Circulating TGF-β1 levels are negatively correlated with sclerostin levels in early postmenopausal women

    Circulating TGF-β1 levels are negatively correlated with sclerostin levels in early postmenopausal women Qun Cheng, Wenjing Tang, Tzong-Jen Sheu, Yanping Du, Jiemin Gan, Huilin Li, Wei Hong, Xiaoying Zhu, Sihong Xue, Xuemei Zhang PII: DOI: Reference:

S0009-8981(16)30032-8 doi: 10.1016/j.cca.2016.01.025 CCA 14258

To appear in:

Clinica Chimica Acta

Received date: Revised date: Accepted date:

15 September 2015 10 December 2015 25 January 2016

Please cite this article as: Cheng Qun, Tang Wenjing, Sheu Tzong-Jen, Du Yanping, Gan Jiemin, Li Huilin, Hong Wei, Zhu Xiaoying, Xue Sihong, Zhang Xuemei, Circulating TGF-β1 levels are negatively correlated with sclerostin levels in early postmenopausal women, Clinica Chimica Acta (2016), doi: 10.1016/j.cca.2016.01.025

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT Circulating TGF-β1 Levels are Negatively Correlated with Sclerostin Levels in Early Postmenopausal Women

IP

T

Qun Cheng a,b, Wenjing Tanga,b, Tzong-Jen Sheuc, Yanping Dua,b, Jiemin Ganb,d, Huilin Lia,b,

a

SC R

Wei Honga,b, Xiaoying Zhua,b, Sihong Xuea,b, Xuemei Zhanga,b

Research Section of Geriatric Metabolic Bone Disease, Shanghai Geriatric Institute,

Department of Osteoporosis and Bone Disease, Huadong Hospital affiliated to Fudan

c

Research Center on Aging and Medicine, Fudan University

MA

b

NU

University

University of Rochester, School of Medicine, USA Central Laboratory, Huadong Hospital affiliated to Fudan University

D

d

TE

Address all correspondence and requests for reprints to: Qun Cheng, M.D., Ph.D.

CE P

Research Section of Geriatric Metabolic Bone Disease, Shanghai Geriatric Institute, Department of Osteoporosis and Bone Disease, Huadong Hospital affiliated with Fudan

AC

University, Research Center on Aging and Medicine, Fudan University, 221 West Yan An Road, Shanghai 200040, China. Cell phone: 86-13918336748. Fax: 86-21-62498319. E-mail: [email protected]

1

ACCEPTED MANUSCRIPT Abstract Background: TGF-β1 regulates bone metabolism and mediates bone turnover during

IP

T

postmenopause. Sclerostin negatively regulates Wnt signaling pathway and also has an

serum TGF-β1 and sclerostin during menopause.

SC R

important role in postmenopausal bone loss. Little is known about relationship between

Methods: We compared serum levels of TGF-β1 and sclerostin in pre- and

NU

postmenopausal women and assessed the potential correlations of these levels with each

MA

other and with serum levels of bone turnover markers and bone mineral density. Results: A total of 176 women (58 premenopausal, 62 early postmenopausal, and 56 late

D

postmenopausal) were included in this study. Serum TGF-β1 level was significantly higher

TE

in early postmenopausal women compared with premenopausal (32.0±7.19 vs. 26.55±6.67

CE P

ng/ml, p=0.01) and late postmenopausal (32.0±7.19 vs. 28.65±7.70 pg/ml, p=0.031) women, and no significant differences in serum sclerostin levels were observed among 3

AC

groups. There was a significant negative correlation between TGF-β1 and sclerostin in early postmenopausal women, but not in other groups of women. Based on multiple regression analysis, only TGF-β1 (β=-0.362; p=0.007) was an independent predictor of sclerostin during early postmenopause. Conclusions: Our findings suggest that serum TGF-β1 level increases during postmenopause and declines in old age. Sclerostin production is inhibited by TGF-β1 during early postmenopause. Keywords: Transforming growth factor β 1; sclerostin; menopause; women; bone 2

ACCEPTED MANUSCRIPT turnover; bone mass Abbreviations

IP

T

TGF-1: transforming growth factor β 1; SOST: sclerostin; L1-4: BMD of lumbar spine 1-4;

SC R

FN: BMD of femur neck; Total hip: BMD of total hip; P1NP: N-terminal propeptide of type I collagen; PTH: parathyroid hormone; VitD: serum 25(OH)D; CTX: cross-linked Ctelopeptide of type I collagen; BGP: osteocalcin; ALP: alkaline phosphatase; FSH: follicle

AC

CE P

TE

D

MA

NU

stimulating hormone; E2: estradiol; Cr: creatinine

3

ACCEPTED MANUSCRIPT 1. Introduction Postmenopausal osteoporosis is characterized by the overproduction of osteoclasts

IP

T

relative to an integrally coupled increase in osteoblastogenesis. Estrogen deficiency may be

SC R

responsible for the rapid bone loss that occurs during the early postmenopausal phase, and this phase of accelerated bone loss may persist for up to 10 or 15 y after menopause in the majority of women [1,2].

NU

Transforming growth factor β 1 (TGF-β1) is a secreted factor that plays an important

MA

role in bone remodeling. It promotes bone formation by augmenting progenitor recruitment, proliferation and differentiation into osteoblasts, helping to maintain the dynamic balance

D

between bone resorption and bone formation [3]. TGF-β1 has long been implicated in the

TE

pathogenesis of postmenopausal bone loss. Some studies have reported that estrogen prevents

CE P

bone loss through a TGF-β1-dependent mechanism, and enhancement of the TGF-β1 level in vivo may constitute a therapeutic approach for preventing bone loss [4]; however, the

AC

presence of TGF-β1 in bone or in serum during menopause remains controversial [5,6]. Sclerostin is produced almost exclusively by osteocytes and regulates bone mass by binding to LRP5/6 to inhibit the canonical Wnt/β-catenin signaling pathway [7]. There are currently clinical trials underway to assess the actions of sclerostin antibodies in therapy for osteoporosis [8]. The circulating sclerostin level has been reported to increase with age and to be negatively correlated with serum-free estrogen in postmenopausal populations [9, 10]. Therefore, studying the mechanisms that regulate sclerostin is important in understanding how estrogen withdrawal regulates bone metabolism and the development of osteoporosis. 4

ACCEPTED MANUSCRIPT Both TGF-β1 and Wnt signaling are important for the regulation of postmenopausal bone turnover. However, whether a correlation exists between TGF-β1 and sclerostin during

SC R

IP

T

menopause has not yet been reported.

2. Subjects and Methods 2.1 Study Population

NU

We conducted a cross-sectional study in healthy women. After excluding subjects with

MA

bone joint-related disease (n=24), ovariectomy (n=9), heart disease (n=5), hyperlipemia (n=6), diabetes (n=8), thyroid disease (n=4) and renal stones (n=2), the study cohort included

D

176 women (age=59.97±15.2 y, BMI=23.13±3.56 kg/m2). To assess the correlation between

TE

TGF-β1 and sclerostin during menopause, we divided our cohort into 3 groups:

CE P

premenopause, early postmenopause, and late postmenopause. Inclusion in the premenopausal group, comprised of 58 women, required the following criteria: greater than

AC

30 years in age, regular menstrual cycles, no history of hysterectomy or bilateral oophorectomy, not pregnant, not taking oral contraceptives. Inclusion in the early postmenopausal group, comprised of 62 women, required menopause to have started <15 y prior to the study. Finally, inclusion in the late postmenopausal group, comprised of 58 women, required menopause to have started greater than 15 years prior to the study. Subjects were recruited for the study by open advertisement at the hospital. All of the subjects were able to walk independently; individuals who were confined to a wheelchair or bed were excluded from the study. According to their recorded medical histories, all participants were 5

ACCEPTED MANUSCRIPT in good health, and none suffered from chronic disease, such as hyperthyroidism, hyperparathyroidism, renal failure, multiple myeloma, leukemia, or chronic arthritis. Patients

IP

T

who suffered from idiopathic bone disease were also excluded from the study, and none of

SC R

the participants were taking any medications that were likely to affect bone or soft tissue metabolism, such as anti-osteoporotic drugs (e.g., glucocorticoids, heparin, warfarin, thyroxine, sex hormones, bisphosphonates, SERMs, calcitonin, PTH analogue, or calcitriol).

NU

The serum levels of TGF-β1, sclerostin, bone turnover markers, calciotropic hormones, and

MA

sex hormones were evaluated in each group, as well as measures of bone mineral density (BMD). All research was conducted at the Osteoporosis and Bone Disease Clinical Research

D

Center of Huadong Hospital, affiliated with Fudan University in Shanghai, China. All of the

TE

subjects provided written informed consent before participating in the study, and the program

CE P

was approved by the Huadong Hospital Ethics Committee. To assess levels of bone turnover markers, calciotropic hormones, sex hormones, TGF-

AC

β1 and sclerostin, serum samples were collected between 0700 and 0900 h after a 10 h fast. For premenopausal women, blood samples were drawn during the first 5-10 days of the menstrual cycle to minimize variability in bone markers and sex hormones that might have been due to the menstrual cycle. Freshly separated serum was then divided into 0.5 ml aliquots and stored at -80°C.

2.2 Serum TGF-β1 and Sclerostin Detection

6

ACCEPTED MANUSCRIPT A commercial enzyme-linked immunosorbent assay (ELISA) kit (TGF-β1 Emax; Promega Corp.) was used for the measurement of total serum TGF-β1 (with intra- and inter-

IP

T

assay coefficients of variability (CVs) of <2.9% and <9.3%, respectively). To measure total

SC R

(active+latent) TGF-β1, the serum samples were acidified using HCl and then reneutralized prior to measurement using NaOH according to the ELISA manufacturer’s instructions [11,12]. The serum sclerostin levels were measured in coded specimens using an ELISA kit

NU

from R&D Systems (with intra- and inter-assay CVs of <2.1% and <10.8%, respectively).

MA

The assay had a minimum detectable dose (MDD) range of 0.37-3.80 pg/ml for sclerostin and of 1.7–15.4 pg/ml for TGF-β1. None of the measured values of TGF-β1 or sclerostin in any

TE

D

of the subjects were below the limits of detection for this assay.

CE P

2.3 Bone Turnover Markers

The assessed bone formation markers included the N-terminal propeptide of type 1

AC

collagen (P1NP; Roche) and serum osteocalcin (BGP; Roche). The assessed bone resorption marker was serum C-telopeptide collagen crosslinks (CTX; Roche). The intra- and inter assay CVs were below 3.5% and 8.4% for CTX, and below 2.6% and 4.1% for P1NP, and below 1.4% and 3.1% for osteocalcin.

2.4 Calciotropic Hormone Measurements

7

ACCEPTED MANUSCRIPT Serum PTH and 25(OH)D were measured by enzyme immunoassays (Roche, with a intra-assay CVs of < 2.7% for PTH and <7.8% for 25(OH)D and a inter-assay CVs of <6.5%

SC R

IP

T

for PTH and <10.7% for 25(OH)D).

2.5 Sex Hormone Measurements

The measured sex hormones included estradiol (E2) and follicle stimulating hormone

NU

(FSH), which were measured using a CL (Beckman Coulter, Inc.), with within-run intra-

MA

assay CVs of 4.3% for E2 and of 3.5% for FSH and between-run inter-assay CVs of 9.1% for E2 and of 7.4% for FSH.

D

2.6 Bone Density Measurements

TE

BMD was measured by dual-energy X-ray absorptiometry (DXA; Hologic Delphi A).

CE P

Based on reproducibility scans, the CVs of the BMD measurements were 1.86% for the

AC

femoral neck, 0.95% for the total hip, and 0.86% for the L1–L4 spine.

2.7 Statistical Analysis Variables are presented as the mean with SD. Differences between the 3 groups were determined either by ANOVA or Kruskal-Wallis tests, depending on the distribution of the variables. Unadjusted and age-adjusted Pearson’s or partial correlation coefficients were used to assess relationships between the serum levels of TGF-β1 and sclerostin and various bone density variables, bone turnover markers, and calciotropic hormone and sex hormone levels. Multivariable linear regression models were developed to calculate the association of serum 8

ACCEPTED MANUSCRIPT sclerostin as a continuous variable with all study variables adjusted for age. A P<0.05 was

IP

T

considered significant. All analyses were performed using SPSS 16.0.

SC R

3. Results

The baseline characteristics of the participants included in the analyses are shown in Table 1. On average, the included premenopausal women were 39.4±7.6 years old, the early

NU

postmenopausal women were 61.6±5.7 y, and the late postmenopausal women were 78.3±5.7

MA

y. As expected, the BMD was significantly lower at the lumbar spine, femoral neck and total hip in the postmenopausal women compared with the premenopausal women (Table 1). The

D

serum levels of the bone formation markers P1NP and BGP and the bone resorption marker

TE

CTX were significantly higher (50-90%) in the postmenopausal women compared with the

CE P

premenopausal women; serum ALP levels were also increased in the same manner (40-60%) (Table 1). The level of E2 was significantly lower and that of FSH was significantly higher in

AC

the postmenopausal women, and the serum PTH and serum creatinine levels increased slightly during the late phase of postmenopause compared with premenopause (p<0.05). However, there were no significant differences in serum levels of 25(OH)D, calcium or phosphorus among the 3 groups (Table 1). The total serum TGF-β1 level was significantly higher in the early postmenopausal women compared with the premenopausal women (32.0±7.19 vs. 26.55±6.67 ng/ml, p=0.01), and it was significantly decreased in the late postmenopausal women relative to the early postmenopausal women (28.65±7.70 vs. 32.0±7.19 ng/ml, p=0.034) but showed no difference 9

ACCEPTED MANUSCRIPT between the premenopausal and late postmenopausal women. In addition, there were no significant differences in the serum sclerostin levels among the 3 groups, as shown in Fig. 1.

IP

T

In all of the groups, after adjusting for age, there was a significant positive correlation

SC R

between serum levels of TGF-β1 and bone turnover markers, such as P1NP (r=0.347; p=0.007), CTX (r=0.328; p=0.011) and ALP (r=0.531; p=0.000). Serum TGF-β1 was also positively correlated with BMI (r=0.563; p=0.007) and serum creatinine levels (r=0.422;

NU

p=0.002), suggesting that serum TGF-β1 levels may be more significantly reduced in late

MA

postmenopausal women compared with early postmenopausal women when renal excretion function is considered. However, serum levels of TGF-β1 were negatively correlated with

D

those of 25(OH)D (r=-0.388; p=0.002) in the entire group when adjusted by age. Serum

TE

sclerostin was significantly positively correlated with BMD in the lumbar spine (r=0.158;

CE P

p=0.035), femur neck (r=0.207; p=0.016), and total hip (r=0.233; p=0.012) in the entire group, even after adjusting for age (Table 2), with a trend toward a significant negative correlation

AC

with the serum PTH level (r=-0.321; p=0.013) after adjustment for age.. Additionally, no significant correlation was observed between the serum TGF-β1 and serum sclerostin levels in the entire group (Table 2). However, further analysis demonstrated that the significant negative relationship between serum TGF-β1 and serum sclerostin was limited to the early postmenopausal women (r=-0.274; p=0.026), even after adjusting for age, BMI, and the serum creatinine level (r=-0.633; p=0.007). No significant correlation between the serum TGF-β1

rved in the other 2 groups (Table 3).

10

ACCEPTED MANUSCRIPT The serum sclerostin level was positively correlated with age (r=0.452; p=0.018) in the premenopausal women and positively correlated with height (r=0.249; p=0.044) in the early

IP

T

postmenopausal women. In addition, it was positively correlated with bone mass (r=0.406-

SC R

0.446; p=0.000-0.001) and the serum FSH level (r=0.244; p=0.044) and negatively correlated with serum ALP and serum phosphorus (r=-0.276, -0.306; p=0.017, 0.008) in the late postmenopausal women. Despite being positively correlated with the serum levels of bone

NU

formation markers in all 3 groups (r=0.247-0.417; p=0.004-0.036), the serum TGF-β1 level

MA

was negatively correlated with age (r=-0.399; p=0.001) in the early postmenopausal women and positively correlated with BMI and serum creatinine (r=0.492, 0.496; p=0.000, 0.003) in

D

the late postmenopausal women (Table 3).

TE

Using multivariate linear regression analysis, serum TGF-β1 (β=-0.362; p=0.007) was

CE P

found to be an independent predictor of serum sclerostin in the early postmenopausal women after adjusting for age; however, the independent predictors of the serum sclerostin level

AC

included both age (β=0.525; p=0.002) and the serum BGP level (β=-0.495; p=0.004) in the premenopausal women and serum phosphorus (β=-0.342; p=0.002), serum ALP (β=-0.319; p=0.003) and bone mass of the lumbar spine (β=0.252; p=0.022) in the late postmenopausal women (Table 4).

4. Discussion TGF-β is one of the most abundant cytokines in the bone matrix (200 g/kg) [13]. Bone extracellular matrix (ECM) is the major storage site for TGF-β in the body. ECM-bound 11

ACCEPTED MANUSCRIPT TGF-β, which is predominantly comprised of the TGF-β1 isoform, is stored in a latent form and can be released and activated by resorbing osteoclasts [14]. TGF-β1 is present at a

IP

T

physiologically significant level in plasma, and this source may contribute to the reservoir

SC R

stored in the bone matrix [15]. Once released from the matrix and activated, TGF-β1 can influence many of the steps in the remodeling pathway, including the inhibition of osteoclast formation and activity, the stimulation of osteoblast precursor recruitment and proliferation,

NU

and the stimulation of mature osteoblasts to produce bone matrix proteins [16]. Thus, TGF-β1

MA

has been implicated as a coupling factor that coordinates the processes of bone resorption and subsequent bone formation. However, the effect of TGF-β1 on osteocytes is still unclear.

D

Sclerostin is secreted nearly exclusively by osteocytes, and measurement of the serum

TE

sclerostin level can reveal the activity and function of osteocytes embedded in bone matrix

CE P

[17]. In this study, we investigated the concentrations of and the correlation between the serum TGF-β1 and sclerostin levels in premenopausal, early and late postmenopausal women

AC

and attempted to determine the effects of TGF-β1 on osteocytes during each phase of menopause.

Our results indicated that the serum TGF-β1 level was relatively low in the premenopausal women and that menopause had a biphasic effect on serum TGF-β1, which increased during the early stage of estrogen deficiency (early postmenopause) but decreased during the late stage of estrogen deficiency (late postmenopause). This finding is consistent with a previous report by Dallas et al. demonstrating that matrix-bound TGF-β1 is released during the early stage of ovariectomy, at which time osteoclasts are activated [14]. Our 12

ACCEPTED MANUSCRIPT results also indicated that after long-term estrogen deficiency, serum TGF-β1 declined to levels similar to those found in premenopausal women (P>0.05), which is consistent with

IP

T

prior reports that endogenous TGF-β1 in bone matrix may be exhausted after long-term bone

SC R

loss [18].

The circulating sclerostin level has been reported to increase with age and to be moderately associated with BMD and bone turnover [19]; however, in our study, no

NU

differences in this level were found among any of the groups. Multiple linear regression

MA

analysis indicated that the serum sclerostin level was positively associated with age in premenopausal women, but not in postmenopausal women. This finding is consistent with a

D

report by Zhang et al., who did not observe any changes in the serum sclerostin level in

TE

relation to aging in postmenopausal women in China [20]. This finding is also consistent with

CE P

the results reported by Karin et al., who found a positive correlation between age and serum sclerostin in premenopausal women (r=0.66; P<0.001) [21]. However, many studies based on

AC

Western populations have noted that the serum sclerostin level is significantly increased after menopause, as a greater quantity of sclerostin is produced by osteocytes during aging [9,19,22]. The differing results of our study may have been due to the examination of different study populations and the use of different study methods. Furthermore, some studies have suggested that the correlation between age and the serum sclerostin level differs between men and women, reporting that sclerostin level is significantly higher in men than in women (33.3 pmol/l vs. 23.7 pmol/l, p<0.001), and over the course of a lifetime, it does not increase to the same degree in women as in men [19]. The higher quantity of circulating 13

ACCEPTED MANUSCRIPT sclerostin present in the elderly population has been associated with higher marrow fat mass in men, but not in women [23]. In individuals with type 2 diabetes mellitus, the serum

IP

T

sclerostin level has been positively correlated with age in males, but not in females [24].

SC R

However, in an animal study, sclerostin expression in bone tissue has been shown to be decreased in OVX mice, and no difference in its serum level was observed between OVX and SHAM mice [25]. Therefore, the changes in the sclerostin level that occur during aging are

NU

complex; since sclerostin production by individual osteocytes increases, but there is a

MA

reduction in bone tissue and therefore reduced osteocytes with aging. In our study, serum sclerostin was positively correlated with BMD in the lumbar spine,

D

femur neck, and total hip and negatively correlated with intact serum PTH in the entire

TE

sample, which was similar to the results of several previous studies [19,26,27]. Because

CE P

sclerostin is produced nearly exclusively by osteocytes, the relationship between the serum sclerostin level and total body bone mass is determined by the osteocyte content in bone

AC

tissue, and PTH can regulate sclerostin production negatively in vivo or in vitro. The level of serum TGF-β1 was significantly related to the level of bone turnover markers, especially serum osteoblast markers. There was a significant positive correlation between the serum TGF-β1 and bone turnover markers in the entire sample, suggesting that serum TGF-β1 level was primarily regulated by bone turnover status with respect to be produced by osteoblasts and released by osteoclasts. However, complex relationships exist between serum sclerostin level and other factors, such as the estrogen status of women, bone turnover levels, and bone mass, and each factor showed a different dominant effect during 14

ACCEPTED MANUSCRIPT different phases. In premenopause, bone turnover and bone mass are relatively stable; thus, age has the most important effect on sclerostin level since sclerostin production increases

IP

T

with aging during premenopause. In early postmenopause, only TGF-β1 was an independent

SC R

predictor of sclerostin, however, age, bone mass and bone turnover markers didn’t show any effect. During late postmenopause, especially in a state of low bone turnover, the sclerostin

NU

level is significantly influenced by bone mass, which determines the number of osteocytes.

Several pathways that control bone development and metabolism regulate SOST

MA

expression. The BMP pathway induces SOST expression during bone development [28].

D

PTH, an essential regulator of mineral homeostasis, represses SOST expression [29]. The

TE

rapid increase in PGE2 following mechanical load also contributes to the mechanosensitive repression of SOST [30]. Recently data from Eric et al. revealed that inhibition of TGF-β

CE P

mediated Smad3 phosphoeylation strongly up-regulates sclerostin mRNA expression; treatment of osteoblast with TGF-β itself represses sclerostin mRNAexpression, and

AC

osteoblast-specific deficiency in TGF-β signaling produces a 3.2 fold increase of sclerostin in bone cells around the epiphysis in mice. ELISA analysis also indicates that there is a significant systemic increase in sclerostin levels in both Smad3 heterozygous and Smad3-KO mice as compared with wild-type controls. Western blot analyses from bone marrow cells of Smad-KO animals further confirm the immunohistochemical findings of elevated sclerostin. These data imply that canonical TGF-β signaling can repress sclerostin production in osteoblasts [31]. Furthermore, intact TGF-β signaling is required for load to repress Sclerostin expression and induce bone formation [32]. 15

ACCEPTED MANUSCRIPT A relationship between TGF-β1 and sclerostin during menopause has not been previously demonstrated. In addition, we are aware of few studie that have examined the

IP

T

effects of TGF-β1 on sclerostin production by osteocytes or on the release of sclerostin from

SC R

bone during menopause in vivo or in vitro. In this study, we found a significant negative relationship between serum TGF-β1 and serum sclerostin in the early postmenopausal women, but not in the premenopausal women or late postmenopausal women, and we did not find a

NU

trend toward a relationship between serum TGF-β1 or sclerostin and the serum E2 or FSH

MA

level. These findings could have been due to the TGF-β1-mediated inhibition of the production of sclerostin by osteocytes or of the release of sclerostin from the skeleton. It is

D

also possible that the increased bone resorption associated with estrogen deficiency causes a

TE

large release of TGF-β1 from bone into the canaliculi; however, TGF-β1 inhibits the

CE P

expression of sclerostin by osteocytes through paracrine activity and relieves depression of the Wnt signaling pathway to promote bone anabolic activity in the process of rapid bone

AC

turnover during postmenopause. TJ et al. [33] analyzed inducible, osteoblast-specific (Collagen-I-Cre ERT2) TGF-βRII knockout mice and found that these mice exhibited significant bone loss and a high sclerostin expression level; however, the levels of the other 3 negative regulators of Wnt signaling (exFRZb, DKK, and AXIN) did not significantly change. Moreover, sclerostin knockout mice can rescue this osteoblast-specific TGF-βRII-/--induced bone loss completely. This result demonstrates that the effects of TGF-β1 on bone metabolism may partially by sclerostin inhibition. 16

ACCEPTED MANUSCRIPT Therefore, TGF-β1 released during osteoclast-mediated bone resorption feeds back on osteoblasts, balancing matrix resorption with new bone deposition in 2 steps. First, TGF-β1-

IP

T

directed migration of bone-derived mesenchymal progenitors to resorptive sites is an

SC R

essential step in the coupling process. Second, TGF-β1 promotes osteogenesis from those mesenchymal progenitors by inhibiting the production of sclerostin in osteocytes. These 2 steps cause the coupling of osteoblasts to osteoclasts during menopause.

NU

The main limitations of our study are its relatively small size and its cross-sectional

MA

design. We presented data from a healthy population with normal hepatic and renal function and without pre-existing medical conditions. Therefore, the conclusions drawn from our data

D

must be applied with caution to other populations, particularly to populations of men and

TE

unhealthy individuals. To further examine these relationships, large and longitudinal studies

CE P

must be performed to determine the temporal relationship between circulating TGF-β1 and sclerostin. However, this is the first study to link the level of serum TGF-β1 with that of

AC

serum sclerostin in healthy premenopausal and postmenopausal women. In conclusion, our study has demonstrated that the serum TGF-β1 level is significantly higher in early postmenopausal women than in premenopausal women and that a significant negative correlation exists between serum TGF-β1 and serum sclerostin during early postmenopause. Our results imply that the induction of TGF-β1 during the early phase after menopause is a compensatory mechanism to promote osteoblast activity, prevent osteoblast apoptosis and balance bone turnover while maintaining bone mass. Additionally, the

17

ACCEPTED MANUSCRIPT relationship that exists between TGF-β1 and sclerostin must be further explored in larger

IP

T

prospective studies.

SC R

Acknowledgments

We thank all participants of the study and the involved laboratory staff. This study was supported by the National Natural Science Foundation of China, Shanghai

NU

Municipal Commission of Health, and Shanghai Key Laboratory of Clinical Geriatric

AC

CE P

TE

D

MA

Medicine.

18

ACCEPTED MANUSCRIPT References [1]

Khosla S, Melton LJ , Atkinson EJ, O'Fallon WM, Klee GG, Riggs BL. Relationship

IP

T

of serum sex steroid levels and bone turnover markers with bone mineral density in men and

SC R

women: a key role for bioavailable estrogen. J Clin Endocrinol Metab 1998;83:2266-74. doi: 10.1210/jcem.83.7.4924. [2]

Nicks KM, Fowler TW, Akel NS, Perrien DS, Suva LJ, Gaddy D. Bone turnover

NU

across the menopause transition : the role of gonadal inhibins. Ann N Y Acad Sci

[3]

MA

2010;1192:153-60. doi: 10.1111/j.1749-6632.2009.05349.x. Alliston T, Piek E, Derynck R. TGF-β family signaling in skeletal development,

Gao Y, Qian WP, Dark K, Toraldo G, Lin AS, Guldberg RE, et al. Estrogen prevents

CE P

[4]

TE

2008;50:667-724.

D

maintenance, and disease; TGF-β family. Cold Spring Harbor Monogragh Archive

bone loss through transforming growth factor β signaling in T cells. Proc Natl Acad Sci U S

[5]

AC

A 2004;101:16618-23. doi: 10.1073/pnas.0404888101. Chen C, Liang MK, Zhang H, Peng YQ, Wu XP, Wu XY, et al. Relationships

between age-related biochemical markers of bone turnover and OPG, TGF-β1 and TGF-β2 in native Chinese women. Endocr Res 2014;39:105-14. doi: 10.3109/07435800.2013.840654. [6]

Pepene CE, Seck T, Diel I, Minne HW, Ziegler R, Pfeilschifter J. Influence of fluor

salts, hormone replacement therapy and calcitonin on the concentration of insulin-like growth factor (IGF)-I, IGF-II and transforming growth factor-β 1 in human iliac crest bone matrix from patients with primary osteoporosis. Eur J Endocrinol 2004;150:81–91. doi: 19

ACCEPTED MANUSCRIPT 10.1530/eje.0.1500081. [7]

Li X, Zhang Y, Kang H, Liu W, Liu P, Zhang J, et al. Sclerostin binds to LRP5/6 and

IP

T

antagonizes canonical Wnt signaling. J Biol Chem 2005;280:19883–7. doi:

[8]

SC R

10.1074/jbc.M413274200.

Baron R, Hesse E. Update on bone anabolics in osteoporosis treatment: rationale,

current status, and perspectives. J Clin Endocrinol Metab 2012;97:311-25. doi:

Ardawi MS, Al-Kadi HA, Rouzi AA, Qari MH. Determinants of serum sclerostin in

MA

[9]

NU

10.1210/jc.2011-2332.

healthy pre- and postmenopausal women. J Bone Miner Res 2011;26:2812-22. doi:

Mirza FS, Padhi ID, Raisz LG, Lorenzo JA. Serum sclerostin levels negatively

TE

[10]

D

10.1002/jbmr.479.

CE P

correlate with parathyroid hormone levels and free estrogen index in postmenopausal women. J Clin Endocrinol Metab 2010;95:1991-7. doi: 10.1210/jc.2009-2283. Wakefield LM, Letterio JJ, Chen T, Danielpour D, Allison RS, Pai LH, et al.

AC

[11]

Transforming growth factor-β1 circulates in normal human plasma and is unchanged in advanced metastatic breast cancer. Clin Cancer Res 1995;1:129–36. [12]

Dallas SL, Sivakumar P, Jones CJ, Chen Q, Peters DM, Mosher DF, et al. Fibronectin

regulates latent transforming growth factor-β (TGF β) by controlling matrix assembly of latentTGF β-binding protein-1. J Biol Chem 2005;280:18871–80. doi: 10.1074/jbc.M410762200. [13]

Seyedin SM, Thomas TC, Thompson AY, Rosen DM, Piez KA. Purification and 20

ACCEPTED MANUSCRIPT characterization of two cartilage-inducing factors from bovine demineralized bone. Proc Natl Acad Sci U S A 1985;82:2267–71. doi: 10.1073/pnas.82.8.2267.

T

Dallas SL, Rosser JL, Mundy GR, Bonewald LF. Proteolysis of latent transforming

IP

[14]

SC R

growth factor-β (TGF-β )-binding protein-1 by osteoclasts. A cellular mechanism for release of TGF-β from bone matrix. J Biol Chem 2002;277:21352-60. doi: 10.1074/jbc.M111663200.

Janssens K, ten Dijke P, Janssens S, Van Hul W. Transforming growth factor-β1 to

NU

[15]

[16]

MA

the bone. Endocr Rev 2005;26:743–74. doi: 10.1210/er.2004-0001. Tang Y, Wu X, Lei W, Pang L, Wan C, Shi Z, et al. TGF-β1-induced migration of

D

bone mesenchymal stem cells couples bone resorption with formation. Nat Med

Durosier C, van Lierop A, Ferrari S, Chevalley T, Papapoulos S, Rizzoli R.

CE P

[17]

TE

2009;15:757-65. doi: 10.1038/nm.1979.

Association of circulating sclerostin with bone mineral mass, microstructure, and turnover

AC

biochemical markers in healthy elderly men and women. J Clin Endocrinol Metab 2013;98:3873-83. doi: 10.1210/jc.2013-2113. [18]

Jia J, Yao W, Amugongo S, Shahnazari M, Dai W, Lay YA, et al. Prolonged

alendronate treatment prevents the decline in serum TGF-β1 levels and reduces cortical bone strength in long-term estrogen deficiency rat model. Bone 2013;52:424-32. doi: 10.1016/j.bone.2012.10.017. [19]

Mödder UI, Hoey KA, Amin S, McCready LK, Achenbach SJ, Riggs BL, et al.

Relation of age, gender, and bone mass to circulating sclerostin levels in women and men. J 21

ACCEPTED MANUSCRIPT Bone Miner Res 2011;26:373–9. doi: 10.1002/jbmr.217. [20]

He J, Zhang H, Wang C, Zhang Z, Yue H, Hu W, et al. Associations of serum

IP

T

sclerostin and polymorphisms in the SOST gene with bone mineral density and markers of

2014;99:E665-73. doi: 10.1210/jc.2013-2086. [21]

SC R

bone metabolism in postmenopausal Chinese women. J Clin Endocrinol Metab

Amrein K, Amrein S, Drexler C, Dimai HP, Dobnig H, Pfeifer K, et al. Sclerostin and

NU

its association with physical activity, age, gender, body composition, and bone mineral

MA

content in healthy adults. J Clin Endocrinol Metab 2012;97:148-54. doi: 10.1210/jc.20112152.

Amrein K, Amrein S, Drexler C, Dimai H, Pieber TR, Tomaschitz A, et al. Sclerostin

D

[22]

TE

is associated with age, BMI, and bone mineral content, but not gender or physical activity in

CE P

healthy adults. Proceedings of the 33rd Annual Meeting of the American Society for Bone and Mineral Research, San Diego, CA; 2011, p. S338 (Abstract SU0346). Ma YH, Schwartz AV, Sigurdsson S, Hue TF, Lang TF, Harris TB, et al. Circulating

AC

[23]

sclerostin associated with vertebral bone marrow fat in older men but not women. J Clin Endocrinol Metab 2014;99:E2584-90. doi: 10.1210/jc.2013-4493. [24]

García-Martín A, Rozas-Moreno P, Reyes-García R, Morales-Santana S, García-

Fontana B, García-Salcedo JA, et al. Circulating levels of sclerostin are increased in patients with type 2 diabetes mellitus. J Clin Endocrinol Metab 2012;97:234–41. doi: 10.1210/jc.2011-2186. [25]

Jastrzebski S, Kalinowski J, Stolina M, Mirza F, Torreggiani E, Kalajzic I, et al. 22

ACCEPTED MANUSCRIPT Changes in bone sclerostin levels in mice after ovariectomy vary independently of changes in serum sclerostin levels. J Bone Miner Res 2013;28:618-26. doi: 10.1002/jbmr.1773.

T

Sheng Z, Tong D, Ou Y, Zhang H, Zhang Z, Li S, et al. Serum sclerostin levels were

IP

[26]

SC R

positively correlated with fat mass and bone mineral density in central south Chinese postmenopausal women. Clin Endocrinol (Oxf) 2012;76:797-801. doi: 10.1111/j.13652265.2011.04315.x.

Polyzos SA, Anastasilakis AD, Bratengeier C, Woloszczuk W, Papatheodorou A,

NU

[27]

MA

Terpos E. Serum sclerostin levels positively correlate with lumbar spinal bone mineral density in postmenopausal women--the six-month effect of risedronate and teriparatide.

Kamiya N, Ye L, Kobayashi T, Mochida Y, Yamauchi M, Kronenberg HM, et al.

TE

[28]

D

Osteoporos Int 2012;23:1171-6. doi: 10.1007/s00198-010-1525-6.

CE P

BMP signaling negatively regulates bone mass through sclerostin by inhibiting the canonical Wnt pathway. Development 2008;135: 3801–11. doi: 10.1242/dev.025825. Keller H, Kneissel M. SOST is a target gene for PTH in bone. Bone 2005;37:148–58.

[30]

Kitase Y, Barragan L, Qing H, Kondoh S, Jiang JX, Johnson ML, et al. Mechanical

AC

[29]

induction of PGE2 in osteocytes blocks glucocorticoid-induced apoptosis through both the bcatenin and PKA pathways. J Bone Miner Res 2010;25: 2657–68. doi: 10.1002/jbmr.168. [31]

Eric E, Tzong S, Deborah D, Jonathan D, Resika U, Philip B, et al. Heavy metal ion

regulation of gene expression mechanisms by which lead inhibits osteoblastic bone forming activity through modulation of the wnt/b-catenin signaling pathway. J Bio Chem 2015;290:18216-26. doi: 10.1074/jbc.M114.629204. 23

ACCEPTED MANUSCRIPT [32]

Nguyen J, Tang S, Nguyen D, Alliston T. Load regulates bone formation and

sclerostin expression through a TGF β dependent mechanism. PloS One 2013;8,e53813. doi:

IP

Hsu SC. I-Hui S, Shi S, Ho H-C, Sheu T-J, Hsu W, et al. Aged-related bone loss

SC R

[33]

T

10.1371/journal.pone.0053813.

(osteopenia) in old male mice results from diminished activity and availability of TGF-b and WNT signaling. Proc 36rd Ann Meet Amer Soc Bone Min Res, Houston, Texas, USA

AC

CE P

TE

D

MA

NU

(Abstract SA0458); 2014, p325-326.

24

ACCEPTED MANUSCRIPT

AC

CE P

TE

D

MA

NU

SC R

IP

T

Disclosure Summary: The authors have no conflicts of interest in association with this manuscript.

25

ACCEPTED MANUSCRIPT

Figure Legends

IP

T

Fig. 1. Median, 25th and 75th percentiles of serum TGF-β1 and SOST in the 3 different

AC

CE P

TE

D

MA

NU

SC R

groups

26

ACCEPTED MANUSCRIPT Table 1. Baseline characteristics Premenopausal women

Early postmenopausal

Late postmenopausal

(n=176)

(n=58)

women

women

(n=62)

(n=56)

T

Total

59.97±15.2

39.4±7.6

61.6±5.7**

78.3±5.7**,$$

Height (m)

156.5±7.2

160.9±5.6

159.3±6.0

149.4±4.0**,$$

Weight (kg)

56.6±9.1

56.5±7.2

BMI (kg/m2)

23.13±3.56

21.8±2.6

BMD (L1-4)

0.826±0.16

0.97±0.12

BMD (FN)

0.633±0.13

BMD (Total hip)

SC R

IP

Age (y)

52.8±8.6$

24.0±4.0

23.6±3.6*

0.81±0.15**

0.700±0.12**

0.75±0.11

0.65±0.10**

0.51±0.07**,$$

0.757±0.15

0.89±0.11

0.77±0.13**

0.61±0.09**,$$

TGF-β1 (ng/ml)

29.48±6.0

26.55±6.67

32.0±7.19*

28.65±7.70$

SOST (pg/ml)

87.33±24.6

101.55±22.6

88.79±23.7

79.34±24.0

PINP (ng/ml)

48.26±16.0

41.13±10.2

50.60±15.9**

55.84±25.4**

PTH (pg/ml)

44.11±16.5

40.89±15.5

43.46±16.7

54.98±42.4*

VitD (ng/ml)

20.76±7.45

22.63±6.8

20.29±6.8

19.05±8.8

CTX (pg/ml)

470.37±215.4

323.71±119.6

494.01±181.3**

613.85±279.3**

BGP (ng/ml)

21.35±7.32

16.19±3.7

23.28±7.2**

28.02±1.4**

75.96±26.8

57.04±11.4

82.69±17.6**

88.79±34.9**

2.33±0.08

2.34±0.1

2.34±0.1

2.32±0.1

1.18±0.13

1.20±0.2

1.17±0.1

1.19±0.1

ALP (IU/l) Serum Ca (mmol/l)

AC

Serum P (mmol/l)

CE P

TE

D

MA

NU

60.7±9.7

Serum FSH (IU/l)

50.74±35.5

15.65±4.7

66.94±21.0**

70.39±18.86**

Serum E2 (pmol/l)

168.83±247.0

348.72±130.1

70.64±24.1**

80.45±14.6**

Serum Cr (mmol/l)

62.33±11.23

55.3±6.7

57.9±8.0

71.3±8.8*,$

Data are presented as the mean±SD. TGF-β1=transforming growth factor β 1; SOST=sclerostin; L1-4=BMD of lumbar spine

1-4; FN=BMD of femur neck; Total hip=BMD of total hip; P1NP=N-terminal propeptide of type I collagen;

PTH=parathyroid hormone; VitD=serum 25(OH)D; CTX=cross-linked C-telopeptide of type I collagen; BGP=osteocalcin;

ALP=alkaline phosphatase; FSH=follicle stimulating hormone; E2=estradiol; Cr=creatinine; * p<0.05, **p<0.01 compared with premenopausal women, $ p<0.05, $$p<0.01 compared with early postmenopausal women

27

ACCEPTED MANUSCRIPT Table 2. Correlation coefficients (unadjusted/age-adjusted) of TGF-β1 and SOST with bone density, bone turnover markers, calciotropic hormones and sex hormones in the

SOST

IP

TGF-β1

T

entire sample 1

SOST

-0.165/-0.195

1

BMI

0.273**/0.563**

0.024/0.070

L1-4

0.123/0.058

0.225**/0.158*

FN

0.083/-0.234

0.242**/0.207*

Total Hip

0.095/-0.170

0.279**/0.233*

PINP

0.126/0.347**

PTH

0.042/0.071

VitD

-0.229**/-0.388** 0.003/-0.094

CTX

0.234**/0.328*

-0.157*/-0.038

BGP

0.181*/0.177

0.034/0.140

ALP

0.249**/0.531**

-0.02/0.129

E2

NU -0.025/0.067

MA

0.029/-0.321*

0.371**/0.422**

-0.122/0.113

0.065/-0.042

-0.116/0.138

0.061/-0.060

-0.176/-0.104

TE

FSH

D

CR

SC R

TGF-β1

CE P

TGF-β1=transforming growth factor β 1; SOST=sclerostin; L1-4=BMD of lumbar spine 1-4; FN=BMD of femur neck; Total Hip=BMD of total hip; P1NP=N-terminal propeptide of type I collagen; PTH=parathyroid

AC

hormone; VitD=serum 25(OH)D; CTX=cross-linked C-telopeptide of type I collagen; BGP=osteocalcin; ALP=alkaline phosphatase; FSH=follicle stimulating hormone; E2=estradiol; CR=creatinine; *p<0.05, **p<0.01.

28

ACCEPTED MANUSCRIPT Table 3. Correlation coefficient of TGF-β1 and SOST with bone density, bone turnover markers, calciotropic hormones and sex hormones in the 3 different groups Early postmenopausal women

TGF-β1

SOST

TGF-β1

SOST

TGF-β1

1

0.047.815

1

-0.274.026

Age

-0.201.314

0.452.018

-0.399.001

0.077.537

Height

0.038.849

0.221.267

-0.030.810

Weight

0.147.464

0.312.113

0.125.318

BMI

0.128.525

0.234.240

0.002.987

L1-4

-0.176.381

0.229.250

FN

-0.076.708

0.301.127

Total Hip

-0.110.584

0.331.091

PINP

0.175.384

0.177.376

PTH

0.092.647

0.255.199

VitD

-0.127.527

CTX

0.020.922

BGP

0.417.030

ALP Ca

Late postmenopausal women

T

Premenopausal women

IP

TGF-β1

SOST 0.156.164

0.015.894

0.136.225

0.249.044

-0.033.770

-0.100.375

0.185.137

0.444.000

-0.151.178

0.217.079

0.492.000

-0.135.231

-0.004.977

0.068.547

0.206.064

0.423.000

0.011.930

-0.083.462

-0.127.258

0.446.000

0.079.527

-0.033.771

0.018.874

0.406.001

0.275.025

-0.175.161

0.195.081

-0.077.494

0.006.961

0.231.062

-0.165.142

0.022.847

-0.152.449

-0.154.218

-0.037.768

-0.122.279

-0.039.731

0.281.155

0.033.792

0.056.654

0.188.092

-0.107.340

-0.123.542

-0.073.562

0.242.051

0.247.026

0.106.348

0.405.036

-0.125.535

0.049.703

0.077.547

0.326.004

-0.276.017

-0.213.286

0.172.390

-0.087.493

0.123.332

0.094.422

-0.176.131

0.199.320

-0.232.244

0.258.040

0.035.784

0.225.052

-0.306.008

NU

MA D

TE

CR

AC

CE P

SC R

1

0.017.951

0.260.350

-0.040.874

0.069.785

0.496.003

0.179.319

FSH

-0.204.361

0.172.444

-0.041.762

-0.055.684

0.179.141

0.244.044

E2

-0.011.961

-0.148.512

-0.019.892

0.021.880

0.202.089

-0.146.222

P

TGF-β1=transforming growth factor β 1; SOST=sclerostin; L1-4=BMD of lumbar spine 1-4; FN=BMD of femur neck; Total Hip=BMD of total hip; P1NP=N-terminal propeptide of type I collagen; PTH=parathyroid hormone; VitD=serum 25(OH)D; CTX=cross-linked C-telopeptide of type I collagen; BGP=osteocalcin; ALP=alkaline phosphatase; FSH=follicle stimulating hormone; E2=estradiol; CR=creatinin

29

ACCEPTED MANUSCRIPT

B

SE

β

t

sig

4.459

1.308

0.525

3.410

0.002

BGP

-8.569

2.667

-0.495

T

0.004

TGF-β1

-0.213

0.076

-0.362

-2.802

0.007

Late postmenopausal

P

-89.932

28.198

-0.342

-3.189

0.002

women

ALP

-0.284

0.094

SC R

Table 4. Multiple regression analysis of serum SOST in the 3 different groups

-0.319

-3.032

0.003

L1-4

57.250

24.454

0.252

2.341

0.022

Premenopausal women Age

Early postmenopausal

NU

women

-3.213

IP

Predictor

MA

TGF-1=transforming growth factor β 1; SOST=sclerostin; BGP=osteocalcin; P=serum phosphorus;

AC

CE P

TE

D

ALP=alkaline phosphatase; L1-4=BMD of lumbar spine 1-4;

30

ACCEPTED MANUSCRIPT Highlights Serum TGF-beta1 level increases during early postmenopause and declines in old age.

T

Serum TGF-beta1 level was significantly related to level of bone turnover markers.

IP

Serum sclerostin level was positively correlated with BMD and negatively with serum PTH.

SC R

Serum sclerostin level was positively associated with age in premenopausal women.

AC

CE P

TE

D

MA

NU

Sclerostin production is inhibited by TGF-beta1 during early postmenopause.

31