Transendothelial Neutrophil Migration
Role of Neutrophil-Derived Proteases and Relationship to Transendothelial Protein Movement
Abstract During an acute inflammatory response polymorphonuclear leukocytes (PMNs) adhere to and emigrate across the venular microvasculature. There is general agreement on the mechanisms involved in PMN adhesive interactions. However, the mechanisms by which PMNs migrate across the endothelial lining remain controversial, particularly with respect to the role of elastase. In the present study, we used human umbilical vein endothelial cells (HUVECs) and PMNs to test the hypothesis that the relative role of PMN-derived elastase may be dependent on the degree of HUVEC retraction within monolayers. A high (10−7 mol/L), but not a low (10−10 mol/L), concentration of platelet-activating factor (PAF) caused HUVEC retraction of sufficient magnitude to increase transendothelial protein movement. Elastase inhibitors prevented PMN transendothelial migration in response to the low, but not the high, concentration of PAF. These findings suggest that PMN migration across confluent endothelial cells is elastase dependent, whereas PMN migration across retracted endothelial cells is elastase independent. However, under the latter condition (high concentration of PAF), the two endogenous proteases, α2-macroglobulin and α1-antitrypsin, could interfere with PAF-induced PMN transendothelial migration. Thus, as the concentration of PAF is increased, migrating PMNs use other proteases, in addition to elastase. We also noted that transendothelial protein movement is closely coupled to PMN migration.
One of the cardinal signs of an acute inflammatory response is neutrophil emigration across the microvasculature and invasion of the affected tissue. This process involves a complex sequence of adhesive interactions between circulating neutrophils and the venular endothelium.1 2 3 The initial event appears to be the local generation of inflammatory mediators, which activate endothelial cells to increase their surface levels of adhesion molecules (P- and E-selectins). As the neutrophils leave the small diameter capillaries and enter the larger diameter venules, hemodynamic forces displace them from the central stream toward the endothelial lining. Once the neutrophils make contact, they begin to roll along the endothelium via weak adhesive interactions mediated by the selectins (L-selectin on neutrophils and P- and E-selectins on endothelium). It is believed that the rolling of neutrophils along the venular endothelium prolongs their close apposition to the endothelium, thereby facilitating their activation by the locally generated inflammatory me- diators. Upon activation, the neutrophils shed their L-selectin and simultaneously upregulate and/or activate their β2 integrins.4 The β2 integrins interact with ICAM-1 on endothelial cells to promote strong adhesive interactions, which arrest the rolling neutrophils. Once firmly adherent, the neutrophils change shape (flatten), extend pseudopodia between endothelial cells, and emigrate into the interstitium.
Although there has been general agreement on the mechanisms involved in neutrophil–endothelial cell adhesive interactions, the mechanisms by which neutrophils extend pseudopodia between endothelial cells and migrate into the interstitium remain unclear. In vivo studies of acute inflammation have provided evidence that elastase inhibitors prevent neutrophil accumulation in the affected tissue.5 6 7 These observations suggest that neutrophils mobilize their endogenous pool of proteases (eg, elastase) to proteolytically traverse the endothelium and invade the interstitium. However, interpretation of the in vivo studies is hampered by the fact that the elastase inhibitors used also interfered with leukocyte adhesion.6 7 Since leukocyte adhesion is a prerequisite for emigration, it is unclear whether the effects of the inhibitors are exerted at the level of emigration or adhesion. By contrast, in vitro studies provide compelling evidence that neutrophil transendothelial migration is independent of proteases; ie, a variety of protease inhibitors did not prevent neutrophil invasion of the subendothelial interstitial matrix.8 9 It is now accepted dogma that neutrophil emigration is independent of proteolytic penetration of the endothelium.10 This has prompted consideration of an active role of the endothelial cytoskeleton (ie, endothelial retraction and disruption of tight junctions) in the emigration process.11 12 Thus, one aim of the present study was to evaluate the relative roles played by neutrophils (eg, neutrophil-derived proteases) and endothelial cells (retraction) in neutrophil transendothelial migration.
Another cardinal sign of acute inflammation is enhanced microvascular protein leakage. Experimental evidence derived from various in vivo models of inflammation indicate that microvascular protein leakage is coupled to leukocyte adhesion/emigration. Intravital microscopy studies have shown that the greater the neutrophil trafficking across microvascular networks, the greater the extent of protein accumulation in the contiguous interstitium.13 14 15 Other studies have shown that phalloidin, an agent that stabilizes endothelial cytoplasmic microfilaments and associated tight junction elements, prevents endothelial cell retraction, neutrophil emigration, and vascular protein leakage.16 17 However, this coupling of vascular protein leakage to neutrophil adhesion and emigration is not universally accepted. Other in vivo studies using electron microscopy approaches have noted that vascular macromolecular leakage is independent of neutrophil emigration.18 19 Similarly, in vitro studies indicate that neutrophil transendothelial migration either does or does not affect transendothelial movement of protein.20 21 Thus, a second major aim of the present study was to evaluate whether neutrophil emigration could lead to increased transendothelial protein movement under specifically controlled conditions.
Our experimental approach was predicated on a potentially attractive explanation for the divergent observations regarding the mechanisms involved in neutrophil emigration and vascular protein leakage. This explanation is based on the premise that both neutrophil emigration and vascular protein leakage use the intercellular pathway to traverse the endothelial barrier and thus are dependent on endothelial cell retraction to open this pathway. Therefore, it is possible that inflammatory mediators that directly activate the endothelium to induce endothelial cell retraction (1) allow neutrophils to emigrate without the necessity of enzymatic digestion and (2) could promote vascular protein leakage independent of neutrophil–endothelial cell interactions. To test these hypotheses, we used an in vitro system to study transendothelial neutrophil migration and albumin movement in response to PAF, an inflammatory mediator known to be capable of endothelial cell retraction.22 We used two concentrations of PAF, one that would and one that would not compromise endothelial cell junctional integrity. Using this approach, we provide evidence that under specifically defined conditions (1) neutrophils use endogenous proteases to migrate across endothelial cell monolayers in culture and (2) transendothelial movement of albumin is directly coupled to neutrophil emigration.
Materials and Methods
HUVECs were harvested from umbilical cords by collagenase treatment as previously described.23 The cells were plated in M199 (GIBCO) supplemented with 10% heat-inactivated FCS (Hyclone Laboratories Inc), thymidine (2.4 mg/L, Sigma Chemical Co), glutamine (230 mg/L, JRH Biosciences), heparin sodium (10 IU/mL, Sigma), antibiotics (100 IU/mL penicillin, 100 μg/mL streptomycin, and 0.125 μg/mL amphotericin B), and endothelial cell growth factor (80 μg/mL Biomedical Technologies Inc). The cell cultures were incubated at 37°C in a humidified atmosphere with 5% CO2 and expanded by brief trypsinization (0.25% trypsin in PBS containing 0.02% EDTA). Primary through third-passage HUVECs were used in the experiments.
Human neutrophilic PMNs were isolated from venous blood of healthy adults using standard dextran sedimentation and gradient separation on Histopaque 1077 (Sigma).23 This procedure yields a PMN population that is 95% to 98% viable (trypan blue exclusion) and 98% pure (acetic acid–crystal violet staining).
HUVECs were grown to confluence on fibronectin (25 μg/mL)–coated Falcon cell culture inserts (3-μm-diameter pores). 51Cr-labeled neutrophils in M199 were added to the HUVEC monolayers (neutrophil–to–endothelial cell ratio, 10:1) and coincubated for up to 2 hours in the absence or presence of PAF in the basal compartment. At different times after coincubation, the monolayers were washed, and the inserts were removed and placed in 2N NaOH. The wash fluid, membrane lysate, and fluid bathing the basal aspect of the inserts were assayed for 51Cr activity. The percentage of added neutrophils that migrated from the apical to the basal aspects of the insert membranes was quantified as follows24 : percent migration=basal fluid (cpm)/[wash fluid (cpm)+membrane lysate (cpm)+basal fluid (cpm)].
HUVECs were grown to confluence on fibronectin-coated 48-well tissue culture plates (GIBCO). 51Cr-labeled neutrophils in M199 were added to the HUVEC monolayers (neutrophil–to–endothelial cell ratio, 10:1) and coincubated for 30 minutes in the presence or absence of PAF. After coincubation, the supernatants were removed, the monolayers were washed, and the remaining cells were lysed (2N NaOH). The supernatants, the wash fluid, and the lysate were assayed for 51Cr activity. The percentage of added neutrophils that adhered to the HUVEC monolayers was quantified as follows23 : percent adhesion=lysate (cpm)/[supernatant (cpm)+wash fluid (cpm)+lysate (cpm)].
Transendothelial Protein Movement
HUVECs were grown to confluence on fibronectin-coated Falcon cell culture inserts. Evan’s blue–labeled albumin in HBSS was added to the monolayers in the absence or presence of PMNs. PAF was added to the basal compartment. At different times after coincubation, samples of the supernatant and the fluid bathing the basal aspects of the inserts were obtained. The percentage of added albumin that entered the basal compartment was calculated from the concentration of Evan’s blue in the supernatant and basal fluid, which was determined spectrophotometrically.
We assessed the effects of PAF on albumin movement across HUVEC monolayers in the absence of PMNs as a functional index of endothelial tight junction integrity. PAF at a high concentration (10−7 mol/L) or low concentration (10−10 mol/L) was added to the basal aspect of HUVEC monolayers. Evan’s blue/albumin (0.5 mg/mL) was added to the apical aspect of HUVEC monolayers, and transendothelial albumin movement was assessed 60 minutes after coincubation. These experiments indicated that the high, but not the low, concentration of PAF increased transendothelial movement of albumin, ie, compromised tight junctional integrity (Fig 1⇓).
Subsequently, we assessed the effects of a MAb against human neutrophil elastase (final concentration, 2.5 μg/mL; Dako) or an elastase inhibitor, L658,758 (final concentration, 250 μmol/L; Merck Sharp & Dohme Research Laboratory) on PAF-induced neutrophil transendothelial migration. PAF at a high concentration (10−7 mol/L) or low concentration (10−10 mol/L) was added to the basal aspect of HUVEC monolayers, and 51Cr-PMNs (1×106) were added to the apical aspect in the presence or absence of MAb or inhibitor. PMN transendothelial cell migration was assessed 60 minutes after coincubation.
Previous in vivo studies indicate that elastase inhibitors may prevent PMN transendothelial migration, in part, by interfering with PMN adhesion to endothelium.6 7 To determine whether the MAb against elastase or the elastase inhibitor (L685,758) could modify PAF-induced PMN–endothelial cell adhesive interactions, they were included in the adhesion assay.
Two additional maneuvers were used to test the importance of an intact endothelial tight junction to elastase-mediated PMN transendothelial migration. In one series of experiments, the endothelial cell monolayers were pretreated with a high concentration (10−7 mol/L) of PAF for 1 hour. This pretreatment regimen compromised tight junction integrity, as evidenced by an increase in transendothelial protein movement (Fig 1⇑) and as assessed in adjacent wells. Subsequently, the effect of a MAb against elastase on PMN transendothelial migration in response to a low concentration (10−10 mol/L) of PAF was assessed. In another series of experiments, the effects of the anti-elastase MAb on PMN migration across naked filters (without endothelial monolayers) in response to the low concentration of PAF was determined.
In another series of experiments, we assessed the effects of two endogenous antiproteases, α2-macroglobulin and α1-antitrypsin, on PMN transendothelial migration induced by the high concentration of PAF. PAF was added to the basal aspect of the HUVEC monolayers, and 51Cr-PMN was added to the apical aspect in the presence of α2-macroglobulin (1 U/mL) or α1-antitrypsin (10 to 1000 μg/mL), and transendothelial migration was assessed as described above. As a control, the effects of α1-antitrypsin and α2-macroglobulin on PMN migration across naked filters were also determined.
To determine whether transendothelial albumin flux was coupled to PMN transendothelial protein movement, we assessed the effects of PAF on both transendothelial PMN migration and albumin movement simultaneously in the same wells. PAF at a low concentration (10−10 mol/L) was added to the basal aspect of HUVEC monolayers. 51Cr-PMN and Evan’s blue/albumin (0.5 mg/mL) were added to the apical aspect of HUVEC monolayers. Transendothelial PMN migration and albumin movement were assessed at 30, 60, and 120 minutes after coincubation. In some experiments, a MAb (IB4) directed against CD11/CD18 was also added to the apical aspect of the monolayers at a final concentration of 40 μg/mL.
All values are expressed as mean+SE. Data were analyzed using ANOVA and Student’s t test (with Bonferroni corrections for multiple comparisons). Regression analysis was performed using Graph Pad InStat software.
The effects of PAF on transendothelial movement of albumin are shown in Fig 1⇑. The high dose of PAF increased transendothelial movement of albumin, whereas the low dose of PAF had no effect. Since it is generally believed that transendothelial protein movement occurs via an intercellular route, these findings indicate that the high concentration of PAF produced endothelial cell retraction to such an extent that protein leakage into the basal compartment was increased.
As shown in Fig 2⇓, both the low and high concentrations of PAF promoted PMN transendothelial migration, but the high concentration of PAF was more potent. The MAb directed against neutrophil elastase completely inhibited the PMN transendothelial migration elicited by the low dose of PAF (Fig 2A⇓). By contrast, the anti-elastase MAb did not affect PMN transendothelial migration induced by the high dose of PAF (Fig 2B⇓). Identical results were obtained with the elastase inhibitor, L658,758 (Fig 3⇓). The elastase inhibitor prevented PMN transendothelial migration induced by the low dose of PAF (Fig 3A⇓) but not by the high dose of PAF (Fig 3B⇓).
In order to determine whether the above observations with PAF could apply to other inflammatory mediators, we repeated these experiments using fMLP, an inflammatory mediator previously used to assess the role of proteases in PMN transendothelial migration in vitro.9 Both a high (10−7 mol/L) and low (10−10 mol/L) concentration of fMLP were used to assess (1) transendothelial protein movement (in the absence of neutrophils) and (2) the role of elastase in PMN transendothelial cell migration. As shown in Fig 4⇓, the results obtained using fMLP were essentially identical to those obtained with PAF (Figs 1⇑ and 2⇑). Transendothelial protein movement was increased with the high concentration, but not the low concentration, of fMLP (Fig 4A⇓). Similarly, the MAb against elastase was effective in preventing fMLP-induced PMN transendothelial migration only when the low concentration of fMLP was used; the elastase inhibitor did not prevent PMN transendothelial migra- tion when the high concentration of fMLP was used (Fig 4B⇓).
The effect of the anti-elastase MAb on PMN adhesion to HUVECs induced by PAF is shown in Fig 5⇓. As expected, the PAF-induced adhesion of PMNs to HUVECs was dose dependent and, based on previous work, may be due to an interaction between PAF and P-selectin.25 Since PMN adhesion is a prerequisite for transendothelial migration, the dose-dependent adhesion interactions induced by PAFmay have contributed to the dose-dependent transendothelial migration induced by PAF (Figs 2⇑ and 3⇑). More important, the MAb against elastase did not affect PMN adhesion to HUVECs induced by either the high or low dose of PAF. Identical results were obtained with the elastase inhibitor L658,758 (data not shown). These findings indicate that the inhibitory effect of MAb against elastase or elastase inhibitors on PAF-induced PMN transendothelial migration cannot be attributed to their ability to prevent PMN adhesion to HUVECs.
The effects of the anti-elastase antibody on PMN transendothelial migration induced by the low concentration of PAF across monolayers pretreated with a high concentration of PAF is shown in Fig 6⇓. This pretreatment regimen compromised the endothelial tight junction integrity with respect to restricting protein movement (Fig 1⇑), which was confirmed in adjacent wells (data not shown). Under these conditions, the MAb against elastase was ineffective in preventing PMN transendothelial migration. Similarly, PMN migration across naked monolayers induced by the low concentration of PAF was unaffected by the anti-elastase MAb (data not shown).
As shown in Fig 7A⇓, both α1-antitrypsin and α2-macroglobulin interfered with PMN transendothelial migration in response to the high concentration of PAF. Further studies with α1-antitrypsin indicated that this inhibitory effect was dose dependent (Fig 7B⇓) and occurred under conditions in which the elastase inhibitors were ineffective (Figs 2B⇑ and 3B⇑). Neither protease inhibitor was able to inhibit PAF-induced PMN migration across naked filters (data not shown).
Since the low concentration of PAF did not alter transendothelial albumin movement (Fig 1⇑), we used this dose to study the relationship between transendothelial PMN migration and albumin movement. Fig 8⇓ illustrates the time course of transendothelial albumin movement (Fig 8A⇓) and PMN migration (Fig 8B⇓) induced by the low dose of PAF as measured simultaneously in the same wells. A greater number of PMNs migrated across the HUVEC monolayers as the duration of coincubation was increased. Similarly, more albumin entered the basal compartment as the duration of coincubation was increased. In the presence of the MAb against CD11/CD18, there was less transendothelial PMN migration and albumin movement.
Fig 9⇓ illustrates the relationship between PAF-induced transendothelial PMN migration and albumin movement when the data from individual inserts were plotted individually. All values are derived from data presented in Fig 8⇑. As shown, there is a linear correlation between the extent of PMN migration and albumin movement. This relationship indicates that in those inserts in which PAF elicited a greater PMN transendothelial migration, there was a correspondingly greater transendothelial movement of albumin.
Neutrophil invasion of inflamed tissue is a complex process involving an initial mild adhesive interaction with the venular endothelium, termed rolling, which allows neutrophils to remain in close apposition to the endothelial cells and to sample the environment for local signals of an ongoing inflammatory process.1 2 3 If the appropriate signals (stimuli) are present, the neutrophils become activated, and a strong adhesive interaction takes place. This results in neutrophil arrest and eventual emigration toward the chemotactic stimulus in the interstitium. Although there is a general consensus on the mechanisms (adhesion molecule activation/expression) involved in neutrophil–endothelial cell adhesive interactions,1 2 3 the mechanisms by which neutrophils penetrate the endothelial cell lining to gain access to the interstitium remain controversial. The barriers to neutrophil movement to the site of chemotactic (or inflammatory) stimuli in the interstitium are (1) the endothelial cells lining the venules, (2) the basement membrane, and (3) the interstitial matrix per se. In the present study, we focused on the initial barrier to neutrophil emigration, the endothelium. Using a reductionist approach (HUVEC monolayers without basement membrane and underlying interstitial matrix), we have provided evidence that neutrophils use endogenous proteases to migrate across endothelial cell monolayers in culture.
In vivo studies indicate that during experimentally induced acute inflammation (ischemia/reperfusion or exogenous administration of PAF), there is an increase in leukocyte emigration across venular endothelium into the interstitium. The emigration process appears to require leukocyte-derived proteases, since elastase inhibitors largely prevent accumulation of PMNs in the interstitium of the affected tissue.5 6 7 Using HUVEC monolayers grown in inserts, we have previously shown that the neutrophil transendothelial migration induced by anoxia/reoxygenation (the in vitro counterpart to ischemia/reperfusion) can be completely prevented by a MAb against elastase.26 In the present study, we provide evidence that PMN transendothelial migration induced by PAF (10−10 mol/L) can be prevented by elastase inhibitors (Figs 2A⇑ and 3A⇑). Thus, our findings using a reductionist approach are in agreement with those from in vivo studies that demonstrate a role for elastase in leukocyte emigration across microvessels.
In order for neutrophils to extend lamellipodia between endothelial cells, the endothelial tight junctional complexes must be breached so that endothelial cells can retract from one another, providing a passageway. It seems plausible to suggest that neutrophils accomplish this task by using endogenous proteases, such as elastase. In support of this contention are the observations that (1) activated neutrophils can induce endothelial cell retraction within confluent monolayers, an event prevented by elastase inhibitors,24 and (2) purified neutrophil elastase can induce endothelial cell retraction in a concentration- and time-dependent manner.24 This elastase-induced retraction occurs without causing injury to the endothelial cells themselves, as evidenced by a normal cobblestone morphology (microscopy) and intact cell membranes (51Cr release). This is analogous to the common practice of expanding endothelial cell cultures with proteases (eg, trypsinization). The exact protein(s) within the tight junctional complex that is attacked by elastase is not clear, but recent studies indicate that cadherins are a likely target.27 Taken together, these findings indicate that neutrophil-derived elastase plays an important role in facilitating migration by proteolytically disrupting endothelial cell tight junctions.
There is evidence in the literature that inflammatory mediators (eg, PAF) and vasoactive amines (eg, histamine) can directly induce endothelial cell retraction and, thereby, facilitate neutrophil transendothelial migration.22 28 It has also been suggested that activated neutrophils can signal the endothelial cells to retract within monolayers.11 29 In these scenarios in which the endothelial cells actively participate in the migration process by retracting from one another, it may be likely that neutrophil migration across denuded areas would be independent of neutrophil-derived proteases. Furthermore, protease inhibitors would not be expected to interfere with PMN migration across retracted monolayers. In the present study, we tested this possibility by using two concentrations of PAF, one that directly caused endothelial cell retraction and one that did not. As shown in Fig 1⇑, PAF at 10−7 mol/L, but not at 10−10 mol/L, increased albumin movement across HUVEC monolayers, providing functional evidence of endothelial cell retraction induced by the higher concentration of PAF. When the higher concentration of PAF was used to stimulate PMN transendothelial migration, both of the elastase inhibitors were ineffective in inhibiting the migration (Figs 2B⇑ and 3B⇑). In another series of experiments, the monolayers were pretreated with the high concentration of PAF (to induce endothelial cell retraction), and PMN transendothelial migration was assessed in response to the low concentration of PAF. In contrast to migration across untreated monolayers (Figs 2A⇑ and 3A⇑), inhibition of elastase did not interfere with PMN transendothelial migration across pretreated monolayers (Fig 6⇑). Finally, inhibition of elastase did not affect PMN migration across naked filters. Taken together, these findings would indicate that if an inflammatory mediator was used at a concentration that would directly cause endothelial cell retraction within the monolayers, PMN transendothelial migration could proceed independent of elastase activity.
One caveat in the above reasoning is that the use of a higher concentration of PAF in the migration assays may have resulted in a greater activation of neutrophils, inciting them to mobilize additional proteases besides elastase. Thus, we assessed the effects of two endogenous antiproteases, α1-antitrypsin (inhibits serine proteases) and α2-macroglobulin (inhibits most endoproteases), on PMN transendothelial migration in response to the high concentration (10−7 mol/L) of PAF. Both of these antiproteases were effective in reducing the PMN migration stimulated by the high concentration of PAF (Fig 7⇑). These latter observations indicate that in response to a high concentration of PAF, PMNs use other proteases besides elastase to digest their way between endothelial cells. Thus, even though the high concentration of PAF can induce endothelial cell retraction (as evidenced by an increase in transendothelial albumin flux), the extent of retraction is insufficient to allow PMNs to migrate across these monolayers without the aid of proteases.
The results of the present study are in marked contrast to other in vitro studies in which PMN migration across HUVEC monolayers was not prevented by elastase inhibitors or, for that matter, a broad spectrum of antiproteases.8 9 The reasons for these disparate observations are not readily apparent. One could argue that the use of different stimuli to promote neutrophil transendothelial migration may be the underlying factor determining whether proteases are necessary for PMN transendothelial migration. To address this issue, we used fMLP (the inflammatory mediator used in Reference 99 ). Essentially identical results were obtained using fMLP (Fig 4⇑) as were noted with PAF (Figs 1⇑ and 2⇑). Thus, the lack of effect of various protease inhibitors on fMLP-induced transendothelial migration in previous studies cannot be simply attributed to the fact that fMLP was used rather than PAF. Indeed, the observation that elastase plays a role in both PAF- and fMLP-induced PMN transendothelial migration suggests that the results of the present study may apply to other inflammatory mediators as well.
Another phenomenon associated with acute inflammation is an increase in vascular protein leakage. Both PMN emigration and vascular protein leakage proceed through interendothelial cell junctions. In the present study, we assessed whether the two phenomenon are linked in our reductionist model using a concentration of PAF (10−10 mol/L) that does not affect the transendothelial movement of albumin per se (Fig 1⇑). Our findings indicate that transendothelial albumin movement and PMN migration are closely coupled (Figs 8⇑ and 9⇑). These findings are in general agreement with another in vitro study using a reductionist approach,21 but not one in which a more complex system was used (HUVECs grown on amnion).20 In the latter study, induction of PMN migration (fMLP) did alter albumin movement across the amnion, an observation most likely attributable to the thickness of the amnion. That the amnion membrane is a significant barrier to albumin diffusion is supported by the fact that complete equilibration of albumin between luminal and basal compartments required 8 hours,20 whereas the same process in our reductionist model (naked membrane) required only 20 minutes.
In vivo studies of acute inflammation indicate that the vascular albumin leakage appears to be coupled to leukocyte emigration.13 14 15 However, in some of these studies, up to 50% of the vascular albumin leakage is independent of leukocyte emigration.13 15 Furthermore, other in vivo studies provide evidence that (1) vascular macromolecular leakage can occur in venules across which minimal leukocyte trafficking is noted and (2) no macromolecular leakage occurs in venules across which substantial leukocyte traffic is observed.18 19 In reconciling these nuances between the in vivo studies and our in vitro studies, it must be remembered that we deliberately chose to use PAF at a concentration (10−10 mol/L) that did not alter transendothelial albumin movement per se. In vivo, it is quite likely that the concentration of inflammatory mediators may achieve levels that can directly cause endothelial cell retraction of a sufficient magnitude to allow protein to leak from the venules into the interstitium. Under these conditions, one would predict that vascular protein leakage may be independent of PMN migration.
In summary, our findings indicate that PAF-induced PMN transendothelial migration is a protease-dependent process. Our use of two different concentrations of PAF, a high concentration (10−7 mol/L), which directly causes endothelial cell retraction (of sufficient magnitude to increase transendothelial albumin movement), and a low concentration (10−10 mol/L), which does not, generated results that can be extrapolated to acute inflammatory responses in vivo. For example, a mild inflammatory response (low concentrations of inflammatory mediators) would promote PMN transendothelial migration, which can be inhibited by elastase. Vascular albumin leakage would be closely coupled to this increased PMN trafficking. When the inflammatory response is more severe (high concentrations of inflammatory mediators), endothelial cells will be directly induced to retract, promoting a more rapid influx of neutrophils into the affected area and allowing vascular protein leakage to occur independent of PMN trafficking. The endothelial retraction can contribute to the increased PMN trafficking in two ways. It can allow PMNs to migrate independent of proteases across a disrupted endothelial barrier and/or allow for an increased diffusion of inflammatory mediators from the interstitium to the endothelial-PMN interface, thereby promoting a greater activation of the PMNs. Our data are more consistent with the latter phenomenon, since PMNs did not freely move across the activated endothelium but required proteases (in addition to or other than elastase) in order to traverse the endothelial barrier. These observations indicate that the PMNs were activated to a greater extent under these conditions and mobilized additional proteases to enhance their movement across the monolayers. Further studies are warranted to identify the additional proteases that are involved in PMN transendothelial migration during a severe inflammatory reaction, since these proteases may provide a rational therapeutic approach to control the adverse effects of increased PMN trafficking in situations in which elastase inhibitors are ineffective.
Selected Abbreviations and Acronyms
|HUVEC||=||human umbilical vein endothelial cell|
|ICAM-1||=||intercellular adhesion molecule-1|
This study was supported by a grant from the Medical Research Council of Canada (MT-12902).
- Received April 11, 1997.
- Accepted August 14, 1997.
- © 1997 American Heart Association, Inc.
Granger DN, Kubes P. The microcirculation and inflammation: modulation of leukocyte-endothelial cell adhesion. J Leukoc Biol. 1994;55:622-675.
Kvietys PR, Cepinskas G, Granger DN. Neutrophil-endothelial cell interactions during ischemia/reperfusion. In: Kamada T, Shiga T, McCuskey RS, eds. Tissue Perfusion and Organ Function: Ischemia/Reperfusion Injury. Amsterdam, Netherlands: Elsevier Science; 1996:179-191.
Kishimoto TK, Julia MA, Berg EL, Butcher EC. Neutrophil Mac-1 and MEL-14 adhesion proteins inversely regulated by chemotactic factors. Science. 1989;245:1238-1241.
Carden DL, Korthuis RJ. Protease inhibition attenuates granulocyte infiltration and microvascular dysfunction in postischemic skeletal muscle. Am J Physiol. 1996;271: H1947-H1951.
Zimmerman BJ, Granger DN. Reperfusion-induced leukocyte infiltration: role of elastase. Am J Physiol. 1990;259:H390-H394.
Woodman RC, Heinhart PH, Kanwar S, Johnston FL, Kubes P. Effects of human neutrophil elastase (HNE) on neutrophil function in vitro and in inflamed microvessels. Blood.. 1993;82:2188-2195.
Huber AR, Weiss SJ. Disruption of the subendothelial basement membrane during neutrophil diapedesis in an in vitro construct of a blood vessel wall. J Clin Invest. 1989;83:1122-1136.
Furie MB, Naprstek BL, Silverstein SC. Migration of neutrophils across monolayers of cultured microvascular endothelial cells: an in vitro model of leukocyte extravasation. J Cell Sci.. 1987;88:161-175.
Cramer EB. Cell biology of phagocyte migration from the bone marrow, out of the bloodstream, and across organ epithelia. In: Gallin JI, Goldstein IM, Snyderman R, eds. Inflammation: Basic Principles and Clinical Correlates. 2nd ed. New York, NY: Raven Press Publishers; 1992:341-351.
Shepro D. Endothelial cytoskeletal regulation of neutrophil diapedesis. In: Granger DN, Schmid-Schonbein GW, eds. Physiology and Pathophysiology of Leukocyte Adhesion. New York, NY: Oxford University Press; 1995:196-216.
Granger HI, Yuan Y, Zawieja DC. Ultrastructural basis of leukocyte migration through the microvascular membrane. In: Granger DN, Schmid-Schonbein GW, eds. Physiology and Pathophysiology of Leukocyte Adhesion. New York, NY: Oxford University Press; 1995:185-195.
Kurose I, Anderson DC, Miyasaka M, Tamatani T, Paulson JC, Todd RF, Rusche JR, Granger DN. Molecular determinants of reperfusion-induced leukocyte adhesion and vascular protein leakage. Circ Res. 1994;74:336-343.
Kurose I, Kubes P, Wolf R, Anderson DC, Paulson J, Miyasaka M, Granger DM. Inhibition of nitric oxide production: mechanisms of vascular albumin leakage. Circ Res. 1993;73: 164-171.
Asako H, Wolf RE, Granger DN, Korthuis RJ. Phalloidin prevents leukocyte emigration induced by proinflammatory stimuli in rat mesentery. Am J Physiol. 1992;263:H1637-H1642.
Korthuis RJ, Carden DL, Kvietys PR, Shepro D, Fuseler J. Phalloidin attenuates postischemic neutrophil infiltration and increased microvascular permeability. Am J Physiol. 1991;71:1261-1269.
Rosengren S, Olofosson AM, Von Andrian UH, Lundgren-Akerlund E, Arfors K-E. Leukotriene B4-induced neutrophil-mediated endothelial leakage in vitro and in vivo. Am J Physiol. 1991;71:1322-1330.
Bussolino F, Camussi G, Aglieta M, Braquet P, Bosia M, Pescarmona G, Sanavio F, D’Urso N, Marchisio PC. Human endothelial cells are target for platelet activating factor, 1: platelet activating factor induces changes in cytoskeletal structures. J Immunol. 1987;139:2439-2446.
Yoshida N, Granger DN, Anderson DC, Rothlein R, Lane C, Kvietys PR. Anoxia/reoxygenation-induced neutrophil adhesion to cultured endothelial cells. Am J Physiol. 1992;262:H1891-H1989.
Lorant DE, Patel KD, McIntyre TM, McEver RP, Prescott SM,Zimmerman GA. Coexpression of GMP-140 and PAF by endothelium stimulated by histamine or thrombin: a juxtacrine system for adhesion and activation of neutrophils. J Cell Biol.. 1991;115:223-234.
Cepinskas G, Granger DN, Kvietys PR. Neutrophil transendothelial migration in an in vitro model of ischemia/reperfusion (I/R): modulation by platelets. FASEB J. 1995;9:A270. Abstract.
Rotrosen D, Gallin JI. Histamine type 1 receptor occupancy increases endothelial cytosolic calcium, reduces F-actin, and promotes albumin diffusion across cultured endothelial monolayers. J Cell Biol. 1986;103:2379-2387.
Huang AJ, Manning JE, Bandak TM, Ratau MC, Hanser KR, Silverstein S. Endothelial cell cytosolic free calcium regulates neutrophil migration across monolayers of endothelial cells. J Cell Biol. 1993;120:1372-1380.