INTRODUCTION

Interleukin-8 (IL-8),¹ a chemokine and also referred to as neutrophil activating peptide-1 and monocyte-derived neutrophil chemotactic factor, is synthesized as a 99-amino acid precursor, secreted after cleavage of a signal sequence of 20 residues, and processed by repeated N-terminal cleavage yielding several biologically active variants (1, 2). The major form consists of 72 amino acid residues with a molecular mass of 8,383 daltons, pI 8.3, and four cysteines that form two disulfide bridges. In addition to well established pleiotropic biological effects of IL-8, which include neutrophil activation, chemotaxis, cell shape change, exocytosis of secretory vesicles and azurophil granules, expression of surface adhesion molecules, production of superoxide and hydrogen peroxide reactive oxygen metabolites (1, 2), and the release of cell matrix resorbing enzymes gelatinase and elastase (2, 3), recent studies are consistent with the hypothesis that IL-8 may prove pivotal in the control of osteoclastogenesis (4).

Bone marrow stromal cells have long been considered as the source of osteoprogenitor cells (5, 6). The renewal of the osteoblast population at the bone surface is considered to occur via differentiation of osteoprogenitor cells along the osteoblast lineage (7, 8). Results from our laboratory have also shown that glucocorticoids induce differentiation of human bone marrow stromal (HBMS) cells into cells that display an osteoblastic phenotype, i.e., increased alkaline phosphatase activity, cAMP production in response to parathyroid hormone (PTH), osteocalcin production in response to 1,25[OH]₂D₃, and mineralization (9). These observations along with others support a role of glucocorticoids in the differentiation of bone marrow osteoprogenitor (10, 11, 12, 13, 14) and human bone-derived (15) cells into osteoblast-like cells. There is also sufficient evidence indicating that two cytokines, interleukin-1 and tumor necrosis factor- (TNF-), play a central role in bone metabolism and regulate osteoblast and osteoclast function (16, 17, 18, 19, 20, 21, 22). Furthermore, IL-1 (23), TNF- (24), granulocyte-macrophage CSF (25), and IL-6 (25, 26) are synthesized by normal human osteoblast-like cells. We recently showed that IL-8 is produced by normal human osteoblast-like (25, 27), HBMS, and human osteosarcoma MG-63 cells in response to IL-1 and TNF- and that dexamethasone inhibited the secretion of IL-8, whereas 17-estradiol had no effect (27). In fact, the magnitude of IL-8 stimulation by IL-1 and TNF- was much greater than the stimulated increments in IL-6 and granulocyte-macrophage CSF (25), suggesting that IL-8 may play a role in the regulation of bone cell function. There is also recent evidence indicating that IL-8 regulates some of the osteogenic functions of osteoclasts and osteoblasts (28, 29). Because molecular mechanisms of IL-1 and IL-8 action, their signal transduction pathways, and regulation of IL-8 gene expression by IL-1 and osteotropic hormones in osteoblastic cells are still virtually unknown, we designed experiments to investigate the regulation of the IL-8 gene in HBMS cells.

EXPERIMENTAL PROCEDURES

Hydrocortisone, dexamethasone, genistein, curcumin, H-7, Denhardt's reagent, salmon sperm DNA, Dulbecco's phosphate-buffered saline, crude bacterial collagenase, trypsin-EDTA, Histopaque-1077, bovine PTH (1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34), fetal bovine serum, and Dulbecco's modified Eagle's medium/Ham's F-12 medium (1:1) were obtained from Sigma. Concentrated stock solutions of hydrocortisone, curcumin, genistein, and dexamethasone were made in absolute ethanol, and appropriate amounts were added directly to the medium and thoroughly mixed. Proteinase K and staurosporine were purchased from Boehringer Mannheim GmbH and Calbiochem-Novabiochem International (San Diego, CA), respectively. Recombinant human IL-1 and human IL-8 cytokine enzyme-linked immunosorbent assay kit were obtained from R & D Systems, Inc. (Minneapolis, MN). Concentrated stock solution of IL-1 was made in phosphate-buffered saline, and required amounts of IL-1 were added directly to the medium.

Human ribs obtained from surgery patients were transported to the laboratory in tissue culture flasks containing Dulbecco's modified Eagle's medium/Ham's F-12 medium and were processed immediately or after storage overnight in the refrigerator. The bone marrow stromal cells were isolated as described previously (9). Briefly, the ribs were cleaned of cartilage and muscle and cracked open with a bone cutter. Bone marrow was harvested by gently flushing the marrow compartment with Dulbecco's modified Eagle's medium/F-12 medium containing heparin (10 units/ml) and DNase (1 µg/ml). Marrow cells were pelleted by centrifugation at 500 × g for 10 min at room temperature. The cells were re-suspended in 20 ml of -minimum essential medium containing 10% fetal bovine serum and transferred to a 50-ml plastic centrifuge tube. 15 ml of Histopaque-1077 (Sigma) was added to the bottom of the marrow cell suspension. The tube was centrifuged at 500 × g at room temperature for 30 min. Cell layer at the interface was harvested and washed three times with medium. The marrow cells were seeded in T-175 culture flasks at a density of 4 × 10⁵ cells/cm² and allowed to attach, without disturbance, for 7 days. After the attachment period, half of the culture medium was replaced with fresh medium, and thereafter cells were fed with the fresh medium at 3-4 day intervals. After cells reached confluency, bone marrow stromal cells were trypsinized and seeded in P-100 cell culture Petri dishes at a density of 1.0 × 10⁶ cells/dish and maintained in a humidified atmosphere of 95% air and 5% CO₂ at 37 °C. Cells were allowed to recover and attach to the plastic for 48 h and then fed with the -minimum essential medium without phenol red and containing 1% charcoal stripped fetal bovine serum for 48 h. Thereafter, cells were treated with the desired agents for specified time periods and used for RNA isolation. First passage cells were used in all experiments.

For the preparation of total cellular RNA, cells were washed with phosphate-buffered saline three times and immediately lysed in harvest buffer composed of 1% SDS, 0.25% NaCl, 30 mM Tris (pH 8.0), 2 mM EDTA (pH 8.0), and 1 mg/ml proteinase K. After 30 min of incubation at 37 °C, the cell lysate was extracted with phenol/chloroform and then with chloroform (30). The total RNA was further purified by precipitation in 2 M LiCl and ethanol, dissolved in diethylpyrocarbonate-treated water, and quantified by measuring absorbance at 260/280 nm on a Beckman spectrophotometer.

Aliquots of total RNA (20 µg) were separated under denaturing conditions on 1% agarose/formaldehyde gel using standard protocols (30) and stained with ethidium bromide to confirm RNA integrity and assess amounts of loading. RNAs were transferred by capillary diffusion onto nylon membranes (Schleicher & Schuell). The blots were dried for 2 h at 80 °C under vacuum and prehybridized overnight at 42 °C in a solution containing 50% formamide, 5 × Denhardt's solution, 3 × SSC, 10 µg/ml poly(A), 0.1% SDS, and 100 µg/ml salmon testes DNA and then hybridized overnight at 42 °C with ³²P-labeled IL-8 cDNA probe (kindly provided by Dr. Steven L. Kunkel, Department of Pathology, University of Michigan Medical School, Ann Arbor, MI) generated by random priming labeling. After hybridization, blots were washed twice with 2 × SSC/0.1% SDS for 15 min at room temperature and then once with 0.2 × SSC/0.1% SDS for 15 min at 52 °C. Blots were exposed to Amersham Hyperfilm-MP at 80 °C for appropriate time, and the film was developed using a Kodak RP X-OMAT processor. The data presented are representative of two or three individual experiments with similar results.

Cells were treated with appropriate agents for 2 h, washed twice with ice-cold phosphate-buffered saline, collected, nuclei prepared, and nuclear run-on performed with modifications (31). Nuclei were suspended in 0.1 ml of buffer (50 mM Tris-HCl, pH 8.3, 5 mM MgCl₂, 0.1 mM EDTA, and 40% glycerol). Elongation of nascent RNA chains was initiated by mixing the nuclear suspension with equal volume of reaction buffer (10 mM Tris-HCl, pH 8.0, 5 mM MgCl₂, 0.3 M KCl, 5 mM dithiothreitol, 1.0 mM each of ATP, CTP, and GTP (Promega, Madison, WI) and 0.1 mCi of [-³²P]UTP (specific activity, 3,000 Ci/mmol; Amersham Corp.). After incubation at 30 °C for 30 min, the nuclear mixture was incubated with RNase-free DNase I (170 units/ml, Promega) for 10 min at 37 °C and then digested with proteinase K (1 mg/ml) in 30 mM Tris-HCl, pH 8.0, 1% SDS, 0.25 M NaCl, and 2 mM EDTA for 30 min at 37 °C. The ³²P-labeled RNA was purified by phenol-chloroform extraction and ethanol precipitation and then dissolved in diethylpyrocarbonate-treated water. The pGEM-7Zf (+) vector alone, pGEM vector with IL-8 cDNA insert and pUC9 vector containing human -actin cDNA insert were linearized and immobilized on nitrocellulose membrane (Schleicher & Schuell). The membranes were prehybridized in a solution of 50% formamide, 4 × SSC, 2 × Denhardt's reagent, 25 mM sodium phosphate, 0.1% SDS overnight at 42 °C. The equal amount of radioactivity (cpm) from control and treated samples containing ³²P-labeled RNA was hybridized with immobilized plasmid DNA at 42 °C for 3 days. The membranes were washed twice, each for 20 min, with 2 × SSC/0.1% SDS at 42 °C, then twice, each for 20 min, with 0.2 × SSC/0.1% SDS at 52 °C, and then exposed to Hyperfilm-MP at 80 °C as described above.

Northern blot autoradiographs were scanned and quantitated by using a computerized ISS SepraScan 2001 (Integrated Separation Systems-Enprotech, Natick, MA). Northern blots were hybridized to IL-8 probe, stripped, and rehybridized to human -actin ³²P-labeled cDNA probe (kindly provided by Dr. Bratin Saha, Emory University, Atlanta, GA). The density readings of IL-8 mRNA were normalized to the respective -actin mRNA density readings. Nuclear run-on autoradiographs were scanned (Sony Multiscan HG) and quantitated by using Image-Pro plus program.

IL-8 was assayed in the conditioned medium by the highly specific quantitative ``sandwich'' enzyme-linked immunoassay using a IL-8 enzyme-linked immunosorbent assay kit (R & D Systems, Minneapolis, MN). The sensitivity of this assay is 3.0 pg/ml. Data shown under ``Results'' are the averages of two individual determinations.

RESULTS

To examine the kinetics of IL-1 stimulation of IL-8 mRNA expression by normal HBMS cells, experiments were conducted to study the effects of IL-1 at different time periods. As shown in Fig. 1, treatment of HBMS cells with IL-1 increased the steady-state level of IL-8 mRNA, which was detectable within 1 h, reached maximal by 4 h, and remained elevated at 24 h. As presented in Fig. 2, IL-1 increased the expression of IL-8 mRNA in a dose-dependent manner. IL-8 protein was also measured in the conditioned medium from the same experiment using the IL-8 enzyme-linked immunosorbent assay kit. Following the treatment with IL-1, IL-8 protein was first detectable in the conditioned medium of HBMS cells at 2 h (482 ± 43) pg/ml) and increased with time at 4 and 24 h (2,661 ± 168 and 61,684 ± 2,963 pg/ml, respectively).

Kinetics of IL-8 mRNA induction by IL-1 in normal HBMS cells.

Dose-response studies of IL-1 on IL-8 mRNA.

We previously demonstrated that dexamethasone inhibited IL-8 protein production by normal HBMS cells (27). In the present study, we determined the effects of IL-1 and dexamethasone on the steady-state levels of IL-8 mRNA as well as at the transcriptional level. As illustrated in Fig. 3 and Table I (experiment A), IL-1-induced IL-8 mRNA expression as well as the secretion of IL-8 protein was inhibited by dexamethasone. Since dexamethasone is a synthetic glucocorticoid with sustained biological effects because it is degraded slowly, we examined the effects of the natural glucocorticoid, hydrocortisone, on IL-8 mRNA levels. In addition, we also investigated the effects of PTH on IL-8 mRNA. As presented in Fig. 4 and Table I (experiment B), hydrocortisone, like dexamethsone, inhibited IL-8 mRNA expression as well as the production of IL-8 protein. PTH had no effect on basal IL-8 mRNA levels (Fig. 4).

Dose-response studies of dexamethasone on IL-1-induced IL-8 mRNA.

Table I. Effect of dexamethasone, hydrocortisone, and cycloheximide on IL-1-induced IL-8 production in HBMS cells HBMS cells were treated with IL-1, dexamethasone (Dex), hydrocortisone (HC), and cycloheximide (CHX; 5 µg/ml) for 4 h. The medium was collected and IL-8 was assayed in the medium by enzyme-linked immunosorbent assay. The values are the means ± S.D. of two individual determinations. The values for control without IL-1 were 11.5 ± 5.3 pg/ml. Treatment IL-8 pg/ml Experiment A IL-1 (1 ng/ml) 2597  ±  148  IL-1 + Dex 107 M 808  ±  21  IL-1 + Dex 108 M 782  ±  5  IL-1 + Dex 109 M 1502  ±  96  Experiment B IL-1 (1 ng/ml) 8981  ±  411  IL-1 + HC 106 M 2337  ±  180  IL-1 + HC 107 M 2978  ±  355  IL-1 + HC 108 M 5891  ±  3  Experiment C IL-1 (1 ng/ml) 1344  ±  161  IL-1 + Dex 107 M 349  ±  11  IL-1 + Dex 107 M + CHX 126  ±  12

Effect of IL-1, hydrocortisone, and PTH on IL-8 mRNA.

To determine whether the suppression of IL-8 mRNA by dexamethasone is independent of new protein synthesis, experiments were performed with cycloheximide. Cycloheximide (5 µg/ml) alone slightly increased the basal levels of IL-8 mRNA in a time-dependent manner and further enhanced the IL-1-induced IL-8 mRNA (Fig. 5). The effect of cycloheximide on the suppression of IL-1-induced IL-8 mRNA and IL-8 protein in HBMS cells by dexamethasone was also determined. Dexamethasone inhibited IL-1-stimulated IL-8 mRNA expression (Fig. 6) and the IL-8 protein production (Table I, experiment C). The concurrent presence of cycloheximide with dexamethasone not only abolished the inhibitory effect of dexamethasone but further enhanced the expression of IL-8 mRNA (Fig. 6). Fetal bovine serum (10%) had no effect on IL-8 mRNA levels (Fig. 6). Cycloheximide also inhibited IL-1-induced IL-8 protein secretion (Table I, experiment C). To examine whether glucocorticoids down-regulated IL-8 expression by decreasing the IL-8 mRNA stability, experiments were performed with actinomycin D. In these studies actinomycin D, dexamethsone or dexamethasone in combination with actinomycin D were added to cells prestimulated for 3 h with IL-1 and treated for an additional 1, 2, and 4 h. The stability of IL-8 mRNA was similar in cells treated with IL-1 or IL-1 plus dexamethasone that had received actinomycin D, demonstrating that this is a very long-lived message and the effects of dexamethasone require mRNA synthesis (Fig. 7). It is important to note that in the same experiment (Fig. 7), dexamethasone strikingly reduced IL-8 mRNA levels in the absence of actinomycin D. 

Kinetics of cycloheximide effect on the IL-1-induced IL-8 mRNA.

Effect of actinomycin D on the suppression of IL-1-induced IL-8 mRNA by dexamethasone.

7

IL-, IL-1

Nuclear run-on experiments were performed to establish if IL-8 gene was induced at the transcriptional level by IL-1. As shown in Fig. 8, the nuclear run-on analysis indicated that the IL-8 transcription was increased by IL-1, whereas dexamethasone had no effect on IL-1-induced transcription of IL-8 gene (Fig. 8). These results indicate that the IL-8 gene is indeed transcriptionally activated by IL-1 and dexamethasone suppresses IL-8 mRNA levels by increasing its degradation without affecting the transcriptional rate. This observation did not appear to be due to nonspecific activation because IL-8 transcriptional activity was not detectable in nuclei from nontreated control cells and because the same results were obtained in each of the four independent experiments. Furthermore, the transcriptional rate of the control -actin gene was not significantly affected.

Effects of IL-1 and dexamethasone on IL-8 gene transcription.

7

To determine whether IL-1 induction of IL-8 mRNA was mediated via protein kinase pathways, we examined the effects of protein kinase inhibitors on IL-1-stimulated IL-8 mRNA expression. As shown in Fig. 9, H-7 (50 µM), and staurosporine (1 µM), potent inhibitors of protein kinase C, blocked IL-1-induced IL-8 gene expression, suggesting that the stimulation of IL-8 gene expression by IL-1 is mediated via protein kinase C. Genistein (100 µM), a specific protein tyrosine kinase inhibitor, also inhibited IL-8 gene expression. Curcumin (20 µM), an inhibitor of c-jun/AP-1, protein kinase C, and protein tyrosine kinase, also blocked IL-1-stimulated IL-8 gene expression (Fig. 9).

Effect of protein kinase inhibitors on IL-1-induced IL-8 mRNA.

DISCUSSION

The results of the present study demonstrate that in human bone marrow stromal cells, IL-1 induces the expression of the IL-8 gene. IL-8 mRNA was not detectable in the unstimulated HBMS cells. Both synthetic and natural (dexamethasone and hydrocortisone) glucocorticoids inhibit IL-1-induced IL-8 mRNA expression and IL-8 protein production by normal human bone marrow osteoprogenitor stromal cells. The induction of IL-8 gene expression by IL-1 appears to be mediated by protein kinases including protein kinase C and protein tyrosine kinase. The accumulated data possibly suggest the involvement of c-JUN/AP-1 in the induction of IL-8 gene expression by IL-1. IL-8 mRNA expression was also increased when IL-1 and cycloheximide were added simultaneously to the cell cultures when compared with results obtained with IL-1 alone. Cycloheximide enhances the expression of IL-8, presumably by inhibiting the synthesis of negative regulatory elements (nucleases and repressors) that are coinduced (32, 33). Cycloheximide did not inhibit but rather enhanced the IL-8 mRNA expression induced by IL-1, suggesting that IL-8 mRNA induction by IL-1 does not require de novo protein synthesis, as has been previously reported for inflammatory cytokines including IL-8 (34, 35, 36) and immediate-early genes (37, 38, 39) primarily by stabilization of mRNAs in different cell types. However, there are differences in the characteristics of IL-8 gene and immediate-early genes. IL-8 gene is induced rapidly but not transiently (Fig. 1), and serum has no effect (Fig. 6), whereas immediate-early genes are induced rapidly and transiently by serum or mitogens (38). The suppressive effect of dexamethasone on IL-1-stimulated IL-8 mRNA was not observed in the presence of cycloheximide. These observations demonstrate that the suppressive effects of dexamethasone on IL-8 gene expression depends on new protein synthesis and suggest that some newly dexamethasone-induced regulatory protein (repressor) is involved.

Our results, which demonstrate that dexamethasone inhibits both IL-8 mRNA and IL-8 protein in normal HBMS cells, are consistent with the observations that glucocorticoids suppress IL-8 mRNA in normal human embryonic lung fibroblasts (35) and porcine macrophages (36). Actinomycin D alone or actinomycin D plus dexamethasone have little effect on IL-1-stimulated IL-8 mRNA, suggesting that IL-8 mRNA has a longer life-span. The life-span of IL-8 mRNA was much shorter in the presence of dexamethasone than when overall transcription was blocked by actinomycin D. The accumulated data from studies with cycloheximide and actinomycin D suggest that glucocorticoids stimulate the transcription and translation of a factor that is pivotal in either causing a dramatic decline in IL-8 transcription and/or mRNA degradation. Because nuclear run-on experiments have demonstrated that IL-1 increases the transcription of IL-8 gene and dexamethasone has no effect on IL-1-stimulated IL-8 gene transcription, the inhibitory action of dexamethasone on IL-8 mRNA appears to be the consequence of increased mRNA degradation. The stability of IL-8 mRNA may be influenced by the RNA instability element, AUUUA, identified in the 3-untranslated region of IL-8, IL-6, and granulocyte-macrophage CSF mRNAs (40, 41). These sequences are critical for mRNA stability because they may represent recognition sites for RNases (41). Because macrophage CSF mRNA does not contain AUUUA sequences in 3-untranslated region macrophage CSF, mRNA is not down-regulated by glucocorticoids in normal human lung fibroblasts (42, 43). Therefore, one possible mechanism whereby glucocorticoids destabilize mRNA of IL-8 in HBMS cells may be through the activation of specific genes that are responsible for the synthesis of certain ribonucleases or other proteins that may bind to defined mRNA instability element, i.e., AUUUA, leading to a rapid degradation of the IL-8 mRNA. This is consistent with the identification of an AUUUA-specific mRNA binding protein in Jurkat cells (44) and that AUUUA-rich sequences in 3-untranslated region mediate the increased turnover of IFN- mRNA induced by dexamethasone (45).

The human IL-8 gene has been sequenced and shown to contain several transcriptional regulatory elements, including AP-1, AP-2, NF-B, NF-IL6, HNF-1, glucocorticoid response element, and interferon regulatory factor-1 (46, 47, 48). Our results, which demonstrated no effect of serum on IL-8 mRNA, are consistent with the fact that the IL-8 gene promoter lacks a serum response element (47). The IL-8 gene promoter has a AP-1 (c-JUN/c-FOS) binding site, and IL-1 stimulates AP-1 and NF-B activity. The 5-flanking region deletion studies of IL-8 gene demonstrated that the nucleotides between 94 and 71 base pairs from the start of the first exon, which contain two cis-elements, NF-B- and C/EBP-like factors, are essential and sufficient for the IL-8 induction by either IL-1, TNF, or phorbol 12-myristate 13-acetate (48). Furthermore, it has been shown that IL-1 rapidly and transiently induces the expression of c-jun, c-fos, and c-myc mRNA in different cell types (49, 50, 51). Thus, one of the effects of IL-1 on HBMS cells may be the stimulation of AP-1, NF-B, and/or C/EBP activities, which in turn activate the transcription of the IL-8 gene. The deletion analysis of IL-8 gene promoter will provide means to identify the specific sites and factors that are responsible for IL-8 gene regulation by IL-1 and glucocorticoids.

The accumulated results also demonstrate that induction of IL-8 mRNA by IL-1 is inhibited by both serine/threonine and tyrosine protein kinase inhibitors, suggesting that protein kinase C and protein tyrosine kinase are involved in the transcriptional activation of IL-8 gene by IL-1. The exact nature of intracellular signaling pathways that link the activation of the IL-1 receptors, which have no intrinsic protein tyrosine kinase activity, to the cellular responses is still unknown. Depending on cell types, studies on IL-1 receptor activation have implicated virtually every second messenger pathway including cAMP, inositol phosphate hydrolysis, diacylglycerol, eicosanoids, immediate-early genes, and protein kinase A and C (52, 53). Others have shown that IL-1 does not transduce signals via protein kinase C, although it does activate other serine/threonine kinases including mitogen-activated protein kinase (54). Also increased phosphorylation of two proteins (65 and 74 kDa) by IL-1 in human peripheral blood mononuclear cells is not mediated by protein kinase C and A, suggesting that another protein kinase may be involved (55). Curcumin (diferuloylmethane) is a major pigment of turmeric that possesses both anti-inflammatory and antioxidant properties (56, 57). It also inhibits the activity of protein kinase C and tyrosine kinase, tumor initiation and promotion, arachidonic acid metabolism, including synthesis of prostaglandins, thromboxanes, and leukotrienes (56, 57, 58, 59). Furthermore, curcumin has been shown to down-regulate the phorbol 12-myristate 13-acetate-induced c-jun gene expression in mouse fibroblasts (60) and inhibits TNF--stimulated c-jun and monocyte chemoattractant JE gene expression in osteoblastic MC3T3-E1 cells without affecting the c-fos gene expression (61). Similarly, the suppression of IL-8 gene expression in HBMS cells by curcumin may be mediated via inhibition of c-JUN/AP-1 activity and protein kinases. Whether a curcumin receptor exits in HBMS cells or whether curcumin can activate some cellular protein(s), which then interact with c-JUN/AP-1 thus inhibiting IL-8 gene expression, is still unknown.

In summary, our results have clearly demonstrated that in normal human bone marrow stromal cells, IL-1 induces IL-8 mRNA by increasing the gene transcription and that IL--induced IL-8 gene expression is down-regulated by glucocorticoids and protein kinase inhibitors. The data also reveal that glucocorticoids suppress IL-8 gene expression mainly by decreasing the mRNA stability without affecting the transcription of the gene and by inducing the synthesis of some protein factor (repressor). Thus, IL-8 must be added to the ever increasing list of cytokines (16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26) that are known to regulate bone remodeling by affecting functions of osteoclasts and osteoblasts. In this regard it is noteworthy that IL-8 inhibits the bone resorbing activity of osteoclasts (28) and alkaline phosphatase activity of osteoblasts (29), suggesting that IL-8 may act as a signaling molecule in the osteoclast-osteoblast interactions regulating bone remodeling by modulating the microenvironment of bone and bone marrow.