Content uploaded by Qi Liu
Author content
All content in this area was uploaded by Qi Liu on Dec 12, 2022
Content may be subject to copyright.
Received: April Revised: November Accepted: November
DOI: ./ahg.
ORIGINAL ARTICLE
Identification of key variants correlated with susceptibility
of primary osteoporosis in the Chinese Han group
Yanjiao Li1Qi Liu2Qiuye Ma3Zhaoxia Ma1Juan Chen1An Yu1
Changguo Ma1Lihua Qiu1Hong Shi4Hongsuo Liang5Min Hu1
Yunnan Key Laboratory for Basic Research on Bone and Joint Diseases & Yunnan Stem Cell Translational Research Center, Kunming University,
Kunming, China
Human Phenome Institute, Zhangjiang Fudan International Innovation Center, Fudan University, Shanghai, China
Orthopedics, Chongqing Jiulongpo District Hospital of Traditional Chinese Medicine, Chongqing, China
State Key Laboratory of Primate Biomedical Research, Institute of Primate Translational Medicine, Kunming University of Science and Technology,
Kunming, China
Joint Surgery Department of the Second People’s Hospital of Nanning City, Guangxi Zhuang Autonomous Region, Nanning, China
Correspondence
Min Hu, Yunnan Key Laboratory for Basic
Research on Bone and Joint Diseases &
Yunnan Stem Cell Translational Research
Center, Kunming University, Kunming
, China. Hongsuo Liang, Joint
Surgery Department of the Second
People’s Hospital of Nanning City,
Guangxi Zhuang Autonomous Region,
Nanning, China. Hong Shi, State Key
Laboratory of Primate Biomedical
Research, Institute of Primate
Translational Medicine, Kunming
University of Science and Technology,
Kunming, China.
Email: huminynkm@.com;
shih@kust.edu.cn;
lianghongsuo@.com
FUNDING INFORMATION: This
research was funded by grants from
Science & Technology Department of
Yunnan Province (No. ZF).
Abstract
Background: Primary osteoporosis is a systemic skeletal disease characterized
by reduced bone mass and vulnerability to fractures. The genetics of osteoporosis
in the Chinese population remain unclear, which hinders the prevention and
treatment of osteoporosis in China. This study aimed to explore the susceptibility
genes and the roles played by their variants in osteoporosis.
Methods: Blood samples were collected from osteoporosis patients and
healthy individuals, and genome-wide association study was performed on array
data. The expression levels of the candidate gene in different genotypes were
further determined by using quantitative real-time PCR. Moreover, the differ-
entiation capacity of bone marrow mesenchymal stem cells under different
genotypes from osteoporosis patients was investigated.
Results: The most significant variant rs located in the upstream ( bp)
region of CRB2, which could down-regulate the expression levels of CRB2 in
genotype-tissue expression database and played an essential role in the regula-
tion of osteoblastic and osteoclastic differentiation during skeletal development.
Another significant variant rs located within the ′UTR region of TBX3
gene. We found that the mRNA levels of TBX3 decreased in the bMSCs of
old osteoporosis patients. Interestingly, osteoblast differentiation capacity and
TBX3 mRNA levels were similar between the young healthy individuals carrying
derived and ancestral allele of rs, whereas the differentiation capacity and
TBX3 mRNA levels dramatically declined in elderly patients with osteoporosis.
Yanjiao Li, Qi Liu and Qiuye Ma contributed equally.
Ann Hum Genet. ;–. © John Wiley & Sons Ltd/University College London. 1wileyonlinelibrary.com/journal/ahg
14691809, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/ahg.12490 by Fudan University, Wiley Online Library on [11/12/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
2LI .
Conclusions: The variant rs might affect the osteogenesis of bMSCs in
an age-dependent manner and that TBX3 may be a key susceptibility gene for
primary osteoporosis. In conclusion, CRB2 and TBX3 may influence the devel-
opment of osteoporosis; additionally, rs and rs, as the key variants
first reported to be associated with primary osteoporosis, may potentially con-
tribute to predicting the risk of osteoporosis (especially for older individuals) and
may serve as therapeutic targets.
KEYWORDS
age-dependent manner, bMSCs, CRB, osteoporosis, TBX
1 INTRODUCTION
Osteoporosis (OP) is a systemic skeletal disease character-
ized by reduced bone mass, degeneration of the bone tissue
microstructure, increased bone fragility and vulnerabil-
ity to fractures (Sambrook, ). Osteoporotic fractures
(OPFs) represent severe forms of osteoporosis that fre-
quently occur in the distal forearm, upper arm, vertebral
region and hip, accounting for approximately % of cases
(Hernlund et al., ; Kanis et al., ). Osteoporosis
threatens human health, especially in aged people. After
fracture, delayed union or non-union of fracture healing
frequently occurs, resulting in a % mortality within the
first year. To date, the mortality rate of osteoporosis has
exceeded that of breast cancer, cervical and uterine cancers
and continues to increase (Bliuc et al., ). According
to the World Health Organization, osteoporosis ranks sev-
enth among common diseases, with approximately
million patients worldwide, most of whom are elderly and
postmenopausal women. In , the International Osteo-
porosis Foundation reported that one case of OPF occurs
every s worldwide, and approximately % of women
and % of men will encounter their first OPF after the
age of years (Kanis et al., ). An increased inci-
dence of osteoporosis has been observed in China owing to
population aging, which has become an intractable health
problem. It is estimated that at least . million osteo-
porosis patients and million people whose bone mass is
below the normal level are at the risk of developing osteo-
porosis. Treatment of osteoporosis demands huge financial
expenditure and places a heavy burden on families and
society (Chen et al., ; Kanis et al., ).
Osteoporosis is a complex disease, caused by poly-
genic and environmental factors that are influenced by
age and sex, and is generally categorized into primary
and secondary osteoporosis (Wang et al., ). Primary
osteoporosis is a polygenic, multi-factorial and age-related
disease that accounts for the majority of osteoporosis
cases, whereas secondary osteoporosis is induced by other
diseases and drugs. Primary osteoporosis is mainly influ-
enced by genetic factors in addition to environmental
and lifestyle factors, including smoking, drinking, exer-
cise and diet. Although great efforts have been devoted
to researching osteoporosis, the genetic variations and
the mechanisms underlying osteoporosis remain elusive
(Huang et al., ).
Owing to their poor osteogenic and fracture healing
potential, the existing treatments for osteoporosis are
unsatisfactory. To date, drugs for osteoporosis include
bisphosphonates, selective oestrogen receptor modula-
tors, recombinant human parathyroid hormone, strontium
reagents, RANKL inhibitors, anti-sclerostin, anti-DKK
and osteocalcin (Raterman et al., ). Although these
drugs significantly alleviate osteoporosis, they are insuffi-
cient to treat patients. For example, bisphosphonates, such
as alendronate, risedronate and ibandronate, are broadly
applied in osteoporosis prevention and therapy and exert
rapid and mild effects. However, bisphosphonates reach
their maximum performance at – months, and gradually,
their effect declines in a time-dependent manner (Watts &
Diab, ). Thus, further explorations of effective drugs
and methods for the diagnosis, therapy and prevention
of osteoporos are required. The severity of osteoporosis
and the response rate to treatment differs among patients
due to diverse genetic backgrounds. Bone mineral den-
sity (BMD) is the main index for the diagnosis and risk
evaluation of osteoporosis, as the disease is characterized
by a reduced BMD and increased bone fragility (Styrkars-
dottir et al., ). Approximately % of the variation in
BMD can be attributed to genetic factors (Zmuda et al.,
). Previous studies identified multiple susceptibility
genes and key variants related to osteoporosis (Wang et al.,
). For example, genetic variants were reported to
be associated with bone size; additionally, computational
algorithm-based studies reported loci to be associated
with osteoporosis (Qin et al., ).
To date, the majority of genetic studies on osteoporo-
sis have been carried out on participants of European and
14691809, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/ahg.12490 by Fudan University, Wiley Online Library on [11/12/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
LI . 3
American ethnicity. However, genetic studies on osteo-
porosis in the Chinese population are limited. It was
reported that the ADAMTS18 rs and the TGFBR3
rs in the Han population are related to osteo-
porosis; however, the study did not verify the functions
of candidate genes (Xiong et al., ). By analysing the
phenotype data of the BMD family, we previously found
that the phenotype of BMD in the Chinese population is
regulated by genes (data unpublished). In addition, BMD
detection is the main method for osteoporosis diagnosis.
However, BMD detection requires equipment such as an
ultrasonic bone density meter and dual-energy X-ray bone
density meter, which are unsuitable for large-scale popu-
lation screening (Lane, ). On the other hand, variants
detection can be performed via high-throughput screen-
ing and used to evaluate osteoporosis risk decades before
osteoporosis occurs in the young population in order to
take preventive measures.
In this study, we utilized a DNA chip and the sequenc-
ing technology to analyse blood samples from patients
with primary osteoporosis in the Chinese Han group.
We explored a series of variants and candidate genes
associated with osteoporosis and showed that the most sig-
nificant variant rs located in the upstream region
of CRB2, which could down-regulate the expression levels
of CRB2,andTBX3 rs was significantly correlated
with osteoporosis in an age-dependent manner.
2 MATERIALS AND METHODS
2.1 Samples collection
Individuals of Chinese Han group between - and -year
old were enrolled for research. Informed consents were
signed by all patients, and sample collection procedures
were carried out according to the guidelines approved by
the Medical Ethics Committee of Medicine Department of
Kunming University, and the ethical approval reference
number was FL No. , Ethical Review . The BMD
in L–L lumbar spine was measured by dual-energy X-
ray absorptiometry scanning (GE PRODIGY) to determine
whether an individual has developed osteoporosis. The
patients were selected based on the WHO diagnosis cri-
teria of osteoporosis in , which is the bone mass T
value ≤−.SD. Other exclusion criteria were as follows:
(a) using drugs that affect calcium absorption, such as
glucocorticoids, anticonvulsants and chemotherapy drugs;
(b) patients with serious basic diseases such as severe
cardiovascular, lung, liver, kidney and hematopoietic sys-
tem diseases; (c) patients with uncontrolled diabetes and
hypertension; (d) patients with various acute and chronic
infectious diseases; (e) patients with malignant tumours;
(f) patients with a history of prescription drug abuse, ille-
gal drug abuse or alcohol abuse within months before
screening. Overall, osteoporosis patients from to -
year old and healthy individuals from to -year old
were included in the study, and blood samples were col-
lected from each one (Table ). DNA of participants was
extracted by the traditional phenol chloroform method.
All the samples were collected in the morning before
fasting.
2.2 Genome-wide association study
Forty-five osteoporosis patients and healthy indi-
viduals were genotyped by the Illumina Infinium
OmniZhongHua- BeadChip with ∼ K SNPs. Geno-
types were clustered and called by Genome Studio
software, achieving overall .% call rates. Variant posi-
tions are reported in hg/GRCh coordinates. PLINK
v. was used to exclude SNPs and individuals with
more than % missing data, markers with minor allele
frequency <%, markers with failed Hardy–Weinberg
deviation (p<.) and individuals who failed the
X-chromosome sex concordance check (Wang et al., ).
Kinship-based Inference for genome-wide association
study(GWAS)(KING)softwarewasusedtoremove
related individuals with Kinship coefficient >.
(i.e. removing second-degree relatives and higher)
(Manichaikul et al., ). Considering the severity of
osteoporosis was influenced by age, individuals with age
< were filtered out. After quality control, , SNPs
and age-matched individuals, including patients and
healthy people, were retained for further analysis.
Considering the population structure as the major
confounder in GWAS, we performed principal compo-
nent analysis using PLINK v. (Chang et al., ).
Then we calculated the point-biserial correlation coef-
ficient among PC, age and phenotype. All PC and age
were not correlated with phenotype (point-biserial cor
<.), so we conducted the case–control GWAS analysis
using with a simple Chi-squared test without covariate
by PLINK v. (Chang et al., ). Significant vari-
ants (p<.) were picked out as candidate variants.
Then we applied ANNOVAR software to annotate variants
(Wang et al., ). Gene regions in which candidate vari-
ants located were annotated with NCBI Refseq database.
Known variants related to osteoporosis in the GWAS Cat-
alog database (version ) were intersected with
significant variants. Evolution constraint in variants was
assessed with the genomic evolutionary rate profiling
(GERP) scores. Conserved transcription factor binding
sites and microRNA binding sites predicted by TargetScan-
Human were used to evaluate the regulatory function of
14691809, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/ahg.12490 by Fudan University, Wiley Online Library on [11/12/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
4LI .
TABLE 1 Information of the enrolled cases for sequencing
Donor information Number of samples collected Age Gender Bone density selection criteria
Osteoporosis patients cases –-year old Male Tvalue ≤−.SD
Normal individuals cases –-year old Male Tvalue >−SD
variants. The genotype-tissue expression (GTEx) database
was used to determine the relationship between variants
and gene expression. To identify candidate genes related
with osteoporosis, functional enrichment for candidate
genes was performed using the annotation tool KOBAS .
(http://kobas.cbi.pku.edu.cn) with the following database:
pathway database (KEGG pathway, Reactome, BioCyc and
PANTHER), disease database (OMIM, KEGG disease and
NHGRI GWAS Catalog) and gene ontology database. To
obtain the functional annotation for each significant gene,
we conducted enrichment analysis and retain all enrich-
ment results. Then the candidate genes related with osteo-
porosis were picked out using ‘bone’, ‘skeletal’, ‘osteoblast’,
‘osteoporosis’, ‘musculoskeletal’ or ‘collagen’ as keywords.
2.3 Cell culture
Bone marrow mesenchymal stem cells (bMSCs) were
isolated from osteoporosis patients and healthy individu-
als with different ages. All sample collection procedures
were approved by the Medical Ethics Committee of the
Department of Medicine of Kunming University(FL No.
, Ethical Review ). The bMSCs were harvested from
some donors undergoing fracture surgery with informed
consent. The discarded bone fragments were collected to
isolate MSC cells according to Zhu’s protocol (Zhu et al.,
). Briefly, excess soft tissues were removed from bone
chips, and bone chips were flushed out and crushed, and
collagenase-treated. Then the bone fragments were cul-
tured for several days, out of which fibroblast-like cells
migrated and grew in basic MSC medium containing low
glucose DMEM (HyClone) and % fetal bovine serum (Bio-
logical Industries). The isolated bMSCs were characterized
according to the International Society for Cellular Therapy
Guidelines.
The information of cell donors was as follows (Table ).
2.4 Quantitative real-time PCR
(qRT-PCR) and sequencing of candidate
genes
The mRNA levels of candidate genes in bMSCs from both
healthy individuals and osteoporosis patients were deter-
mined by quantitative real-time PCR (qRT-PCR). Total
RNA was extracted from bMSCs using TRIzol reagent
TABLE 2 Information of cell donors
Sample Age (years) Gender
bMSC ( years) Male
bMSC ( years) Female
bMSC ( years) Male
bMSC ( years-#) Female
bMSC ( years-#) Female
bMSC ( years-#) Female
bMSC ( years-#) Female
bMSC ( years-#) Female
Abbreviation: bMSC, bone marrow mesenchymal stem cells.
(Takara Bio) according to the manufacturer’s instructions.
RNA (. μg) was reverse transcribed with PrimeScript RT
reagent Kit (Takara Bio). The cDNA was used for qRT-PCR
detection with an SYBR Premix EX Taq kit (Takara Bio).
The cDNA was amplified using gene-specific primers
listed as follows:
ALP primers:
Forward: ′-GTACGAGCTGAACAGGAACAACG-,
Reverse: ′-CTTGGCTTTTCCTTCATGGTG-′;
Runx2 primers:
Forward: ′-GCTGTTATGAAAAACCAAGT-′,
Reverse: ′-GGGAGGATTTGTGAAGAC-′;
TBX3 primers:
Forward: ′-CCCGGTTCCACATTGTAAGAG-′,
Reverse: ′-GTATGCAGTCACAGCGATGAAT-′;
CDH20 primers:
Forward: ′-GGCTGATGGACCTTACGACC-′,
Reverse: ′-TGTCTGACAGGAGTGCTTCTAAT-′;
ACTB primers:
Forward: ′-CCCTGGAGAAGAGCTACGAG-,
Reverse: ′-GGAAGGAAGGCTGGAAGAGT-.
GAPDH primers:
Forward: ′-TCTGCCCCCTCTGCTGATG-,
Reverse: ′-GTCATGAGTCCTTCCACGATACC-.
The relative mRNA expression levels were normalized
to both ACTB and GAPDH that were used as the internal
control.
For qRT-PCR analysis results, data were shown as
mean ±SD, and statistical analyses were performed by
SPSS .. Significances were determined by one-way
ANOVA or Student’s t-test, for p<. (*) was signifi-
cant, p<. (**) was very significant, p<. (***)
14691809, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/ahg.12490 by Fudan University, Wiley Online Library on [11/12/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
LI . 5
FIGURE 1 Genome-wide association study (GWAS) analysis of variants. (a) Manhattan map of the distribution and statistical values of
variants in the whole genome. The red line denotes the cut-off of the p-value (.). (b) quantile-quantile plot(QQplot) used to analyse the
relevant variation. Using chip data, a Chi-squared test was performed for case–control association analysis. ANNOVAR software was used to
annotate the variation and analyse the position of the variant on the genome, including specific gene regions or inter-gene regions, variation
type of variant located in exon region and variation annotation of regulatory region (promoter region, enhancer region, insulator region etc.).
was extremely significant and p>. (ns) represented
non-significant.
For Sanger sequencing, genomic DNA was extracted
from whole blood or bMSCs. PCR was performed to
amplify the interested region of the target variant in bMSCs
from healthy individuals and osteoporosis patients with
age from to . The PCR products were purified and
subjected to Sanger sequencing for the candidate variants.
2.5 Osteogenic differentiation and
alizarin red S staining
Osteogenic differentiations of bMSCs carrying either wild
type TBX3 or its variants were tested. On day , bMSCs
were seeded onto a -well plate ( ×cells/well). The
medium was refreshed complete medium supplemented
with μg/ml ascorbic acid, mM β-glycerol phosphate
and nM dexamethasone (osteogenic medium) on day
. The medium was changed every – days without
trypsinization and reseeding. The cells were fixed and
stained with alizarin red S solution (Sigma-Aldrich) to
observe osteogenic differentiation on day .
3RESULTS
3.1 Candidate variants and genes
associated with osteoporosis
To explore the potential susceptibility genes and their
variants related to osteoporosis, the samples of osteo-
porosis patients and healthy individuals were genotyped
by using Illumina Infinium OmniZhongHua- BeadChip.
After quality control, the samples of age-matched indi-
viduals, including osteoporosis patients and healthy
individuals, were used to detect candidate variants using
case–control GWAS analysis (Figure ). A total of
variants showed significant association (p<.) with
osteoporosis (Table S). Through enrichment analysis and
keyword search (see detail in Section ), osteoporosis-
related genes were detected (Table ). In these genes,
variants were significantly associated with osteoporosis
(Tables and S).
Next, we filtered out the intronic variants and focused
on the variants located at the upstream region or ′UTR
of the candidate genes, which may regulate gene expres-
sion. The most significant variant detected was rs
(p=. and OR =., % CI =[., .]),
located in the upstream ( bp) region of the crumbs
cell polarity complex component (CRB2) gene (Table ).
This variant showed strongly evolutionary conservation,
as indicated by a GERP score of ., which showed its
regulatory function. In addition, GTEx results indicated
that the derived allele (A) of rs down-regulated
CRB2 expression in four tissues (whole blood, nerve-tibial,
thyroid and uterus; Figure S). CRB2 is involved in the
positive regulation of bone morphogenetic proteins (BMP)
signalling pathway (Table ), which plays an essential role
in the regulation of osteoblastic and osteoclastic differen-
tiation during skeletal development, bone formation and
bone homeostasis. These results indicate that the rs
variant could regulate CRB2 gene expression and influence
the development of osteoporosis.
Another significant variant detected was rs
(p=. and OR =., % CI =[., .]),
located at the ′UTR of the T-Box transcription factor
(TBX3)gene(Tableand Figure S); it showed strongly
evolutionary conservation, as indicated by a GERP score of
., which showed that its potential function is to desta-
bilize mRNA and that it may play an essential role in
regulating TBX3 expression. TBX3 is involved in osteoblast
differentiation during skeletal development (Table ).
These results indicated that the rs variant could reg-
ulate TBX3 gene expression and influence development of
osteoporosis.
14691809, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/ahg.12490 by Fudan University, Wiley Online Library on [11/12/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
6LI .
TABLE 3 Twenty-six candidate genes and their relationship to osteoporosis
Genes Functional term Database Database ID
SOX5|ACVRL1|SMAD3 Positive regulation of chondrocyte
differentiation
Gene ontology GO:
RBPJ|ACVRL1|CRB2 Positive regulation of BMP
signalling pathway
Gene ontology GO:
COL22A1|COL23A1|COL4A2|COL4A1 Collagen formation Reactome R-HSA-
NELL1|SMAD3 Negative regulation of osteoblast
proliferation
Gene ontology GO:
TBX3|PLCB4|GPC6|EVC2|ERBB3 Congenital malformations of the
musculoskeletal system
KEGG DISEASE
COL4A2|COL4A1 Crosslinking of collagen fibrils Reactome R-HSA-
RERE|ADAMTS18|MEPE Bone mineral density NHGRI GWAS Catalog
COL4A1|ANXA2|GPC6|SPARCL1|COL4A2|COL23A1 Collagen-containing extracellular
matrix
Gene ontology GO:
ANXA2 Osteoclast development Gene ontology GO:
LRRC4C Osteoporosis-related phenotypes NHGRI GWAS Catalog
BNC2 Endochondral bone growth Gene ontology GO:
SPARCL1|SMAD3 Collagen binding Gene ontology GO:
MEPE Bone mineral density (paediatric,
total body less head)
NHGRI GWAS Catalog
INPP5D Negative regulation of bone
resorption
Gene ontology GO:
SMAD3 Osteoblast development Gene ontology GO:
TBX3|MEPE Skeletal system development Gene ontology GO:
WWOX Skeletal system morphogenesis Gene ontology GO:
ANO5 Skeletal diseases KEGG DISEASE
ANXA2 Collagen fibril organization Gene ontology GO:
KCNAB1 Skeletal muscle tissue
development
Gene ontology GO:
NELL1 Positive regulation of osteoblast
differentiation
Gene ontology GO:
Abbreviations: BMP, bone morphogenetic proteins; GWAS, genome-wide association study.
3.2 Variant rs1061657 inhibited the
osteoblast differentiation of bMSCs and
reduced TBX3 expression in an
age-dependent manner
To further investigate the effects of the rs variant
on TBX3 expression, we detected TBX3 mRNA levels in
the bMSCs derived from osteoporosis patients at an old
age. Compared with those of the ancestral allele (A) of
the variant rs, the mRNA levels of TBX3 with its
derived allele (G) were significantly reduced using the
housekeeping gene ACTB (Figure )orGAPDH (Figure S)
for normalization. This result suggests that the rs
variant may influence TBX3 expression.
Considering that osteoporosis severity is influenced by
age, we divided the osteoporosis patients and healthy
individuals into young ( >age ≥), old ( >age >)
andsenior(age≥) groups (Figure a). Then, we
detected the osteoblast differentiation capability of bMSCs
and mRNA levels of TBX3 in each group. We found that
the osteogenic capability of bMSCs from both individ-
uals carrying the ancestral allele (A) or derived allele
(G) of rs presented with a similar pattern in the
young healthy group, whereas the osteogenic capabil-
ity of bMSCs from donors carrying the derived allele
(G) of rs was significantly lower than that of the
bMSCs from donors carrying the ancestral allele (A) of
rs in the old and senior groups (Figure b). Con-
sistent with TBX3 expression levels, the mRNA levels of
the osteogenic differentiation genes RUNX2 and ALP in
the donors carrying the derived allele (G) of rs were
remarkably reduced in the old and senior groups compared
to those in the individuals carrying the ancestral allele
(A) using the housekeeping gene ACTB (Figure c–e)) or
14691809, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/ahg.12490 by Fudan University, Wiley Online Library on [11/12/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
LI . 7
TABLE 4 Thirty-seven variants significantly associated with osteoporosis and located in the osteoporosis-related genes
dbSNP ID CHR
BP
(GRCh37
position)
Targe t
allele Freq_Affected Freq_Unaffectedp-Value OR L95 U95 Func Gene
rs A . . . . . . Upstream CRB2
rs T. . . . . . Intronic INPP5D
rs A . . . . . . Intronic SOX5
rs C. . . . . . Intronic PLCB4
rs C . . – – – Intronic RERE
rs C. . –––Intronic BNC2
rs A . . – – – Intronic BNC2
rs C . . . . . . Intronic WWOX
rs A . . . . . . Intronic COL4A2
rs A . . –––Intronic COL23A1
rs T . . . . . . Intronic CRB2
rs C. . . . . Intronic ANO5
rs A . . . . . Intronic ERBB3
rs A. . . . . . Intronic RERE
rs C . . . . . . Intronic RERE
rs C. . . . . . Intronic SMAD3
rs T . . – – – Intronic COL23A1
rs G. . . . . . Intronic KCNAB1
rs C . . . . . . Intronic NELL1
rs C. . . . . . Intronic LRRC4C
rs A . . . . . . Intronic ACVRL1
rs C. . . . . . Intronic KCNAB1
rs A . . . . . Intronic COL22A1
rs A . . . . . Intronic COL4A1
rs A . . . . . Intronic COL4A1
rs T. . . . . . Intronic SPARCL1
rs T . . . . . . Intronic MEPE
rs T. . . . . . Intronic ANO5
rs G . . . . . . Intronic ANO5
rs C. . . . . . Intronic ERBB3
rs C . . . . . . Intronic ADAMTS18
(Continues)
14691809, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/ahg.12490 by Fudan University, Wiley Online Library on [11/12/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
8LI .
TABLE 4 (Continued)
dbSNP ID CHR
BP
(GRCh37
position)
Targe t
allele Freq_Affected Freq_Unaffectedp-Value OR L95 U95 Func Gene
rs C . . . . . . Intronic ANXA2
rs G. . . . . . ′UTR TBX3
rs A . . . . . . Intronic PLCB4
rs T . . – – – Intronic GPC6
rs C . . . . . . Intronic RBPJ
rs T. . . . . . Intronic EVC2
Note:Variantsweresortedbyp-value.
FIGURE 2 Quantitative real-time PCR (qRT-PCR) analysis of
TBX3 expression in the bone marrow mesenchymal stem cells
(bMSCs) derived from osteoporosis patients. Expression of TBX3
carrying a different allele of the rs variant during osteogenic
induction. ACTB was used as the internal reference in the qRT-PCR
test. *p<., **p<., n= (each sample was tested three
times), Student’s t-test, compared to bMSCs (ancestral)−#
GAPDH (Figure SA–C) for normalization. Overall, our
in vitro experiments suggested that the rs vari-
ant inhibited the osteoblast differentiation of bMSCs and
repressed TBX3 mRNA expression in an age-dependent
manner.
4DISCUSSION
Osteoporosis is a major health-threatening disease, partic-
ularly in the elderly (Rodrigues et al., ), and is charac-
terized by deteriorated bone strength and an increased risk
of fracture (Lorentzon & Cummings, ). Due to its high
prevalence, serious outcomes and heavy economic bur-
den, early diagnosis and prevention of osteoporosis have
become particularly important (Morris, ). Although
considerable progress has been made in developing detec-
tion methods and treatment strategies for osteoporosis,
more researches are required for the long-term prediction
of fracture risk and the development of widely applicable
methods and targeted interventions to reduce osteoporosis
risk, particularly in high-risk groups (Wark, ). It is
well known that primary osteoporosis is influenced by
diverse factors, including menopause, glucocorticoid ther-
apy and calcium intake, among which genetic factors are
a fundamental effector of primary osteoporosis. With the
development of sequencing technology, SNPs have been
widely used in diagnostic and therapeutic applications and
14691809, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/ahg.12490 by Fudan University, Wiley Online Library on [11/12/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
LI . 9
FIGURE 3 Osteogenic differentiation of bone marrow mesenchymal stem cells (bMSCs) carrying the rs variant in TBX3 from
different aged individuals. (a) Sequence analysis of bMSCs containing the rs variant. (b)Alizarin red staining of bMSCs on the st day
after osteogenic induction. mRNA levels of RUNX2 (c), ALP (d) and TBX3 (e) in the bMSCs were analysed using quantitative real-time PCR
(qRT-PCR) after osteogenic induction. ACTB was used as the internal reference in the qRT-PCR test. The bMSCs were harvested from eight
different donors who were , , , or -year old. The three -year-old individuals were osteoporosis patients. *p<., **p<.,
***p<., n=; one-way ANOVA, compared to bMSCs ( years) or pairwise comparison labelled with a bar
can be used for the prevention of osteoporosis. Numerous
osteoporosis variants and susceptibility genes have been
reported through GWAS analysis (Li et al., ; Zhu et al.,
). However, only few studies have been conducted
on primary osteoporosis in Asians, and in particular, the
Chinese population; this has limited the development of
preventive and treatment options for osteoporosis.
In this study, we explored primary osteoporosis suscep-
tibility genes in the Chinese Han group and laid emphases
on bone formation with no gender discriminations – as
14691809, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/ahg.12490 by Fudan University, Wiley Online Library on [11/12/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
10 LI .
the oestrogen deficiency due to menopause usually leads
to reduced bone density in women. Thirty-seven vari-
ants in osteoporosis-related genes, including CRB2 and
TBX3, that were significantly related to primary osteo-
porosis in the Chinese Han group were screened out via
GWAS analysis. Furthermore, we found that the most sig-
nificant variant, rs, which down-regulated CRB2
expression, was potentially involved in the development
of osteoporosis. Previous studies have only reported that
the variant rs was associated with colorectal cancer
susceptibility (Fernández-Rozadilla et al., ) and osteol-
ysis susceptibility (Macinnes, ). This study is the first
report about the rs variant as a new signal associ-
ated with osteoporosis. We also showed that another key
variant, rs, in the ′UTR of the TBX3 gene signif-
icantly reduced its mRNA levels in bMSCs, implying its
potential regulatory effect on TBX3 expression. Previous
studies have reported that variants in TBX3 are related to
ulnar mammary syndrome, characterized by limb malfor-
mation (Brummelkamp et al., ). The TBX3 gene can
be induced by WNT and BMP- signalling (Eblaghie et al.,
; Tumpel et al., ), in addition to growth hormone,
to regulate osteoblast proliferation in the bone (Govoni
et al., ). TBX3 is involved in the osteogenic differen-
tiation of human adipose stromal cells (Lee et al., ).
Although TBX3 was identified as a candidate gene in knee
osteoarthritis with site-matched cartilage, there have been
no reports on its relationship with osteoporosis (Zhang
et al., ). So far, the rs variant has only been
reported to be associated with blood pressure (Zhu et al.,
), breast cancer (Zhang et al., ), appendicular lean
mass (Pei et al., ) and insulin-like growth factor levels
(Sinnott-Armstrong et al., ). Our results first suggest
that the derived allele of variant rs antagonized
osteoblast differentiation of bMSCs and suppressed TBX3
mRNA expression, compared with its ancestral allele, in an
age-dependent manner in vitro.
However, GWAS analysis based on a small sample size
has some limitations, including low detection efficiency
and robustness; more samples are needed to investi-
gate and verify the relationships between the candidate
genes/variants found to be involved in osteoporosis in
the present study. In addition, only case–control analy-
sis was conducted to scan candidate susceptibility genes.
Due to the limitation of the sample source, our in vitro
cell experiments detected the effects of the rs vari-
ant within TBX3 on the osteogenic differentiation of the
bMSC from female donors rather than male in a small
sample size (– individuals in each group). Therefore, the
results only provided a preliminary relationship between
TBX3 and osteoporosis without considering the effects of
the rs variant on osteogenic differentiation in both
genders. The analysis with detailed phenotypes is required
to explore the genetic mechanisms underlying osteoporo-
sis. Further research is required to explore the roles of
TBX3 and the rs variant in osteoporosis, for exam-
ple expression quantitative trait loci analysis to determine
the effect of TBX3 expression on osteoporosis in a large
sample size. Furthermore, in order to rule out the impact
of other factors using gene functional tests, gene editing of
the same cell line may be a viable option for future study.
In conclusion, the susceptibility genes, CRB2 and TBX3,
may influence the development of osteoporosis; addi-
tionally, the key variants rs and rs may
potentially contribute to predicting the risk of osteoporo-
sis (especially for older individuals) and may serve as
therapeutic targets.
ACKNOWLEDGMENTS
This work was supported by the Science & Technology
Department of Yunnan Province (No. ZF).
AUTHOR CONTRIBUTIONS
Study design: Min Hu; sample recruitment and selection:
Min Hu, Hongsuo Liang, Qiuye Ma, Zhaoxia Ma; data col-
lection, analysis and interpretation: Qi Liu, Yanjiao Li, Min
Hu, Lihua Qiu; manuscript preparation: Qi Liu, Yanjiao Li,
Juan Chen, An Yu, Changguo Ma; critically reviewed by
Min Hu, Hong Shi. All the authors read and approved the
manuscript.
DATA AVAILABILITY STATEMENT
The data that support the findings of this study are
available from the corresponding author upon reasonable
request.
ORCID
Min Hu https://orcid.org/---
REFERENCES
Bliuc, D., Nguyen, N. D., Milch, V. E., Nguyen, T. V., Eisman, J. A.,
& Center, J. R. (). Mortality risk associated with low-trauma
osteoporotic fracture and subsequent fracture in men and women.
JAMA,301(), –.
Brummelkamp, T. R., Kortlever, R. M., Lingbeek, M., Trettel, F.,
MacDonald, M. E., van Lohuizen, M., & Bernards, R. (). TBX-
, the gene mutated in Ulnar-Mammary Syndrome, is a negative
regulator of pARF and inhibits senescence. JournalofBiological
Chemistry,277(), –.
Chang, C. C., Chow, C. C., Tellier, L. C., Vattikuti, S., Purcell, S. M., &
Lee, J. J. (). Second-generation PLINK: Rising to the challenge
of larger and richer datasets. GigaScience,4,.
Chen, P., Li, Z., & Hu, Y. (). Prevalence of osteoporosis in
China: A meta-analysis and systematic review. BMC Public Health
[Electronic Resource],16(), .
14691809, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/ahg.12490 by Fudan University, Wiley Online Library on [11/12/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
LI . 11
Eblaghie, M. C., Song, S. J., Kim, J. Y., Akita, K., Tickle, C., & Jung,
H. S. (). Interactions between FGF and Wnt signals and Tbx
gene expression in mammary gland initiation in mouse embryos.
Journal of Anatomy,205(), –.
Fernández-Rozadilla, C., Tarrío, R., Clofent, J., de Castro, L., Brea-
Fernández, A., Bessa, X., Abulí, A., Andreu, M., Jover, R., Xicola,
R., Llor, X., Castells, A., Castellví-Bel, S., Carracedo, A., &
Ruiz-Ponte, C., Gastrointestinal Oncology Group of the Spanish
Gastroenterological Association. (). Colorectal cancer sus-
ceptibility quantitative trait loci in mice as a novel approach
to detect low-penetrance variants in humans: A two-stage case-
control study. Cancer Epidemiology and Prevention Biomarkers,
19(), –.
Govoni, K. E., Lee, S. K., Chadwick, R. B., Yu, H., Kasukawa, Y.,
Baylink, D. J., & Mohan, S. (). Whole genome microarray anal-
ysis of growth hormone-induced gene expression in bone: T-box,
a novel transcription factor, regulates osteoblast proliferation.
American Journal of Physiology: Endocrinology and Metabolism,
291(), E–E.
Hernlund, E., Svedbom, A., Ivergård, M., Compston, J., Cooper,
C., Stenmark, J., McCloskey, E. V., Jönsson, B., & Kanis, J. A.
(). Osteoporosis in the European Union: Medical manage-
ment, epidemiology and economic burden. A report prepared
in collaboration with the International Osteoporosis Founda-
tion (IOF) and the European Federation of Pharmaceutical
Industry Associations (EFPIA). Archives of Osteoporosis,8,
.
Huang, Q. Y., Recker, R. R., & Deng, H. W. (). Searchingfor osteo-
porosis genes in the post-genome era: Progress and challenges.
Osteoporosis International,14(), –.
Kanis, J. A., McCloskey, E. V., Johansson, H., Cooper, C., Rizzoli,
R., & Yves Reginster, J. (). European guidance for the diagno-
sis and management of osteoporosis in postmenopausal women.
Osteoporosis International,24(), –.
Kanis, J. A., Oden, A., Johnell, O., Jonsson, B., de Laet, C., & Dawson,
A. (). The burden of osteoporotic fractures: A method for
setting intervention thresholds. Osteoporosis International,12(),
–.
Lane, N. E. (). Epidemiology, etiology, and diagnosis of osteo-
porosis. American Journal of Obstetrics and Gynecology,194(Suppl
), S–S.
Lee, H. S., Cho, H. H., Kim, H. K., Bae, Y. C., Baik, H. S., & Jung,
J. S. (). Tbx, a transcriptional factor, involves in prolifera-
tion and osteogenic differentiation of human adipose stromal cells.
Molecular and Cellular Biochemistry,296(–), –.
Li,S.M.,Guo,H.,Yang,H.J.,Lv,M.Q.,&Zhou,D.X.().HLA-
A gene polymorphisms contribute to osteoporosis susceptibility
in postmenopausal Han Chinese women. Genetics and Molecular
Research [Electronic Resource],14(), –.
Lorentzon, M., & Cummings, S. R. (). Osteoporosis: The evo-
lution of a diagnosis. Journal of Internal Medicine,277(), –
.
Manichaikul, A., Mychaleckyj, J. C., Rich, S. S., Daly, K., Sale, M., &
Chen, W. M. (). Robust relationship inference in genome-wide
association studies. Bioinformatics,26(), –.
Morris, H. A. (). Osteoporosis prevention—A worthy and achiev-
able strategy. Nutrients,2(), –.
Pei, Y. F., Liu, Y. Z., Yang, X. L., Zhang, H., Feng, G. J., Wei, X.
T., & Zhang, L. (). The genetic architecture of appendicular
lean mass characterized by association analysis in the UK Biobank
study. Communications Biology,3(), .
Qin, L., Liu, Y., Wang, Y., Wu, G., Chen, J., Ye, W., Yang, J., &
Huang, Q. (). Computational characterization of osteoporosis
associated SNPs and genes identified by genome-wide association
studies. PLoS One,11(), e.
Raterman, H. G., Bultink, I. E. M., & Lems, W. F. ().
Current treatments and new developments in the manage-
ment of glucocorticoid-induced osteoporosis. Drugs,79(), –
.
Rodrigues, A. M., Eusébio, M., Santos, M. J., Gouveia, N., Tavares, V.,
Coelho, P. S., Mendes, J. M., Branco, J. C., & Canhão, H. (). The
burden and undertreatment of fragility fractures among senior
women. Archives of Osteoporosis,13(), .
Sambrook, P. N. (). Osteoporosis. Medical Journal of Australia,
165(), –.
Macinnes, S. J. (). The genetics of osteolysis and heterotopic ossi-
fication after total HIP arthroplasty[D]. University of Sheffield
School of Medicine, Sheffield, United Kingdom.
Sinnott-Armstrong, N., Tanigawa, Y., Amar, D., Mars, N., Benner,
C., Aguirre, M., Venkataraman, G. R., Wainberg, M., Ollila, H.
M., Kiiskinen, T., Havulinna, A. S., Pirruccello, J. P., Qian, J., &
Shcherbina, A. (). Genetics of blood and urine biomarkers
in the UK Biobank. Nature Genetics,53(), –.
Styrkarsdottir, U., Stefansson, O. A., Gunnarsdottir, K., Thorleifsson,
G., Lund, S. H., Stefansdottir, L., Juliusson, K., Agustsdottir, A.
B., Zink, F., Halldorsson, G. H., Ivarsdottir, E. V., Benonisdottir,
S., Jonsson, H., Gylfason, A., Norland, K., Trajanoska, K., Boer,
C. G., Southam, L., Leung, J. C. S., .. . Kwok, T. C. Y. ().
GWAS of bone size yields twelve loci that also affect height, BMD,
osteoarthritis or fractures. Nature Communications,10(), .
Tumpel, S., Sanz-Ezquerro, J. J., Isaac, A., Eblaghie, M. C., Dobson,
J., & Tickle, C. (). Regulation of Tbx expression by anteropos-
terior signalling in vertebrate limb development. Developmental
Biology,250(), –.
Wang, K., Li, M., & Hakonarson, H. (). ANNOVAR: Functional
annotation of genetic variants from high-throughput sequencing
data. Nucleic Acids Research,38(), e.
Wang, X., Diao, L., Sun, D., Wang, D., Zhu, J., He, Y., Liu, Y., Xu, H.,
Zhang, Y., Liu, J., Wang, Y., He, F., Li, Y., & Li, D. (). Osteo-
porosAtlas: A human osteoporosis-related gene database. PeerJ,7,
e.
Wark, J. D. (). Osteoporosis: Pathogenesis, diagnosis, preven-
tion and management. Bailliere’s Clinical Endocrinology and
Metabolism,7(), –.
Watts, N. B., & Diab, D. L. (). Long-term use of bisphosphonates
in osteoporosis. Journal of Clinical Endocrinology and Metabolism,
95(), –.
Xiong, D. H., Liu, X. G., Guo, Y. F., Tan, L. J., Wang, L., Sha, B. Y.,
Tang, Z. H., Pan,F., Yang, T. L., Chen, X. D., Lei, S. F., Yerges, L. M.,
Zhu, X. Z., Wheeler, V. W., Patrick, A. L., & Bunker, C. H. ().
Genome-wide association and follow-up replication studies iden-
tified ADAMTS and TGFBR as bone mass candidate genes
in different ethnic groups. American Journal of Human Genetics,
84(), –.
Zhang, H., Ahearn, T. U., Lecarpentier, J., Barnes, D., Beesley, J., Qi,
G., Jiang, X., O’Mara, T. A., Zhao, N., Bolla, M. K., Dunning, A.
M., Dennis, J., Wang, Q., Ful, Z. A., Aittomäki, K., Andrulis, I. L.,
Anton-Culver, H., Arndt, V., Aronson, K. J., .. . Becher, H. ().
14691809, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/ahg.12490 by Fudan University, Wiley Online Library on [11/12/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
12 LI .
Genome-wide association study identifies novel breast can-
cer susceptibility loci from overall and subtype-specific analyses.
Nature Genetics,52(), –.
Zhang, Y., Fukui, N., Yahata, M., Katsuragawa, Y., Tashiro, T.,
Ikegawa, S., & Lee, M. T. (). Identification of DNA methyla-
tion changes associated with disease progression in subchondral
bone with site-matched cartilage in knee osteoarthritis. Scientific
Reports,6, .
Zhu, H., Guo, Z. K., Jiang, X. X., Li, H., Wang, X. Y., Yao, H. Y., Zhang,
Y., & Mao, N. (). A protocol for isolation and culture of mes-
enchymal stem cells from mouse compact bone. Nature Protocols,
5(), –.
Zhu, H., Jiang, J., Wang, Q., Zong, J., Zhang, L., Ma, T., Xu, Y.,
& Zhang, L. (). Associations between ERalpha/beta gene
polymorphisms and osteoporosis susceptibility and bone min-
eral density in postmenopausal women: A systematic review and
meta-analysis. BMC Endocrine Disorders,18(), .
Zhu, X., Hou, J., Zhou, B., Xiao, X., & Wang, J. (). Genome-wide
pleiotropy analysis identifies novel blood pressure variants and
improves its polygenic risk scores. Genetic Epidemiology,46(),
–.
Zmuda, J. M., Sheu, Y. T., & Moffett, S. P. (). Genetic epidemiol-
ogy of osteoporosis: Past, present, and future. Current Osteoporosis
Reports,3(), –.
SUPPORTING INFORMATION
Additional supporting information can be found online
in the Supporting Information section at the end of this
article.
How to cite this article: Li, Y., Liu, Q., Ma, Q.,
Ma, Z., Chen, J., Yu, A., Ma, C., Qiu, L., Shi, H.,
Liang, H., & Hu, M. (). Identification of key
variants correlated with susceptibility of primary
osteoporosis in the Chinese Han group. Annals of
Human Genetics,–.
https://doi.org/./ahg.
14691809, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/ahg.12490 by Fudan University, Wiley Online Library on [11/12/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
