Microvascular Research 71 (2006) 40 – 47 www.elsevier.com/locate/ymvre
Albumin leak across human pulmonary microvascular vs. umbilical vein endothelial cells under septic conditions Jennifer L. Shelton, Lefeng Wang, Gediminas Cepinskas, Martin Sandig, Richard Inculet, David G. McCormack, Sanjay Mehta * Centre for Critical Illness Research, Lawson Health Research Institute, Division of Respirology, Departments of Medicine, Physiology and Pharmacology, London Health Sciences Center, University of Western Ontario, South Street Campus, Colborne Building Room 216, 375 South Street, London, Ontario, Canada N6A 4G5 Received 11 March 2005; revised 2 September 2005; accepted 1 November 2005 Available online 22 December 2005
Abstract Human pulmonary microvascular endothelial cell (HPMVEC) injury is central to the pathophysiology of human lung injury. However, septic HPMVEC barrier dysfunction and the contribution of neutrophils have not been directly addressed in vitro. Instead, human EC responses are often extrapolated from studies of human umbilical vein EC (HUVEC). We hypothesized that HUVEC was not a good model for investigating HPMVEC barrier function under septic conditions. HPMVEC was isolated from lung tissue resected from lung cancer patients using magnetic bead-bound anti-PECAM-1 antibody. In confluent monolayers in 3-Am cell-culture inserts, we assessed trans-EC Evans-Blue (EB)-conjugated albumin leak under basal, unstimulated conditions and following stimulation with either lipopolysaccharide or a mixture of equal concentrations of TNF-a, IL-1h and IFN-g (cytomix). Basal EB-albumin leak was significantly lower across HPMVEC than HUVEC (0.64 T 0.06% vs. 1.13 T 0.10%, respectively, P < 0.001). Lipopolysaccharide and cytomix increased leak across both HPMVEC and HUVEC in a dose-dependent manner, with a similar increase relative to basal leak in both cell types. The presence of neutrophils markedly and dose-dependently enhanced cytomixinduced EB-albumin leak across HPMVEC ( P < 0.01), but had no effect on EB-albumin leak across HUVEC. Both cytomix and lipopolysaccharide-induced albumin leak was not associated with a loss of cell viability. In conclusion, HPMVEC barrier dysfunction under septic conditions is dramatically enhanced by neutrophil presence, and HUVEC is not a suitable model for studying HPMVEC septic barrier responses. The direct study of HPMVEC septic responses will lead to a better understanding of human lung injury. D 2005 Elsevier Inc. All rights reserved. Keywords: Sepsis; Edema; Acute lung injury; Endothelial cell; Neutrophil
Introduction Acute lung injury (ALI) is a common condition characterized by high-protein pulmonary edema and severe hypoxemia, which in its most severe form, the acute respiratory distress syndrome (ARDS), has a mortality of 40 –50% (Ware and Matthay, 2000; Groeneveld, 2002). The most common cause of ALI is sepsis, a systemic inflammatory reaction to infection, which occurs in 1% of hospitalized patients (Bone et al., 1992; Angus et al., 2001). The pathophysiology of ALI is characterized by a complex, integrated response of inflammatory mediators (e.g. TNF-a, nitric oxide [NO]) and cells (e.g. neutrophils, macrophages) to * Corresponding author. Fax: +1 519 685 8406. E-mail address:
[email protected] (S. Mehta). 0026-2862/$ - see front matter D 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.mvr.2005.11.003
an insult (Bone et al., 1992; Abraham et al., 2000; Bellingan, 2002). This inflammation is not addressed by current ALI therapy, which is limited to supportive care, e.g. supplemental oxygen and mechanical ventilation (Abraham, 1999; Glauser, 2000; Wheeler and Bernard, 1999). One of the key targets of the inflammatory response in sepsis and ALI is the microvascular endothelial cell (EC) (Curzen et al., 1994; Reinhart et al., 2002). Septic EC demonstrates upregulation of surface cell adhesion molecules, increased production of cytokines and chemotaxins and barrier dysfunction, resulting in high-protein pulmonary edema, as shown by us and others (Wang et al., 2002; Razavi et al., 2002; Granger and Kubes, 1994; Garcia et al., 1998; Wagner and Roth, 1999). The importance of EC injury in human ALI has recently been highlighted, as plasma markers of EC injury in
J.L. Shelton et al. / Microvascular Research 71 (2006) 40 – 47
patients correlate with survival in ARDS (Ware et al., 2001). Neutrophils are central to the microvascular and tissue injury of ALI (Tate and Repine, 1983; Downey et al., 1995; Wagner and Roth, 1999; Doerschuk, 2001), and infiltration of neutrophils into the lung is facilitated by EC barrier dysfunction (Garcia et al., 1998; Dull and Garcia, 2002; Razavi et al., 2004). A common cell type used to investigate human EC biology is the human umbilical vein EC (HUVEC), largely because of accessibility and its ease of isolation. However, there is significant inter-EC heterogeneity, based on the organ or tissue and the size or type of blood vessel from which EC is isolated (Beck et al., 1999; Burg et al., 2002; Hillyer et al., 2003; Otto et al., 2001). The isolation and culture of human pulmonary microvascular EC (HPMVEC) has recently been described (Wagner et al., 1999b; Hewett and Murray, 1993). However, the effects of septic stimulation and exposure to neutrophils on HPMVEC barrier function have not been reported. (De Staercke et al., 2003; King et al., 2004). Thus, we hypothesized that HUVEC is not a good model for investigating septic HPMVEC barrier dysfunction. To address this, we compared albumin leak, via the Evans-Blue (EB) dye-conjugated albumin method, across HPMVEC and HUVEC monolayers in response to septic stimuli, including lipopolysaccharide (LPS) and cytomix, a mixture of equal concentrations of cytokines relevant to human sepsis and ALI (TNF-a, IL-1h and IFN-g). Moreover, we compared barrier function of HPMVEC and HUVEC monolayers under cytomix stimulation in the presence of human neutrophils. In order to address whether septic albumin leak was due to cell death, we assessed cell viability under all of these conditions. Materials and methods The use of human material was approved by the Institutional Research Ethics Board. Informed consent was obtained from volunteer blood donors, but was not required or obtained for use of discarded tissues. All identifying data were removed to protect anonymity.
41
under direct microscopy, in order to allow visually identified HPMVEC clusters to grow. When HPMVEC had proliferated to 50% confluence, they were harvested with trypsin/EDTA, and were re-purified with anti-CD31-coated magnetic microbeads as above. HPMVEC was seeded in a 25 cm2 flask and passaged at a ratio of 1:3. Using this method, we obtained HPMVEC at >99% purity, as confirmed by diI-LDL uptake and CD31 (PECAM) staining (Fig. 1). HPMVEC was used for experiments at passages 4 – 5. Primary HUVEC was isolated from three fresh umbilical cord veins as previously described (Jaffe et al., 1973). Briefly, the umbilical vein lumen was instilled with 300 U/ml type II collagenase at 37-C for 10 min. HUVEC was collected by flushing the vein with PBS, and were pelleted and resuspended in EGM-2 containing 20 AM HEPES (Sigma), 5% heat inactivated fetal calf serum and 1% Penicillin – Streptomycin (Gibco, New York, NY). HUVEC was used for experiments at passages 4 – 5.
PECAM-1 immunostaining HPMVEC was grown on fibronectin-coated glass coverslips until confluence. Subsequently, HPMVEC was washed and fixed with 3% paraformaldehyde in PBS for 15 min at room temperature. PECAM-1 (CD31) immunostaining was performed as described (Ionescu et al., 2003). Briefly, fixed HPMVEC was incubated for 30 min at room temperature with mouse antihuman PECAM-1 antibody (hec 7; a gift from Dr. W. Muller, Cornell Univ.) at a concentration of 5 Ag/ml in PBS buffer containing 1% BSA. Subsequently, HPMVEC was washed and incubated for another 30 min with the secondary, fluorescein isothiocyanate (FITC)-conjugated goat-anti-mouse IgG (Molecular Probes, Burlington, Ontario) at a dilution of 1:500. As a negative control, the same protocol was performed in the absence of the primary antibody. The samples were mounted in Vectashield (Vector Laboratories, Burlingame, CA) on microscope slides and analyzed by fluorescence microscopy (DMR Leica Upright Fluorescent Microscope, Quorum, Guelph, Ontario) and images digitally recorded (Q-Imaging Retiga Ex CCD camera, Quorum; Openlab 4.0.2 software, Toronto, Ontario).
Isolation of human blood neutrophils Neutrophils were isolated from normal human blood donors using Lymphocyte Separation Medium as per manufacturer’s instructions (ICN Biomedicals, Aurora, OH).
Trans-EC EB-albumin leak To establish confluent monolayers, 105 HUVEC and HPMVEC were seeded in 10% gelatin-coated Transwell Inserts (3.0 Am pore; BD Falcon,
Isolation and culture of HPMVEC and HUVEC HPMVEC was isolated from human lung as we previously reported for murine lung (Razavi et al., 2004), using a modification of a published protocol (Hewett and Murray, 1993; Wagner et al., 1999b). Briefly, 5 – 10 g of peripheral human lung tissue was isolated from a grossly normal-appearing region farthest away from the cancer following resectional surgery for localized lung cancer in three patients. Lung tissue was rinsed in 10 mM phosphate-buffered saline (PBS, pH 7.4), finely minced into 2 – 3 mm pieces and digested in 0.3% type II collagenase (Sigma Chemical Co., Oakville, Ontario) for 45 min at 37-C with occasional agitation. The digested suspension was filtered through 100 Am mesh, centrifuged at 500g and washed twice in phosphate-buffered saline (PBS). The cell pellet was resuspended in binding buffer (2 AM Na-citrate, 1.2 AM NaH2PO4IH2O, 5.6 AM Na2HPO4, 138.6 AM NaCl, 1% Bovine serum albumin) and incubated for 20 min at 4-C with magnetic microbeads (Dynal Inc., Lake Success, NY) coated with anti-human CD31 antibody. Cells bound to microbeads were magnetically captured and washed 5 times with binding buffer. Isolated cells were resuspended in 10% Endothelial Growth Medium-2 (EGM-2, Cambrex Bio Science Inc., Walkersville, MD), and incubated for 8 – 12 h at 37-C in 5% CO2/21% O2. The culture medium was then changed to remove unattached cells, and the medium was subsequently changed every 3 days. From 7 – 10 days after seeding the cells, the quickly growing fibroblasts in the flask were repeatedly removed mechanically using a sharp-end rubber policeman
Fig. 1. Characterization of >99% purity of human pulmonary microvascular endothelial cells by immunostaining with anti-PECAM (CD31). Negative staining in the absence of the primary anti-PECAM antibody showed no fluorescence (not shown). Magnification 100.
42
J.L. Shelton et al. / Microvascular Research 71 (2006) 40 – 47
Franklin Lakes, NJ) and cultured at 37-C in 5% CO2 for up to 4 days in EGM-2 containing 5% heat-inactivated fetal calf serum, 20 AM HEPES and 1% Penicillin – Streptomycin. Media were replaced after 3 days. EB-conjugated albumin (final concentration 0.67 mg/ml) was prepared by diluting a stock solution of 2% EB in a 60-fold excess of Bovine Serum Albumin (BSA, 4%) in order to eliminate any free EB, according to a method previously described (Patterson et al., 1992). To measure trans-EC EB-albumin leak, EB-albumin was added to the upper chamber for 1 h, while 4% BSA was added to the lower chamber, such that albumin movement due to an oncotic pressure gradient was eliminated. Moreover, the heights of the medium in the upper and lower compartments were at the same level, such that bulk flow due to a hydrostatic pressure gradient was not a factor. After incubation at 37-C for 1 h, media in the lower chamber were collected, and the absorbance of EB was measured at 595 nm and referenced to a standard curve in order to report trans-EC EB-albumin leak as a flux of albumin in ng/min. Trans-EC EB-albumin leak was assessed in unstimulated HPMVEC and HUVEC, and at various time points following stimulation with LPS (0.001 Ag/ml – 10 Ag/ml) or cytomix (0.01 ng/ml – 10 ng/ml doses of each of TNFa, IL-1h and IFN-g). At each of the different time points, discrete EC monolayers/inserts were used, such that EB-albumin leak was never measured twice in the same monolayer at different time points. EB-albumin was added to the upper chamber exactly 1 h prior to each time point being assessed. The effect of co-culture of HPMVEC and HUVEC with human neutrophils on cytomix-stimulated trans-EC EB-albumin leak was also assessed.
Assessment of HPMVEC and HUVEC viability Cell viability was assessed by incubating cells with 0.5 mg/ml metallothionein (MTT; Sigma) for 3 – 4 h at 37-C in the dark. Excess MTT was removed, and oxidized MTT extracted by incubation in DMSO for 5 min. 100% viability was defined by measuring OD570 in supernatant from unstimulated cells. In LPS or cytomix-treated cells, viability was calculated as OD570experimental / OD570unstimulated 100%. Treatment of HPMVEC and HUVEC with 1 mM H2O2 was used as a positive control for loss of cell viability. In the presence of neutrophils, HPMVEC and HUVEC viability was determined by incubating with 2% Trypan Blue/EGM for 3 min. Cell death was assessed microscopically as number of blue-stained cells / total cell number 100%.
Statistical analysis Data are presented as mean T SEM. Between-group differences were assessed by ANOVA and post hoc t test with Bonferroni correction, where appropriate. Time-course studies were analyzed by repeated-measures ANOVA. Significance was accepted for two-tailed P < 0.05.
Results Basal and stimulated albumin leak across HPMVEC and HUVEC Basal, unstimulated EB-albumin leak was significantly lower across HPMVEC than HUVEC monolayers (Fig. 2). Of the total EB-albumin added to the upper chamber, 0.64 T 0.06% and 1.13 T 0.10% leaked across HPMVEC and HUVEC monolayers, respectively ( P < 0.001). This basal level of albumin leak was stable over at least 8 h in both HPMVEC and HUVEC (data not shown). Basal level of albumin leak was not significantly different between HPMVEC from the three individual patients, and was also consistent across HUVEC from different umbilical veins.
Fig. 2. Basal, unstimulated leak of Evans-Blue-conjugated albumin across human pulmonary microvascular (HPMVEC) and human umbilical vein endothelial cells (HUVEC). *P < 0.001 vs. HPMVEC. n = 26 – 29/group.
Treatment of both HPMVEC and HUVEC with a high dose of LPS (10 Ag/ml) was associated with increased transEC albumin leak, which was maximal following 8 h of stimulation (Fig. 3). After 16 h of stimulation with LPS, trans-EC albumin leak across both HPMVEC and HUVEC was no longer significantly different from basal leak (data not shown). We observed a dose-dependent effect of LPS stimulation on albumin leak across both HPMVEC and HUVEC (Fig. 4). The relative increase in LPS-stimulated leak was similar in HPMVEC and HUVEC, when normalized for differences in respective basal albumin leak. Albumin leak across both HPMVEC and HUVEC in response to doses of LPS up to 10 Ag/ml was not associated with any loss of cell viability as determined by the MTT assay (Table 1). Treatment with a higher dose (100 Ag/ml) of LPS induced markedly greater leak in both cell types (55 T 2-fold and 22 T 4-fold greater than unstimulated leak in HPMVEC and HUVEC, respectively; P < 0.01 for both vs. respective unstimulated group). Treatment with 100 Ag/ml LPS was also associated with greater loss of cell viability, although this did not achieve significance (Table 1). Treatment of HPMVEC with a high dose of cytomix (10 ng/ml) also increased albumin leak maximally after 8 h of stimulation (Fig. 5). In HUVEC treated with 10 ng/ml cytomix, albumin leak peaked slightly earlier, being maximal 4– 8 h after stimulation. Albumin leak was also assessed after 16 h of stimulation with cytomix. At this time point, cytomixstimulated trans-EC albumin leak was more than 10-fold greater than baseline leak in both HPMVEC and HUVEC (data not shown). Since loss of cell viability after stimulation for more than 8 h was of concern, trans-EC albumin leak was only assessed after 8 h of stimulation in subsequent dose – response and neutrophil co-culture studies. As with LPS, cytomix also increased albumin leak across both HPMVEC and HUVEC in a dose-dependent manner (Fig. 6), with a similar relative increase in leak in HPMVEC and HUVEC. At doses up to 10 ng/ml, cytomix-induced albumin leak was not associated with loss of cell viability (Table 1). However, upon stimulation with a higher dose of cytomix (100 ng/ml), albumin leak was markedly increased (59 T 16-fold and 46 T 9-fold unstimulated leak in HPMVEC and HUVEC, respectively; P < 0.05 for both vs. respective unstimulated leak) and was associated with significant loss of cell viability (Table 1).
J.L. Shelton et al. / Microvascular Research 71 (2006) 40 – 47
43
Table 1 Effect of varying doses of cytomix and LPS on HPMVEC and HUVEC cell survival following 8 h stimulation Treatment
Dose
HPMVEC
HUVEC
% survival T SEM % survival T SEM MTT assay LPS
Cytomix Fig. 3. Time course of LPS (lipopolysaccharide 10 Ag/ml) induced albumin leak in HPMVEC and HUVEC monolayers. For each time point assessed, individual EC monolayers/inserts were used, and EB-albumin was added to the upper chamber exactly 1 h prior to the time point being assessed. *P < 0.05; **P < 0.01 vs. respective baseline (time 0). n = 3 – 6/group.
Effect of neutrophil presence on cytomix-stimulated albumin leak across HPMVEC and HUVEC The presence of neutrophils in co-culture with either HPMVEC or HUVEC, in the absence of cytomix-stimulation, had no effect on albumin leak over 8 h (Fig. 7). In contrast, the presence of neutrophils (5:1 ratio relative to the number of HPMVEC) was associated with a marked increase ( P < 0.01) in albumin leak across HPMVEC after stimulation with a low, sub-threshold dose (0.3 ng/ml) of cytomix (Fig. 7). In the absence of neutrophils, this dose of cytomix alone did not induce albumin leak across HPMVEC monolayers. Similarly, following stimulation with 0.1 ng/ml cytomix, albumin leak was also slightly, but not significantly, increased in the presence of neutrophils. In contrast, the presence of neutrophils (5:1 relative to HUVEC) had no effect on cytomix-induced albumin leak across HUVEC monolayers. Furthermore, an increased number of neutrophils (10:1 ratio relative to HUVEC) did not accentuate cytomix (0.3 ng/ml)-induced albumin leak in neutrophil-HUVEC co-culture (data not shown). In cytomix-treated neutrophil-HPMVEC co-cultures, the increase in albumin leak was dependent on the number of neutrophils present (Fig. 8). As the MTT assay for measurement of cell viability cannot be selectively applied to a single cell type (e.g. HPMVEC or HUVEC) in co-culture with neutrophils, the Trypan Blue
Fig. 4. Dose response of LPS-induced albumin leak in HPMVEC and HUVEC following 8 h stimulation. *P < 0.05; **P < 0.01 vs. respective unstimulated group (PBS). n = 3 – 6/group. PBS = phosphate-buffered saline.
H2O2 PBS
100 Ag/ml 10 Ag/ml 1 Ag/ml 0.1 Ag/ml 100 ng/ml 10 ng/ml 1 ng/ml 0.1 ng/ml 1 mM
Trypan blue exclusion assay Cytomix 10 ng/ml 0.3 ng/ml Neutrophils + cytomix 0.3 ng/ml Neutrophils alone H2O2 1 mM PBS
90 99 98 94 81 89 93 97 6 100
T T T T T T T T T T
3 2 0.4 4 4* 4 8 3 1* 8
88 100 102 109 81 96 93 100 5 100
90 97 98 96 0 98
T T T T T T
2 1 0.4 0.3 0* 0.3
92 96 97 96 0 98
T2 T3 T6 T3 T 1* T4 T 0.4 T3 T 0.4* T8
T T T T T T
1 0.3 0.5 0.2 0* 1
n = 3/group. Abbreviations: HPMVEC: human pulmonary microvascular endothelial cells, HUVEC: human umbilical vein endothelial cells, LPS: lipopolysaccharide, cytomix: equal concentrations of TNF-a, IL-1h, IFN-g, PBS: phosphatebuffered saline. * P < 0.05 vs. respective unstimulated group (PBS).
exclusion assay was used to assess HPMVEC and HUVEC viability following cytomix-stimulation in the presence of neutrophils. In HPMVEC and HUVEC monolayers in the absence of neutrophils, cell viability assessed by the Trypan Blue assay was similar to that measured by the MTT assay under conditions of no stimulation, and following H2O2 or cytomix treatment (Table 1). In HPMVEC or HUVEC cocultured with neutrophils, cytomix treatment was not associated with any significant loss of cell viability. Discussion We assessed the leak of albumin across HPMVEC and HUVEC monolayers under basal conditions and septic stimulation in the presence and absence of human neutrophils.
Fig. 5. Time course of cytomix (10 ng/ml of each of TNF-a, IL-1h and IFN-g)induced albumin leak in HPMVEC and HUVEC monolayers. *P < 0.05 vs. respective baseline (time 0). n = 3 – 6/group.
44
J.L. Shelton et al. / Microvascular Research 71 (2006) 40 – 47
Fig. 6. Dose response of cytomix-induced albumin leak in HPMVEC and HUVEC following 8 h stimulation. *P < 0.05; **P < 0.01 vs. respective unstimulated group (PBS). n = 3 – 6/group.
Under basal, unstimulated conditions, HPMVEC was less permeable to albumin than HUVEC monolayers. Relative to this basal leak, LPS stimulation increased albumin leak dosedependently across HPMVEC and HUVEC to a similar degree and with a similar time course. Cytomix stimulation also increased albumin leak across HPMVEC and HUVEC dosedependently and to a similar degree. At doses up to 10 Ag/ml and 10 ng/ml, respectively, LPS and cytomix-stimulated albumin leak across both HPMVEC and HUVEC was not due to loss of cell viability. However, at very high doses of both LPS and cytomix, cell viability was impaired and resulted in a dramatic increase in albumin leak across both HPMVEC and HUVEC. Treatment of neutrophil-HPMVEC co-cultures with a low, sub-threshold dose of cytomix induced a marked increase in albumin leak that was proportional to the number of neutrophils present. This neutrophil-dependent increase in cytomix-stimulated albumin leak across HPMVEC was not associated with loss of HPMVEC viability. In sharp contrast, the presence of neutrophils had no effect on cytomixstimulated albumin leak across HUVEC monolayers. ALI is a common and important clinical problem associated with significant morbidity and mortality (Ware and Matthay, 2000; Abraham et al., 2000; Angus et al., 2001). The pathologic hallmarks of ALI are high-protein pulmonary edema and pulmonary neutrophil infiltration (Bellingan, 2002). Based mainly on animal studies, injury and dysfunction of EC specifically in the pulmonary microvasculature are
Fig. 7. The effect of neutrophils on cytomix-induced albumin leak across HPMVEC and HUVEC monolayers. Trans-EC EB-albumin leak was assessed after 8 h incubation in the presence/absence of cytomix and/or neutrophils. **P < 0.01 vs. respective unstimulated group. n = 3 – 6/group.
Fig. 8. Effect of increasing ratio of neutrophils:HPMVEC on cytomixstimulated albumin leak following 8 h stimulation. *P < 0.01 vs. unstimulated HPMVEC in the absence of neutrophils. n = 3/group.
thought to be responsible for the key pathologic features of ALI. Studies of biologic responses and signaling pathways in these pulmonary microvascular EC under septic conditions in vitro are greatly increasing our understanding of the pathophysiology of sepsis-induced ALI (Razavi et al., 2004; Hadkar et al., 2004; Usatyuk and Natarajan, 2004; Speyer et al., 2003). However, EC from different species may demonstrate significant heterogeneity, such as a differential susceptibility to injury induced by activated neutrophils (Murphy et al., 1998). The availability of human pulmonary microvascular EC (HPMVEC) will permit direct study in vitro of the mechanisms of human ALI and endothelial injury. In the present study, HPMVEC formed a tight permeability barrier to the leak of albumin under basal conditions. Septic stimulation with either LPS or cytomix induced significant, dose-dependent trans-HPMVEC albumin leak. HPMVEC barrier function has previously been assessed under inflammatory conditions, such as exposure to human burn serum or individual cytokines, such as TNF-a (Murphy and Duffy, 2003; Sedgwick et al., 2002). The effects of inflammatory stimuli on HPMVEC adhesion molecule expression, interaction with neutrophils and viability have also been reported (Hamacher et al., 2002; Jiang et al., 2005; Wang et al., 2005; Wyman et al., 2002). However, the effect of human neutrophils on HPMVEC permeability under septic stimulation has never been assessed. We found that at low, sub-threshold doses of cytomix (no effect in the absence of neutrophils), HPMVEC was exquisitely sensitive to the presence of neutrophils, resulting in dramatically greater trans-HPMVEC albumin leak. Thus, the increase in HPMVEC permeability was mediated through cytokine-dependent neutrophil activation. Our findings are consistent with our understanding of the role of neutrophils in human ALI/ARDS (Tate and Repine, 1983; Downey et al., 1995; Wagner et al., 1999a; Bellingan, 2002). Although ALI/ ARDS can also occur in neutropenic patients, it is usually observed during resolution of neutropenia (Rinaldo and Borovetz, 1985; Azoulay et al., 2002). We also report significant differences in trans-EC albumin leak between HPMVEC and HUVEC. For example, HPMVEC established a tighter permeability barrier under basal, unstimulated conditions than did HUVEC. However, the relative increase in septic albumin leak (corrected for differences in basal, unstimulated leak) was similar between HPMVEC and
J.L. Shelton et al. / Microvascular Research 71 (2006) 40 – 47
HUVEC. The most striking difference found between HPMVEC and HUVEC was the susceptibility to human neutrophil-induced barrier dysfunction under septic conditions, which was unique to HPMVEC. Our findings are consistent with previous reports of differences between HPMVEC and HUVEC (Beck et al., 1999; Burg et al., 2002; Hillyer et al., 2003). For example, HPMVEC exhibited significantly greater PMN adhesion with TNF-a stimulation under venous flow conditions as compared to HUVEC (Otto et al., 2001). Our findings are also consistent with other reports of biologic differences between EC from different sources. This includes differences between microvascular EC, such as HPMVEC, and macrovascular EC, such as HUVEC (Beck et al., 1999; Chetham et al., 1999; Kelly et al., 1998; Moldobaeva and Wagner, 2002; Muth et al., 2004; Irwin et al., 2004), as well as between microvascular EC from different organs (Murphy et al., 1998; Invernici et al., 2005). Of relevance to our observations on neutrophil-induced leak, there are also important differences in the sensitivity of different EC to neutrophil-induced injury and cytotoxicity (Wang and Doerschuk, 2002; Murphy et al., 1998). For example, important differences were observed in neutrophil-dependent cytotoxicity between dermal, lung and renal microvascular EC (Murphy et al., 1998). Isolating, culturing and studying HPMVEC in vitro clearly do not capture the complex in vivo situation of human ALI, in which HPMVEC is exposed to blood flow, rapidly changing luminal mediator levels, and the influences of multiple other cell types. Co-culture of HPMVEC with relevant inflammatory and pulmonary parenchymal cells (Mul et al., 2000) is an in vitro construct of the pulmonary microvascular – interstitial interface, the key site of pathophysiologic events in ALI. In vitro studies of HPMVEC appear to be consistent with our in vivo understanding of the pathophysiology of ALI. For example, in vitro ICAM-1 expression in HPMVEC was found to be consistent with in vivo expression under inflammatory conditions (Muller et al., 2002). As well, significant cytotoxicity was found in HPMVEC exposed to bronchoalveolar lavage fluid supernatant from patients with early stage ARDS compared to that from control patients (Hamacher et al., 2002). We also recognize a limitation in studying HPMVEC ‘‘septic’’ barrier responses in vitro using LPS or cytomix. However, the use of plasma isolated from septic patients is also problematic, since significant differences in cytokine levels are found between patients, including the balance of pro-inflammatory (e.g. TNF-a) and anti-inflammatory (e.g. IL-10) cytokines (Cavaillon et al., 2003). In addition, the routine clinical use of multiple medications in sepsis, including antibiotics and vasopressors, may also confound the use of septic plasma for stimulation of HPVMEC in vitro. The use of LPS in animal models and in human cells is an acceptable surrogate for ‘‘septic’’ conditions, and the three component cytokines (TNF-a, IL-1h and IFN-g) used in ‘‘cytomix’’ are all relevant to the pathophysiology of sepsis (Takala et al., 2002; Scott et al., 2002). Another limitation of our study is not establishing the mechanism of ‘‘septic’’ leak. In patients with ALI, leak of
45
plasma proteins (e.g. albumin) into the lung may be due to death and shedding of pulmonary microvascular EC, EC contraction and generation of inter-EC gaps, or to trans-EC albumin transport. In the present report, both LPS and cytomix were cytotoxic for HPMVEC and HUVEC at very high doses, which was associated with dramatically increased albumin leak. Significant levels of albumin leak were observed across both HPMVEC and HUVEC at doses of LPS and cytomix that did not induce any loss of cell viability. Moreover, the enhanced cytomix-stimulated albumin leak across HPMVEC in the presence of neutrophils also was not simply due to increased HPMVEC cell death. Thus, our data suggest that septic and neutrophil-dependent leak across HPMVEC is a specific biological response, perhaps similar to the in vivo situation of HPMVEC injury in ALI leading to protein-rich pulmonary edema and pulmonary neutrophil infiltration. In summary, HUVEC is perhaps a less suitable model than HPMVEC for research into the molecular mechanisms of septic EC injury and neutrophil-induced injury of relevance to human ALI. Given that isolated EC may not be a perfect model for studying in vivo EC responses, direct in vivo studies will be helpful for a better understanding of the pathophysiology of human ALI and pulmonary EC responses. We believe that both in vitro and in vivo EC studies will ultimately lead to new therapeutic approaches to the common and serious clinical condition of human ALI. Acknowledgments Sources of support: Ontario Thoracic Society, Lawson Health Research Institute, UWO, LHSC Department of Medicine and HSFO (NA 5580). References Abraham, E., 1999. Why immunomodulatory therapies have not worked in sepsis. Intensive Care Med. 25, 556 – 566. Abraham, E., Matthay, M.A., Dinarello, C.A., Vincent, J.L., Cohen, J., Opal, S.M., Glauser, M., Parsons, P., Fisher, C.J.J., Repine, J.E., 2000. Consensus conference definitions for sepsis, septic shock, acute lung injury, and acute respiratory distress syndrome: time for a reevaluation. Crit. Care Med. 28, 232 – 235. Angus, D.C., Linde-Zwirble, W.T., Lidicker, J., Clermont, G., Carcillo, J., Pinsky, M.R., 2001. Epidemiology of severe sepsis in the United States: analysis of incidence, outcome, and associated costs of care. Crit. Care Med. 29, 1303 – 1310. Azoulay, E., Darmon, M., Delclaux, C., Fieux, F., Bornstain, C., Moreau, D., Attalah, H., Le Gall, J.R., Schlemmer, B., 2002. Deterioration of previous acute lung injury during neutropenia recovery. Crit. Care Med. 30, 781 – 786. Beck, G.C., Yard, B.A., Breedijk, A.J., Van Der Woude, F.J., 1999. Release of CXC-chemokines by human lung microvascular endothelial cells (LMVEC) compared with macrovascular umbilical vein endothelial cells. Clin. Exp. Immunol. 118, 298 – 303. Bellingan, G.J., 2002. The pathogenesis of ALI/ARDS. Thorax 57, 540 – 546. Bone, R.C., Balk, R.A., Cerra, F.B., Dellinger, R.P., Fein, A.M., Knaus, W.A., Schein, R.M., Sibbald, W.J., 1992. Definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. The ACCP/SCCM Consensus Conference Committee. American College of Chest Physicians/Society of Critical Care Medicine. Chest 101, 1644 – 1655.
46
J.L. Shelton et al. / Microvascular Research 71 (2006) 40 – 47
Burg, J., Krump-Konvalinkova, V., Bittinger, F., Kirkpatrick, C.J., 2002. GMCSF expression by human lung microvascular endothelial cells: in vitro and in vivo findings. Am. J. Physiol.: Lung Cell. Mol. Physiol. 283, L460 – L467. Cavaillon, J.M., Adib-Conquy, M., Fitting, C., Adrie, C., Payen, D., 2003. Cytokine cascade in sepsis. Scand. J. Infect. Dis. 35, 535 – 544. Chetham, P.M., Babal, P., Bridges, J.P., Moore, T.M., Stevens, T., 1999. Segmental regulation of pulmonary vascular permeability by. Am. J. Physiol: Lung Cell. Mol. Physiol. 276, L41 – L50. Curzen, N.P., Griffiths, M.J., Evans, T.W., 1994. Role of endothelium in modulating the vascular response to sepsis. Clin. Sci. 86, 359 – 374. De Staercke, C., Phillips, D.J., Hooper, W.C., 2003. Differential responses of human umbilical and coronary artery endothelial cells to apoptosis. Endothelium 10, 71 – 78. Doerschuk, C.M., 2001. Mechanisms of leukocyte sequestration in inflamed lungs. Microcirculation 8, 71 – 88. Downey, G.P., Fialkow, L., Fukushima, T., 1995. Initial interaction of leukocytes within the microvasculature: deformability, adhesion, and transmigration. New Horiz. 3, 219 – 228. Dull, R.O., Garcia, J.G.N., 2002. Leukocyte-induced microvascular permeability: how contractile tweaks lead to leaks. Circ. Res. 90, 1143 – 1144. Garcia, J.G., Verin, A.D., Herenyiova, M., English, D., 1998. Adherent neutrophils activate endothelial myosin light chain kinase: role in transendothelial migration. J. Appl. Physiol. 84, 1817 – 1821. Glauser, M.P., 2000. Pathophysiologic basis of sepsis: considerations for future strategies of intervention. Crit. Care Med. 28, S4 – S8. Granger, D.N., Kubes, P., 1994. The microcirculation and inflammation: modulation of leukocyte-endothelial cell adhesion. J. Leukocyte Biol. 55, 662 – 675. Groeneveld, A.B., 2002. Vascular pharmacology of acute lung injury and acute respiratory distress syndrome. Vascul. Pharmacol. 39, 247 – 256. Hadkar, V., Sangsree, S., Vogel, S.M., Brovkovych, V., Skidgel, R.A., 2004. Carboxypeptidase-mediated enhancement of nitric oxide production in rat lungs and microvascular endothelial cells. Am. J. Physiol.: Lung Cell. Mol. Physiol. 287, L35 – L45. Hamacher, J., Lucas, R., Lijnen, H.R., Buschke, S., Dunant, Y., Wendel, A., Grau, G.E., Suter, P.M., Ricou, B., 2002. Tumor necrosis factor-alpha and angiostatin are mediators of endothelial cytotoxicity in bronchoalveolar lavages of patients with acute respiratory distress syndrome. Am. J. Respir. Crit. Care Med. 166, 651 – 656. Hewett, P.W., Murray, J.C., 1993. Human lung microvessel endothelial cells: isolation, culture, and characterization. Microvasc. Res. 46, 89 – 102. Hillyer, P., Mordelet, E., Flynn, G., Male, D., 2003. Chemokines, chemokine receptors and adhesion molecules on different human endothelia: discriminating the tissue-specific functions that affect leucocyte migration. Clin. Exp. Immunol. 134, 431 – 441. Invernici, G., Ponti, D., Corsini, E., Cristini, S., Frigerio, S., Colombo, A., Parati, E., Alessandri, G., 2005. Human microvascular endothelial cells from different fetal organs demonstrate organ-specific CAM expression. Exp. Cell Res., 273 – 282. Ionescu, C.V., Cepinskas, G., Savickiene, J., Sandig, M., Kvietys, P.R., 2003. Neutrophils induce sequential focal changes in endothelial adherens junction components: role of elastase. Microcirculation 10, 205 – 220. Irwin, D.C., Tissot van Patot, M.C., Tucker, A., Bowen, R., 2004. Direct ANP inhibition of hypoxia-induced inflammatory pathways in pulmonary microvascular and macrovascular endothelial monolayers. Am. J. Physiol.: Lung Cell. Mol. Physiol., 849 – 859 (eprint). Jaffe, E.A., Nachman, R.L., Becker, C.G., Minick, C.R., 1973. Culture of human endothelial cells derived from umbilical veins. Identification by morphologic and immunologic criteria. J. Clin. Invest. 52, 2745 – 2756. Jiang, M.Z., Tsukahara, H., Hayakawa, K., Todoroki, Y., Tamura, S., Ohshima, Y., Hiraoka, M., Mayumi, M., 2005. Effects of antioxidants and NO on TNF-alpha-induced adhesion molecule expression in human pulmonary microvascular endothelial cells. Respir. Med. 99, 580 – 591. Kelly, J.J., Moore, T.M., Babal, P., Diwan, A.H., Stevens, T., Thompson, W.J., 1998. Pulmonary microvascular and macrovascular endothelial cells: differential regulation of Ca2+ and permeability. Am. J. Physiol: Lung Cell. Mol. Physiol. 274, L810 – L819.
King, J., Hamil, T., Creighton, J., Wu, S., Bhat, P., McDonald, F., Stevens, T., 2004. Structural and functional characteristics of lung macro- and microvascular endothelial cell phenotypes. Microvasc. Res. 67, 139 – 151. Moldobaeva, A., Wagner, E.M., 2002. Heterogeneity of bronchial endothelial cell permeability. Am. J. Physiol.: Lung Cell. Mol. Physiol. 283, L520 – L527. Mul, F.P., Zuurbier, A.E., Janssen, H., Calafat, J., van Wetering, S., Hiemstra, P.S., Roos, D., Hordijk, P.L., 2000. Sequential migration of neutrophils across monolayers of endothelial. J. Leukocyte Biol. 68, 529 – 537. Muller, A.M., Hermanns, M.I., Cronen, C., Kirkpatrick, C.J., 2002. Comparative study of adhesion molecule expression in cultured human macro- and microvascular endothelial cells. Exp. Mol. Pathol. 73, 171 – 180. Murphy, J.T., Duffy, S., 2003. ZO-1 redistribution and F-actin stress fiber formation in pulmonary endothelial cells after thermal injury. J. Trauma 54, 81 – 89. Murphy, M.P., Packer, M.A., Scarlett, J.L., Martin, S.W., 1998. Peroxynitrite—A biologically significant oxidant [review]. Gen. Pharmacol. 31, 179 – 186. Muth, H., Maus, U., Wygrecka, M., Lohmeyer, J., Grimminger, F., Seeger, W., Gunther, A., 2004. Pro- and antifibrinolytic properties of human pulmonary microvascular versus artery endothelial cells: impact of endotoxin and tumor. Crit. Care Med. 32, 217 – 226. Otto, M., Bittinger, F., Kriegsmann, J., Kirkpatrick, C.J., 2001. Differential adhesion of polymorphous neutrophilic granulocytes to macro- and microvascular endothelial cells under flow conditions. Pathobiology 69, 159 – 171. Patterson, C.E., Rhoades, R.A., Garcia, J.G., 1992. Evans blue dye as a marker of albumin clearance in cultured endothelial monolayer and isolated lungs. J. Appl. Physiol. 72, 865 – 873. Razavi, H.M., Werhun, R., Scott, J.A., Weicker, S., Wang, L.F., McCormack, D.G., Mehta, S., 2002. Effects of inhaled nitric oxide in a mouse model of sepsis-induced acute lung injury. Crit. Care Med. 30, 868 – 873. Razavi, H.M., Wang, L.F., Weicker, S., Rohan, M., Law, C., McCormack, D.G., Mehta, S., 2004. Pulmonary neutrophil infiltration in murine sepsis: role of inducible nitric oxide synthase. Am. J. Respir. Crit. Care Med. 170, 227 – 233. Reinhart, K., Bayer, O., Brunkhorst, F., Meisner, M., 2002. Markers of endothelial damage in organ dysfunction and sepsis. Crit. Care Med. 30, S302 – S312. Rinaldo, J.E., Borovetz, H., 1985. Deterioration of oxygenation and abnormal lung microvascular permeability during resolution of leukopenia in patients with diffuse lung injury. Am. Rev. Respir. Dis. 131, 579 – 583. Scott, M.J., Godshall, C.J., Cheadle, W.G., 2002. Jaks, STATs, cytokines, and sepsis. Clin. Diagn. Lab. Immunol. 9, 1153 – 1159. Sedgwick, J.B., Menon, I., Gern, J.E., Busse, W.W., 2002. Effects of inflammatory cytokines on the permeability of human lung microvascular endothelial cell monolayers and differential eosinophil transmigration. J. Allergy. Clin. Immunol. 110, 752 – 756. Speyer, C.L., Neff, T.A., Warner, R.L., Guo, R.F., Sarma, J.V., Riedemann, N.C., Murphy, M.E., Murphy, H.S., Ward, P.A., 2003. Regulatory effects of iNOS on acute lung inflammatory responses in mice. Am. J. Pathol. 163, 2319 – 2328. Takala, A., Nupponen, I., Kylanpaa-Back, M.L., Repo, H., 2002. Markers of inflammation in sepsis. Ann. Med. 34, 614 – 623. Tate, R.M., Repine, J.E., 1983. Neutrophils and the adult respiratory distress syndrome. Am. Rev. Respir. Dis. 128, 552 – 559. Usatyuk, P.V., Natarajan, V., 2004. Role of mitogen-activated protein kinases in 4-hydroxy-2-nonenal-induced actin remodeling and barrier function in endothelial cells. J. Biol. Chem. 279, 11789 – 11797. Wagner, J.G., Roth, R.A., 1999. Neutrophil migration during endotoxemia. J. Leukocyte Biol. 66, 10 – 24. Wagner, J.G., Driscoll, K.E., Roth, R.A., 1999a. Inhibition of pulmonary neutrophil trafficking during endotoxemia is dependent on the stimulus for migration. Am. J. Respir. Cell Mol. Biol. 20, 769 – 776. Wagner, M., Hermanns, I., Bittinger, F., Kirkpatrick, C.J., 1999b. Induction of stress proteins in human endothelial cells by heavy metal ions and heat shock. Am. J. Physiol. 277, L1026 – L1033.
J.L. Shelton et al. / Microvascular Research 71 (2006) 40 – 47 Wang, Q., Doerschuk, C.M., 2002. The signaling pathways induced by neutrophil – endothelial cell adhesion. Antioxid. Redox Signal. 4, 39 – 47. Wang, L.F., Patel, M., Razavi, H.M., Weicker, S., Joseph, M.G., McCormack, D.G., Mehta, S., 2002. Role of inducible nitric oxide synthase in pulmonary microvascular protein leak in murine sepsis. Am. J. Respir. Crit. Care Med. 165, 1634 – 1639. Wang, Q., Yerukhimovich, M., Gaarde, W.A., Popoff, I.J., Doerschuk, C.M., 2005. MKK3 and -6-dependent activation of p38alpha MAP kinase is required for cytoskeletal changes in pulmonary microvascular endothelial cells induced by ICAM-1 ligation. Am. J. Physiol.: Lung Cell. Mol. Physiol. 288, L359 – L369.
47
Ware, L.B., Matthay, M.A., 2000. The acute respiratory distress syndrome. N. Engl. J. Med. 342, 1334 – 1349. Ware, L.B., Conner, E.R., Matthay, M.A., 2001. von Willebrand factor antigen is an independent marker of poor outcome in patients with early acute lung injury. Crit. Care Med. 29, 2325 – 2331. Wheeler, A.P., Bernard, G.R., 1999. Treating patients with severe sepsis. N. Engl. J. Med. 340, 207 – 214. Wyman, T.H., Bjornsen, A.J., Elzi, D.J., Smith, C.W., England, K.M., Kelher, M., Silliman, C.C., 2002. A two-insult in vitro model of PMN-mediated pulmonary endothelial damage: requirements for adherence and chemokine release. Am. J. Physiol.: Cell Physiol. 283, C1592 – C1603.
