chain (CD18) with a unique a chain (CD11a, b, and. In contrast, monocytes stimulated with exogenous IFN- c) to form receptors first shown to mediate cellâcell.
EXPERIMENTAL CELL RESEARCH
222, 209–217 (1996)
Article No. 0026
TGF-b1-Stimulated Adhesion of Human Mononuclear Phagocytes to Fibronectin and Laminin Is Abolished by IFN-g: Dependence on a5b1 and b2 Integrins BRIGITTE BAUVOIS,*,†,1 JOHAN VAN WEYENBERGH,*,† DANY ROUILLARD,†
AND
JUANA WIETZERBIN*,†
*Unite´ 365 INSERM, †Institut Curie-Section de Recherches, Paris, France
Monocyte migration within the extravascular space of inflamed tissues is controlled by adhesion molecules and inflammatory cytokines. In this study, we analyzed the capacity of TGF-b1 and IFN-g to regulate adhesion of human activated monocytes to fibronectin (FN) and to laminin (LM), two components of the extracellular matrix. When cultured in the absence of any of these two stimuli, human monocytes underwent ‘‘spontaneous activation’’ and adhered to both FN and LM. Adhesion to FN was inhibited in the presence of a5 and b1 integrin blocking antibodies, whereas b2 blocking antibody blocked attachment to LM. Exogenous TGF-b1 increased the adhesive ability of monocytes to FN and to LM, respectively, linked to the increase of a5 and b2 mRNA and protein synthesis levels. Moreover, an increase in a5 expression at the monocyte cell surface was observed. In contrast, monocytes stimulated with exogenous IFNg lost their capacity to bind to FN and this coincided with the down-regulation of surface a5 expression which occurred at the posttranscriptional level of a5 synthesis. Although IFN-g-treated monocytes also showed a decreased ability to adhere to LM, no alteration of b2 mRNA levels, b2 protein synthesis, and b2 cell surface expression was detectable, thus suggesting a modification of the functional state of surface b2 integrins. Furthermore, when stimulated with TGF-b1, IFN-g-pretreated monocytes reacquired the ability to bind to FN and LM. Conversely, IFN-g reduced adhesiveness to FN and LM of monocytes initially stimulated with TGF-b1. These in vitro adhesive–deadhesive responses of monocytes to TGF-b1 and IFN-g modulation may reflect mononuclear phagocyte motility within sites of inflammation. q 1996 Academic Press, Inc.
INTRODUCTION
During their migration from the capillary bed to and within the extravascular tissues, blood monocytes need 1 To whom correspondence and reprint requests should be addressed at Unite´ 365 INSERM, Institut Curie, 26, rue d’Ulm, 75231 Paris cedex 05, France. Fax: 44-07-07-85.
to transiently adhere to different cellular and extracellular matrix proteins including proteoglycans, collagens, and glycoproteins. Thus, signals resulting from the interactions between the tissue adhesive molecules and the adhesion counterreceptors present on their surface presumably play a role in the further immune and nonimmune functions displayed by tissue mononuclear phagocytes [1]. The major adhesion receptors which are involved in cell–cell and cell–extracellular matrix interactions belong to the b1, b2, and b3 integrin subfamilies [2]. The b1 subfamily, distributed on many different cell lineages and including seven heterodimers composed by unique a1– a6 and av peptides associated to the b1 peptide, mediates the interactions with collagens, laminin, and fibronectin [2]. The b2 subfamily, found only on leukocytes, associates the b2 chain (CD18) with a unique a chain (CD11a, b, and c) to form receptors first shown to mediate cell–cell interactions [2]. Recent studies have established that cell adhesion to extracellular matrix can also be mediated by the CD11b/CD18 and CD11c/CD18 integrins [2]. The b3 subfamily consists of the platelet glycoprotein aIIbb3 and the vitronectin receptor avb3, and both members are able to bind several matrix proteins [2]. Although they express most anb1 members and all b2 members of integrin subfamilies, human blood monocytes do not attach efficiently to the extracellular matrix (ECM)2 proteins, suggesting that the affinity of the integrins for their ligands is highly regulated [3]. However, exposure of cells to inflammatory cytokines results in increased monocyte adhesion to ECM. Upon incubation with transforming growth factor-b1 (TGFb1), monocyte a5 and b1 integrins are up-regulated and this regulation is a prerequisite for the further adhesion of monocytes to fibronectin [4]. Treatment of human monocytes with granulocyte–macrophage col2 Abbreviations used: BSA, bovine serum albumin; ECM, extracellular matrix; FACS, flow-activated cell sorter; FCS, fetal calf serum; FN, fibronectin; IFN-g, interferon-g; LM, laminin; mAb, monoclonal antibody; PMA, phorbol myristate acetate; SSPE, phosphate-buffered saline containing 1 mM EDTA; TGF-b1, transforming growth factor-b1.
209
/ m4861$3037
12-21-95 20:20:01
eca
0014-4827/96 $12.00 Copyright q 1996 by Academic Press, Inc. All rights of reproduction in any form reserved.
AP: Exp Cell
210
BAUVOIS ET AL.
ony-stimulating factor (GM-CSF) induces avb3 integrin surface expression and a migratory phenotype on vitronectin [5]. In contrast, macrophage colony-stimulating factor (M-CSF) increased monocyte adhesion on vitronectin by turning on the surface expression of avb5 [5]. Interferon-g (IFN-g), a NK- and T-cell-derived cytokine, has been shown to exert pleiotropic effects on several cell types including monocytes [6]. Treatment of monocytes or their precursors with IFN-g induces maturation into macrophages [7]. Moreover, it is an effective activator of monocyte/macrophage functions including tumoricidal/antimicrobial activity and oxidative metabolism [6]. IFN-g modulates surface expression of a variety of molecules including FcgRI, CRI (CD35), CD14, and HLA class II molecules on monocytes [6]. TGF-b is a multifunctional cytokine produced by various types of cells including platelets, activated T-cells, B cells, monocytes, and macrophages [8]. It can act as a monocyte–macrophage deactivating factor as assessed by decreased hydrogen peroxide production, TNF-a, and IL-6 production and can induce macrophage chemotaxis [8]. The aim of the present study was to examine whether IFN-g and TGF-b1, which can be locally released in an inflammatory site, can influence human monocyte– extracellular matrix interactions, thus determining monocyte motility within the extracellular compartment. Our results indicated that IFN-g and TGF-b1 differentially regulate the a5 and b2 components of integrins involved in their interaction with fibronectin and laminin, respectively. The combined action of IFNg and TGF-b1 on integrin receptors led us to propose a cytokine-mediated mechanism which could be involved in the motility of monocytes within inflamed tissues. MATERIALS AND METHODS Cytokines, mAbs, and other reagents. TGF-b1 purified from human platelets was obtained from British Bio Technology (Oxon, UK). Human recombinant IFN-g (2 1 107 U/mg) was kindly provided by Roussel-Uclaf (Romainville, France). The irrelevant mouse immunoglobulins IgG1, IgG2a, IgM and monoclonal antibodies specific for b1 (4B4, mIgG1) and ap150/95 (CD11c) (MO1, mIgM) were obtained from Coulter Immunology (Hialeah, FL). Goat F(ab*)2 fragment antimouse and anti-rat fluorescein isothiocyanate (FITC)-conjugated Ig were purchased from Cappel Laboratories (Malvern, PA). The irrelevant rat isotype IgG2a and the mouse monoclonal antibodies specific for a2 (Gi9, mIgG1), a3 (M-KID 2, mIgG1), a4 (HP2/1, mIgG1), a5 (SAM1, mIgG2a), a6 (GoH3, rIgG2a), av (AMF7, mIgG1), aL(CD11a) (25.3.1, mIgG1), and aM(CD11b) (BEAR 1, mIgG1) were purchased from Immunotech (Luminy, France). Mouse monoclonal antibody anti-b2(CD18) (MHM23, mIgG1) was obtained from Dako Ltd. (Glostrup, Denmark). Monoclonal antibodies 4B4, Gi9, HP2/1, SAM1, GoH3, AMF7, and MHM23 were assayed for their ability to block cell adhesion to extracellular matrix [9–15]. Laminin purified from Englebreth-Holm Swarm mouse tumor, human plasma fibronectin, bovine serum albumin, and phorbol myris-
/ m4861$3037
12-21-95 20:20:01
eca
tate acetate were obtained from Sigma Chemical Co. (St. Louis, MO). cDNA for fibronectin receptor a5 subunit was purchased from Gibco BRL (Cergy-Pontoise, France). The synthetic b2(CD18) oligonucleotide (5*-CTGGTCACCGCGAAGATGGGCTGG-3*) was obtained from Eurogentec (Seraing, Belgium). Cell culture. Peripheral blood leukocytes were obtained from heparinized whole venous blood. Monocytes were separated by Ficoll– Hypaque density gradient (1.077 g/ml) centrifugation. Mononuclear cells concentrated at the interface were collected, washed twice in RPMI 1640 plus 5% fetal calf serum (FCS), and allowed to adhere in tissue culture dishes (Nunc, Uppsala, Sweden) for 30 min at 377C. Nonadherent cells were removed by aspiration and two washes with phosphate-buffered saline (PBS). The adherent cell population was removed from the plastic by gentle scraping and represented ú90% monocytes as determined by immunofluorescence flow cytometry analysis of the CD14 antigen. Adherent monocytes (1–5 1 106/ml) were cultured in RPMI 1640 medium, 10% FCS, 2 mM glutamine, and 40 mg/ml gentamycin at 377C in Nunc flasks, in the absence or in the presence of 100 U/ml IFN-g or 1 ng/ml TGF-b1 (initially dissolved in 5 mM HCl at a concentration of 10 mg/ml). We previously showed that these cytokine concentrations were optimal for inducing monocytic cell activation [12]. After various periods of incubation, cells were collected by repeated medium aspirations followed by scraping if necessary, washed twice with PBS, counted with a Coulter Counter ZM equipped with a Coultronic 256 channelizer, and tested for cell viability (determined by trypan blue exclusion), morphology (May-Gru¨nwald Giemsa and subsequent light microscope examination), flow cytometry analysis, and adhesion assays. Flow cytometry analysis. All staining steps were done in PBS, supplemented with 1% BSA and 0.1% sodium azide and were carried out in 96-microwell plates (Nunc) on ice. Cells to be stained (2 1 105) were centrifuged to form a pellet and incubated with the appropriate antibodies under saturating conditions. After 30 min of incubation, the cells were washed twice and subsequently incubated for 30 min with FITC-conjugated goat F(ab*)2 anti-mouse IgG or with FITCconjugated goat F(ab*)2 anti-rat IgG (10 ml of 1:100 dilution for 2 1 105 cells), washed twice, and fixed with 0.4% formaldehyde in PBS. For cell permeabilization, cells were fixed in 4% paraformaldehyde in PBS at room temperature. After washing with PBS, the cells were incubated in PBS containing 0.05% Triton X-100 for 10 min at room temperature. Cells were washed and further stained as outlined above. Cell fluorescent intensity was measured using a FACScan fluorescence-activated cell analyzer (Becton–Dickinson, Mountain View, CA). For each sample, data from 5000–10,000 cells were collected. Fluorescence data are expressed in relative fluorescence intensity (%). Specific fluorescence density (SF) was obtained by substracting the peak channel number of the negative control from the peak channel number of the corresponding experimental sample. Cell adhesion assays. Twenty-four-well flat-bottomed microtiter plates (Nunc) were coated at room temperature overnight with 500 ml of fibronectin (FN) or laminin (LM) (20 mg/ml), in PBS, pH 7.4. The plates were washed twice with PBS and nonspecific sites were blocked by the addition of 500 ml of 1% BSA in PBS for 1 h at 377C. Wells were washed twice with PBS. Cells obtained as described above were resuspended in complete RPMI without FCS to yield a final concentration of 106 cells/ml. A total of 500 ml of this suspension (5 1 105) was added to each substrate-coated well. Cells were also added to wells that had been coated with BSA alone. Plates were incubated at 377C for 30 min in a 5% CO2-humidified atmosphere, after which the wells were gently washed twice with 0.5 ml PBS. To remove adherent cells from the wells, 500 ml of 0.05% trypsin–EDTA buffer was added to each well. After 20 min at 377C, the 500 ml of trypsin– EDTA containing the detached cells was removed and the cell number was determined using a cell Coulter Counter channelizer 256. The counting window was 8.2–13 mm, which corresponds to cells of the monocytic type. Each cell adhesion assay was carried out in duplicate. The number of attached cells/total number of cells 1 100
AP: Exp Cell
211
MONOCYTE ADHESION RESPONSES TO TGF-b1 AND IFN-g
TABLE 1 Effects of IFN-g and TGF-b1 on the Expression of Human Monocyte Integrins Percentage of positive cells { SD (SF) Incubated monocytes Control monocytes
b1 a3 a4 a5 a6 b2 CD11a CD11b CD11c av
98 64 81 72 83 99 91 87 91 6
{ { { { { { { { { {
2 9 6 2 3 1 0 4 0 6
(194) (69) (51) (34) (43) (274) (123) (76) (133) (0)
/ IFN-g
0 96 77 80 54 68 97 82 70 90 64
{ { { { { { { { { {
1 7 3 19 2 1 9 13 04 16
(167) (42) (32) (43) (20) (432) (67) (65) (101) (41)
96 62 92 13 68 96 75 78 91 69
{ { { { { { { { { {
01 9 4 10 7 01 17 15 3 21
(228) (47) (57) (3) (25) (438) (87) (87) (103) (53)
/ TGF-b1 94 65 88 87 75 96 76 77 93 74
{ { { { { { { { { {
4 6 0 2 2 2 2 8 2 7
(231) (36) (55) (60) (40) (381) (87) (80) (97) (50)
Note. Fresh monocytes (control) or monocytes incubated in the absence (0) or in the presence of 100 U/ml IFN-g or 1 ng/ml TGF-b1 for 1 day were assayed for expression of ab1, ab2, and av integrins by immunofluorescence in a Becton flow cytometer, as described under Materials and Methods. Percent values represent the means of two to five experiments { SD. Specific fluorescence SF is shown in parentheses.
was determined as the percentage of cell adhesion. Specific adhesion to FN and LM was determined by substracting the nonspecific attachment of cells to BSA-coated surfaces from the total cell attachment to FN- or LM-coated surfaces. In other experiments, adherent cells were fixed in PBS containing 1% formaldehyde, stained with 10% Giemsa, and photographed under high magnification with a Wild MPS 46/52 inverted microscope (Heerbrugg, Switzerland). For the inhibition of attachment to FN or LM by antibodies, 5 1 105 cells were preincubated with 10 mg of the antibody preparation in a total volume of 100 ml RPMI for 30 min at 47C. The final volume was adjusted to 500 ml and the cells were then placed in substrate-coated wells and adhesion was assessed as described above. Northern analysis. Total RNA was prepared from cell cultures by acid guanidium thiocyanate–phenol–chloroform extraction [16]. RNA was fractionated (20 mg/lane) in a formaldehyde-containing 1% agarose gel, then transferred to Biodyne nylon membrane, and prehybridized. The cDNA probes for human a5 and actin were radiolabeled with [a-32P]dCTP (Du Pont NEN) using a random primed DNA labeling system (Amersham) and hybridized at 427C in a hybridization buffer containing 50% formamide, 61 SSPE, 51 Denhardt’s, 0.1% SDS, and washed to a final stringency of 0.1% SDS, 0.11 SSPE at 607C. End-labeling of the b2 oligonucleotide was performed using T4 polynucleotide kinase (Gibco, BRL) and [g-32P]ATP (NEN). Hybridization and subsequent washings were carried out in the same hybridization buffer without formamide at 107C below Tm (707C), as calculated according to Itakura et al. [17]. Signal intensity of mRNA bands was quantified using a phosphorimager system (Molecular Dynamics, Sunnyvale, CA) and normalized for actin transcripts. a5 mRNA enrichment factor was determined as the ratio of the a5 mRNA value of stimulated cells to the a5 mRNA value of unstimulated cells.
RESULTS
In agreement with previous reports [18, 19], freshly isolated monocytes were found to express high levels of (a3– a6) b1 integrins and all b2 integrins as determined by FACS analysis (Table 1). No detectable level
/ m4861$3037
12-21-95 20:20:01
eca
of av subunit was observed (Table 1). In order to determine whether IFN-g (100 U/ml) or TGF-b1 (1 ng/ml) affects integrin expression, integrin levels on cytokine treated-monocytes were initially determined after 24 or 48 h stimulation and compared to that of untreated monocytes which are known to undergo a spontaneous and time-dependent maturation [20]. Expression of most of the integrins tested, i.e., a3, a4, a6, b2, CD11a, CD11b, and CD11c, was not significantly altered in spontaneously maturing monocytes or in IFN-g- or TGF-b-treated ones (Table 1). However, in all cases, a consistent increase in av expression was observed (Table 1). Most interestingly, the amounts of a5 were strongly depressed in a monocyte population cultured in the presence of IFN-g (Table 1). In contrast, a significant increase in the amount of a5 on TGF-b1treated monocytes was observed (Table 1). We therefore determined the time course of surface a5 changes on untreated and cytokine-treated monocytes (Fig. 1). Stimulation by IFN-g rapidly down-regulated the amount of surface a5 (decrease in percent positivity values), whereas TGF-b1 induced a marked enhancement in the average number of a5 subunits per cell (increase in a5 density SF). The modulating effects of IFN-g and TGF-b1 on a5 and av molecules of cell adhesion receptors for fibronectin and laminin/vitronectin, respectively, prompted us to determine whether both cytokines regulate monocyte adhesion to extracellular matrix proteins. Cells were allowed to adhere to fibronectin (FN)- and to laminin (LM)-coated substrata for 30 min at 377C. Monocytes freshly isolated from blood demonstrated low adherence to FN (Fig. 2A) and LM (Fig. 2B). Monocytes precultured for 1 day in the absence of any stimulus
AP: Exp Cell
212
BAUVOIS ET AL.
FIG. 1. Time-course study of cell surface expression of a5 in spontaneously or cytokine-treated monocytes. Fresh monocytes were cultured in the absence (spontaneous activation) or in the presence of 100 U/ml IFN-g or 1 ng/ml TGF-b1 at 377C. At the indicated time points, cells were assayed by FACS for expression of a5 chain as described under Materials and Methods. The open curves represent background fluorescence obtained with the irrelevant antibody and the filled curves specific fluorescence obtained with anti-a5 mAb.
showed a slight increase in their attachment to FN (Fig. 2A). Cells cultured in the presence of IFN-g showed a marked decrease in their attachment to FN (Fig. 2A) and LM (Fig. 2B). In contrast, treatment of monocytes with TGF-b1 for 24 h resulted in a significant increase in cell binding to FN (Fig. 2A) and LM (Fig. 2B) which was sustained for at least 48 h (see Fig. 4). In parallel, visual examination of adherent monocytes showed different morphological changes associated with FN and LM (Fig. 3). Adherent fresh monocytes retained a round and spherical morphology on FN and LM (Figs. 3a and 3b), whereas monocytes spontaneously maturing after 1 day culture exhibited strong spreading on FN (Fig. 3c), which was not observed on LM (Fig. 3d). Both FN and LM supported pronounced spreading of monocytes incubated with TGF-b1 for 1 day (Figs. 3e and 3f). In contrast, monocytes stimulated with IFN-g adopted a round/stellate morphology on FN (Fig. 3g) and LM (Fig. 3h). To investigate the nature of adhesion molecules that mediate monocyte binding to FN and LM, the effects of function blocking monoclonal antibodies (mAbs) directed to b1, a4, a5, a6, av, and b2 integrins were
/ m4861$3037
12-21-95 20:20:01
eca
examined (Table 2). The attachment of spontaneously activated monocytes and TGF-b1-activated monocytes to FN was not affected by preincubation of these cells with mAbs against a4 and b2 subunits, but was almost completely inhibited by a5 and b1 subunit antibodies, while control isotype antibodies had no blocking effect (Table 2). This result indicated the involvement of the a5b1 receptor in the adhesion of monocytes to FN all along the activation pathway of monocytes. MAbs against b1, a6, and av subunits did not block activated monocyte adhesion to LM (Table 2). Moreover, no inhibition of cell adherence to LM was noted in the presence of anti-CD44 or anti-b3 mAbs which recognize the cell surface laminin receptors CD44 and avb3 present on activated monocytes (data not shown). In contrast, adhesion of monocytes spontaneously activated or incubated in the presence of TGF-b1 to LM was almost totally inhibited by anti-b2 mAb (Table 2). To determine whether the regulation of a5 and b2 integrins occurred at transcriptional or posttranscriptional levels, Northern blot analysis was performed. Total RNAs from fresh monocytes or monocytes cultured in the absence or in the presence of IFN-g or
AP: Exp Cell
MONOCYTE ADHESION RESPONSES TO TGF-b1 AND IFN-g
213
synthesis at the posttranscriptional level. On the other hand, no alteration in the amount of b2 protein was detectable in IFN-g-treated monocytes (Table 3). To further study the relative importance of IFN-g and TGF-b1 on monocyte adhesion to FN and LM, we examined the adhesion modulating capacity of one cy-
FIG. 2. Monocyte adhesion to FN and LM in the absence or in the presence of IFN-g or TGF-b1. Freshly isolated monocytes (j) or monocytes spontaneously activated (…) or activated with 100 U/ml IFN-g (h) or 1 ng/ml TGF-b1 (Ω) at 377C for 24 h were added to either FN- or LM-coated wells and the plates were incubated at 377C for 30 min. Attached cells were quantitated as described under Materials and Methods. Each determination was done in duplicate and each value represents the mean of four to six separate experiments. (A) Adhesion to FN, (B) adhesion to LM.
TGF-b1 were hybridized with probes specific for a5 and b2 subunits. a5 mRNA (4.9 kb) was detected in fresh monocytes (data not shown). No change in a5 mRNA expression was noted in spontaneously activated monocytes (Fig. 4a) or in IFN-g-treated cells (Fig. 4b). When the filters were probed for the b2 subunit, no detectable levels of b2 mRNA were observed in fresh monocytes (data not shown) or in monocytes cultured for 1 day in the absence (Fig. 4a) or in the presence of IFN-g for 1 day (Fig. 4b). However, TGF-b1 was found to induce b2 mRNA (3 kb) (Fig. 3c) and to up-regulate a5 mRNA level (4.4-fold increase, Fig. 4c) in Day 1 monocytes. In addition, total amounts of a5 and b2 subunits were determined in permeabilized monocytes cultured with or without IFN-g or TGF-b1 (Table 3). When compared to control untreated monocytes, increased levels of a5 and b2 proteins were observed in cells exposed to TGF-b1 (Table 3), thus indicating that TGF-b1-increased a5 and b2 mRNA levels were reflected by synthesis of a5 and b2 proteins. In contrast, the total amount of a5 was much lower in IFN-g treated cells than in control cells (Table 3), indicating that IFNg produced a selective effect on the regulation of a5
/ m4861$3037
12-21-95 20:20:01
eca
FIG. 3. Monocyte morphology on FN- and LM-coated wells. Fresh monocytes (a, b) or monocytes cultured in the absence (c, d) or in the presence of 1 ng/ml TGF-b1 (e, f) or 100 U/ml IFN-g (g, h) for 1 day were added to wells coated with FN (a, c, e, g) or LM (b, d, f, h). After incubation at 377C for 30 min, adherent cells were fixed, stained with Giemsa, and photographed under phase-contrast conditions. Bar, 10 mm.
AP: Exp Cell
214
BAUVOIS ET AL.
TABLE 3 Amounts of a5 and b2 Subunits in Permeabilized Monocytes Percentage of positive cells (SF)
FIG. 4. mRNA expression of a5 and b2 subunits in human monocytes cultured in the absence or in the presence of IFN-g or TGFb1. Total RNA extracted from monocytes cultured for 1 day in the absence (a) or in the presence of 100 U/ml IFN-g (b) or 1 ng/ml TGF-b1 (c) was prepared as described under Materials and Methods. Northern blots were hybridized with 32P-labeled a5 or b actin cDNAs or with 32P-labeled b2 oligonucleotide. Each set of tracks is from the same exposure of the same gel.
tokine on monocytes pretreated with the second cytokine. Adhesion levels of monocytes to FN and LM being variable between individual normal donors, Fig. 5 shows a representative experiment which was reproduced in three separate independent assays. There was a significant increase in adhesion to FN and to LM of IFN-g-pretreated monocytes subsequently incubated in the presence of TGF-b1 (Fig. 5A). MAbs against a5, b1, or b2 molecules were again the only mAbs capable of affecting cell adhesion to FN or LM, respectively (data not shown). Conversely, IFN-g treatment of TGFb1-prestimulated monocytes reduced adhesion levels on FN and LM (Fig. 5B). As expected, the decrease in
Integrin
Control
/IFN-g
/TGF-b1
a5 b2
51 (43) 66 (302)
7 (18) 66 (302)
68 (64) 88 (214)
Note. Monocytes were cultured for 1 day in the absence (control) or in the presence of 100 U/ml IFN-g or 1 ng/ml TGF-b1. Total levels of a5 and b2 (percentage and SF) in permeabilized cells (cell surface plus intracellular expression) were determined by immunofluorescence analysis as described under Materials and Methods.
cell adhesion to FN was correlated to the down-regulation of a5 subunit from the cell surface (data not shown). Taken together, these data showed that TGFb1 reversed the effect of IFN-g and vice versa. DISCUSSION
It is known that the cell surface expression and synthesis of integrins mediating cell–matrix interactions
TABLE 2 Effect of mAb Anti-integrins on Monocyte Adhesion to FN and LM Inhibition of cell attachment (% { SD) Adhesive protein FN Anti-b1 Anti-a5 Anti-a4 Anti-b2 LM Anti-b1 Anti-a6 Anti-av Anti-b2
/TGF-b1
0
60 66 0 0
{ { { {
20 20 0 0
99 68 0 0
{ { { {
0 04 0 0
0 0 0 79
{ { { {
0 0 0 19
0 0 0 90
{ { { {
0 0 0 10
Note. Adhesion of spontaneously activated monocytes [column (0)] and monocytes incubated with TGF-b1 (24 h) to FN and LM was assayed as described under Materials and Methods in the absence or in the presence of 20 mg/ml of the different mAbs. When used at the same concentration, isotype-matched antibodies (controls) were not inhibitory and adhesion measured in the presence of these controls was set to be 100%. Values represent the means of two experiments { SD.
/ m4861$3037
12-21-95 20:20:01
eca
FIG. 5. Combined effects of IFN-g and TGF-b1 on human monocyte adhesion to FN and LM. (A) Monocytes were cultured in the presence of 100 U/ml IFN-g for 20 h followed by another 20 h culture without or with a pulse of 1 ng/ml TGF-b1. (B) Monocytes were cultured with a pulse of 1 ng/ml TGF-b1 for 20 h followed or not by a second pulse of IFN-g for another 20 h. Adhesion of cells to FN (j) or LM (h) was tested as described under Materials and Methods. Each determination was done in duplicate { SD.
AP: Exp Cell
MONOCYTE ADHESION RESPONSES TO TGF-b1 AND IFN-g
can be modulated by cytokines. IL-1b and TNF-a upregulate a1b1 integrin in fibroblasts and osteosarcoma cells [21]. GM-CSF induces b3 on human monocytes, while M-CSF enhances b5 on the same cells [5]. TGFb1 increases the expression of a1–5, b1 on a large variety of cells including fibroblasts, osteosarcoma cells, and monocytes [4, 22, 23]. Among cytokines able to induce alterations in the expression of receptors for extracellular matrix components, TGF-b1 has been shown to down-regulate a3 in osteosarcoma cells [24], whereas TNF-a in synergy with IFN-g decreases expression of b3 subunit of the vitronectin receptor in human endothelial cells [25]. We have previously shown that IFN-g reduced the levels of a4 and a5 subunits at the surface of the human monoblastic leukemic cell line U937 [12]. Whether IFN-g could induce changes in the integrin profile of human peripheral blood monocytes had not yet been reported to our knowledge. The functional receptor for FN on human resting monocytes is the a5b1 integrin [26]. Whether this receptor may participate in the interaction with FN all along the maturation pathway of monocytes is unknown. In the present report, we show that both spontaneously and cytokine-treated monocytes attach to FN via the a5b1 receptor. Moreover, we demonstrated that adhesion of activated monocytes to LM was dependent on b2 integrins since anti-b2 mAbs inhibited adhesion while mAbs against a6, av, and b1 subunits of LM receptors were ineffective. Consistent with our data, human monocytes stimulated with monocyte chemoattractant protein-1 have recently been shown to attach to LM through the CD11b/b2 receptor [27]. Adherence of stimulated human neutrophils to LM and collagen I involves the b2 receptor [15, 28–31], whereas activated myeloblastic HL-60 cells adhere to LM in a b2 integrindependent manner, possibly in combination with both CD11b and CD11c subunits [24]. We had recent evidence that a mAb to the CD11c subunit (MO1, Coulter Immunology) partially blocked monocyte adhesion to LM (data not shown). Whether monocyte adhesion to LM can be mediated by both CD11b and CD11c integrins remains to be determined. We demonstrate here that TGF-b1 and IFN-g differentially modulate the expression of a5 and b2 subunits which are essential for the adhesive properties of human monocytes. The increased ability of TGF-b1treated monocytes to bind to FN correlates to an increase in the levels of a5 mRNA and a5 protein synthesis as well as a5 subunit at the cell surface. In contrast, the enhanced b2-dependent adhesion of TGF-b1-stimulated monocytes to LM was not correlated with an apparent increase in surface expression of the b2 chain. However, Northern blot analysis provides the first observation of enhanced b2 mRNA and b2 protein synthesis in TGF-b1-treated human monocytes. In previ-
/ m4861$3037
12-21-95 20:20:01
eca
215
ous studies performed on myeloid cell lines, enhanced transcription of b2 gene occurred in differentiating HL60 and U937 cells [32–34], while no modulation in transcription of b2 gene in TGF-b1-stimulated THP-1 cells was observed [35]. Thus, the regulated expression of b2 integrin in human cells of the monocytic lineage appears to be controlled both at the transcriptional and posttranscriptional levels [22, 34]. There are various mechanisms by which integrin function may be regulated in the absence of apparent changes in the level of expression at the cell surface, including the altered glycosylation of the extracellular domain of integrins, phosphorylation of their cytoplasmic domains, or changes in integrin conformation [2, 3]. It has recently been suggested that stimulation of human neutrophils via FcgRII/III evokes a tyrosine kinase-dependent and actin filament-dependent intracellular signal that enhances adherence of neutrophils to fibrinogen- and BSA-coated surfaces through the activation of the CD11b/b2 integrin [36]. Interestingly, TGF-b is able to rapidly induce FcgRIII on blood monocytes in vitro [37] and the tyrosine kinase activity involved in this process is now shown to be associated to the fgr protein kinase (p58fgr) belonging to the src family [38]. The striking effect of IFN-g on monocytes was the dramatic decrease of intracellular and surface expression of the a5 subunit. We found a correlation between such decrease in the relative abundance of the a5b1 receptor at the cell surface and the reduction in monocyte adhesiveness to FN. Commitment to terminal differentiation of human epidermal keratinocytes has been shown to result in a decrease in adhesiveness to FN explained by a loss of the a5b1 receptor from the cell surface due to inhibition of transcription of the a5 and b1 genes [39]. However, in the present study, IFNg down-regulation of a5 integrin did not occur at the mRNA level but at the posttranscriptional level. Similarly, IFN-g has been found to synergize with TNF-a to inhibit the expression of the vitronectin receptor avb3 on human endothelial cells by reducing b3 subunit synthesis at the translational level, leading to a decrease in cell ability to adhere to vitronectin [25]. After IFN-g treatment, the specific loss of monocyte adhesion to LM was, however, not paralleled by any alteration in the amount of b2 (mRNA, protein synthesis, and cell surface occupancy). In line with these data, the effect of IFN-g on b2 may be linked to the modification of the functional state of cell surface b2 (glycosylation, phosphorylation, conformation), as mentioned above. Moreover, the interactions of the intracytoplasmic domain of integrins with components of the cytoskeleton or plasma membrane can modulate their functions [2, 3, 40]. Interestingly, IFN-g is a potent regulator of cytoskeletal proteins [41] and it is thus possible that IFN-g alters the interaction between the cytoskeleton and b2 integrin.
AP: Exp Cell
216
BAUVOIS ET AL.
FIG. 6. Model of monocyte attachment to extracellular matrix FN and LM. Our findings suggest that binding capacities of a5b1 and ab2 integrins are differentially modulated on motile monocytes (IFN-g-activated cells) and on nonmotile monocytes (TGF-b1-activated cells). ECM, extracellular matrix composed of fibronectin and laminin.
Monocyte entry within inflamed tissues and further migration within ECM are likely a multistep process involving attachment to ECM followed by detachment and subsequent migration through ECM. Our present results on the sequential actions of IFN-g and TGF-b1 on monocytes can be combined in a model for the role of extracellular matrix interactions in adhesion and migration of monocytes through inflammatory tissues (Fig. 6). TGF-b1-induced or IFN-g-abrogated adhesion of monocytes to FN and LM and their associated alterations in binding capacity of a5b1 and ab2 integrins (with or without changing their level of expression) may implicate a role in monocyte motility: monocytes may tend to accumulate in a TGF-b1-rich environment, but become motile in an IFN-g-rich environment, providing a rapid deadhesion (Fig. 6). An excess of TGFb1 has been associated with various tissue pathologies including arthritis, glomerulonephritis, and skin/brain scars [8]. Similarly, an increase in the levels of IFNg is observed in several immunological and infectious diseases such as lupus erythematosus, rheumatoid arthritis, multiple sclerosis, and HIV infection [6, 42, 43]. Thus, our model may be relevant in several pathological states in which dysregulations of cytokines such as IFN-g and TGF-b can be presumed to play a role in disordered monocyte trafficking. Supported by grants from the Institut National de la Sante´ et de la Recherche Me´dicale and the Association pour la Recherche sur le Cancer. The authors thank Mrs. A. Birot for her secretarial expertise. Recombinant human IFN-g was kindly provided by Roussel-Uclaf (Romainville, France).
1. Juliano, R. L., and Haskill, S. (1993) J. Cell. Biol. 120, 577– 585. 2. Hynes, R. O. (1992) Cell 69, 11–25.
12-21-95 20:20:01
14. De Vries, J. E., Keizer, G. D., Te Velde, A. A., Voordouw, A., Ruitter, D., Rumke, P., Spits, H., and Figdor, C. G. (1986) Int. J. Cancer 38, 465–473. 15. Monboisse, J. C., Garnotel, R., Randoux, A., Dufer, J., and Borel, J. P. (1990) J. Leukocyte Biol. 50, 373–380. 16. Chomczynski, P., and Sacchi, N. (1987) Anal. Biochem. 162, 156–159. 17. Itakura, K., Rossi, J. J., and Wallace, R. (1984) Annu. Rev. Biochem. 53, 323–356. 18. Hemler, M. E., Huang, C., and Schwarz, L. (1987) J. Biol. Chem. 262, 3300–3308. 19. Pietro, J., Eklund, A., and Patarroyo, M. (1994) Cell. Immunol. 156, 191–211. 20. Johnston, R. B. (1988) N. Engl. J. Med. 318, 747–752. 21. Santala, P., and Heino, J. (1991) J. Biol. Chem., 266, 23505– 23509. 22. Ignotz, R. A., and Massague´, J. (1987) Cell 51, 189–197. 23. Heino, J., Ignotz, R. A., Hemler, M. E., Crouse, C., and Massague, J. (1989) J. Biol. Chem. 264, 380–388. 24. Heino, J., and Massague´, J. (1989) J. Biol. Chem. 264, 21806– 21811. 25. Defilippi, P., Truffa, G., Stefanuto, G., Altruda, F., Silengo, L., and Tarone, G. (1989) J. Biol. Chem. 266, 7638–7645. 26. Brown, D. L., Phillips, D. R., Damsky, C. H., and Charo, I. F. (1989) J. Clin. Invest. 84, 366–370. 27. Jiang, Y., Zhu, J., Luscinskas, F. W., and Graves, D. T. (1994) Am. J. Physiol. 267, 1112–1118. 28. Nathan, C., Srimal, S., Farber, C., Sanchez, E., Kabbash, L., Asch, A., Gailit, J., and Wright, S. D. (1989) J. Cell. Biol. 109, 1341–1349. 29. Bohnsack, J. F., Akiyama, S. K., Damsky, C. H., Knape, W. A., and Zimmerman, G. A. (1990) J. Exp. Med. 171, 1221–1237. 30. Davis, G. E. (1992) Exp. Cell Res. 200, 242–252. 31. Lundgren-Akerlund, E., Olofsson, A. M., Berger, E., and Arfors, K. E. (1993) Scand. J. Immunol. 37, 569–574.
REFERENCES
/ m4861$3037
3. Diamond, M. S., and Springer, T. (1994) Current Biol. 4, 506– 517. 4. Wahl, S. M., Allen, J. B., Weeks, B. S., Wong, H. L., and Klotman, P. E. (1993) Proc. Natl. Acad. Sci. USA 90, 4577–4581. 5. De Nicholo, M. O., and Burns, G. F. (1993) Proc. Natl. Acad. Sci. USA 90, 2517–2521. 6. Farrar, M. A., and Schreiber, R. D. (1993) Annu. Rev. Immunol. 11, 571–611. 7. Harris, P., and Ralph, P. (1985) J. Leukocyte Biol. 37, 407–422. 8. McCartney-Francis, N. L., and Wahl, S. M. (1994) J. Leukocyte Biol. 55, 401–409. 9. Liesveld, J. L., Winslow, J. M., Frediani, K. E., Ryan, D. H., and Abboud, C. N. (1993) Blood 81, 112–121. 10. Staatz, W. D., Rajpara, S. M., Wayner, E. A., Carter, W. G., and Santoro, S. A. (1989) J. Cell Biol. 108, 1917–1925. 11. Gismondi, A., Morrone, S., Humphries, M. J., Piccoli, M., Frati, L., and Santoni, A. (1991) J. Immunol. 146, 384–392. 12. Bauvois, B., Rouillard, D., Sance´au, J., and Wietzerbin, J. (1992) J. Immunol. 148, 3912–3919. 13. Shaw, L. M., Messier, J. M., and Mercurio, A. M. (1990) J. Cell Biol. 110, 2167–2174.
eca
32. Hickstein, D. D., Hickey, M. J., and Collins, S. J. (1988) J. Biol. Chem. 263, 13863–13867. 33. Hickstein, D. D., Back, A. L., and Collins, S. J. (1989) J. Biol. Chem. 264, 21812–21817.
AP: Exp Cell
MONOCYTE ADHESION RESPONSES TO TGF-b1 AND IFN-g 34. Bellon, T., Lopez-Rodriguez, C., Rubio, M. A., Jochems, G., Bernabeu, C., and Corbi, A. L. (1994) Eur. J. Immunol. 24, 41–47. 35. Ignotz, R. A., Heino, J., and Massague´, J. (1989) J. Biol. Chem. 264, 389–397. 36. Kusunoki, T., Tsuruta, S., Higashi, H., Hosoi, S., Hata, D., Sugie, K., Mayumi, M., and Mikawa, H. (1994) J. Leukocyte Biol. 55, 735–742. 37. Welch, G. R., Wong, H. L., and Wahl, S. M. (1990) J. Immunol., 144, 3444–3448.
38. Berton, G., Fumagalli, L., Laudanna, C., and Sorio, C. (1994) J. Cell. Biol. 126, 1111–1121. 39. Hotchin, N. A., and Watt, F. M. (1992) J. Biol. Chem. 267, 14852–14858. 40. Ruoslahti, E. (1991) J. Clin. Invest. 87, 1–5. 41. Parodi, M. T., Cornaglia-Ferraris, P., and Ponzoni, M. (1989) Exp. Cell. Res. 185, 327–341. 42. Border, W. A., and Ruoslahti, E. (1992) J. Clin. Invest. 90, 1–7. 43. Panitch, H. S. (1992) Drugs 44, 946–962.
Received May 31, 1995 Revised version received September 18, 1995
/ m4861$3037
12-21-95 20:20:01
eca
217
AP: Exp Cell
