P | Beta (SE) | P | P |
| rs61620093 | 2p23.2 | W/M | - | - | - | - | SPDYA | 0.067 (0.013) | 2.55 × 10−7 | 0.047 (0.013) | 3.00 × 10−4 | 1.93 × 10−8 | 0.029 (0.006) | 2.68 × 10−7 | 0.2 (0.005) | 2.10 × 10−5 | 1.68 × 10−7 |
| rs17817712 | 16q12.2 | G/A | 0.95 | 0.75 | 0.84 | 0.58 | FTO | 0.027 (0.013) | 0.04 | 0.082 (0.013) | 2.83 × 10−10 | 2.60 × 10−9 | 0.038 (0.006) | 1.08 × 10−9 | 0.142 (0.005) | 2.02 × 10−177 | 1.23 × 10−121 |
| rs10871777 | 18q21.32 | G/A | 0.70 | 0.85 | 0.79 | 0.76 | MC4R | 0.035 (0.014) | 0.01 | 0.083 (0.015) | 3.14 × 10−8 | 2.82 × 10−8 | 0.021 (0.007) | 3.0 × 10−3 | 0.103 (0.005) | 2.74 × 10−94 | 1.86 × 10−52 |
Notes: Meta-analysis included seven GWA samples. Lead SNP of each loci was reported. An independent locus was defined as a genomic region of 500 kb length from either side of the lead SNP. EA/OA, effect allele/non-effect allele; Allele frequencies (EAF) are reported for the effect allele. The EAF of different ethic/race participants was derived from 1000 Genomes Project data; bold italic fonts represent the P value reach genome wide significance level (P < 5 × 10−8).
Abbreviations: BMD, bone mineral density; BMI, body mass index; GWA, genome-wide association.
The three replicated lead SNPs and their neighbor SNPs (LD r2 > 0.8) were annotated through HaploReg.2 SNPs at 2p23.2 are strongly associated with expression of TRNA Methyltransferase 61B (TRMT61B) gene in subcutaneous adipose (P = 1.21 × 10−13), visceral omentum adipose (P = 1.61 × 10−16) and lymphoblastoid (P = 7.96 × 10−25). The lead SNP rs17817712 at 16q12.2 is located in an intron of Fat mass and Obesity-associated protein (FTO) gene. Its flanking region has the enhancer activity in adipocyte cultured cells and osteoblast primary cells. FTO gene has been well described in relation to obesity phenotypes. Functional study demonstrates that FTO functions intrinsically in osteoblasts through Hspa1a-NF-κB signaling to enhance the stability of mRNA of proteins that function to protect cells from genotoxic damage.3 rs10871777 at 18q21.32 is located in the 3’-end of the Melanocortin 4 Receptor (MC4R) gene, which regulates food intake, body weight and caloric efficiency. Functional study demonstrates that osteoblast-derived lipocalin 2 (LCN2) crosses the blood-brain barrier, binds to the MC4R in the paraventricular and ventromedial neurons of the hypothalamus and activates an MC4R-dependent anorexigenic (appetite-suppressing) pathway.4
In conclusion, statistical evidence together with previously reported biological functions suggest that 3 loci 2p23.2, 16q12.2 and 18q21.32 may play pleiotropic roles in both bone and body mass metabolism. Functional annotations highlight several candidate genes, including TRMT61B, FTO and MC4R. Our findings enhance the understanding of genetic interplay between osteoporosis and obesity.
ACKNOWLEDGEMENTS
FHS program: conducted and supported by the National Heart, Lung, and Blood Institute (NHLBI) in collaboration with Boston University (Contract No. N01-HC-25195, N02-HL-64278), NIH (R01 AR/AG 41398). SHARe Illumina genotyping was provided under an agreement between Illumina and Boston University. Manuscript not prepared in collaboration with investigators of the FHS and does not necessarily reflect the opinions or views of the FHS, Boston University, or NHLBI. The datasets used obtained from dbGaP at http://www.ncbi.nlm.nih.gov/sites/entrez?db=gap through dbGaP accession phs000342.v14.p10. WHI program: funded by the NHLBI, NIH, U.S. Department of Health and Human Services. Contracts N01WH22110, 24152, 32100-2, 32105-6, 32108-9, 32111-13, 32115, 32118-32119, 32122, 42107-26, 42129-32 and 44221. Manuscript not prepared in collaboration with investigators of WHI, not reviewed and/or approved by WHI, and not necessarily reflecting the opinions of WHI or NHLBI. WHI SHARe genotyping: funded by NHLBI Contract N02-HL-64278. The datasets used obtained from dbGaP at http://www.ncbi.nlm.nih.gov/sites/entrez?db=gap through dbGaP accession phs000200.v10.p3. IFS study: funded by the NIA Division of Geriatrics and Clinical Gerontology (DGCG). Assistance with phenotype harmonization and genotype cleaning, as well as with general study coordination provided by the NIA DGCG and the NIA Division of Aging Biology. Support for the collection of datasets and samples provided by the parent grant: Genetic Determinants of Bone Fragility (P01AG018397). Funding support for the genotyping which was performed at the Johns Hopkins University Center for Inherited Diseases Research was provided by the NIA.
Footnotes
CONFLICT OF INTEREST
The authors declare no potential conflict of interest.
DATA AVAILABILITY STATEMENT
The summary data are available from the corresponding authors upon reasonable request.
REFERENCES
- 1.Pei YF, Zhang L, Liu Y, et al. Meta-analysis of genome-wide association data identifies novel susceptibility loci for obesity. Hum Mol Genet. 2014;23(3):820–830. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Ward LD, Kellis M. HaploReg: a resource for exploring chromatin states, conservation, and regulatory motif alterations within sets of genetically linked variants. Nucleic Acids Res. 2012;40(Database issue): D930–D934. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Zhang Q, Riddle RC, Yang Q, et al. The RNA demethylase FTO is required for maintenance of bone mass and functions to protect osteoblasts from genotoxic damage. Proc Natl Acad Sci USA. 2019;116(36): 17980–17989. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Mosialou I, Shikhel S, Liu JM, et al. MC4R-dependent suppression of appetite by bone-derived lipocalin 2. Nature. 2017;543(7645): 385–390. [DOI] [PMC free article] [PubMed] [Google Scholar]