Osteoporosis International

, Volume 22, Issue 1, pp 327–337

Green tea polyphenols attenuate deterioration of bone microarchitecture in female rats with systemic chronic inflammation

Authors

    • Department of PathologyTexas Tech University Health Sciences Center
    • Laura W. Bush Institute for Women’s HealthTexas Tech University Health Sciences Center
  • J. K. Yeh
    • Applied Bench Core LaboratoryWinthrop-University Hospital
  • C. Samathanam
    • Department of PathologyTexas Tech University Health Sciences Center
  • J. J. Cao
    • USDA ARS Grand Forks Human Nutrition Research Center
  • B. J. Stoecker
    • Nutritional SciencesOklahoma State University
  • R. Y. Dagda
    • Department of PathologyTexas Tech University Health Sciences Center
  • M.-C. Chyu
    • Department of PathologyTexas Tech University Health Sciences Center
    • Department of Mechanical EngineeringTexas Tech University
    • Graduate Healthcare Engineering OptionTexas Tech University
  • D. M. Dunn
    • Department of PathologyTexas Tech University Health Sciences Center
  • J.-S. Wang
    • Department of Environmental Health ScienceUniversity of Georgia, Athens
Original Article

DOI: 10.1007/s00198-010-1209-2

Cite this article as:
Shen, C., Yeh, J.K., Samathanam, C. et al. Osteoporos Int (2011) 22: 327. doi:10.1007/s00198-010-1209-2

Abstract

Summary

Green tea polyphenols (GTP) are promising agents for preventing bone loss. GTP supplementation sustained microarchitecture and improved bone quality via a decrease in inflammation. Findings suggest a significant role for GTP in skeletal health of patients with chronic inflammation.

Introduction

This study evaluated whether GTP can restore bone microstructure along with a molecular mechanism in rats with chronic inflammation. A 2 [placebo vs. lipopolysaccharide (LPS)]× 2 [no GTP vs. 0.5% GTP (w/v) in drinking water] factorial design was employed.

Methods

Female rats were assigned to four groups: placebo, LPS, placebo + GTP, and LPS + GTP for 12 weeks. Efficacy was evaluated by examining changes in bone microarchitecture using histomorphometric and microcomputed tomographic analyses and by bone strength using the three-point bending test. A possible mechanism was studied by assessing the difference in tumor necrosis factor-α (TNF-α) expression in tibia using immunohistochemistry.

Results

LPS lowered trabecular volume fraction, thickness, and bone formation in proximal tibia while increasing osteoclast number and surface perimeter in proximal tibia and eroded surface in endocortical tibial shafts. GTP increased trabecular volume fraction and number in both femur and tibia and periosteal bone formation rate in tibial shafts while decreasing trabecular separation in proximal tibia and eroded surface in endocortical tibial shafts. There was an interaction between LPS and GTP in trabecular number, separation, bone formation, and osteoclast number in proximal tibia, and trabecular thickness and number in femur. GTP improved the strength of femur, while suppressing TNF-α expression in tibia.

Conclusion

In conclusion, GTP supplementation mitigated deterioration of bone microarchitecture and improved bone integrity in rats with chronic inflammation by suppressing bone erosion and modulating cancellous and endocortical bone compartments, resulting in a larger net bone volume. Such a protective role of GTP may be due to a suppression of TNF-α.

Keywords

Bone qualityDietary supplementHistomorphometryInflammationMicro-CTTea

Introduction

Low bone mass, also called osteopenia, has been reported in patients with different types of chronic inflammation (i.e., chronic periodontitis [1] and pancreatitis [2], inflammatory bowel disease [3], rheumatoid arthritis [4], and lupus erythematosus [5]). The pathogenesis of osteopenia in these patients involves oxidative stress [6] and overproduction of pro-inflammatory cytokine mediators (i.e., tumor necrosis factor-α (TNF-α), cyclooxygenase-2 (COX-2), and interleukin-1β [4, 7, 8]), glucocorticoid treatment [9], and decreased muscular function, resulting in decreased bone formation, increased bone resorption, increased risk for falls, and therefore, increased risk for bone fracture [10].

Oxidative stress has been associated with the progression of bone loss [1114]. Oxidative stress resulting from excessive production of reactive oxygen species perturbs the normal redox balance of osteogenesis including bone formation and resorption. Recently, Shen et al. [15] demonstrated that oxidative stress, as shown by an increase in urinary excretion of 8-hydroxydeoxyguanosine (an oxidative stress biomarker), is involved in the pathogenesis of bone loss in female rats due to chronic inflammation. Oxidative stress results in (1) an increase in osteoblast and osteocyte apoptosis [11], (2) a decrease in osteoblast number via extracellular signal-regulated kinases (ERK) and ERK-dependent nuclear factor-κB signaling pathways [12], (3) a decrease in the rate of bone formation via Wnt/β-catenin signaling [13], and (4) an increase in the differentiation and function of osteoclasts [14].

Green tea, a popular beverage consumed worldwide, has received considerable attention because of its many scientifically supported beneficial effects on human health, including maintaining bone mass [16]. The most widely recognized properties of green tea are their antioxidant activities, arising from their ability to scavenge reactive oxygen species [17]. Our recent study demonstrated a protective action of green tea polyphenols (GTP, green tea extract) against chronic inflammation-induced bone loss in female rats through reducing oxidative stress [15]. However, no data are available showing an impact of GTP on bone quality and microachitecture of rats with chronic inflammation. Therefore, the present study was designed to investigate the effect of GTP on bone microarchitecture and quality in rats with chronic inflammation to further elucidate the benefits of GTP in skeletal health and prevention of pathological bone loss during chronic inflammation. Based on the reported protective effect of GTP supplementation on bone microarchitecture in ovariectomized middle-aged female rats [18], we hypothesized that GTP may have anabolic properties sustaining bone microstructure in the female animals with chronic inflammation. In addressing the study objectives, bone mineral density (BMD), bone histomorphometry, microcomputed tomography (μCT), and bone quality were analyzed. Furthermore, the possible mechanistic profile for how GTP supplementation may attenuate bone microstructure was investigated by evaluating the expression of the pro-inflammatory cytokine mediator, TNF-α, in tibia of rats.

Materials and methods

Animals and GTP treatments

Forty 3-month-old virgin Sprague Dawley (SD) female rats (Charles River, Wilmington, MA, USA) were allowed to acclimate for 5 days to a rodent chow diet and distilled water ad libitum. After acclimation, rats were randomized by weight and assigned to placebo implantation (P), lipopolysaccharide (LPS) administration (L), P + 0.5% GTP (PG), or LPS + 0.5% GTP (LG) for 12 weeks. This 2 (placebo vs. LPS administration) × 2 (no GTP vs. 0.5% GTP in drinking water) factorial design allowed to evaluate effects of LPS administration, GTP levels, and LPS × GTP interaction.

Twenty rats in the LPS-administered groups were subjected to the modified procedures of Smith et al. [19]: LPS (Escherichia coli Serotype 0127:B8, Sigma, St Louis, MO, USA) was incorporated into time-release pellets (Innovative Research of America, Sarasota, FL, USA), designed to deliver a consistent dose for 12 weeks. For LPS animals, the dorsal neck area was shaved and sterile techniques were utilized. A small incision equal in diameter to that of the pellet (2.25 mm) was made at the back of the neck and a horizontal pocket for LPS pellet (33.3 μg/day) implantation (approximately 2 cm beyond the incision site) was formed using forceps. The incision site was closed with surgical glue. Rats had free access to drinking water with no GTP (L group) or 0.5% GTP (LG group) throughout the 12-week study period. The remaining 20 rats in the placebo-administered group received a pellet-containing matrix only using the same procedures for administration described above. The placebo rats also had free access to no GTP (P group) or 0.5% GTP drinking water (PG group) throughout the study period. The 0.5% concentration of GTP in drinking water daily mimics human consumption of green tea of 4 cups a day based on our previous human [20] and animal studies [15, 21]. All rats were fed a rodent chow diet ad libitum during the 12-week feeding period.

Distilled water mixed with GTP was prepared fresh daily, and the amount of water consumed was recorded for each rat. GTP was purchased from the same source as that used in our previous studies (Shili Natural Product Company, Inc., Guangxi, China), with a purity higher than 98.5%. Every 1,000 mg of GTP contained 464 mg of (-)-epigallocatechin gallate (EGCG), 112 mg of (-)-epicatechin gallate, 100 mg of (-)-epicatechin, 78 mg of (-)-epigallocatechin, 96 mg of (-)-gallocatechin gallate, and 44 mg of catechin according to the high performance liquid chromatography-electrochemical detector (HPLC-ECD) and high performance liquid chromatography-ultraviolet detector (HPLC-UV) analyses. Rats were housed in individual stainless steel cages under a controlled temperature of 21 ± 2°C with a 12-h light–dark cycle. Rats were weighed weekly and examined daily. All procedures were approved by the local Institutional Animal Care and Use Committee.

Sample preparation

Each animal was given an intraperitoneal injection of calcein green (10 mg/kg of body weight; Sigma Co., St. Louis, MO, USA) at 14 and 4 days before euthanasia and anesthetized at the end of the study with sodium pentobarbital (50 mg/kg, i.p.). The final body weight was recorded. After animals were anesthetized and euthanized, femora and tibiae were harvested and cleaned of adhering soft tissues. The right tibia samples were kept in 70% ethanol and then processed for histomorphometric assays. The left tibia samples were kept in 10% formalin and then processed for immunohistochemistry. Left femur samples were kept in phosphate-buffered saline (PBS) solution at 4°C for bone scans and the bone strength test.

Bone mass assessment

Left femur bone mass was determined by dual-energy X-ray absorptiometry (DEXA) (HOLOGIC QDR-2000 plus DXA, Waltham, MA, USA) [21]. The machine was set at an ultra-high resolution mode with line spacing of 0.0254 cm, resolution of 0.0127 cm, and a collimator diameter of 0.9 cm.

Histomorphometric analysis

Preparation of right tibiae for the static and dynamic bone histomorphometric analysis was described previously [18]. Undecalficied frontal sections of proximal tibia were embedded in methylmethacrylate (Eastman Organic Chemicals, Rochester, NY, USA) and cut (5 µm thickness) using a microtome (Leica RM 2155, Germany) for metaphyseal bone histomorphometric analysis. The adjacent section was stained with Goldner’s Trichrome method for osteoclast surface and osteoclast cell number measurement [22].

Undecalcified tibial shaft was embedded and a cross-section of the proximal of the tibiofibular junction (8 µm thickness) was cut using microtome for cortical bone histomorphometric analysis. All sections were coverslipped with Eukitt (Calibrated Instruments, Hawthorne, NY, USA) for static and dynamic histomorphometric analysis using a semiautomatic image analysis system (Osteomeasure Histomorphometry System, Osteometrics, Atlanta, GA, USA).

A digitizing morphometric system was used to measure bone histomorphometric parameters. The system consisted of an epifluorescence microscope (Nikon E-400, OsteoMetrics, Atlanta, GA, USA), an Osteomeasure High Resolution Color Subsystem (OsteoMetrics) coupled to an IBM computer, and a morphometry program (OsteoMetrics). The measured parameters for cancellous bone included total tissue volume (TV), bone volume (BV), bone surface (BS), single- and double-labeled surfaces, interlabel width, osteoclast surface, osteoclast cell number, and eroded surface. These data were used to calculate standard morphometric parameters analyzed in bone studies, including percent cancellous bone volume (BV/TV, percent), trabecular thickness (Tb.Th, micrometer), trabecular number (Tb.N, number per millimeter), trabecular separation (Tb.Sp, micrometer), and trabecular bone formation rate (BFR/BS, cubic micrometer per square micrometer per day), osteoclast surface per bone surface (OcS/BS, percent), and osteoclast cell number per bone surface (N.Oc/BS, number per millimeter) according to the standard nomenclature recommended by the American Society for Bone and Mineral Research Nomenclature Committee [23]. The region of bone measured in all groups was 1 to 4 mm from the growth plate in the proximal tibia.

Measurements in cortical bone included periosteal mineral total bone area, periosteal perimeter, marrow area, endocortical perimeter, periosteal and endocortical single- and double-labeled perimeters, interlabeled widths, and endocortical eroded surface. These measures were then used to calculate percent cortical bone area (Ct.Ar, percent), percent marrow area (Ma.Ar, percent), percent periosteal mineralized surface/bone surface (Ps-MS/BS, percent), periosteal mineral apposition rate (Ps-MAR, micrometer per day), periosteal bone formation rate (Ps-BFR/BS, cubic micrometer per square micrometer per day), percent endocortical mineralized surface/bone surface (Ec-MS/BS, percent), endocortical mineral apposition rate (Ec-MAR, micrometer per day), endocortical bone formation rate (Ec-BFR/BS, cubic micrometer per square micrometer per day), and endocortical eroded surface/bone surface (Ec-ES/BS, percent) [18].

Bone microarchitecture assessment by microcomputed tomography

Bone microarchitecture in femur was assessed using microcomputed tomography (μCT) (MicroCT40, SCANCO Medical, Switzerland) following the procedure of Shen et al. [18]. Trabecular bone of the femur was scanned so that 250 images were acquired. The volume of interest comprised the secondary spongiosa in 1,000 cross-sectional slices of the distal femur beginning 25 slices from the growth plate region. All scans were performed in a 1,024 × 1,024 matrix resulting in an isotropic voxel resolution of 16 μm3. An integration time of 150 ms per projection was used. Trabecular parameters in femur included trabecular bone volume fraction (BV/TV, percent), number (Tb.N, number per millimeter), thickness (Tb.Th, micrometer), and separation (Tb.Sp, micrometer). Coefficients of variation were 2.0% (BV/TV), 1.1% (Tb.N), 0.66% (Tb.Th), and 1.30% (Tb.Sp) for morphometric parameters.

Bone quality assessment

Femoral quality was evaluated by a three-point bending test using a custom-designed and built apparatus according to the procedures of Nielsen [24]. Descriptions of the terms used for the assessment of bone strength have been described previously [25]. Maximum force (newton) and yield point force (newton) to break bones and modulus of elasticity were assessed.

Expression of TNF-α in proximal tibia

Seven-micron tissue sections were cut from decalcified (Immunocal, American Master*Tech Scientific, Lodi, CA, USA), formalin-fixed, paraffin-embedded blocks to Superfrost/Plus slide (Fisher Scientific, Fair Lwan, NJ, USA); deparaffinized in xylene; rinsed in 100%, 95%, and 70% ethanol; and rehydrated in distilled water. Tissue sections were treated with peroxidase blocking reagent (DAKO, Carpenteria, CA, USA); blocked with normal serum (Vector Laboratories, Burlingame, CA, USA) at room temperature; incubated with primary antibody TNF-α (AbD Serotec., Raleigh, NC, USA) at dilution of 1:100; washed with PBS; treated with biotin-labeled secondary antibody (Vector Laboratories) treated with ABC solution for 30 min followed by incubation with NovaRd™ (Vector Laboratories). Counterstaining was performed with Immuno*Master Hematoxylin (Zymed Laboratories Inc., South San Francisco, CA, USA). Expression of TNF-α was confirmed by comparing control tissue section performed following the same procedures with the omission of the primary antibody. All slides were evaluated by the study pathologist for intensity in a blinded manner according to the following scoring system: normal (0), low (1), medium (2), and high (3).

Statistical analysis

Data are expressed as mean ± standard error of the mean (SEM). All data were analyzed using SigmaStat, version 2.03 (Systat Software, Inc., San Jose, CA, USA). Normality of distribution and homogeneity of variance were tested. Data of body weight were analyzed by three-way analysis of variance (ANOVA) (LPS administration × GTP levels × time). Data of bone mass, microstructure, dynamics, osteoclast parameters, quality, and TNF-α expression intensity were analyzed by two-way ANOVA to evaluate the effect of LPS administration, GTP levels, or interaction. Significant interactions between LPS and GTP were tested using Fisher’s least significant difference (LSD) tests to further define treatment effects (the P, PG, L, and LG groups). The level of significance was set at P < 0.05 for all statistical tests, and statistical trends (P < 0.10) were also indicated.

Results

Body weight

There was no significant difference in initial body weight among all treatment groups (P group, 229.0 ± 23.0 g; PG, 249.0 ± 5.2 g; L group, 249.9 ± 5.6 g, LG group, 249.9 ± 5.8 g). Over the course of the 12-week study, all animals gained body weight and neither LPS administration nor GTP supplementation significantly affected body weights (P group, 327.4 ± 9.7 g; PG, 306.5 ± 8.4 g; L group, 315.6 ± 5.9 g, LG group, 314.8 ± 6.6 g).

Bone mass

Neither LPS administration nor GTP levels significantly affected femoral bone area after 12 weeks (Table 1). Based on the results of two-way ANOVA, after 12 weeks of treatment, LPS administration resulted in a decrease in the values for femur bone mineral content (BMC) and bone mineral density (BMD) of rats, while GTP supplementation led to an increase in the values for both parameters. There was no interaction between LPS administration and GTP supplementation affecting BMC or BMD.
Table 1

Bone measurement in trabecular bone of femur and proximal tibia in placebo- and lipopolysaccharide-administered female rats supplemented with green tea polyphenols (GTP) in drinking water for 12 weeks

Parameters

-LPS

+LPS

Two-way ANOVA

P value

No GTP (P group)

0.5% GTP (PG group)

No GTP (L group)

0.5% GTP (LG group)

LPS

GTP

LPS × GTP

DEXA analysis of femur

 Area (cm2)

1.938 ± 0.031

1.931 ± 0.036

1.916 ± 0.035

1.899 ± 0.029

0.804

0.536

0.818

 BMC (mg)

505 ± 13xb

533 ± 9xa

488 ± 11yb

506 ± 6ya

0.044

0.034

0.596

 BMD (mg/cm2)

265.6 ± 2.9xb

270.0 ± 2.2xa

258.6 ± 2.7yb

265.7 ± 2.2ya

0.039

0.036

0.606

Histomorphometric analysis of proximal tibia

 BV/TV (%)

23.82 ± 0.90xbA

24.14 ± 1.22xaA

17.54 ± 1.21ybB

23.82 ± 1.08yaA

0.006

0.006

0.012

 Tb.Th (μm)

57.05 ± 1.23x

58.01 ± 1.38x

51.97 ± 1.34y

53.83 ± 1.14y

0.001

0.289

0.731

 Tb.N (n/mm)

4.17 ± 0.13bA

4.16 ± 0.17aA

3.37 ± 0.22bB

4.42 ± 0.16aA

0.133

0.007

0.005

 Tb.Sp (μm)

265 ± 11yaB

267 ± 14ybB

368 ± 30xaA

251 ± 14xbB

0.034

0.006

0.005

 BFR/BS (μm3/μm2/day)

11.97 ± 1.23xA

13.98 ± 1.69xA

12.15 ± 1.23yA

8.01 ± 1.14yB

0.043

0.445

0.032

Results are expressed as mean values ± SEM. Treatments were analyzed by two-way analysis of variance (ANOVA) to evaluate the effect of LPS administration, GTP supplementation, or interaction. Significant interactions between LPS and GTP were tested using Fisher’s LSD tests to further define treatment effects (the P, PG, L, and LG groups). The level of significance was set at P < 0.05 for all statistical tests

Within a row having different superscripts (x and y for LPS effect; a and b for GTP effect; capital letters for interaction effect) are significantly different by two-way ANOVA and Fisher’s LSD test (P < 0.05)

BMC, bone mineral content; BMD, bone mineral density; BV/TV, percent trabecular bone volume; Tb.Th, trabecular thickness; Tb.N, trabecular number; Tb.Sp, trabecular separation; and BFR/BS, trabecular bone formation rate

Histomorphometric changes in proximal tibia

The effects of LPS administration or GTP supplementation on histomorphometric changes in proximal tibia are described in Table 1. The results of two-way ANOVA show that LPS administration resulted significantly in lower values for BV/TV, Tb.Th, and BFR/BS, higher values for Tb.Sp, while no significant effect on Tb.N in proximal tibia. In contrast, supplementation of GTP in the drinking water significantly increased BV/TV and Tb.N, decreased Tb.Sp, but had no effect on Tb.Th and BFR/BS. Significant interactions between LPS administration and GTP supplementation were observed in BV/TV, Tb.N, Tb.Sp, and BFR/BS.

Among the significant interaction, we further performed post hoc test to show that rats in the L group had the lowest value for BV/TV and Tb.N and the highest value for Tb.Sp than those in other groups. On the other hand, the rats in the LG group had the lowest value for BFR/BS compared to others.

Alteration in dynamic parameters in cortical bone of tibia shaft

After 12 weeks, neither LPS administration nor GTP supplementation significantly affected total area (T.Ar), percent cortical bone area (Ct.Ar), and percent marrow area (Ma.Ar) in tibial shaft (Table 2). Based on the results of two-way ANOVA, after 12 weeks of treatment, (1) LPS administration significantly increased percent endocortical mineralized surface/bone surface (Ec-MS/BS), endocortical mineral apposition rate (Ec-MAR), and eroded surface/bone surface (Ec-ES/BS); and (2) GTP supplementation significantly increased percent periosteal mineralized surface/bone surface (Ps-MS/BS), while significantly suppressing Ec-ES/BS. There was no significant interaction between LPS and GTP in all parameters (Table 2).
Table 2

Bone histomorphometric measurement in cortical bone of tibia shaft in placebo- and lipopolysaccharide-administered female rats supplemented with GTP in drinking water for 12 weeks

Parameters

-LPS

+LPS

Two-way ANOVA

P value

No GTP (P group)

0.5% GTP (PG group)

No GTP (L group)

0.5% GTP (LG group)

LPS

GTP

LPS × GTP

T.Ar (mm2)

5.33 ± 0.08

5.30 ± 0.10

5.33 ± 0.07

5.22 ± 0.05

0.635

0.436

0.666

Ct.Ar (%)

85.22 ± 0.58

84.33 ± 0.50

84.04 ± 0.55

84.87 ± 0.66

0.583

0.956

0.145

Ma.Ar (%)

14.78 ± 0.58

15.67 ± 0.50

15.96 ± 0.55

15.13 ± 0.67

0.583

0.956

0.145

Ps-MS/BS (%)

18.20 ± 1.09b

19.13 ± 1.29a

14.39 ± 1.33b

19.66 ± 1.93a

0.250

0.034

0.131

Ps-MAR (μm/day)

1.039 ± 0.045

1.046 ± 0.074

0.886 ± 0.079

0.959 ± 0.063

0.087

0.562

0.632

Ps-BFR/BS (μm3/μm2/day)

19.05 ± 1.66

20.14 ± 2.06

12.82 ± 1.64

18.95 ± 2.27

0.060

0.067

0.195

Ec-MS/BS (%)

1.68 ± 0.67y

1.94 ± 0.82y

5.37 ± 1.01x

2.51 ± 0.98x

0.041

0.240

0.149

Ec-MAR (μm/day)

0.140 ± 0.000y

0.140 ± 0.000y

0.259 ± 0.082x

0.214 ± 0.048x

0.047

0.632

0.632

Ec-BFR/BS (μm3/μm2/day)

0.352 ± 0.140

0.277 ± 0.118

1.591 ± 0.795

0.797 ± 0.461

0.072

0.366

0.454

Ec-ES/BS (%)

7.82 ± 0.88ya

6.58 ± 0.51yb

14.62 ± 0.97xa

11.27 ± 0.88xb

<0.001

0.009

0.216

Results are expressed as mean values ± SEM. Treatments were analyzed by two-way ANOVA to evaluate the effects of LPS administration, GTP supplementation, or interaction. Significant interactions between LPS and GTP were tested using Fisher’s LSD tests to further define treatment effects (the P, PG, L, and LG groups). The level of significance was set at P < 0.05 for all statistical tests

Within a row having different superscripts (x and y for LPS effect; a and b for GTP effect; capital letters for interaction effect) are significantly different by two-way ANOVA and Fisher’s LSD test (P < 0.05)

T.Ar total area, Ct.Ar percent cortical bone area (cortical area/total area), Ma.Ar percent marrow area, Ps-MS/BS percent periosteal mineralized surface/bone surface, Ps-MAR periosteal mineral apposition rate, Ps-BFR/BS periosteal bone formation rate, Ec-MS/BS percent endocortical mineralized surface/bone surface, Ec-MAR endocortical mineral apposition rate, Ec-BFR/BS endocortical bone formation rate, Ec-ES/BS endocortical eroded surface/bone surface

Changes in osteoclast number

The effect of LPS administration and GTP supplementation on osteoclast number in proximal tibia is shown in Fig. 1. Administration of LPS to the rats significantly increased osteoclast parameters, in terms of N.Oc/BS (P < 0.001; Fig. 1a) and OcS/BS (P < 0.001; Fig. 1b), while supplementation of GTP in the drinking water suppressed both osteoclast parameters (P < 0.001 for both parameters). A significant interaction between LPS administration and GTP supplementation was found in both parameters (P = 0.009 for N.Oc/BS; P < 0.001 for OcS/BS).
https://static-content.springer.com/image/art%3A10.1007%2Fs00198-010-1209-2/MediaObjects/198_2010_1209_Fig1_HTML.gif
Fig. 1

Osteoclast parameters measured at proximal tibia in placebo- and LPS-administered female rats supplemented with GTP in drinking water (0.5%, w/v) for 12 weeks. Osteoclast cell number per bone surface (N.Oc/BS, n/mm) (a) and osteoclast surface per bone surface (OcS/BS, percent) (b). Values are mean (n = 10) with their SEM represented by vertical bars. Data was evaluated by two-way ANOVA (LPS administration × GTP level). Significant interactions between LPS and GTP were tested using Fisher’s LSD tests to further define treatment effects (the P, PG, L, and LG groups). Having different letters (x and y for LPS effect; a and b for GTP effect; capital letters for interaction effect) are significant different by two-way ANOVA and Fisher’s LSD test, P < 0.05. LPS administration increased osteoclast parameters; whereas GTP supplementation suppressed both parameters. There was a significant interaction between LPS administration and GTP supplementation on both parameters

Microarchitectural parameters of femur

Data for trabecular bone microarchitecture of the femur are exhibited in Table 3. Results of two-way ANOVA show that GTP supplementation, not LPS administration, significantly increased BV/TV, Tb.Th, and Tb.N of femur. There was a significant interaction between LPS and GTP affecting Tb.Th. and Tb.N. Among these significant interactions, the post hoc tests further show that the LG group has the highest Tb.Th, while the L group has the lowest Tb.N among the four groups. Moreover, at the midshaft of the femur, neither LPS administration nor GTP supplementation affected cortical bone volume fraction (BV/TV), thickness or area, or the medullary area (Table 3).
Table 3

Bone microarchitectural properties of femur in placebo- and lipopolysaccharide-administered female rats supplemented with GTP in drinking water for 12 weeks as determined by µCT

Parameters

-LPS

+LPS

Two-way ANOVA

P value

No GTP (P group)

0.5% GTP (PG group)

No GTP (L group)

0.5% GTP (LG group)

LPS

GTP

LPS × GTP

Trabecular bone

 BV/TV (%)

35.80 ± 2.2b

39.17 ± 4.59a

27.30 ± 1.66b

44.79 ± 6.25a

0.716

0.012

0.082

 Tb.Th (mm)

0.081 ± 0.003bAB

0.077 ± 0.005aB

0.070 ± 0.001bB

0.093 ± 0.009aA

0.710

0.029

0.028

 Tb.N (n/mm)

6.12 ± 0.16bA

6.06 ± 0.23aA

5.37 ± 0.20bB

6.36 ± 0.31aA

0.408

0.042

0.045

 Tb.Sp (mm)

0.153 ± 0.008

0.161 ± 0.014

0.169 ± 0.007

0.143 ± 0.011

0.967

0.401

0.117

Cortical bone at midshaft of femur

 BV/TV (%)

95.48 ± 0.08

95.34 ± 0.09

95.43 ± 0.07

95.53 ± 0.04

0.402

0.811

0.158

 Cortical thickness (mm)

0.700 ± 0.013

0.684 ± 0.012

0.692 ± 0.015

0.708 ± 0.007

0.517

0.977

0.231

 Cortical area (mm2)

7.25 ± 0.21

7.08 ± 0.19

7.21 ± 0.14

7.30 ± 0.09

0.290

0.794

0.661

 Medullary area (mm2)

5.32 ± 0.20

5.37 ± 0.15

5.41 ± 0.14

5.21 ± 0.05

0.848

0.619

0.415

Results are expressed as mean values ± SEM. Treatments were analyzed by two-way ANOVA to evaluate the effect of LPS administration, GTP levels, or interaction. Significant interactions between LPS and GTP were tested using Fisher’s LSD tests to further define treatment effects (the P, PG, L, and LG groups)

Within a row having different superscripts (x and y for LPS effect; a and b for GTP effect; capital letters for interaction effect) are significantly different by two-way ANOVA and Fisher’s LSD test (P < 0.05)

BV/TV percent bone volume fraction, Tb.Th trabecular thickness, Tb.N trabecular number, and Tb.Sp trabecular separation

Bone quality

Figure 2 shows the impact of LPS administration or GTP supplementation on bone strength of femur parameters, as determined by three-point bending. The results of two-way ANOVA analysis show that (1) after the 12-week study period LPS administration significantly decreased maximum force (P < 0.001; Fig. 2a) and yield point force (P = 0.012; Fig. 2b), (2) GTP supplementation significantly improved these parameters (P = 0.003 for maximum force; P = 0.001 for yield point force), and (3) an interaction between LPS administration and GTP levels was observed in yield point force (P = 0.018), but not in maximum force (P = 0.089).
https://static-content.springer.com/image/art%3A10.1007%2Fs00198-010-1209-2/MediaObjects/198_2010_1209_Fig2_HTML.gif
Fig. 2

Maximum force (a) and yield point force (b) of femur in placebo- and LPS-administered female rats supplemented with GTP in drinking water (0.5%, w/v) for 12 weeks. Values are mean (n = 10) with their SEM represented by vertical bars. Data was evaluated by two-way ANOVA (LPS administration × GTP level). Significant interactions between LPS and GTP were tested using Fisher’s LSD tests to further define treatment effects (the P, PG, L, and LG groups). Having different letters (x and y for LPS effect; a and b for GTP effect; capital letters for interaction effect) are significantly different by two-way ANOVA and Fisher’s LSD test, P < 0.05. LPS administration decreased maximum force and yield point force; whereas GTP supplementation improved both parameters. There was a significant interaction between LPS administration and GTP supplementation in yield point force

Expression of TNF-α in proximal tibia

The expression intensity of TNF-α in proximal tibia is presented in Fig. 3. Based on two-way ANOVA analysis, after 12 weeks intervention, (1) LPS administration significantly induced the expression of TNF-α in proximal tibia with the greatest extent within chondrocytes in the growth plate region; (2) GTP supplementation significantly suppressed that of TNF-α in proximal tibia (P = 0.024); and (3) there was a significant interaction between LPS administration and GTP supplementation. Figure 4 demonstrates that the L group has the strongest expression intensity of TNF-α than the other groups.
https://static-content.springer.com/image/art%3A10.1007%2Fs00198-010-1209-2/MediaObjects/198_2010_1209_Fig3_HTML.gif
Fig. 3

TNF-α in the tibia of placebo- and LPS-administered female rats supplemented with GTP in drinking water (0.5%, w/v) for 12 weeks. Values are mean (n = 10) with their SEM represented by vertical bars. Data was evaluated by two-way ANOVA (LPS administration × GTP level). Significant interactions between LPS and GTP were tested using Fisher’s LSD tests to further define treatment effects (the P, PG, L, and LG groups). Having different letters (x and y for LPS effect; a and b for GTP effect; capital letters for interaction effect) are significantly different by two-way ANOVA and Fisher’s LSD test, P < 0.05. LPS administration increased TNF-α expression intensity; whereas GTP supplementation decreased TNF-α expression intensity. There was a significant interaction between LPS administration and GTP supplementation in TNF-α expression intensity

https://static-content.springer.com/image/art%3A10.1007%2Fs00198-010-1209-2/MediaObjects/198_2010_1209_Fig4_HTML.gif
Fig. 4

Immunohistochemical staining for TNF-α in the tibia of placebo- and LPS-administered female rats supplemented with GTP in drinking water (0.5%, w/v) for 12 weeks. Section obtained from the proximal tibia metaphysic area (×40) showing the P group (a), the PG group (b), the L group (c), and the LG group (d)

Discussion

In this study, a model of LPS administration to female rats was successfully utilized to investigate the impact of GTP supplementation in drinking water in chronic inflammation-induced deterioration of bone microarchitecture. The results of histomorphometric analyses show that LPS administration lowered trabecular volume and thickness, but increased trabecular separation in proximal tibia. Such findings demonstrate that chronic inflammation produced a detrimental effect on bone microarchitecture, a result consistent with a previous study where bone microarchitecture was measured by µCT [19].

The present study demonstrated potent effects of GTP supplementation in preserving bone mass and microarchitecture in female rats during chronic inflammation. Supplementation of GTP in drinking water attenuates chronic inflammation (LPS)-induced decrease in trabecular bone volume and number in proximal tibia, with lower values for trabecular separation in proximal tibia (Table 1). These changes in bone microarchitecture of GTP-supplemented rats may be mediated primarily through a suppression of bone turnover rate, as shown by lower osteoclastic activity in proximal tibia (Fig. 1a, b) as well as by lower endocortical eroded surface at midshaft tibia (Table 2), resulting in a larger net bone mass. Such a suppressive effect on bone resorption by GTP supplementation is observed, particularly significant under the LPS-induced chronic inflammation.

The observation that GTP supplementation primarily influenced bone resorption coincides with a report of Shen et al. [15] that mitigating bone loss in LPS-treated rats by GTP supplementation in drinking water was due to the suppression of bone resorption (as shown in lower serum tartrate-resistant acid phosphate (TRAP)), but not due to bone formation (as no change shown in serum osteocalcin, P > 0.05). Therefore, a net balance of osteocalcin and TRAP levels results in a higher ratio of bone formation to resorption in GTP-supplemented groups to benefit bone remodeling [15].

The capacity of GTP to decrease indices of bone resorption in placebo- and LPS-treated rats is in agreement with the findings by Shen et al. [18] that GTP inhibited bone resorption at cancellous and endocortical bone compartments in both intact and ovariectomized middle-aged female rats. The inhibitory impact of GTP in osteoclastic activity in the present study is supported by several previous studies, as described in a recently published comprehensive review [16]. Cellular studies demonstrated that (1) EGCG, active component in GTP, stimulated osteoclastic cell death via Fenton reaction [26] and caspase activation [27]; (2) EGCG depressed bone resorption either via inhibiting interleukin-6 production or via suppressing of p44/p42 mitogen-activated protein kinase activation in osteoblastic-like MC3T3-E1 cells [28]; and (3) EGCG inhibited osteoclast formation by inhibiting the expression of matrix metalloproteinase-9 in osteoblasts [29]. Furthermore, studies also suggest that green tea bioactive components may modulate osteoimmunological activity (a) to inhibit differentiation of osteoclasts through the receptor activator of nuclear factor-κB ligand signaling pathway [30] or c-Jun N-terminal kinase (JNK)/c-Jun signaling pathways [31], and (b) to suppress the production of cytokines by immune cells [32].

In addition to GTP’s impact in sustaining bone mass and microarchitecture, our data also demonstrated that GTP supplementation has a beneficial effect on bone strength in rats with continuously chronic inflammation. Continuous administration of LPS reduced bone strength characterized by a decrease in maximum force to break the femoral bone of rats (Fig. 2a). Such a detrimental effect of LPS on bone biomechanics in female rats has been reported previously [19]. Our data show that both maximum force and yield point force in the LG group were significantly higher than those in the L group. These results confirm our hypothesis that LPS-induced deterioration in bone mechanical properties in female rats can be mitigated by GTP supplementation. However, in the present study, bone strength was measured only for cortical bone, not for femoral distal metaphysic (rich in cancellous bone).

Previous studies suggest that increased reactive oxidative species (ROS) production may exacerbate the chronic inflammation-induced bone deterioration by elevating the production of pro-inflammatory cytokine mediators, such as TNF-α or COX-2. In this study, immunohistochemistry was utilized to evaluate alterations in expression of TNF-α, a local pro-inflammatory mediator of bone metabolism in proximal tibia of rats with chronic inflammation. TNF-α has been reported to stimulate bone resorption [33] via enhancing osteoclast differentiation and activity and to suppress bone formation by inhibiting osteoblast progenitor cell recruitment and increasing osteoblast apoptosis [34, 35]. Our findings that during the chronic inflammation up-regulation of TNF-α in bone tissue, especially in the area of growth plates along with a low bone mass, agree with results reported by Smith et al. [19].

One of the goals of this study was to explore the molecular mechanism of GTP in attenuating the deterioration of bone microarchitecture in rats during chronic inflammation. GTP has been characterized as an anti-inflammatory agent, suggesting that GTP supplementation in drinking water may have a protective role in bone microstrucure through a reduction of inflammation. In this study, we explored the relationship between GTP supplementation and TNF-α in the tibia in a model of chronic inflammation-induced deterioration of bone microstructure. We showed that GTP supplementation significantly down-regulated protein expression of TNF-α induced by chronic inflammation. Such a bone-protective effect of GTP supplementation on chronic inflammation agrees with other antioxidant such as soy isoflavones [15, 36] using the same model of bone loss. Our current data show that the down-regulation of TNF-α and improvement in bone microarchitecture and quality by GTP supplementation further support the anti-inflammatory role of GTP in bone health which may reduce the risk of bone loss. Further study is therefore needed to investigate the mechanisms regarding the components of the GTP that affect bone metabolism, and especially whether such effect is mediated by the modulation of local bone factors and/or systemic hormones.

Conclusions

This study demonstrates the beneficial effects of GTP supplementation on skeletal remodeling in models of LPS-induced deterioration of bone microarchitecture. In general, supplementation of GTP to LPS-administered female rats for 12 weeks had a beneficial effect on maintaining cancellous and cortical bone compartment as well as improving bone integrity via suppressing bone turnover rate. Such a protective role of GTP may, in part, be attributed to a suppression of the pro-inflammatory cytokine mediator, TNF-α.

Acknowledgements

This study was supported by the Laura W. Bush Institute for Women’s Health and National Institutes of Health/National Center for Complementary and Alternative Medicine grant R21AT003735 (CLS) and the National Institutes of Health/National Cancer Institute grant CA90997 (JSW).

Conflicts of interest

None.

Copyright information

© International Osteoporosis Foundation and National Osteoporosis Foundation 2010