Alveolar macrophage urokinase receptors localize enzyme activity to the cell surface HAROLD A. CHAPMAN, PAUL BERTOZZI, LAURA ZENZIUS SAILOR, AND A. ROOM1 Departments and Harvard

NUSRAT

of Medicine, Brigham and Women’s Hospital Medical School, Boston, Massachusetts 02115

CHAPMAN, HAROLD A., PAUL BERTOZZI, LAURA ZENZIUS because recent evidence indicates that the cell-associated SAILOR,AND A. ROOMI NUSRAT.Alveolar macrophage urokiuPA activity of alveolar macrophages correlates with nase receptors localize enzyme activity to the cell surface. Am. J. Physiol. 259 (Lung Cell. Mol. Physiol. 3): L432-L438, 1990.Human alveolar macrophages are known to synthesize urokinase (uPA) and a specific plasminogen activator inhibitor, PAI2. In this study we have identified a uPA receptor expressed by these cells and defined the influence of PAI- on the interaction of uPA with its receptor. Alveolar macrophages from four normal volunteers were incubated with 55 kDa “‘I-labeled uPA (0.24-8 nM) in the presence or absence of excess unlabeled uPA. Specific and saturable binding was demonstrable in all cases. Scatchard plots were linear; regression analysis revealed a mean Kd of 5.25 nM (range 3.2-6.7) and mean B,,, of 30.7 femtomoles/lO’ cells (range 21.5-34.5). The structure of the uPA receptor was defined by electroblotting membrane fractions of macrophages and sequentially exposing filters to uPA and uPA antibodies. Membranes from macrophages demonstrated binding of either uPA or a 15-kDa amino-terminal fragment of uPA to a 55- to 60-kDa glycosylated membrane protein. Binding of uPA to filters was blocked by a synthetic oligopeptide containing the known receptor binding region of native uPA. Preincubation of 12”1-uPA with PAI- dramatically reduced the rate of association of uPA with macrophage uPA receptor. Conversely, receptor-bound uPA activity was less susceptible to inhibition by PAIthan soluble uPA activity. These data indicate that normal alveolar macrophages express uPA receptors. The receptor preferentially binds and protects free uPA over complexed enzyme, indicating that one function of the receptor is to allow the cells to express active uPA in an inhibitor-rich environment. plasminogen

activator

inhibitor;

protease

inhibitors

HUMAN ALVEOLAR MACROPHAGESare knowntoexpress a cell-associated form of a urokinase (uPA)-type plasminogen activator (4,9). As is the case with most migratory cells, this enzyme activity is viewed as a basic mechanism for the cell to move through connective tissue barriers (12, 25). In recent years accumulating evidence indicates that on most cell types, including mononuclear phagocytes, the cell-associated form of uPA is at least in part due to binding of uPA to a specific receptor(s) (2, 3, 29). However, no data are available with respect to receptor expression by human alveolar macrophages. We judged this question important not only because alveolar macrophages must migrate through several anatomic barriers in route to their ultimate anatomic locus but L432

1040-0605/90 $1.50 Copyright

0

lung function (Z3), suggesting a pathobiological role for macrophage uPA. Alveolar macrophages can also be induced to express a specific urokinase inhibitor (7). Previous reports from this laboratory have suggested that the cell-associated form of uPA on murine macrophages is relatively resistant to inhibition by soluble uPA inhibitors (10). A recent study of human blood monocytes reached a similar conclusion with respect to receptor-bound uPA (18). In this study we sought to first characterize functionally and morphologically a uPA receptor of human alveolar macrophages. We then explored the question of whether uPA inhibitors influence binding of uPA to this receptor. METHODS Reagents. uPA, obtained from the American Red Cross (Winkinase), was iodinated and used for saturation binding studies. Recombinant urokinase expressed in Chinese hamster ovary cells was used for all other experiments involving purified urokinase. A 15kDa amino-terminal fragment of urokinase was kindly provided by Dr. J. Henkin (Abbott Laboratories, Abbott Park, IL). This fragment of uPA is known to contain the receptor binding domain but is devoid of uPA activity (27). A synthetic peptide containing amino acids 12-32 of uPA was prepared by the Harvard Core Sequencing Laboratory. This peptide has been reported to compete with native uPA for binding to its receptor in U937 cells (1). Partially purified rabbit antibodies to uPA were obtained from Alpha Therapeutics (Los Angeles, CA). Iodinated goat anti-rabbit antibodies were obtained from New England Nuclear (Boston, MA). Methoxy-succinyl-alanine-alanine-proline-valine-chloromethylketone (Meosuc-AlaAla-pro-Val-CH&l), an elastase inhibitor, and benzoxylvaline-leucine-lysine-aminomethylcoumarin (Z-ValLeu-Lys-AMC), a plasmin substrate, were obtained from Enzyme Systems Products (Torrence, CA). N-glycanase was obtained from Genzyme (Boston, MA). Cell culture and preparation of macrophage membranes.

Human alveolar macrophages were obtained by bronchoalveolar lavage, counted, and incubated in vitro as previously described (9). The myelomonocytic cell line, HL-60, was maintained in RPM1 tissue culture media

1990 the American

Physiological

Society

Downloaded from www.physiology.org/journal/ajplung at Lunds Univ Medicinska Fak Biblio (130.235.066.010) on February 14, 2019.

ALVEOLAR

MACROPHAGE

supplemented with 15% fetal bovine serum (Hyclone, Logan, UT). For some experiments cells were incubated at lo6 cells/ml and stimulated with 16 nM phorbol myristate acetate (PMA) to initiate differentiation toward macrophages (24). Membrane fractions of alveolar macrophages and undifferentiated and differentiated HL-60 cells were prepared as previously described (10) but with the additional protease inhibitors: 0.1 mM Meosuc-AlaAla-pro-Val-CHz-Cl, trasylol (100 KU/ml), and 2 mM phenylmethanesulfonylfluoride. In experiments in which membrane uPA activity was to be measured, inhibitors were omitted. Binding assays. Binding of uPA to freshly lavaged alveolar macrophages was assessed essentially as previously described by Stopelli and Nielsen for A541 and U937 cells, respectively (19,28). Urokinase was iodinated by a chloramine-T method (0.07 mg/ml final concentration for 45 s), and labeled enzyme was separated from unbound 12’1 by gel chromatography in Sepharose (GlO, Pharmacia) in phosphate saline, pH 7.4, containing 0.25 mg/ml bovine serum albumin as carrier (8). The activity of iodinated uPA was determined by a flourometric assay as previously described except fluorescence was assayed against Z-Val-Leu-Lys-AMC, 0.1 mM final concentration, in the presence or absence of added plasminogen, 8 pg/ml (23). On average, the iodinated uPA had -50% the activity of noniodinated uPA. The specific activity of the probe was -2-4 X lo6 counts per minute (counts/ min) per microgram. As judged by autoradiography of polyacrylamide gels following electrophoresis nonreduced in sodium dodecyl sulfate (SDS), the iodinated uPA was over 90% the 55 kDa form of uPA. Equilibrium binding was performed by incubating freshly adherent alveolar macrophages (105/well) with varying concentrations of uPA, 0.24-8 nM, in tissue culture medium containing 0.25 mg/ml bovine serum albumin. In some experiments the range of uPA was extended to 16 nM to confirm saturation. After 1 h at room temperature the cell monolayers were rinsed three times with phosphatebuffered saline (PBS), pH 7.4, and lysed in 0.2% Triton X-100, 0.2% SDS in PBS. The rinse required -2 min to perform. Lysates were counted for bound radioactivity by P-scintillation. Initial experiments compared the effect of varying concentrations of unlabeled uPA on the fraction bound. These experiments indicated that -2540% of the total label was nonspecifically bound and that the addition of greater than 3O-fold M excess of unlabeled uPA had no further effect on bound radioactivity. To determine the stability of receptor-bound uPA, monolayers exposed to 1251-~PA were washed after 1 h as described above and placed in fresh medium without ligand. Cell-free media and lysed cells were sampled at various times after further incubation at either 4 or 37°C to assess the fraction of originally bound label remaining on the cells. Fibrin plate assay. Plasminogen activator activity was quantitated by a fibrin plate assay as previously described (9). Experiments were done to titrate the inhibitory effect of PAIon membrane-bound and soluble forms of uPA. For these experiments 24-h culture supernatants (tissue culture medium supplemented with 0.25 mg/ml bovine serum albumin) of endotoxin-stimulated

UROKINASE

RECEPTOR

L433

human monocytes were used as a source of PAI(26). The uPA inhibitory activity of the media was determined by titration of known amounts of purified uPA in the fibrin plate assay. Aliquots were frozen at -70°C until further use. The monocytic cell line, HL-60, was used as a source of uPA receptor following 48 h differentiation toward macrophages by stimulation with 16 nM PMA. The uPA activity of suspensions of membrane in PBS, pH 7.4, was quantitated by comparison with a uPA standard curve. Equivalent units of membrane-bound and soluble uPA were then mixed with increasing amounts of PAIand the inhibition of activity was determined by a 3-h fibrin plate assay. In some experiments the membrane preparations were acid-washed (28) at pH 3.0 in 0.05 M glycine and 0.1 M NaCl for 3 min followed by neutralization with 0.5 M N-2-hydroxyethylpiperazine-N’-2-ethanesulfonic acid (HEPES) to remove endogenous uPA. Membranes were incubated with excess soluble uPA to saturate binding sites, washed twice by 100,000 g 45-min centrifugations, and the subsequent suspensions were tested in the assays as described above. SDS gel electrophoresis and ligand blotting. For experiments examining 1251-~PA binding to whole cells, cells were lysed, at various times after incubation with 1251uPA, directly in the SDS gel sample buffer and electrophoresed in 10% polyacrylamide gels (19). The gels were fixed, stained, dried and exposed to Kodak XAR film for 1 day at -70°C. Ligand blotting was performed with modifications of methods previously reported to reveal the structure of low-density lipoproteins (LDL) and follicle-stimulating hormone (FSH) receptors (11, 13). Briefly, the total membrane fraction of macrophages was solubilized on ice in 40 mM glucopyranoside, 125 mM Tris maleate, pH 6.0, in the presence of proteinase inhibitors as described above and centrifuged at 12,000 g for 15 min. Solubilized membrane protein was measured with a BioRad (Piscataway, NJ) protein assay using a bovine serum albumin standard. Twenty-five to 100 pg of membrane protein were electrophoresed in 10% polyacrylamide gels under nonreducing conditions. Coomassie blue staining of the electrophoresed protein showed multiple discrete protein bands indicating that the solution contained solubilized membrane protein. The proteins were electroblotted onto nitrocellulose by “Western” blotting in buffers described by Burnette (5). The filters were then incubated for l-2 h in 5% nonfat dry milk to prevent nonspecific adsorption of proteins and then gently agitated for 16-20 h at 4°C in the presence or absence of 1 nM uPA in fresh milk. Filters were then washed thoroughly in milk and incubated sequentially with polyclonal uPA antibodies followed by iodinated goat anti-rabbit IgG. The filters were dried and exposed to film as described above. In some experiments solubilized membrane protein was preincubated overnight at 37°C with N-glycanase to remove A/-linked oligosaccharides (21). The reaction mixture for deglycosylation was 125 mM Tris, pH 8.0 containing 160 mM NaCl, 40 mM glucopyranoside, 10 mM 1,lO phenanthroline and 10 units N-glycanase. Preliminary experiments indicated that the addition of reducing agents to the sample buffer abolished all subsequent uPA binding.

Downloaded from www.physiology.org/journal/ajplung at Lunds Univ Medicinska Fak Biblio (130.235.066.010) on February 14, 2019.

L434

ALVEOLAR

MACROPHAGE

UROKINASE

tor-bound uPA (lane 2). Lanes 3 and 4 are identical to 1 and 2 except no uPA was added to the filters before development. The faint band at 55 kDa in the non-acid washed samples (lanes 3, 6) indicates that endogenous uPA is demonstrable on membranes of freshly isolated macrophages. To confirm the specificity of the binding pattern of uPA to membranes electroblotted onto nitrocellulose, we have performed two additional controls. These are illustrated in the remaining lanes of Fig. 2. First we have examined the staining pattern observed when the amino-terminal 15-kDa portion of the uPA molecule is substituted for uPA itself during the ligand binding. This fragment of the uPA molecule is known to contain the receptor binding region of uPA for previously described uPA receptors (27). Lanes 6 and 7 are controls again showing the binding pattern of native uPA (added only to lane 7) to macrophage membranes. Lane 5 illustrates that exactly the same pattern of staining is seen with the amino-terminal fragment (20 nM final concentration) of uPA as with uPA itself. We have also tested the ability of a synthetic peptide containing amino acid residues 12-32 of the uPA molecule to compete with uPA itself in the ligand blotting experiment. This peptide has been reported to compete with native uPA for binding to uPA receptors on either U937 cells or A415 cells (l), and we have confirmed this result with adherent alveolar macrophages (not shown). A representative result of this type of experiment is illustrated in lanes 8-10. The peptide itself, being only 21 residues, is not visualized by the uPA antibodies in this assay (lane 8). Lane 9 shows the pattern of binding of native uPA, 1 nM. The addition of the peptide (lane 10) in a final concentration of 1 mM markedly decreased the binding of uPA to the membrane protein. Taken together, these experiments indicate that uPA will specifically bind by its known receptor binding domain to a membrane protein(s) of 55-60 kDa on alveolar macrophages. Previous reports employing lectin chromatography have provided evidence for glycosylation of the uPA receptor in PMA-stimulated U937 cells (17, 24, 25). To explore this question we preincubated the solubilized macrophage membrane protein with N-glycanase, a fun-

RESULTS

Binding of 1251-uPA to human alveolar macrophages. Freshly isolated human alveolar macrophages from each of four separate subjects were incubated at room temperature for 1 h with various concentrations of uPA from 0.24 nM to 8 nM in the presence or absence of excess unlabeled uPA. Binding was maximal after -20-30 min and that total binding was maximally displaced by approximately 3O-fold excess cold uPA. Specific binding was calculated as the difference between total bound label per lo5 cells after 1 h and that bound in the presence of 50-fold excess cold uPA. Data in Fig. 1 illustrate the mean t SD specific binding of labeled uPA to macrophages obtained from the four individuals as a function of uPA concentration. A representative Scatchard plot of the binding data is shown as an inset in Fig. 1. All the Scatchard analyses were linear, suggesting one major class of binding sites with a mean & of 5.2 nM. These analyses also indicated that, on average, there are -190,000 binding sites per cell. In additional experiments the time course of turnover of the bound label following washing of the cells to remove unbound label and the influence of temperature were also measured. In five separate experiments the average fraction of bound ligand still remaining on the cell after 2 h was 54% (range 33%-75%) whether the cells were incubated at 4 or 37°C. Initial equilibrium binding was also similar at 4 and 23°C (not shown). These data are in agreement with previously reported values for turnover and temperature dependence of uPA receptors on U937, A421, and HeLa cells (15, 24, 28). Demonstration of receptor structure by Western blotting. A total membrane preparation (50 pg total protein) of freshly isolated human macrophages was electrophoresed, electroblotted, and then exposed sequentially to uPA (1 nM), rabbit uPA antibodies, and iodinated goat anti-rabbit antibodies. Data in Fig. 2 lanes 1, 7, and 9, illustrate representative results. The binding of exogenous uPA appears as a relatively broad band of 55-60 kDa. Binding is still evident when membranes have been acid-washed before electrophoresis to remove any recep18-

I a r

RECEPTOR

,

16.

,

14.

,

12a

FIG. 1. A: binding of lz51-uPA to human alveolar macrophages. Freshly adherent alveolar macrophages were exposed at room temperature to varying concentrations of iodinated uPA, washed, lysed, and bound radioactivity measured. Specific binding and mean k SD, in fmol/lO” cells of uPA as a function of uPA concentration are shown. Inset: representative Scatchard plot of saturation binding. See text for definitions.

,

R2 = .827

0 s

IO

IS

20

2s fm

0

1

2

3

4

5

6

7

80 bound/

3s

40

4s

so

s

10 5 HAH

0

9

UPA, nH Downloaded from www.physiology.org/journal/ajplung at Lunds Univ Medicinska Fak Biblio (130.235.066.010) on February 14, 2019.

ALVEOLAR

ATF

MACROPHAGE

Peptide

UROKINASE

RECEPTOR

L435

Peptide

kDa 92 66

45 30

12

34

5

6

7

8

9

IO

FIG. 2. Visualization of uPA receptor by immunoblotting of human alveolar macrophage membranes. Total membrane fractions of freshly isolated alveolar macrophages were electrophoresed, electroblotted, and stained for uPA receptor by sequential exposure of the filters to uPA (indicated by a + at top of figure), uPA antibodies, and secondary iodinated antibodies. Lanes l-4 represent electrophoresis of 50 fig of total membrane protein either with, 1 and 2, or without, 3 and 4, addition of uPA, 1 nM. Membranes loaded into lanes 2 and 4 had been acid-washed to remove any receptor-bound uPA. Lanes 5-7 show pattern of staining when 20 nM amino-terminal uPA (lane 5) is substituted for native uPA (lane 7). Lane 6 is again a control to which no exogenous uPA was added. Lanes 8-10 show effect of addition of excess (1.5 PM) receptor binding peptide on ability of native uPA to bind to macrophage membrane protein. Lane 9 is again the pattern of staining with native uPA, lane 10 the staining with uPA plus peptide, and lane 8 is the peptide alone.

gal derived enzyme that cleaves core N-linkages of oligosaccharides (21). Fig. 3, lane 2, illustrates the effect of N-glycanase pretreatment on the subsequent binding of uPA to macrophage membrane protein relative to control membrane protein (lane I). Clearly the size distribution is altered by preincubation with N-glycanase to a lower molecular mass range, i.e., 40-50 kDa (lane 2). Although the size range is lower, the broad nature of the binding band is not abrogated by pretreatment with the enzyme. It should be noted that the membrane proteins may not be fully deglycosylated, as no reducing agents were used in the denaturation of proteins subjected to N-glycanase treatment. As discussed earlier, reducing agents blocked all uPA binding in this assay. The reduction in the extent of binding of uPA apparent in lane 2 vs. lane 1 was reproducible in three separate experiments. Lanes 1, 2, 4, and 5 also show a discrete band at -85 kDa that is not affected by N-glycanase treatment. The 85-kDa band is not a uPA binding protein, as the staining is not dependent on the addition of exogenous uPA to the assay; its nature is unknown. The 85-kDa band is presumably not glycosylated by N-linked saccharides as N-glycanase pretreatment had no effect on this band. The pattern of uPA binding to membrane protein

FIG. 3. Comparison of HL-60 and alveolar macrophage uPA receptors: effect of PMA stimulation and partial dezlvcosvlation. Total membrane preparations of unstimulated and 72yh PMA-stimulated HL-60 cells and freshly lavaged alveolar macrophages were electrophoresed and analyzed for receptor as described in text. Lanes contain the following: lanes 1 and 2 show control and N-glycanase-treated alveolar macrophage membrane (50 pg each), respectively; lane 3, undifferentiated HL-60 membrane (100 rre): lane 4. PMA-stimulated HL-60 (25 pg); lane 5, PMA-stimulated H’LY60 membranes (25 pg) pretreated with N-glycanase.

derived from the unstimulated myelomonocytic cell, HL60, was compared with that of cells stimulated with PMA for 72 h. Fig. 3, lanes 3-5, illustrates representative results obtained with these cells. Undifferentiated cells (100 pg total protein) have a discrete band visualized at -55 kDa (lane jl), and the appearance of this band is entirely dependent on incubation with exogenous uPA. In contrast, PMA stimulation results in a marked increase in the heterogeneity of staining with uPA (lune 4). This increase in heterogeneity is not simply a function of more receptor protein, as a similar pattern was observed over a range (lo-75 pg) of electrophoresed protein. The pattern of binding of uPA to membranes of PMA-stimulated HL-60 cells was similar to that seen with membrane derived from human alveolar macrophages. Similar findings were observed with membrane protein derived from either U937 or THP-1 cells, two other human monocytic leukemic cell lines (not shown). We again tested the effect of N-glycanase on the pattern of binding of uPA to membrane preparations of differentiated HL-60 cells (lane 5) and found a reduction in size of the protein bound by uPA. The increased heterogeneity observed with membrane protein derived from PMA-stimulated cells over that of unstimulated cells was not abrogated by N-glycanase pretreatment. Preferential binding of free UPA over UPA complexed with the specific UPA inhibitor, PAI-2. Because macro-

phages are known to synthesize and release the uPA inhibitor, PAI-2, as well as uPA, we have tested the ability of uPA complexed to PAI- to bind to the macrophage uPA receptor. Partially purified placental PA inhibitor was used as a source of PAI(8). For these experiments uPA was iodinated with more dilute chloramine-T (0.02 mg/ml) and the 1251-~PA was -80% active. Iodinated uPA was preincubated for 2-4 h at 37°C with

Downloaded from www.physiology.org/journal/ajplung at Lunds Univ Medicinska Fak Biblio (130.235.066.010) on February 14, 2019.

L436

ALVEOLAR

MACROPHAGE

UROKINASE

molar excess PAI- and then tested for its ability to bind to freshly adherent human alveolar macrophages. Complex formation was monitored by SDS gel electrophoresis of free and PAI-2-treated 1251-~PA (lanes 1 and 2, respectively, Fig. 4B). Initial experiments were done after a 2-h incubation of either free or precomplexed uPA with the macrophages. No obvious differences in binding were noted. Therefore we performed time course experiments to determine the rate of association of free and complexed uPA with the macrophage uPA receptor. Data in Fig. 4A show that free uPA, 1 nM, associates much more rapidly than precomplexed enzyme with macrophages. Similar results were seen in each of three separate experiments. In additional experiments, companion wells for the binding experiments were lysed at various times, 110. p

100.

Gc .C m

90.

A

80.

z

E .C

-IO.

: r

60.

n

50.

l

d

lb

ii

3b

46

Time

B UPA

+

+

PA1 a?* - bw* +

-IO min + +++

uPA+PAI

5b

6b

7b

(mid

30 -

min

5

6

-+-+

RECEPTOR

electrophoresed, and visualized by autography to ascertain whether free and precomplexed enzyme associate the same or differently with the macrophages. The autoradiograph illustrated of Figure 4B, lanes 3-6, confirmed the binding data shown in Fig. 4A that only the free enzyme effectively associates with macrophages during the first 30 min of incubation. As can be discerned from the figure, some complexed enzyme does eventually bind to the membrane. Densitometry of this autoradiograph indicated that -20% as much complexed enzyme bound to the cell at 30 min as did free enzyme. Differential inhibition of bound and free UPA by the urokinase inhibitor PAI-2. Murine macrophage cell-as-

sociated plasminogen activator is relatively protected from inhibitor by murine macrophage-derived uPA inhibitors by comparison with soluble human urokinase activity (10). Because the data described above demonstrate that the human macrophage uPA receptor preferentially binds free over complexed enzyme, we next determined whether free enzyme, once bound, was protected from inhibition by PAI-2. For these experiments membrane protein derived from PMA-stimulated THP1 or HL-60 cells was used as a source of receptor and monocyte conditioned media as a source of PAI-2. Moncytes are known to exclusively release PAIand the form released is mostly the 60- to 64-kDa extracellular form of PAI(16, 26). Data in Fig. 5 compare the fractional inhibition of receptor bound uPA (closed circles) and soluble uPA (open circles) by increasing amounts of PAI-2. The data show that both forms of the enzyme can be inhibited by PAI-2. However, the membrane-bound form of uPA is inhibited considerably less effectively than free enzyme. In the experiment shown when at least 80% of soluble uPA is inhibited, ~20% of the membrane-bound form of the enzyme is blocked. Similar results were obtained in five separate experiments, and the average percent inhibition of membranebound enzyme was -8% (range 2-30) when the soluble enzyme was 50% inhibited. An identical pattern of preferential inhibition of free over membrane-bound uPA was observed when placental PAI- was substituted for monocyte PAI- and when membranes were acid-washed

55 kDo

I

2

3

4

4. Rate of association of free uPA and uPA complexed with the urokinase inhibitor, PAI-2, with uPA receptors. A: human alveolar macrophages were allowed to adhere for 1 h at 37°C and then exposed to ‘““I-uPA (open circles) or ““I-uPA precomplexed with PAI(closed circles), 1 nM final concentration in each case, for various times as indicated. Data shown are specific binding of uPA as a function of time in minutes. ““I-uPA was allowed to complex at 37°C for 4 h with placental PAI-2, 25 M excess of inhibitor over enzyme. B: an autoradiograph of lysates of macrophages exposed to free or complexed irsIuPA at 10 and 30 min. Cell monolayers were washed 3 times with PBS and immediately lysed in sample buffer for electrophoresis. FIG.

FIG. 5. Differential inhibition of membrane-bound and free uPA by the urokinase inhibitor, PAI-2. Equivalent uPA units of activity of purified, soluble uPA (open circles) and suspensions of membrane fractions (closed circles) prepared from PMA-stimulated HL-60 cells were titrated with increasing amounts (microliters) of PAI-2-rich supernatants conditioned by overnight stimulation of human monocytes with endotoxin, 100 rig/ml. Data are expressed as percent inhibition of control activity. Similar results were obtained when membrane fractions were acid-treated and neutralized before exposure to uPA. See text for definitions.

Downloaded from www.physiology.org/journal/ajplung at Lunds Univ Medicinska Fak Biblio (130.235.066.010) on February 14, 2019.

ALVEOLAR

and then saturated (not shown).

with

soluble uPA prior

MACROPHAGE

to analysis

DISCUSSION

Observations that human alveolar macrophages express a uPA receptor having a -5 nM & and 1-2 X lo5 binding sites per cell are consistent with the characteristics of previously reported uPA receptors on monocytic cell lines (3, 19, 27, 29). Also consistent with prior reports, we find that the turnover of bound macrophage uPA is a matter of hours rather than minutes. Although the binding characteristics of the uPA receptor have been extensively studied, little information is available regarding the influence of PA inhibitors on uPA binding. Our data indicate that the uPA receptor preferentially binds free over inhibitor-complexed enzyme (Fig. 4) and that, once bound, free uPA is relatively protected from soluble uPA inhibitors (Fig. 5). One caveat in the method we used to demonstrate preferential inhibition of soluble over membrane-bound uPA (Fig. 5) is that our analyses are based on the assumption that equivalent activities of the soluble and membrane-bound forms reflect equivalent molar amounts of these two forms. Based on available data this assumption seems reasonable. Ellis and co-workers (14) recently examined the kinetics of plasminogen activation by soluble and membrane bound forms of uPA. These investigators found that membranebound uPA converted plasminogen to plasmin with similar or faster reaction rates than the soluble form of the enzyme. Thus it is unlikely that our result of preferential inhibition of soluble over bound enzyme could be explained by higher molar amounts of less active bound enzyme in the comparative assays we performed. In addition the preferential inhibition of free enzyme function (Fig. 5) is consistent with the preferential binding of free over complexed enzyme (Fig. 4). Recently reported studies of receptor-bound uPA on human blood monocytes also concluded that bound uPA is relatively resistant to PAI- and that uPA/PAI-2 complexes bind to uPA receptors with much lower affinity (18). The ability of the uPA receptor to protect uPA activity should allow macrophages to maintain some degree of active cell uPA even in an inhibitor-rich environment. This appears to be the case. We have previously found that alveolar macrophages lavaged from lungs of patients with active alveolitis due to either idiopathic pulmonary fibrosis or sarcoidosis have cell-associated uPA in the range of that observed normally. However, the bronchoalveolar lining layer in these patients has little free uPA, whereas free uPA inhibitors are demonstrable (6). It should be noted that our binding data (Fig. 4) also indicate that uPA complexed to inhibitor is slowly bound and that bound active enzyme can be eventually inhibited if inhibitor concentrations are high enough (Fig. 5). Therefore other cooperative mechanisms may exist to protect bound uPA from inhibition by soluble inhibitors. Pollanen and colleagues (22) have reported localization of fibroblast uPA receptors along the ventral surface of these adherent cells, whereas uPA inhibitors were excluded from this region (22). Similarly, Herbert and Baker have reported linkage of the uPA receptor to the

UROKINASE

L437

RECEPTOR

cellular cytoskeleton (17). Thus it is possible that uPA receptors can move within the membrane of the cell to zones of contact between cells and surfaces and thereby exclude soluble inhibitors from access to uPA. Taken together, the available data suggest a modified version of previously proposed models (3) for functioning of the uPA receptor on tissue macrophages. The receptor preferentially binds either pro-urokinase or free uPA (Fig. 4) and, once uPA is bound (and activated in the case of pro-uPA), the receptor provides a relatively protected environment. Since plasmin is the only known physiological activator of pro-uPA (3)) the preferential binding of active uPA should allow the continued activation of plasmin at the cell surface necessary for positive feedback activation of new pro-uPA. Continued synthesis of pro-uPA by the cells then allows the continued presence of active cell-associated enzyme even in an inhibitor-rich environment. Adherence may further protect and prolong the life of surface uPA activity, although we have no direct evidence that this is the case. The structure of the uPA receptor has been previously analyzed by cross-linking of surface receptors with iodinated uPA (or a fragment thereof) (15,20,28). Although the immunoblotting technique we have adapted for visualization of the uPA receptor does not discriminate between intracellular pools of receptor and surface receptor, this technique does appear to complement data derived from cross-linking experiments. For example, results obtained with ligand blotting imply that disulfide bonds are critical to receptor function, as we observe no binding of uPA to membrane protein electrophoresed under reducing conditions. This technique also further supports prior suggestions of a correlation between the structural properties of the receptor and its binding characteristics. Previously reported data indicate that the differentiation of U937 cells by stimulation with PMA is accompanied by increases in B,,, and & as well as an increase in heterogeneity of the receptor visualized by cross-linking (19). The increase in both amount (per microgram of total membrane protein) and heterogeneity of the uPA binding protein(s) of PMA-stimulated HL60 cells compared with that of undifferentiated HL-60 cells is quite apparent by electroblotting (Fig. 3). Finally, the marked reduction in uPA binding to N-glycanase pretreated receptor protein of human alveolar macrophages (Fig. 2, lane 2) suggests that carbohydrate may be an important determinant of receptor affinity in these cells. The availability of a simple method to visualize the receptor in gels should facilitate further elucidation of the structural correlates of uPA receptor function. This work was supported by National Heart, Lung, and Blood Institute Grant HL-35505 and a grant from the Council for Tobacco Research. H. A. Chapman is a Career Investigator of the American Lung Association. A. R. Nusrat is a recipient of an Individual National Research Service Award (HL-08002). Address for reprint requests: H. A. Chapman, Respiratory Division, Brigham and Women’s Hospital, 75 Francis St., Boston, MA 02115. Received

6 November

1989; accepted

in final

form

23 May

1990.

REFERENCES 1. APPELLA, E., E. A. ROBINSON, S. J. ULLRICH, M. P. STOPPELLI, A. CORTI, G. CASSANI, AND F. BLASI. The receptor-binding se-

Downloaded from www.physiology.org/journal/ajplung at Lunds Univ Medicinska Fak Biblio (130.235.066.010) on February 14, 2019.

L438

ALVEOLAR

MACROPHAGE

quence of urokinase. J. Biol. Chem. 262: 4437-4440, 1987. 2. BLASI, F. Surface receptors for urokinase plasminogen activator. Fibrinolysis 2: 73-84, 1988. 3. BLASI, F., J-K. VASSALLI, AND K. DANO. Urokinase-type plasminogen activator: proenzyme, receptor, and inhibitors. J. Cell Biol. 104: 801-804,1987. 4. BOWEN, R. M., J. R. HOIDAL, AND R. D. ESTENSEN. Urokinasetype plasminogen activator in alveolar macrophages and bronchoalveolar lavage fluid from normal and smoke-exposed hamsters and humans. J. Lab. Clin. Med. 106: 667-673,1985. 5. BURNETTE, W. N. “Western Blotting”; electrophoretic transfer of proteins from SDS-polyacrylamide gels to unmodified nitrocellulose and radiographic detection with antibody and radioiodinated protein A. Anal. Biochem. 112: 195-203, 1981. 6. CHAPMAN, H. A., C. L. ALLEN, AND 0. L. STONE. Abnormalities in pathways of alveolar fibrin turnover among patients with interstitial lung disease. Am. Rev. Respir. Dis. 133: 437-444, 1986. 7. CHAPMAN, H. A., AND 0. L. STONE. A fibrinolytic inhibitor of human alveolar macrophages. Induction with endotoxin. Am. Rev. Respir. Dis. 132: 569-575, 1985. 8. CHAPMAN, H. A., AND 0. L. STONE. Characterization of a macrophage-derived plasminogen-activator inhibitor. Similarities with placental urokinase inhibitor. Biochem. J. 230: 109-116, 1985. 9. CHAPMAN, H. A., 0. L. STONE, AND Z. VAVRIN. Degradation of fibrin and elastin by intact human alveolar macrophages in vitro: characterization of a plasminogen activator and its role in matrix degradation. J. Clin. Invest. 73: 806-815, 1984. 10. CHAPMAN, H. A., JR., Z. VAVRIN, AND J. B. HIBBS, JR. Macrophage fibrinolytic activity: Two pathways of plasmin formation by intact cells and an inhibitor of plasminogen activator. Cell 28: 653-662, 1982. 11. DANIEL, T. O., W. J. SCHNEIDER, J. L. GOLDSTEIN, AND M. S. BRO w N. Visualization of lipoprotein receptors by ligand blotting. J. Biol. Chem. 258: 4606-4611,1983. 12. DANO, K., P. A. ANDERSON, J. GRONDAHL-HANSEN, P. KRISTENSEN, L. S. NIELSEN, AND L. SKRIVER. Plasminogen activators, tissue degradation, and cancer. Adv. Cancer Res. 49: 139-266,1985. 13. EIDNE, K. A., H. T. HENDRICKS, AND R. P. MILLAR. Demonstration of a 60K molecular weight luteinizing hormone-releasing hormone receptor in solubilized adrenal membranes by a ligandimmunoblotting technique. Endocrinology 116: 1792-1795, 1985. 14. ELLIS, V., M. F. SCULLY, AND V. V. KAKKAR. Plasminogen activation initiated by single-chain urokinase-type plasminogen activator. J. Biol. Chem. 264: 2185-2188, 1989. 15. ESTREICHER, A., A. WOHLWEND, D. BELIN, W-D. SCHLEUNING, AND J-D. VASSALLI. Characterization of the cellular binding site for the Urokinase-type plasminogen activator. J. Biol. Chem. 264: 1180-1189,1989. 16. GENTON, C., E. K. 0. KRUITHOF, AND W-D. SCHLEUNING. Phorbol ester induces the biosynthesis of glycosylated and nonglycosylated plasminogen activator inhibitor 2 in high excess over urokinasetype plasminogen activator in human U-937 lymphoma cells. J. Cell BioZ. 104: 705-712, 1987.

UROKINASE

RECEPTOR

17. HEBERT, C. A., AND J. B. BAKER. Linkage of extracellular plasminogen activator to the fibroblast cytoskeleton: colocalization of cell surface urokinase with vinculin. J. Cell Biol. 106: 1241-1247, 1988. 18. KIRCHHEIMER, J. C., AND H. G. REMOLD. Functional characteristics of receptor-bound urokinase on human monocytes: catalytic efficiency and susceptibility to inactivation by plasminogen activator inhibitors. Blood 74: 1396-1402, 1989. 19. NIELSEN, L. S., G. M. KELLERMAN, N. BEHRENDT, R. PICONE, K. DANO, AND F. BLASI. A 55,000-60,000 Mr Receptor Protein for Urokinase-type Plasminogen Activator. Identification in Human Tumor Cell Lines and Partial Purification. J. Biol. Chem. 263: 2358-2363,1988. 20. PICONE, R., E. L. KAJTANIAK, L. S. NIELSEN, N. BEHRENDT, M. R. MASTRONICOLA, M. V. CUBELLIS, M. P. STOPPELLI, S. PEDERSEN, K. DANO, AND F. BLASI. Regulation of urokinase receptors in monocyte-like U937 cells by the phorbol ester, phorbol myristate acetate. J. Cell Biol. 108: 693-702, 1989. 21. PLUMMER, T. H., J. H. ELDER, S. ALEXANDER, A. W. PHELAN, AND A. L. TARENTINO. Demonstration of peptide: N-glycosidase F activity in endo-beta-N-acetylglucosaminidase F preparations. J. Biol. Chem. 259: 10700-10704,1984. 22. POLLANEN, J., 0. SAKSELA, E. M. SALONEN, P. ANDREASEN, L. NIELSEN, K. DANO, AND A. VAHERI. Distinct localizations of Urokinase-type plasminogen activator and its Type 1 inhibitor under cultured human fibroblasts and sarcoma cells. J. Cell Biol. 104: 1085-1096,1987. 23. REILLY, J. J., JR., AND H. A. CHAPMAN, JR. Association between macrophage plasminogen activator activity and lung function in young cigarette smokers. Am. Rev. Respir. Dis. 138: 1422-1428, 1988. 24. ROVERA, G., D. SANTOLI, AND C. DAMSKY. Human promyelocytic leukemia cells in culture differentiate into macrophage-like cells when treated with a phorbol diester. Proc. Natl. Acad. Sci. USA 76: 2779-2783, 1979. 25. SAKSELA, 0. Plasminogen activation and regulation of pericellular proteolysis. Biochem. Biophys. Acta 823: 35-65, 1985. 26. SCHWARTZ, B. S., M. C. MONROE, AND E. G. LEVIN. Increased release of plasminogen activator inhibitor Type 2 accompanies the human mononuclear tissue factor response to lipopolysaccharide. Blood 71: 734-741,1988. 27. STOPPELLI, M. P., A. CORTI, A. SOFFIENTINI, G. CASSANI, F. BLASI, AND R. K. ASSOIAN. Differentiation-enhanced binding of the amino-terminal fragment of human urokinase plasminogen activator to a specific receptor on U937 monocytes. Proc. Natl. Acad. Sci. USA 82: 4939-4943,1985. 28. STOPPELLI, M. P., C. TACCHETTI, M. V. CUKELLIS, A. CORTI, V. J. HEARING, G. CASSANI, E. APPELLA, AND F. BLASI. Autocrine saturation of pro-urokinase receptors on human A431 cells. Cell 45: 675-684,1986. 29. VASSALLI, J. D., D. BACCINO, AND D. BELIN. A cellular binding site for the Mr 55,000 form of the human plasminogen activator, urokinase. J. Cell Biol. 100: 86-92, 1985.

Downloaded from www.physiology.org/journal/ajplung at Lunds Univ Medicinska Fak Biblio (130.235.066.010) on February 14, 2019.