Kitagawa, Yuko, Stephan F. Van Eeden, Darlene M. Redenbach, Maleki Daya, Blair A. M. Walker, Maria E. Klut, Barry R. Wiggs, and James C. Hogg. Effect of mechanical deformation on structure and function of polymorphonuclear leukocytes. J. Appl. Physiol. 82(5): 1397–1405, 1997.—The present studies were designed to test the hypothesis that mechanical deformation of polymorphonuclear leukocytes (PMN) leads to functional changes that might influence their transit in the pulmonary capillaries. Human leukocytes were passed through 5- or 3-μm-pore polycarbonate filters under controlled conditions. Morphometric analysis showed that the majority of PMN were deformed and that this deformation persisted longer after filtration through 3-μm filters than through 5-μm filters (P < 0.05) but did not result in the cytoskeletal polarization characteristic of migrating cells. Flow cytometric studies of the filtered PMN showed that there was a transient increase in the cytosolic free Ca2+ concentration after both 3- and 5-μm filtration (P< 0.01) with an increase in F-actin content after 3-μm filtration (P < 0.05). AlthoughL-selectin expression on PMN was not changed by either 5- or 3-μm filtration, CD18 and CD11b were increased by 3-μm filtration (P < 0.05). Priming of the PMN withN-formyl-methionyl-leucyl-phenylalanine (0.5 nM) before filtration resulted in an increase of CD11b by both 5 (P < 0.05)- and 3-μm (P < 0.01) filtration. Neither 5- nor 3-μm filtration induced hydrogen peroxide production. We conclude that mechanical deformation of PMN, similar to what occurs in the pulmonary microvessels, induces both structural and functional changes in the cells, which might influence their passage through the pulmonary capillary bed.
- adhesion molecules
the pulmonary microvessels restrict the passage of polymorphonuclear leukocytes (PMN) because of the discrepancy between the size of the PMN and the size of the lung capillary segments (5, 14,15). Although erythrocytes [red blood cells (RBC)] have similar maximum dimensions to PMN, their greater deformability allows them to negotiate these restrictions more quickly (14). This results in pulmonary capillary transit times that are 60–100 times longer for PMN than for RBC, and this concentrates PMN with respect to RBC in pulmonary capillaries (14, 15).
Stimuli such as peptide chemoattractants, endotoxin, and smoking are known to decrease PMN deformability and further increase the concentration of PMN in pulmonary capillaries (6, 18, 27, 30). Several in vitro filtration studies have measured the biophysical properties of PMN, particularly their size and deformability, and have shown that these factors determine the magnitude of PMN sequestration in the pulmonary microvessels (6, 18, 27, 30). The importance of the discrepancy between PMN and pulmonary capillary dimensions in causing PMN sequestration in the lung has also been demonstrated during forced expiratory maneuvers in which raised intra-alveolar pressure compresses alveolar microvessels and delays PMN (21). Studies in humans undergoing cardiac catheterization show that alveolar compression results in an immediate arteriovenous difference of PMN across the lung that reverses when alveolar pressure is returned to normal (21). Mechanical deformation of PMN has been demonstrated in pulmonary microvessels (5), and Wiggs et al. (29) have shown that PMN of ∼7 μm in diameter negotiate the capillary bed in the same way as 4-μm nondeformable beads. These findings indicate that the PMN deform in the pulmonary capillary bed, but whether this deformation is an active or passive process remains to be determined.
Our working hypothesis is that the mechanical deformation of PMN in the pulmonary capillaries may induce functional changes that could influence their passage through these restrictions. To test this hypothesis, we used an in vitro filtration system with polycarbonate membranes to simulate the in vivo mechanical deformation of PMN. This system has been used by others to study the deformability of PMN (6, 8,10, 11, 18, 23, 27, 30), and we have modified it to study the effect of PMN deformation on their structure and function. We measured the deformation that occurs as the cells pass through the polycarbonate filters and assessed cell polarity for evidence of active locomotion. Flow cytometry was used to determine whether this mechanical deformation produced changes in the conversion of globular actin (G-actin) to filamentous actin (F-actin), cytoplasmic free Ca2+ concentration ([Ca2+]i), the expression of cell adhesion molecules, and degranulation with oxygen radical production.
Human leukocyte-rich plasma (LRP) was prepared from citrate-anticoagulated venous blood obtained from healthy volunteers (n = 4) by dextran (molecular wt 100–200,000, 1.7% final concentration, Sigma Chemical, St. Louis, MO) sedimentation of RBC in PMN buffer [(in mM) 138 NaCl, 27 KCl, 8.1 Na2HPO4 ⋅ 7H2O, 1.5 KH2PO4, and 5.5 glucose, pH 7.4]. PMN were purified from the resulting LRP. Briefly, LRP was centrifuged with hypotonic lysis of residual RBC and diluted in sterile water with 2× phosphate-buffered saline (PBS; 2× PBS is 27 mM Na2HPO4, 132 mM KH2PO4, and 2.74 M NaCl). PMN were then separated from the mononuclear cells by centrifugation in Histopaque (Sigma Chemical), with a density of 1.077 g/ml at 150 g for 13 min and resuspended in Hanks’ balanced salt solution (HBSS; StemCell Technologies, Vancouver, BC, Canada). The isolated PMN were 95–98% pure, with a viability of 98% as assessed by trypan blue exclusion.
In Vitro Filtration of PMN
LRP or purified PMN were filtered in vitro according to the filtration method described by Lennie et al. (20). Briefly, a 20-ml polypropylene syringe (Sherwood Medical, St. Louis, MO) was filled with sample solution (LRP or purified PMN) and filtered through polycarbonate filters (Poretics, Livermore, CA) with defined pore size [pore diameter 5 μm, length 10 μm, and pore density 4 × 105/cm2or pore diameter 3 μm, length 9 μm, and pore density 2 × 106/cm2(manufacturer’s data)] by using a syringe-infusion pump (pump 22, Harvard Apparatus, Millis, MA), which provides a constant flow rate of solution across the filter. Hydrostatic pressure was continuously monitored upstream from the filter by using a pressure transducer (Validyne Engineering, Northridge, CA) connected to a recording system. The pressure-sensing system was calibrated by using a water manometer under conditions of no flow before each filtration. Cell viability after filtration was 98% as tested by trypan blue exclusion.
Morphometric Analysis of PMN
LRP (4 × 105 cells/ml) prepared by the methods described above was filtered through polycarbonate filters with pore sizes of 5 and 3 μm at two different flow rates (3.0 or 1.0 ml/min). Samples were fixed in paraformaldehyde (final 1.6%) before filtration, immediately after (time 0), and 10, 30, 60, and 180 s after filtration. Cells allowed to settle were stained with Giemsa in suspension. The degree of deformation was quantified by dividing the longest by the shortest diameter (perpendicular at midpoint of longest diameter) to obtain a long-to-short ratio (L/S) in 20 randomly selected PMN per time point (n = 4). These measurements were made on a standardized image-analysis system (Bioquant System IV, R&M Biometrics, Nashville, TN).
Analysis of PMN Polarity
The distribution of microtubules and F-actin was determined in PMN after either 3- or 5-μm filtration at 3.0 ml/min at time = 0, 10, 30, 60, or 180 s (n = 3). The PMN were double stained with anti-tubulin (5A6 mouse monoclonal antibody raised against α-tubulin, gift from Dr. David Brown, final dilution 1:4,000) (16) to localize microtubules or fluorescein isothiocyanate (FITC)-conjugated phalloidin (1:200 dilution; Molecular Probes, Eugene, OR) to localize F-actin. Experiments were carried out with materials in suspension at room temperature, except if otherwise noted. Cells were permeabilized with a mixture of methanol-acetone (1:1) at −20°C for 10 min, followed by sequential incubation in 2 μMl-α-lysophosphatidylcholine palmitoyl (Sigma Chemical) and 0.2% Triton X-100 in PBS for 15 min at 37°C. Non-specific binding was blocked with 5% normal goat serum in 0.1% bovine serum albumin (BSA)/PBS for 30 min at 37°C. Samples were incubated in 200 μl of anti-tubulin antibody 5A6 diluted in 1% normal goat serum in 0.1% BSA/PBS for 60 min at 37°C, followed by tetramethylrhoadamine isothiocyanate (TRITC)-conjugated goat anti-mouse immunoglobulin (Ig) G secondary antibody for 60 min at 37°C (final dilution 1:200). For F-actin, cells were incubated with FITC-conjugated phalloidin for 20 min (final dilution 1:200) at 37°C. Cell suspensions were rinsed three times in 500 μl of 0.1% BSA/PBS between all steps. Cells were visualized on a Zeiss Axiophot microscope equipped for epifluorescence with filters for FITC (excitation 450–490; barrier 510; emission 515–556 nm) and TRITC (excitation 546; barrier 580; emission 590 nm).
Cytoskeletal polarity was evaluated without knowledge of the sample source. Twenty cells were scored from each time point. Cells were classified visually as either “deformed” or “round.” The microtubule organizing center (MTOC) was scored as either “polarized,” when microtubules converged at any pole of the nucleus of a deformed PMN, or “nonpolar” when they did not. Cells were scored for F-actin polarity as either “nonpolarized” (symmetrical distribution) or polarized (asymmetrical distribution) to identify F-actin-rich lamellapodia or filipodia as an indication of active cell migration.
F-actin content of PMN.
F-actin content of PMN was determined immediately after filtration by using a modification of a previously described method (17). Leukocytes filtered through polycarbonate filters (5- or 3-μm pore) at a flow rate of 3.0 ml/min were fixed immediately for 30 min at room temperature with 3% paraformaldehyde in PBS (pH 7.3) and washed for 5 min with PBS. The cells were simultaneously permeabilized and stained, in the dark at 37°C for 30 min, with a fresh mixture of 2 μMl-α-lysophosphatidylcholine palmitoyl (Sigma Chemical) and 0.1 μg/ml N -(7-nitrobenz-2-oxa-1,3-diazol-4-yl)-phallacidin (NBD-phallacidin; Molecular Probes). After staining, samples were washed and resuspended in PBS. F-actin content was measured with a flow cytometer (model Profile Epics II, Coulter Electronics, Hialeah, FL). Analysis gates for PMN were established by using distinctive forward and side scatter profiles and were expressed as mean fluorescent intensity (log) of 3,000–6,000 cells. As a positive control, PMN were stimulated withN-formyl-methionyl-leucyl-phenylalanine (FMLP; 100 nM, Sigma Chemical) for 10 s, fixed with 3% paraformaldehyde, and analyzed as above.
Changes in [Ca2+]iof PMN were determined by using flow cytometry with a fluorescent Ca2+ indicator, fluo 3-aminomethyl ester (fluo 3-AM; Molecular Probes), at excitation and emission wavelengths of 488 and 525 nm, respectively (24). Purified PMN were suspended in HBSS (4 × 105 cells/ml) and incubated with 2.2 μM fluo 3-AM for 30 min in the dark at room temperature. Cells were washed twice and resuspended in HBSS. Before the filtration or stimulation, baseline [Ca2+]iwas measured in triplicate. Changes in fluo 3 fluorescence, representing [Ca2+]i, were measured for up to 3 min after filtration through 5- or 3-μm-pore filters (flow rate 3.0 ml/min) or after addition of stimuli. Ionomycin (10 μM, Calbiochem, La Jolla, CA) and 100 nM FMLP stimulation were used as positive controls.
Expression of l-selectin, CD11b, and CD18 on PMN.
After filtration of leukocytes (4 × 105 cells/ml) through 5- or 3-μm-pore filters by using a flow rate of either 1.0 or 3.0 ml/min, the expression of L-selectin and CD11b/CD18 was measured by indirect immunofluorescence. One hundred microliters of prefiltered or filtered LRP were incubated with 200 μl of PBS (pH 7.3) containing the primary antibodies against human L-selectin (1 μg/ml, Leu-8, Becton-Dickinson, San Jose, CA), CD11b (1.6 μg/ml, DAKO-CD11b, Dakopatts, Glostrup, Denmark), or CD18 (6.9 μg/ml, DAKO-CD18, Dakopatts) for 10 min in the dark at room temperature. LRP incubated with mouse IgG1 (2 μg/ml, Sigma Chemical) served as negative control. Cells were washed with PBS and incubated with FITC-conjugated anti-mouse IgG (5 μg/ml, Sigma Chemical) as a secondary antibody for 10 min at room temperature in the dark. The remaining RBC in the samples were lysed (Immunolyse and Coulter Clone, Coulter Electronics), and the leukocytes were fixed with 1% paraformaldehyde and stored at 4°C. The immunofluorescence was measured by using flow cytometry as described above. Cells incubated with 100 nM FMLP or 1 μM phorbol 12-myristate 13-acetate (PMA; Sigma Chemical) served as positive controls. To test the effect of mechanical deformation of primed PMN on the expression of adhesion molecules, PMN exposed to priming doses of FMLP (0.5 nM) were filtered through the polycarbonate filters (5- or 3-μm pore) by using a 3.0 ml/min flow rate. The expression ofL-selectin and CD11b was determined by immunofluorescence as described above. The changes in fluorescence after filtration on primed PMN were compared with those of untreated PMN.
Hydrogen peroxide (H2O2) production by PMN.
The respiratory burst activity of PMN after mechanical deformation was determined by measuring PMN H2O2production according to the method previously described by Bass et al. (4). Briefly, PMN are loaded with dichlorofluorescin-diacetate (DCFH-DA), which diffuses into the cells and is hydrolyzed intracellularly to 2′,7′-dichlorofluorescin (DCFH), a nonfluorescent compound. H2O2produced during the respiratory burst oxidizes the DCFH to highly fluorescent 2′,7′-dichlorofluorescein (DCF), which can be measured by flow cytometry. LRP was incubated with PMN buffer consisting of 5 μM DCFH-DA (prediluted in 95% ethanol; final concentration of ethanol was 0.1%) and 0.1% gelatin and horizontal agitation in a shaking water bath at 37°C and was kept at this temperature for the whole experiment. Baseline DCF fluorescence was measured as well as fluorescence of PMN after filtration through 5- or 3-μm-pore filters for up to 40 min. As positive controls, 100 nM FMLP and 1 μM PMA (Sigma Chemical) were used.
Changes in deformation and polarity between PMN filtered through 3- and 5-μm filters and over time were analyzed by using χ2. Recovery of cell shape to prefilter shape was examined by comparison of deformation at each time point with prefilter values.
Differences between prefiltered or prestimulated control values for each condition were treated as individual parametric values, and groups of these values were compared with postfiltered or poststimulated values by using paired t-tests. Changes over time were analyzed by using analysis of variance for repeated measures. Corrections for multiple tests and comparisons were performed by using the sequential rejective Bonferroni procedure (18). Corrected P < 0.05 were considered significant. All values are expressed as the means ± SE.
Pressures and Shear Stress During PMN Filtration
The pressures 1 min after starting filtration of the LRP at a concentration of 4 × 105cells/ml were 0.6 ± 0.2 (1.0 ml/min) and 3.9 ± 0.7 cmH2O (3.0 ml/min) through 5-μm pores and 5.6 ± 1.3 (1.0 ml/min) and 22.5 ± 3.7 cmH2O (3.0 ml/min) through 3-μm pores. We calculated the shear stress on leukocytes at the wall by using the method of Downey and Worthen (10). In this technique the wall shear stress (τω) is related to the diameter (d) and length (l) of the tube as well as the pressure drop across the tube (ΔP) by the relationship τω = [ΔPd/(4l)]. The calculated shear stress in the 5-μm pores with a length of 10 μm was 487.5 dyn/cm2 at high flow rates and 75 dyn/cm2 for the low flow rates.
Morphometric Analysis of PMN
Figure 1 shows the representative features of PMN deformed by filtration through 5- or 3-μm-pore polycarbonate filters. The filtered PMN have elongated shapes, and their nuclei were often located off center after 3-μm filtration. Exposure to the same pressures without filtration had no effect on cell shape. Figure2 shows the degree of PMN deformation measured by L/S. This ratio was greater for PMN filtered through 3- than through 5-μm polycarbonate filters at each time point (P < 0.01). PMN also remained deformed for a longer period when filtered through 3- than through 5-μm filters (Fig. 2).
Analysis of PMN Polarity
Table 1 shows the fractions of deformed PMN, MTOC polarization, and F-actin redistribution. Samples at different time points after 3- and 5-μm filtration at 3.0 ml/min were pooled (n = 3). These data show that 78 and 87% of cells were deformed immediately after 3- and 5-μm filtration, respectively, and recovery of spherical shape was significantly delayed in 3-μm compared with 5-μm filtration at each time point. There was an increased brightness in F-actin staining only at t = 0 s in the 3-μm filtered cells. Although many cells were greatly deformed, there was no immunocytochemical evidence of microtubular or microfilament reorganization or polarization characteristic of migrating cells.
F-actin content of PMN.
Filtration of PMN through 3-μm-pore filters increased their F-actin content by 94 ± 44% from prefiltered values (P < 0.05,n = 4) immediately after filtration (t = 0 s), which was similar to that seen with FMLP stimulation (83 ± 17%,P < 0.01,n = 4, Fig.3). Filtration through 5-μm filters did not change F-actin content of PMN at t= 0 s. The exposure of PMN to the pressure required to force them through the filters had no effect on their F-actin content.
Free [Ca2+]iin PMN.
Filtration of PMN through both 5- and 3-μm pores caused an increase in cytosolic free [Ca2+]i(Fig.4 A). The degree of this increase was greater in 3-μm filtration than in 5-μm filtration (P < 0.05,n = 4) but was smaller than the value obtained with FMLP or ionomycin stimulation (P < 0.01,n = 4). Unlike the immediate increase of [Ca2+]iin PMN stimulated by FMLP or ionomycin, the peak increase of [Ca2+]iin filtered PMN was observed 30 s after filtration (Fig.4 B). The increase in [Ca2+]iwas sustained for up to 3 min after stimulation by ionomycin, in contrast to a rapid decline to baseline after 3 min in the [Ca2+]isignal after FMLP stimulation or filtration of PMN.
To determine whether [Ca2+]iincrease by filtration induces cross-desensitization to activation by FMLP, PMN were sequentially activated by either filtration or 100 nM FMLP, followed by 100 nM FMLP at 3 min. Three minutes after stimulation, fluo 3 fluorescence was back to baseline, and this time point was selected for the second stimulation. Figure5 A shows that sequential stimulation of PMN after initial FMLP stimulation did not result in a second increase in [Ca2+]i. In contrast, FMLP stimulation after 3-μm filtration induced a significant second [Ca2+]isignal (Fig. 5 B).
Expression of L-selectin, CD11b, and CD18 on PMN.
Changes in the surface expression of the cell adhesion moleculesL-selectin and CD11b/CD18 on PMN after deformation by filtration through 5- or 3-μm-pore filters are shown in Fig. 6. PMN filtration through 5- or 3-μm-pore filters did not change the expression ofl-selectin, in contrast to stimulation with either 100 nM FMLP or 1 μM PMA (Fig.6 A). CD11b and CD18 were significantly upregulated by 3-μm filtration (Fig. 6,B andC) by using both 1.0 and 3.0 ml/min flow rate (n = 4). Five-micrometer filtration did not change the expression level of CD11b/CD18 on PMN (Fig. 6, B andC). The exposure to the pressure required to pass the PMN through the filters did not affect expression of L-selectin or CD11b/CD18 (data not shown).
Neither a suboptimal (priming) dose of 0.5 nM FMLP nor 5-μm filtration alone induced upregulation of CD11b on PMN, as shown in Fig.7. However, PMN primed with 0.5 nM FMLP showed a significant increase in their expression of CD11b when passed through 5-μm filters (P < 0.05,n = 4). Moreover, the upregulation of CD11b was enhanced when primed PMN were filtered through 3-μm filters (P < 0.01,n = 4). The filtration of primed PMN had no effect on the expression level ofL-selectin (data not shown).
H2O2production by PMN.
The time-related changes in DCF fluorescence (H2O2production) in PMN are shown in Fig. 8. Significant increases in DCF fluorescence were induced by 100 nM FMLP (5 min after stimulation, P < 0.01,n = 4) or 1 μM PMA (30 min after stimulation, P < 0.05,n = 4). PMN filtration did not induce an increase in DCF fluorescence after either 5- or 3-μm filtration.
The cells circulating through the pulmonary microvessels must negotiate an average of 50–100 capillary segments between the arterial to the venous side of the pulmonary vascular bed (14, 15). The average diameter of each segment is 7.4 ± 2.3 μm in humans with a range between 2 and 15 μm (14). We have used filters with 3- and 5-μm pore sizes to simulate the smaller 20–30% of pulmonary capillary segments to determine the effect of deformation on the phenotypic and functional characteristics of PMN (5, 16). The experimental procedure chosen to examine the problem has been extensively used by other investigators (6, 8, 10, 11, 18, 23, 27, 30), and Selby et al. (27) have demonstrated, by using a similar system, a relationship between in vitro deformability of PMN and the in vivo sequestration of radiolabeled PMN in the lungs of human subjects. Downey and Worthen (10) have investigated the factors that determine the passage of PMN through polycarbonate filters and have calculated that the shear stress in their in vitro filtration system (80–800 dyn/s2) was remarkably similar to the estimated wall shear stress in vivo of between 100 and 500 dyn/s2 (10, 12). In our experiments we have used similar flow rates (1.0 and 3.0 ml/min) and calculated similar wall shear stress values. All filters were coated with albumin, and cells were prepared and filtered by using techniques to eliminate any significant exposure of cells to endotoxin that could result in cell activation. Hydrodynamic forces of flow, driving pressure, shear stress, and pore size, as well as cell size and deformability, have been shown to be important determinants in PMN passage through filters of a defined pore size (10-12). In this study we have determined the functional and structural changes that occur when PMN are exposed to hydrodynamic forces similar to those when they pass through smaller capillary segments in the lung.
The morphometrical analysis showed greater and more prolonged deformation after filtration through 3-μm compared with 5-μm-pore filters. The degree of PMN deformation estimated by L/S in the present study is within the range of deformation that was previously observed in human pulmonary capillaries and rabbit lungs (5). The delayed recovery of PMN to a rounded shape is also consistent with recent studies from our laboratory on isolated perfused rabbit lungs. In these studies PMN deformed in the capillaries maintained this elongated shape as they exited the pulmonary circulation, indicating that they retain this change in shape after their exit from the capillary bed (26). Varying the flow rates produced different driving pressures and shear stress on the PMN but did not affect the degree to which the cells deform or their recovery time to a spherical shape. Therefore, we conclude that the degree of deformation is mainly determined by pore size.
F-actin assembly is important for cytoskeletal reorganization and stabilization necessary to induce a migratory phenotype in PMN and acts as a critical factor in the mechanisms leading to protrusion of lamellapodia and provision of motor functions for chemotaxis (7). In response to stimulation by chemotactic factors, there is a rapid and transient increase in the assembly of F-actin from the G-actin pool (7). In several other cell systems, such as with endothelial cells in vitro (28) and in vivo (19), mechanical stress caused reorganization of actin filaments. Our data demonstrate an immediate short-lived increase in PMN F-actin content only after filtration through 3-μm-pore filters. This observation was supported by our immunofluoresence data showing brighter F-actin staining only after 3-μm filtration att = 0 s (Table 1). No evidence of actin-rich lamellapodia or filipodia, which would indicate active migration, was observed. Similarly, the MTOC did not show evidence of organelle polarization, indicating active migration of PMN (Table 1). The prolonged deformation of PMN associated with the lack of MTOC and F-actin polarity is more consistent with deformation-induced stabilization of the actin network by cross-linking events rather than actin polymerization in preparation for migration.
An increase in free [Ca2+]iis an important second messenger in cell signal transduction pathways by activation of Ca2+-dependent proteases, protein kinases, and phosphatases. Recent studies have shown that mechanical stress induced a Ca2+ signal in several other cell systems (1). We observed a modest but significant elevation in [Ca2+]iin these studies, but the amplitude and time course of the Ca2+ response associated with mechanical deformation were markedly different from those observed after FMLP treatment. This suggests that a different mechanism may be responsible for Ca2+ mobilization after deformation, and this is supported by the observation that, in contrast to sequential stimulation by FMLP, the Ca2+ response associated with mechanical deformation of PMN did not abolish or impair a subsequent Ca2+ response by FMLP stimulation. PMN deformation may induce a Ca2+signal through a nonreceptor mechanism similar to the Ca2+ signal in RBC. The Ca2+ signal resulting from PMN filtration is also different from the signaling pathway induced by PMA, which can induce oxidative burst activity of PMN without cytoplasmic Ca2+ flux. This increase of [Ca2+]iafter PMN filtration may contribute to activation of Ca2+-dependent protein kinases, proteases, or phosphatases.
The expression of adhesion molecules acts as a component of a molecular cascade in leukocyte-endothelial interaction and is regulated according to the activation status of PMN (2). Chemotactic stimulation of PMN results in shedding ofL-selectin and translocation of CD11b/CD18 (MAC-1) from intracellular storage pools to the cell surface (17). Therefore, the quantitative evaluation of these molecules is a useful indicator for PMN activation status. PMN deformation causes a differential change of cell adhesion molecules. The upregulation of CD11b/CD18 induced by 3-μm filtration shows that PMN are activated by mechanical deformation. These changes were flow rate independent and underline the importance of pore size and the degree of deformation, rather than shear stress, as a determinant of CD11b/CD18 upregulation. Furthermore, the calculated shear stress in our system (80–800 dyn/cm2) is three orders of magnitude larger than that required to inhibit adherence to flat endothelial surfaces (10). Unlike the upregulation of CD18/CD11b, PMN deformation did not affect the expression ofL-selectin. This differential effect of PMN deformation on cell adhesion regulation suggests that there are different signaling pathways that modify these molecules (22). Molad et al. (22) recently demonstrated that immune complexes induced the same type of differential effect on adhesion molecule expression of PMN (impairedL-selectin shedding after stimulation). They speculated that upregulation of β2-integrins with impairedL-selectin shedding causes tighter adhesion of PMN onto microvascular endothelium without migration, which may cause vascular endothelial injury if associated with production of active oxygen metabolites. Our data show that mechanical deformation of PMN did not cause H2O2production even in 3-μm filtration with flow rates of 3.0 ml/min. This result suggests that the mechanical deformation induced by filtration through small pores was not able to induce the respiratory burst associated with damage to the pulmonary microvessels.
The physiological significance of the structural and functional changes observed in PMN after mechanical deformation remains to be established in vivo. The difference in CD11b after 5- and 3-μm filtration suggests there is a threshold for the degree of deformation that produces CD11b upregulation and that priming the PMN with low doses of FMLP lowers this threshold (Fig. 7). This suggests that primed PMN may be more sensitive to the activation by mechanical stress, and we speculate that mechanical stress is one of the factors that influences PMN function at inflammatory sites.
In conclusion, this study establishes that mechanical deformation of PMN causes functional changes that include reorganization and stabilization of the cytoskeleton that could contribute to their passage through small restrictions in the pulmonary capillary bed. It also results in signal transduction with an increase in free [Ca2+]iand upregulation of cell-surface expression of CD11b/CD18. It further shows that the effect of deformation can be enhanced by priming PMN. These changes may be of relatively little importance under normal physiological conditions but could enhance the response of PMN sequestrated in the pulmonary capillary beds compressed by positive end-expiratory pressure induced by either external sources in mechanically ventilated patients or internal sources in patients with chronic airway obstruction. If these modest effects were amplified by other inflammatory stimuli, they might enhance the recruitment of PMN and contribute to lung injury.
We especially thank Stuart Greene and Beth Whalen for technical support and Lorraine Verburgt for statistical advice and analysis.
Address for reprint requests: S. F. Van Eeden, Univ. of British Columbia, Pulmonary Research Laboratory, St. Paul’s Hospital, 1081 Burrard St., Vancouver, British Columbia, Canada V6Z 1Y6.
This work was supported by the British Columbia Lung Association and by Medical Research Council of Canada Grant 4219.
- Copyright © 1997 the American Physiological Society