Febrile-range hyperthermia accelerates caspase-dependent apoptosis in human neutrophils

Abstract

Human neutrophilic polymorphonuclear leukocytes (PMNs) are central to innate immunity and are responsible for clearance of pathogens. PMNs undergo a tightly regulated apoptosis program that allows for timely clearance of PMNs without extravasation of toxic intracellular contents. We investigated the rate of spontaneous apoptosis of human peripheral blood PMNs cultured at basal (37°C) and febrile-range (39.5°C) temperatures (FRT). We found that PMN apoptosis is accelerated at FRT, reaching ~90% completion by 8 h at 39.5°C vs. 18 h at 37°C based on morphologic criteria. Caspase-8 activation peaked within 15 minutes of PMN exposure to FRT and subsequent activation of caspase-3 and -9, cleavage of the BH3 only protein Bid, and mitochondrial release of cytochrome c were also greater in FRT-exposed PMNs. Inhibition of caspase-3, -8, and -9 conferred comparable protection from apoptosis in FRT-exposed PMNs. These results demonstrate that exposure to FRT enhances caspase-8 activation and subsequent mitochondrial-dependent and mitochondrial–independent apoptosis pathways. The PMN survival factors G-CSF, GM-CSF, and IL-8, each prolonged PMN survival at 37°C and 39.5°C, but did reduce the difference in survival at the two temperatures. In a mouse model of intratracheal endotoxin-induced alveolitis, co-exposure to FRT (core temperature ~39.5°C) doubled the proportion of bronchoalveolar PMNs undergoing apoptosis compared with euthermic mice. This process may play an important role in limiting inflammation and tissue injury during febrile illnesses.

Introduction

Neutrophilic polymorphonuclear leukocytes (PMNs) are phagocytic cells that constitute an integral component of innate immune defense (1). Early in the acute inflammatory response to infection and injury, PMNs migrate to sites of inflammation where they eliminate pathogens through phagocytosis and release of cytotoxic effector molecules into the phagolysosomes (2, 3). However, PMNs may also increase host collateral tissue injury through extravasation of these intracellular toxins into the extracellular microenvironment (4, 5) and by secreting inflammatory cytokines (6). The capacity of PMNs to injure tissue is limited, in part, by spontaneous, Fas-dependent PMN apoptosis (7, 8). Apoptosis not only allows PMNs to undergo cell death without loss of plasma membrane integrity and leak of intracellular contents into the extracellular microenvironment, but phagocytosis of apoptotic PMNs by macrophages reprograms the macrophage gene expression program to an anti-inflammatory profile (9). Thus, appropriate PMN apoptosis may be crucial for resolution of inflammation during infections (10, 11). Once PMNs have emigrated from the bone marrow, they survive for only 12-36 h (12). However, survival time is prolonged by exposure to soluble and cell-associated PMN-survival factors, many of which are released during inflammatory reactions (13-15).

Fever, a temporary, regulated increase in core temperature, is widely believed to confer cytoprotection, but the mechanisms underlying these effects are incompletely understood (16). Our laboratory has focused on the immunomodulatory effects of febrile-range hyperthermia (FRH) and the participation of elements of the heat shock response in mediating these effects. We have shown that exposing mice to FRH (core temperature ~39.5°C) accelerates pathogen clearance (17), but also increases PMN accumulation and collateral tissue injury, especially in the lungs (18). We have shown that FRH augments expression of neutrophil chemoattractants and increases endothelial capacity for transendothelial migration of neutrophils (19, 20).

These observations led us to hypothesize that exposing PMNs to FRH prolongs PMN survival, which contributes to PMN accumulation. We tested this hypothesis in freshly isolated human PMNs cultured in vitro, in the absence and presence of defined exogenous PMN survival factors. Surprisingly, we found that culturing human PMNs at 39.5°C greatly accelerated caspase-dependent apoptotic cell death, thereby identifying a potentially important mechanism that may limit collateral tissue injury during febrile illnesses.

Materials and Methods

Materials

Rabbit anti-human caspase-3 antibody was purchased from Cell Signaling Technology (Beverly, MA). Mouse monoclonal antibody recognizing the active cleavage product of human caspase-3 antibody was purchased from Chemicon (Temecula, CA). Rabbit antibody raised against amino acids 120-180 in human Bid was purchased from Bethyl Laboratories (Montgomery, TX). Mouse anti-human p38 MAPK antibody, goat anti-mouse IgG, HRP-conjugated, and goat anti-rabbit IgG HRP-conjugated were purchased from Santa Cruz Biotechnologies (Santa Cruz, CA). Peptide inhibitors of caspase-3 (Ac-DQMD-CHO), caspase-8 (Ac-IETD-CHO), and caspase-9 (Ac-LEHD-CHO) and rabbit polyclonal antibody raised against amino acids 120-180 of human Bid were purchased from Alexis Biochemicals (Lausen, Switzerland). HEPES, EGTA, EDTA, 4-(2-aminoethyl) benzenesulfonyl fluoride hydrochloride (AEBSF), leupeptin, and aprotinin, were from Sigma Chemical Company; St. Louis, MO). Recombinant human IL-8 (rhIL-8), rhGCSF, and rhGMCSF were purchased from BioSource (Camarillo, CA), Cell Sciences (Canton, MA), and GenScript (Piscataway, NJ), respectively.

Isolation and culture of PMN

All sample collection procedures were approved by and performed according to institutional guidelines (IRB/IACUC). Fifty milliliters of venous blood was collected from healthy subjects, anticoagulated with citrate, and PMNs were isolated by dextran sedimentation and density centrifugation through Histopaque 1083 (Sigma) at 20°C exactly as we have previously described (21). Contaminating erythrocytes were lysed by resuspending the pellet in 4°C water for 5 seconds, then adjusting the sodium chloride concentration to 150 mM with 3.5 M sodium chloride. The cells were collected by centrifugation at 700 × g for 15 min at 4°C, washed in 4°C Dulbecco's PBS, and suspended in RPMI 1640 containing 10% fetal bovine serum (FBS) at 3 × 10⁶ cells per ml. The resultant cell population comprised at least 95% PMNs based on their characteristic morphology. PMNs were cultured in polypropylene culture tubes and incubated in 5% CO₂ atmosphere in incubators set at either 37°C or 39.5°C with 0.1°C precision and re-calibrated prior to each experiment using an electronic thermometer (FLUKE Instruments model 5211, Everett, WA).

Morphologic analysis of apoptosis

Wright-Giemsa-stained cytospin preparations were analyzed blindly by one of us (AN) based on cell morphology. Apoptotic PMNs were identified based on replacement of the normal multi-segmented nuclear architecture with condensed, spherical nuclei (22).

Flow cytometric quantification of apoptosis by Terminal deoxynucleotidyl transferase biotin-dUTP Nick End Labeling (TUNEL) staining

An In Situ Cell Death Detection Kit (Roche, Indianapolis, IN) was used to fluorescein-label DNA strand breaks in PMNs according to the manufacturer's instructions. Positive controls were obtained by treating PMN with DNAse I (300 U/ml) for 10 min at room temperature. PMNs were analyzed with a FACScan flow cytometer (Beckton-Dickinson, Franklin Lakes, NJ).

Caspase activity assays

Caspase-3, -8, and -9 activities in PMN cell lysates were measured using fluorimetric activity measurement kits for caspase-3 [substrate: (Ac-DEVD-7-amino-4-methylcoumarin (AMC)], caspase-8 [substrate: IETD-(7-amino-4-trifluoromethyl coumarin), AFC] and caspase-9 (substrate: Ac-LEHD-AFC) from Calbiochem (La Jolla, CA) following the manufacturer's protocols. Briefly, to measure caspase-3 activity, PMNs were lysed in the supplier's lysis buffer after incubation at 37°C or 39.5°C. Lysates were incubated with the fluorimetric caspase-3 substrate Ac-DEVD-AMC for 2 h at 37°C and fluorescence was measured using excitation at 346 nm and measuring emission at 442 nm. To measure caspase-8 and-9, PMN lysates were incubated with the fluorimetric substrates IETD-AFC or LEHD-AFC, respectively, for 2 h at 37°C and fluorescence was measured using excitation at 400 nm and measuring emission at 505 nm. Caspase-3 activity in intact PMNs was measured using the FLICA Apoptosis Detection Kit (FAM-DEVD-FMK) from Immunochemistry Technologies, LLC (Bloomington, MN). PMNs were incubated at 37°C or 39.5°C for the indicated time, the FLICA substrate was then added, and cells were incubated at 37°C for an additional 4 h. Fluorescence was measured using excitation and emission wavelengths of 488 nm and 530 nm, respectively. To analyze the temperature-dependence of caspase-8 activity, the activity of recombinant caspase-8 was compared at 37°C and 39.5°C using a kit based on the colorimetric caspase-8 substrate Ac-IETD-pNA (CASP-8C, Sigma) according to the manufacturer's protocol using an ELISA plate reader equipped with a plate warmer (VERSAMax; Molecular Devices, Sunnyvale, CA) set at either 37°C or 39.5°C.

Western blotting

PMNs were lysed in RIPA buffer (50 mM TrisHCl pH7.4, 150 mM NaCl, 2 mM EDTA, 1% NP-40, 0.1% SDS) containing Protease Inhibitor Cocktail and Phosphatase Inhibitor Cocktail (both from Sigma), boiled for 5 min in loading buffer containing ß-mercaptoethanol, and resolved on 7.5, 10, 12, 15, or 10-20% Tris HCl gels (BioRad). The proteins were electrostatically transferred to Immobilon PVDF membranes, blocked with TBST (25mM Tris, 140mM NaCl, 3mM KCl, 0.05% Tween-20, pH 8.0) containing 5% nonfat milk, probed with primary antibodies, then with secondary peroxidase conjugates, and detected by chemiluminescence (Perkin-Elmer; Boston, MA) as we have described(23). The Western blots were imaged using a Fuji LAS-1000 gel documentation system and ImageQuant™ software.

Measurement of cytosolic cytochrome c

Following incubation, PMNs were washed twice with ice-cold PBS, resuspended in 100 μl of Mitochondria Buffer (20 mM HEPES (pH 7.5), 1 mM EGTA, 1 mM EDTA, 10 mM KCl, 1.5 mM MgCl₂, 1 mM dithiothreitol, 250 mM sucrose, 0.1 mM AEBSF [4-(2-aminoethyl) benzenesulfonyl fluoride hydrochloride], 2 μg/ml leupeptin, 2 μg/ml pepstatin, and 2 μg/ml aprotinin), and lysed using a microfuge tube-pestle homogenizer (Kontes) on ice. After centrifugation at 800 × g for 20 min at 4°C to remove cell debris, mitochondria and intact nuclei, supernatants were further clarified by centrifugation at 14,000 × g for 20 min at 4°C, and cytochrome c was analyzed using an ELISA kit (Calbiochem) and the manufacturer's protocol.

Temperature clamping and i.t. administration of LPS

Mice were adapted to standard plastic cages for at least 4 days before study. To avoid the influence of diurnal cycling, all experiments were started at approximately the same time each day (between 8:00 and 10:00 a.m.). Mice were placed in either 24°C (euthermia) or 34°C (FRH) infant incubators immediately after intratracheal instillation of 50 μg LPS (prepared by trichloroacetic acid extraction from Escherichia coli O111:B4 and 2,2,2-tribromoethyl; Sigma-Aldrich) in 50 μl PBS, exposures which we previously showed maintains core temperatures at approximately 37°C and 39.5°C, respectively (18). Twenty four hours after LPS instillation, the mice were euthanized by isoflurane inhalation and cervical dislocation and lung lavage was performed with a total of 2 ml PBS as previously described (18). Cells were collected and analyzed for morphologic features of apoptosis and TUNEL staining as described above for human PMNs except the cells were co-stained with Phycoerythrin-conjugated anti-Gr-1 (Serotec Biotechnology; Raleigh, NC; cat. no. MCA2387PE) to facilitate gating in the flow cytometric analysis. All protocols were approved by the Institutional Animal Care and Use Committee of the University of Maryland, Baltimore.

Statistical analysis

Data are displayed as mean±SE. Differences between two groups was analyzed using a Student t test. Differences among multiple groups was analyzed by applying The Tukey-Kramer honestly significant difference test to a one-way ANOVA using the JMP statistical software program (SAS Institute).

Results

Exposure to 39.5°C accelerates PMN apoptosis

We compared the kinetics of spontaneous apoptosis in human PMNs cultured at 37°C and 39.5°C in RPMI 1640 containing 10% FBS, but without other exogenous survival factors. The percent survival was calculated based on the characteristic morphologic features of apoptosis (22) (Figure 1A, B). PMNs in 39.5°C culture exhibited a much shorter lag time before onset of apoptosis, 4 h compared with 8 h in 37°C PMN cultures. Once apoptosis began, it preceded at similar rates at both temperatures. The times required for half the 39.5°C and 37°C PMNs to exhibit apoptotic morphology were approximately 6.5 h and 11.5 h, respectively. We analyzed the effect of incubation temperatures intermediate to 37° and 39.5°C on the rate of neutrophil apoptosis based on morphologic criteria and found the rank order of apoptosis rates to be 39.5°C > 39°C > 38°C > 37°C. (Figure 1C).

Figure 1. The effects of FRT on the kinetics of PMN apoptosis.

A, B. Morphologic analysis of apoptosis. 3 × 10⁶ PMN/ml were cultured at 37°C or 39.5°C in RPMI containing 10% FBS. A. Representative photomicrographs of Wright-Giemsa-stained PMNs after 6 h culture at either 37°C or 39.5°C. PMNs demonstrating apoptotic morphology are indicated by arrows. B. Time-course of PMN apoptosis based on morphology. The number of non-apoptotic PMNs in 37°C and 39.5°C culture was calculated by multiplying the total number of cells recovered by the non-apoptotic fraction and the percent survival calculated by dividing the number of surviving cells by the number of cells initially placed in culture. Data represent mean ± SE of 4 experiments. For all graphs in this figure, * denotes p < 0.05 vs. 37°C at the same time point and † denotes p < 0.05 vs. time 0. C. Neutrophils were incubated at 37°C, 38°C, 39°C, or 39.5°C and sequentially analyzed for morphologic features of apoptosis. The data from 4 experiments is displayed as in 1B. * denotes p < 0.05 vs. 37°C, 38°C, and 39°C, † vs. 37°C and 38°C, and ¶ vs. 37°C. D. Time-course of PMN apoptosis based on TUNEL staining. PMNs were incubated at 37°C or 39.5°C and analyzed at each time point by TUNEL staining and flow cytometry. The percentage of TUNEL-negative cells is shown at each time point. Data represent mean ± SE of 3 experiments. E. Proteolytic activation of pro-caspase-3. PMNs incubated at 37°C or 39.5°C were sequentially lysed and analyzed for loss of full-length caspase-3 cleavage by Western blotting. The density of the pro-caspase-3 bands was measured by direct imaging of the chemilluminescent signal and normalized to time 0 values. A representative blot and the mean ± SE of 3 experiments are shown. F. Generation of active caspase-3 was measured by Western blotting with an antibody against active caspase-3 and band intensities were normalized to time 0. A representative blot and the mean ± SE of 3 experiments are shown. G. Analysis of Caspase-3 activity in cell lysates. PMNs were incubated at 37°C or 39.5°C, sequentially lysed and the caspase-3 activity determined in a cell-free assay by measuring the generation of the fluorescent cleavage product from Ac-DEVD-AFC at 37°C. Mean ± SE of 3 experiments. H. Caspase-3 activity was measured in intact PMNs following 8h incubation at 37°C or 39.5°C by adding FLICA substrate, incubating at 37°C for an additional 4 h. and analyzing fluorescence. Mean ± SE of 3 experiments.

We confirmed the morphologic analysis of PMN apoptosis by TUNEL-staining and flow cytometry (Figure 1D). The appearance of TUNEL staining was delayed compared with onset of apoptotic morphology, but exhibited the same temperature-dependence, occurring much earlier in 39.5°C than in 37°C PMN cultures. The lag time until appearance of TUNEL staining was markedly shortened in the 39.5°C PMNs, but then proceeded at similar rates in both 37°C and 39.5°C cells. The time required for 50% of the PMNs to exhibit TUNEL staining was approximately 11 h in 39.5°C culture vs. 18 h in 37°C culture.

Exposure to 39.5ºC accelerates caspase-3 activation

Caspase-3, the predominant executioner caspase mediating spontaneous PMN apoptosis (24-26), is cleaved to its active form by the actions of the initiator caspases, caspase-8 and -9. The kinetics of caspase-3 cleavage were analyzed by measuring the loss of full-length caspase-3 by Western blotting for full-length caspase-3 (Figure 1E) or by Western blotting with an antibody that recognizes the active caspase-3 cleavage product (Figure 1F). Compared with PMNs in 39.5°C culture, PMNs at 37°C exhibit a 4 h delay before loss of full-length caspase-3 was detectable (Figure 1E). Active caspase-3 forms were first detectable after 6 h at 39.5°C compared with 16 h at 37°C (Figure 1F).

The Western blot analysis of caspase-3 activation was confirmed using a functional assay of caspase-3 in PMN cell lysates based on cleavage of a caspase-3-specific fluorimetric substrate (Ac-DEVD-AMC) (Figure 1G). This assay demonstrated a pattern that parallels onset of apoptotic morphology (Figure 1A). The lag time until caspase-3 activation was 4h in the 39.5°C PMN cultures vs. 12 h in the 37°C PMN cultures. The analysis of caspase-3 activity in PMN cell lysates was complemented by measuring caspase-3 activation in intact PMNs using an intracellular fluorimetric caspase-3 substrate, FAM-DEVD-FMK, added after 8h incubation at 37°C or 39.5°C, the time of maximal difference in morphologic evidence of apoptosis (Figure 1A). Caspase-3 activity increased 4.6-fold vs. baseline levels after 8 h incubation at 39.5°C, but did not change from baseline in PMNs incubated in parallel at 37°C (Figure 1H).

Exposure to 39.5°C accelerates caspase-8 and -9 activation and Bid cleavage

Spontaneous PMN apoptosis is known to be triggered by activation of Fas, which activates the initiator caspase, caspase-8, and triggers the extrinsic apoptosis pathway (7, 8). Caspase-8 can directly cleave pro-caspase-3 to its active form. Alternatively, caspase-8 can indirectly activate caspase-3 by activating the intrinsic apoptosis pathway through the sequential cleavage of Bid, induction of mitochondrial cytochrome c leak, and activation of caspase-9.

Since caspase-8 and -9 may be activated without detectable proteolytic cleavage (27, 28), we analyzed activation of caspase-8 and -9 in PMNs using functional assays with fluorimetric substrates (Ac-IETD-AFC for caspase-8; Ac-LEHD-AFC for caspase-9) (Figure 2) rather than by measuring cleavage by Western blotting. Caspase-8 activity peaked after 15 min incubation in 39.5°C PMNs at 2.3-fold above basal levels and at least a 2-fold increase was sustained for 90 min (Figure 2A). In contrast, caspase-8 activity in the PMNs cultured at 37°C were not statistically different than baseline levels at any time over an 18h incubation. Caspase-9 activity increased in PMNs cultured at both temperatures, but the increase in caspase-9 activity began earlier, 7 h vs. 12 h, and reached similar maximal 6-fold increases earlier, 7 vs. 18 h, at 39.5°C compared with 37°C cells (Figure 2B).

Figure 2. Analysis of Caspase-8 and -9 activity.

PMNs were incubated at 37°C or 39.5°C, sequentially lysed and the caspase-8 (A) and -9 (B) activity determined in a cell-free assay by measuring the generation of the fluorescent cleavage product from IETD-AFC and LEHD-AFC, respectively, at 37°C. Mean ± SE of 3 experiments is shown. * denotes p < 0.05 vs. 37°C at the same time point and † denotes p < 0.05 vs. time 0.

In addition to participating in the extrinsic apoptosis pathway by directly cleaving caspase-3 to its active form, caspase-8 can activate the intrinsic apoptosis pathway by cleaving the BH3 only protein, Bid (29). The C-terminal tBid fragment that is generated translocates to mitochondria and stimulates release of cytochrome c (30, 31), a stimulus for caspase-9 activation (32, 33). The sequential activation of caspase-8 and caspase-9 in the 39.5°C PMN cultures suggests that this pathway may be activated in PMNs exposed to FRT. To further analyze the potential participation of this pathway to PMN apoptosis at basal and febrile temperatures, we analyzed Bid cleavage in PMNs cultured at 37°C and 39.5°C by Western blotting and calculating the 15kDa tBid:20kDa full-length band density ratios for each lane (Figure 3A). The tBid:full-length Bid ratio increased gradually during the first 3h and more rapidly over the subsequent 3 to 6h in culture in the 39.5°C PMN cultures. In contrast, there was no detectable increase in tBid:full-length ratio during a 6h incubation at 37°C. To confirm that Bid cleavage resulted in mitochondrial cytochrome c release, we analyzed cytosolic cytochrome c levels in PMNs cultured at 37°C and 39.5°C (Figure 3B). After 3h incubation, cytochrome c levels in mitochondria-free cytosol were 2-fold higher in PMNs cultured for 39.5°C than in cells cultured at 37°C.

Figure 3. Proteolytic cleavage of Bid and release of cytochrome c.

A. Bid cleavage: 3 × 10⁶ PMNs incubated at 37°C or 39.5°C were sequentially lysed and analyzed for Bid cleavage by Western blotting. The density of the full-length (p20) and tBid (p15) bands was measured by direct imaging of the chemilluminescent signal and plotted as the mean ± SE of tBid:full-length Bid ratio. A. Representative Western blot and mean ± SE of 3 experiments are shown. * denotes p < 0.05 vs. 37°C at the same time point and † denotes p < 0.05 vs. time 0. B. Cytochrome c release: 3 × 10⁶ PMN/ml were cultured at 37°C or 39.5°C in RPMI containing 10% FBS. Cells were lysed in mitochondria buffer and centrifuged to remove cell debris as well as intact mitochondria. Cytochrome c was assayed from the mitochondria-free supernatants using ELISA. Mean ± SE of 3 experiments. * denotes p < 0.05 vs. 37°C at the same time point and † denotes p < 0.05 vs. time 0.

Caspase-3, -8, and -9 inhibitors each reduce PMN apoptosis at both basal and febrile levels

To evaluate the contribution of extrinsic and intrinsic apoptosis pathways to PMN apoptosis at basal and febrile temperatures, we analyzed the anti-apoptotic effects of caspase inhibitors in 37°C and 39.5°C PMN cultures (Figure 4). Freshly isolated PMNs were treated with 10 μM concentration of each inhibitor, incubated at either 37°C or 39.5°C, and the kinetics of apoptosis analyzed based on morphologic criteria. Our preliminary dose-response studies demonstrated that inhibitor concentrations greater than 10 μM increased cell death in human PMNs at both temperatures. Compared with vehicle-treated controls, each of the three inhibitors caused similar delays in appearance of apoptotic morphology in the 37°C PMN cultures, extending the time required for 50% apoptosis from 16 h to 19.5-21.5 h. Like the 37°C cells each of the three inhibitors caused similar delays in apoptosis in the 39.5°C PMN cultures, extending the time required for 50% apoptosis from 3 h to 8 h. at 39.5°C. To confirm the anti-apoptotic effects of the caspase inhibitors, we analyzed TUNEL expression in PMNs cultured with each of the three caspase inhibitors at 37°C for 18h or at 39.5°C for 6h. The three inhibitors exhibited similar anti-apoptotic activity in the 37°C and 39.5°C PMN cultures, increasing survival 1.9 to 2.9-fold compared with DMSO-treated control cells.

Figure 4. Effect of caspase inhibitors on PMN.

3 × 10⁶ PMN/ml were cultured at 37°C (A) or 39.5°C (B) in RPMI containing 10% FBS and DMSO (vehicle control) or peptide inhibitor of caspase-3, -8, or -9, apoptosis was analyzed based on morphologic criteria, and the percent survival calculated at each time point. Mean ± SE of three experiments. * denotes p < 0.05 for all three treatment groups vs. DMSO-treated cells at the same time point. C. PMNs treated with DMSO or each of the caspase inhibitors were analyzed after 18 h at 37°C or 6 h at 39.5°C by TUNEL staining and flow cytometry. The percentage of TUNEL-negative cells is shown; mean ± SE of 3 experiments. * denotes p < 0.05 vs. control for each temperature.

The pro-survival cytokines GCSF, GMCSF, and IL-8 prolong PMN survival at both basal and febrile levels

Treatment with IL-8 (10 ng/ml), G-CSF (500 U/ml), and GM-CSF (500 U/ml) increased the 50% survival time of PMNs in from 12 h to 20.5 h, 37 h, and 38.5 h, respectively, in 37°C culture (Figure 5A), and from 6.5 h to 10 h, 11.5 h, and 14 h, respectively, in 39.5°C culture (Figure 5B). An analysis of active caspase-3 levels using the fluorimetric substrate, Ac-DEVD-AMC, confirmed the morphologic analysis (Figure 5C). IL-8, G-CSF, and GM-CSF reduced active caspase-3 levels by 83%, 74%, and 88% in 37°C PMN cultures and by 62%, 92%, and 88% in 39.5°C PMN cultures after 24 h in culture.

Figure 5. Effect of cytokines IL-8, GCSF and GMCSF on PMN apoptosis at 37°C and 39.5°C.

3 × 10⁶ PMN/ml were cultured at 37°C (A) or 39.5°C (B) in RPMI containing 10% FBS, with or without IL-8 (10 ng/ml), G-CSF (500 U/ml), or GM-CSF (500 U/ml), apoptosis was sequentially analyzed based on morphologic criteria, and the percent survival calculated as described in figure 1. Mean ± 4 experiments. * denotes p < 0.05 vs. 37°C at the same time point and † denotes p < 0.05 vs. time 0. (C) PMNs were incubated at 37°C or 39.5°C with or without IL-8, G-CSF, or GM-CSF for 24h, cells were lysed, active caspase-3 levels were measured by immunoblotting with an antibody against active caspase-3, and band densities were normalized to time 0 levels. Mean ± SE of 3 experiments. † denotes p < 0.05 vs. untreated control cells cultured at the same temperature.

Effect of Hyperthermia on PMN apoptosis in an in vivo lung injury model

To analyze the effect of hyperthermia on PMN apoptosis in a clinically-relevant in vivo model, we analyzed the proportion of PMNs undergoing apoptosis in a previously described intratracheal LPS-challenged mouse model of pneumonia (18). We have previously shown that co-exposing mice to FRH by increasing ambient temperature from 25° to 34°C increased PMN accumulation in the bronchoalveolar compartment. In this study, we exposed mice to FRH for 24h beginning immediately after intratracheal instillation of 50 μg LPS, then collected cells from lung lavage and analyzed them for TUNEL staining by flow cytometry and morphometric features of apoptosis by light microscopy (Table 1). We found a similar effect of hyperthermia on PMN apoptosis in vivo as we found in human PMNs in vitro. When measured 24h after instillation of LPS, approximately twice as many PMNs exhibited evidence of apoptosis in the warmer mice. The low numbers of PMNs in lung lavage did not permit a similar analysis in the absence of LPS challenge.

Table 1.

Effect of in vivo hyperthermia on apoptosis of lung PMNs in LPS-challenged mice^a.

Treatment	% TUNEL-positiveb	% apoptotic morphologyb
Euthermia	2.22 ± 0.05	11.8± 0.2
Hyperthermia	3.79 ± 0.45c	29.2 ± 2.7d

^aMice received 50 μg LPS in 50 μl PBS via intratracheal instillation, then were maintained with core temperature 37°C (Euthermia) or 39.5°C (Hyperthermia) for 24h. Lung lavage was collected and analyzed for apoptosis by TUNEL staining using flow cytometry and gating on GR-1-staining cells and for morphologic features of apoptosis.

^bData from 4 mice per group are expressed as mean ± SE percent PMNs that express TUNEL staining or apoptotic morphology.

^cp < 0.05 vs. Euthermia

^dp < 0.03 vs. Euthermia

Discussion

We have previously demonstrated that co-exposure to FRH increases PMN accumulation in mouse models of pneumonia and pulmonary oxygen toxicity (18, 20). Additional data in these mouse models and in in vitro cell culture models demonstrated that FRH exerts multiple effects that contribute to enhanced PMN recruitment, including: (1) induction of G-CSF expression and expansion of the circulating PMN pool (34), (2) augmented expression of CXC chemokines (18, 20), and (3) increased endothelial capacity to facilitate PMN transmigration (19). In the present study, we tested the hypothesis that FRH would additionally contribute to PMN accumulation by prolonging PMN survival. We compared the kinetics of apoptosis in human PMNs cultured at 37°C and 39.5°C using apoptotic morphology, TUNEL staining, and caspase-3 activation as criteria. Characteristic morphologic changes, including cell rounding, shrinkage, blebbing, nuclear condensation, and chromatin fragmentation remain the most specific criteria for identifying apoptotic cell death (35-38). Since PMNs have multisegmented nuclei, nuclear condensation is easily assessed by light microscopy in these cells and correlates well with other characteristics of apoptotic death (22). We found that human PMNs incubated at 37°C in medium containing bovine serum without defined exogenous survival factors completed spontaneous apoptosis within 22 h. This spontaneous apoptosis rate is within the range of spontaneous PMN apoptosis rates reported by other laboratories (7, 14, 25, 39, 40). We found that the development of apoptotic morphology in the 37°C PMN cultures was accompanied by caspase-3 activation and was modestly attenuated by treatment with the caspase-3 inhibitory peptide, Ac-DQMD-CHO. These findings are also similar to those of previous studies of spontaneous apoptosis in human PMNs (25, 41). However, some investigators (40) have reported that treatment with pan-caspase inhibitors fails to block apoptosis. Collectively these data suggest that caspase-dependent mechanisms may play only a minor role in spontaneous PMN cell death at 37°C.

Surprisingly, we found that the appearance of apoptotic morphology and TUNEL-staining was greatly accelerated in PMNs incubated at 39.5°C and was similar to the reported rate of apoptosis in PMNs treated with TNF-α, a potent pro-apoptotic agonist (42). We used multiple complementary assays to demonstrate that PMNs incubated at 39.5°C, exhibited greatly enhanced and accelerated activation of caspase-3 compared with 37°C PMN cultures. Full-length caspase-3 is cleaved to an active fragment by caspase-8 or –9. The time required for 50% reduction of full-length caspase-3 was 8h in 39.5°C cells vs. 14h in 37°C cells. Utilizing an antibody specific for the active caspase-3 fragment, we showed generation of the active caspase-3 fragment is detectable by 8h at 39.5°C, but not until 16h 37°C. Casapase-3 activity in PMN lysates reached 50% maximal levels after only 4h in 39.5°C cells vs. 9h in 37°C cells. Using an independent measurement of caspase-3 activity in intact PMNs with a cell-permeable fluorimetric substrate, we showed that capsase-3 activity increased by 4.6-fold after 8h incubation at 39.5°C but had not yet increased above baseline levels in PMNS cultured for 8h at 37°C. These independent measures of caspase-3 activation each show a similar pattern in which caspase-3 activation in PMNs occurs more rapidly at 39.5°C than 37°C.

Although caspase-3 activation and onset of TUNEL staining, and nuclear condensation occurred much earlier in 39.5°C PMN cultures, the caspase-3 inhibitory peptide, Ac-DQMD-CHO, exerted similar anti-apoptotic activity in 37°C and 39.5°C PMN cultures. Collectively, these data indicate that the enhanced apoptosis caused by exposure to 39.5°C is caused, at least in part, by accelerated activation of caspase-3. This distinguishes the effects of FRT from the accelerated PMN death triggered by TNF-α, which proceeds in the presence of caspase inhibitors and lacks the classic morphologic features of apoptosis (42). Previous studies have demonstrated that the caspase-dependent apoptosis that does occur in cultured human PMNs is triggered in part by autocrine activation of the death domain receptor, Fas, and subsequent activation of caspase-8 (7, 8). Scaffidi et al.(43) described two pathways through which active caspase-8 causes caspase-3 activation. Type I cells exclusively use the extrinsic pathway in which capase-8 directly cleaves pro-caspase-3 to its active form, whereas type II cells utilize an alternative pathway in which caspase-8 activates the intrinsic apoptosis pathway by cleaving the pro-apoptotic Bcl family member, Bid. The C-terminal, BH3-containing Bid fragment generated by the actions of caspase-8, tBid, translocates to the mitochondria and triggers efflux of mitochondrial cytochrome c (29, 44). Subsequent studies of fas-dependent apoptosis have demonstrated that Bid cleavage, mitochondrial leak of cyctochrome c, and activation of caspase-9 occur in both type I and II cells, but that caspase-8 activation occurs more rapidly in type I cells (45). In the current study, caspase-8 was maximally activated within 15 min of warming PMNs to 39.5°C. Generation of tBid and release of mitochondrial cytochrome c were each detectable within 3 h of warming to 39.5°C. Caspase-9 activity occurred after 7 h of warming but an upward trend in activity was evident within 5 h of warming. Caspase-3 was activated after 4 h, and morphologic characteristics of apoptosis appeared after 6 h of warming PMNs to 39.5°C. Treatment with the caspase-8 inhibitory peptide, Ac-IETD-CHO, reduced apoptosis in the 39.5°C PMN cultures as effectively as caspase-3 inhibition. In contrast, in 37°C PMN cultures, an increase in caspase-8 activation was not detectable during an 18 h incubation and activation of caspase-9 and caspase-3 and appearance of apoptosis did not occur until 12 h in culture. Treatment with peptide inhibitors of caspase-3, -8, and –9 each exerted similar anti-apoptotic effects in 39.5°C and 37°C PMN cultures. Collectively, these data suggest that accelerated apoptosis in 39.5°C PMN cultures occurs, in part, via enhanced activation of caspase-8 and activation of type II apoptosis signaling pathway.

Although apoptosis is classically directed by activation of caspases, caspase-independent cell death with some features of apoptosis has been observed in several cell lines and may participate significantly to spontaneous death of PMNs cultured at 37°C (46) Furthermore, some pro-apoptotic agonists, such as TNF-α, have been shown to activate caspase-independent death pathways in PMNs (42). While our results do not rule out smaller contributions by caspase-independent pathways, the protection conferred by caspase inhibitors in against early onset of apoptosis in 39.5°C PMN cultures suggest that caspase-dependent apoptosis pathways are enhanced and predominate in PMNs exposed to FRT.

We extended the studies in human PMNs in vitro to a clinically relevant mouse model of pneumonia and lung injury. The mouse model not only analyzes the effect of hyperthermia on PMN apoptosis in vivo, but analyzes PMNs that have emigrated from the vasculature into tissue, a process that can modify PMN apoptosis (47). When analyzed 24 h after LPS challenge, the proportion of PMNs at each temperature that exhibited TUNEL staining was much lower than the proportion of PMNs that exhibited morphologic features of apoptosis. A similar relationship between these two measures of apoptosis was found in our analysis of human PMNs in which morphologic features of apoptosis preceded TUNEL staining by several hours (compare figures 1B and D). However, utilizing either TUNEL staining or morphologic criteria, we found that the proportion of lung lavage PMNs exhibiting features of apoptosis was approximately twice as great in the FRH-exposed mice. The difference is proportion of apoptotic PMNs in the warmer mice at this time point is even more meaningful when one considers the effect of FRH on the kinetics of PMN accumulation in lung. We previously showed that the total number of PMNs in lung lavage plateau between 6 and 24 h after LPS challenge in eutherrmic mice, but PMNs continue to accumulate between over this time in hyperthermic mice. This suggests that the PMNs recovered from the hyperthermic mice not only contain a greater proportion of apoptotic cells, but that these PMNs are younger compared with PMNs from euthermic mice. Taken together, these data suggest that the PMNs in the hyperthermic mice undergo earlier apoptosis.

Based on our previous in vivo studies of mouse lung injury demonstrating enhanced PMN accumulation in FRT-exposed animals (18, 20), we predicted that PMN survival would be prolonged at FRT. A possible explanation for the disparity between our previous in vivo studies and the present study is that the enhanced generation of PMN survival factors that is stimulated by FRT might compensate for the direct pro-apoptotic effects of FRT on PMNs. We have shown that mice exposed to FRT increase expression of several known PMN survival factors, including G-CSF (34), GM-CSF (18), and CXC chemokines (18). In the present study, we showed that supplementing 39.5°C PMN cultures with exogenous G-CSF, GM-CSF, or the principle human CXC chemokine, IL-8, restored PMN survival to 37°C levels. However, each of these cytokines exhibited greater anti-apoptotic effects in 37°C than 39.5°C PMN cultures. Since, our analysis of PMN apoptosis in the in vivo LPS challenged mouse model revealed a higher proportion of apoptotic PMNs in the lungs of hyperthermic than euthermic mice, we conclude that the enhanced expression of PMN survival factors that occurs in the warmer mice (18, 20), does not appear to compensate for the direct pro-apoptotic effects of hyperthermia.

In summary, we have shown that caspase-dependent apoptosis of human PMNs is accelerated at FRT (39.5°C) and that the effect is mediated by enhanced activation of caspase-8. The combined effect of FRT to enhance PMN recruitment (18, 20) and accelerate PMN apoptosis will result in accumulation of a younger PMN population at sites of infection and more rapid clearance of older PMNs through the apoptotic pathway. This will enable rapid pathogen elimination, minimize dysregulated release of toxic contents (7, 8) from older PMNs, and trigger the reprogramming of macrophages to an anti-inflammatory cytokine expression profile (9). Such effects of accelerated PMN apoptosis may improve the balance between pathogen clearance and collateral tissue injury. However, the consequences of these effects of hyperthermia on PMN apoptotic signaling have not yet been elucidated in vivo.