British /oirrrrul o/ Harmutolog!/. 1992, 81, 8 1-8 5
Specific binding between human neutrophils and heparin C. LECULIEK, 0. BENZERARA,* N. C O U P R I E . A . FRANCINA,Y. LASNE,*E. A R C H I M B A U D t A N D
D. F I E R E t
Laboratoire de Biochimie B, Hbpital Edouard Herriot. 6 9 4 3 7 Lyon Cedex 03, *Laboratoire Central des Radio-lsotopes, Hbpital Edouard Herriot, 6 94 37 Lyon Cedex 03, and ?Service d'Hkmatologie Clinique, Hbpital Edouard Herriot, 6 9 4 3 7 Lyon Cedex 03, France Received 17 October 1 99 I , accepted for publication 3 January 1992
Summary. Heparin binding on polymorphonuclear leucocytes (PMNL) was characterized. Heparin binding was specific. rapid, saturable and reversible. One single class of heparin binding sites was found with a dissociation constant of 1.22 pmol/l and 7 . 7 x 10h sites per PMNL. The binding was independent of the anticoagulant activity of heparin. Heparin affinity chromatography on radio-jodinated cell lysates followed by polyacrylamide gel electrophoresis in the presence of sodium dodecylsulphate revealed a 130 kD heparin
binding protein. Heparin binding was inhibited by disodium salt of ethylenediamine tetraacetic acid. Cell surface bound heparin was functionally inactive and did not affect the inactivation of thrombin by antithrombin 111. Our study demonstrates that heparin interacts with PMNL by a cellsurface binding protein. These instructions could be consistent with the modifications of some PMNL functional properties in the presence of heparin.
Heparin affects the oxidative metabolism and degranulation of polymorphonuclear leucocytes (PMNL) (Laghi-Pasinirt ul, 1983. 1984). Heparin increases PMNI, adherence in vitro and potentiates the effect ofchemotactic factor formylmethionylleucyl-phenylalanine (FMLP) (Cairo ct ul. 198 3 ) . No specific binding of heparin to PMNL has been reported and identifed so far. In the present study we have characterized the specific binding of heparin to PMNI,. The heparin binding was found to be time-dependent, saturable, specific and reversible. The binding was inhibited by the bivalent metal chelator ethylenediamine tetraacetic acid (disodium salt) (EDTA). Neutrophil cell surface-bound heparin was demonstrated as functionally inactive. The protein interacting with heparin was revealed by cell surface radioiodination and heparin affinity chromatography, followed by polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulphate (PACK-SDS) on chromatographic fractions and autoradiograph y.
rombin 111 were purchased from Boehringer Mannheim (Meylan, France). Thrombin was purchased from Sigma (Saint-Louis. Mo.. U.S.A.) and thrombin substrate (S 2238) from Kabi (Stockholm, Sweden). Heparin-Sepharose CL-6B was purchased from Pharmacia (Saint-Quentin-en Yvelines, France). All reagents and buffers used in the present study were free of edotoxin. Gl!~c.osuririnogl!~c.uns. [N-s~lphonate-'~S]heparin was purchased from Amersham. The specific activity was 1.48 GBq/g and the molecular weight (MW) I6 000. Sodium heparin (grade I). 170 LISP units/mg (United States Pharmacopoeia), with a MW of 16 000 and chondroitin-4-sulphate (sodium salt) with a MW of 50000 were purchased from Sigma. Dermatan sulphate (sodium salt) with a MW of 14 500 was purchased from Calbiochem (San Diego, Calif., U.S.A.). Heparin (sodium salt, MW 1 7 500, anticoagulant activity: 5000 (IU) International Units/ml) was purchased from Roche (Neuilly, France) and Fraxiparine (calcium salt, MW 4500. anticoagulant activity 10 250 anti-Xa IU/ml) from Choay Institute (Paris. France). Humun nmtrophils. Whole blood from healthy donors was collected on ACD (citric acid-sodium citrate-dextrose). Human peripheral blood neutrophils were isolated by dextran sedimentation followed by centrifugation on FicollHypaque according to published techniques (Boyum. 1968). Remaining red blood cells were lysed by hypotonic treatment (0.2%NaCI). Purity of PMNL preparation was >95% as judged by Wright-Giemsa staining. PMNL suspensions were
MATERIALS AND METHODS Reagvnfs. Carrier free ['2iI]sodium iodide was purchased from Amersham (Les tllis. France). EDTA. 3 - [ ( 3-cholamidopropy1)-dimethylammonio]-1-propane sulphonate (CHAPS), phosphatidylinositol phospholipase C (PI-PLC) and antithCorrespondence: Dr A. Francina. Laboratoire de Biochimie B. HBpital Edouard Herriot. 69437 Lyon Cedex 0 3 . France.
81
82
C. Leculier et a1
determined to be free of endotoxin, using the Limulus test (Amilabo. Paris, France). Heparin binding to human neutrophils. The procedure was identical to that described elsewhere (Leung et al, 1989). Briefly, neutrophils (7 x 10h/ml) were incubated in phosphate-buffered saline, pH 7.40, containing 2% bovine serum albumin (PBS/BSA), for 1 h at 4OC, 3SS-heparin. in the presence or absence of 100-fold molar excess of unlabelled heparin, was added to cell samples and incubated for 1 h at 4' C. Final cell concentration was 3.6 x 1Ohce1k/mlin a final volume of 300 pl. Cell separation was performed immediately after incubation using a dot-blot apparatus (Biorad, Richmond, Calif., U.S.A.) in which a nitrocellulose membrane was inserted (pore size 0.45 pm;Biorad). 300 pI of cell suspension were poured in each well. The membrane was carefully cut after repeated washings in PBS. The respective pieces of membrane were placed into scintillation tubes containing Instagel@(Packard. Pans, France). The cell-bound radioactivity was counted in a Packard beta-counter. Each measurement was the mean of two assays. lnternalization of bound heparin by ncutrophils. The incubations were carried out as described for heparin binding to monocytes (Leung et a!, 1989). Binding of 35S-heparin(5.2 p~ in final concentration) was assayed at two different temperatures ( 4 O C and 37OC) with incubation times of 1 h and 3 h. Following incubation, neutrophils were washed in cold PBS. Cells were then treated with trypsin (0.25% for 30 min) after the addition of a 100-fold molar excess of unlabelled heparin. Fractionation of 35S-heparin by AT Ill affinity chromatography. 1 g of AT 111-agarose (Sigma) was suspended in PBS pH 7.40. An aliquot of 3SS-heparin(5.20 p ~corresponding ) to about 10% of column capacity of fixation in heparin, was applied on an AT 111 column equilibrated in PBS. Elution was carried out with PBS, followed by PBS containing 2 M NaCI. The corresponding fractions were measured by radioactive counting (low affinity and high affiity heparin fractions), separately pooled. dialysed and lyophilized. Heparin binding to human neutrophils experiments were performed with unfractionated heparin, low affinity and high affinity fractions after adjusting the respective radioactivities. Binding of heparin in the presence of EDTA. Heparin binding to human neutrophils was studied in the presence of EDTA at final concentrations of lo-' and M. ISSheparin final concentration was 1.3 p ~ Results . were expressed in per cent of controls without EDTA. '%-heparin binding curve was also assayed in the presence of EDTA at final concentration of lo-' M corresponding to maximal effect of EDTA. Heparin affinity chromatography of surface iodinated neutrophi/ cell lysates. Neutrophils (2 x 1OY/mlin 0.2 M sodium phosphate. pH 7.20) were surface radioiodinated using Enzymobeads method (Biorad) according to the instructions of the manufacturer and using a total activity of Na12SIof 74 MBq. PI-PLC treatment of neutrophils was performed before cell lysis with 20 mU PI-PLC/107 cells in PBS, pH 7.40. at 3 7°C for 45 min. The labelled cells were washed three times in PBS and solubilizedby 15 mmol/l CHAPS with PMSF 2 m~ and aprotinin 3 p~ in PBS for 1 h at 4°C. Insoluble materials
were removed by centrifugation and cell lysates were applied on a heparin-Sepharose 6B column, equilibrated in PBS. and after extensive washings, eluted with NaCl solutions (0.152 M). The eluates were concentrated and analysed by PAGESDS (Laemmli, 1970) and autoradiography. Control experiments were performed after addition of unlabelled heparin in excess into PMNL lysates. that were incubated overnight at 4 O C on heparin-Sepharose column and eluted by NaCl
solutions. Eflect of heparin bound to neutrophils on inactivation of thrombin. The functional role of neutrophil-bound heparin was investigated as previously described for the effect of heparin-pretreated U93 7 cells on thrombin inactivation by AT I11 (Leung et al, 1989). Briefly, PMNL and control red blood cells (RBC) (1.2 x 107/ml)were incubated with RPMI/ BSA. Cells were washed with PBS/BSA and fixed with glutaraldehyde. Thrombin (final concentrations 10 U/ml) and AT 111(final concentration, 10 pg/ml) in 40 mmol/l TrisHCI, 0.15 M NaCI, pH 8.30, were added to the fixed cells preincubated or not with heparin at concentrations of 0.11000 p ~ After . 30 s of incubation, thrombin substrate S 22 38 (final concentration, 0.5 mmol/l) was added. After 30 s, acetic acid (final concentration 15.6%)was added to stop the reaction. The reaction mixtures were centrifuged and the absorbance of the supernatants read at 405 nm. Each experiment was repeated three times with duplicate measurement of each point. RESULTS Heparin binding to human neutrophils 35S-heparinbound to neutrophils in a concentration-depenHEPARI! (ng/lO
BOUND cells)
I 250
200
150
100
50
0 0
1
2
3 4 [HEPARIN] (PM)
5
6
Fig ](a). Heparin binding to neutrophils. Neutrophils ( 3 . 6 x loh, h a 1 volume: 300 pl) were incubated for 1 h in PBS/BSA. at 4% "Sheparin was added in the presence or in the absence of unlabelled heparin for 1 h at 4°C. The cell suspensions were filtered through nitrocellulose membranes and cell-bound radioactivities were measured. Each point represents the mean of duplicate determinations. A: Typical total %-heparin binding to neutrophils.B Non-specific lSSheparin binding. C: Total 35S-heparinbinding minus non-specific %-heparin binding. D Total 3sS-heparinbinding minus non-specific "S-heparin binding in the presence of EDTA (lo-' M).
Binding between Human Neutrophils and Heparin B/AT
83
HEPARIN BOUND (percent)
0.041 120 100
I
0.03 80
0.02
60
40 0.01 20
1
0.00
0
20
10
30
40
o !
50
0
B (m)
-
t
0
o
I
1
20
40
I
40
MOLAR EXCESS OF UNLABELED LIGAND
Fig l(b). Scatchard representation of ljs-heparin binding to neutrophils. Linear regression analysis of data using the least square method gave the following equation: Y = -0~000816X+O~O377 (r=0.992).Kd (dissociation constant)= 1.22 pmot/t), n=7.7 x sites per cell. (n=4 determinations).
inn
I
20
I
A
Fig 3. Specificity of the binding of heparin on neutrophils. Neutrophils were incubated with PBS/BSA for 1 h at 4°C and '%-heparin (final concentration 5.2 p ~was ) added. Glycosaminoglycans were also added with radioactive heparin at molar ratios of 10.20 and 30. Glycosaminoglycans tested: (O), heparin Roche: ( A )heparin Sigma; (a),Fraxiparine:(A),dermatan-sulphate. (8). chondroitin-sulphate. Each point was the mean of double separate determinations. 'jSheparin binding without glycosaminoglycans was considered as 100% binding. Table I. Internalization of surface-bound heparin by neutrophils
I
Incubation
Trypsin + unlabelled heparin
l h
-
3h
-
+
+
Heparin bound (ng/lO" cells) 4°C
37OC
183.75 5.65
170.95 34.82
187.03 14.28
141.62 38.28
60
TIME (min)
Fig 2. Time-dependence and reversibility of '%heparin binding to neutrophils. Cells were incubated in PBS/BSA at 4°C for 1 h. "Sheparin (final concentration: 5.2 PM) was added.Cell-bound radioactivities were measured (curve A). Unlabelled heparin (arrow) was added 30 min after the start of incubation with "S-heparin (curveB).
dent and saturable manner (Fig l a ) . Non-specific binding, as determined by binding in the excess of unlabelled heparin, was < 2070, except for high concentrations of %-heparin. near the saturation. Scatchard analysis was performed on binding data. On n =4 determinations, the represented linear regression equation (Fig 1b) obtained by least square analysis of data was: Y = - 0 4 0 0 8 1 6 X + 0 . 0 3 7 7 , r = 0 . 9 9 2 . The dissociation constant ( K d ) was 1 . 2 2 prnol/l and there were 7.7 x 1Oh binding sites per cell (Fig 1b). Binding of heparin to neutrophils was found to be time-dependent. Approximately 90% of the surface bound labelled heparin was displaced by excess of unlabelled ligand (Fig 2 ) . The specificity of binding for heparin was evaluated in the presence of a n excess of unlabelled heparin (heparin Sigma and heparin Roche).
Neutrophils were incubated in PBS/BSA buffer for 1 h at 4°C and 37°C. Binding of '%-heparin ( 5 . 2 pnol/l. final concentration) was performed at these two temperatures during 1 h or 3 h. When the incubation time was achieved. cell suspensions were washed with PBS/BSA and treated by trypsin and unlabelled heparin at 4 T during 30 min. Finally, cell-bound heparin was then determined (n= 4 determinations).
dermatan sulphate, chondroitin-4-sulphate. and low molecular weight heparin (Fraxiparine) (Fig 3 ) . Unlabelled heparin inhibited the binding of j5S-heparin, whereas no inhibition was observed with the other glycosaminoglycans even at large molar excess.
lriternalization a/ bound heparin by neutrophils Internalization of bound heparin by neutrophils was carried out a t 4°C and 37°C (Table I). At 4OC. < 10% of heparin remained cell-bound after the addition of unlabelled heparin followed by trypsin treatment, suggesting minimal internalization. At 3 7"C, 20% and 2 7% were internalized at 1h and 3 h. respectively.
-
84
-
C. Leculier et al Table 11. Binding of heparin fractions to neutrophils
1
2
3
Heparin bound (ng/10"cells) Unfractionated heparin High affinity heparin Low affinity heparin
35 . 3 33.1 18.9
0-130 kDa
3SS-heparin was separated on a AT-I11 column in high and low affinity fractions. Heparin binding was carried out as described in Fig 1. Data presented are the mean of two experiments.
Table 111. Effect of heparin bound to neutrophils on AT 111 activation of thrombin Reaction mixture
Absorbance 405 nm
Thrombin Thrombin + AT 111 Thrombin + AT 111 +heparin Thrombin +AT 111 + PMNL Thrombin +AT I11 + 0 . 1 PM heparin/PMNL Thrombin+AT H I + 1 P M heparin/PMNL Thrombin+ AT 111 + 1 0 I(M heparin/PMNI, Thrombin + AT 111+ 100 PM heparin/PMNL Thrombin + AT 111+ 1000 p~ heparin/PMNL Thrombin + AT 111 + heparin/RBC
0.438 0.471 0.09 3 0.436 0.438 0.527 0-471 0.469 0.443 0.436
Binding of heparin fractions to neutrophils In order to determine whether heparin binding to neutrophils was related to its anticoagulant property, 35S-heparinwas fractionated into a low and high affinity fractions on an AT 111 column. The low and high affinity fractions gave significant binding, indicating that heparin binding to neutrophils was independent of the heparin binding site on AT I11 (Table 11). Binding of heparin in the presence of EDTA Binding of heparin was inhibited in the presence of EDTA: 38%. 52%. 65% and 94% of heparin remaining bound at concentrationsof 1 O - I q lO-'and 10-9~,respectively. Y3-heparin binding curve in the presence of 10- M EDTA is shown in Fig l ( a ) (curve D). Eflect of bound heparin on AT ZZZ activation of thrombin The effect of heparin (unfractionated heparin from Sigma) bound to neutrophils on AT 111 activation of thrombin is indicated in Table 111,showing that cell-bound heparin lost its anticoagulant activity. It was checked, using radioactive labelled heparin, that cell-bound heparin was not released in the medium. Characterization of heparin-binding proteins on neutrophil cell surface Neutrophil cell lysates after cell surface radioiodination and lysis were applied to a heparin-Sepharose affinity column.
+ Fig 4. Heparin binding proteins on cell-surface "'1 radiolabelled neutrophils. Neutrophils were radiolabelled using Enzymobeadm technique. Detergent lysates were prepared and applied to a heparinSepharose column in the presence and the absence of excess heparin. Heparin binding protein was eluted with NaCl after extensive washings. Finally, eluates were analysed by SDS-PACE and autoradiography. Lane 1 : cell-surface l L 5 1 radiolabelled unfractionated lysate. Lane 2: heparin column eluate in the presence of heparin in excess. Lane 3: heparin column eluate in the absence of heparin.
SDS-PAGE analysis of eluate showed a major protein band with a molecular weight of 1 30 kDa (Fig4).PI-PLC treatment of neutrophils before cell lysis induced no change in molecular weight of this protein (data not shown). DISCUSSION Heparin has been found to bind to various non-haemopoietic cells, including human endothelial cells (Glimelius et al, 1978). rat liver cells (Kjellen et al. 1977) and smooth muscle cells (Castellot et al. 1985). Heparin binding on mouse peritoneal macrophages has also been reported (Bleiberget al, 1983). Heparin can also bind to human platelets (Horne & Hart, 1986: Gogstad et al, 1983) and murine splenocytes (Bradbury & Parish, 1991). In a recent report on heparin binding to human monocytic cell line U93 7 ( h u n g et al, 1989), binding was found specific, rapid, saturable and reversible with a minimal internalization at 4°C and an internalization of 18%and 28% at 1 h and 3 h at 37OC,
-
Binding between Human Neutrophils and Heparin respectively. Heparin binding to U 9 3 7 cells was independent of its affinity for AT I11 and cell surface-bound heparin was found as functionally active. A I 2 0 kDa binding protein was identified. In our study, heparin binding to PMNL was also found saturable. time-dependent, specific and reversible. Low molecular weight heparin did not displace high molecular weight heparin from its specific sites. Scatchard analysis also showed a single class of heparin binding sites. The values of K d and n and the molecular weight of heparin binding protein were in the same range as found for U937 cells. The use of unlabelled heparin followed by trypsin treatment was justified by the high level of non-specific radioactive background obtained with cold heparin only a t 37°C. This finding strongly suggested important internalization by PMNL at 37°C (Table I). By contrast with results obtained in 11937 cells. heparin fixed on PMNL was found to be functionally inactive. The possibility of inactivation of heparin function by the fixative agent (glutaraldehyde) has been ruled out by similar results with non-fixed cells. The binding was inhibited by EDTA: this observation suggests that the binding of heparin to PMNL is probably calcium dependent. The impairment of the functional properties of PMNL surface bound-heparin could be associated with a masking of the heparin functional site by its binding protein. In human endothelial cells. heparin was internalized in endocytotic vesicles (Barau (it nl. 1987). These vesicles fused with primary granules (Barzu rt a/. 1987). I’MNI, are known to contain glycosaminoglycans (GAGs), particularly chondroitin sulphate and heparin sulphate (Parmley ct ul, 1983).The role of these GAGs could be to inactivate some enzymes containing within lysosomes and facilitate granule storage (Parmley er ul, 1 9 8 3; Avila. 1978: Avila & Convit, 1976). Heparin was also found to stimulate macrophages with increasing secretion of a mononuclear cell-factor (Yoffe et al, 1985). interferon (Schultz et a/. 1 9 7 7 ) and the secretion of lysosomal enzymes (Schorlemmer et a / . 1977). In PMNL, heparin was shown to potentiate lactoferrin secretion induced by FMLP (Cairo rt a/, 1 9 8 3). On monocytes. heparin induced protein secretion of 1 7 kDa and 1hO kDa, proteins of yet unknown function (Leung et al, 1989). We demonstrate that heparin interacts with PMNL by a cell-surface binding protein. The biological significance of these interactions remains to be investigated. These interactions could be consistent with the modification of some PMNL functional properties in the presence of heparin (Laghi-Pasini ct nl, 198 3 , 19x4; Cairo et al. 1 9 8 3 ) .
A CK N 0 W IXDG M E N TS This work was supported by ‘La Ligue Nationale Franqaise contre le Cancer. Comites Departementaux de la HauteSavoie et de la SaBne-et-Loire’. KEFEKENCES Avila. 1.L. ( 1 9 7 X ) The influence of the type of sulphate bond and degrce of sulphation of glycosaminoglycans on their interaction with lysosomal enzymes. Bioc.Jtrmica/ Journal. 171, 489-491. Avila. 1.L. & Convit. 1. (1976)Physicochemical characteristics of the
85
gtycosaminogtycan-tysosomal enzyme interaction in vitro. A model of control of leucocytic lysosornal activity. Biochemical Journal. 160, 129-1 36. Barzu.T.. Rijn, J.L.M.L.. Petitou. M.. Tobelem, G. & Caen. J.P. (1987) Heparin degradation in the endothelial cells. Thrombosis Research. 47, 601-609.
Bleiberg, I.. MacGregor, I. & Aronson. M. (1983) Heparin receptors on mouse macrophages. Thrombosis Research, 29, 53-61. Boyum. A. (1968) Isolation of mononuclear cells and granulocytes from human blood. Scandinavian Iournal of Clinical and Laboratory Investigation. 2 1 , (Suppl. 97). 77-89. Bradbury. M.G. & Parish. C.R. (199 1 ) Characterization of lymphocytereceptorsfor glycosaminoglycans.Immunology, 72,231 -238. Cairo. M.S.. Allen. J.. Higgins, C.. Baehner. K.L. &Boxer,L.A. (1983) Synergistic effect of heparin and chemotactic factor on polymorphonuclear leukocyte aggregation and degranulation. American Journal of Pathology. 1 1 3 , 67-74. Castellot. J.J.. Wong. K.. Herman, B.. Hoover, R.I.. Albertini. D.F.. Wright. T.C.. Caleb. B.L. & Karnovsky. M.J. (1985) Binding and internalization of heparin by vascular smooth muscle cells. Journal of Cellular Physiology. 124, 1 3-20. Glimelius. B.. Busch. C. & Hook. M. (1978) Binding ofheparin on the surface of cultured human endothelial cells. Thrombosis Research, 12. 773-782.
Cogstad,G.O.. Solum. N.O. & Krutnes. M.B. ( 198 3) Heparin-binding platelet proteins demonstrated by crossed affinity immuno-electrophoresis. British Journal of Haematology. 5 3 . 563-573. Horne. M.K. & Hart. J.S. ( 1 986) Heparin binding to platelets. Blood, 68, (Suppl. 1 ) . 31 7a. Kiellen. L.. Oldberg. A.. Rubin. K. L Hook. M. (1977) Binding of heparin and heparan sulfate to rat liver cells. Biochemical and Biophysical Research Communications. 74, 126-1 33. Laemmli. U.K. ( 1 970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature, 227,680-685. Laghi-Pasini,F.. Pasqui. A.L.. Ceccatelli. L.. Capecchi. P.L.. Orrico. A. & Di Perri, T. ( I 984) Heparin inhibition of polymorphonuclear leukocyte activation in vitro. A possible pharmacological approach to granulocyte-mediated vascular damage. Thrombosis Research. 3 5 , 527-537. Laghi-Pasini. F.. Pasqui. A.L.. Ceccatelli. I,. & Di Perri, T. (1983) Heparin inhibits FMLP and Con A dependent activation of human polymorphonuclear leukocytes in vitro. International journal of Tissue Reactants. 5 , 1 4 5 - 1 5 1 . Leung. L.. Saigo. K. & Grant, D. ( 1 9 8 9 ) Heparin binds to human monocytes and modulates their procoagulant activities and secretory phenotypes. Effect of histidine-rich-glycoprotein. Blood. 73, 177-184.
Parmley. R.T.. Hurst. R.E.. Takagi. M.. Spicer. S.S. & Austin, R.L. ( 1 98 3) Glycosaminoglycans in human neutrophils and leukemic myeloblasts: ultastructural. cytochemical. immunologic and biochemical characterization. Blood. 61, 257-266. Schorlemmer. H.U.. Burger, R.. Hylton, W. & Allison, A.C. (1977) Induction of lysosomal enzyme release from cultured macrophage by dextran sulfate. Clinical Immunology and Immunopathology. 7, 88-96.
Schultz. R.M.. Papamatheakis. J.D.& Chirigos. M.A. (1977) Interferon: an inducer of macrophage activation by polyanions.Science. 197,674-676.
Yoffe.I.R., Taylor,D.J.& Woolley. D.E. ( 1 985) Mast-cell products and heparin stimulate the production of mononuclear-cell factor by cultured human monocyte/macrophages. Biochemical journal. 230, 83-88.
Specific binding between human neutrophils and heparin C. LECULIEK, 0. BENZERARA,* N. C O U P R I E . A . FRANCINA,Y. LASNE,*E. A R C H I M B A U D t A N D
D. F I E R E t
Laboratoire de Biochimie B, Hbpital Edouard Herriot. 6 9 4 3 7 Lyon Cedex 03, *Laboratoire Central des Radio-lsotopes, Hbpital Edouard Herriot, 6 94 37 Lyon Cedex 03, and ?Service d'Hkmatologie Clinique, Hbpital Edouard Herriot, 6 9 4 3 7 Lyon Cedex 03, France Received 17 October 1 99 I , accepted for publication 3 January 1992
Summary. Heparin binding on polymorphonuclear leucocytes (PMNL) was characterized. Heparin binding was specific. rapid, saturable and reversible. One single class of heparin binding sites was found with a dissociation constant of 1.22 pmol/l and 7 . 7 x 10h sites per PMNL. The binding was independent of the anticoagulant activity of heparin. Heparin affinity chromatography on radio-jodinated cell lysates followed by polyacrylamide gel electrophoresis in the presence of sodium dodecylsulphate revealed a 130 kD heparin
binding protein. Heparin binding was inhibited by disodium salt of ethylenediamine tetraacetic acid. Cell surface bound heparin was functionally inactive and did not affect the inactivation of thrombin by antithrombin 111. Our study demonstrates that heparin interacts with PMNL by a cellsurface binding protein. These instructions could be consistent with the modifications of some PMNL functional properties in the presence of heparin.
Heparin affects the oxidative metabolism and degranulation of polymorphonuclear leucocytes (PMNL) (Laghi-Pasinirt ul, 1983. 1984). Heparin increases PMNI, adherence in vitro and potentiates the effect ofchemotactic factor formylmethionylleucyl-phenylalanine (FMLP) (Cairo ct ul. 198 3 ) . No specific binding of heparin to PMNL has been reported and identifed so far. In the present study we have characterized the specific binding of heparin to PMNI,. The heparin binding was found to be time-dependent, saturable, specific and reversible. The binding was inhibited by the bivalent metal chelator ethylenediamine tetraacetic acid (disodium salt) (EDTA). Neutrophil cell surface-bound heparin was demonstrated as functionally inactive. The protein interacting with heparin was revealed by cell surface radioiodination and heparin affinity chromatography, followed by polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulphate (PACK-SDS) on chromatographic fractions and autoradiograph y.
rombin 111 were purchased from Boehringer Mannheim (Meylan, France). Thrombin was purchased from Sigma (Saint-Louis. Mo.. U.S.A.) and thrombin substrate (S 2238) from Kabi (Stockholm, Sweden). Heparin-Sepharose CL-6B was purchased from Pharmacia (Saint-Quentin-en Yvelines, France). All reagents and buffers used in the present study were free of edotoxin. Gl!~c.osuririnogl!~c.uns. [N-s~lphonate-'~S]heparin was purchased from Amersham. The specific activity was 1.48 GBq/g and the molecular weight (MW) I6 000. Sodium heparin (grade I). 170 LISP units/mg (United States Pharmacopoeia), with a MW of 16 000 and chondroitin-4-sulphate (sodium salt) with a MW of 50000 were purchased from Sigma. Dermatan sulphate (sodium salt) with a MW of 14 500 was purchased from Calbiochem (San Diego, Calif., U.S.A.). Heparin (sodium salt, MW 1 7 500, anticoagulant activity: 5000 (IU) International Units/ml) was purchased from Roche (Neuilly, France) and Fraxiparine (calcium salt, MW 4500. anticoagulant activity 10 250 anti-Xa IU/ml) from Choay Institute (Paris. France). Humun nmtrophils. Whole blood from healthy donors was collected on ACD (citric acid-sodium citrate-dextrose). Human peripheral blood neutrophils were isolated by dextran sedimentation followed by centrifugation on FicollHypaque according to published techniques (Boyum. 1968). Remaining red blood cells were lysed by hypotonic treatment (0.2%NaCI). Purity of PMNL preparation was >95% as judged by Wright-Giemsa staining. PMNL suspensions were
MATERIALS AND METHODS Reagvnfs. Carrier free ['2iI]sodium iodide was purchased from Amersham (Les tllis. France). EDTA. 3 - [ ( 3-cholamidopropy1)-dimethylammonio]-1-propane sulphonate (CHAPS), phosphatidylinositol phospholipase C (PI-PLC) and antithCorrespondence: Dr A. Francina. Laboratoire de Biochimie B. HBpital Edouard Herriot. 69437 Lyon Cedex 0 3 . France.
81
82
C. Leculier et a1
determined to be free of endotoxin, using the Limulus test (Amilabo. Paris, France). Heparin binding to human neutrophils. The procedure was identical to that described elsewhere (Leung et al, 1989). Briefly, neutrophils (7 x 10h/ml) were incubated in phosphate-buffered saline, pH 7.40, containing 2% bovine serum albumin (PBS/BSA), for 1 h at 4OC, 3SS-heparin. in the presence or absence of 100-fold molar excess of unlabelled heparin, was added to cell samples and incubated for 1 h at 4' C. Final cell concentration was 3.6 x 1Ohce1k/mlin a final volume of 300 pl. Cell separation was performed immediately after incubation using a dot-blot apparatus (Biorad, Richmond, Calif., U.S.A.) in which a nitrocellulose membrane was inserted (pore size 0.45 pm;Biorad). 300 pI of cell suspension were poured in each well. The membrane was carefully cut after repeated washings in PBS. The respective pieces of membrane were placed into scintillation tubes containing Instagel@(Packard. Pans, France). The cell-bound radioactivity was counted in a Packard beta-counter. Each measurement was the mean of two assays. lnternalization of bound heparin by ncutrophils. The incubations were carried out as described for heparin binding to monocytes (Leung et a!, 1989). Binding of 35S-heparin(5.2 p~ in final concentration) was assayed at two different temperatures ( 4 O C and 37OC) with incubation times of 1 h and 3 h. Following incubation, neutrophils were washed in cold PBS. Cells were then treated with trypsin (0.25% for 30 min) after the addition of a 100-fold molar excess of unlabelled heparin. Fractionation of 35S-heparin by AT Ill affinity chromatography. 1 g of AT 111-agarose (Sigma) was suspended in PBS pH 7.40. An aliquot of 3SS-heparin(5.20 p ~corresponding ) to about 10% of column capacity of fixation in heparin, was applied on an AT 111 column equilibrated in PBS. Elution was carried out with PBS, followed by PBS containing 2 M NaCI. The corresponding fractions were measured by radioactive counting (low affinity and high affiity heparin fractions), separately pooled. dialysed and lyophilized. Heparin binding to human neutrophils experiments were performed with unfractionated heparin, low affinity and high affinity fractions after adjusting the respective radioactivities. Binding of heparin in the presence of EDTA. Heparin binding to human neutrophils was studied in the presence of EDTA at final concentrations of lo-' and M. ISSheparin final concentration was 1.3 p ~ Results . were expressed in per cent of controls without EDTA. '%-heparin binding curve was also assayed in the presence of EDTA at final concentration of lo-' M corresponding to maximal effect of EDTA. Heparin affinity chromatography of surface iodinated neutrophi/ cell lysates. Neutrophils (2 x 1OY/mlin 0.2 M sodium phosphate. pH 7.20) were surface radioiodinated using Enzymobeads method (Biorad) according to the instructions of the manufacturer and using a total activity of Na12SIof 74 MBq. PI-PLC treatment of neutrophils was performed before cell lysis with 20 mU PI-PLC/107 cells in PBS, pH 7.40. at 3 7°C for 45 min. The labelled cells were washed three times in PBS and solubilizedby 15 mmol/l CHAPS with PMSF 2 m~ and aprotinin 3 p~ in PBS for 1 h at 4°C. Insoluble materials
were removed by centrifugation and cell lysates were applied on a heparin-Sepharose 6B column, equilibrated in PBS. and after extensive washings, eluted with NaCl solutions (0.152 M). The eluates were concentrated and analysed by PAGESDS (Laemmli, 1970) and autoradiography. Control experiments were performed after addition of unlabelled heparin in excess into PMNL lysates. that were incubated overnight at 4 O C on heparin-Sepharose column and eluted by NaCl
solutions. Eflect of heparin bound to neutrophils on inactivation of thrombin. The functional role of neutrophil-bound heparin was investigated as previously described for the effect of heparin-pretreated U93 7 cells on thrombin inactivation by AT I11 (Leung et al, 1989). Briefly, PMNL and control red blood cells (RBC) (1.2 x 107/ml)were incubated with RPMI/ BSA. Cells were washed with PBS/BSA and fixed with glutaraldehyde. Thrombin (final concentrations 10 U/ml) and AT 111(final concentration, 10 pg/ml) in 40 mmol/l TrisHCI, 0.15 M NaCI, pH 8.30, were added to the fixed cells preincubated or not with heparin at concentrations of 0.11000 p ~ After . 30 s of incubation, thrombin substrate S 22 38 (final concentration, 0.5 mmol/l) was added. After 30 s, acetic acid (final concentration 15.6%)was added to stop the reaction. The reaction mixtures were centrifuged and the absorbance of the supernatants read at 405 nm. Each experiment was repeated three times with duplicate measurement of each point. RESULTS Heparin binding to human neutrophils 35S-heparinbound to neutrophils in a concentration-depenHEPARI! (ng/lO
BOUND cells)
I 250
200
150
100
50
0 0
1
2
3 4 [HEPARIN] (PM)
5
6
Fig ](a). Heparin binding to neutrophils. Neutrophils ( 3 . 6 x loh, h a 1 volume: 300 pl) were incubated for 1 h in PBS/BSA. at 4% "Sheparin was added in the presence or in the absence of unlabelled heparin for 1 h at 4°C. The cell suspensions were filtered through nitrocellulose membranes and cell-bound radioactivities were measured. Each point represents the mean of duplicate determinations. A: Typical total %-heparin binding to neutrophils.B Non-specific lSSheparin binding. C: Total 35S-heparinbinding minus non-specific %-heparin binding. D Total 3sS-heparinbinding minus non-specific "S-heparin binding in the presence of EDTA (lo-' M).
Binding between Human Neutrophils and Heparin B/AT
83
HEPARIN BOUND (percent)
0.041 120 100
I
0.03 80
0.02
60
40 0.01 20
1
0.00
0
20
10
30
40
o !
50
0
B (m)
-
t
0
o
I
1
20
40
I
40
MOLAR EXCESS OF UNLABELED LIGAND
Fig l(b). Scatchard representation of ljs-heparin binding to neutrophils. Linear regression analysis of data using the least square method gave the following equation: Y = -0~000816X+O~O377 (r=0.992).Kd (dissociation constant)= 1.22 pmot/t), n=7.7 x sites per cell. (n=4 determinations).
inn
I
20
I
A
Fig 3. Specificity of the binding of heparin on neutrophils. Neutrophils were incubated with PBS/BSA for 1 h at 4°C and '%-heparin (final concentration 5.2 p ~was ) added. Glycosaminoglycans were also added with radioactive heparin at molar ratios of 10.20 and 30. Glycosaminoglycans tested: (O), heparin Roche: ( A )heparin Sigma; (a),Fraxiparine:(A),dermatan-sulphate. (8). chondroitin-sulphate. Each point was the mean of double separate determinations. 'jSheparin binding without glycosaminoglycans was considered as 100% binding. Table I. Internalization of surface-bound heparin by neutrophils
I
Incubation
Trypsin + unlabelled heparin
l h
-
3h
-
+
+
Heparin bound (ng/lO" cells) 4°C
37OC
183.75 5.65
170.95 34.82
187.03 14.28
141.62 38.28
60
TIME (min)
Fig 2. Time-dependence and reversibility of '%heparin binding to neutrophils. Cells were incubated in PBS/BSA at 4°C for 1 h. "Sheparin (final concentration: 5.2 PM) was added.Cell-bound radioactivities were measured (curve A). Unlabelled heparin (arrow) was added 30 min after the start of incubation with "S-heparin (curveB).
dent and saturable manner (Fig l a ) . Non-specific binding, as determined by binding in the excess of unlabelled heparin, was < 2070, except for high concentrations of %-heparin. near the saturation. Scatchard analysis was performed on binding data. On n =4 determinations, the represented linear regression equation (Fig 1b) obtained by least square analysis of data was: Y = - 0 4 0 0 8 1 6 X + 0 . 0 3 7 7 , r = 0 . 9 9 2 . The dissociation constant ( K d ) was 1 . 2 2 prnol/l and there were 7.7 x 1Oh binding sites per cell (Fig 1b). Binding of heparin to neutrophils was found to be time-dependent. Approximately 90% of the surface bound labelled heparin was displaced by excess of unlabelled ligand (Fig 2 ) . The specificity of binding for heparin was evaluated in the presence of a n excess of unlabelled heparin (heparin Sigma and heparin Roche).
Neutrophils were incubated in PBS/BSA buffer for 1 h at 4°C and 37°C. Binding of '%-heparin ( 5 . 2 pnol/l. final concentration) was performed at these two temperatures during 1 h or 3 h. When the incubation time was achieved. cell suspensions were washed with PBS/BSA and treated by trypsin and unlabelled heparin at 4 T during 30 min. Finally, cell-bound heparin was then determined (n= 4 determinations).
dermatan sulphate, chondroitin-4-sulphate. and low molecular weight heparin (Fraxiparine) (Fig 3 ) . Unlabelled heparin inhibited the binding of j5S-heparin, whereas no inhibition was observed with the other glycosaminoglycans even at large molar excess.
lriternalization a/ bound heparin by neutrophils Internalization of bound heparin by neutrophils was carried out a t 4°C and 37°C (Table I). At 4OC. < 10% of heparin remained cell-bound after the addition of unlabelled heparin followed by trypsin treatment, suggesting minimal internalization. At 3 7"C, 20% and 2 7% were internalized at 1h and 3 h. respectively.
-
84
-
C. Leculier et al Table 11. Binding of heparin fractions to neutrophils
1
2
3
Heparin bound (ng/10"cells) Unfractionated heparin High affinity heparin Low affinity heparin
35 . 3 33.1 18.9
0-130 kDa
3SS-heparin was separated on a AT-I11 column in high and low affinity fractions. Heparin binding was carried out as described in Fig 1. Data presented are the mean of two experiments.
Table 111. Effect of heparin bound to neutrophils on AT 111 activation of thrombin Reaction mixture
Absorbance 405 nm
Thrombin Thrombin + AT 111 Thrombin + AT 111 +heparin Thrombin +AT 111 + PMNL Thrombin +AT I11 + 0 . 1 PM heparin/PMNL Thrombin+AT H I + 1 P M heparin/PMNL Thrombin+ AT 111 + 1 0 I(M heparin/PMNI, Thrombin + AT 111+ 100 PM heparin/PMNL Thrombin + AT 111+ 1000 p~ heparin/PMNL Thrombin + AT 111 + heparin/RBC
0.438 0.471 0.09 3 0.436 0.438 0.527 0-471 0.469 0.443 0.436
Binding of heparin fractions to neutrophils In order to determine whether heparin binding to neutrophils was related to its anticoagulant property, 35S-heparinwas fractionated into a low and high affinity fractions on an AT 111 column. The low and high affinity fractions gave significant binding, indicating that heparin binding to neutrophils was independent of the heparin binding site on AT I11 (Table 11). Binding of heparin in the presence of EDTA Binding of heparin was inhibited in the presence of EDTA: 38%. 52%. 65% and 94% of heparin remaining bound at concentrationsof 1 O - I q lO-'and 10-9~,respectively. Y3-heparin binding curve in the presence of 10- M EDTA is shown in Fig l ( a ) (curve D). Eflect of bound heparin on AT ZZZ activation of thrombin The effect of heparin (unfractionated heparin from Sigma) bound to neutrophils on AT 111 activation of thrombin is indicated in Table 111,showing that cell-bound heparin lost its anticoagulant activity. It was checked, using radioactive labelled heparin, that cell-bound heparin was not released in the medium. Characterization of heparin-binding proteins on neutrophil cell surface Neutrophil cell lysates after cell surface radioiodination and lysis were applied to a heparin-Sepharose affinity column.
+ Fig 4. Heparin binding proteins on cell-surface "'1 radiolabelled neutrophils. Neutrophils were radiolabelled using Enzymobeadm technique. Detergent lysates were prepared and applied to a heparinSepharose column in the presence and the absence of excess heparin. Heparin binding protein was eluted with NaCl after extensive washings. Finally, eluates were analysed by SDS-PACE and autoradiography. Lane 1 : cell-surface l L 5 1 radiolabelled unfractionated lysate. Lane 2: heparin column eluate in the presence of heparin in excess. Lane 3: heparin column eluate in the absence of heparin.
SDS-PAGE analysis of eluate showed a major protein band with a molecular weight of 1 30 kDa (Fig4).PI-PLC treatment of neutrophils before cell lysis induced no change in molecular weight of this protein (data not shown). DISCUSSION Heparin has been found to bind to various non-haemopoietic cells, including human endothelial cells (Glimelius et al, 1978). rat liver cells (Kjellen et al. 1977) and smooth muscle cells (Castellot et al. 1985). Heparin binding on mouse peritoneal macrophages has also been reported (Bleiberget al, 1983). Heparin can also bind to human platelets (Horne & Hart, 1986: Gogstad et al, 1983) and murine splenocytes (Bradbury & Parish, 1991). In a recent report on heparin binding to human monocytic cell line U93 7 ( h u n g et al, 1989), binding was found specific, rapid, saturable and reversible with a minimal internalization at 4°C and an internalization of 18%and 28% at 1 h and 3 h at 37OC,
-
Binding between Human Neutrophils and Heparin respectively. Heparin binding to U 9 3 7 cells was independent of its affinity for AT I11 and cell surface-bound heparin was found as functionally active. A I 2 0 kDa binding protein was identified. In our study, heparin binding to PMNL was also found saturable. time-dependent, specific and reversible. Low molecular weight heparin did not displace high molecular weight heparin from its specific sites. Scatchard analysis also showed a single class of heparin binding sites. The values of K d and n and the molecular weight of heparin binding protein were in the same range as found for U937 cells. The use of unlabelled heparin followed by trypsin treatment was justified by the high level of non-specific radioactive background obtained with cold heparin only a t 37°C. This finding strongly suggested important internalization by PMNL at 37°C (Table I). By contrast with results obtained in 11937 cells. heparin fixed on PMNL was found to be functionally inactive. The possibility of inactivation of heparin function by the fixative agent (glutaraldehyde) has been ruled out by similar results with non-fixed cells. The binding was inhibited by EDTA: this observation suggests that the binding of heparin to PMNL is probably calcium dependent. The impairment of the functional properties of PMNL surface bound-heparin could be associated with a masking of the heparin functional site by its binding protein. In human endothelial cells. heparin was internalized in endocytotic vesicles (Barau (it nl. 1987). These vesicles fused with primary granules (Barzu rt a/. 1987). I’MNI, are known to contain glycosaminoglycans (GAGs), particularly chondroitin sulphate and heparin sulphate (Parmley ct ul, 1983).The role of these GAGs could be to inactivate some enzymes containing within lysosomes and facilitate granule storage (Parmley er ul, 1 9 8 3; Avila. 1978: Avila & Convit, 1976). Heparin was also found to stimulate macrophages with increasing secretion of a mononuclear cell-factor (Yoffe et al, 1985). interferon (Schultz et a/. 1 9 7 7 ) and the secretion of lysosomal enzymes (Schorlemmer et a / . 1977). In PMNL, heparin was shown to potentiate lactoferrin secretion induced by FMLP (Cairo rt a/, 1 9 8 3). On monocytes. heparin induced protein secretion of 1 7 kDa and 1hO kDa, proteins of yet unknown function (Leung et al, 1989). We demonstrate that heparin interacts with PMNL by a cell-surface binding protein. The biological significance of these interactions remains to be investigated. These interactions could be consistent with the modification of some PMNL functional properties in the presence of heparin (Laghi-Pasini ct nl, 198 3 , 19x4; Cairo et al. 1 9 8 3 ) .
A CK N 0 W IXDG M E N TS This work was supported by ‘La Ligue Nationale Franqaise contre le Cancer. Comites Departementaux de la HauteSavoie et de la SaBne-et-Loire’. KEFEKENCES Avila. 1.L. ( 1 9 7 X ) The influence of the type of sulphate bond and degrce of sulphation of glycosaminoglycans on their interaction with lysosomal enzymes. Bioc.Jtrmica/ Journal. 171, 489-491. Avila. 1.L. & Convit. 1. (1976)Physicochemical characteristics of the
85
gtycosaminogtycan-tysosomal enzyme interaction in vitro. A model of control of leucocytic lysosornal activity. Biochemical Journal. 160, 129-1 36. Barzu.T.. Rijn, J.L.M.L.. Petitou. M.. Tobelem, G. & Caen. J.P. (1987) Heparin degradation in the endothelial cells. Thrombosis Research. 47, 601-609.
Bleiberg, I.. MacGregor, I. & Aronson. M. (1983) Heparin receptors on mouse macrophages. Thrombosis Research, 29, 53-61. Boyum. A. (1968) Isolation of mononuclear cells and granulocytes from human blood. Scandinavian Iournal of Clinical and Laboratory Investigation. 2 1 , (Suppl. 97). 77-89. Bradbury. M.G. & Parish. C.R. (199 1 ) Characterization of lymphocytereceptorsfor glycosaminoglycans.Immunology, 72,231 -238. Cairo. M.S.. Allen. J.. Higgins, C.. Baehner. K.L. &Boxer,L.A. (1983) Synergistic effect of heparin and chemotactic factor on polymorphonuclear leukocyte aggregation and degranulation. American Journal of Pathology. 1 1 3 , 67-74. Castellot. J.J.. Wong. K.. Herman, B.. Hoover, R.I.. Albertini. D.F.. Wright. T.C.. Caleb. B.L. & Karnovsky. M.J. (1985) Binding and internalization of heparin by vascular smooth muscle cells. Journal of Cellular Physiology. 124, 1 3-20. Glimelius. B.. Busch. C. & Hook. M. (1978) Binding ofheparin on the surface of cultured human endothelial cells. Thrombosis Research, 12. 773-782.
Cogstad,G.O.. Solum. N.O. & Krutnes. M.B. ( 198 3) Heparin-binding platelet proteins demonstrated by crossed affinity immuno-electrophoresis. British Journal of Haematology. 5 3 . 563-573. Horne. M.K. & Hart. J.S. ( 1 986) Heparin binding to platelets. Blood, 68, (Suppl. 1 ) . 31 7a. Kiellen. L.. Oldberg. A.. Rubin. K. L Hook. M. (1977) Binding of heparin and heparan sulfate to rat liver cells. Biochemical and Biophysical Research Communications. 74, 126-1 33. Laemmli. U.K. ( 1 970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature, 227,680-685. Laghi-Pasini,F.. Pasqui. A.L.. Ceccatelli. L.. Capecchi. P.L.. Orrico. A. & Di Perri, T. ( I 984) Heparin inhibition of polymorphonuclear leukocyte activation in vitro. A possible pharmacological approach to granulocyte-mediated vascular damage. Thrombosis Research. 3 5 , 527-537. Laghi-Pasini. F.. Pasqui. A.L.. Ceccatelli. I,. & Di Perri, T. (1983) Heparin inhibits FMLP and Con A dependent activation of human polymorphonuclear leukocytes in vitro. International journal of Tissue Reactants. 5 , 1 4 5 - 1 5 1 . Leung. L.. Saigo. K. & Grant, D. ( 1 9 8 9 ) Heparin binds to human monocytes and modulates their procoagulant activities and secretory phenotypes. Effect of histidine-rich-glycoprotein. Blood. 73, 177-184.
Parmley. R.T.. Hurst. R.E.. Takagi. M.. Spicer. S.S. & Austin, R.L. ( 1 98 3) Glycosaminoglycans in human neutrophils and leukemic myeloblasts: ultastructural. cytochemical. immunologic and biochemical characterization. Blood. 61, 257-266. Schorlemmer. H.U.. Burger, R.. Hylton, W. & Allison, A.C. (1977) Induction of lysosomal enzyme release from cultured macrophage by dextran sulfate. Clinical Immunology and Immunopathology. 7, 88-96.
Schultz. R.M.. Papamatheakis. J.D.& Chirigos. M.A. (1977) Interferon: an inducer of macrophage activation by polyanions.Science. 197,674-676.
Yoffe.I.R., Taylor,D.J.& Woolley. D.E. ( 1 985) Mast-cell products and heparin stimulate the production of mononuclear-cell factor by cultured human monocyte/macrophages. Biochemical journal. 230, 83-88.
