Effect of divalent cations on adhesion of polymorphonuclear leukocytes to matrix molecules in vitro Evy Lundgren-Akerlund, La Jolla

Institute

for

Experimental

Elaine Medicine,

Berger,

and

La Jolla,

Abstract: Adhesion of N-formyl-methionyl-leucylphenylalanine-stimulated human polymorphonuclear leukocytes (PMNs) to dishes coated with laminin, fibronectin, or collagen types I and IV was dependent on the presence of magnesium (Mg) but not calcium (Ca2). Addition of manganese (Mn ) in combination with Ca2 and Mg further increased the number of PMNs adhering to the matrix proteins. Monoclonal antibody 60.3 (mAb 60.3) was equally effective at inhibiting adhesion of PMNs to all the matrix proteins. The presence of Mn2 (50 tM), in addition to 1 mM Ca2 and Mg required higher concentrations of mAb 60.3 to inhibit adhesion of PMNs to collagens type I or IV, suggesting increased affinity of PMNs for these substrates. These findings suggest that the PMNs may regulate the affinity of CD11/CD18 lor multiple ligands by binding different divalent cations to the receptor. J. Leukoc. Biol. 51:

,

603-608;

Key 60. 3

1992.

Words: .

Mn2”

neutrophil

#{149} CDJJ/CDJ8

#{149} monoclonal

antibody

collagen

Karl-E

that is common for all the different ligands or if other mechanisms are involved. The present investigation was undertaken to increase the understanding of the molecular mechanisms involved in PMN adhesion in vitro. Adhesion of PMNs to various matrix proteins was studied, and requirements for different divalent cations were investigated.

MATERIALS

During the acute inflammatory response polymorphonuclean leukocytes (PMNs) adhere to the endothelial cells ofthe postcapillary venules and migrate through the extnacellular matrix toward the focus ofinflammation [14]. The molecular mechanisms supporting adhesion and migration are not yet elucidated, although several proteins that participate in these interactions have been identified [23, 26, 28]. In vitro adhesion of stimulated PMNs has been studied on cultured endothelial cells [19, 20, 27, 32, 35, 38] and protein-coated plastic dishes [4, 19, 21, 37]. This adhesion is mediated primarily by the heterodimeric leukocyte adhesion receptors CD11/CD18 (LFA-1, Mac-i, gpiSO/95) [1, 22, 24, 27, 35, 38]. These neceptons are members of the integnin superfamily of matrix receptors [16-18, 25] and have the same 13 subunit (CD18) but different a subunits (CDiia, CD11b, CDlic). Binding of integnins to their ligands has been shown to require the presence ofdivalent cations such as Ca2 and Mg2 [26]. Furthermore, the presence of Mn2 in addition to Ca2 and Mg2 has been shown to increase the affinity of integnins for their ligands [9]. Leukocyte adhesion also appears to depend on the presence ofdivalent cations [4, 11, 16, 33], and the amino acid sequence of CD11b from cDNA cloning predicts three divalent cation binding sites on this molecule [2, 6]. Monoclonal antibodies such as mAb 60.3 [3, 35] and IB4 [36], recognizing epitopes on the /3 chain (CD18), effectively inhibit the adhesion of activated PMNs to cultured endothelial matrix if these

cells [19, proteins antibodies

20] and to immobilized [4, 21] and albumin [11, block a specific binding

proteins including 19]. It is not known site on CD11/CD18

AND METHODS

Materials Fibronectin (FN), collagen type I (CI), bovine serum albu(BSA), N-formyl-L-methionyl-L-leucyl-L-phenylalanine (fMLP), and phorbol myristate acetate (PMA) were obtamed from Sigma Chemical Co., St. Louis, MO; laminin (LM) from Bethesda Research Laboratories, Bethesda, MD; and collagen type IV (CIV) from Collaborative Research, Bedford, MA. Monoclonal antibody 60.3 was prepared as described earlier [3].

mm

PMN

INTRODUCTION

Arfors

Calfomia

Isolation

Human PMNs were isolated from freshly drawn hepaninized (10 U/mi) venous blood collected from healthy adult donors. Erythnocyte sedimentation was accelerated using Macrodex (Pharmacia, Piscataway, NJ) followed by Percoll (Pharmacia) gradient separation of leukocyte-nich plasma as descnibed [15]. The PMN suspension was washed twice in isotonic saline and brought to a final concentration of 5 x 106/ml by adding adhesion buffer [Dulbecco’s phosphate-buffered saline (PBS) with 0.1% BSA].

Endothelial

Cell Culture

Endothelial cells were harvested and cultured from human umbilical veins as described previously [31, 34]. Briefly, endothelial cells were isolated by collagenase digestion and suspended in culture medium 199 with Earle’s salt (MA. Bioproducts, Walkersville, MD) supplemented with HEPES buffer (25 mM; Gibco, Grand Island, NY), fetal bovine serum (17%; M.A. Bioproducts), endothelial growth factor (25 jg/ml; Sigma), heparmn (20 U/mI), and antibiotics. The

Abbreviations:

I and

IV;

BSA,

FACS,

bovine

serum

fluorescence-activated

albumin;

CI

cell

and

sorter;

CIV,

collagen

Flit,

types

fluorescein

isothiocyanate; IMLP, N-formyl-L-methionyl-L-leucyl-L-phenylalanine; FN, fibronectin; HUVEC, human umbilical vein endothelial cell; LM, laminin; PBS, phosphate-buffered saline; PMA, phorbol myristate acetate; PMN, polyrnorphonuclear leukocyte. Reprint requests: Dr. Evy LundgrenAkerlund, Department of Medical and Physiological Chemistry, Lund University, P.O. Box 94, 5-221 00 Lund, Sweden. Received June 18, 1991; accepted January 2, 1992.

Journal

of Leukocyte

Biology

Volume

51, June

1992

603

harvested cells were seeded onto 75-mm2 culture flasks (Corning, Corning, NY) coated with 2% gelatin (Sigma) in PBS. Once confluent, the cells were mechanically scraped off, diluted three times with culture medium, and plated onto tissue culture treated 48-well plastic dishes (Costar, Cambridge, MA) coated with gelatin. Confluent monolayers were formed 5-7 days after plating.

PMN Adhesion Tissue culture treated 48-well plastic dishes (Costar) were coated overnight at room temperature with 20 g/ml of either LM, FN, CI, CIV, on BSA diluted in PBS. Before the adhesion assay the dishes were washed three times with PBS. Endothelial cell cultures were washed twice with PBS prior to the adhesion experiment. The PMNs, suspended in adhesion buffer, were added to the wells (2.5 x 10 cells/well) together with the different reagents and allowed to adhere for 30 mm at 37#{176}C.After this incubation period the wells were gently washed twice with PBS to remove nonadherent cells and adherence was quantitated as myeloperoxidase extracted from the adherent PMNs.

Myeloperoxidase

Assay

Mycloperoxidase was determined as described by Suzuki et al. [30]. Briefly, PMNs were solubilized by addition ofO.5 ml of 0.5% hexadecyltnimethylammonium bromide in 50 mM potassium phosphate buffer (pH 6.0) to each well. Samples of the extracts were added to 96-well microtiter plates (Dynatech) containing 80 mM potassium phosphate buffer (pH 5.4), 0.5 mM hydrogen peroxide, 0.16 mM tetramethylbenzidine, and 8% N,N-dimethylformamide. The samples were incubated for 15-20 mm at 37#{176}Con an orbital shaker and the reaction was stopped by addition of 0.2 M sodium acetate (pH 3.0). The absorbance of the developed color was measured at 650 nm in a microtiten plate reader (Flow Laboratonies, McLean, VA). A known number of PMNs in suspension were serially diluted, extracted, and analyzed for mycloperoxidase content on the micnotiter plates and used for calculation of a standard curve. Adherence is expressed as a percentage of the total number of cells present in the adherence assay.

Statistics Results are presented as means ben of individual experiments. was used to determine statistical

± SEM and The Wilcoxon significance.

n = rank

the numsum test

RESULTS Adhesion of PMNs stimulated with fMLP (106 M) to the immobilized matrix proteins LM, FN, CI, and CIV in the presence of 1 mM CaC12 and MgCl2 is shown in Table 1. fMLP-stimulated cells also adhered to dishes coated with BSA or to uncoated plastic dishes as shown in Table 1. Nonstimulated cells did not adhere to any of the protein-coated dishes but adhered to uncoated plastic dishes as well as did fMLP-activated PMNs (data not shown). Furthermore, adhesion to the different substrates was studied in the absence of Ca2 or Mg2 or when either of the divalent cations alone was added. In the absence of divalent cations, there was only slight adhesion of PMNs to the various matrix proteins (Fig. 1). When Ca2 (1 mM) was added to the adhesion buffer as the only divalent cation, a minor increase in adhesion to the matrix proteins was seen compared to adhesion in the ab-

604

Journal

of Leukocyte

Biology

Volume

51, June

1992

TABLE

1.

Adhesion

of fM LP-stimulated unc oated plastic

Substrate

and

strates and allowed a percentage of the numbers represent

sence added

25.0 39.7 37.5 27.4 50.2 82.2

I IV albumin

fMLP

to protein-coate

(%)

Adhesion

Laminin Fibronectin Collagen type Collagen type Bovine serum Plastic apMNs

PMNs dishesa

(106

M)

were

added

d and

4.0 3.8 2.6 6.2 1 .4 4.5

simultaneously

to the various

to adhere for 30 mm at 37#{176}C.Adhesion is expressed total number of cells added in the adherence assay. the average of n individual experiments ± SEM.

of divalent cations. as the only divalent

However, cation,

n

SEM

when Mg2 the adhesion

(1 mM) increased

6 6 6 6 4 3 subas The

was to

levels seen in experiments in which both Ca2 and Mg2 were present (Fig. 1). Adhesion in the presence of both Ca2 and Mg2 is expressed as 100% in Fig. 1. We also studied adhesion to the different matrix proteins by PMNs stimulated with PMA (108 M) instead of fMLP (data not shown). Although more cells adhered to the immobilized substrates when activated with PMA, the results showed that the adhesion to the different matrix proteins was dependent on the presence of Mg2 but not Ca2. Furthermore, we studied the effect of Ca2 and Mg2 on the adhesion of stimulated PMNs to confluent cultures of human umbilical vein endothelial cells (HUVECs). In these experiments the PMNs were stimulated with PMA (108 M). It was found that PMN adhesion to HUVECs was also dependent on the presence of Mg2. When Ca2 was added to the adhesion buffer as the only divalent cation, 4.5 ± 0.8% of the PMNs adhered to the HUVECs. In contrast, 24.1 ± 0.2% of the cells adhered when Mg2 was added as the only divalent cation. We also found that adhesion to HUVECs, in the presence of both Ca2 and Mg2, was slightly lower (13.7 ± 0.9%) than adhesion in the presence of Mg2 only. The numbers represent averages of five individual experiments ± SEM. The nequirement for Mg2 in adhesion of PMNs to HUVECs and to immobilized matrix proteins might indicate that similar mechanisms are involved in supporting adhesion of PMNs to these substrates. In contrast to these experiments, it was found that BSA-coated dishes and uncoated plastic dishes were able to mediate adhesion of PMNs both in the absence of divalent cations and divalent cation, although strates when both Ca2

when Ca2 was added more cells adhered and Mg2 were present

as

to

the only these sub(Fig. 1).

Figure 2 shows the results of experiments on the effect of 50 M Mn2 on adhesion offMLP-stimulated PMNs to LM, FN, CI, and CIV in addition to Ca2 and Mg2. Because others have shown that Mn2 can increase the affinity between some matrix receptors (integnins) and their ligands ( see Introduction) we tested whether Mn2 had any effect on PMN adhesion to the various matrix proteins. When 50 M Mn2 was added to the adhesion buffer, the attachment increased on all substrates. This dose of Mn2 had no detectable effect on superoxide production by the PMNs (data not shown), suggesting that it does not induce damage to the PMNs. To study the affinity of PMNs for the different substrates, mAb 60.3, which recognizes the /3 chain of the adhesion complex CD11/CD18, was used to inhibit the adhesion. The monoclonal was added at different concentrations together with PMNs and fMLP. It was found that mAb 60.3 inhibited PMN adhesion in a dose-dependent manner and that the

0

120

C 0 0

100

0

80 C 0 C,)

60

0

40

0 (8

z

20

ci-

0

LM

FN

CI

CIV

BSA

plastic

Fig.

1. Effect of 1 mM Ca2 and Mg2 on adhesion of fML}tstimulated (106 M) PMNs to plastic dishes coated with laminin (LM), fibronectin (FN), collagen type I (CI), collagen type IV (CIV), or bovine serum albumin (BSA) and to uncoated dishes (plastic). The PMNs were allowed to adhere for 30 mm at 37#{176}Cin the absence of divalent cations (0) or in the presence of either 1 mM Ca2 () or 1 mM Mg2 (0). Adherence of PMNs in the presence of both divalent cations is set to 100% and is represented by the dashed line. The numbers represent the average of at least three individual experiments ± SEM.

efficiency

was

similar

on the

different

substrates

(Fig.

3a and

b). At an mAb 60.3 concentration of 15 tg/ml the inhibition of adhesion was almost maximal on all the substrates. In parallel experiments, 50 cM Mn2 was added together with PMNs, mAb 60.3, and #{163}MLPin the Ca2, Mg-containing buffer.

It was for the

PMNs

found that immobilized

Mn2 could substrates

alter because

the

only at 5 g/ml of mAb 60.3. Mn2 had no significant effect on the other two substrates, LM or FN, at any concentration of mAb 60.3. The possibility that the altered affinity of PMNs for collagen was a result of increased expression of CD11/CD18 on the PMN surface was tested by fluorescence activated cell sorter (FACS) analyses. It was found that 50 M Mn did not increase the expression of the receptor CD11/CD18 as studied by binding of fluorescein isothiocyanate (FITh)-labeled mAb 60.3 to fMLP-stimulated PMNs (data not shown). Taken together, these results showed that CD11/CD18 is the major adhesion receptor on fMLP-stimulated PMNs and that binding of the receptor to the immobilized matrix proteins LM, FN, CI, and CIV requires Mg2 but not Ca2. In contrast, adhesion of PMNs to immobilized BSA and to uncoated plastic dishes occurs in the absence of divalent cations. Furthermore, Mn2 can synergize with Mg2 to increase the affinity of CD11/CD18 for immobilized collagen, especially for CI.

DISCUSSION The present study of the effect of divalent cations on PMN adhesion found that adhesion to matrix proteins immobilized on plastic dishes required the presence of Mg. We found that the absence of any divalent cations or the presence of Ca2 alone supported only minimal PMN adhesion to matrix proteins. These results are somewhat contradictory to those previously described by Bohnsack et al. [4], who reported

affinity of higher doses

ofmAb 60.3 were required to inhibit attachment (Fig. 4a-d). However, only the affinity for the collagen substrates increased to a significant degree and especially the affinity of PMNs for CI appeared to be changed. As shown in Fig. 4c, mAb 60.3 inhibition of PMN attachment to CI was significantly lower when Mn2 was present, indicating an increased affinity of PMNs for this substrate. This was seen at concentrations of mAb 60.3 up to 15 g/ml. At higher concentrations of mAb 60.3 (40 tg/ml), Mn2 had no significant effect on the PMN affinity for CI. Addition of Mn2 also appeaned to increase the affinity of PMNs for CIV; however, the change in inhibition was highly significant (P < .01)

40

30

20

0

a a

10

-C

z 0

ci0

b

a .0

40

E

80

0

C

30

0 -C 0

60

Co

20

z ci-

40

10

0 0

:::T

20

.0

E

,

.

.

0

C

0

10

20

30

40

0

LM

Fig.

2.

plastic

(CI),

Effect dishes

or

of Mn2 coated

collagen

type

FN on

adhesion

with

laminin

IV

Cl

CIV

of fMLP-stimulated (LM),

(CIV).

The

(106

fibronectin PMNs

were

mAb

M)

PMNs

(FN),

collagen

allowed

to

adhere

Fig. 3. Adhesion

to

type

I for

ious

collagen adhere

numbers

the

the

average

of

six

individual

experiments

±

SEM.

Lundgren-Akerlund

of mAb

simultaneously

30 mm at 37#{176}C in the presence of 1 mM Ca2 and Mg2 () and in the presence of 50 iM Mn2 in addition to Ca2 and Mg2 (0). Adherence is cxpressed as a percentage of the total of cells added in the adhesion assay. The represent

ofIMLP-stimulated

concentrations

total

to

plastic

(106

60.3. dishes

coated

number

et al.

of of

cells six

Divalent

added

individual

cations

in

the

PMNs

in the

antibody, with

adhesion

and

laminin

IV (A) and is expressed

experiments

and

(pg/mI)

M)

PMNs,

type I (L:), or collagen type for 30 mm at 37#{176}C.Adherence

average

60.3

(0),

the

assay.

presence

fMLP

of var-

were

added

(#{149}),

fibronectin

cells were allowed to as a percentage of the

The

numbers

represent

± SEM.

PMN

adhesion

in vitro

605

100 80

60 40

20 0

C 0 .0 -C C

100 80 60

40 20

0

10

20

30

40

0

mAb

.,

Fig. 4. Inhibition PMNs, antibody,

of fMLIstimulated

fMLP, + Mn2); (c) collagen at 37#{176}C. Inhibition was

of antibody. determined experiments

Statistical

significance

with the Wilcoxon ±

(106

and Mn’ type I (L calculated

M)

adhesion

by

various

added simultaneously to plastic , - Mn2; A, + Mn2), or (d) collagen for each concentration of mAb 60.3 of differences

rank

PMN

were

sum

in inhibition,

test for each

ofmAb

lated PMN adhesion to BSA-coated or uncoated plastic dishes was seen both in the absence of divalent cations and in the presence of either Ca2 or Mg2, different mechanisms are probably involved in adhesion to these substrates as compared to matrix proteins or endothelial cells. It was reported by Dransfield and Hogg [7] that Mg can regulate the binding activity of a specific CD11/CD18 epitope present on all Journal

60.3.

of

with , -

in the absence or in the presence of Mn2 (50 sM). (0, - Mn2; + Mn); (b) fibronectin ((), - Mn2; V , + Mn2) and the cells were allowed to adhere for 30 mm in the presence of antibody with adherence in the absence

Mn2; the

.01;

40

#{149},

mAb 60.3,

(a)laminin

adherence in

P