|Home | About | Journals | Submit | Contact Us | Français|
Despite the female predilection for joint diseases, and the known effects of female hormones in regulating chondrocyte function, the various female hormone receptor subtypes in joints are not well characterized, and comparisons in receptor profiles between joints and genders is lacking. This investigation characterized and compared the relative levels of estrogen receptors (ER) –α and -β, relaxin receptors LGR7 and LGR8, and progesterone receptor (PR) in the temporomandibular joint (TMJ) disc, knee meniscus and pubic symphysis fibrocartilages.
Fibrocartilaginous cells from 12-week old mice were maintained in serum-containing α-MEM until confluence. Total RNA and cell lysates were assayed by RT-PCR, qRT-PCR, immunocytochemistry and Western blots, and joint sections subjected to immunohistochemistry.
All hormone receptors assayed were present in the three joints, but showed substantial differences in expression levels between joints. TMJ cells had higher ER-α (>2.8-fold), ER-β (>2.2-fold), LGR7 (>3-fold) and PR (>1.8-fold), and lower LGR8 (0.5-fold) gene expression levels than knee meniscus cells. The ratio of ER-α:ER-β and LGR7:LGR8 was 1.8- and 7.5-fold higher, respectively, in TMJ than in knee meniscus cells. The profile of hormone receptors in the TMJ disc were similar to those in the pubic symphysis. Immunochemistry confirmed the differential expression patterns of these receptors in the three tissues. The TMJ cells demonstrated sexual dimorphism in the levels of PR.
The findings suggest that these fibrocartilages are putative target tissues for actions of female hormones. The differential expression profiles of the hormone receptors in the three joint fibrocartilages and the sexual dimorphism in ERs in TMJ disc cells are likely to result in varied downstream effects in response to hormones within these fibrocartilaginous tissues.
The high female to male preponderance of degenerative joint diseases is well established,1,2 suggesting a potential role of female sex hormones in causing or predisposing to these disorders. Indeed, it has been shown that estrogen modulates immune responses, plasma extravasation, pain responsiveness, and matrix synthesis and degradation in joints in vitro and in vivo.3-6 of these mechanisms, the modulation of extracellular matrix (ECM) remodeling activities by estrogen, progesterone and another hormone relaxin may constitute one of the key mechanisms for predisposing joints to degenerative changes. While the effects of estrogen on tissue turnover appear to be dose- and tissue-dependent7-10, in general it enhances regulating the gene expression of various tissue degrading matrix metalloproteinases (MMPs),7,8 in various cell types and causing matrix loss in specific tissues including that of the temporomandibular joint (TMJ) fibrocartilage9. Estrogen's effects on cartilaginous tissues are likely mediated by their cytosolic steroid receptors ER-α and –β. Although, some evidence has been presented on the presence of ER-α in cartilaginous tissues,11,12 these are largely qualitative data. Additionally, although the comparison between several investigations suggest variations in estrogen receptor levels in different cartilages,13-15 these observations require validation by studies in which different cartilages are studied concurrently. Finally, the presence of ER-β in joints is not well characterized, the relative concentrations of ER-α to –β are not known, and the difference in relative levels of these two receptors in various joints has not been determined. The determination of the relative cellular concentrations of ER-α and –β is important since these receptors often have varied and sometimes opposite downstream effects, they often regulate different target genes, and ER-β has the capacity to repress transcriptional activity of ER-α with the same ligand.16-18 Within various reproductive tissues, the net response of cells to estrogen is determined by the ratio of ER-α to ER-β,16,19,20 emphasizing the potential importance of determining this ratio in cartilage.
As with estrogen, relaxin is known to be a potent mediator of ECM turnover. Relaxin H2, the major stored and circulating form of relaxin in humans, is a 6-kDa polypeptide belonging to the insulin family of structurally related hormones but its activities are distinct from other members of this family. It has long term effects on connective tissue composition by altering the concentrations and organization of key matrix macromolecules through modulation of their synthesis,21,22 or their increased degradation by enhancing the expression of MMPs,23-25 or both.22 Specifically, within the fibrocartilaginous TMJ disc, relaxin induces MMP-1 (collagenase-1) and -3 (stromelysin-1), which is paralleled by loss of collagen and proteoglycan from this tissue.9,26,27 Relaxin has two known leucine-rich guanine nucleotide-binding (G protein)-coupled (LGR) receptors, designated LGR7 and LGR8. 28-30 Relaxin binds with high affinity to LGR7 receptor and with low affinity to LGR8 receptor.28,31 Relaxin receptors and LGR7 transcripts have been identified in reproductive and non-reproductive tissues such as the brain, kidney, lung, and the anterior cruciate ligament of knee joint.28,32 However, there is currently no evidence of relaxin receptors in fibrocartilaginous tissues of joints.
In contrast to the effects of estrogen and relaxin in enhancing MMP expression, progesterone has been shown to down-regulate several MMPs33-35 in reproductive tissues, and more recently of MMP-3, -9 and -13 in TMJ fibrocartilage.36 The physiologic effects of progesterone are mediated by interaction of the hormone with specific intracellular progesterone receptors (PRs). The PRs are members of a large family of ligand-activated nuclear transcription regulators, which include specific receptors for several other steroid hormones, but also for retinoids, thyroid hormone, and the active Vitamin D metabolite. Although progesterone is known to mitigate tissue degradation in reproductive and non-reproductive tissues including TMJ fibrocartilage,9,33-36 its role and that of its receptor in joint tissue turnover and their potential contribution to joint diseases have not been fully characterized.
Despite the likely importance of hormone activity on cartilaginous tissues of joints in health and disease, the presence of specific hormone receptors in joints has not been fully determined, the different ER and relaxin receptor subtypes and their relative expression levels are not characterized, and comparisons in receptor profiles between joints and genders is lacking. Therefore, the purpose of this investigation was to compare both qualitatively and quantitatively the profiles and levels of estrogen, relaxin and progesterone receptors in the TMJ and the knee meniscus. The pubic symphysis fibrocartilage was used as a reference positive control. The findings on relative receptor levels and profiles in joint tissues could provide important insights not only on whether these cartilages are putative target tissues for these hormones, but also on the relative metabolic activities in each of these tissues in response to the respective hormones.
All cell culture reagents and media were purchased from Invitrogen Corp. (Carlsbad, CA) and chemicals were from Sigma-Aldrich Corp. (St. Louis, MO) unless otherwise mentioned. Total RNA extraction and reverse-transcriptional polymerase chain reaction (RT-PCR) kits were from Qiagen Corp. (Valencia, CA), and quantitative real time RT-PCR (qRT-PCR) reagents were obtained from Applied Biosystems (Carlsbad, CA). BCA protein assay kit was purchased from Pierce Biotechnology (Rockford, IL) and Western blot reagents were from Bio-Rad Laboratories (Hercules, CA). Primary antibodies to mouse ER-α, ER-β, LGR8, PR, and actin were all from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA) and the primary antibody to mouse LGR7 was from Phoenix Pharmaceuticals, Inc. (Burlingame, CA). C57BL/6J mice were used in this study. All animal procedures were conducted in compliance with federal and institutional guidelines and approved by the University of Michigan Institutional Animal Care and Use Committee.
Fibrochondrocytes were isolated from 12-week old male and female C57BL/6J mouse TMJs, and female knee meniscus and pubic symphysis fibrocartilages. The tissues were retrieved under dissection microscope in aseptic conditions after euthanasia with an overdose of CO2. The tissues were washed with Hank's balanced salt solution (HBSS) containing 1% antibiotics, minced and incubated in α-MEM phenol-free media supplemented with 10% fetal bovine serum (FBS) for 2 to 4 weeks. Passage 4 to 6 cells cultured at 1×106 cells in a 100-mm dish and maintained in serum-containing medium until confluent were used for all subsequent experiments.
We first quantified the gene expression levels for each of the receptor isoforms in the cultured fibrochondrocytes by qRT-PCR (ABI PRISM 7500). Upon confluence, total RNA was extracted from fibrochondrocytes of female and male mice using RNeasy Mini kit according to manufacturer's protocol (Qiagen). Total RNA (1 μg) was reverse transcribed using Ominiscript RT kit (Qiagen) with random hexamers as primers. A 1:100 (v/v) dilution of the resulting cDNA from the reverse transcription was utilized as the template and the relative content of mRNA for specific receptors in each sample was quantified using SYBR green master mix and the specific primers designed with Primer Express 2.0 (Applied Biosystems) (Table 1) to detect gene expression of ER-α, ER-β, LGR7, LGR8 and PR. The successful qRT-PCR amplification was confirmed by performing the melting curve test and observing a single peak for each gene. qRT-PCR analyses were also conducted to compare the relative expression levels of these hormone receptors in male and female TMJ fibrochondrocytes. 18 S rRNA was used as an internal control. Cycle threshold values of the genes of interest and the quantitative gene expression levels relative to 18s rRNA in the samples were determined.
Next, in order to derive to ratios for ER-α:ER-β and LGR7:LGR8 in the fibrochondrocytes, we performed semi-quantitative RT-PCR for these receptors. A 1:10 (v/v) dilution cDNA from the previous experiment was used as a template to amplify for female hormone receptors using a PCR master mix (Qiagen) and specific primers for mouse ER-α, ER-β, LGR7, LGR8, and PR (Table 1). Mouse glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was amplified at the same time as an internal control. Equal volumes of PCR product from each group were resolved by electrophoresis on a 1% (w/v) agarose gel and visualized with ethidium bromide. The optical densities of PCR products were quantified using an image analysis system (NIH Image J 1.32), and standardized by GAPDH signal.
Confluent female TMJ, knee meniscus and pubic symphysis fibrochondrocytes were lysed using RIPA lysis buffer (Santa Cruz Biotechnology, Inc.) and the protein concentration in the lysate was determined by BCA protein assay (Pierce Biotechnology). The cell lysates, standardized by total protein were mixed with 4x sample buffer, and electrophoretically resolved on 10% (w/v) SDS-PAGE. The proteins were transferred to PVDF membrane, blocked overnight with 5% (w/v) milk in PBS, incubated with primary polyclonal antibodies specific for ER-α, ER-β, LGR7, LGR8, or PR for 2 hours, washed and incubated for 2 hours with horseradish peroxidase-conjugated secondary antibody at room temperature. Following further washes, the protein bands were visualized by incubation of the blot with a chemiluminescent substrate (Pierce Biotechnology) and exposing the blot to an X-ray film. After incubation with stripping buffer (Pierce Biotechnology) for 40 minutes, the blot was re-probed with primary antibody specific for actin to confirm equal loading of samples.
The presence and distribution of ER-α, ER-β, LGR7, LGR8, and PR in female mouse TMJ, knee meniscus and pubic symphysis fibrochondrocytes were further validated using immunocytochemistry. Early passage fibrochondrocytes (5×10C4 cells/chamber) were seeded into slide chamber and maintained in serum containing α-MEM. Upon 70% confluence, cells were washed twice with PBS, fixed with 75% (v/v) ethanol for 15 minutes, and washed twice with PBS. The cells were incubated with 1 mg/ml hyaluronidase for 30 minutes, washed with PBS and blocked with 1% (w/v) bovine serum albumin for 30 minutes. The cells were incubated with primary antibodies specific for ER-α, ER-β, LGR7, LGR8, PR or non-immune IgG for 3 hours. After washing with PBS twice, cells were incubated with fluorescent-labeled secondary antibody (1:400 v/v) (Invitrogen Corp.) for 1 hour. The nuclei were stained with Dapi. Following further washing, the cells were mounted with coverslips and observed under a fluorescent microscope (Nikon TS100, Tokyo, Japan).
To localize female hormone receptors in the fibrocartilages, the entire TMJ, knee and pubic symphysis were retrieved from female mice, fixed in 4% (w/v) paraformaldehyde, and decalcified with 0.2 M EDTA/PBS solution (pH 7.4). The tissues were routinely embedded and sectioned into 5 μm paraffin sections. Immunostaining was performed using Histostain-SP kit (Invitrogen Corp.) according to manufacturer's instructions. Briefly, after deparaffinization with xylene, rehydration, quenching of endogenous peroxidase with 3% (v/v) hydrogen peroxide, and incubation with bovine blocking serum, the sections were incubated with primary antibodies specific for ER-α, ER-β, LGR7, LGR8, PR or non-immune serum for 3 hours. Sections were washed and incubated with biotinylated secondary antibody for 10 minutes followed by incubation with a streptavidin-peroxidase conjugate for 10 minutes. The sections were washed and incubated with diaminobenzidine solution in distilled water for 5 minutes, and counterstained with hematoxylin for 3 minutes. After mounting with coverslips, the specimens were viewed and analyzed under a microscope (Nikon TS100).
Numerical data were plotted as means ± standard deviations (SD). The statistical significance of any differences in quantitative data was determined by single factorial analysis of variance (ANOVA), and the inter-group differences were determined by Tukey's test. A p value of less than 0.05 was considered statistically significant.
Fluorescence immunocytochemistry revealed conspicuous positive staining for both ER-α and ER-β in the TMJ and pubic symphysis cells, which was largely confined to the cytoplasm and nucleus (Fig. 1). However, only a weak signal of both ER isoforms was detected in knee meniscus cells. Similarly, although the TMJ disc cells and pubic symphysis cells demonstrated high levels of cell surface and cytoplasmic staining for LGR7, staining for this receptor was barely perceptible in knee meniscus cells. Finally, while LGR8 expression was strong in the pubic symphysis and knee meniscus cells, those from the TMJ stained weakly for this receptor.
Since in vitro cell culture system may not accurately represent the in vivo expression of hormone receptors due to an altered cell state through dedifferentiation and the lack of matrices, we performed immunohistochemistry on intact tissues to verify results obtained from cell culture and to characterize the in situ distribution of these receptors. Subjectively, we found higher expression levels of ER-α, ER-β, and LGR7 in the TMJ disc and pubic symphysis as compared to the knee meniscus, while LGR8 was more highly expressed in the knee meniscus and pubic symphysis versus the TMJ disc (Fig. 2)- findings that are similar to our observations in cell culture.
We next quantitatively and semi-quantitatively analyzed the mRNA and protein expression levels of estrogen, relaxin and progesterone receptors from TMJ, pubic symphysis and knee meniscus cells. Since the knee meniscus had the lowest levels of most of the receptors, the qRT-PCR data was standardized to expression levels in the knee meniscus cells. We found that TMJ disc fibrochondroctyes had significantly greater ER-α (>2.8-fold), ER-β (>2.2-fold), LGR7(>3-fold) and PR (>1.8-fold) gene expression levels, and significantly lower (<0.5-fold) LGR8 gene expression than knee meniscus fibrochondrocytes (Fig. 3). Similarly, the pubic symphysis cells had significantly greater gene expression levels for ER-α (>12-fold), ER-β (>3-fold), LGR7 (>5-fold) and PR (>2.5-fold) than knee meniscus fibrochondrocytes. The expression level for LGR8 was statistically lower in pubic symphysis cells (<0.7-fold) than knee meniscus cells, whereas its levels were not significantly different between the TMJ and pubic symphysis cells.
The findings on qRT-PCR were confirmed by semi-quantitative RT-PCR and Western blots. These assays showed that the TMJ and pubic symphysis cells have higher expression levels of ER-α, ER-β, LGR7 and PR than the knee meniscus cells, while LGR8 is most highly expressed in the knee meniscus fibrochondrocytes (Fig. 4A and B). Further analysis of the RT-PCR gels using videodensitometry was used to derive ER-α:ER-β and the LGR7:LGR8 ratios. These analysis revealed that the ER-α:ER-β in TMJ disc and pubic symphysis cells were significantly higher (2-fold and 2.5-fold, respectively) than that in knee meniscus cells (Fig. 5). Similarly, the LGR7:LGR8 ratio was approximately 7.5-fold greater in TMJ disc cells and 2.5-fold higher in pubic symphysis cells than in the knee meniscus cells.
In order to understand whether the gender differences in TMJ diseases may result at least in part from differences in concentrations of hormone receptors in target tissues between males and females, qRT-PCR was performed on RNA extracted from isolated TMJ disc cells. We found that ER-α and ER-β levels were significantly lower (67% and 25%, respectively) in cells from male than female mice (Fig. 6). In contrast, the expression levels of LGR7 and LGR8 were slightly but not significantly lower (88% and 76%, respectively), while PR receptor levels were slightly, but not significantly greater (116%) in TMJ cells from males and female mice.
Our findings show substantial differences in the quantities of estrogen, relaxin and progesterone receptors expressed in cells from the mouse TMJ, knee meniscus and pubic symphysis fibrocartilages. In general, the TMJ disc and pubic symphysis cells express higher concentrations of ER-α, ER-β, LGR7 and PR than the knee meniscus cells, while that of LGR8 are highest in the knee meniscus followed by the pubic symphysis and TMJ disc cells. Accordingly, the TMJ disc and pubic symphysis cells have higher ratios of ER-α:ER-β and LGR7:LGR8 than knee meniscus cells. Additionally, TMJ disc cells from female mice had higher concentrations of ER-α and ER-β than cells from male mice. Since the binding of female hormone ligands to their respective receptors are known to affect cell proliferation, differentiation, matrix turnover and tissue phenotype,6,37-41 these dissimilarities in absolute and relative concentrations of receptors between the fibrocartilages are likely to result in differential downstream effects in response to hormones within these three fibrocartilaginous tissues.
Although no studies have been done to show the presence of relaxin receptors in cartilage, several previous investigations to characterize steroid hormone receptors in cartilaginous tissues have produced varied findings. Thus, while chondrocytes within the knee articular cartilage have been shown to be positive for estrogen and progesterone receptors,14,15 those from thyroid cartilage of male and female subjects did not react with ER-α and PR antibodies.13 These and other studies also show differences in distribution of mRNA and proteins for ERs and PR in articular cartilages and in chondrocytes at various stages of differentiation in developing endochondral bone.42,43 Together, these studies highlight the need for concurrent comparative analysis of receptor levels in cartilages in different joints, as performed in the present study. Also, the presence and distribution of these receptors in fibrocartilaginous tissues is not well characterized. In a preliminary study on symptomatic and asymptomatic human subjects ER, whose isoform was not determined, and PR were localized in TMJ discs.44 The single study done to identify and localize the ER-α isoform in the TMJ showed the presence of this receptor in numerous tissues of the TMJ including the articular disc.11 The present study iS the first to show the presence of relaxin receptors in any type of cartilaginous tissues, and conclusively demonstrate the differential expression of both ER isoforms, PR and relaxin receptors in fibrocartilaginous tissues from the TMJ and other joints.
The localization of these hormone receptors in developing endochondral bone, mature cartilage and, as shown by the present studies, in fibrocartilage attests to the importance of the receptors and their respective hormones in cartilage development and physiology.12,14,42,43 It is well known that these hormones regulate the proliferation, differentiation, maturation and programmed cell death of chondrocytes.6,39-41 However, different cartilages also show site and age dependent divergent responses to hormones. For example, estrogen and progesterone together increase mesenchymal cell proliferation, but not proteoglycan synthesis during matrix induced endochondral bone formation.40 Additionally, estrogen enhances endochondral bone growth in fetal rat metatarsal bones, while progesterone has no effects.45,46 Conversely, ovariectomy leads to decreased number of growth plate chondrocytes resulting from decreased cell proliferation and increased apoptosis in rabbits41 - a finding that is in concurrence with the effects of the SERM tamoxifen on chondrocytes in the resting and hypertrophic zones of cultured metatarsal bones and chondrocyte cell lines.47 In contrast to the findings on long bones, in the neonatal mandibular condylar cartilage, progesterone increases proteoglycan synthesis.48 Also, as opposed to developing long bones, estrogen treatment of mature rat TMJ condyle explants results in significant reduction in cartilage thickness, mitotic index, decrease in proteoglycan content and an increase in type X collagen in hypertrophic chondrocytes suggesting that estrogen decreases cartilage thickness by inhibiting chondrocyte differentiation and increasing chondrocyte maturation.6 Taken together with the findings of the present study, such site and developmentally specific differences in responses to hormones may result from differences in absolute and relative local concentrations of hormone receptors. Therefore, the varied effects of each of these hormones on cellular responses and our current findings showing different hormone receptor profiles in various fibrocartilages suggest a substantial complexity in the hormonal control of cell proliferation, differentiation, and maturation and therefore on the resultant tissue phenotype in different cartilaginous tissues.
Besides the contribution of female hormones in regulating cell proliferation and differentiation, estrogen, progesterone and relaxin are important regulators of ECM turnover both in development42,43,49 and possibly also in disease.7,44,50 Evidence for such a matrix regulatory role of hormones are provided by studies showing that estrogen induces the expression of MMP-3, -9, and -13 in various cells including fibrocartilaginous cells.7,8,36 Although little is known of the effects of progesterone in regulating MMPs in cartilage, it produces a dose-dependent receptor-mediated decrease in MMP-9 expression in trophoblasts.33 Finally, relaxin is known to enhance ECM turnover in reproductive21,22,51 and non-reproductive tissues including fibrocartilage9,36,38 likely through its upregulation of various MMPs.26,27,36,38 These findings together with those of the present study on the presence of estrogen, relaxin and progesterone receptors in joint tissues point to an important role of these hormones and their receptors in cartilage metabolism and matrix remodeling activities in health and possibly in disease.
Degenerative joint diseases are frequently associated with enhanced matrix turnover modulated by MMPs.52 It is plausible that the aberrant modulation of ECM turnover by female hormones, which could be dependent on both the systemic levels of hormones and the receptor profiles in target tissues, may predispose to degenerative changes and progressive disease in target joints. Indirect evidence for such a link is provided by our current and previous studies showing that the most responsive target tissues, namely the pubic symphysis and TMJ disc, that show enhanced matrix loss by estrogen and relaxin treatments in ovariectomized rabbits, have similar levels and profiles of the respective hormone receptors, which differ substantially from that in the non-responsive knee meniscus.9 Similarly, our findings on the higher levels of ERs in female vs male TMJ disc cells, which have also been shown in preliminary studies in non-human primates,53 may partially explain the preponderance of TMJ disorders in women of reproductive age. Further studies are likely to help decipher any relationship between hormone receptors and altered remodeling activities in cartilaginous tissues.
In concurrence with the differences in receptor profiles between the TMJ disc and pubic symphyis on one hand, and the knee meniscus on the other, we found similar ER-α to ER-β and LGR7 to LGR8 ratios in the TMJ disc and pubic symphysis cells that were significantly greater than those in the knee meniscus. The determination of ratios of various receptors and/or their isoforms is important because previous studies have shown that the relative receptor concentrations in target cells often contribute to diverse cell responses. Thus, in cells containing both ERs, ER-β has the capacity to repress transcriptional activity of ER-α,16,17 and therefore may act as a dominant negative regulator of ER activity. Also, microarray analysis have shown that ER-α and ER-β regulate different target genes in response to estrogens and SERMs and will produce distinct phenotypes in cells that express predominantly ER-α compared with those expressing ER-β.18 Therefore, any change in the ratio of ER-α:ER-β in tissues that occurs with aging or disease states may alter the tissue response to estrogens.16,19,20 Although the precise contributions of LGR7 and LGR8 and their ratios to changes in cartilaginous tissues on activation by relaxin are not known, important evidence for the role of LGR7 to in vivo remodeling of matrices is provided by the phenotypic characteristics of the female LGR7 null mice. These mice show several abnormal characteristics similar to those described in relaxin-deficient mice that include increased local density of collagen of the nipple and an unrelaxed pubic symphysis with densely packed collagen fibers during parturition.54,55 While these studies suggest that LGR7 is a likely candidate receptor in the modulation of tissue remodeling by relaxin, its role and that of LGR8 in these process remains to be determined. Our findings and others16,18-20 suggest the intriguing possibilities that not only the systemic hormone concentrations, but also the local profiles and ratios of receptors may have on the net cellular responses and the resultant tissue phenotype.
It has previously been shown that cultured fibrochondrocytes undergo increasing dedifferentiation of the chondrocytic cells with increasing number of cell passages.56,57 It is plausible that this dedifferentiation into a greater fibroblastic phenotype may affect the expression of hormone receptors on these cells. However, since our previous findings show similar MMP responses of TMJ cells whether in their natural matrix environment within tissue explants or in monolayer cultures to relaxin 26,27, it is likely that the use of cells in early passages as in the present experiments produces minimal changes in this system. This point is further validated by our findings on the similarities in relative levels of receptors in cell culture and those of in situ localization in tissue sections in the three types of tissues.
In summary, our findings on the localization and quantification of estrogen, progesterone and relaxin receptors in fibrocartilaginous tissues suggest that cartilage is a likely target tissue for the actions of these three hormones. Also, the differences in receptor types and ratios in different fibrocartilages together with the different effects of their respective hormones on cell proliferation, differentiation, apoptosis and matrix remodeling activities may be critical in determining the ultimate tissue phenotype.
This study was supported by NIH/NIDCR RO1 DE018455 and KO2 DE00458. The authors declare that the funding agency did not provide any input in the study design, data collection, analysis and interpretation, manuscript preparation, and in the decision to submit the manuscript for publication.
Publisher's Disclaimer: 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 citable 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.