Ca²⁺-Dependent Production and Release of IL-8 in Human Neutrophils¹

Interleukin-8 is a potent C-X-C chemokine that induces exocytosis of specific granules of neutrophils (1), up-regulation of integrin expression on the surface of neutrophils (2), and increased adherence of neutrophils to endothelial cells (3). Exudative neutrophils harvested from an inflammatory site have ∼200-fold more cell-associated IL-8 than isolated peripheral blood neutrophils (4). These observations, together with the demonstration that neutrophils undergoing phagocytosis release IL-8 (5), suggest that stimulation of IL-8 production by the neutrophil may be a paracrine mechanism for recruitment of additional neutrophils, as well as other immune cells such as lymphocytes (6) and basophils (7), to inflammatory sites.

While the mechanism(s) controlling IL-8 production and release have not been fully delineated, the observation that the Ca²⁺ ionophore, A23187, increased neutrophil IL-8 synthesis and release (4) implicates a role for elevation of the intracellular Ca²⁺ concentration, [Ca²⁺]_i,³ in the signaling pathway. To further characterize the role of [Ca²⁺]_i in IL-8 production by neutrophils, this report examines the effect of thapsigargin on the production and release of IL-8 and other cytokines by neutrophils. Thapsigargin is a naturally occurring sesquiterpene lactone that inhibits a microsomal Ca²⁺-ATPase, and thus results in the release of microsomal Ca²⁺ stores. In neutrophils (8) and many other cells (9, 10, 11), thapsigargin not only causes the release of intracellular Ca²⁺ stores, but also opens a Ca²⁺ influx pathway. These events can lead to cell activation in the absence of receptor-ligand interactions, e.g., the induction of IL-2 (12) and the α-chain of the IL-2R in lymphocytes (13) and IL-6 production in murine macrophages (14).

This report demonstrates that the thapsigargin-induced synthesis and release of IL-8 in neutrophils is relatively specific and requires a sustained Ca²⁺ influx. Studies with inhibitors of the immunophilins that target calcineurin implicate this calmodulin-dependent phosphatase in the signaling cascade leading to thapsigargin-induced IL-8 synthesis.

The following reagents were purchased from the indicated sources: recombinant human IL-1β (R&D Systems, Minneapolis, MN); recombinant human IL-8 and recombinant human TNF-α (PeproTech, Rocky Hill, NJ); HEPES and HBSS (BioWhittaker, Walkersville, MD); and 1-[2-amino-5-(6-carboxyindol-2-yl)phenoxy]-2-(2′-amino-5′-methylphenoxy)ethane-N,N,N′,N′-tetraacetic acid pentaacetoxymethyl ester (Indo1-AM) and BAPTA-AM, 1,2-bis-(2-aminophenoxy)ethane-N,N,N′,N′-tetraacetic acid tetra(acetoxymethyl) ester (BAPTA-AM) (Molecular Probes, Eugene, OR). Anti-human cytokine neutralizing Abs were obtained from the following sources: MAB201 (anti-IL-1β) and MAB208 (anti-IL-8, R&D Systems); and mAb 1 (anti-TNF-α) and 107.3 (purified mouse IgG1 isotype control Ig, PharMingen, San Diego, CA). All other chemicals used were of reagent grade and were purchased from Sigma (St. Louis, MO). Stock thapsigargin was dissolved in tissue culture grade, endotoxin-free DMSO at 1 mM and stored until use at −80°C.

Peripheral blood neutrophils were isolated from blood obtained from normal volunteers (anticoagulant citrate dextrose solution USP formula A, Baxter Healthcare, Deerfield, IL). Whole blood was diluted with HBSS without divalent cations and layered over a discontinuous gradient cushion (Histopaque-1083; Sigma). The sample was spun for 30 min at 500 × g at room temperature. After discarding the supernatant fluid, the neutrophil/erythrocyte pellet was suspended in an equal volume of HBSS. The cell suspension was then diluted with dextran ([1.5%]_final, m.w. 200,000–500,000 g/mol; Pharmacia, Uppsala, Sweden) and allowed to sediment at 1 × g for 20 min. The neutrophil-rich supernatant fluid was harvested and spun at 200 × g for 10 min. Contaminating erythrocytes in the pellet were removed by two sequential hypotonic lyses using 0.25× PBS for 30 s followed by an equal volume of 1.75× PBS to restore isotonicity. The neutrophils were then counted on an automated cell counter (model T540, Coulter, Hialeah, FL). The final preparation of neutrophils was >95% pure with 4% eosinophils and <1% monocytes and lymphocytes as assessed by differential staining.

Isolated peripheral blood neutrophils (2 × 10⁶) were suspended in HBSS with 10 mM HEPES (pH 7.35) and incubated in 2 ml polypropylene screw cap tubes (Sarstedt, Newton, NC) for up to 8 h at 37°C in the presence or absence of thapsigargin. In all inhibitor studies, with the exception of the cell permeant inhibitor BAPTA-AM, inhibitors were added 10 min before the addition of thapsigargin and were present throughout the experiment. In studies with BAPTA-AM, neutrophils (1 × 10⁷/ml HBSS without divalent cations) were incubated with the indicated concentration of BAPTA-AM for 45 min at 37°C, washed twice, and resuspended at the indicated cell concentration in HBSS with HEPES. Neutrophils loaded with as high as 75 μM BAPTA-AM remained viable, evidenced by their ability to a maintain a Ca²⁺ gradient.

After incubation with thapsigargin for the indicated times, the cell suspensions were spun at 4°C and the supernatant fluid harvested for analysis. Cell pellets were solubilized in 0.2% Triton X-100 using 20, 1-s pulses (minimum setting) with a microtip sonicator (Sonifier II, Branson Ultrasonics, Danbury, CT).

All of the following cytokines were measured using commercial enzyme-linked immunosorbent assays (R&D Systems); manufacturer’s limits of detection are included in parentheses: IL-8 (4.7 pg/ml), IL-6 (0.7 pg/ml), IL-1β (1.0 pg/ml), IL-1R antagonist (IL-1Ra; 6.5 pg/ml), TNF-α (4.4 pg/ml), RANTES (2.5 pg/ml), growth-related protein α (GRO-α; 5.0 pg/ml), and macrophage inflammatory protein-1α (MIP-1α; 7.0 pg/ml). Lactate dehydrogenase activity was determined by monitoring the reduction of NADH at 340 nm using a molar extinction coefficient of 6220.

Total cellular RNA was isolated from neutrophil preparations by a single step phenol/chloroform extraction procedure using TRIzol (Life Technologies, Gaithersburg, MD). The OD₂₆₀/OD₂₈₀ of the extracted RNA was >1.6; total RNA yields were 20 to 30 μg/5 × 10⁷ neutrophils. RNA concentrations were also determined using the RiboGreen Quantitation Kit with a ribosomal RNA standard (Molecular Probes). In general, the concentrations obtained spectrophotometrically were 2-fold higher than those obtained using the commercial kit. Total cellular RNA (10 μg per sample) was size-fractionated on formaldehyde-denaturing 0.8% agarose gel and transferred to Magnagraph (Micron Separations, Westborough, MA). After UV crosslinking, blots were hybridized in NyloHybe (Fast Pair, Silver Spring, MD) to ³²P-labeled cDNA probes prepared by random priming utilizing a commercially available kit (Stratagene, La Jolla, CA). Blots were hybridized for 24 to 36 h at 42°C and then washed for 10 min at room temperature in 2× SSC/0.1% SDS followed by a 10-min wash at 65°C in 0.2× SSC/0.1% SDS. After hybridization with one probe, the blots were stripped by placing in boiling 0.02× SSC/0.01% SDS for 20 min. Blots were then placed in hybridization buffer for 1 to 2 h and the next probe added. All cDNA probes had a specific activity of 2 to 8 × 10⁸ cpm/μg and all hybridizations were performed with 1 × 10⁶ cpm/ml. Blots were exposed to Kodak (Rochester, NY) X-OMAT x-ray film for 2 to 18 h at −70°C. Human IL-8 cDNA was obtained from Dr. Joost Oppenheim (National Cancer Institute/Frederick Cancer Research and Development Center, Frederick, MD) and chicken β-actin cDNA was obtained from Dr. Donald Cleveland (Johns Hopkins University, Baltimore, MD). Northern blots were quantitated using a imaging densitometer (model GS 670, Bio-Rad, Hercules, CA). Relative levels of IL-8 mRNA were adjusted for unequal loading using β-actin mRNA expression.

RiboQuant multiprobe RNase protection assays (hCK-5, hCK-2, and hCK-3; PharMingen) were performed as described by the manufacturer. One microgram of neutrophil RNA was added to the reaction mix. A plot of the mobility of the probes vs the nucleotide length was used to predict the migration of the protected probe fragments.

Neutrophils (1 × 10⁷/ml of HBSS/HEPES) were incubated with the cell permeant dye, Indo1-AM, in the dark at 37°C for 45 min. The neutrophils were then pelleted by centrifugation at 200 × g, resuspended in HBSS/HEPES, and the procedure repeated to remove the residual extracellular Indo1-AM. The cells were resuspended at 2.5 × 10⁶/ml HBSS/HEPES. Changes in [Ca²⁺]_i were monitored on a DeltaScan spectrophotometer (Photon Technology, South Brunswick, NJ) using a thermostatically controlled cuvette holder. Data were collected as the ratio (R) of the λ_emissions [λ_{402 nm}/λ_{486 nm}] using an λ_excitation = 358 nm. [Ca²⁺]_i was determined as described previously (15). R_max and R_min were empirically determined by addition of ionomycin (1 μM) and EGTA (12.5 mM), respectively.

The equilibrium composition of free Ca²⁺ in an EGTA-containing buffer solution was derived using the software EQCAL (BIOSOFT, Cambridge, U.K.) and equilibrium constants cited in the software manual.

The significance of difference between test and control groups was analyzed using either Student’s t test or analysis of variance (ANOVA).

Maintenance of unstimulated isolated peripheral blood neutrophils for 8 h in HBSS/HEPES resulted in the accumulation of a total of 0.6 ng of IL-8/10⁶ neutrophils (n = 2), with only a small fraction of (<10%) found in the extracellular fluid. In contrast, incubation with thapsigargin (50–100 nM) for 8 h induced the accumulation of 26.4 ng of IL-8/10⁶ neutrophils (n = 2), equally partitioned between the cellular and extracellular compartments. The induction of IL-8 was detectable by 15 min and continued to rise throughout the 8-h measurement period (Fig. 1). The rate of total IL-8 production was greatest 1 to 2 h after addition of thapsigargin. In the first hour, most of the IL-8 (>90%) was found in detergent extracts of the cells. At 4 h, cellular IL-8 reached a plateau, while levels of IL-8 in the extracellular fluid continued to increase. The viability of thapsigargin-treated neutrophils at 4 h was within 90% of the viability of untreated neutrophils when assessed by both trypan blue exclusion and lactate dehydrogenase release. In addition, despite incubation of Indo1-loaded neutrophils with 100 nM of thapsigargin for 4 h, addition of 1.0 μM ionomycin caused a further increase in [Ca²⁺]_i (data not shown), indicating that the cells still maintained a Ca²⁺ gradient across the plasma membrane.

Treatment of neutrophils with the protein synthesis inhibitor, cycloheximide (added at 10 μg/ml at t = −10 min before the addition of thapsigargin and present throughout the course of the experiment), inhibited thapsigargin-induced IL-8 production (measured at t = 4 h; 15.21 ± 2.92 ng IL-8/10⁶ neutrophils in the absence of cycloheximide vs 1.18 ± 0.56 ng IL-8/10⁶ neutrophils in the presence of cycloheximide, p < 0.05), indicating that protein synthesis de novo was responsible for the marked increase in IL-8 production.

Northern blot analysis of total cellular RNA isolated from neutrophils stimulated with 100 nM thapsigargin showed an increase in mRNA for IL-8 that was detectable at the first time point (5 min), reached a maximum at 1 h, and remained elevated at 4 h (Fig. 2), suggesting that IL-8 production resulted from a rapid and sustained expression of IL-8 mRNA levels. As shown in Figure 2, incubation of neutrophils in buffer alone (resting) for 1 h caused a slight increase in IL-8 mRNA relative to β-actin mRNA compared with the effect observed with thapsigargin. Densitometric analysis of the Northern blots indicated that neutrophils treated with thapsigargin exhibited a 7- to 19-fold increase (n = 3) in the expression of IL-8 mRNA relative to freshly isolated control neutrophils. Neutrophils incubated in buffer alone exhibited a 3- to 4-fold increase in IL-8 mRNA.

Thapsigargin induced a dose-dependent (ED₅₀ = 19.4 ± 4.6 nM) increase in IL-8 production (Fig. 3). Concentrations of thapsigargin as low as 1 to 10 nM caused significant increases in the levels of cellular IL-8 with very little extracellular release of IL-8. Higher levels of thapsigargin (20–100 nM) were associated with increases in both the cellular and extracellular IL-8. Peak IL-8 production measured at 4 h (18.5 ng/10⁶ neutrophils) was observed with 50 to 100 nM thapsigargin.

Northern blot analysis of total cellular RNA isolated from neutrophils showed a similar dose-response curve for induction of IL-8-specific mRNA. Neutrophils stimulated with thapsigargin showed a detectable increase in mRNA for IL-8 with concentrations of thapsigargin as low as 2 nM while maximal increases in mRNA for IL-8 occurred at 50–100 nM thapsigargin (Fig. 4).

To determine whether the stimulatory effect of thapsigargin was specific for IL-8, or if other inflammatory cytokines also were up-regulated by this Ca²⁺-ATPase inhibitor, cells were incubated with 50 nM of thapsigargin for 4 h, and both the cellular and extracellular levels of several different inflammatory cytokines were measured. The results shown in Table I indicate that the levels of another C-X-C chemokine, GRO-α, and the C-C chemokines, RANTES and MIP-1α, or other cytokines/antagonists such as IL-6 and IL-1Ra were unchanged by incubation with thapsigargin. However, while IL-8 was the predominant cytokine induced by thapsigargin, the levels of TNF-α and possibly IL-1β also increased in response to thapsigargin. These increases were to a much smaller degree (in terms of total mass) than the increases measured in the level of IL-8. Because mononuclear cells can produce large quantities of IL-1β and TNF-α (16), the increased levels of these cytokines could be accounted for by the contamination of the neutrophil preparation by as little as 1% mononuclear cells (the estimate of contamination of our preparations).

Regardless of the source of the IL-1β and TNF-α, it was possible that thapsigargin-induction of IL-8 production was a consequence of secondary stimulation by either cytokine. Several experimental protocols were performed to examine this possibility. The first set of studies demonstrated that incubation of neutrophils with exogenous IL-1β or TNF-α (at concentrations 1–2 logs higher than those produced after stimulation of neutrophils with thapsigargin) failed to induce the IL-8 levels observed with thapsigargin (Table II). The second set of studies demonstrated that incubation of neutrophils with IL-1β or TNF-α did not synergize with the effect of a suboptimum dose (20 nM) of thapsigargin on IL-8 production (Table II). Lastly, the addition of neutralizing Abs against both IL-1β or TNF-α (at concentrations which blocked the priming of neutrophils for enhanced FMLP-induced O₂⁻ generation by IL-1β and TNF-α) had no effect on thapsigargin-induced IL-8 production.

To determine whether there was up-regulation of mRNA for genes other than IL-8, RNase protection assays were performed on RNA isolated from thapsigargin-treated neutrophils. Thapsigargin induced increased expression of mRNA for MIP-1α and MIP-1β, as well as IL-8. There was no detectable expression in resting cells, nor increased expression in thapsigargin-treated neutrophils, of mRNA for the chemokines, lymphotactin, RANTES, IFN-γ-inducible protein 10 (IP-10), monocyte chemotactic protein-1 (MCP-1), and I309 (Fig. 5). In a single experiment assessing the expression of mRNA for other cytokines, there was detectable expression of mRNA for IL-1α, IL-1β, IL-1Ra, TNF-α, and TNF-β in resting neutrophils. Neutrophils treated with thapsigargin exhibited increased expression of mRNA for both IL-1α and IL-β. There was no detectable expression in resting neutrophils, nor increased expression in thapsigargin-treated neutrophils, of mRNA for IL-12p35, IL-12p40, IL-6, IL-10, IFN-γ, IFN-β, lymphotoxin-β, or TGF-β1–3 (data not shown).

Addition of a low dose of thapsigargin (5–10 nM) to Indo1-loaded neutrophils, a dose that caused a significant albeit submaximal increase in cellular IL-8 and little IL-8 secretion, caused a small (50–100 nM) elevation in [Ca²⁺]_i (Fig. 6, inset). In contrast, as shown in Figure 6, a higher concentration of thapsigargin, 20 to 100 nM, produced a gradual, biphasic rise in the [Ca²⁺]_i that peaked within 2 to 3 min, dropped slightly, and then rose again to a maximum level ([Ca²⁺]_i = 1.0 μM) by 10 min that persisted for at least 4 h (data not shown). As shown by the R_max, ionomycin induced an even greater elevation in the [Ca²⁺]_i, indicating that neutrophils were still capable of maintaining their Ca²⁺ gradient across the plasma membrane. Dose-response studies showed that thapsigargin-induced changes in [Ca²⁺]_i exhibited an ED₅₀ similar to thapsigargin-induced IL-8 production (Fig. 6 inset compared with Fig. 3).

Addition of 2.5 mM EGTA (a dose that negates the Ca²⁺ gradient across the plasma membrane) before the addition of thapsigargin (50 nM) had no effect on the thapsigargin-induced elevation in [Ca²⁺]_i observed within the first 2 min but blocked a sustained rise in [Ca²⁺]_i, suggesting that early changes in [Ca²⁺]_i resulted from the release of intracellular Ca²⁺ stores (Fig. 7). Addition of EGTA at ≥2 min after thapsigargin caused a rapid drop in the [Ca²⁺]_i to levels observed in the presence of EGTA added at t = 0, suggesting that later increases in [Ca²⁺]_i involved the influx of Ca²⁺ across the plasma membrane.

To determine whether an influx of extracellular Ca²⁺ induced by thapsigargin was involved in the stimulation of IL-8 production, experiments were performed in which neutrophils were exposed to EGTA (2.5 mM) before or at various times after the addition of thapsigargin (Table III). The addition of EGTA (2.5 mM) before the addition of thapsigargin caused >95% inhibition of the thapsigargin-induced increase in IL-8 production. EGTA also caused a similar inhibition of the thapsigargin-stimulated induction of mRNA for IL-8 (Fig. 8). Addition of EGTA up to 2 min after the addition of thapsigargin (and concurrent with the EGTA-insensitive elevation of [Ca²⁺]_i]) continued to block 95% of the thapsigargin-induced IL-8 production (Table III). Addition of EGTA at later times (t ≥ 5 min, and subsequent to activation of the Ca²⁺ influx) extended the duration of the Ca²⁺ influx and had progressively less of an inhibitory effect, demonstrating that a sustained influx of extracellular Ca²⁺ was required for maximum thapsigargin-induced IL-8 production. In addition to blocking the synthesis of IL-8, the addition of EGTA also inhibited the secretion of IL-8 into the extracellular fluid. When EGTA was added 120 min after the addition of thapsigargin, a time at which EGTA no longer significantly altered the synthesis of IL-8, the secretion of IL-8 was still significantly inhibited (35 ± 7% of the total IL-8 in neutrophils treated with thapsigargin alone vs 25 ± 7% of the total IL-8 in neutrophils treated with thapsigargin plus EGTA, p < 0.01).

In related studies, neutrophils were incubated with the intracellular Ca²⁺ chelator, BAPTA-AM, to determine whether the sustained elevation of [Ca²⁺]_i was required to support the thapsigargin-induction of IL-8 production. Loading with BAPTA-AM resulted in a dose-dependent inhibition of thapsigargin-induced IL-8 production (Fig. 9). However, BAPTA-AM loading did not significantly alter the percentage of synthesized IL-8 that was secreted at any of the concentrations tested (data not shown).

Cyclopiazonic acid is a mycotoxin that is structurally unrelated to thapsigargin. However, both cyclopiazonic acid and thapsigargin are reversible inhibitors of the endoplasmic reticulum Ca²⁺-ATPase and promote the release of internal stores of Ca²⁺ (17, 18). Cyclopiazonic acid caused an EGTA-inhibitable production of IL-8, similar to that observed with thapsigargin. An EGTA-inhibitable increase in IL-8 was also induced by the Ca²⁺ ionophore, ionomycin (Table IV).

Cyclosporin A and ascomycin are immunosuppressive drugs that bind to specific immunophilins and interfere with the activation of the Ca²⁺-, calmodulin-dependent serine/threonine phosphatase, calcineurin, or protein phosphatase 2B (19). Preincubation of neutrophils for 10 min with ascomycin and cyclosporin A, at concentrations achievable therapeutically, caused significant inhibition (IC₅₀ values of 2.8 ± 0.1 and 54.2 ± 8.4 nM, respectively) of thapsigargin-induced IL-8 production (Fig. 10) and IL-8 mRNA synthesis (Fig. 8). Densitometric analyses of the Northern blots indicated that cyclosporin A and ascomycin reduced the levels of thapsigargin-induced IL-8 mRNA to ∼50% the level observed in the absence of the inhibitors. Rapamycin, a member of the same family of immunosuppressants that complexes with an immunophilin but does not interfere with the activation of calcineurin (20), did not inhibit thapsigargin-induced IL-8 gene expression. None of the inhibitors caused any significant alteration in the percentage of synthesized IL-8 that was secreted.

To determine the specificity of the protein phosphatase 2A inhibitors, okadaic acid and calyculin A, inhibitors of protein phosphatases 1 and 2A, were tested and found to have no effect on thapsigargin-induced IL-8 synthesis at concentrations of 20 and 2 nM, respectively.

Previous studies demonstrated that while unstimulated peripheral blood neutrophils produced minimal levels of IL-8, the Ca²⁺ ionophore, A23187, increased IL-8 to levels comparable to those found in exudative neutrophils isolated from an inflammatory site in vivo (4). The data suggested elevation of [Ca²⁺]_i may be an important physiologic signal regulating the production of the inflammatory cytokine IL-8 by neutrophils. The studies presented in the current report extend those observations by demonstrating that incubation of freshly isolated peripheral blood neutrophils with thapsigargin caused a dose-dependent increase in the production of the chemokine, IL-8, to levels comparable to those previously observed in exudative neutrophils (4).

Thapsigargin stimulation of IL-8 levels was evident 1 h after exposure, continued for the entire 8-h measurement period, and was blocked by cycloheximide. Northern blot analysis of total RNA isolated from peripheral blood neutrophils demonstrated that the levels of IL-8 mRNA were low in both freshly isolated neutrophils and control neutrophils incubated at 37°C for 1 h, whereas exposure to thapsigargin caused a rapid induction (within 5 min) of IL-8 mRNA that persisted for up to 4 h. Thus, thapsigargin treatment of neutrophils induced both the transcription and the subsequent translation and secretion of IL-8.

These findings are consistent with other studies demonstrating thapsigargin induces the production of cytokines in both macrophages and lymphocytes (12, 14). To determine whether thapsigargin selectively induced production and release of IL-8 from neutrophils, measurement of cellular and extracellular levels of other inflammatory cytokines were performed. Thapsigargin increases in IL-8 were relatively specific, with no detectable increases in the levels of GRO-α, RANTES, MIP-1α, IL-6, or IL-1Ra. Small increases in the levels of IL-1β and TNF-α were observed, but these were at much lower levels (pg of IL-1β or TNF-α/10⁶ cells vs ng of IL-8/10⁶ cells) and could occur from the 1% contaminating mononuclear cells (16). Based on studies using Abs that neutralized the activity of IL-1β and TNF-α, we excluded the possibility that induction of IL-8 by thapsigargin was due to the stimulatory effect of these cytokines.

Although the effect of thapsigargin on the production of cytokines was relatively specific for IL-8, notable increases were observed in the concentrations of both TNF-α and IL-1β. However, at the level of mRNA, RNase protection assays revealed marked increases in the mRNA levels of MIP-1α, MIP-1β, IL-1α, IL-1β, and TNF-α. These data can be explained by two possibilities. It is possible that these other cytokines are synthesized but at levels too low for detection. Alternatively, it is possible that the regulatory mechanisms for synthesis of these other cytokines differ at the translational level.

Numerous studies have demonstrated that thapsigargin induces depletion of Ca²⁺ stores, elevating [Ca²⁺]_i. In many cell types, including neutrophils, thapsigargin-induced depletion of Ca²⁺ stores activates a Ca²⁺ influx pathway (9, 21, 22, 23). Because thapsigargin can activate cells in the absence of receptor ligand interactions by modulating Ca²⁺ homeostasis, it has been a very useful tool for dissecting signal transduction pathways. Our data confirm other studies in neutrophils showing that thapsigargin induces an initial EGTA-insensitive [Ca²⁺]_i rise followed by a sustained [Ca²⁺]_i elevation that is blocked by EGTA (23). Addition of EGTA at various times after exposure to thapsigargin rapidly restored [Ca²⁺]_i to baseline levels, indicating that the influx of Ca²⁺ determined the level of [Ca²⁺]_i at times > 2 min. While thapsigargin has been shown to induce a Na⁺ influx, as well as a Ca²⁺ influx, in neutrophils (24), the observations that either BAPTA-AM-loading of neutrophils or the simultaneous addition of EGTA and thapsigargin inhibited IL-8 production and mRNA synthesis indicates that the influx of Ca²⁺, not Na⁺ or another cation, triggers IL-8 production. The data obtained using EGTA also indicate that the release of Ca²⁺ from thapsigargin-sensitive intracellular stores only minimally triggers IL-8 production (Table IV). A sustained Ca²⁺ influx is necessary for maximal release as shown by the experiments in which addition of EGTA as late as 60 min after thapsigargin caused some inhibition of the IL-8 response. Similarly, the secretion of synthesized IL-8 was inhibited by the addition of EGTA, suggesting that the influx of Ca²⁺ was necessary for not only IL-8 synthesis but its secretion as well. These findings are similar to those in lymphocytes where a sustained elevation of [Ca²⁺]_i is required for the thapsigargin-induced expression of the early immune response genes such as IL-2, IL-3, and IL-4 (25).

In lymphocytes, thapsigargin-induced proliferation and differentiation are blocked by the immunosuppressive drug, cyclosporin A, which inhibits calcineurin, a Ca²⁺-activated serine/threonine phosphatase necessary for the nuclear transport of nuclear factor of activated T cells (NF-AT) (25). Increases in [Ca²⁺]_i activate calcineurin, leading to dephosphorylation of the cytosolic form of NF-AT and its migration to the nucleus where it is involved in the regulation of cytokine gene expression. In lymphocytes, this process can occur within minutes, resulting in the rapid up-regulation of gene expression.

The data indicate that thapsigargin-induced IL-8 production in neutrophils results in the activation of a biochemical pathway that is similar to that activated in thapsigargin-treated lymphocytes. Like lymphocytes, thapsigargin-induced up-regulation of the IL-8 gene in neutrophils has a rapid (within 5 min) time course, requires a prolonged increase in [Ca²⁺]_i levels, and is inhibited by ascomycin and cyclosporin A. These results suggest that calcineurin and NF-AT are involved in thapsigargin-induced IL-8 production in neutrophils. It is possible that inhibition of IL-8 synthesis is an integral part of the immunosuppressive action of cyclosporin A and ascomycin in vivo.

In conclusion, the production of IL-8 by peripheral blood neutrophils provides a paracrine mechanism to recruit more neutrophils into an acutely inflamed site. This mechanism can be triggered by the thapsigargin-induced release of intracellular Ca²⁺ stores and subsequent Ca²⁺ influx.

We thank Rhonda DaSilva and Debra Long Priel for their excellent technical assistance.