| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Regular Article |
, Prostaglandin E2, and Prostaglandin F2
a Departments of Obstetrics and Gynecology and
b Pathology, Mie University School of Medicine, Tsu, Mie 514-8507, Japan
ABSTRACT
Tenascin-C (TN-C), an extracellular matrix glycoprotein, is known to be expressed in uterine stroma in the peri-implantation period. Examination of the spatiotemporal pattern during early pregnancy using immunohistochemistry and in situ hybridization revealed TN-C expression in the stroma beneath the luminal epithelia of the murine endometrium on Days 0 and 1 of pregnancy, subsequent disappearance, and reappearance on Day 4. After decidualization, tissue around the deciduoma was positive. In situ hybridization demonstrated TN-C production by the stromal cells adjacent to the epithelia. To investigate the regulation of TN-C expression in vitro, murine uterine stromal and epithelial cells were isolated and cultured. Addition of interleukin-1
(IL-1) and prostaglandin E2 (PGE2) and F2 (PGF2) induced TN-C expression in the stromal cells at both protein and mRNA levels, while the sex steroid hormones, progesterone and ß-estradiol, exerted little effect. Immunohistochemistry using anti-IL-1 antibody showed epithelial cells to be positive on Days 2–4 of pregnancy, and addition of progesterone but not ß-estradiol enhanced IL-1 expression in epithelial cells in vitro. In a culture insert system, TN-C expression by stromal cells cocultured with epithelial cells was induced by addition of progesterone alone that was blocked by additions of anti-IL-1 antibody. Collectively, these findings indicate that TN-C expression in the preimplantation period is under the control of progesterone, but not directly, possibly by the paracrine and autocrine intervention of IL-1 secreted by epithelial cells and PGE2 and PGF2 secreted by stromal cells.
decidua, estradiol, implantation, progesterone, uterus
INTRODUCTION
The process of implantation is dependent upon a series of cell-cell and cell-matrix interactions. In the rodent uterus, epithelial cells adopt a receptive state permissive to blastocysts [1, 2], and stromal cells undergo extensive morphological and physiological changes, socalled decidualization, in preparation for the implanting blastocyst [3, 4]. These changes are under the control of ovarian steroid hormones and various substances like cytokines [5–9], growth factors [10], and extracellular matrix [11–13], interacting with each other. Prostaglandin E2 (PGE2) and PGF2 secreted by uterine stromal cells [5] and epithelial cells [14] are also associated with the interactions occurring during implantation [5, 14].
Tenascin (TN), now designated TN-C, is an extracellular matrix glycoprotein that is expressed in association with processes linked to embryogenesis, suggesting an involvement in modulation of epithelial-mesenchymal interactions taking place during tissue development. In the adult, expression is very restricted in normal tissues but may increase under pathological conditions requiring remodeling, like wound healing [15–17] and cancer [18–21]. In the human endometrium, TN-C expression is physiologically changed during the menstrual cycle [19, 22, 23], usually with deposition in the stroma surrounding the endometrial glands in the proliferative, but not the secretory, phase [22, 23]. In the rodent uterus, intense TN-C expression is observed in the stroma beneath the uterine luminal epithelial cells in diestrous, becoming spotty in estrous, and lost in proestrous [24]. Furthermore, it changes with the preimplantation period [24, 25], possibly facilitating embryo penetration by disrupting uterine epithelial cell adhesion to the underlying basal lamina [26]. However, the underlying regulatory mechanisms have yet to be elucidated.
In cultured fibroblast cell lines derived from different organs, it is well known that growth factors and cytokines can induce TN-C expression. For example, this is the case with interleukin-1 (IL-1) or ß, IL-4, and tumor necrosis factor- (TNF-) acting on fetal lung fibroblasts [27]. With fetal leptomeningeal fibroblasts, acidic- and basic-fibroblast growth factor (a- and b-FGF), transforming growth factor-ß (TGF-ß), epidermal growth factors, and bone morphogenetic protein are active [27], and for mouse embryonic fibroblasts, TGF-ß and b-FGF cause induction [28]. When progesterone supplements are given to ovariectomized female rats, TN-C accumulation occurs in the subluminal stroma and the circular smooth muscle, indicating that expression is regulated by the hormonal environment [24]. However, it remains unclear which substances have the greatest effects on TN-C expression in uterine endometrial cells.
In the present studies, TN-C expression was investigated in the murine uterus during early pregnancy in vivo. In addition, substances inducing TN-C expression were identified using cultured endometrial stromal cells, and interactions between epithelial and stromal elements of the endometrium were evaluated by coculture of these cells. It could thereby be demonstrated that the interactions via cytokines and prostaglandins regulate TN-C expression in the preimplantation period.
MATERIALS AND METHODS
Animals
The C3H mice, purchased from SLC Japan Co. (Shizuoka, Japan), were maintained in the animal facility of Mie University School of Medicine. These mice were naturally mated overnight, and the date of pregnancy was determined by physical examination with the day when a vaginal plug was recognized defined as Day 0. They were killed by cervical dislocation and the uteri were collected. All procedures and animal care taken in the study were in accordance with the guidelines approved by the Mie University Animal Experiment and Care Committee.
Tissue Sections and Immunohistochemistry
Murine uteri for TN-C immunohistochemistry were fixed in 4% paraformaldehyde, dehydrated, and embedded in paraffin. Sections cut at 3 µm were placed on glass slides coated with Silan (Dako Japan, Kyoto, Japan) and incubated in 0.4% pepsin/0.01 N HCl solution at 37°C for 15 min, and then in 0.6% H2O2 in methanol for 25 min to block endogenous peroxidase activity. After equilibration in 10 mM PBS and incubation in 10% normal goat serum for 30 min, they were treated with affinity purified rabbit anti-human TN-C polyclonal antibody (1 µg/ml) [29] overnight at 4°C. Following washing in PBS for 10 min, they were treated with goat anti-rabbit peroxidase-labeled antibody (MBL, Nagoya, Japan) for 1 h at room temperature. Color development was achieved with diaminobenzidine 4HCl/H2O2 solution. As a negative control, polyclonal antibody preparations were passed through a TN-C conjugated column prior to use immunohistochemistry. Tissues for IL-1 immunohistochemistry were fixed in methanol-Carnoy solution. After H2O2/methanol treatment, sections were incubated with polyclonal rabbit anti-mouse IL-1 antibody (diluted 1:100, Genzyme, Cambridge, MA). As a negative control, normal rabbit serum at the same dilution was used instead of the first antibody. Three animals or more were evaluated for each of the days of pregnancy.
Preparation of Digoxigenin-Labeled cRNA and In Situ Hybridization
Antisense and sense cRNA probes were prepared by in vitro transcription of mouse TN-C cDNA [30], using a digoxigenin RNA labeling kit (SP6/T7; Boehringer Mannheim, Mannheim, Germany) [31]. TN-C mRNA expression was detected by in situ hybridization using the previously described method of Ishihara et al. [32]. Tissue sections were treated with proteinase K for 10 min and hybridization signals visualized using alkaline phosphatase-conjugated antidigoxigenin antibody and incubation in nitrotetrazolium blue/5-bromo-4-chloro-3-indolyl phosphate solution.
Primary Cultures of Uterine Stroma and Epithelia
Primary cultures of mouse endometrial stromal cells and epithelial cells were prepared from uteri of randomly cycling nonpregnant mice, according to the methods of Jacobs et al. [13, 14]. Each preparation was isolated from more than three mice. The cells were incubated in culture medium consisting of Dulbecco modified Eagle medium/Ham F12, 1:1 (v/v), plus 10% heat-inactivated fetal calf serum (FCS) and 1% antibiotic-antimycotic (Gibco BRL, Gaithersburg, NY). The purity of the cells, checked by immunofluorescent staining of vimentin (Dako Japan) and desmin (Dako Japan) for stromal cells and of pan-cytokeratin (Sigma, St. Louis, MO) for epithelial cells, was over 90% in both cases (data not shown). Cells after two to three passages were used in the experiments. Progesterone (1–100 ng/ml; 3.2 x 10-9 to 3.2 x 10-7 M), ß-estradiol (0.01–1 ng/ml; 3.8 x 10-11 to 3.8 x 10-9 M), b-FGF (100 ng/ml) (Wako Life Science, Osaka, Japan), TGF-ß1 (5 ng/ml) (Boehringer Mannheim), recombinant mouse IL-1 (10–100 ng/ml) (Genzyme), recombinant mouse IL-1ß (200 ng/ml) (UBI, Lake Placid, NY), PGE2 and PGF2 (1–100 ng/ml; 2.8 x 10-9 2.8 x 10-7 M; Calbiochem, San Diego, CA), were added to the medium for induction of TN-C. Treatments were performed after the cells had been incubated in medium plus 0.1% FCS instead of 10% FCS for 48 h, because a high concentration of FCS itself induces TN-C expression. To reconstruct interactions between epithelial cells and stromal cells, the Cell Culture Insert System (Becton Dickinson Labware, Franklin Lakes, NJ), 0.4 µm pore size, was employed. Stromal cells (1 x 106 cells/well) were grown on the bottom of six-well plates. Epithelial cells (1 x 106 cells/well) were separately cultured on inserts, coated with Matrigel (Becton Dickinson Labware), that were placed in the wells of the stromal cells after 2 days to reconstruct cell-cell interactions. All experiments were performed using, at least, three different preparations of the cultured cells.
Immunofluorescence of Cultured Cells
For immunofluorescent staining, the cells were cultured on CultureSlide (Becton and Dickinson) and treated with 1 µM monensin for 5 h before fixation to block secretion and thus cause accumulation of proteins in the cytoplasm. The cells were fixed in 4% paraformaldehyde with 0.25% Triton-X, incubated in 10% normal goat serum for 30 min, and incubated with rabbit antihuman TN-C polyclonal antibody (2.5 µg/ml) overnight at 4°C, washed in PBS, and incubated with goat fluorescein isothiocyanate-labeled antirabbit IgG antibody (MBL) for 2 h at room temperature. Expression of TN-C protein was assessed by fluorescence microscopy.
Quantitative Analysis of TN-C and IL-1 mRNAs Using Reverse Transcription Polymerase Chain Reaction
Total RNA was isolated from cultured stromal cells and epithelial cells using the acid guanidinium-phenol-chloroform method. Total RNA concentrations were determined by spectrophotometry at 260 nm and quality assessed by agarose gel electrophoresis and ethidium bromide staining. Messenger RNA was converted to cDNA utilizing the NotI-d(T)18 as a primer and Moloney muring leukemia virus reverse transcriptase in first-strand cDNA reaction mix buffer (Amersham Pharmacia Biotech, Piscataway, NJ). Complementary DNA was amplified by polymerase chain reaction (PCR) with appropriate sets of primers for TN-C, ß-actin and IL-1 (Table 1). The PCR was initiated by heating to 95°C for 10 min with hot starting DNA polymerase, Amplitaq Gold DNA Polymerase (Perkin-Elmer Biosystems, Foster, CA), and carried out for 25 cycles of 30 sec at 94°C, 45 sec at 55°C, and 30 sec at 72°C. The amplified products were analyzed by electrophoresis in 2% agarose gels, stained with ethidium bromide (0.1 µg/ml), and detected by ultraviolet light. The images were stored on a floppy disk using Documentation System (Pharmacia Biotech) and quantitative analysis was performed with NIH Image (version 1.56). There was a linear increase in the PCR product when varied concentrations of reverse transcriptase product were subjected to the PCR assay (data not shown). The sequences of the PCR products could be confirmed.
|
Effects of Anti-IL-1 Antibody and Indomethacin on TN-C Expression Under the Coculture Conditions
To investigate the pathway of TN-C induction by progesterone treatment with cocultures, we tested effects of anti-IL-1 and indomethacin. With addition of progesterone, a neutralizing antibody against IL-1 (monoclonal hamster, Genzyme) and control IgG was added into the medium at a final concentration of 0.1 mg/ml. Indomethacin was added at 10 µM. The RNAs of stromal cells were isolated 24 h after the treatment.
RESULTS
Expression of TN-C in Murine Uterine Endometrium During Early Pregnancy
On immunohistochemical analysis, TN-C was not detected in uterine endometrium of nonpregnant mice (Fig. 1a). In pregnant animals, expression was evident in stroma along the basement membrane beneath the luminal epithelium on Day 0 (Fig. 1c), this becoming weaker on Day 2 (Fig. 1e) and disappearing on Day 3 (Fig. 1g), but reappearing on Day 4 (Fig. 1i). On Day 5, immediately after decidual formation, it was present in stromal tissues surrounding the deciduomas (Fig. 1k). Throughout the pregnant and nonpregnant period, the inner smooth muscle layer was positive for TN-C (Fig. 1k). When the IgG fraction adsorbed with TN-C was used as the first antibody, the staining was abolished. In situ hybridization analysis revealed TN-C mRNA expression by stromal cells in early pregnancy. Although TN-C mRNA was not detected in uterine endometrium of nonpregnant mice (Fig. 1b), it was localized in stromal cells near the basement membrane on Days 0 (Fig. 1d) and 2 (Fig. 1f), but disappeared on Day 3 (Fig. 1h) of pregnancy. Signals were increased again on Day 4 (Fig. 1j), and were evident around deciduomas on Day 5 (Fig. 1l). In pseudopregnant mice mated with males having ligated seminal tubes, almost the same pattern of TN-C expression was observed during Days 0 to 5 until the completion of decidualization (data not shown). Thus, TN-C is produced and secreted by uterine stromal cells during early pregnancy. These signals of TN-C mRNA were not detected in using a sense cRNA probe (data not shown).
|
Induction of TN-C Expression in Uterine Stromal Cells by Sex Steroids, Cytokines, and Prostaglandins
In order to examine factors inducing TN-C in uterine stromal cells, we selected IL-1, PGE2, PGF2, progesterone, and ß-estradiol as candidates. The effects of b-FGF and TGF-ß, general TN-C inducers for mesenchymal cells, were also tested. When cells were cultured in medium containing 0.1% serum, they did not show positive staining for TN-C (Fig. 2a). In the cells treated with TGF-ß, accumulation of TN-C-positive granules was observed in perinuclear regions (Fig. 2b). The effect of b-FGF was stronger than TGF-ß, with numerous TN-C-positive granules accumulation throughout the cytoplasm (Fig. 2c). Interleukin-1 stimulated TN-C expression as strongly as b-FGF (Fig. 2d), while PGE2 (Fig. 2e) and PGF2 (Fig. 2f) exerted the effects comparable with that of TGF-ß. Progesterone and ß-estradiol did not induce TN-C synthesis (data not shown). The results of immunofluorescence staining after addition of various factors are summarized in Table 2.
|
|
We also examined induction of TN-C by these substances at the mRNA level to confirm immunocytochemical findings. Total RNA was isolated from cultured stromal cells after treatment with the substances for 6 h and 24 h. The TN-C mRNA signals were increased by addition of IL-1, PGE2, and PGF2, in comparison with the control level (Fig. 3). Treatment with progesterone or ß-estradiol exerted no significant effect. The same results were obtained after both 6-h and 24-h treatments (data not shown). Expression of TN-C was increased by addition of IL-1 in a dose-dependent manner (Fig. 4a). The PGE2 and PGF2 were effective at TN-C induction over 1 ng/ml (Fig. 4b).
|
|
Expression of IL-1 in Murine Uterus During Early Pregnancy
Because TN-C was induced by IL-1, we performed an analysis of IL-1 expression in murine uterus during the preimplantation period, on pregnancy Days 0 to 5. Interleukin-1 was immunohistochemically detected in the epithelial cells, and the labeling became more intense day by day, while the epithelium of nonpregnant mice was only faintly labeled (Fig. 5a). From Day 0, labeling of endometrial luminal epithelia appeared positive (Fig. 5b), increased on Day 2, was intense on Day 4 (Fig. 5c), and decreased on Day 5 (Fig. 5d). No IL-1 staining was detected in endometrial glands or stromal cells in either the nonpregnant or pregnant cases.
|
Induction of IL-1 in Uterine Epithelial Cells by Ovarian Steroid Hormones
To examine which steroids induce IL-1 expression, total RNAs were isolated from cultured epithelial cells, and reverse transcription (RT)-PCR was performed for IL-1. Interleukin-1 mRNA expression was very low in the epithelial cells without treatment. Expression proved inducible by treatment with progesterone (2.4 ± 0.16-fold increased), but not with ß-estradiol (1.06 ± 0.25-fold) in comparison with control (Fig. 6). In stromal cells, IL-1 mRNA was not detected, even after progesterone treatment (data not shown).
|
Induction of TN-C by Interaction of Epithelial and Stromal Cells after Steroid Treatment
Our results suggested that progesterone induces secretion of IL-1 by epithelial cells, which in turn results in TN-C production by stromal cells. To clarify regulation of TN-C expression by interactions between the two cell types, stromal cells on well bottoms were cocultured with epithelial cells on culture inserts coated with Matrigel. After treatment with progesterone, TN-C mRNA levels in the stromal cells under the coculture conditions were significantly increased in comparison with no or ß-estradiol treatment (Figs. 7 and 8). Expression of TN-C was not detected in epithelial cells (data not shown). When stromal cells alone were cultured, addition of progesterone or ß-estradiol did not affect TN-C expression (Fig. 8).
|
|
To prove the involvement of IL-1 in TN-C expression, we added monoclonal hamster anti-IL-1 antibody into the coculture system. Relative amounts of TN-C mRNA after progesterone treatment decreased to the 0.57 ± 0.21-fold, that using control IgG (Fig. 9a). A previous report demonstrated increased synthesis of prostaglandins in stromal cells stimulated by IL-1 [5]. Because the effects of IL-1 on TN-C expression could be due to prostaglandins, we tested the influence of indomethacin (10 µM), an inhibitor of prostaglandin synthesis. The treatment had no significant impact on relative amounts of TN-C mRNA induced by IL-1 and progesterone in the coculture system (3.74 ± 0.23-fold with progesterone alone versus 3.86 ± 0.19-fold with progesterone plus indomethacin) (Fig. 9b).
|
DISCUSSION
Previous analyses of the distribution of TN-C in human uterine endometrium [22, 23] and rodent uterine endometrium during early pregnancy [24–26] suggested that TN-C plays an important role in implantation. Its function may be to modulate adhesion of uterine endometrial cells, thereby promoting embryo penetration [26]. Our present study demonstrated that, in agreement with the reports of Michie and Head [24] and Kida et al. [25], TN-C accumulates in the endometrium stroma immediately beneath the luminal epithelium in early pregnancy, being produced by uterine stromal cells, as shown by in situ hybridization. In addition, our sequential observations demonstrated that the expression disappears on Day 3 and reappears just before embryo attachment, on Day 4. Strong expression by stromal cells was also observed around deciduomas. Expression of TN-C after Day 4 appears to be under hormonal control, while that on Day 0 could be due to mechanical stimulation. In the normal estrous cycle, expression has been detected in estrous and diestrous phases, corresponding with the progesterone peak [24, 33, 34]. Furthermore, in ovariectomized female rats, progesterone supplements induce TN-C expression [24]. Thus, the hormonal environment could control TN-C expression of stromal cells around implantation. However, there is no evidence that steroids directly induce TN-C.
In order to investigate the regulatory mechanism of the TN-C expression around the embryo attachment stage, we isolated and purified stromal cells from murine uteri. Although TN-C was not expressed in nonpregnant endometrium in vivo, the stromal cells showed expression in vitro, possibly due to alteration of the environment and/or phenotype during passage. Serum stimulation is also known to induce TN-C expression [26]. With the medium supplemented with 0.1% FCS used here, low expression of TN-C was detected. Interleukin-1, PGE2, and PGF2 significantly upregulated TN-C expression in stromal cells in culture, whereas progesterone or ß-estradiol did not. However, when the stromal cells were cocultured with epithelial cells, TN-C expression was stimulated by progesterone. It has been reported that immunolabeling of IL-1 in the rabbit endometrial epithelium is intense from Days 3 to 6 of pregnancy [7]. In the murine endometrium, as we immunohistochemically showed in the present report, IL-1 expression in epithelial cells is also increased during the blastocyst stage during preimplantation [6, 8], when both progesterone and estradiol are transiently high [35, 36]. As we confirmed in this study, progesterone treatment upregulates IL-1 mRNA in cultured epithelial cells [5, 37].
Furthermore, IL-1 upregulates synthesis of PGE2 and PGF2 by uterine stromal cells [5]. Taken together, the findings are consistent with the hypothesis that a high level of steroid hormone during the preimplantation period upregulates IL-1 expression in epithelial cells, that could also stimulate increased synthesis of PGE2 and PGF2 in stromal cells. An increased level of IL-1, directly or mediated by PGE2 and PGF2, could induce TN-C expression in stromal cells (Fig. 10). When IL-1 was neutralized by the antibody, TN-C mRNA expression of the stromal cells in the coculture system simulated by progesterone decreased about 40%. Inhibition of prostaglandin synthesis by indomethacin, in contrast, did not alter the expression level. These findings indicate that IL-1 is involved in the interaction between epithelial cells and stromal cells, and that the direct stimulation by IL-1 secreted from epithelial cells could be a main regulatory mechanism of TN-C expression in stromal cells. In conclusion, TN-C expression in murine uterine stromal cells during early pregnancy appears to be regulated by interaction of epithelial and stromal cells. This presumably also impacts on other genes related to remodeling of the endometrium around implantation.
|
FOOTNOTES
First decision: 30 March 2000.
1 Correspondence: Toshimichi Yoshida, Department of Pathology, Mie University School of Medicine, 2-174 Edobashi, Tsu, Mie 514-8507, Japan. FAX: 81 59 231 5210; t-yosida{at}doc.medic.mie-u.ac.jp 
Accepted: July 27, 2000.
Received: February 29, 2000.
REFERENCES
induces prostaglandin E2 (PGE2) and PGF2 secretion by uterine stromal cells in vitro. Endocrinology 1993; 132:300–308.[Abstract] in mouse uterus during the peri-implantation period of pregnancy. J Reprod Fertil 1993; 97:83–89.[Abstract] in the rabbit uterus during early pregnancy. Early Pregnancy 1995; 1:201–205., interleukin-1ß, and interleukin-1 receptor type I mRNA during preimplantation mouse development. J Reprod Immunol 1996; 32:27–35.[CrossRef][Medline] opposes progesterone-mediated suppression of MMP-7. Ann N Y Acad Sci 1997; 828:137–145. are produced in the mouse uterus during the estrous cycle and are induced by estrogen and progesterone. Dev Biol 1992; 151:297–305.[CrossRef][Medline]This article has been cited by other articles:
|
|
 |
|
  A.A Fouladi-Nashta, L Mohamet, J.K Heath, and S.J Kimber Interleukin 1 Signaling Is Regulated by Leukemia Inhibitory Factor (LIF) and Is Aberrant in Lif-/- Mouse Uterus Biol Reprod, July 1, 2008; 79(1): 142 - 153. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E.A. Campbell, L. O'Hara, R.D. Catalano, A.M. Sharkey, T.C. Freeman, and M. H. Johnson Temporal expression profiling of the uterine luminal epithelium of the pseudo-pregnant mouse suggests receptivity to the fertilized egg is associated with complex transcriptional changes Hum. Reprod., October 1, 2006; 21(10): 2495 - 2513. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. J. Han, J. Jeong, F. J. DeMayo, J. Xu, S. Y. Tsai, M.-J. Tsai, and B. W. O'Malley Dynamic Cell Type Specificity of SRC-1 Coactivator in Modulating Uterine Progesterone Receptor Function in Mice Mol. Cell. Biol., September 15, 2005; 25(18): 8150 - 8165. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. J Kimber Leukaemia inhibitory factor in implantation and uterine biology Reproduction, August 1, 2005; 130(2): 131 - 145. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Fujinaga, K. Onoda, K. Yamamoto, K. Imanaka-Yoshida, M. Takao, T. Shimono, H. Shimpo, T. Yoshida, and I. Yada Locally applied cilostazol suppresses neointimal hyperplasia by inhibiting tenascin-c synthesis and smooth muscle cell proliferation in free artery grafts J. Thorac. Cardiovasc. Surg., September 1, 2004; 128(3): 357 - 363. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D Pfander, N Heinz, P Rothe, H-D Carl, and B Swoboda Tenascin and aggrecan expression by articular chondrocytes is influenced by interleukin 1{beta}: a possible explanation for the changes in matrix synthesis during osteoarthritis Ann Rheum Dis, March 1, 2004; 63(3): 240 - 244. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
