Eur. J. Immunol. 1990. 20: 2775-2781
Leukocyte/endothelial cell interactions
2775
Betty C. Hakkert, Jeannette M. Rentenaar, Willem G.Van Aken, Dirk Roos and Jan A.Van Mourik
A three-dimensional model system to study the interactions between human leukocytes and endothelial cells*
Central Laboratory of the Netherlands Red Cross Blood Transfusion Service, Amsterdam
Leukocyte adhesion to endothelial cells and migration into the subendothelial matrix was studied with a three-dimensional model system, consisting of human endothelial cells cultured on a loose collagen matrix.We developed a new method to separate the endothelial cell monolayer and adhering leukocytes from the subendothelial matrix, allowing simultaneous analysis of leukocyte adhesion and transendothelial migration. Monocytes adhered more avidly to untreated endothelial cells than did neutrophils (2.5 k 0.3 vs. 1.0 k 0.2 leukocytes per endothelial cell). Only a small fraction (10Y0-20%) of these leukocytes migrated into the subendothelium. Pretreatment of endothelial cells with interleukin 1 (IL 1) enhanced adhesion (20%), but not migration of monocytes. In contrast, neutrophil adhesion was markedly and in a time-dependent manner increased by IL1 treatment (i.e. 200% after 6 h and 110% after 24 h of IL1 treatment). Moreover, IL 1pretreatment enhanced neutrophil migration twofold. Activation of leukocytes with formyl-methionyl-leucyl-phenylalanine(fMLP) enhanced both monocyte and neutrophil adhesion, but did not affect leukocyte migration. Under all conditions, monocyte adhesion was only partly (30%-40%) inhibited by monoclonal antibodies (mAb) against the common fi subunit of the leukocyte-cell adhesion molecules (LeuCAM; CD18) and 25% -30% by mAb against the a subunit of LFA-1 (CDlla). In contrast, mAb against the a subunits of Mac-1 (CDllb) and p150,95 (CDllc) were hardly effective. fMLP-mediated neutrophil adhesion was reduced to below baseline levels by anti-LeuCAM (CD18) mAb, whereas the LeuCAM contribution in IL 1-mediated neutrophil adhesion was less pronounced and varied in time. IL 1-mediated neutrophil migration, however, was completely blocked by anti-LeuCAM mAb. fMLPmediated neutrophil adhesion was inhibited by mAb against the a subunits of Mac, while mAb against the a subunits of LFA-1 and Mac-1 both reduced IL 1-mediated adherence. In summary, we describe a novel leukocyte adhesiodmigration method and demonstrate that the contribution of the LeuCAM complex in leukocyte-endothelium interaction varies depending on cell type and stimulus used.
1 Introduction Investigation of the interactions between leukocytes and endothelial cells (EC) is essential for our understanding of normal vascular biology and the pathogenesis of such various conditions as artherosclerosis, tumor invasion and inflammation [l,21. The study of the early events in these interactions, which include the adhesion of leukocytes to EC and their subsequent diapedesis into the subendothelial compartment, requires an appropriate model system. Several investigators have studied the interaction of leukocytes and EC by using EC cultured on various substrates, such as fibronectin, gelatin and porous filters [3-51. These in vitro
[I 87461
* This study was supported by the Netherlands Heart Foundation (grant no. 86.068).
Correspondence: Jan A. Van Mourik, c/o Publication Secretariat, Central Laboratory of the Netherlands Red Cross Blood Transfusion Service, PO. Box9406; NL-1006 AK Amsterdam, The Netherlands
Abbreviations: ICAM-1: Intercellular cell adhesion molecule-1 LeuCAM: Leukocyte-cell adhesion molecules LFA-1: Leukocyte function-associated antigen-1 EC: Endothelial cell(s) fMLP: Formyl-methionyl-leucyl-phenylalanine 0 VCH Verlagsgesellschaft mbH, D-6940 Weinheim, 1990
studies have shown that basal adhesion of monocytes ( 5 to 8 monocytes/EC) far exceeds that of neutrophils (1 neutrophil/EC) [6]. Recent attempts to define the molecular basis for leukocyte adhesion to E C have focused on a family of three structurally related molecules, the leukocyte-cell adhesion molecules (LeuCAM, CD18) [3, 4, 71. The LeuCAM comprise LFA-1 (CDlla, CD18), Mac-1 ( C D l l b , CD18) and p150,95 (CDllc, CD18) [8]. These three antigens have distinct a subunits (180, 170 and 150 kDa, respectively) noncovalently linked to an identical 95-kDa fi subunit (CD18) @].Therelevance of the LeuCAM in monocyte and neutrophil adhesion to EC is reflected by the finding that leukocytes from patients genetically deficient in this glycoprotein family show an impaired adhesion to EC in vitro [9]. Previous studies have shown that mAb directed against the common fi subunit of the LeuCAM partly inhibit monocyte and neutrophil adhesion to EC [7, 10, 111. Leukocyte adhesion to EC cells is enhanced by several inflammatory stimuli [12, 131. Among these, phorbol esters and chemotactic factors such as formyl-methionyl-leucyl-phenylalanine (fMLP) and leukotriene B4 (LTBJ) primarily act on leukocytes [14, 151, whereas I L 1 and TNF increase the intercellular binding by acting on EC [ 16, 171. IL 1and TNF induce the expression of at least three endothelial adhesion molecules, intercellular adhesion molecule-1 (ICAM-1) 0014-2980/90/1212-2775$3S O + .25/0
B. C. Hakkert, J. M. Rentenaar, W. G. Van Aken et al.
Eur. J. Irnmunol. 1990. 20: 2775-2781
[18], endothelial leukocyte adhesion molecule-1 (ELAM1) [19, 201 and vascular cell adhesion molecule 1 (VCAM1) [21, 221.
ty was > 98% (exclusion of ethidium bromide and hydrolysis of fluorescein diacetate). FCM analysis with the platelet-specific mAb MB 45 against CD42b and CLBthromb/6 (C-2) against CD62 [28] revealed that the monocyte preparations were not contaminated with platelets. Neutrophils were isolated from the pellet fraction of the Percoll gradient. The erythrocytes were lysed with isotonic ammonium chloride at 4°C [29]. Neutrophils were > 95% pure, and their viability was > 97%. Purified leukocytes were resuspended in “labeling medium” (equal mixture of RPMI 1640 and M 199, with 2 mM L-glutamine and 0.1% HSA (Central Laboratory of the Netherlands Red Cross Blood Transfusion Service, Amsterdam, The Netherlands).
2776
The above-mentioned studies were performed with EC cultured on either fibronectin or gelatin, and it is still unknown whether these processes also occur when E C are grown on more physiological culture substrates such as collagen. Therefore, we have used a 3-D model system consisting of human EC cultured on a loose collagen matrix to study the adhesion and migration of leukocytes. Because of its spatial arrangement, this model system approximates the architecture of the vessel wall. Recent studies, regarding the secretion of Von Willebrand factor [23] have shown that EC cultured under these conditions behave in a polar way. In the present report we describe a new method to quantify both leukocyte adhesion to EC and their migration into the subendothelial matrix.
2 Materials and methods 2.1 Cell culture EC were isolated from human umbilical cord veins according to Jaffe et al. [24],with some minor modifications [25]. EC were cultured in medium consisting of an equal mixture of RPMI 1640 and Medium 199 (both from Gibco, Paisley, Scotland) supplemented with 20% (v/v) human serum (pool of 24 healthy donors), 2 mM L-glutamine (Merck, Darmstadt, FRG), penicillin (100 U/ml) and streptomycin (100 pg/ml). The cells were identified by their typical characteristics [25]. Confluent EC cultures (either first or second passage) were harvested with trypsin/EDTA (Gibco) and subcultured on collagen lattices.
2.2 Preparation of collagen matrices A solution of calf skin collagen (0.25 ml, 0.3% w/v; type 111, Sigma Chemical Co., St. Louis, MO) was allowed to polymerize and form a gel at 37°C for 60 min [26] in 48-well microtiter plates (no. 3548, Costar Europe, Badhoevedorp,The Netherlands). After gelation, the collagen gels were sterilized for 30 min by ultraviolet radiation. EC (4 x lo4 cells/cm2) were seeded on top of the gel and were grown to confluence. Confluency was determined by May-Griinwald/Giemsa staining.
2.3 Leukocytes Monocytes and neutrophils (PMN) were isolated from buffy coats of 500 ml of human blood anticoagulated with 0.4% (w/v) trisodium citrate (pH 7.4) obtained from healthy volunteers, as described previously [27]. Briefly, the buffy coat was diluted three times with PBS containing 13 mM trisodium citrate and was centrifuged (20 min, 1000 x g) at room temperature over Percoll (Pharmacia, Uppsala, Sweden; at room temperature e = 1.077 g/cm3). For the isolation of monocytes the monunuclear cells were further purified by countercurrent centrifugal elutriation. The monocyte preparation was > 95% pure (nonspecific esterase and May-Grunwald/Giemsa staining), and viabili-
2.4 Labeling of leukocytes
Freshly purified monocytes and neutrophils were radiolabeled with 51Cr according to the method of Gallin et al. [30]. Briefly, leukocytes (1 X 107/ml) were incubated with 1 pCi = 37 kBq 51Cr/106cells (sodium chromate, 200 to 500 Ci/g, New England Nuclear, Boston, MA) at 37 “C for 1 h with gentle agitation. After incubation, the unbound 51Cr was removed in three washes with warm (37°C) “incubation medium” (M 199/RPMI 1640,2 m M L-glutamine, 0.5% HSA). Labeled leukocytes were resuspended in incubation medium at a final concentration of 8 x 1OS/ml and were kept on ice until use.Viability after labeling was > 95%. 2.5 Adhesion/migration assay
Confluent EC monolayers cultured on collagen gels were washed twice with warm (37 “C) incubation medium. Radiolabeled leukocytes (2 X lo5cells/well, i.e. five leukocytesEC) were added to each well, and culture plates were gently agitated for 1 min to promote contact of the leukocyte suspension with the EC monolayers. All incubations were run in quadruplicate. Plates were incubated for 20 min or as indicated, at 37°C in a 5% COz incubator. After incubation, the medium with unattached cells was collected, and the EC monolayers were subsequently washed twice with 0.25 ml warm (37°C) incubation medium. These three fractions were pooled (lumenal medium). The intact EC monolayers together with adhering leukocytes were harvested by incubation with 0.1 ml of collagenase (100 U/ml; Worthington Biochem. Corp., Freehold, NJ) for 10 min at 37°C (EC fraction). Subsequently, the collagen matrices were digested by collagenase (2 h, 37”C), incubated with 0.2 ml of 1% Triton X-100 (30 min) and collected (subendothelial matrix). Radioactivity was measured in the three fractions, and the number of leukocytes in the fractions was calculated from the total number of leukocytes (counts) incubated.The recovery was > 92%. 2.6 mAb CLB-LFA-1/1 (murine IgG1, [31]) and IBJ (murine IgGl, kindly provided by Dr. S. Wright, Rockefeller University, New York, NY) [32]), are directed against the common chain (CD18) of LFA-1, Mac-1 and p150,90. CLB-LFA-1R
Eur. J. Immunol. 1990. 20: 2775-2781
(murine IgGI, [31]) is directed against the a subunit of LFA-1 (CDlla). CLB-B2.12 (murine IgM, [33]) recognizes the a chain of Mac-1 (CD1lb). Leu-MS (murine IgGZh, Becton Dickinson, Mountain View, CA) is directed against the a chain of p150,95 (CDllc). CLB-MONO-1 (murine IgG2,) directed against CD14 antigen, CLB-B4.3 (murine IgM. [34]) against fucosyl-N-acetyllactosamine on neutrophils and an mAb (murine IgG,) against an irrelevant allergen (anti-pollen) were used as control mAb. mAb were used as purified Ig at a concentration of 15 yg/ml. WCAM-1 (IgGZb, a gift from Dr. A.W. Boyd,Walter and Eliza Hall Institute, Melbourne Hospital, Melbourne, Australia, [35]) is directed against ICAM-1 and was used as ascites (dilution 1/100). mAb were preincubated with leukocytes or EC monolayers (W-CAM-1) for 15 min at 37"C, and were not removed during the assay.
2.7 Pretreatment of EC and leukocytes fMLP ( l o p hM, Sigma) was added to the leukocytes 10 min after the addition of the mAb, thereafter the leukocytes were incubated with the EC monolayers. Confluent E C monolayers were preincubated with 10 U/ml of rIL 1p (gift from Dr. P.T. Lomedico, Hoffmann-La Roche, Nutley, NJ; spec. act. 5 x lohU/rng) in incubation medium for 6 or 24 h. Thereafter, the monolayers were washed twice with incubation medium and the adherence assay was performed.
Leukocyte/endothcliaI cell interactions 100 I
+EC
60
0
40
120
Figure 1. Monocyte adhesion to E C grown on a collagcn gel and migration into the subendothelial matrix. Lumenal medium (0-O), E C fraction (4-4) and subendothelial matrix (0-0). In the prcsence of EC (A) and in the absence o f EC (B). Results are means of six independent observations SD.
+
MONOCYTES
NEUTROPHILS
3-
3
2 -
el W
I?
z
0 W I
9
n
W
3 1 -
s
0 0
.
For statistical analysis two-tailed Student's r-tests were performed; p values exceeding 0.05 were considered not significant.
c c I
0 -
3 Results
-
INCUBATION TIME WIN)
?.i
2.8 Statistical analysis
2777
h J
T 0
3.1 Baseline leukocyte adhesion and migration in the 3D model system Kinetic studies showed that monocytes adhere avidly to EC monolayers cultured on collagen matrices (Fig. 1A). After 60 min of incubation, the mean monocyte adhesion was 60% -70%. Light microscopic examination of the E C monolayers showed that most of the monocytes were located at the margins of the EC (not shown). A maximum of adhesion was reached after 1.5-2 h of incubation. At that time only a small fraction (15 f 5%, mean f SD) of the monocytes had migrated into the subendothelial matrix (Fig. 1A).Visual counting of the number of adhering and migrated monocytes revealed the same percentages as obtained by the chromium-labeling method. Similar results were obtained when EC and monocytes were incubated at different ratios varying from 1/1 to 1/8 (data not shown). Control experiments in the absence of EC revealed that monocytes hardly adhered to or migrated into the collagen matrices (Fig., 1B). Even after 2 h of incubation, more than 80% of the monocytes were recovered in the lumenal medium. Baseline monocyte adhesion to EC cultured on collagen far exceeded that of neutrophils: 2.5 f 0.3 monocyteslEC and
Figure 2. Effect of mAb against the LeuCAM complex on baseline leukocyte adhesion to, and migration through EC monolayers. Leukocytes and EC were incubated fgr 20 min. Bars represent the number of adheringlmigrated leukocytes/EC. Values are means f SEM of six independent experiments. ( * ) p < 0.05 compared to isotype-matched control mAb.
1 . 0 f 0 . 2 (PMN/EC, i.e. 5 0 + 5 % and 2 0 f 3 % (mean f SEM) of the leukocytes added, respectively (Fig. 2). As was observed for monocytes (Fig. 1A), transendothelial neutrophil migration was very low in the unstimulatcd situation (Fig. 2 ) .
3.2 Effect of anti-LeuCAM mAb on baseline leukocyte adhesion and migration Monocyte adhesion was significantly inhibited by mAb against the common fi subunit of the LeuCAM (Fig. 2). Both antibodies, CLB-LFA-1/1 and IB-4, had an identical inhibitory effect of about 30% compared to isotypematched control antibodies.The mAb against the a subunit
B. C. Hakkert, J. M. Rentenaar, W. G.Van Aken et al.
Eur. J. Immunol. 1990.20: 2775-2781
of LFA-1 also inhibited monocyte adherence to EC (25 k 5% inhibition). The mAb against the a subunits of Mac-1 and p150,90, CLB-B2.12 and Leu-M5, however, had hardly any effect on monocyte adherence to EC.Thus, basal monocyte adhesion was only partly blocked by mAb against the LeuCAM. Even in experiments in which combinations of the anti-LeuCAM mAb were used, inhibition never exceeded 30%. Similar results were obtained when EC were cultured on fibronectin instead of collagen matrices (data not shown).The effects of LeuCAM mAb on baseline neutrophil adhesion were not extensively examined because of the low binding affinity of neutrophils to untreated EC. CLB-LFA-111, an antibody against the common LeuCAM p chain, had n o significant effect on basal neutrophil adherence.
ence was partly reduced by the mAb against the common LeuCAM fi chain (49 f 6% and 51 k 9% (p I
I Y 4
s ..: .....
i.:.:'.
n
u
:UI.: a:
.
n
5:
c
~
-i
j :-: !..+:.:'i UI:
,ei
-
i
7
1
I
:.:...:; ...:.:.:... :...:.:.... .... , *.:.:
z
P 7 -1
Figure 5. Effect of fMLP stimulation on leukocyte adhesion to, and migration through, E C monolayers. fMLP was added to the leukocytes 5 min prior to their incubation with EC. (W) Controls without stimulation and (0)fMLP stimulation. Leukocytes and EC were incubated for 20 min. Bars represent number of adhering/migrated leukocytes per EC. Values are means f SEM of five independent experiments. ( * ) p < 0.05 and ( * * ) p < 0.01 compared to isotype-matched control mAb.
2779
Neutrophil adhesion was strongly enhanced ( 150Y0) by fMLP addition. In contrast to the effects observed with IL 1-treated EC, fMLP did not affect neutrophil migration (Fig. 5). Furthermore, the contribution of the LeuCAM in neutrophil adhesion to cytokine-activated EC also differed from that of the fMLP-mediated neutrophil adherence. fMLP-mediated neutrophil adhesion was reduced to about 20% below baseline levels by anti-CD18 mAb (p < 0.01). With respect to the contribution of the members of the LeuCAM complex, fMLP-mediated neutrophil adhesion was predominantly inhibited by the mAb against the a subunit of Mac-1 (47 k 9%; p
Leukocyte/endothelial cell interactions
2775
Betty C. Hakkert, Jeannette M. Rentenaar, Willem G.Van Aken, Dirk Roos and Jan A.Van Mourik
A three-dimensional model system to study the interactions between human leukocytes and endothelial cells*
Central Laboratory of the Netherlands Red Cross Blood Transfusion Service, Amsterdam
Leukocyte adhesion to endothelial cells and migration into the subendothelial matrix was studied with a three-dimensional model system, consisting of human endothelial cells cultured on a loose collagen matrix.We developed a new method to separate the endothelial cell monolayer and adhering leukocytes from the subendothelial matrix, allowing simultaneous analysis of leukocyte adhesion and transendothelial migration. Monocytes adhered more avidly to untreated endothelial cells than did neutrophils (2.5 k 0.3 vs. 1.0 k 0.2 leukocytes per endothelial cell). Only a small fraction (10Y0-20%) of these leukocytes migrated into the subendothelium. Pretreatment of endothelial cells with interleukin 1 (IL 1) enhanced adhesion (20%), but not migration of monocytes. In contrast, neutrophil adhesion was markedly and in a time-dependent manner increased by IL1 treatment (i.e. 200% after 6 h and 110% after 24 h of IL1 treatment). Moreover, IL 1pretreatment enhanced neutrophil migration twofold. Activation of leukocytes with formyl-methionyl-leucyl-phenylalanine(fMLP) enhanced both monocyte and neutrophil adhesion, but did not affect leukocyte migration. Under all conditions, monocyte adhesion was only partly (30%-40%) inhibited by monoclonal antibodies (mAb) against the common fi subunit of the leukocyte-cell adhesion molecules (LeuCAM; CD18) and 25% -30% by mAb against the a subunit of LFA-1 (CDlla). In contrast, mAb against the a subunits of Mac-1 (CDllb) and p150,95 (CDllc) were hardly effective. fMLP-mediated neutrophil adhesion was reduced to below baseline levels by anti-LeuCAM (CD18) mAb, whereas the LeuCAM contribution in IL 1-mediated neutrophil adhesion was less pronounced and varied in time. IL 1-mediated neutrophil migration, however, was completely blocked by anti-LeuCAM mAb. fMLPmediated neutrophil adhesion was inhibited by mAb against the a subunits of Mac, while mAb against the a subunits of LFA-1 and Mac-1 both reduced IL 1-mediated adherence. In summary, we describe a novel leukocyte adhesiodmigration method and demonstrate that the contribution of the LeuCAM complex in leukocyte-endothelium interaction varies depending on cell type and stimulus used.
1 Introduction Investigation of the interactions between leukocytes and endothelial cells (EC) is essential for our understanding of normal vascular biology and the pathogenesis of such various conditions as artherosclerosis, tumor invasion and inflammation [l,21. The study of the early events in these interactions, which include the adhesion of leukocytes to EC and their subsequent diapedesis into the subendothelial compartment, requires an appropriate model system. Several investigators have studied the interaction of leukocytes and EC by using EC cultured on various substrates, such as fibronectin, gelatin and porous filters [3-51. These in vitro
[I 87461
* This study was supported by the Netherlands Heart Foundation (grant no. 86.068).
Correspondence: Jan A. Van Mourik, c/o Publication Secretariat, Central Laboratory of the Netherlands Red Cross Blood Transfusion Service, PO. Box9406; NL-1006 AK Amsterdam, The Netherlands
Abbreviations: ICAM-1: Intercellular cell adhesion molecule-1 LeuCAM: Leukocyte-cell adhesion molecules LFA-1: Leukocyte function-associated antigen-1 EC: Endothelial cell(s) fMLP: Formyl-methionyl-leucyl-phenylalanine 0 VCH Verlagsgesellschaft mbH, D-6940 Weinheim, 1990
studies have shown that basal adhesion of monocytes ( 5 to 8 monocytes/EC) far exceeds that of neutrophils (1 neutrophil/EC) [6]. Recent attempts to define the molecular basis for leukocyte adhesion to E C have focused on a family of three structurally related molecules, the leukocyte-cell adhesion molecules (LeuCAM, CD18) [3, 4, 71. The LeuCAM comprise LFA-1 (CDlla, CD18), Mac-1 ( C D l l b , CD18) and p150,95 (CDllc, CD18) [8]. These three antigens have distinct a subunits (180, 170 and 150 kDa, respectively) noncovalently linked to an identical 95-kDa fi subunit (CD18) @].Therelevance of the LeuCAM in monocyte and neutrophil adhesion to EC is reflected by the finding that leukocytes from patients genetically deficient in this glycoprotein family show an impaired adhesion to EC in vitro [9]. Previous studies have shown that mAb directed against the common fi subunit of the LeuCAM partly inhibit monocyte and neutrophil adhesion to EC [7, 10, 111. Leukocyte adhesion to EC cells is enhanced by several inflammatory stimuli [12, 131. Among these, phorbol esters and chemotactic factors such as formyl-methionyl-leucyl-phenylalanine (fMLP) and leukotriene B4 (LTBJ) primarily act on leukocytes [14, 151, whereas I L 1 and TNF increase the intercellular binding by acting on EC [ 16, 171. IL 1and TNF induce the expression of at least three endothelial adhesion molecules, intercellular adhesion molecule-1 (ICAM-1) 0014-2980/90/1212-2775$3S O + .25/0
B. C. Hakkert, J. M. Rentenaar, W. G. Van Aken et al.
Eur. J. Irnmunol. 1990. 20: 2775-2781
[18], endothelial leukocyte adhesion molecule-1 (ELAM1) [19, 201 and vascular cell adhesion molecule 1 (VCAM1) [21, 221.
ty was > 98% (exclusion of ethidium bromide and hydrolysis of fluorescein diacetate). FCM analysis with the platelet-specific mAb MB 45 against CD42b and CLBthromb/6 (C-2) against CD62 [28] revealed that the monocyte preparations were not contaminated with platelets. Neutrophils were isolated from the pellet fraction of the Percoll gradient. The erythrocytes were lysed with isotonic ammonium chloride at 4°C [29]. Neutrophils were > 95% pure, and their viability was > 97%. Purified leukocytes were resuspended in “labeling medium” (equal mixture of RPMI 1640 and M 199, with 2 mM L-glutamine and 0.1% HSA (Central Laboratory of the Netherlands Red Cross Blood Transfusion Service, Amsterdam, The Netherlands).
2776
The above-mentioned studies were performed with EC cultured on either fibronectin or gelatin, and it is still unknown whether these processes also occur when E C are grown on more physiological culture substrates such as collagen. Therefore, we have used a 3-D model system consisting of human EC cultured on a loose collagen matrix to study the adhesion and migration of leukocytes. Because of its spatial arrangement, this model system approximates the architecture of the vessel wall. Recent studies, regarding the secretion of Von Willebrand factor [23] have shown that EC cultured under these conditions behave in a polar way. In the present report we describe a new method to quantify both leukocyte adhesion to EC and their migration into the subendothelial matrix.
2 Materials and methods 2.1 Cell culture EC were isolated from human umbilical cord veins according to Jaffe et al. [24],with some minor modifications [25]. EC were cultured in medium consisting of an equal mixture of RPMI 1640 and Medium 199 (both from Gibco, Paisley, Scotland) supplemented with 20% (v/v) human serum (pool of 24 healthy donors), 2 mM L-glutamine (Merck, Darmstadt, FRG), penicillin (100 U/ml) and streptomycin (100 pg/ml). The cells were identified by their typical characteristics [25]. Confluent EC cultures (either first or second passage) were harvested with trypsin/EDTA (Gibco) and subcultured on collagen lattices.
2.2 Preparation of collagen matrices A solution of calf skin collagen (0.25 ml, 0.3% w/v; type 111, Sigma Chemical Co., St. Louis, MO) was allowed to polymerize and form a gel at 37°C for 60 min [26] in 48-well microtiter plates (no. 3548, Costar Europe, Badhoevedorp,The Netherlands). After gelation, the collagen gels were sterilized for 30 min by ultraviolet radiation. EC (4 x lo4 cells/cm2) were seeded on top of the gel and were grown to confluence. Confluency was determined by May-Griinwald/Giemsa staining.
2.3 Leukocytes Monocytes and neutrophils (PMN) were isolated from buffy coats of 500 ml of human blood anticoagulated with 0.4% (w/v) trisodium citrate (pH 7.4) obtained from healthy volunteers, as described previously [27]. Briefly, the buffy coat was diluted three times with PBS containing 13 mM trisodium citrate and was centrifuged (20 min, 1000 x g) at room temperature over Percoll (Pharmacia, Uppsala, Sweden; at room temperature e = 1.077 g/cm3). For the isolation of monocytes the monunuclear cells were further purified by countercurrent centrifugal elutriation. The monocyte preparation was > 95% pure (nonspecific esterase and May-Grunwald/Giemsa staining), and viabili-
2.4 Labeling of leukocytes
Freshly purified monocytes and neutrophils were radiolabeled with 51Cr according to the method of Gallin et al. [30]. Briefly, leukocytes (1 X 107/ml) were incubated with 1 pCi = 37 kBq 51Cr/106cells (sodium chromate, 200 to 500 Ci/g, New England Nuclear, Boston, MA) at 37 “C for 1 h with gentle agitation. After incubation, the unbound 51Cr was removed in three washes with warm (37°C) “incubation medium” (M 199/RPMI 1640,2 m M L-glutamine, 0.5% HSA). Labeled leukocytes were resuspended in incubation medium at a final concentration of 8 x 1OS/ml and were kept on ice until use.Viability after labeling was > 95%. 2.5 Adhesion/migration assay
Confluent EC monolayers cultured on collagen gels were washed twice with warm (37 “C) incubation medium. Radiolabeled leukocytes (2 X lo5cells/well, i.e. five leukocytesEC) were added to each well, and culture plates were gently agitated for 1 min to promote contact of the leukocyte suspension with the EC monolayers. All incubations were run in quadruplicate. Plates were incubated for 20 min or as indicated, at 37°C in a 5% COz incubator. After incubation, the medium with unattached cells was collected, and the EC monolayers were subsequently washed twice with 0.25 ml warm (37°C) incubation medium. These three fractions were pooled (lumenal medium). The intact EC monolayers together with adhering leukocytes were harvested by incubation with 0.1 ml of collagenase (100 U/ml; Worthington Biochem. Corp., Freehold, NJ) for 10 min at 37°C (EC fraction). Subsequently, the collagen matrices were digested by collagenase (2 h, 37”C), incubated with 0.2 ml of 1% Triton X-100 (30 min) and collected (subendothelial matrix). Radioactivity was measured in the three fractions, and the number of leukocytes in the fractions was calculated from the total number of leukocytes (counts) incubated.The recovery was > 92%. 2.6 mAb CLB-LFA-1/1 (murine IgG1, [31]) and IBJ (murine IgGl, kindly provided by Dr. S. Wright, Rockefeller University, New York, NY) [32]), are directed against the common chain (CD18) of LFA-1, Mac-1 and p150,90. CLB-LFA-1R
Eur. J. Immunol. 1990. 20: 2775-2781
(murine IgGI, [31]) is directed against the a subunit of LFA-1 (CDlla). CLB-B2.12 (murine IgM, [33]) recognizes the a chain of Mac-1 (CD1lb). Leu-MS (murine IgGZh, Becton Dickinson, Mountain View, CA) is directed against the a chain of p150,95 (CDllc). CLB-MONO-1 (murine IgG2,) directed against CD14 antigen, CLB-B4.3 (murine IgM. [34]) against fucosyl-N-acetyllactosamine on neutrophils and an mAb (murine IgG,) against an irrelevant allergen (anti-pollen) were used as control mAb. mAb were used as purified Ig at a concentration of 15 yg/ml. WCAM-1 (IgGZb, a gift from Dr. A.W. Boyd,Walter and Eliza Hall Institute, Melbourne Hospital, Melbourne, Australia, [35]) is directed against ICAM-1 and was used as ascites (dilution 1/100). mAb were preincubated with leukocytes or EC monolayers (W-CAM-1) for 15 min at 37"C, and were not removed during the assay.
2.7 Pretreatment of EC and leukocytes fMLP ( l o p hM, Sigma) was added to the leukocytes 10 min after the addition of the mAb, thereafter the leukocytes were incubated with the EC monolayers. Confluent E C monolayers were preincubated with 10 U/ml of rIL 1p (gift from Dr. P.T. Lomedico, Hoffmann-La Roche, Nutley, NJ; spec. act. 5 x lohU/rng) in incubation medium for 6 or 24 h. Thereafter, the monolayers were washed twice with incubation medium and the adherence assay was performed.
Leukocyte/endothcliaI cell interactions 100 I
+EC
60
0
40
120
Figure 1. Monocyte adhesion to E C grown on a collagcn gel and migration into the subendothelial matrix. Lumenal medium (0-O), E C fraction (4-4) and subendothelial matrix (0-0). In the prcsence of EC (A) and in the absence o f EC (B). Results are means of six independent observations SD.
+
MONOCYTES
NEUTROPHILS
3-
3
2 -
el W
I?
z
0 W I
9
n
W
3 1 -
s
0 0
.
For statistical analysis two-tailed Student's r-tests were performed; p values exceeding 0.05 were considered not significant.
c c I
0 -
3 Results
-
INCUBATION TIME WIN)
?.i
2.8 Statistical analysis
2777
h J
T 0
3.1 Baseline leukocyte adhesion and migration in the 3D model system Kinetic studies showed that monocytes adhere avidly to EC monolayers cultured on collagen matrices (Fig. 1A). After 60 min of incubation, the mean monocyte adhesion was 60% -70%. Light microscopic examination of the E C monolayers showed that most of the monocytes were located at the margins of the EC (not shown). A maximum of adhesion was reached after 1.5-2 h of incubation. At that time only a small fraction (15 f 5%, mean f SD) of the monocytes had migrated into the subendothelial matrix (Fig. 1A).Visual counting of the number of adhering and migrated monocytes revealed the same percentages as obtained by the chromium-labeling method. Similar results were obtained when EC and monocytes were incubated at different ratios varying from 1/1 to 1/8 (data not shown). Control experiments in the absence of EC revealed that monocytes hardly adhered to or migrated into the collagen matrices (Fig., 1B). Even after 2 h of incubation, more than 80% of the monocytes were recovered in the lumenal medium. Baseline monocyte adhesion to EC cultured on collagen far exceeded that of neutrophils: 2.5 f 0.3 monocyteslEC and
Figure 2. Effect of mAb against the LeuCAM complex on baseline leukocyte adhesion to, and migration through EC monolayers. Leukocytes and EC were incubated fgr 20 min. Bars represent the number of adheringlmigrated leukocytes/EC. Values are means f SEM of six independent experiments. ( * ) p < 0.05 compared to isotype-matched control mAb.
1 . 0 f 0 . 2 (PMN/EC, i.e. 5 0 + 5 % and 2 0 f 3 % (mean f SEM) of the leukocytes added, respectively (Fig. 2). As was observed for monocytes (Fig. 1A), transendothelial neutrophil migration was very low in the unstimulatcd situation (Fig. 2 ) .
3.2 Effect of anti-LeuCAM mAb on baseline leukocyte adhesion and migration Monocyte adhesion was significantly inhibited by mAb against the common fi subunit of the LeuCAM (Fig. 2). Both antibodies, CLB-LFA-1/1 and IB-4, had an identical inhibitory effect of about 30% compared to isotypematched control antibodies.The mAb against the a subunit
B. C. Hakkert, J. M. Rentenaar, W. G.Van Aken et al.
Eur. J. Immunol. 1990.20: 2775-2781
of LFA-1 also inhibited monocyte adherence to EC (25 k 5% inhibition). The mAb against the a subunits of Mac-1 and p150,90, CLB-B2.12 and Leu-M5, however, had hardly any effect on monocyte adherence to EC.Thus, basal monocyte adhesion was only partly blocked by mAb against the LeuCAM. Even in experiments in which combinations of the anti-LeuCAM mAb were used, inhibition never exceeded 30%. Similar results were obtained when EC were cultured on fibronectin instead of collagen matrices (data not shown).The effects of LeuCAM mAb on baseline neutrophil adhesion were not extensively examined because of the low binding affinity of neutrophils to untreated EC. CLB-LFA-111, an antibody against the common LeuCAM p chain, had n o significant effect on basal neutrophil adherence.
ence was partly reduced by the mAb against the common LeuCAM fi chain (49 f 6% and 51 k 9% (p I
I Y 4
s ..: .....
i.:.:'.
n
u
:UI.: a:
.
n
5:
c
~
-i
j :-: !..+:.:'i UI:
,ei
-
i
7
1
I
:.:...:; ...:.:.:... :...:.:.... .... , *.:.:
z
P 7 -1
Figure 5. Effect of fMLP stimulation on leukocyte adhesion to, and migration through, E C monolayers. fMLP was added to the leukocytes 5 min prior to their incubation with EC. (W) Controls without stimulation and (0)fMLP stimulation. Leukocytes and EC were incubated for 20 min. Bars represent number of adhering/migrated leukocytes per EC. Values are means f SEM of five independent experiments. ( * ) p < 0.05 and ( * * ) p < 0.01 compared to isotype-matched control mAb.
2779
Neutrophil adhesion was strongly enhanced ( 150Y0) by fMLP addition. In contrast to the effects observed with IL 1-treated EC, fMLP did not affect neutrophil migration (Fig. 5). Furthermore, the contribution of the LeuCAM in neutrophil adhesion to cytokine-activated EC also differed from that of the fMLP-mediated neutrophil adherence. fMLP-mediated neutrophil adhesion was reduced to about 20% below baseline levels by anti-CD18 mAb (p < 0.01). With respect to the contribution of the members of the LeuCAM complex, fMLP-mediated neutrophil adhesion was predominantly inhibited by the mAb against the a subunit of Mac-1 (47 k 9%; p
