Platelet-activating Factor Primes Human Eosinophil Generation of Superoxide Edward M. Zoratti, Julie B. Sedgwick, Mary Ellen Bates, Rose F. Vrtis, Kristen Geiger, and William W. Busse Department of Medicine, Section of Allergy and Immunology, University of Wisconsin Medical School, Madison, Wisconsin
Platelet-activating factor (PAF) is a potent inflammatory mediator that can cause airway obstruction and hyperresponsiveness; these processes are also associated with pulmonary eosinophilia, suggesting a link between these two events. Thus, PAP's interaction with eosinophils may provide a mechanism for airway damage. However, direct in vitro activation of eosinophils by PAF requires concentrations that are likely higher than those achieved in vivo. As a result, we investigated whether lower, more physiologic concentrations of PAF could prime eosinophils for subsequent activation to another receptor-stimulated factor, in this case formylmethionylleucylphenylalanine (FMLP). To test this hypothesis, eosinophils were preincubated (1 and 15 min) with low concentrations of PAF (1 x 10-8 and 1 x lO-IOM); this exposure to PAF resulted in enhanced generation of superoxide anion to FMLP stimulation. Moreover, similar concentrations of PAF decreased eosinophil density and increased expression of cell surface CR3 receptors. Finally, low, nonactivating concentrations of PAF (1 x 10-10 to 1 X 10-8 M) caused transient increases in eosinophil cytosolic free Ca> concentrations. Collectively, these responses are consistent with the hypothesis that short-term exposure to low concentrations of PAF primes eosinophils to cause an enhanced inflammatory response upon subsequent activation to another receptor agonist. The consequences of this PAF-associated phenomenon can produce an enhanced inflammatory response and airway injury.
Platelet-activating factor (PAF) is generated by many cells, including mast cells, macrophages, neutrophils, and eosinophils (1-3). The pharmacologic properties of PAF suggest an important role in the pathophysiology of asthma (4, 5). For example, PAF can cause airway obstruction (6) and bronchial hyperresponsiveness (7). Interestingly, these effects are associated with airway eosinophilia and inflammation (8, 9), suggesting a possible interrelationship between the actions of PAF and eventual eosinophilic bronchial injury. There is, in fact, considerable evidence that PAF influences eosinophil function in the development of inflammation. Intradermal administration of PAF causes eosinophilic infiltration of the skin of atopic subjects (10). Furthermore, bronchial challenge of primates with PAF produces an eosinophilic accumulation in the lungs (11). In vitro, PAF is a potent chemoattractant for eosinophils, and increases eosinophil adhesion to vascular endothelium (12, 13). In addition, PAF stimulates eosinophil degranulation (14), prostanoid (Received in original form April 11, 1991 and in revisedform July 16. 1991) Address correspondence to: William W. Busse, M.D., University of Wisconsin Hospital, 600 Highland Avenue, H6/360 CSC, Madison, WI 53792. Abbreviations: bovine serum albumin, BSA; intracellular calcium concentration, [CaHI,; formylmethionylleucylphenylalanine, FMLP; Hanks' balanced salt solution, HBSS; superoxide, 0,; superoxide dismutase, SOD; platelet-activating factor, PAF; phorbol myristate acetate, PMA. Am. J. Respir. Cell Mol. BioI. Vol. 6. pp. 111O-106. 1992
formation (15), and superoxide (0,) generation (14). Finally, PAF-activated eosinophils cause significantly more damage to guinea pig respiratory epithelia than do nonactivated eosinophils (16). Thus, direct activation of eosinophils by PAF can contribute to airway inflammation. There is, however, reason to suspect that PAP's contribution to eosinophilic-derived bronchial inflammation is not limited to direct activation of the cell. First, the concentrations of PAF often required to activate eosinophils in vitro are high and may not reflect in vivo conditions (14). Second, although PAF is rapidly metabolized, changes in bronchial responsiveness to this agonist can persist beyond the initial response (7). Finally, in vitro exposure to low, nonactivating concentrations of PAF primes neutrophils; these primed cells have enhanced protein kinase activity (17), oxidative metabolism (18-21), aggregation (20), degranulation (20), and actin filament assembly (22). Based upon these observations with neutrophils, we evaluated whether low concentrations of PAF also prime eosinophil generation of O2 to another activator, the chemotactic peptide formylmethionylleucylphenylalanine (FMLP).
Materials and Methods Reagents Percoll was purchased from Pharmacia Fine Chemicals (Piscataway, NJ). Hanks' balanced salt solution (HBSS) and new-
Zoratti, Sedgwick, Bates et al.: PAF Primes Eosinophils
born calf serum were obtained from GIBCO (Grand Island, NY). Horse-heart ferricytochrome C (type Vl), Tris (hydroxymethyl) aminomethane (Tris), Triton, bovine serum albumin (BSA), sodium azide, phenylmethylsulfanyl fluoride, FMLP, i-o-pbosphatidylcholine-B acetyl-v-o-octadecy 1-9-cis-enyI (PAF; CI8:1), and superoxide dismutase (SOD) were purchased from Sigma Chemical Co. (St. Louis, MO). Indo I-AM was purchased from Molecular Probes Corp. (Eugene, OR). Stock solutions of FMLP (10-2 M) and cytochalasin B (10 mg/ml; Aldrich Chemical Co., Milwaukee, WI) were made in dimethyl sulfoxide and stored in aliquots at -70° C. Immediately before use, individual aliquots were thawed and adjusted to appropriate concentrations with HBSS. The concentration of dimethyl sulfoxide was always less than 0.1% and did not interfere with cell function. Stock solutions of PAF were made in phosphate-buffered saline with 0.2 % BSA and stored at -70° C. The following monoclonal antibodies to CR3 were also obtained: OKMI (Ortho, Raritan, NJ), anti-MAC-l (Hybritech, San Diego, CA), and anti-CD18 (Biodesign, Kennebunkport, ME). Fluorescein conjugated F (ab), fragments of goat and anti-mouse (or rat) IgG were obtained from Cappel (Cochraneville, PA) and mouse immunoglobulin standards were obtained from Pharmingen (San Diego, CA). Patient Population Blood for eosinophils was obtained from patients with either skin test-verified allergic rhinitis or documented asthma. All patients were between 19 and 42 yr of age, with an equal gender distribution. Medications for the treatment of asthma or allergic rhinitis were withheld for at least a 12-h period before blood drawing. None of the subjects received oral corticosteroid therapy for at least 4 wk before the time of study. Isolation of Eosinophil Suspensions Eosinophil isolation followed previously described methods (23). Briefly, EDTA-anticoagulated blood was sedimented in 4.5% dextran in Tris-buffered saline solution for 45 min at 22° C. The plasma-leukocyte suspension was removed, layered over Ficoll-Hypaque, and then centrifuged (400 x g for 15 min at 22° C). After centrifugation, the granulocyte pellet was washed twice, and resuspended in HBSS (with I mM Ca 2+) with 5 % newborn calf serum. Two milliliters of the granulocyte suspension (10 to 20 x lQ6 cells/ml) was carefully layered onto multiple discontinuous density Percoll gradients and centrifuged for 20 min at 700 x gat 22° C. The centrifuged cells formed five distinct density bands on the Percoll gradients; the band containing normal density eosinophils (density of 1.095 to 1.100 g/ml) (23) was carefully removed, and the cells were counted (Coulter Counter; Coulter Electronics, Hialeah, FL) and differentiated by Wright-stained cytospins. Residual red blood cells were lysed by hypotonic shock, and the purified (> 88 %) normaldensity eosinophil fraction was washed and resuspended in HBSS with 0.1% gelatin (HBSS/gel). O2 Generation Generation of O2 was measured as the SOD-inhibitable reduction of ferricytochrome C, as previously described (24). Briefly, using a 96-well microtiter plate (Immulon II;
101
Dynatech, Alexandria, VA), 100 p.l of eosinophil suspensions (I x ]06/ml) was added to a 100-p.l volume containing 200 p.m cytochrome C, 2 x 10-7 M FMLP, and 10 p.g/ml cytochalasin B in HBSS/gel. Wells without FMLP, or containing only the corresponding concentration of PAF, served as controls to measure spontaneous and PAF-induced O 2 production, respectively, Immediately after the addition of eosinophils, the absorbance in the individual reaction wells was measured at 550 nm in a Microplate Autoreader (EL 309; Bio-Tek Instruments, Burlington, VT), followed by repetitive readings over the next 50 min. Between readings, the microtiter plates were incubated in 5 % CO 2 at 3r c. Each reaction was performed in duplicate and against an identical control reaction, which contained 20 p.g/ml of SOD. O 2 generation was calculated with an extinct coefficient of 21.1 x ]03 mol/liter/crrr' as nmoles of cytochrome C reduced per 5 X 105 cells per time (minutes) minus SOD control and spontaneous O 2 generation. Eosinophil viability (by trypan blue exclusion) exceeded 95 % for all reaction conditions. In experiments to evaluate priming, eosinophils (1 x lQ6/ml) were preincubated with 1 x 10-8 or 1 X 10-10 M PAF (or HBSS/gel only, as a control) for I or 15 min at 37° C. Determination of Eosinophil Density after PAF Incubation Suspensions of normal density eosinophils (1.5 to 4 x 106 cells/ml) were incubated for 15 min at 3r C with buffer or PAF (1 x 10-6 to 1 X 10-8 M). Following incubation, the cells were layered onto discontinuous density Percoll gradients and centrifuged as described above. The cell bands at each density interface were collected and counted to determine the change in eosinophil density. The density distribution is expressed as a percentage of the total number of eosinophils recovered from each Percoll gradient. Determination of Eosinophil Surface Receptors Purified eosinophils were suspended in HBSS/gel and incubated (3r C) with buffer or PAF (1 x ]0-8 M) for 15 min. The cells were then washed and stained with monoclonal antibodies to CR3 for flow cytometry, according to the method of Berger and associates (25). Samples were analyzed in duplicate; for each treatment, a control tube was stained with an immunoglobulin standard of the appropriate subclass instead of monoclonal antibody to CR3. Stained cells were fixed overnight in 1% paraformaldehyde, washed once in HBSS, and resuspended in HBSS containing 0.1% BSA and 0.05% NaN 3 • Cell suspensions were kept on ice until evaluated by flow cytometer. Cell fluorescence was determined using a FACScan (Becton-Dickinson, San Jose, CA). Data on 10,000 cells were collected for each sample with contaminating erythrocytes excluded from analysis based on their light-scattering properties. Fluorescent channels were determined in a logarithmic mode and geometric means were calculated by Consort 30 software (Becton-Dickinson), Linear equivalents were calculated, and corresponding subclass controls were subtracted. The results of duplicate samples were averaged.
102
AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 6 1992
Measurement of Cytosolic Free Ca2+ Concentrations ([Ca H ];) Purified eosinophils were loaded with the fluorescent calcium indicator Indo I-AM (37 0 C X 30 min) at a cell concentration of 5 X IQ6 cells/ml (26). After the loading procedure, the cells were washed twice and resuspended in HBSS (with 1 mM Ca2+). Before the measurements of fluorescence, the cells (2 to 5 X 106/ml) were transferred to a cuvette and equilibrated to 37 0 C while magnetically stirred. Changes in fluorescence were measured with an SLM 8000 C spectrofluorometer (SLM Inc., Urbana, IL) using an excitation wavelength of 355 om. The ratio of fluorescence at 405 and 485 om was measured simultaneously. Maximal and minimal fluorescence values were determined following cell lysis in the presence of excess extracellular calcium (HBSS with 1 mM Ca H ) or absence of extracellular calcium (EGTA-chelated HBSS), respectively. Measurements of fluorescence were also determined after stimulation with PAF (1 x 10-11 to 1 X 10-6 M). [Ca H ] , was calculated by the following formula: [CaHJ
=
Kd( Rmax R- Rmin )( Sf - R Sb
2
30
A
PAF 1xl0- 10M PREINCUBATION
VJ
-'
1 MIN
-' w
15 MIN
0
20
ll)
0
x
L{)
-
a::: w >
80
a:::
~
(SJ
0
u w
-
PAF CONCENT RATION
60
_1000
I x IO- 6 M I x IO-7 M I x IO-8 M
::!:
.5
.J
,..---,
+
[=:J CONTROL
~
...J
0
TIME (SEC)
~ ~
w
I
Q.
20 40 60 80 100
NO
a.
40
500
Z
Vi
0
w
20
~
....
(n=2)
1.080
1.085
1.090
1.095
1.100
(n=2)
(n=5)
o DENSITY ( g/ml)
Figure 3. The effectsof PAF on eosinophil density. Normal dense eosinophils were incubated with PAF for 15 min at 37° C. The cells were then fractionated on Percoll density gradients. The percentage of total eosinophils recovered at each density interface is given. Values are mean ± SEM.
,I I o "Xlo- ll 10- 10
10-9
I 10- 8
I 10-7
I 10- 6
PAF 1M)
Figure 5. Effect of PAF on eosinophil intracellular calcium concentration ([CaH ];) . Inset: Representative [Ca"], response to different concentrations of PAF. Values are mean ± SEM. * P < 0.05 compared with baseline values.
104
AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 6 1992
dothelia (12, 28, 29), a phenomenon that becomes less prominent at higher concentrations (29). When purified suspensions of eosinophi1s are exposed to higher concentrations of PAF, 1 X 10-9 to 1 X 10-6 M, chemotaxis (13), prostanoid generation (15, 30), and degranulation (14) occur. Finally, 02" generation occurs when eosinophils are activated with micromolar concentrations of PAF (14). These direct cellular responses to PAF are likely to contribute to eosinophildependent inflammation. PAF has another important property that serves to promote granulocyte-dependent tissue injury: priming. To illustrate this phenomenon, Koenderman and co-workers (18) incubated human neutrophils with nonactivating concentrations of PAF, 1 i-tM, and observed a transient increase in [Ca2+], and enhanced oxygen consumption following subsequent activation by the FMLP. Furthermore, Vercellotti and colleagues (20) found that preincubation of neutrophi1s with 10-12 M PAF for 5 min enhanced the cell's respiratory burst, degranulation, and aggregation when stimulated by FMLP or phorbol myristate acetate (PMA). In neutrophils, priming by PAF was associated with both an enhanced expression of surface receptors, identified by OKM1 antibody, and increased [Ca2+J; PAF-primed cells were more effective in their ability to damage endothelial cells. Gay and associates (21) reported a synergistic enhancement of PMAactivated 02" generation and protein kinase C activity that was Ca2+ dependent (17). Finally, preincubation of neutrophils with 10-11 to 10-8 M PAF for 2 to 60 min increased FMLP-stimulated action on filament assembly (22). Although priming observed by Gay and associates (21) and Koenderman and co-workers (18)required higher concentrations (10-7 M or greater) of PAF than were used in either our experiment or that of Vercellotti and colleagues (20), their observations indicate that PAF primes human neutrophils in a concentration- and time-dependent manner. We found that PAF also primes isolated human eosinophi1s. When normal-density eosinophils were incubated with either 1 X 10-10 or 1 X 10-8 M PAF, 02" generation did not occur. However, when PAF-exposed eosinophils were subsequently incubated with a different activator, in our studies the chemotactic peptide FMLP, 02 generation was significantly enhanced. A number of phenotypic changes occurred in eosinophils exposed to 1 X 10-8 M PAF that are commonly associated with a cell that has undergone regulation: increased activity, change in density, and heightened expression of membrane receptors. When evaluated by density, function, or surface membranes, eosinophils are a heterogeneous collection of cells. Low-density eosinophils are phenotypically distinct from their normal-dense counterpart. First, low-density eosinophils have enhanced inflammatory properties, including antibody-dependent cellular cytotoxicity and leukotriene C. generation (31, 32). A number of factors have been shown to regulate cell density and function. For example, Rothenberg and co-workers and Owen and associates (33-35) found that cytokines, i.e., granulocyte/macrophage colony-stimulating factor, interleukin-3, and interleukin-5, decrease eosinophil density, prolong in vitro survival, and enhance selected cell functions. Our present report indicates that low doses of PAF also decrease eosinophil density and enhance functional activity (e.g., 02" generation). Therefore, the effect of low, yet
biologically relevant concentrations of PAF, may have similar effects on eosinophil function as has been noted with certain cytokines. Furthermore, we demonstrated that PAF increases the expression of CR3 receptors on eosinophils, a response previously noted with neutrophils (36) and another phenotypic characteristic of an upregulated eosinophil. CR3 receptors enhance binding to the opsonin complement particle C3bi, promote internalization of parasites, and also may playa role in adherence to endothelial cells (28); these receptor-linked functions are all instrumental to the eosinophil's ability to mount an inflammatory response. Concomitant PAF priming and enhanced CR3 receptor expression have been shown to occur in the neutrophil (37); we have now shown a parallel change in the eosinophil, suggesting that enhanced surface CR3 expression marks priming of this cell. Of the three individuals studied, only one showed even a slight increase in CD18 expression (Figure 4C). This change was very small compared with that observed when the 01 chain was studied (MAC-1 and OKM-1). The finding of unchanged ,6-chainexpression after PAF exposure remains unexplained. However, because the ,6chain is shared by several adhesion molecules (28), which may be unaffected by PAF priming, it is possible that our assay was not sufficiently sensitive to detect a small increase in the total cellular expression of CDI8. Nathan and colleagues (38) have reported that expression of CDU/CD18 adherence molecules is necessary for adherent neutrophil activation in response to cytokines. Adherent neutrophils from patients with leukocyte adhesion deficiency (LAD) who are deficient in CD18 were unable to generate hydrogen peroxide in response to cytokine stimulation. The binding of neutrophil CD11ICD18 to extracellular matrix proteins may promote responses to cytokines. In addition, the ability of adherent neutrophils to secrete hydrogen peroxide in response to cytokines, but not FMLP, was associated with cell spreading and blocked by dihydrocytochalasin B, suggesting the involvement of microfilaments (39, 40). Our findings that PAF can upregulate eosinophil adherence molecules suggest that PAF priming may facilitate both subsequent cell adherence and functional responses to cytokines such as tissue necrosis factor and colony-stimulating factors. FMLP was the only eosinophil activator used in this study because its function as an agonist and primer of neutrophils has been extensively studied. However, other agonists such as C5a and cytokines, as mentioned above, may have similar enhanced or desensitized responses following PAF incubation and need to be assessed. Similarly, the effect of PAF on activator receptor expression may be one mechanism of eosinophil priming that could effect both FMLP and cytokine responses. Finally, we established that priming concentrations of PAF, I X 10-10 and I x 10-8 M, transiently increased eosinophil [Ca2+]" even though these doses did not stimulate the respiratory burst. Furthermore, the time relationship of an intracellular rise in calcium was similar to that seen with eosinophil priming, as both occurred within I min of exposure to PAE Consequently, our findings parallel those of Koendermann and co-workers (18), who found that PAFprimed neutrophils also had a transient increase in [CaH ] , . Calcium transients partially, if not entirely, modulate the agonist concentration-dependent upregulation, or desensiti-
Zoratti, Sedgwick, Bates et al.: PAF Primes Eosinophils
zation, of key enzymes of the respiratory burst and degranulation (41, 42). We found that PAF not only caused an increase in [Ca2+], at high activating doses (1 ItM) but also at doses that primed cosinophils, 1 X 10-8 to 1 X 10-10 M. Although these lower doses may not initiate functional activity (i.c., 02" generation), their mechanism of priming may be related to an influx of extracellular or redistribution of intracellular Ca2+ in the eosinophil; furthermore, it is likely that the degree to which the [Ca2+J eventually increases determines whether activation (large increase in [Ca2+];) or priming (small rise in [Ca2+],) occurs. As previously noted by McPhail and associates (41), a single activator may provide signals for priming, activation, and downregulation, depending on its concentration. PAP's potentiation of eosinophil functions may have specific relevance to airway inflammation in asthma and atopic disease. In vivo, local effects of PAF in primate airways (11) and human skin (43) include eosinophil accumulation. Similarly, antigen challenge of atopic subjects results in a specific and dramatic airway infiltration by eosinophils (44). It is our hypothesis that recruitment of eosinophils to sites of antigen activation causes these cells to become functionally upregulated as a consequence of exposure to PAF, which is actively synthesized by a variety of inflammatory cells (1-3) and other cell-derived factors (i.e., cytokines). There is debate as to whether the levels of PAF (1 ItM) required for in vitro stimulation of 02" production by eosinophils are physiologically attainable; however, it is likely that nanomolar levels of PAF arc achieved in vivo and can prime eosinophils to enhance their inflammatory ability to cause airway injury. Although we have focused on the response to stimulation with a formyl peptide, studies with neutrophils suggest that primed cells undergo an exaggerated response to a heterologous collection of stimuli, such as concanavalin A, PMA, zymosan particles, calcium ionophore, C5a, and digitonin (17, 41, 45, 46). In summary, we have shown that PAF potentiates eosinophil 02" production at concentrations that cause a concomitant decrease in cell density as well as an increase in [Ca2+J and CR3 receptor expression. These phenotypic changes of the eosinophil represent a "primed" cell that can then generate an enhanced inflammatory response. If this occurs in the airway of the asthma patient, the result could be increased bronchial injury, obstruction, and hyperresponsivenessevents following the inhalation of PAF (6, 7). Our observations are further evidence that PAF can promote tissue injury through many mechanisms, including functional priming of eosinophil. These results further support a role for PAF and eosinophils in the pathogenesis of asthma. Acknowledgments: This work was supported by Public Health Service Grants AI23181, HL-44098, and AI-26609 from the National Institutes of Health.
105
4. 5. 6. 7.
8. 9. 10. I I.
12. 13. 14. 15.
16. 17. 18. 19. 20.
21. 22. 23. 24.
25. 26.
References 27. I. Lee, T.-C., D. J. Lenihans, B. Malone, L. L. Roddy, and S. I. Wasserman. 1984. Increased biosynthesis of platelet-activating factor in activated human eosinophils. J. Bioi. Chern. 259:5526-5530. 2. Arnoux, B., D. David, and J. Benveniste. 1980. Release of plateletactivating factor (PAF-acether) from alveolar macrophages by the calcium ionophore A23187 and phagocytosis. Eur. J. CUn. Invest. 10:437441. 3. Sanchez-Crespo, M., F. Alonso, and J. Figaro. 1980. Platelet-activating factor in anaphylaxis and phagocytosis. I. Release from human peripheral
28. 29. 30.
polymorphonuclear and monocytes during the stimulation by ionophore A23187 and phagocytosis but not from degranulating basophils. Immunology 40:645-655. Barnes, P. J. 1988. Platelet-activating factor and asthma. J. Allergy CUn. Immunol. 81:152-162. Page, C. P. 1988. The role of platelet-activating factor in asthma. J. ALlergy CUn. Immunol. 81:144-150. Rubin, A. E., L. J. Smith, and R. Patterson. 1987. The bronchoconstrictor properties of platelet activating factor in man. Am. Rev. Respir. Dis. 136:1145-1151. Cuss, F. M., D. M. S. Dixon, and P. J. Barnes. 1986. Effect of inhaled platelet activating factor on pulmonary function and bronchial responsiveness in man. Lancer 2:189-192. Godard, P., J. Bousquet, B. Lebel, and F. B. Michel. 1987. Bronchoalveolar lavage in the asthmatic. Bull. Eur. Physioparhol. Respir. 23:78-83. de Monchy, J. G. R., H. F. Kauffman, P. Vengeeral. 1985. Bronchoalveolar eosinophilia during allergen-induced late asthmatic reactions. Am. Rev. Respir. Dis. 131 :373-376. Archer, C. B., C. P. Page, W. Paul, 1. Morley, and D. M. MacDonald. 1984. Inflammatory characteristics of platelet activating factor (PAF acether) in human skin. Br. J. Dermarol. 110:45-50. Arnoux, B., A. Denjean, C. P. Page, D. Nolibe, J. Morley, and 1. Benveniste. 1988. Accumulation of platelets and eosinophils in baboon lung after PAF-acether challenge: inhibition by ketotifen. Am. Rev. Respir. Dis. 137:855-860. Kimani, G., M. G. Tonnesen, and P. M. Henson. 1988. Stimulation of eosinophil adherence to human vascular endothelial cells in vitro by plateletactivating factor. J. Immunol. 140:3161-3166. Wardlaw, A. J., R. Moqbel, O. Cromwell, and A. B. Kay. 1986. Platelet activity factor- potent chemotactic and chemokinetic factor for human eosinophils. J. CUn. Invest. 78:1701-1706. Kroegel, C., T. Yukawa, G. Dent., P. Venge, K. F. Chung, and P. J. Barnes. 1989. Stimulation of degranulation from human eosinophils by platelet activity factor. J. Immunol. 142:3518-3526. Bruynzeel, P. L., L. Koendermann, P. T. M. Kok, M. L. Hameling, and J. Verhagen. 1984. Platelet activating factor induced LTC. formation and luminol dependent chemiluminescence by human eosinophils. Pharmacol. Res. Commun. 18:61-69. Yukawa, T., R. C. Read, C. Kraegel et al. 1990. The effects of activated eosinophils and neutrophils on guinea pig airway epithelium in vitro. Am. J. Respir. Cell Mol. Bioi. 2:341-353. Gay, J. C., and E. S. Stitt. 1988. Enhancement of phorbol ester-induced protein kinase activity in human neutrophils by platelet-activating factor. J. Cell. Physiol. 137:439-447. Koenderman, L., M. Yazdombakksh, D. Roos, and A. J. Verhoeven. 1989. Dual mechanisms in priming of the chemoattractant-Induced respiratory burst in human granulocytes. J. Immunol. 142:623-628. Dewald, B., and M. Baggiolini. 1985. Activation of NADPH oxidase in human neutrophils: synergism between FMLP and the neutrophil products PAF and LTB•. Biochem. Biophys. Res. Commun. 128:297-304. Vercellotti, G. M., H. Q. Yin, K. S. Gustafson, R. D. Nelson, and J. S. Jacob. 1988. Platelet-activating factor primes neutrophil responses to agonists: role in promoting neutrophil-mediated endothelium damage. Blood 71:1100-1107. Gay, J. C., J. K. Beckman, K. A. Zabog, andJ. N. Lukens. 1986. Modulation of neutrophil oxidative responses to soluble stimuli by PAF. Blood 67:931-936. Shalit, M., G. A. Dabini, and F. S. Southwick. 1987. Platelet-activating factor both stimulates and "primes" human polymorphonuclear leukocyte actin filament assembly. Blood 70:1921-1927. Frick, W. E., J. B. Sedgwick, and W. W. Busse. 1989. The appearance of hypodense eosinophils in antigen dependent late phase asthma. Am. Rev. Respir. Dis. 139:1401-1406. Sedgwick, J. B., R. F. Vritis, M. F. Gourley, and W. W. Busse. 1988. Stimulus-dependent differences in superoxide anion generation by normal human eosinophils and neutrophils. J. Allergy Clin. Immunol. 81:876883. Berger, M., J. O'Shea, A. S. Cross etal. 1984. Human neutrophils increase expression of C3bi as well as C3b receptors upon activation. J. CUn. Invest. 74: 1566-1571. Grynkiewicz, G., M. Poenie, and R. Y. Tsien. 1985. A new generation of calcium indicators with improved fluorescent properties. J. Bioi. Chern. 260:3440-3450. Zoratti, E. M., J. B. Sedgwick, R. E. Vritis, and W. w. Busse. 1991. The effect of platelet activating factor on the generation of superoxide anion in human eosinophils and neutrophils. J. Allergy CUn. Immunol. (In press) Demers, P. A., and S. D. Wright. 1988. Adhesion promoting receptors on leukocytes. Current Opin. Immunol. 1:10-15. Dillon, P. K., M. F. Fitzpatrick, A. B. Rittar , and W. N. Duran. 1988. Effect of PAF on leukocyte adhesion to microvascular endothelium. Inflammation 12:563-573. Giembycz, M. A., C. Kroegel, and P. J. Barnes. 1990. Platelet activating
106
31.
32. 33. 34. 35.
36. 37.
AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 6 1992
factor stimulates cyclooxygenase activity in guinea pig eosinophils. J. Immunol. 144:3489-3497. Prin, L., M. Capron, A. B. Tonnel, O. Bletry, and A. Capron. 1983. Heterogeneity of human peripheral blood eosinophils: variability in cell density and cytotoxic ability in relation to the level and the origin of hypereosinophilia. Int. Arch. Allergy Appl. Immunol. 72:336-346. Kajita, T., Y. Yui, H. Mita et al. 1985. Release of leukotriene C 4 from human eosinophils and its relation to cell density. Int. Arch. Allergy Appl. Immunol. 78:406-410. Rothenberg. M. E., W. F. Owen, D. S. Silberstein et al. 1988. Human eosinophils have prolonged survival, enhanced functional properties and become hypodense when exposed to IL-3. J. CUn. Invest. 81: 1986-1992. Owen, W. F., M. E. Rothenberg, D. S. Silberstein etal. 1987. Regulation of human eosinophil viability, density and function by GM-CSF in the presence of 3T3 fibroblasts. J. Exp. Med. 166:129-141. Owen, W. F., Jr., J. Petersen, D. M. Sheff et al. 1990. Hypodense eosinophils and interleukin 5 activity in the blood of patients with the eosinophilia-myalgia syndrome. Proc. Natl. Acad. Sci. USA 87:86478651. Shalit,.M., C. V. Allmen, P. C. Atkin, and B. Zweiman. 1988. Platelet activating factor increases expansion of complement receptors on human neutrophils. J. Leukoc. Bioi. 44:212-217. Zimmerli, W., B. Selgrnann, and J. I. Gallin. 1986. Erudition primes human and guinea pig neutrophils for subsequent responsiveness to the chemotactic peptide nFMLP and increases complement component C3bi receptor expansion. J. CUn. Invest. 77:925-933.
38. Nathan, C. F., S. Srimal, C. Farberetal. 1989. Cytokine-induced respiratory burst of human neutrophils: dependence on extracellular matrix proteins and CDll/CDI8 integrins. J. Cell Bioi. 109:1341-1349. 39. Nathan, C. F. 1987. Neutrophil activation on biological surfaces: massive secretion of hydrogen peroxide in response to products of macrophages and lymphocytes. J. CUn. Invest. 80:1550-1560. 40. Nathan, C. F. 1989. Respiratory burst in adherent human neutrophils: triggering by colony-stimulating factors CSF-GM and eSF-G. Blood 73: 301-306. 41. McPhail, L. M., C. C. Clayton, R. Snyderman. 1984. The NADPH oxidase of human PMNs. J. Bioi. Chern. 259:5768-5775. 42. O'Flaherty, J. T., D. P. Jacobson, and J. F. Redman. 1989. Bidirectional effects of protein kinase C activators. J. Bioi. Chern. 264:6836-6843. 43. Archer, C. B., C. P. Page, J. Morley, and D. M. MacDonald. 1985. Accumulation of inflammatory cells in response to intracutaneous plateletactivating factor (PAF-acether) in man. Br. J. Dermato/. 112:285-290. 44. Sedgwick, J. B., W. J. Calhoun, G. J. Gleich et al. 1991. Immediate and late airway response of allergic rhinitis patients to segmental antigen challenge: characterization of eosinophil and mast cell mediators. Am. Rev. Respir. Dis. 144:1274-1281. 45. English, D., J. S. Roloff. and J. N. Lukens. 1981. Regulation of human PMN superoxide release by cellular responses to chemotactic peptides. J. Immunol. 126:165-171. 46. VanEpps, D. E.,andM. L. Garcia. 1980. Enhanccmcnt of ncutrophil function as a result of prior exposure to chemotactic factor. J. CUn. Invest. 66:167-175.
Platelet-activating factor (PAF) is a potent inflammatory mediator that can cause airway obstruction and hyperresponsiveness; these processes are also associated with pulmonary eosinophilia, suggesting a link between these two events. Thus, PAP's interaction with eosinophils may provide a mechanism for airway damage. However, direct in vitro activation of eosinophils by PAF requires concentrations that are likely higher than those achieved in vivo. As a result, we investigated whether lower, more physiologic concentrations of PAF could prime eosinophils for subsequent activation to another receptor-stimulated factor, in this case formylmethionylleucylphenylalanine (FMLP). To test this hypothesis, eosinophils were preincubated (1 and 15 min) with low concentrations of PAF (1 x 10-8 and 1 x lO-IOM); this exposure to PAF resulted in enhanced generation of superoxide anion to FMLP stimulation. Moreover, similar concentrations of PAF decreased eosinophil density and increased expression of cell surface CR3 receptors. Finally, low, nonactivating concentrations of PAF (1 x 10-10 to 1 X 10-8 M) caused transient increases in eosinophil cytosolic free Ca> concentrations. Collectively, these responses are consistent with the hypothesis that short-term exposure to low concentrations of PAF primes eosinophils to cause an enhanced inflammatory response upon subsequent activation to another receptor agonist. The consequences of this PAF-associated phenomenon can produce an enhanced inflammatory response and airway injury.
Platelet-activating factor (PAF) is generated by many cells, including mast cells, macrophages, neutrophils, and eosinophils (1-3). The pharmacologic properties of PAF suggest an important role in the pathophysiology of asthma (4, 5). For example, PAF can cause airway obstruction (6) and bronchial hyperresponsiveness (7). Interestingly, these effects are associated with airway eosinophilia and inflammation (8, 9), suggesting a possible interrelationship between the actions of PAF and eventual eosinophilic bronchial injury. There is, in fact, considerable evidence that PAF influences eosinophil function in the development of inflammation. Intradermal administration of PAF causes eosinophilic infiltration of the skin of atopic subjects (10). Furthermore, bronchial challenge of primates with PAF produces an eosinophilic accumulation in the lungs (11). In vitro, PAF is a potent chemoattractant for eosinophils, and increases eosinophil adhesion to vascular endothelium (12, 13). In addition, PAF stimulates eosinophil degranulation (14), prostanoid (Received in original form April 11, 1991 and in revisedform July 16. 1991) Address correspondence to: William W. Busse, M.D., University of Wisconsin Hospital, 600 Highland Avenue, H6/360 CSC, Madison, WI 53792. Abbreviations: bovine serum albumin, BSA; intracellular calcium concentration, [CaHI,; formylmethionylleucylphenylalanine, FMLP; Hanks' balanced salt solution, HBSS; superoxide, 0,; superoxide dismutase, SOD; platelet-activating factor, PAF; phorbol myristate acetate, PMA. Am. J. Respir. Cell Mol. BioI. Vol. 6. pp. 111O-106. 1992
formation (15), and superoxide (0,) generation (14). Finally, PAF-activated eosinophils cause significantly more damage to guinea pig respiratory epithelia than do nonactivated eosinophils (16). Thus, direct activation of eosinophils by PAF can contribute to airway inflammation. There is, however, reason to suspect that PAP's contribution to eosinophilic-derived bronchial inflammation is not limited to direct activation of the cell. First, the concentrations of PAF often required to activate eosinophils in vitro are high and may not reflect in vivo conditions (14). Second, although PAF is rapidly metabolized, changes in bronchial responsiveness to this agonist can persist beyond the initial response (7). Finally, in vitro exposure to low, nonactivating concentrations of PAF primes neutrophils; these primed cells have enhanced protein kinase activity (17), oxidative metabolism (18-21), aggregation (20), degranulation (20), and actin filament assembly (22). Based upon these observations with neutrophils, we evaluated whether low concentrations of PAF also prime eosinophil generation of O2 to another activator, the chemotactic peptide formylmethionylleucylphenylalanine (FMLP).
Materials and Methods Reagents Percoll was purchased from Pharmacia Fine Chemicals (Piscataway, NJ). Hanks' balanced salt solution (HBSS) and new-
Zoratti, Sedgwick, Bates et al.: PAF Primes Eosinophils
born calf serum were obtained from GIBCO (Grand Island, NY). Horse-heart ferricytochrome C (type Vl), Tris (hydroxymethyl) aminomethane (Tris), Triton, bovine serum albumin (BSA), sodium azide, phenylmethylsulfanyl fluoride, FMLP, i-o-pbosphatidylcholine-B acetyl-v-o-octadecy 1-9-cis-enyI (PAF; CI8:1), and superoxide dismutase (SOD) were purchased from Sigma Chemical Co. (St. Louis, MO). Indo I-AM was purchased from Molecular Probes Corp. (Eugene, OR). Stock solutions of FMLP (10-2 M) and cytochalasin B (10 mg/ml; Aldrich Chemical Co., Milwaukee, WI) were made in dimethyl sulfoxide and stored in aliquots at -70° C. Immediately before use, individual aliquots were thawed and adjusted to appropriate concentrations with HBSS. The concentration of dimethyl sulfoxide was always less than 0.1% and did not interfere with cell function. Stock solutions of PAF were made in phosphate-buffered saline with 0.2 % BSA and stored at -70° C. The following monoclonal antibodies to CR3 were also obtained: OKMI (Ortho, Raritan, NJ), anti-MAC-l (Hybritech, San Diego, CA), and anti-CD18 (Biodesign, Kennebunkport, ME). Fluorescein conjugated F (ab), fragments of goat and anti-mouse (or rat) IgG were obtained from Cappel (Cochraneville, PA) and mouse immunoglobulin standards were obtained from Pharmingen (San Diego, CA). Patient Population Blood for eosinophils was obtained from patients with either skin test-verified allergic rhinitis or documented asthma. All patients were between 19 and 42 yr of age, with an equal gender distribution. Medications for the treatment of asthma or allergic rhinitis were withheld for at least a 12-h period before blood drawing. None of the subjects received oral corticosteroid therapy for at least 4 wk before the time of study. Isolation of Eosinophil Suspensions Eosinophil isolation followed previously described methods (23). Briefly, EDTA-anticoagulated blood was sedimented in 4.5% dextran in Tris-buffered saline solution for 45 min at 22° C. The plasma-leukocyte suspension was removed, layered over Ficoll-Hypaque, and then centrifuged (400 x g for 15 min at 22° C). After centrifugation, the granulocyte pellet was washed twice, and resuspended in HBSS (with I mM Ca 2+) with 5 % newborn calf serum. Two milliliters of the granulocyte suspension (10 to 20 x lQ6 cells/ml) was carefully layered onto multiple discontinuous density Percoll gradients and centrifuged for 20 min at 700 x gat 22° C. The centrifuged cells formed five distinct density bands on the Percoll gradients; the band containing normal density eosinophils (density of 1.095 to 1.100 g/ml) (23) was carefully removed, and the cells were counted (Coulter Counter; Coulter Electronics, Hialeah, FL) and differentiated by Wright-stained cytospins. Residual red blood cells were lysed by hypotonic shock, and the purified (> 88 %) normaldensity eosinophil fraction was washed and resuspended in HBSS with 0.1% gelatin (HBSS/gel). O2 Generation Generation of O2 was measured as the SOD-inhibitable reduction of ferricytochrome C, as previously described (24). Briefly, using a 96-well microtiter plate (Immulon II;
101
Dynatech, Alexandria, VA), 100 p.l of eosinophil suspensions (I x ]06/ml) was added to a 100-p.l volume containing 200 p.m cytochrome C, 2 x 10-7 M FMLP, and 10 p.g/ml cytochalasin B in HBSS/gel. Wells without FMLP, or containing only the corresponding concentration of PAF, served as controls to measure spontaneous and PAF-induced O 2 production, respectively, Immediately after the addition of eosinophils, the absorbance in the individual reaction wells was measured at 550 nm in a Microplate Autoreader (EL 309; Bio-Tek Instruments, Burlington, VT), followed by repetitive readings over the next 50 min. Between readings, the microtiter plates were incubated in 5 % CO 2 at 3r c. Each reaction was performed in duplicate and against an identical control reaction, which contained 20 p.g/ml of SOD. O 2 generation was calculated with an extinct coefficient of 21.1 x ]03 mol/liter/crrr' as nmoles of cytochrome C reduced per 5 X 105 cells per time (minutes) minus SOD control and spontaneous O 2 generation. Eosinophil viability (by trypan blue exclusion) exceeded 95 % for all reaction conditions. In experiments to evaluate priming, eosinophils (1 x lQ6/ml) were preincubated with 1 x 10-8 or 1 X 10-10 M PAF (or HBSS/gel only, as a control) for I or 15 min at 37° C. Determination of Eosinophil Density after PAF Incubation Suspensions of normal density eosinophils (1.5 to 4 x 106 cells/ml) were incubated for 15 min at 3r C with buffer or PAF (1 x 10-6 to 1 X 10-8 M). Following incubation, the cells were layered onto discontinuous density Percoll gradients and centrifuged as described above. The cell bands at each density interface were collected and counted to determine the change in eosinophil density. The density distribution is expressed as a percentage of the total number of eosinophils recovered from each Percoll gradient. Determination of Eosinophil Surface Receptors Purified eosinophils were suspended in HBSS/gel and incubated (3r C) with buffer or PAF (1 x ]0-8 M) for 15 min. The cells were then washed and stained with monoclonal antibodies to CR3 for flow cytometry, according to the method of Berger and associates (25). Samples were analyzed in duplicate; for each treatment, a control tube was stained with an immunoglobulin standard of the appropriate subclass instead of monoclonal antibody to CR3. Stained cells were fixed overnight in 1% paraformaldehyde, washed once in HBSS, and resuspended in HBSS containing 0.1% BSA and 0.05% NaN 3 • Cell suspensions were kept on ice until evaluated by flow cytometer. Cell fluorescence was determined using a FACScan (Becton-Dickinson, San Jose, CA). Data on 10,000 cells were collected for each sample with contaminating erythrocytes excluded from analysis based on their light-scattering properties. Fluorescent channels were determined in a logarithmic mode and geometric means were calculated by Consort 30 software (Becton-Dickinson), Linear equivalents were calculated, and corresponding subclass controls were subtracted. The results of duplicate samples were averaged.
102
AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 6 1992
Measurement of Cytosolic Free Ca2+ Concentrations ([Ca H ];) Purified eosinophils were loaded with the fluorescent calcium indicator Indo I-AM (37 0 C X 30 min) at a cell concentration of 5 X IQ6 cells/ml (26). After the loading procedure, the cells were washed twice and resuspended in HBSS (with 1 mM Ca2+). Before the measurements of fluorescence, the cells (2 to 5 X 106/ml) were transferred to a cuvette and equilibrated to 37 0 C while magnetically stirred. Changes in fluorescence were measured with an SLM 8000 C spectrofluorometer (SLM Inc., Urbana, IL) using an excitation wavelength of 355 om. The ratio of fluorescence at 405 and 485 om was measured simultaneously. Maximal and minimal fluorescence values were determined following cell lysis in the presence of excess extracellular calcium (HBSS with 1 mM Ca H ) or absence of extracellular calcium (EGTA-chelated HBSS), respectively. Measurements of fluorescence were also determined after stimulation with PAF (1 x 10-11 to 1 X 10-6 M). [Ca H ] , was calculated by the following formula: [CaHJ
=
Kd( Rmax R- Rmin )( Sf - R Sb
2
30
A
PAF 1xl0- 10M PREINCUBATION
VJ
-'
1 MIN
-' w
15 MIN
0
20
ll)
0
x
L{)
-
a::: w >
80
a:::
~
(SJ
0
u w
-
PAF CONCENT RATION
60
_1000
I x IO- 6 M I x IO-7 M I x IO-8 M
::!:
.5
.J
,..---,
+
[=:J CONTROL
~
...J
0
TIME (SEC)
~ ~
w
I
Q.
20 40 60 80 100
NO
a.
40
500
Z
Vi
0
w
20
~
....
(n=2)
1.080
1.085
1.090
1.095
1.100
(n=2)
(n=5)
o DENSITY ( g/ml)
Figure 3. The effectsof PAF on eosinophil density. Normal dense eosinophils were incubated with PAF for 15 min at 37° C. The cells were then fractionated on Percoll density gradients. The percentage of total eosinophils recovered at each density interface is given. Values are mean ± SEM.
,I I o "Xlo- ll 10- 10
10-9
I 10- 8
I 10-7
I 10- 6
PAF 1M)
Figure 5. Effect of PAF on eosinophil intracellular calcium concentration ([CaH ];) . Inset: Representative [Ca"], response to different concentrations of PAF. Values are mean ± SEM. * P < 0.05 compared with baseline values.
104
AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 6 1992
dothelia (12, 28, 29), a phenomenon that becomes less prominent at higher concentrations (29). When purified suspensions of eosinophi1s are exposed to higher concentrations of PAF, 1 X 10-9 to 1 X 10-6 M, chemotaxis (13), prostanoid generation (15, 30), and degranulation (14) occur. Finally, 02" generation occurs when eosinophils are activated with micromolar concentrations of PAF (14). These direct cellular responses to PAF are likely to contribute to eosinophildependent inflammation. PAF has another important property that serves to promote granulocyte-dependent tissue injury: priming. To illustrate this phenomenon, Koenderman and co-workers (18) incubated human neutrophils with nonactivating concentrations of PAF, 1 i-tM, and observed a transient increase in [Ca2+], and enhanced oxygen consumption following subsequent activation by the FMLP. Furthermore, Vercellotti and colleagues (20) found that preincubation of neutrophi1s with 10-12 M PAF for 5 min enhanced the cell's respiratory burst, degranulation, and aggregation when stimulated by FMLP or phorbol myristate acetate (PMA). In neutrophils, priming by PAF was associated with both an enhanced expression of surface receptors, identified by OKM1 antibody, and increased [Ca2+J; PAF-primed cells were more effective in their ability to damage endothelial cells. Gay and associates (21) reported a synergistic enhancement of PMAactivated 02" generation and protein kinase C activity that was Ca2+ dependent (17). Finally, preincubation of neutrophils with 10-11 to 10-8 M PAF for 2 to 60 min increased FMLP-stimulated action on filament assembly (22). Although priming observed by Gay and associates (21) and Koenderman and co-workers (18)required higher concentrations (10-7 M or greater) of PAF than were used in either our experiment or that of Vercellotti and colleagues (20), their observations indicate that PAF primes human neutrophils in a concentration- and time-dependent manner. We found that PAF also primes isolated human eosinophi1s. When normal-density eosinophils were incubated with either 1 X 10-10 or 1 X 10-8 M PAF, 02" generation did not occur. However, when PAF-exposed eosinophils were subsequently incubated with a different activator, in our studies the chemotactic peptide FMLP, 02 generation was significantly enhanced. A number of phenotypic changes occurred in eosinophils exposed to 1 X 10-8 M PAF that are commonly associated with a cell that has undergone regulation: increased activity, change in density, and heightened expression of membrane receptors. When evaluated by density, function, or surface membranes, eosinophils are a heterogeneous collection of cells. Low-density eosinophils are phenotypically distinct from their normal-dense counterpart. First, low-density eosinophils have enhanced inflammatory properties, including antibody-dependent cellular cytotoxicity and leukotriene C. generation (31, 32). A number of factors have been shown to regulate cell density and function. For example, Rothenberg and co-workers and Owen and associates (33-35) found that cytokines, i.e., granulocyte/macrophage colony-stimulating factor, interleukin-3, and interleukin-5, decrease eosinophil density, prolong in vitro survival, and enhance selected cell functions. Our present report indicates that low doses of PAF also decrease eosinophil density and enhance functional activity (e.g., 02" generation). Therefore, the effect of low, yet
biologically relevant concentrations of PAF, may have similar effects on eosinophil function as has been noted with certain cytokines. Furthermore, we demonstrated that PAF increases the expression of CR3 receptors on eosinophils, a response previously noted with neutrophils (36) and another phenotypic characteristic of an upregulated eosinophil. CR3 receptors enhance binding to the opsonin complement particle C3bi, promote internalization of parasites, and also may playa role in adherence to endothelial cells (28); these receptor-linked functions are all instrumental to the eosinophil's ability to mount an inflammatory response. Concomitant PAF priming and enhanced CR3 receptor expression have been shown to occur in the neutrophil (37); we have now shown a parallel change in the eosinophil, suggesting that enhanced surface CR3 expression marks priming of this cell. Of the three individuals studied, only one showed even a slight increase in CD18 expression (Figure 4C). This change was very small compared with that observed when the 01 chain was studied (MAC-1 and OKM-1). The finding of unchanged ,6-chainexpression after PAF exposure remains unexplained. However, because the ,6chain is shared by several adhesion molecules (28), which may be unaffected by PAF priming, it is possible that our assay was not sufficiently sensitive to detect a small increase in the total cellular expression of CDI8. Nathan and colleagues (38) have reported that expression of CDU/CD18 adherence molecules is necessary for adherent neutrophil activation in response to cytokines. Adherent neutrophils from patients with leukocyte adhesion deficiency (LAD) who are deficient in CD18 were unable to generate hydrogen peroxide in response to cytokine stimulation. The binding of neutrophil CD11ICD18 to extracellular matrix proteins may promote responses to cytokines. In addition, the ability of adherent neutrophils to secrete hydrogen peroxide in response to cytokines, but not FMLP, was associated with cell spreading and blocked by dihydrocytochalasin B, suggesting the involvement of microfilaments (39, 40). Our findings that PAF can upregulate eosinophil adherence molecules suggest that PAF priming may facilitate both subsequent cell adherence and functional responses to cytokines such as tissue necrosis factor and colony-stimulating factors. FMLP was the only eosinophil activator used in this study because its function as an agonist and primer of neutrophils has been extensively studied. However, other agonists such as C5a and cytokines, as mentioned above, may have similar enhanced or desensitized responses following PAF incubation and need to be assessed. Similarly, the effect of PAF on activator receptor expression may be one mechanism of eosinophil priming that could effect both FMLP and cytokine responses. Finally, we established that priming concentrations of PAF, I X 10-10 and I x 10-8 M, transiently increased eosinophil [Ca2+]" even though these doses did not stimulate the respiratory burst. Furthermore, the time relationship of an intracellular rise in calcium was similar to that seen with eosinophil priming, as both occurred within I min of exposure to PAE Consequently, our findings parallel those of Koendermann and co-workers (18), who found that PAFprimed neutrophils also had a transient increase in [CaH ] , . Calcium transients partially, if not entirely, modulate the agonist concentration-dependent upregulation, or desensiti-
Zoratti, Sedgwick, Bates et al.: PAF Primes Eosinophils
zation, of key enzymes of the respiratory burst and degranulation (41, 42). We found that PAF not only caused an increase in [Ca2+], at high activating doses (1 ItM) but also at doses that primed cosinophils, 1 X 10-8 to 1 X 10-10 M. Although these lower doses may not initiate functional activity (i.c., 02" generation), their mechanism of priming may be related to an influx of extracellular or redistribution of intracellular Ca2+ in the eosinophil; furthermore, it is likely that the degree to which the [Ca2+J eventually increases determines whether activation (large increase in [Ca2+];) or priming (small rise in [Ca2+],) occurs. As previously noted by McPhail and associates (41), a single activator may provide signals for priming, activation, and downregulation, depending on its concentration. PAP's potentiation of eosinophil functions may have specific relevance to airway inflammation in asthma and atopic disease. In vivo, local effects of PAF in primate airways (11) and human skin (43) include eosinophil accumulation. Similarly, antigen challenge of atopic subjects results in a specific and dramatic airway infiltration by eosinophils (44). It is our hypothesis that recruitment of eosinophils to sites of antigen activation causes these cells to become functionally upregulated as a consequence of exposure to PAF, which is actively synthesized by a variety of inflammatory cells (1-3) and other cell-derived factors (i.e., cytokines). There is debate as to whether the levels of PAF (1 ItM) required for in vitro stimulation of 02" production by eosinophils are physiologically attainable; however, it is likely that nanomolar levels of PAF arc achieved in vivo and can prime eosinophils to enhance their inflammatory ability to cause airway injury. Although we have focused on the response to stimulation with a formyl peptide, studies with neutrophils suggest that primed cells undergo an exaggerated response to a heterologous collection of stimuli, such as concanavalin A, PMA, zymosan particles, calcium ionophore, C5a, and digitonin (17, 41, 45, 46). In summary, we have shown that PAF potentiates eosinophil 02" production at concentrations that cause a concomitant decrease in cell density as well as an increase in [Ca2+J and CR3 receptor expression. These phenotypic changes of the eosinophil represent a "primed" cell that can then generate an enhanced inflammatory response. If this occurs in the airway of the asthma patient, the result could be increased bronchial injury, obstruction, and hyperresponsivenessevents following the inhalation of PAF (6, 7). Our observations are further evidence that PAF can promote tissue injury through many mechanisms, including functional priming of eosinophil. These results further support a role for PAF and eosinophils in the pathogenesis of asthma. Acknowledgments: This work was supported by Public Health Service Grants AI23181, HL-44098, and AI-26609 from the National Institutes of Health.
105
4. 5. 6. 7.
8. 9. 10. I I.
12. 13. 14. 15.
16. 17. 18. 19. 20.
21. 22. 23. 24.
25. 26.
References 27. I. Lee, T.-C., D. J. Lenihans, B. Malone, L. L. Roddy, and S. I. Wasserman. 1984. Increased biosynthesis of platelet-activating factor in activated human eosinophils. J. Bioi. Chern. 259:5526-5530. 2. Arnoux, B., D. David, and J. Benveniste. 1980. Release of plateletactivating factor (PAF-acether) from alveolar macrophages by the calcium ionophore A23187 and phagocytosis. Eur. J. CUn. Invest. 10:437441. 3. Sanchez-Crespo, M., F. Alonso, and J. Figaro. 1980. Platelet-activating factor in anaphylaxis and phagocytosis. I. Release from human peripheral
28. 29. 30.
polymorphonuclear and monocytes during the stimulation by ionophore A23187 and phagocytosis but not from degranulating basophils. Immunology 40:645-655. Barnes, P. J. 1988. Platelet-activating factor and asthma. J. Allergy CUn. Immunol. 81:152-162. Page, C. P. 1988. The role of platelet-activating factor in asthma. J. ALlergy CUn. Immunol. 81:144-150. Rubin, A. E., L. J. Smith, and R. Patterson. 1987. The bronchoconstrictor properties of platelet activating factor in man. Am. Rev. Respir. Dis. 136:1145-1151. Cuss, F. M., D. M. S. Dixon, and P. J. Barnes. 1986. Effect of inhaled platelet activating factor on pulmonary function and bronchial responsiveness in man. Lancer 2:189-192. Godard, P., J. Bousquet, B. Lebel, and F. B. Michel. 1987. Bronchoalveolar lavage in the asthmatic. Bull. Eur. Physioparhol. Respir. 23:78-83. de Monchy, J. G. R., H. F. Kauffman, P. Vengeeral. 1985. Bronchoalveolar eosinophilia during allergen-induced late asthmatic reactions. Am. Rev. Respir. Dis. 131 :373-376. Archer, C. B., C. P. Page, W. Paul, 1. Morley, and D. M. MacDonald. 1984. Inflammatory characteristics of platelet activating factor (PAF acether) in human skin. Br. J. Dermarol. 110:45-50. Arnoux, B., A. Denjean, C. P. Page, D. Nolibe, J. Morley, and 1. Benveniste. 1988. Accumulation of platelets and eosinophils in baboon lung after PAF-acether challenge: inhibition by ketotifen. Am. Rev. Respir. Dis. 137:855-860. Kimani, G., M. G. Tonnesen, and P. M. Henson. 1988. Stimulation of eosinophil adherence to human vascular endothelial cells in vitro by plateletactivating factor. J. Immunol. 140:3161-3166. Wardlaw, A. J., R. Moqbel, O. Cromwell, and A. B. Kay. 1986. Platelet activity factor- potent chemotactic and chemokinetic factor for human eosinophils. J. CUn. Invest. 78:1701-1706. Kroegel, C., T. Yukawa, G. Dent., P. Venge, K. F. Chung, and P. J. Barnes. 1989. Stimulation of degranulation from human eosinophils by platelet activity factor. J. Immunol. 142:3518-3526. Bruynzeel, P. L., L. Koendermann, P. T. M. Kok, M. L. Hameling, and J. Verhagen. 1984. Platelet activating factor induced LTC. formation and luminol dependent chemiluminescence by human eosinophils. Pharmacol. Res. Commun. 18:61-69. Yukawa, T., R. C. Read, C. Kraegel et al. 1990. The effects of activated eosinophils and neutrophils on guinea pig airway epithelium in vitro. Am. J. Respir. Cell Mol. Bioi. 2:341-353. Gay, J. C., and E. S. Stitt. 1988. Enhancement of phorbol ester-induced protein kinase activity in human neutrophils by platelet-activating factor. J. Cell. Physiol. 137:439-447. Koenderman, L., M. Yazdombakksh, D. Roos, and A. J. Verhoeven. 1989. Dual mechanisms in priming of the chemoattractant-Induced respiratory burst in human granulocytes. J. Immunol. 142:623-628. Dewald, B., and M. Baggiolini. 1985. Activation of NADPH oxidase in human neutrophils: synergism between FMLP and the neutrophil products PAF and LTB•. Biochem. Biophys. Res. Commun. 128:297-304. Vercellotti, G. M., H. Q. Yin, K. S. Gustafson, R. D. Nelson, and J. S. Jacob. 1988. Platelet-activating factor primes neutrophil responses to agonists: role in promoting neutrophil-mediated endothelium damage. Blood 71:1100-1107. Gay, J. C., J. K. Beckman, K. A. Zabog, andJ. N. Lukens. 1986. Modulation of neutrophil oxidative responses to soluble stimuli by PAF. Blood 67:931-936. Shalit, M., G. A. Dabini, and F. S. Southwick. 1987. Platelet-activating factor both stimulates and "primes" human polymorphonuclear leukocyte actin filament assembly. Blood 70:1921-1927. Frick, W. E., J. B. Sedgwick, and W. W. Busse. 1989. The appearance of hypodense eosinophils in antigen dependent late phase asthma. Am. Rev. Respir. Dis. 139:1401-1406. Sedgwick, J. B., R. F. Vritis, M. F. Gourley, and W. W. Busse. 1988. Stimulus-dependent differences in superoxide anion generation by normal human eosinophils and neutrophils. J. Allergy Clin. Immunol. 81:876883. Berger, M., J. O'Shea, A. S. Cross etal. 1984. Human neutrophils increase expression of C3bi as well as C3b receptors upon activation. J. CUn. Invest. 74: 1566-1571. Grynkiewicz, G., M. Poenie, and R. Y. Tsien. 1985. A new generation of calcium indicators with improved fluorescent properties. J. Bioi. Chern. 260:3440-3450. Zoratti, E. M., J. B. Sedgwick, R. E. Vritis, and W. w. Busse. 1991. The effect of platelet activating factor on the generation of superoxide anion in human eosinophils and neutrophils. J. Allergy CUn. Immunol. (In press) Demers, P. A., and S. D. Wright. 1988. Adhesion promoting receptors on leukocytes. Current Opin. Immunol. 1:10-15. Dillon, P. K., M. F. Fitzpatrick, A. B. Rittar , and W. N. Duran. 1988. Effect of PAF on leukocyte adhesion to microvascular endothelium. Inflammation 12:563-573. Giembycz, M. A., C. Kroegel, and P. J. Barnes. 1990. Platelet activating
106
31.
32. 33. 34. 35.
36. 37.
AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 6 1992
factor stimulates cyclooxygenase activity in guinea pig eosinophils. J. Immunol. 144:3489-3497. Prin, L., M. Capron, A. B. Tonnel, O. Bletry, and A. Capron. 1983. Heterogeneity of human peripheral blood eosinophils: variability in cell density and cytotoxic ability in relation to the level and the origin of hypereosinophilia. Int. Arch. Allergy Appl. Immunol. 72:336-346. Kajita, T., Y. Yui, H. Mita et al. 1985. Release of leukotriene C 4 from human eosinophils and its relation to cell density. Int. Arch. Allergy Appl. Immunol. 78:406-410. Rothenberg. M. E., W. F. Owen, D. S. Silberstein et al. 1988. Human eosinophils have prolonged survival, enhanced functional properties and become hypodense when exposed to IL-3. J. CUn. Invest. 81: 1986-1992. Owen, W. F., M. E. Rothenberg, D. S. Silberstein etal. 1987. Regulation of human eosinophil viability, density and function by GM-CSF in the presence of 3T3 fibroblasts. J. Exp. Med. 166:129-141. Owen, W. F., Jr., J. Petersen, D. M. Sheff et al. 1990. Hypodense eosinophils and interleukin 5 activity in the blood of patients with the eosinophilia-myalgia syndrome. Proc. Natl. Acad. Sci. USA 87:86478651. Shalit,.M., C. V. Allmen, P. C. Atkin, and B. Zweiman. 1988. Platelet activating factor increases expansion of complement receptors on human neutrophils. J. Leukoc. Bioi. 44:212-217. Zimmerli, W., B. Selgrnann, and J. I. Gallin. 1986. Erudition primes human and guinea pig neutrophils for subsequent responsiveness to the chemotactic peptide nFMLP and increases complement component C3bi receptor expansion. J. CUn. Invest. 77:925-933.
38. Nathan, C. F., S. Srimal, C. Farberetal. 1989. Cytokine-induced respiratory burst of human neutrophils: dependence on extracellular matrix proteins and CDll/CDI8 integrins. J. Cell Bioi. 109:1341-1349. 39. Nathan, C. F. 1987. Neutrophil activation on biological surfaces: massive secretion of hydrogen peroxide in response to products of macrophages and lymphocytes. J. CUn. Invest. 80:1550-1560. 40. Nathan, C. F. 1989. Respiratory burst in adherent human neutrophils: triggering by colony-stimulating factors CSF-GM and eSF-G. Blood 73: 301-306. 41. McPhail, L. M., C. C. Clayton, R. Snyderman. 1984. The NADPH oxidase of human PMNs. J. Bioi. Chern. 259:5768-5775. 42. O'Flaherty, J. T., D. P. Jacobson, and J. F. Redman. 1989. Bidirectional effects of protein kinase C activators. J. Bioi. Chern. 264:6836-6843. 43. Archer, C. B., C. P. Page, J. Morley, and D. M. MacDonald. 1985. Accumulation of inflammatory cells in response to intracutaneous plateletactivating factor (PAF-acether) in man. Br. J. Dermato/. 112:285-290. 44. Sedgwick, J. B., W. J. Calhoun, G. J. Gleich et al. 1991. Immediate and late airway response of allergic rhinitis patients to segmental antigen challenge: characterization of eosinophil and mast cell mediators. Am. Rev. Respir. Dis. 144:1274-1281. 45. English, D., J. S. Roloff. and J. N. Lukens. 1981. Regulation of human PMN superoxide release by cellular responses to chemotactic peptides. J. Immunol. 126:165-171. 46. VanEpps, D. E.,andM. L. Garcia. 1980. Enhanccmcnt of ncutrophil function as a result of prior exposure to chemotactic factor. J. CUn. Invest. 66:167-175.
