LGR4 Gene Polymorphisms Are Associated With Bone and Obesity Phenotypes in Chinese Female Nuclear Families

Objective The current study was conducted to determine whether peak bone mineral density (BMD) and obesity phenotypes are associated with certain LGR4 gene polymorphisms found in Chinese nuclear families with female children. Methods A total of 22 single nucleotide polymorphisms (SNPs) located in and around the LGR4 gene were identified in 1,300 subjects who were members of 390 Chinese nuclear families with female children. Then, BMD readings of the femoral neck, total hip, and lumbar spine as well as measurements of the total lean mass (TLM), total fat mass (TFM), and trunk fat mass were obtained via dual-energy X-ray absorptiometry. The quantitative transmission disequilibrium test was used to analyze the associations between specific SNPs and LGR4 haplotypes and peak BMD as well as between LGR4 haplotypes and TLM, percent lean mass, TFM, percent fat mass, trunk fat mass, and body mass index (BMI). Results Here, rs7936621 was significantly associated with the BMD values for the total hip and lumbar spine, while rs10835171 and rs6484295 were associated with the trunk fat mass and BMI, respectively. Regarding the haplotypes, we found significant associations between GAA in block 2 and trunk fat mass and BMI, between AGCGT in block 3 and total hip BMD, between TGCTCC in block 5 and femoral neck BMD, and between TACTTC in block 5 and both lumbar spine and femoral neck BMD (all P-values < 0.05). Conclusion Genetic variations of the LGR4 gene are related to peak BMD, BMI, and trunk fat mass.


INTRODUCTION
Osteoporosis and obesity are common occurrences during the aging process, and both are subject to environmental and genetic influences. Individuals with osteoporosis have a low bone mineral density (BMD) and an increased risk of fracture due to degradation of their bone microstructure. It is currently estimated that >200 million people have osteoporosis worldwide. In the United States, osteoporosis is an underlying cause of 1.3 million fractures, which cost $10 billion per year to repair (1). As a complex quantitative trait that is 50%-85% heritable, BMD can serve as a surrogate marker for osteoporosis (2)(3)(4). Obesity occurs due to an excessive accumulation of body fat and is closely associated with metabolic, cardiovascular, and physical disabilities (5). While body mass index (BMI) is the most commonly used index of obesity, it does not directly measure the total amount of body fat. Alternative ways of measuring obesity include measurements of the total fat mass, trunk fat mass, and percent body fat. Studies of twins and families in several populations have shown that BMI is strongly influenced by genetics and is 40%-70% inheritable (6).
While osteoporosis and obesity were once viewed as nonrelated disorders, they are now thought to be closely associated with each other. Increased body weight, as a form of mechanical loading, can induce osteogenesis and protect against bone loss (7). Mesenchymal stem cells (MSCs) in bone marrow give rise to both osteoblasts and adipocytes. Obesity induces the MSCs to generate greater numbers of adipocytes than osteoblasts, which leads to decreased bone formation (8). As an endocrine organ, adipose tissue secretes estrogen-synthesizing enzymes that increase estrogen levels. The resulting increased levels of estrogen then enhance osteoclast apoptosis (9) and inhibit osteoblast apoptosis (10), thereby decreasing bone tissue resorption. Moreover, adipose tissue also produces various cytokines and adipokines, whose dysregulation promotes a cascade of metabolic alterations that lead to bone loss (11,12), possibly by creating and maintaining a low-grade inflammatory environment (13). Moreover, certain molecules produced by bone (e.g., osteopontin and osteocalcin) are known to help regulate body weight and maintain appropriate glucose levels (14). When examined at the genetic level, BMD and BMI share a 10%-20% genetic variation (15). Several studies have identified pleiotropic genes [e.g., VDR, ESR1 (16), and FTO (17)] that assist in regulating both body composition and the osteoporosis process.
The gene that encodes the LGR4 protein, a leucine-rich protein also known as GPR48, maps to chromosome 11p14.1, which comprises 18 exons and 19 introns. The LGR4 protein consists of 951 amino acids, including 17 N-terminal leucinerich repeats. The protein also contains a seven-transmembrane region (18). It contributes to the physiological function of several organs through the role it plays in both Wnt and cyclic AMP/ protein kinase A (PKA) signaling (19,20). A recent large-scale study conducted in Iceland identified a strong association between the nonsense mutation c.376C>T [p.R126Xc.367C-T; arg126 to ter; R126Xarg126 to ter; R126Xarg126 to ter; R126Xarg126 to ter; R126Xarg126 to ter; R126Xarg126 to ter; R126Xarg126 to ter; R126X)] found within the LGR4 gene and both lower BMD values and osteoporotic fracture rates (21). Moreover, LGR4 has been found to help control body weight by regulating the conversion of white fat into brown fat (22). Furthermore, a Chinese study that compared obese subjects with control subjects found that a heterogenous variant of c.2248G>A within LGR4 is associated with obesity (23). These results suggest that LGR4 can influence the phenotype of both osteoporosis and obesity. To investigate the validity of this hypothesis, we used the quantitative transmission disequilibrium test (QTDT) to examine whether LGR4 polymorphisms are associated with variations in peak BMD and obesity-related phenotypes in a large cohort of female nuclear family members residing in China.

Subjects
The study protocol was reviewed and approved by the Shanghai Jiao Tong University Affiliated Sixth People's Hospital Ethics Committee. Each study participant signed a written informed consent form prior to enrollment. The study participants resided in Shanghai City and were of Chinese Han ethnic descent.
From 2008 to 2015, we recruited 1,300 individuals who were members of 390 Chinese female-offspring nuclear families (24). The DNA samples obtained from 38 of the individuals were of poor quality and could not be sufficiently amplified to allow for discrimination of the genotypes. Thus, 1,262 individuals who were members of 379 different female nuclear families were included in our final analysis. Among them were 379 parents (mean age, 61.6 ± 7.5 years), 467 female children (mean age, 34.1 ± 6.8 years), and 37 male children (mean age, 34.6 ± 4.7 years). The families had a mean size of 3.33 individuals, as 268, 97, and 14 families included one, two, and three children, respectively. Each enrolled subject was asked to answer questions concerning their demographic characteristics, lifestyle, dietary habits, and tobacco/alcohol use. The study participant exclusion criteria were as follows: (1) significant health problems due to cerebral vascular disease; (2) diabetes; (3) chronic kidney, liver, or lung disease; (4) alcoholism; (5) pharmacologic doses of corticosteroids for >3 months; (6) anticonvulsant therapy for >6 months; (7) evidence of other hereditary or metabolic bone diseases; (8) rheumatoid arthritis or a collagen disease; (9) major gastrointestinal disease within the last year; (10) any major endocrinopathy that would affect bone mass such as hyperthyroidism, hypogonadism, or hypercortisolism; (11) any neurological or musculoskeletal condition that might cause secondary low bone mass (16,(24)(25)(26)(27)(28)(29)(30).

Phenotype Measurements
A dual-energy X-ray absorptiometry (DXA) densitometer manufactured by Lunar Corporation (Madison, WI, U.S.A.) and operated in fan-beam mode was used to measure the bone density in the lumbar spine (L1-4) and left proximal femur (including the femoral neck and total hip) as well as to obtain measurements of the total lean mass (TLM), total fat mass (TFM), and trunk fat mass. The individual ratios of TFM to body weight and TLM to body weight were used as measurements of the percent fat mass (PFM) and percent lean mass (PLM), respectively. The densitometer was calibrated on a daily basis. Fifteen different individuals were each measured three times, and the results were used to calculate the coefficients of variability (CVs). The CV values of the measurements for the lumbar spine, femoral neck, total hip, and trochanter were 1.39%, 2.22%, 0.7%, and 1.41%, respectively. With regard to body composition, the CV values for the TLM, TFM, and trunk fat mass were 1.18%, 3.72%, and 2.52% respectively (28). The weekly phantom measurements taken during the course of our study showed that the DXA data had a reproducibility of 0.45% (16,28). Standard methods were used to measure each subject's height and weight, and the BMI was calculated as the weight in kilograms divided by the square of the height in meters.

Single Nucleotide Polymorphism Selection and Genotyping
The 376 C>T mutation sites and 21 tagSNPs located in and around the LGR4 gene were chosen for evaluation. The tagSNPs were selected based on information contained in the HapMap and dbSNP databases as well as factors that included: (1) whether the tag SNP had been validated in a Chinese population; (2) the amount of heterozygosity (minor allele frequency [MAF] >0.05); (3) whether the pairwise linkage disequilibrium (LD) of the algorithm bin tagSNP had an r 2 value >0.8; (4) the reported functional importance of the tagSNP. Each study subject fasted overnight, after which a blood sample (5 mL) was taken and treated with disodium ethylenediaminetetraacetic acid. A chloroform-phenol solution was used to extract genomic DNA from peripheral blood leukocytes. Genotyping was performed with the combined use of an imLDR Multiplex Kit (Genesky Biotechnologies, Shanghai), an RT

Haplotype and LD Analysis
A process described by Stephens et al. (2001) was used in combination with the PHASE software (version 2.1) (31) to obtain haplotypes from the study population's genotype data. The Haploview software (version 4.2) was used to assess the significance of the LD that existed between the LGR4 gene markers (32). We also examined Lewontin's D' and LD coefficients (r 2 ) between all pairs of biallelic loci. The genotype and haplotype frequencies were determined using genotype information obtained from non-related parents in the nuclear families.

Statistical Analysis
The genotype occurrences of each of the 22 polymorphic sites identified in the parents in each nuclear family were analyzed with the c 2 test to assess their conformity to the Hardy-Weinberg equilibrium. The QTDT (orthogonal model) was performed to evaluate the parameters of population stratification, total association, linkage, and within-family associations between the various SNPs, haplotypes, BMD phenotypes, and obesity-related phenotypes. This method of using the QTDT to analyze data relevant to quantitative traits in conjunction with genotype data obtained from siblings and parents is very effective and was implemented herein using QTDT software (33). All children in the nuclear families were daughters, and the possible effects of the parental phenotypes were not analyzed by the QTDT; therefore, we did not use sex as a covariate when adjusting for variations in phenotype. Instead, the raw BMD values were adjusted using covariates that included height, age, and weight, and age was used to adjust the obesity phenotypes. To avoid false-positive results being obtained after multiple tests, the QTDT software package was used to perform 1,000 Monte Carlo permutations to obtain P-values that could be used to determine the accuracy of our results (34)(35)(36). The statistical power for each SNP is 0.8. A P-value <0.05 was shown to be statistically significant for all our analyses. The QTDT software package used can be downloaded at http://www. sph.umich.edu/csg/abecasis/QTDT/.

Clinical Characteristics
This study enrolled 1,262 subjects who were members of 379 nuclear families with female children. The subjects included 758 parents and 467 female offspring. As we did not use the QTDT to analyze the effects of parental phenotypes, only the body compositions of the daughters were obtained. The characteristics of the enrolled subjects are listed in Table 1.

SNP Characterization and LD
We initially examined 22 SNPs located within and around LGR4. Because our genotype analysis showed that only one C376T genotype (GG) was present, C376T was excluded from our final statistical analysis. Moreover, rs11029986 had a MAF <0.05 and was also excluded. However, 20 SNPs had a MAF ≥0.07 and displayed Hardy-Weinberg equilibrium. Data obtained from 758 unrelated nuclear family parents were used to calculate the MAF values of the SNPs examined in our study. Detailed information concerning the SNPs is provided in Table 2.

Association Between Haplotypes and Peak BMD/Body Composition
The genotype data obtained from our SNP analysis were used to perform a haplotype analysis that identified groups of three, four, four, four, and five haplotypes in blocks 1, 2, 3, 4, and 5, respectively; these haplotypes accounted for 99.8%, 99.9%, 98.6%, 100%, and 99.0%, respectively, of the unrelated parents in the blocks. We next used the QTDT program to examine how different haplotypes might be associated with peak BMD and obesity phenotypes. .011, and 0.023, respectively). We also found significant total associations between trunk fat mass and haplotypes CC, CT, and TC in block 1 (P = 1.00E-28, 9.00E-29, and 6.00E-29, respectively), between AGCGT in block 3 and both femoral neck and total hip BMD (P = 0.009 and 0.026, respectively), between TGTCTT and total hip BMD (P = 0.029), between TACTTC and femoral neck BMD (P = 0.027), and between CGCCTC and femoral neck, total hip, and lumbar spine BMD in block 5 (P = 0.013, 0.031, and 0.038, respectively). Significant within-family associations were detected between trunk fat mass and haplotypes CC, CT, and TC in block 1 (P = 1.00E-28, 7.00E-29, and 5.00E-29, respectively), between GAA in block 2 and both BMI and trunk fat mass (P = 0.031 and 0.043, respectively), between AGCGT in block 3 and total hip BMD (P = 0.032), between GACCC in block 3 and trunk fat mass (P = 0.049), between TTA in block 4 and TFM (P = 0.033), between TGCTCC in block 5 and both femoral neck and total hip BMD (P = 0.015 and 0.041, respectively), and between TACTTC in block 5 and lumbar spine BMD (P = 0.040). Furthermore, after 1,000 permutations had been performed, significant within-family associations remained between GAA in block 2 and both trunk fat mass and BMI (P = 0.027 and 0.019, respectively), between AGCGT in block 3 and total hip BMD (P = 0.043), between TGCTCC in block 5 and femoral neck BMD (P = 0.025), and between TACTTC in block 5 and both femoral neck and lumbar spine BMD (P = 0.024 and 0.037, respectively). In contrast, no significant within-family associations after 1,000 permutations were found in either block 1 or block 4. The significant haplotype P-values obtained with the QTDT are listed in Table 4, and the non-significant P-values are listed in Table S2.

DISCUSSION
Few studies have examined the association between LGR4 gene polymorphisms and osteoporosis and obesity phenotypes. In 2013, Styrkarsdottir et al. (21) reported the presence of a nonsense mutation [c.376C>T (p.R126X)] in the LCR4 gene of an Icelandic population. This mutation resulted in the production of a truncated and non-functional LGR4 protein, with the resultant physiological effect being a reduction in peak bone mass instead of an increased rate of bone loss, related to increasing age. Luo et al. (37) found that LGR4 -/and LGR4 CKO mice exhibit low levels of BMD, and a subsequent morphometric study showed that the mean number, size, and surface area of the mouse osteoclasts as well as the amounts of eroded bone surface area in those mice were all significantly higher compared with those parameters in the control mice. These investigators also proposed that LGR4 and RANK compete for RANKL binding sites on osteoclasts; thus, LGR4 inhibits osteoclast differentiation and bone tissue remodeling. In contrast, LGR4 appears to function differently in osteoblasts, where it acts through the cAMP-PKA-CREB pathway to regulate ATF4 expression levels. Thus, LGR4 promotes both the differentiation of osteoblasts and the formation of new bone tissue (38). Based on those studies, we focused on examining whether LGR4 SNPs could affect BMD in humans. However, none of our study subjects carried the c.376C>T mutation, consistent with results reported by Zou et al. (23). Our study found significant associations between the polymorphism rs7936621 and the BMD values for the femoral neck, total hip, and lumbar spine. An examination of the haplotype associations in five blocks with BMD values revealed strong associations between haplotype AGCGT in block 3 and total hip BMD, haplotype TGCTCC in block 5 and femoral neck BMD, and haplotype TACTTC in block 5 and both femoral neck and lumbar spine BMD. Based on this biological and statistical evidence, we propose that LGR4 plays a role in regulating BMD or the osteoporotic process.
Additionally, an in vitro study showed that LGR4 ablation potentiates the white-to-brown fat transition that occurs in areas of visceral fat. Meanwhile, a significant correlation has been identified between the LGR4 c.2248G>A variant and both the waist circumference and waist-to-height ratio of young individuals as well as the amount of abdominal visceral fat in young obese subjects (23). Our study identified an association between rs10835171 and trunk fat mass as well as between rs6484295 and BMI, and our Haploview analysis showed that both of these SNPs were located in block 2. We also found that haplotype GAA in block 2 displayed significant within-family associations with trunk fat mass and BMI. Moreover, rs10835172, which is near rs10831571 and was found in block 2, has been reported to significantly correlate with BMI in Chinese individuals (22). In fact, we also observed a relationship between LGR4 SNPs and appendicular fat mass but did not find any association between them. Therefore, we can suppose that either rs10835171 and rs6484295 by themselves or the region around them in block 2 may influence fat metabolism, especially central obesity, which is a known risk factor for both metabolic syndrome and type 2 diabetes.
Further studies are required to elucidate how these SNPs may alter protein function and contribute to human central obesity.
This study has several strengths worth mentioning. First, to ensure a comprehensive analysis of the entire LGR4 gene, we chose to study all 21 of its tagSNPs instead of only a small subset of the SNPs. Second, we chose peak BMD as the osteoporosis phenotype for examination, as this parameter is believed to be genetically controlled. Moreover, a larger number of phenotypes related to obesity were examined in this study than in previous studies. In addition to BMI, our study also examined other parameters including TFM, TLM, PFM, PLM, and trunk fat mass. A recent study showed that trunk fat mass as measured by DAX can be a reliable indicator of total abdominal fat (39), which is closely associated with various metabolic and cardiovascular diseases (38,40). Third, due to the large number of Chinese nuclear families included in our study, our QTDT analysis was able to identify 281, 285, 288, 240, 205, and 163 informative nuclear families suitable for transition disequilibrium analysis at rs1083517, rs6484295, and haplotype GAA in block 2, haplotype AGCGT in block 3, TGCTCC in block 5, and haplotype TACTTC in block 5, respectively. This large sample size enhances the validity of our results. However, the study also has limitations that should be mentioned: First, DAX cannot distinguish between trunk subcutaneous and visceral mass. Second, information concerning the effects of environmental factors such as dietary fat intake and physical exercise, as well as the reproductive history (number of pregnancies, number of children and duration of breastfeeding), years since menopause, physical activity, estrogen replacement therapy or calcium supplements is lacking.
In conclusion, we identified common polymorphisms and haplotypes of LGR4 that are associated with peak BMD, trunk fat mass, and BMI in young Chinese females.

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are publicly available. This data can be found here: https://www.ncbi.nlm.nih. gov/SNP/snp_viewTable.cgi?handle=LGR4-SNP.

ETHICS STATEMENT
The study protocol was reviewed and approved by the Shanghai Jiao Tong University Affiliated Sixth People's Hospital Ethics Committee. Written informed consent to participate in this study was provided by the participants' legal guardian/next of kin.

AUTHOR CONTRIBUTIONS
S-qS, L-lZ and Z-lZ conceived the idea and conceptualized the study. S-sL,X-yZ, ZW, W-zF and J-wH collected the data. ML and Y-qH analyzed the data. S-qS, L-lZ, and Z-lZ drafted the manuscript. All authors contributed to the article and approved the submitted version.

FUNDING
This study was supported by the National Natural Science Foundation of China (81370978) and the Science and Technology Commission of Shanghai Municipality (16411954500). The funding body had no role in the design of the study and collection, analysis, and interpretation of data and in writing the manuscript.