INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

Periodontal ligament (PDL) cells reside between the cementum of the roots of teeth and the alveolar bone. In this location PDL cells are uniquely situated to maintain the overall integrity of the periodontal ligament (23). Phenotypic features such as the expression of high levels of alkaline phosphatase (AP), transforming growth factor 1 (TGF-1), and osteocalcin, the synthesis of cyclic AMP (cAMP) in response to prostaglandin E₂, and the formation of calcium phosphate nodules categorize them as osteoblast-like cells (4, 15, 17, 18). However, their capacity to take part in osteogenesis and cementogenesis by differentiating into either osteoblasts or cementoblasts distinguishes them from osteoblasts (23). Additionally, PDL cells differ functionally from committed osteoblasts in their ability to synthesize extracellular matrix proteins in the formation of the PDL.

Gingival infections frequently result in the exposure of PDL cells to microbial products as well as to proinflammatory cytokines liberated from neighboring tissue. These cytokines affect the functions of PDL cells; tumor necrosis factor-alpha (TNF-), for example, has been shown to modulate the PDL cell osteoblast-like phenotype and functions (31). Additionally, TNF- and interleukin-1 (IL-1) have been shown to alter the phenotypic characteristics of osteoblasts by inducing down-regulation of AP (16) and by the modulation of collagen, collagenase, proteoglycan, and prostaglandin syntheses (2, 7, 13, 15, 19-21, 33). Overall, these observations suggest that proinflammatory cytokines alter the osteoblast-like characteristics of osteogenic cells; however, the significance of this change in the host is not known.

Here we show that at concentrations as low as 1 ng/ml, IL-1 induces phenotypic changes in PDL cells. PDL cells from healthy periodontium do not recognize bacterial lipopolysaccharide (LPS) nor do they elicit proinflammatory cytokines in response to LPS. Following IL-1 treatment, PDL cells lose their osteoblast-like characteristics while assuming a new LPS-responsive phenotype. Thus, IL-1 is an important regulator of PDL cell functions and directs these cells to participate actively in an immune response during infections, at the expense of their normal osteoblast-like functions. The altered PDL cell phenotype and functions are transient; these cells reacquire their original characteristics following removal of IL-1. Taken together, these findings suggest that proinflammatory cytokines control the homeostasis of the PDL, a function that may be pivotal to the integrity of the PDL as well as to the host immune response during inflammation.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Reagents. All tissue culture media and supplements were purchased from Sigma Chemical Co., St. Louis, Mo. Low-endotoxin fetal calf serum was purchased from HyClone Inc., Logan, Utah. Recombinant human (rHu)IL-1 (specific activity, 1.67 × 10⁷ U/mg of protein) was kindly provided by Hoffmann-La Roche, Nutley, N.J. The LPS from Escherichia coli O127:B8 was purchased from Difco Labs, Ann Arbor, Mich., and the LPS from Actinobacillus actinomycetemcomitans Y4 was kindly provided by Mark Wilson, SUNY Buffalo (32). The oligonucleotide primers were synthesized by the DNA Synthesis Facility of the University of Pittsburgh according to published sequences (35). All reagents for mRNA analysis by reverse transcriptase (RT)-PCR were obtained from Perkin-Elmer (Foster City, Calif.). Enzyme-linked immunosorbent assay (ELISA) kits for IL-1, IL-6, IL-8, and TNF- were purchased from Medgenix Labs (Brussels, Belgium).

Isolation and characterization of PDL cells. Root surfaces of disease-free, erupted third molars from healthy human subjects were scraped to obtain tissue as a source of PDL cells. This tissue was minced and cultured as explants in tissue culture medium (TCM) (RPMI 1640 containing 10% low-endotoxin fetal calf serum, 2 mM glutamine, 100 U of penicillin/ml, 100 µg of streptomycin/ml, and 80 µg of tylosin/ml). The semiconfluent cultures were then grown in tylosin-free medium and cloned by limiting dilutions, and cultures between the 6th and 20th passages were used for experimentation as described earlier (24). PDL cell clones, designated PL-1 (obtained from a 19-year-old white male) and JP-5 (obtained from a 27-year-old white male), were characterized to ensure their PDL cell phenotype by the presence of AP, the formation of calcium phosphate nodules (4), the constitutive expression of mRNA for TGF-1 (22) and osteocalcin, and prostaglandin-induced cAMP formation (27, 28). Cells from both of these lines between the 6th and 20th passages were used. No significant differences were observed in AP activity and calcium phosphate nodule formation for these passages.

Regulation of PDL cell phenotype by rHuIL-1. To determine the possible regulation of PDL cell phenotypic characteristics by rHuIL-1, PL-1 or JP-5 cells (5 × 10⁴/100 µl) were grown in 96-well microtiter plates with various concentrations of rHuIL-1 (0, 0.03, 0.1, 0.3, 1, 3, and 10 ng/ml) in a final volume of 200 µl of TCM at 37°C in 5% CO₂ and 95% humidity. The cultures were replenished with 50% fresh TCM containing identical concentrations of rHuIL-1 on day 6. The AP activities in the rHuIL-1-treated and control cells were analyzed by spectrophotometric analysis at 405 nm over a period of 30 min on days 5 and 28 in digitonin (10 µg/ml)-permeabilized cell cultures in the presence of p-nitrophenyl phosphate (1 mg/ml). As markers for differentiation, the expression of TGF-1- and osteocalcin-specific mRNAs (5, 6) was assessed by RT-PCR following incubation with rHuIL-1 for 6 days.

LPS responsiveness of PDL cells and expression of mRNA for IL-1, IL-6, IL-8, and TNF-. To examine the effect of rHuIL-1 on LPS responsiveness in PDL cells, PL-1 or JP-5 cells were seeded in 6-well plates at a rate of 10⁶ cells/well and pretreated with rHuIL-1 (1 ng/ml) for 5 days. Subsequently, the cells were washed twice and activated with rHuIL-1 (5 ng/ml) or LPS from A. actinomycetemcomitans Y4 or E. coli O127:B8 (100 ng/ml) for 5 h, and mRNA was extracted for the analysis of IL-8-, IL-6-, IL-1-, or TNF--specific mRNA by RNA-PCR (35). Alternatively, cells were washed after rHuIL-1 treatment and grown for another 14 days. Thereafter, the LPS responsiveness was assessed by RT-PCR as described above.

Analysis of mRNA expression by RT-PCR. Total RNA was extracted according to the method described by Chomczynski and Sacchi (8). For preparation of cDNA by reverse transcription, a total of 1 µg of RNA was denatured at 65°C for 15 min. The RNA was then mixed with 4 µl of 25 mM MgCl₂, 2 µl of 10× PCR buffer (Promega, Madison, Wis.), 1 µl of 10 mM oligo(dT₁₅), 1 µl of murine Maloney leukemia virus RT, 5 µl of RNasin, 2 µl of 10 mM (each) deoxynucleoside triphosphate, and distilled water to make the final volume 20 µl. The mixture was incubated at room temperature for 10 min and then placed in a Perkin-Elmer Gene Amp PCR system 9600. The oligonucleotide primers used were made according to previously published sequences for TGF-1 (14), osteocalcin (11), TNF-, IL-1, IL-6, and IL-8 (35). The reverse transcription was carried out at 42°C for 15 min, 92°C for 5 min, and 5°C for 5 min. Each PCR cycle run consisted of the following conditions: DNA denaturation at 95°C for 1.45 min, primer annealing at 60°C for 30 s, and DNA extension at 72°C for 1 min. After 35 cycles of amplification, the reaction was terminated at 72°C for 7 min. The PCR products were stored at 4°C until further analysis.

Synthesis of IL-1, IL-6, IL-8, and TNF- by PDL cells in response to LPS. To determine if the expression of mRNA for proinflammatory cytokines by PDL cells is accompanied by the synthesis of their proteins, one set of 25-cm² flasks containing PL-1 cells (2 × 10⁶ cells/8 ml of TCM) was treated with or without rHuIL-1 at a concentration of 1 ng/ml. The cells were replenished with fresh TCM with or without 1 ng of rHuIL-1 per ml. On days 6 and 21, the treated cells were challenged with LPS from A. actinomycetemcomitans Y4 or E. coli O127:B8 (100 ng/ml) or with rHuIL-1 (5 ng/ml) for 10 h at 37°C. The synthesis of IL-1, IL-6, IL-8, and TNF- in the culture supernatants was assessed by ELISA according to the manufacturer's recommended procedures. The concentrations of each cytokine were calculated by using a standard curve generated for that cytokine at concentrations between 10 pg/ml and 10 ng/ml.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Phenotypic characteristics of osteoblast-like PDL cells. Both PL-1 and JP-5 cells exhibited an osteoblast-like PDL cell phenotype according to previously established criteria (4, 27, 28). Briefly, PL-1 and JP-5 cells were both AP positive (Fig. 1) and expressed TGF-1- and osteocalcin-specific mRNAs (Fig. 2). These cell lines also exhibited a rise in intracellular cAMP in response to prostaglandin E₂ and formed calcium phosphate nodules, as assessed by von Kossa staining (24). In addition, PL-1 and JP-5 cells both exhibited the lack of contact inhibition typically observed in gingival fibroblasts (24).

View larger version (20K): [in this window] [in a new window]   FIG. 1.   Down-regulation of AP activity and TGF-1 mRNA expression in PDL cells by rHuIL-1. (A) PL-1 () and JP-5 () cells were incubated for 6 days in the presence of various concentrations of rHuIL-1 or rHuIL-1 (10 ng/ml) pretreated with 10 µg of rabbit anti-rHuIL-1 immunoglobulin G (, PL-1; , JP-5). The AP activities were measured as described in Materials and Methods. The data represent means and standard errors of the means (SEM) of triplicate values. In some instances small SEM bars are obscured by the symbols. (B) PL-1 (shaded bars) and JP-5 (dotted bars) cells were treated with various concentrations of rHuIL-1 for 6 days as described above for panel A. Thereafter, the presence of TGF-1 mRNA was assessed by RT-PCR. Each point on the y axis represents the average luminescence of duplicate values with less than 5% variation.

View larger version (21K): [in this window] [in a new window]   FIG. 2.   Regulation of PDL cell phenotype by rHuIL-1. (A and B) PL-1 cells (5 × 104/200 µl of TCM/well) were treated with rHuIL-1 (1 ng/ml) for 6 days, followed by removal of rHuIL-1 and replenishment with TCM alone on day 6 and every 6th day thereafter (A). For controls PL-1 cells were treated with TCM alone on day 6 and every 6th day thereafter (B). Cell growth () and AP activity () were determined as described in Materials and Methods. Data are means and standard errors of the means of triplicate values for one of three separate experiments. OD, optical density. (C) Expression of TGF-1 mRNA in PL-1 and JP-5 cells treated for various time intervals with rHuIL-1 as described for panel A or with TCM alone as described for panel B. (D) Expression of osteocalcin mRNA in PL-1 and JP-5 cells treated with rHuIL-1 as described for panel A or with TCM alone. The arrows indicate the day of rHuIL-1 removal from the medium. TGF-1- and osteocalcin-specific mRNA expression were measured as described in Materials and Methods, and values are represented as ratios relative to -actin expression. The data represent one of three separate experiments.

rHuIL-1 modulates phenotypic characteristics of PDL cells. The modulation of phenotypic characteristics of PDL cells by rHuIL-1 in two clonally derived PDL cell lines, PL-1 and JP-5, is shown in Fig. 1. Although these cell lines differed in their constitutive levels of AP activity, incubation of either cell line for 5 days with rHuIL-1 at concentrations greater than 50 pg/ml resulted in a concentration-dependent loss of AP activity, which reached a maximum at 1 ng of rHuIL-1/ml (Fig. 1A). Treatment with rHuIL-1 also down-regulated the expression of TGF-1 mRNA in both PL-1 and JP-5 cells (Fig. 1B). Since PL-1 and JP-5 cells both exhibited phenotypic changes in the presence of rHuIL-1, the next experiments were aimed at investigating whether this phenotypic change (i) was permanent or transient and (ii) had an effect on the LPS responsiveness of these cells.

rHuIL-1-induced phenotypic changes in PDL cells are transient. To determine whether the effects of rHuIL-1 on PDL cells were permanent or transient, the PL-1 cells were first incubated with rHuIL-1 (1 ng/ml) for 6 days. This incubation resulted in the inhibition of more than 98% of the AP activity. However, rHuIL-1 treatment produced no significant change in cell growth rate compared to that of the untreated controls (Fig. 2A and B). The removal of rHuIL-1 from the PL-1 cell cultures resulted in reexpression of AP activity in 12 to 15 days, and this expression reached an optimal level 3 weeks after rHuIL-1 removal (Fig. 2A).

rHuIL-1 treatment renders PDL cells LPS responsive. PDL cells (PL-1) were not LPS responsive and did not express mRNA or proteins specific for IL-8, IL-1, IL-6, or TNF- when challenged with LPS at a concentration of 100 ng/ml (Fig. 3A and 4A). However, these cells were IL-1 responsive in that incubation of PL-1 cells with rHuIL-1 (5 ng/ml) for 5 h resulted in not only expression of IL-6-, IL-8-, and TNF--specific mRNAs but also expression of mRNA for IL-1 itself (Fig. 3A).

View larger version (42K): [in this window] [in a new window]   FIG. 3.   Modulation of PDL cell osteoblast-like phenotype to an LPS-responsive cell phenotype by rHuIL-1 as assessed by induction of mRNA expression for proinflammatory cytokines. (A and B) To examine the effect of rHuIL-1 on the modulation of the PDL cell phenotype, PL-1 cells were either incubated with TCM alone (A) or treated with rHuIL-1 (1 ng/ml) in TCM for 6 days (B). Subsequently, the cells were washed and activated with LPS from A. actinomycetemcomitans Y4 (Aa) or E. coli O127:B8 (Ec) (100 ng/ml) or with rHuIL-1 (5 ng/ml) for 5 h. IL-1-, IL-6-, IL-8-, or TNF--specific mRNA expression was assessed by RT-PCR. The results in panel B show that pretreatment with rHuIL-1 induced LPS responsiveness in PDL cells. (C) To determine whether the LPS responsiveness induced by the rHuIL-1 was permanent or transient, PDL cells were treated for 6 days with rHuIL-1 (1 ng/ml), washed, and further incubated for an additional 15 days in TCM alone. Thereafter, the cells were washed and activated with LPS or rHuIL-1 as described for panels A and B. IL-1-, IL-6-, IL-8-, or TNF--specific mRNA expression was assessed by RT-PCR. The luminescence values of the PCR products in the lanes at the top are expressed relative to -actin mRNA in the histogram at the bottom. The data represent one of three separate experiments.

View larger version (28K): [in this window] [in a new window]   FIG. 4.   Modulation of PDL cell osteoblast-like phenotype to an LPS-responsive cell phenotype by rHuIL-1 as assessed by the synthesis of proinflammatory cytokines. To examine the effect of rHuIL-1 on the LPS responsiveness of PDL cells, PL-1 cells were incubated with TCM alone (A), treated with rHuIL-1 (1 ng/ml) in TCM for 6 days (B), or treated with rHuIL-1 (1 ng/ml) in TCM for 6 days, washed, and replenished with TCM alone for 15 days (C). Subsequently, the cells were washed and activated with LPS from A. actinomycetemcomitans Y4 (Aa) or E. coli O127:B8 (Ec) (100 ng/ml) or with rHuIL-1 (5 ng/ml) for 10 h. The culture supernatants were then harvested, and the synthesis of each cytokine was measured by ELISA. Asterisks signify cytokine concentrations of less than 15 pg/ml. The bars represent means and standard errors of the means of triplicate values.

rHuIL-1-induced LPS responsiveness is transient. In the next series of experiments, we determined whether the rHuIL-1-induced LPS responsiveness in PDL cells was transient or permanent and whether there was an inverse relationship between the osteoblast-like phenotype and LPS responsiveness. PL-1 cells, grown for 6 days in the presence of rHuIL-1 (1 ng/ml), were then washed to remove rHuIL-1 and allowed to grow in fresh medium for 12 days. The cells were then analyzed for the presence of AP activity and for TGF-1- and osteocalcin-specific mRNAs (Fig. 2). Culture of PL-1 cells for 12 days after removal of rHuIL-1 resulted in the reexpression of AP activity and the expression of mRNA for osteocalcin and TGF-1. Furthermore, these cells, when activated with LPS from Y4 or O127:B8 (100 ng/ml) for 5 h, did not exhibit mRNA expression for IL-8, IL-6, TNF-, or IL-1. However, the reactivation of PL-1 cells by rHuIL-1 resulted in IL-6-, IL-8-, TNF--, and IL-1-specific mRNA expression (Fig. 3C). Hence, although rHuIL-1 treatment modulated the LPS responsiveness, these cells remained IL-1 responsive regardless of the length of rHuIL-1 treatment (Fig. 3).

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

Factors critical to understanding the homeostatic mechanisms regulating the regeneration of PDL during inflammation are poorly understood. Among the most important factors controlling tissue destruction and remodeling are proinflammatory cytokines and the potent immunostimulatory agent LPS. During inflammation, the PDL, due to its proximity to the gingival sulcus, is exposed to inflammatory exudates containing proinflammatory cytokines and bacterial products. Since IL-1 is known to profoundly affect both immune and nonimmune cell functions and is one of the major proinflammatory cytokines elaborated during periodontal infections (19, 29), we examined its role on the regulation of PDL cell phenotype and functions.

We show that rHuIL-1 is a potent regulator of PDL cell phenotype and functions. It down-regulates multiple characteristics that define the osteoblast-like PDL cell phenotype, such as AP activity and the expression of TGF-1 and osteocalcin mRNA (4, 10, 22, 25, 28). The observations that down-regulation of the osteoblastic phenotype in PDL cells is rHuIL-1 dose dependent and is inhibited by neutralizing antibodies to rHuIL-1 indicate that these effects are IL-1 specific. It is of interest that modulation of all osteoblast-like phenotypic characteristics of PDL cells tested in this study were directly dependent upon the presence of IL-1 and that its removal resulted in the reversion of PDL cells to an osteoblast-like phenotype. This transient regulation of PDL cell phenotypic characteristics by IL-1 closely resembles the effects of TNF- (24). More importantly, the concentrations of rHuIL-1 that maximally inhibited AP activity in PDL cells are similar to those frequently found in biological fluids surrounding chronic infections in the periodontium (19, 29). Thus, these observations have clinical significance and suggest that the modulation of PDL cell functions by IL-1 may occur in vivo during inflammation, thereby affecting the structural integrity and attachment of this tissue to tooth and bone surfaces.

rHuIL-1-mediated down-regulation of the osteoblast-like phenotype has also been observed in bone cells (15, 16); however, the significance of this change to the host has not been addressed. Since osteoblasts have been shown to synthesize chemoattractant protein-1 (MCP-1) in response to IL-1 (22, 34), we speculated that IL-1 may also regulate PDL cell functions that may include the shift to a cell phenotype which is advantageous to the host immune response at the nidus of infection. To this end, we first determined if PDL cells could participate directly in the immune response through the action of proinflammatory cytokines elaborated in response to rHuIL-1. Our experiments demonstrate that rHuIL-1 is a potent inducer of mRNA and protein synthesis for IL-6, IL-8, TNF-, and IL-1 in PDL cells. Nevertheless, the expression of these proinflammatory cytokines in PDL cells in response to rHuIL-1 is short lived, and their expression is not observed when rHuIL-1 is present for a longer period of time. This autocrine regulation of IL-1, similar to that in macrophages and gingival fibroblasts (1, 9), may be important in the augmentation of an inflammatory cascade in neighboring PDL cells during infection.

It has been shown previously that gingival fibroblasts function as accessory immune cells and elicit proinflammatory cytokines in response to LPS (1, 26). Therefore, we examined the ability of PDL cells to elicit proinflammatory cytokines in response to LPS. Our experiments demonstrate that, unlike gingival fibroblasts, PDL cells are not constitutively LPS responsive, since LPS does not induce expression of proinflammatory cytokines. However, in the continuous presence of rHuIL-1 PDL cells undergo phenotypic changes that render these cells LPS responsive. Hence, the down-regulation of the osteoblast-like phenotype in PDL cells by rHuIL-1 is concomitant with the acquisition of LPS responsiveness, indicating a crucial role for IL-1 in the regulation of PDL cell functions and the maintenance of homeostasis in the PDL. Furthermore, rHuIL-1-mediated modulation of the PDL cell phenotype is transient, in that the removal of rHuIL-1 from the PDL cell cultures results in reacquisition of the osteoblast-like phenotype and functions. The transience of the rHuIL-1-induced LPS-responsive phenotype also suggests that osteoblast-like characteristics are only suppressed by IL-1 and are not permanently altered. Interestingly, PDL cells express only one set of phenotypic characteristics at a time, indicating that rHuIL-1 regulates the phenotype as a whole. However, during expression of the osteoblast-like phenotype and the LPS-responsive phenotype, PDL cells remain IL-1 responsive. Thus, IL-1 responsiveness is not a part of the phenotypic alteration that occurs during inflammation. Confirmation of these results in two cell lines, each derived from a different individual, suggests that rHuIL-1-dependent modulation of phenotypes may be a general characteristic of PDL cells. TNF- has been shown to affect PDL cell phenotype in a similar manner (24), indicating that proinflammatory cytokines play a pivotal role in the regulation of periodontal homeostasis during inflammation.

The mechanisms of the acquisition of LPS responsiveness are not yet clear. It is possible that IL-1 induces LPS responsiveness in PDL cells through induction of LPS receptors. Alternatively, the regulation of responsiveness to LPS may arise through a transcription factor such as NF-B, known to regulate LPS-receptor interactions. However, the consequences of these phenotypic changes do not appear to be associated with corresponding changes in PDL cell morphology.

Taken together, our results demonstrate that IL-1 may provide the mechanism by which PDL cells can participate in the disease process. Since IL-1 is an immunoregulatory cytokine, it is not surprising that its effects on the PDL cells favor host defense rather than preserving the structural integrity of the periodontium during infection. Based on present observations and the levels of IL-1 shown to be present in the inflamed gingiva, it is tempting to speculate that, during an infection, locally produced IL-1 may be sufficient to induce an LPS-responsive phenotype in PDL cells. These cells can then respond to a bacterial insult, thereby exerting cytotoxicity to pathogens through the recruitment of immune cells into the infected area of periodontium (3). Although this process assists in the elimination of infection, it occurs at the expense of the integrity of the PDL. This point is illustrated by the observation that, in the presence of IL-1 in vitro, PDL cells do not synthesize TGF-1 or proteins associated with mineralization, such as AP and osteocalcin. This loss of osteoblast-like properties of PDL cells at the inflamed site may result in the failure of these cells to maintain or restore attachment. However, following resolution of the infection, they revert to their osteoblast-like phenotype and may be able to take part in the process of regeneration. It is conceivable that, similar to PDL cells, osteoblasts may also undergo functional changes leading to bone degradation at an inflammatory site. Since IL-1 has been shown to down-regulate osteoblastic phenotype in bone cells, it is possible that bone degradation at an inflammatory site may not be due only to the activity of osteoclasts but may also be due to an alteration in the phenotype of osteoblasts to that of LPS-responsive cells. Nevertheless, data presented here suggest that cytokines serve as important regulators of disparate cellular functions during tissue homeostasis as well as during inflammation.