Interleukin-17: A New Bone Acting Cytokine In Vitro†
Rutger L. Van Bezooijen,
Department of Endocrinology and Metabolic Diseases, Leiden University Medical Center, Leiden, The Netherlands
Address reprint requests to: Rutger L. van Bezooijen, Department of Endocrinology and Metabolic Diseases, Building 1, C4-R89, Leiden University Medical Center, P.O. Box 9600, 2300 RC Leiden, The Netherlands
Presented in part at the 19th Annual Meeting of the American Society for Bone and Mineral Research, September 10–14, 1997, Cincinnati, OH, U.S.A. (J Bone Miner Res 12 (Suppl 1):S342) and the Second Joint Meeting of the American Society for Bone and Mineral Research and the International Bone and Mineral Society, December 1–6, 1998, San Francisco, CA, U.S.A. (Bone 23 (Suppl):S441).
Interleukin-17 (IL-17) is a recently cloned cytokine that is exclusively produced by activated T cells, but its receptor has been found on several cells and tissues. Like other proinflammatory cytokines produced by activated T cells, IL-17 may affect osteoclastic resorption and thereby mediate bone destruction accompanying some inflammatory diseases. In the present study, we investigated whether osteogenic cells possess the receptor for IL-17 (IL-17R) and whether IL-17 affects osteoclastic resorption. We found that IL-17R mRNA is expressed both in mouse MC3T3-E1 osteoblastic cells and fetal mouse long bones, suggesting that osteogenic cells may be responsive to IL-17. In fetal mouse long bones, IL-17 had no effect on basal and IL-1β–stimulated osteoclastic bone resorption, but when given together with tumor necrosis factor-α (TNF-α) it increased bone resorption dose dependently in serum-free conditions. In addition, IL-17 increased TNF-α–induced IL-1α, IL-1β, and IL-6 mRNA expression in fetal mouse metatarsals and IL-1α and IL-6 mRNA expression in MC3T3-E1 cells. In conclusion, IL-17R mRNA was expressed by mouse osteoblastic cells and fetal mouse long bones, and IL-17 in combination with TNF-α, but not IL-1β, increased osteoclastic resorption in vitro. IL-17 may therefore affect bone metabolism in pathological conditions characterized by the presence of activated T cells and TNF-α production such as rheumatoid arthritis and loosening of bone implants.
PROINFLAMMATORY CYTOKINES play an important role in the modulation of bone metabolism.1–5 Recently, a new cytokine, interleukin-17 (IL-17), was cloned from a CD4+ T cell library.6 This human IL-17 exhibits 63% amino acid identity with the previously reported cDNA encoding for murine cytotoxic T-lymphocyte associated antigen-8 (CTLA-8),7 which was later on found to be the rat homolog of IL-17.8 CTLA-8 exhibits 57% amino acid identity with the open reading frame 13 (ORF13) of herpesvirus saimiri, a T-lymphotropic virus.7 IL-17 therefore seems to be another virus-captured gene incorporated into cellular DNA.9 The murine homolog of IL-178 shares 62.5% amino acid sequence identity with human IL-17.6,10
Production of IL-17 is restricted to activated human memory or mouse αβTCR+CD4−CD8− T cells.6,8,10–12 In contrast to the restricted set of IL-17–producing cells, a wide variety of cells express mRNA encoding for the IL-17 receptor (IL-17R), which is unrelated to previously identified cytokine receptor families.13,14 The amino acid sequences for murine and human IL-17R share 69% identity.
In inflammatory disease characterized by the presence of activated T cells, such as rheumatoid arthritis, bone resorption can be increased.15 Furthermore, loosening and failure of artificial bone implants is related to an inflammatory reaction in which activated T cells are involved.16–22 Interestingly, IL-17 has been found to induce the expression of potent modulators of bone resorption as NF-κB, IL-1, IL-6, tumor necrosis factor-α (TNF-α), leukemia inhibitory factor, granulocyte-macrophage colony stimulating factor (GM-CSF), granulocyte-colony stimulating factor (G-CSF), monocyte chemotactic protein-1 (MCP-1), and prostaglandin E2, and nitric oxide in fibroblasts, stromal, epithelial, endothelial cells, macrophages, keratinocytes, articular chondrocytes, and osteoarthritis cartilage.6,8,10,13–14,23–29 We therefore examined whether IL-17 can have an effect on bone resorption. For this, we tested IL-17 alone or in combination with other cytokines produced by activated T cells and known to affect bone resorption, for example TNF-α and IL-1β, in fetal mouse bone explants in vitro. In addition, because osteoblasts are important regulatory cells in bone metabolism, we examined whether these cells expressed mRNA encoding for IL-17R. At present, it is unknown whether any of the bone cells possess the IL-17R and respond to IL-17.
MATERIALS AND METHODS
Alpha modified essential medium (α-MEM) was purchased from GIBCO BRL (Breda, The Netherlands), penicillin, and streptomycin were from Flow Laboratories (Amstelstad, Zwanenburg, The Netherlands), fetal calf serum (FCS) was from Integro (Zaandam, The Netherlands), and bovine serum albumin (BSA) was from ICN Biomedicals, Inc. (Asse-Relegem, Belgium). Recombinant human IL-17 was from R&D Systems Europe, Ltd. (Abingdon, U.K.). Human IL-17, similar to mouse IL-17, induced IL-6 secretion in mouse bone marrow cells.8,30,31 It, however, was about 6-fold less effective. Recombinant human IL-1β was a generous gift of Dr. M. Ponec (Department of Dermatology, Leiden University Medical Center, The Netherlands) and recombinant murine TNF-α of Dr. A. van de Voorde (Innogenetics, Antwerpen, Belgium). The restriction enzyme Pst1 was purchased from GIBCO BRL, parathyroid hormone (PTH)–related hormone from Bacham AG (Budendorf, Switzerland), and indomethacin from ICN Biomedicals, Inc. (Aurora, OH, U.S.A.).
The synthetic internal standard pMUS was a generous gift of Dr. D. Shire (Sanofi Recherche, Labège, France).32 Specific sense and antisense primers were purchased from Isogen Bioscience BV (Maarssen, The Netherlands). If possible, the primer sets used crossed intron/exon boundaries with such large introns that eventual contaminations with genomic DNA will not be amplified in the amplification process. The primer sets used were as follows: β2-microglobulin sense, 5′-tgaccggcttgtatgctatc-3′, nucleotides 64–83, antisense, 5′-cagtgtgagccaggatatag-3′, nucleotides 3159–317833; IL-17R sense, 5′-ccactctgtagcaccccaat-3′, nucleotides 2302–2321, antisense, 5′-gagagactcaacgggctcac-3′, nucleotides 2586–2605 (cDNA sequence)13; IL-1α sense, 5′-cagttctgccattgaccatc-3′, nucleotides 123–142, antisense, 5′-tctcactgaaactcagccgt-3′, nucleotides 321–340 (cDNA sequence)34; IL-1β sense, 5′-ttgacggaccccaaaagatg-3′, nucleotides 1915–34, antisense, 5′-agaaggtgctcatgtcctca-3′, nucleotides 3639–365835; IL-6 sense, 5′-gttctctgggaaatcgtgga-3′, nucleotides 1661–1680, antisense,5′-tgtactccaggtagctatgg-3′, nucleotides 6083–6102.36 Deoxynucleoside triphosphates (dNTPs) were obtained from Pharmacia Biotech (Roosendaal, The Netherlands), random hexanucleotide primers and RNasin from Promega (Leiden, The Netherlands), Goldstar DNA polymerase from Eurogentec (Seraing, Belgium), and Moloney murine leukemia virus reverse transcriptase (M-MLV) and the 100 bp DNA ladder from GIBCO BRL (Paisley, U.K.).
Murine osteoblastic MC3T3-E1 cells were seeded at a density of 50,000 cells/1.8 cm2 and cultured until confluence (4 days) in α-MEM (500 μl/1.8 cm2 well) (Greiner, Alphen a/d Rijn, The Netherlands) supplemented with 10% heat-inactivated FCS. At confluence, cells were stimulated with IL-17, TNF-α, or the combination of these two cytokines at the concentrations and for the time periods mentioned in the Results section.
Bone resorption assay
Pregnant Swiss Albino mice were injected with 30 μCi45calcium (45Ca) (1 Ci/mmol; Amersham, Aylesbury, Buckinghamshire, U.K.) on day 16 or 18 of gestation and sacrificed at day 17 or 19 of gestation, respectively. The45Ca prelabeled fetal metatarsals were aseptically excised in phosphate buffered 0.9% NaCl containing 10% heat-inactivated FCS. The metatarsals were precultured in 1 ml of α-MEM supplemented with either 10% heat-inactivated FCS or 0.1% BSA in 6-well culture plates (Costar, Cambridge, MA, U.S.A.) for 24 h to allow45Ca exchange with the culture medium. Thereafter, metatarsals were cultured in 250 μl of α-MEM supplemented with either 10% heat-inactivated FCS or 0.1% BSA in 24-well culture plates. At the end of the culture time, residual calcium was extracted from the metatarsals in 0.5 ml of 5% trichloracetic acid for 24 h. The amount of45Ca in the culture media and in the decalcification fluid was determined by liquid scintillation in a β-counter (1600 TR; Packard, Groningen, The Netherlands). Resorption was expressed as percentage of45Ca that was released from the prelabeled metatarsals during incubation (%45Ca-release). The values were corrected for physicochemical calcium exchange which is the amount of45Ca released into the medium that is not due to osteoclastic resorption (%kCo). This is determined by the release from bones which were killed by three cycles of freeze-thawing. This animal protocol was approved by the Leiden University Committee for Animal Experiments (UDEC).
Osteoclastic resorption is thus calculated as:
Isolation of total cytoplasmic RNA
RNA isolation was carried out according to the method developed by Chomczynski and Sacchi.37 All solutions and glassware were autoclaved. Briefly, MC3T3-E1 cells or 19-day-old fetal metatarsals were homogenized in 4 M guanidinium isothiocyanate lysis buffer, extracted with phenol and chloroform, precipitated at −20°C with 100% iso-propanol, resuspended in autoclaved denaturated water, and stored at −80°C. RNA concentrations were determined spectrophotometrically assuming 40 μg/ml/optical density at a wavelength of 260 nm (1 cm path-length).
Denatured RNA (1 μg; 5 minutes at 70°C and quick chilled on ice) was reverse transcribed into cDNA in a 50-μl reaction volume containing first-strand buffer (75 mM KCl, 3 mM MgCl2, 50 mM Tris-HCl, pH 8.3), 10 mM DTT, 0.5 mM dNTPs, 200 ng of random hexanucleotide primers, 1 U of RNasin/μl, and 2.5 U of M-MLV/μl. Reverse transcription was performed at 37°C for 90 minutes, 85°C for 3 minutes, followed by quick chilling on ice. Obtained cDNA was diluted to a theoretical (assuming 100% efficiency of reverse transcription) concentration of 10 ng/μl in autoclaved denaturated water and stored at −20°C until subsequent amplification.
Competitive polymerase chain reaction
Competitive polymerase chain reaction (PCR) was performed as previously described.38 To correct for any variation in RNA content and cDNA synthesis between the different preparations, samples were equalized on basis of their β2-microglobulin content. Five nanograms of cDNA was coamplified over 33 cycles in the presence of 4-fold serial dilutions of internal standard pMUS. Competitive PCR was performed in a 25 μl reaction volume containing reaction buffer (75 mM Tris-HCl, pH 9.0), 20 mM (NH4)2SO4, 0.01% (w/v) Tween-20, 1.5 mM MgCl2, 200 μM dNTPs, 0.25 μM sense and antisense primer, 0.125 U of Goldstar DNA polymerase, and cDNA and internal standard pMUS to be coamplified. Negative controls in which cDNA or both cDNA and internal standard were omitted were run in parallel in each experiment and found to be negative. Mixtures were overlayed with light white mineral oil (Sigma, Zwijndrecht, The Netherlands) to reduce evaporation. After one cycle of 30 s at 94°C, cDNA and internal standard were coamplified by repeated cycles of which one cycle consisted of a 30 s denaturating step at 94°C, a 30 s annealing step at 56°C, and a 1 minute primer extension step at 72°C on a Hybaid Omnigene thermal cycler (Biozym, Landgraaf, The Netherlands). This was followed by one cycle of 2 minutes at 72°C. Aliquots of 20 μl of each amplified sample were subjected to electrophoresis on a 1% agarose gel containing 0.5 μg of EtBr/ml and photographed. The intensity of each band was measured using computerized densitometry. The intensity of the cDNA and internal standard pMUS amplicons could be measured separately on gel due to their difference in size. The internal standard amplicons had a size of 300 bp, while that of the cDNA amplicon was 222 bp. The internal standard to be coamplified contained the same specific nucleotide sequence for β2-microglobulin and could therefore be amplified with the corresponding primer set. The use of one primer set allowed equal amplification efficiency of the cDNA and internal standard pMUS. The cDNA/internal standard ratio therefore remained constant throughout the amplification process even into the plateau phase. This implies that in the case of a cDNA/internal standard ratio of 1, the initial amount of cDNA was equal to the initial amount of the internal standard. Coamplification of cDNA with serial dilutions of the internal standard and plotting the cDNA/internal standard ratio against the number of copies of internal standard allowed us to determine the number of copies of internal standard with which the cDNA sample could compete. The point at which the cDNA/internal standard ratio is equal to 1 is independent of the linearity of ethidium bromide staining because loss of linearity will flatten the slope of the ratio between the intensities of the amplicons but will not affect the point at which the intensities of the amplicons are equal to each other.
Cytokine and IL-17R mRNA expression were analyzed using semiquantitative PCR. The amplification process was performed as with competitive PCR, except that no internal standard was added and 20 ng instead of 5 ng of cDNA was amplified.
Experiments were performed in quintuplicate except for the experiment specified in the legend of Table 1. Values are expressed as mean ± SEM Statistical differences between values were examined by one-way analysis of variance followed by Fischer's projected least signifcant difference test.
Table Table 1.. Effect of IL-17 in Combination with TNF-α or IL-1β on Osteoclastic Bone Resorption in Fetal Mouse Metatarsals in Media Supplemented with 0.1% BSA
IL-17R mRNA expression by MC3T3-E1 osteoblastic cells and fetal mouse long bones
Total RNA was isolated from mouse MC3T3-E1 osteoblastic cells that were cultured until confluence (4 days) and from whole 19-day-old fetal mouse metatarsal bones. Using RT-PCR, an IL-17R amplicon with the predicted size of 285 bp was found (Fig. 1). This suggests that the receptor for IL-17 was present in both MC3T3-E1 cells and fetal mouse long bones. Cleavage of the obtained IL-17R amplicons with restriction enzyme Pst1 gave rise to two smaller fragments of 89 and 196 bp. This demonstrated that the obtained IL-17R amplicons contained the specific restriction site for Pst1, i.e., ctgca/g, which is present at position 2390 of the mouse IL-17R nucleotide sequence.13 A strong IL-17R mRNA signal was already observed after 30 cycles (Fig. 2). To visualize the small 89 bp fragment, however, the amplicons used to analyze the presence of the restriction site for Pst1 were amplified over 35 cycles.
Effect of IL-17 on osteoclastic bone resorption in fetal mouse long bones
Fetal mouse long bones, in which osteoclastic bone resorption was measured by the release of45Ca, were cultured in media supplemented with either 10% heat-inactivated FCS or 0.1% BSA. In the presence of 10% heat-inactivated FCS, osteoclastic resorption is more susceptible to inhibition, whereas in the presence of 0.1% BSA it is more susceptible to stimulation. PTH–related protein (1 U/ml) elevated osteoclastic resorption by only 142.0 ± 8.3% in cultures supplemented with 10% heat-inactivated FCS, while it increased resorption by 606.3 ± 13.1% (p < 0.0001) in long bones cultured in media supplemented with 0.1% BSA. However, TNF-α (100 ng/ml) inhibited (p < 0.01) osteoclastic resorption by 66.7% in cultures supplemented with 10% FCS (Fig. 3A), while it had no effect on resorption in cultures supplemented with 0.1% BSA (Fig. 3B).
IL-17, up to a concentration of 100 ng/ml, had no effect on osteoclastic resorption in 19-day-old fetal mouse metatarsals under both culture conditions tested (Figs. 3A and 3B). In cultures supplemented with 10% FCS, however, IL-17 (100 ng/ml) prevented the inhibitory effect of TNF-α (Fig. 3A). This prevention of TNF-α inhibited osteoclastic resorption by IL-17 was dose dependent; inhibition of45Ca-release by 10 ng and 100 ng TNF-α/ml was prevented by 1 ng and 10 ng of IL-17/ml, respectively (Table 2). In media supplemented with 0.1% BSA, i.e., culture conditions whereby TNF-α had no effect on osteoclastic resorption, addition of IL-17 (100 ng/ml) to TNF-α–treated cultures increased osteoclastic resorption significantly (p < 0.0001) at doses of 10 ng and 100 ng of TNF-α/ml (Fig. 3). This increase was also dose dependent with regard to IL-17;45Ca release was increased by 1 ng and 0.1 ng of IL-17/ml in combination with 10 ng and 100 ng of TNF-α/ml, respectively (Table 1).
Table Table 2.. Effect of IL-17 in Combination with TNF-α or IL-1β on Osteoclastic Bone Resorption in Fetal Mouse Metatarsals in Media Supplemented with 10% Heat-Inactivated FCS
In fetal mouse metatarsals, IL-1β had a biphasic stimulating effect on osteoclastic bone resorption in cultures supplemented with 0.1% BSA (Table 1), but not in cultures supplemented with 10% FCS (Table 2). At concentrations of 1 ng/ml and 10 ng/ml, IL-1β increased osteoclastic resorption significantly (p < 0.001 and p < 0.0001, respectively), while 100 ng of IL-1β/ml increased it but not significantly. Addition of IL-17 (100 ng/ml) to IL-1β–treated cultures had no effect on osteoclastic resorption (Tables 1 and 2). In cultures supplemented with 0.1% BSA, IL-17 (100 ng/ml) prevented the biphasic effect of the high IL-1β concentration: osteoclastic resorption was increased significantly (p < 0.001) by the cytokine combination IL-1β (100 ng/ml) + IL-17 (100 ng/ml).
Identical results were obtained with respect to the effects of the cytokines TNF-α and IL-17 and their combination on osteoclastic resorption in 17-day-old metatarsals, an in vitro model in which osteoclastic bone resorption depends on the recruitment and activity of osteoclasts (data not shown).39,40
Prostaglandins probably did not mediate the effect of the cytokine combination TNF-α (100 ng/ml) + IL-17 (100 ng/ml) on osteoclastic bone resorption. Indomethacin at concentrations of 1 μM and 10 μM had no effect on45Ca-release from 19-day-old metatarsals cultured in the presence of 10% FCS (control, 47.8 ± 6.0; indomethacin (1 μM), 51.4 ± 10.5; indomethacin (10 μM), 53.0 ± 6.3), nor could it prevent the abolishment of TNF-α inhibited resorption by IL-17 (TNF-α, 26.2 ± 1.2; TNF-α + IL-17, 48.1 ± 11.5; TNF-α + IL-17 + indomethacin (1 μM), 53.0 ± 10.6, TNF-α + IL-17 + indomethacin (10 μM), 47.6 ± 11.5).
Effects of IL-17, alone or in combination with TNF-α, on cytokine mRNA expression
The stimulatory effect of the cytokine combination TNF-α + IL-17 on osteoclastic resorption may have been mediated by the induction of bone resorption stimulating cytokines. To study this, mRNA steady-state levels of the cytokines IL-1α, IL-1β, and IL-6 were determined in the presence or absence of IL-17 (100 ng/ml), TNF-α (100 ng/ml), or the combination of both cytokines in 19-day-old fetal mouse metatarsals cultured in the presence of 10% FCS or 0.1% BSA and in mouse osteoblastic MC3T3-E1 cells cultured until confluence (4 days). Total RNA was isolated at several time points after stimulation and reverse transcribed into cDNA. All samples, fetal metatarsals cultured in 10% FCS (Fig. 4A) or 0.1% BSA, and MC3T3-E1 cells (data not shown), expressed equal amounts of β2-microglobulin mRNA as determined by competitive PCR using the synthetic internal standard pMUS.
Using semiquantitative RT-PCR, IL-17R mRNA expression was found not to be affected by the cytokine combination TNF-α + IL-17 in fetal mouse bone explants (data not shown). Stimulation with IL-17 did not induce IL-1α and IL-1β mRNA expression nor did it increase IL-6 mRNA expression (Figs. 4B and 4C). TNF-α induced the expression of IL-1α and IL-1β and elevated the expression of IL-6 mRNA. Simultaneous treatment with TNF-α and IL-17 strongly increased TNF-α–induced mRNA levels of all three cytokines, i.e., IL-1α, IL-1β, and IL-6. Whereas increased IL-1β and IL-6 mRNA levels were found 4 h after stimulation with the cytokine combination TNF-α + IL-17, IL-1α mRNA expression was not found until 24 h after stimulation in fetal mouse long bones cultured in the presence of 10% FCS (data not shown). The mRNA levels of all three cytokines remained elevated at least 72 h after stimulation.
Similarly, IL-17 did not induce IL-1α and IL-1β mRNA expression in MC3T3-E1 cells, but slightly elevated IL-6 mRNA expression (Fig. 2). TNF-α induced IL-1α mRNA expression at 1 h and 72 h after stimulation and increased IL-6 mRNA expression during the whole culture period. Simultaneous treatment with TNF-α and IL-17 strongly increased TNF-α–induced mRNA levels of IL-1α and IL-6. IL-6 mRNA expression was increased 1 h after stimulation with the cytokine combination TNF-α + IL-17 and remained constant thereafter, whereas IL-1α mRNA expression gradually increased until at least 72 h after stimulation. The expression of IL-1β mRNA was not induced by TNF-α nor the cytokine combination (data not shown). IL-17R mRNA expression was found not to be affected by IL-17, but to be down-regulated by TNF-α after 72 h of stimulation (Fig. 2). IL-17 accelerated this TNF-α induced down-regulation of IL-17R mRNA to 24 h after stimulation.
In the present study, we identified for the first time osteogenic cells as target cells for IL-17. Mouse MC3T3-E1 osteoblastic cells and fetal mouse long bones expressed IL-17R mRNA and stimulated IL-6 mRNA levels in MC3T3-E1 cells, adding osteoblasts to the list of cells in which production of IL-6 is stimulated by IL-17. These include fibroblasts, stromal, epithelial, endothelial cells, macrophages, keratinocytes, and articular chondrocytes.6,8,10,13–14,23–26,28,29 Furthermore, in both osteoblast-like cells and fetal mouse bone explants, IL-17 increased the TNF-α–induced expression of IL-1α and IL-6 mRNA. Similarly, IL-17 was recently reported to synergize with TNF-α, CD40 ligand, IL-1, and Interferon-γ in rheumatoid synoviocytes, kidney epithelial cells, synovial fibroblasts, and keratinocytes, respectively. In rheumatoid synoviocytes, IL-17 synergized with TNF-α for the induction of GM-CSF.10 In kidney epithelial cells, it synergized with CD40-ligand for the induction of IL-6, IL-8, RANTES (regulated on activation normal T cell-expressed and secreted), and MCP-1.12 A synergism between IL-17 and IL-1 was found for IL-6 production by synovial fibroblasts25 and between IL-17 and interferon-γ for IL-6 and IL-8 production in keratinocytes.29
Activated T cells produce multiple cytokines, including TNF-α, and are the exclusive source of IL-17.6,10 It may therefore be that IL-17 exerts its effect in concert with such cytokines. This may further depend on the target cells. We show here, for example, that IL-17 stimulated IL-1α and IL-1β mRNA in bone only when given together with TNF-α, while it has been recently reported that IL-17 alone can stimulate IL-1α and IL-1β mRNA in macrophages and articular chondrocytes.16,28 We also found differences in IL-6 mRNA expression in response to IL-17 and TNF-α between osteoblast-like cells and fetal bone explants.
Pathological conditions characterized by the presence of activated T cells, such as rheumatoid arthritis, can be associated with increased osteoclastic bone resorption.15 This is thought to be mediated by proinflammatory cytokines released by activated T cells, such as TNF-α and IL-1β.41–44 Because IL-17R mRNA was found in bone cells and IL-17 together with TNF-α was shown here to increase the mRNA expression of other proinflammatory cytokines, we further examined whether it can also affect bone resorption. When given alone, IL-17 had no effect on osteoclastic resorption in fetal mouse bone explants and did not affect the IL-1β–stimulated resorption. However, when given together with TNF-α, it stimulated osteoclastic resorption significantly. This was demonstrated under two different experimental conditions. When tested in bones cultured in the absence of serum,45Ca-release was stimulated by 400–600% while the two cytokines given alone had no significant effect on resorption. In bones cultured in the presence of 10% serum, IL-17 abolished the inhibitory effect of TNF-α on osteoclastic resorption. It may thus be that IL-17 together with TNF-α are at least partly responsible for loosening and failure of artificial bone implants and the development of bone changes occurring in diseases such as rheumatoid arthritis. Interestingly, supernatants collected from cultured rheumatoid arthritis synovium pieces that were incubated with blocking antibodies to IL-17 gave about 50% lower induction of IL-6 and leukemia inhibitory factor production by cultured synoviocytes than nontreated supernatants.25 This suggested that the rheumatoid arthritis synovium pieces contained IL-17 producing T cell infiltrates. The mechanism responsible for the induction of osteoclastic resorption by the combination of IL-17 and TNF-α is not immediately obvious, but it may be mediated either directly or by the induction of other cytokines. We have previously shown that IL-6 stimulates the development but not the activity of osteoclasts in mouse fetal long bones,39 while in the present study the combination of IL-17 and TNF-α stimulated existing osteoclasts making IL-6 an unlikely mediator of the effect of IL-17 and TNF-α on osteoclastic resorption. In addition, in preliminary studies with neutralizing antibodies against IL-1α and IL-1β, we found that IL-1β did not contribute while IL-1α contributed only partly to IL-17 and TNF-α stimulation of osteoclastic resorption.45 Thus, IL-1 seems also not to be the mediator of the effect.
Previous studies have shown that TNF-α increases IL-1β and IL-6 mRNA expression in mouse MC3T3-E1 and human osteoblast-like cells46–48 and stimulates osteoclastic bone resorption in vitro.49–54 The last effect, however, appears to depend on culture conditions and dose of TNF-α used. For example, we have previously reported that in cultures of 17-day-old mouse radii in which resorption depends mainly on the activity of mature osteoclasts, TNF-α had no effect on osteoclastic resorption.40 In contrast in 17-day-old metatarsals cultured in BGJ medium in which resorption depends on both the recruitment and activity of osteoclasts, TNF-α had a biphasic effect stimulating resorption at low concentrations and inhibiting it at high concentrations. In addition, TNF-α has been shown to stimulate osteoclastic resorption in fetal rat long bones cultured in BGJ medium.49,50 In cultures of calvarial cells,52 neonatal mouse calvariae,51 and human bone marrow stromal cells sedimented onto devitalized bone slices,54 TNF-α stimulated osteoclastic resorption by a prostaglandin-mediated mechanism. In contrast, Lerner and Ohlin53 reported that TNF-α and TNF-β can stimulate bone resorption in mouse calvariae by a prostaglandin-independent mechanism. In the present study, using fetal mouse long bones representing different stages of osteoclast development, we found that TNF-α had either no effect or suppressed osteoclastic resorption. A small but not significant increase in resorption was found by 0.1 ng and 1 ng of TNF-α in 17-day-old and by 1 ng of TNF-α in 19-day-old metatarsals cultured in media supplemented with 10% FCS. A possible explanation for the observed differences may be the use of different TNF-α preparations. The source of TNF-α was mouse (m) in the present study versus human (h) in the other studies. mTNF-α and hTNF-α may have different effects on osteoclastic resorption in these fetal mouse metatarsals as mTNF-α binds to both mTNF-receptor 1 (TNF-R1) and 2 (TNF-R2), while hTNF-α binds to mTNF-R1 only.55 This may be significant because the two receptors have been reported to interact with different cellular proteins and may therefore utilize different signaling pathways.56–59 Differences between culture media may also be responsible for the different effects reported for TNF-α on osteoclastic resorption. The rate of osteoclastic resorption (46.8 ± 6.2 and 59.0 ± 9.5%), for example, is much higher in metatarsals cultured in α-MEM supplemented with 10% FCS in the present study than in metatarsals cultured in BGJ medium supplemented with 10% FCS previously reported (12.5 ± 3.5%).40 The lack of a stimulatory effect of TNF-α on osteoclastic resorption in metatarsals cultured in α-MEM supplemented with 10% FCS may therefore be due to the inability of TNF-α to further increase the already stimulated osteoclastic activity. Removal of FCS from the α-MEM medium decreased basal osteoclastic resorption dramatically, as shown here. In addition, this may have removed essential cofactors needed for the stimulatory effect of TNF-α on osteoclastic resorption. Nevertheless, the serum-free culture system provides the most appropriate conditions to study stimulation of osteoclastic resorption, because PTH stimulated resorption by only 1.4-fold in cultures supplemented with FCS and by 6.1-fold in serum-free cultures. This is the reason why most of the fetal long bone resorption cultures are performed in the absence of serum as originally described by Raisz et al.60,61
The combination of TNF-α with IL-17 affected osteoclastic resorption similarly in 17-day-old and in 19-day-old metatarsals. In 17-day-old metatarsals, osteoclastic bone resorption depends on the recruitment and activity of osteoclasts, while resorption depends mainly on the activity of mature osteoclasts in 19-day-old metatarsals. These findings suggest that the cytokine combination stimulated the activity of mature osteoclasts and/or the survival of mature osteoclasts. The effect of TNF-α + IL-17 on osteoclastic resorption was not prevented by the addition of indomethacin, suggesting that prostaglandin production was not involved. Probably prostaglandins do not mediate osteoclastic resorption in this in vitro bone resorption model of fetal mouse long bones.62,63
In conclusion, mouse osteoblast-like cells and fetal bone explants express IL-17R mRNA and increase IL-1α, IL-1β, and IL-6 mRNA expression after stimulation with IL-17 in combination with TNF-α. The exclusive production of IL-17 by activated T cells and the present observation that IL-17 increased osteoclastic bone resorption in combination with TNF-α only, restricts the action of IL-17 on bone resorption to pathological conditions characterized by the presence of activated T cells and TNF-α production.
This work was supported by a grant from N.V. Organon, Oss, The Netherlands.