Low molecular weight kininogen binds to platelets to modulate thrombin-induced platelet activation

The kininogens, high molecular weight kininogen (HK) and low molecular weight kininogen (LK), are multifunctional, single-gene products that contain bradykinin and identical amino-terminal heavy chains. Studies were performed to determine if LK would bind directly to platelets. 125I-LK specifically bound to gel-filtered platelets in the presence of 50 microM Zn2+. HK effectively competed with 125I-LK for the same binding site (Ki = 27 +/- 9 nM, n = 5). Similarly, the Ki for LK inhibition of 125I-LK binding was 12 +/- 1 nM (n = 3). Albumin, fibrinogen, factor XIII, and kallikrein did not inhibit 125I-LK binding to unstimulated platelets. 125I-LK (66 kDa) was not cleaved upon binding to platelets. The binding of 125I-LK to unstimulated platelets was found to be fully reversible by the addition of a 50 molar excess of unlabeled LK at both 10 and 20 min. LK binding to platelets was saturable with an apparent Kd of 27 +/- 2 nM (mean +/- S.E., n = 9) and 647 +/- 147 binding sites/platelet. Both LK and HK at plasma concentrations inhibited thrombin-induced platelet aggregation. LK and HK at about 5% of plasma concentration also inhibited thrombin-induced secretion of both stirred and unstirred platelets. Both kininogens were found to be noncompetitive inhibitors of proteolytically active thrombin binding to platelets. The kininogens did not inhibit D-phenylalanyl-prolyl-arginine chloromethyl ketone-treated thrombin from binding to platelets. These studies indicated that both kininogens have a region on their heavy chain which allows them to bind to platelets. Further, kininogen binding by its heavy chain modulates thrombin activation of platelets since it prevents proteolytically active thrombin from binding to its receptor.     

Properties of chemically oxidized kininogens

Kininogens are multifunctional proteins involved in a variety of regulatory processes including the kinin-formation cascade, blood coagulation, fibrynolysis, inhibition of cysteine proteinases etc. A working hypothesis of this work was that the properties of kininogens may be altered by oxidation of their methionine residues by reactive oxygen species that are released at the inflammatory foci during phagocytosis of pathogen particles by recruited neutrophil cells. Two methionine-specific oxidizing reagents, N-chlorosuccinimide (NCS) and chloramine-T (CT), were used to oxidize the high molecular mass (HK) and low molecular mass (LK) forms of human kininogen. A nearly complete conversion of methionine residues to methionine sulfoxide residues in the modified proteins was determined by amino acid analysis. Production of kinins from oxidized kininogens by plasma and tissue kallikreins was significantly lower (by at least 70%) than that from native kininogens. This quenching effect on kinin release could primarily be assigned to the modification of the critical Met-361 residue adjacent to the internal kinin sequence in kininogen. However, virtually no kinin could be formed by human plasma kallikrein from NCS-modified HK. This observation suggests involvement of other structural effects detrimental for kinin production. Indeed, NCS-oxidized HK was unable to bind (pre)kallikrein, probably due to the modification of methionine and/or tryptophan residues at the region on the kininogen molecule responsible for the (pro)enzyme binding. Tests on papain inhibition by native and oxidized kininogens indicated that the inhibitory activity of kininogens against cysteine proteinases is essentially insensitive to oxidation.     

Isolation and partial characterization of a novel form of low molecular weight kininogen from guinea pig plasma

Two kinds of low molecular weight kininogen (termed LK1 and LK2) were isolated from pooled plasma of guinea pigs. When polyclonal antisera raised against the individual proteins were used, immunological cross-reactions were observed between LK1 and high molecular weight kininogen (HK), but not either between LK1 and LK2 or between LK2 and HK. After tissue injury, plasma level of LK1 doubled while those of LK2 and HK remained relatively unchanged.     

Purification and characterization of two kinds of low molecular weight kininogens from rat (non-inflamed) plasma. One resistant and the second sensitive to rat glandular kallikreins

Two kinds of low molecular weight kininogens (identified as A and B) were isolated from pooled plasma of Sprague-Dawley rats. They show a single band on sodium dodecyl sulfate-polyacrylamide gel electrophoresis in the presence and absence of 2-mercaptoethanol, and the molecular weights are 68,000 for low Mr kininogen A and 73,000 for low Mr kininogen B. Although the molecular weights and amino acid compositions of the low Mr kininogens are similar, rat submaxillary and urinary kallikreins released bradykinin from low Mr kininogen B, whereas low Mr kininogen A was resistant to these enzymes. The COOH-terminal portion of low Mr kininogen A was isolated after cyanogen bromide treatment, and the amino acid sequence of the COOH-terminal 55 residues including the T-kinin (Ile-Ser-bradykinin) was determined. The COOH-terminal portion consists of two sequences with substitution of 4 residues. One peptide corresponds to alpha 1-major acute phase protein (Cole, T., Inglis, A. S., Roxburgh, C. M., Howlett, G. J., and Schreiber, G. (1985) FEBS Lett. 182, 57-61) and the other to the TI-kininogen predicted from a cDNA study (Furuto-Kato, S., Matsumoto, A., Kitamura, N., and Nakanishi, S. (1985) J. Biol. Chem. 260, 12054-12059). The results demonstrate that there exist at least two kinds of low Mr kininogens with clearly different function in rat plasma: one of them, low Mr kininogen A, is a precursor of T-kinin and is resistant to kallikreins, and the second, low Mr kininogen B, is sensitive to tissue kallikreins and shares properties with bovine and human low Mr kininogens. The results also demonstrate that T-kininogen is a mixture of two isoproteins which correspond to alpha 1-major acute phase protein or TI-kininogen, respectively. We could not detect the low Mr kininogen corresponding to the TII-kininogen predicted from the cDNA study of Furuto-Kato et al.     

Attenuated kinin release from human neutrophil elastase-pretreated kininogens by tissue and plasma kallikreins

Components of kinin-forming systems operating at inflammatory sites are likely to interact with elastase that is released by recruited neutrophils and may, at least temporarily, constitute the major proteolytic activity present at these sites. The aim of this work was to determine the effect of kininogen degradation by human neutrophil elastase (HNE) on kinin generation by tissue and plasma kallikreins. We show that the digestion of both low molecular mass (LK) and high molecular mass (HK) forms of human kininogen by HNE renders them essentially unsusceptible to processing by human urinary kallikrein (tissue-type) and also significantly quenches the kinin release from HK by plasma kallikrein. Studies with synthetic model heptadecapeptide substrates, ISLMKRPPGFSPFRSSR and SLMKRPPGFSPFRSSRI, confirmed the inability of tissue kallikrein to process peptides at either termini of the internal kinin sequence, while plasma kallikrein was shown to release the kinin C-terminus relatively easily. The HNE-generated fragments of kininogens were separated by HPLC and the fractions containing internal kinin sequences were identified by a kinin-specific immunoenzymatic test after trypsin digestion. These fractions were analyzed by electrospray-ionization mass spectrometry. In this way, multiple peptides containing the kinin sequence flanked by only a few amino acid residues at each terminus were identified in elastase digests of both LK and HK. These results suggest that elastase may be involved in quenching the kinin-release cascade at the late stages of the inflammatory reaction.     

Domain 3 of kininogens contains a cell-binding site and a site that modifies thrombin activation of platelets.

High and low molecular weight kininogens (HK and LK) are able to bind to platelets to inhibit thrombin binding to and activation of platelets. The heavy chain domain on the kininogens that contains these functions has been determined. Domain 3 (D3) but not domains 1 or 2, completely inhibited 125I-HK binding to platelets (Ki = 24 +/- 7 nM, n = 4). 125I-D3 specifically bound to unstimulated platelets and human umbilical vein endothelial cells. On platelets, it was blocked by unlabeled D3 and HK but not prekallikrein, factor XII, C1s, or C1 inhibitor. Further, one monoclonal antibody (HKH13) directed to kininogens' D3 blocked 125I-HK and 125I-D3 binding to platelets. The binding of 125I-D3 to platelets was fully reversible by addition of 35 molar excess of unlabeled D3. D3 binding to platelets was saturable with an apparent Kd of 39 +/- 8 nM (n = 4) and 1227 +/- 404 binding sites/platelet. D3, like HK and LK, inhibited thrombin-induced platelet activation by preventing thrombin binding to platelets. Another monoclonal antibody (HKH12), directed to D3, which did not block HK binding to platelets, reduced HK's ability to inhibit 125I-alpha-thrombin binding. This result suggests that the region on D3 that inhibits 125I-alpha-thrombin binding to platelets is different from that which directly binds to platelets. These studies indicate that D3 of the kininogens contains both a binding region for platelets and endothelial cells and another region that inhibits thrombin-induced platelet activation.     

Inhibition of cell adhesion by high molecular weight kininogen

An anti-cell adhesion globulin was purified from human plasma by heparin-affinity chromatography. The purified globulin inhibited spreading of osteosarcoma and melanoma cells on vitronectin, and of endothelial cells, platelets, and mononuclear blood cells on vitronectin or fibrinogen. It did not inhibit cell spreading on fibronectin. The protein had the strongest antiadhesive effect when preadsorbed onto the otherwise adhesive surfaces. Amino acid sequence analysis revealed that the globulin is cleaved (kinin-free) high molecular weight kininogen (HKa). Globulin fractions from normal plasma immunodepleted of high molecular weight kininogen (HK) or from an individual deficient of HK lacked adhesive activity. Uncleaved single-chain HK preadsorbed at neutral pH, HKa preadsorbed at pH greater than 8.0, and HKa degraded further to release its histidine-rich domain had little anti-adhesive activity. These results indicate that the cationic histidine-rich domain is critical for anti-adhesive activity and is somehow mobilized upon cleavage. Vitronectin was not displaced from the surface by HKa. Thus, cleavage of HK by kallikrein results in both release of bradykinin, a potent vasoactive and growth-promoting peptide, and formation of a potent anti-adhesive protein.     

Canine Plasma Kininogen: Evidence for the Presence of Two Kininogens and Purification of High Molecular Weight Kininogen and Characterization as Multi-Functional Molecule

The ratio of kininogen that is substrate of plasma kallikrein to kininogen, which is not substrate of plasma kallikrein in canine plasma, was about 1:3.6 by differential assay of kininogens. When the plasma was gel-filtered through a column of Sephacryl S-300 superfine, two fractions, which released kinin by trypsin, were obtained. These results indicate that two kininogens with different molecular weights are present in the plasma and they show different susceptibility to plasma kallikrein. One kininogen was purified by ion-exchange and zinc-chelating affinity chromatographies. Purified kininogen showed a single band in sodium dodecyl sulfate-polyacrylamide gel electrophoresis under reducing condition and its molecular weight was 125 kDa. Released kinin from the kininogen by trypsin was bradykinin. The kininogen inhibited papain and ficin but did not inhibit bromelain at the concentration used. The kininogen bound to carboxymethylated-papain and this binding was dissociated by 3M NaSCN. Canine plasma shortened the abnormal clotting time of human high molecular weight kininogen-deficint plasma. The kininogen also shortened the abnormal clotting time of the plasma. From these results, the purified kininogen was high molecular weight kininogen and it was multi-functional protein.     

The Occluding Loop of Cathepsin B Prevents Its Effective Inhibition by Human Kininogens

Kininogens, the major plasma cystatin-like inhibitors of cysteine cathepsins, are degraded at sites of inflammation, and cathepsin B has been identified as a prominent mediator of this process. Cathepsin B, in contrast to cathepsins L and S, is poorly inhibited by kininogens. This led us to delineate the molecular interactions between this protease and kininogens (high molecular weight kininogen and low molecular weight kininogen) and to elucidate the dual role of the occluding loop in this weak inhibition. Cathepsin B cleaves high molecular weight kininogen within the N-terminal region of the D2 and D3 cystatin-like domains and close to the consensus QVVAG inhibitory pentapeptide of the D3 domain. The His110Ala mutant, unlike His111Ala cathepsin B, fails to hydrolyze kininogens, but rather forms a tight-binding complex as observed by gel-filtration analysis. K_i values (picomolar range) as well as association rate constants for the His110Ala cathepsin B variant compare to those reported for cathepsin L for both kininogens. Homology modeling of isolated inhibitory (D2 and D3) domains and molecular dynamics simulations of the D2 domain complexed with wild-type cathepsin B and its mutants indicate that additional weak interactions, due to the lack of the salt bridge (Asp22-His110) and the subsequent open position of the occluding loop, increase the inhibitory potential of kininogens on His110Ala cathepsin B.     

T-kininogen--the major plasma kininogen in rat adjuvant arthritis

Total kininogen in plasma of Freund's adjuvant treated rats increased 20-fold 7 days following the injection. Analysis of the kininogens demonstrated that increases in T-kininogen was the major reason for the rise in kininogen. High molecular weight and low molecular weight kininogens showed little or no change. The increase in T-kininogen paralleled the inflammatory condition. Anti-inflammatory agents which reduced paw swelling also reduced plasma T-kininogen levels. Unidentified peaks on HPLC of kinin following plasma treatment by trypsin were shown to be oligopeptides containing T-kinin (Ile-serbradykinin). The relationship of T-kininogen to the inflammatory response is discussed.     

Local bradykinin formation is controlled by glycosaminoglycans

Bradykinin is a potent inflammatory mediator that induces vasodilation, vascular leakage, and pain sensations. This short-lived peptide hormone is liberated from its large precursor protein high molecular weight kininogen (HK) through the contact system cascade involving coagulation factor XII and plasma kallikrein. Although bradykinin release is well established in vitro, the factors and mechanisms controlling bradykinin generation in vivo are still incompletely understood. In this study we demonstrate that binding of HK to glycosaminoglycans (GAGs) of the heparan and chondroitin sulfate type efficiently interferes with bradykinin release in plasma and on endothelial surfaces. Proteolytic bradykinin production on endothelial cells is restored following degradation of cell surface GAG through heparinase. Alternatively, application of HK fragments D3 or light chain, which compete with uncleaved HK for cell binding, promote kininogen proteolysis and bradykinin release. Intravital microscopy revealed that HK fragments increase bradykinin-mediated mesentery microvascular leakage. Topical application of D3 or light chain enhanced bradykinin generation and edema formation in the mouse skin. Our results demonstrate that bradykinin formation is controlled by HK binding to and detachment from GAGs. Separation of the precursor from cell surfaces is a prerequisite for its efficient proteolytic processing. By this means, fragments arising from HK processing propagate bradykinin generation, revealing a novel regulatory level for the kallikrein-kinin system.     

The contact system--a novel branch of innate immunity generating antibacterial peptides

Activation of the contact system has two classical consequences: initiation of the intrinsic pathway of coagulation, and cleavage of high molecular weight kininogen (HK) leading to the release of bradykinin, a potent proinflammatory peptide. In human plasma, activation of the contact system at the surface of significant bacterial pathogens was found to result in further HK processing and bacterial killing. A fragment comprising the D3 domain of HK is generated, and within this fragment a sequence of 26 amino acids is mainly responsible for the antibacterial activity. A synthetic peptide covering this sequence kills several bacterial species, also at physiological salt concentration, as effectively as the classical human antibacterial peptide LL-37. Moreover, in an animal model of infection, inhibition of the contact system promotes bacterial dissemination and growth. These data identify a novel and important role for the contact system in the defence against invasive bacterial infection.     

Assembly of human contact phase proteins and release of bradykinin at the surface of curli-expressing Escherichia coli

Previous work has demonstrated that most strains of the human pathogen Streptococcus pyogenes bind kininogens through M protein, a fibrous surface protein and virulence determinant. Here we find that strains of several other pathogenic bacterial species, both Gram-positive and Gram-negative, isolated from patients with sepsis, also bind kininogens, especially H-kininogen (HK). The most pronounced interaction was seen between HK and Escherichia coli. Among clinical isolates of E. coli, the majority of the entero-haemorrhagic, enterotoxigenic, and sepsis strains, but none of the enteroinvasive and enteropathogenic strains, bound HK. Binding of HK to E. coli correlated with the expression of curli, another fibrous bacterial surface protein, and the binding of HK to purified curli was specific, saturable, and of high affinity; Ka = 9 107M-1. Other contact phase proteins such as factor XI, factor XII, and prekallikrein bound to curliated E. coli, but not to an isogenic curli-deficient mutant strain, suggesting that contact phase activation may occur at the surface of curliated bacteria. Kininogens are also precursor molecules of the vasoactive kinins. When incubated with human plasma, curli-expressing bacteria absorbed HK. Addition of purified plasma kallikrein to the HK-loaded bacteria resulted in a rapid and efficient release of bradykinin from surface-bound HK. The assembly of contact phase factors at the surface of pathogenic bacteria and the release of the potent proinflammatory and vasoactive peptide bradykinin, should have a major impact on the host-microbe relationship and may contribute to bacterial pathogencity and virulence.     

Studies on human high molecular weight (HMW) kininogen. II. Structural change of HMW kininogen by the action of human plasma kallikrein

We have investigated in detail the cleavage of human high molecular weight (HMW) kininogen by human plasma kallikrein and revealed the formation of a nicked kininogen and a novel kinin-free protein (KFP) as intermediate cleavage products. The cleavage of a single chain HMW kininogen (Mr=120,000) by plasma kallikrein was a three-step reaction. The first cleavage yielded a nicked kininogen composed of two disulfide-linked 62,000 and 56,000 daltons chains. The second cleavage yielded kinin and an intermediate kinin-free protein, KFP-I, which was apparently of equal size to the nicked kininogen. The third cleavage yielded a stable kinin-free protein, KFP-II, composed of two disulfide-linked 62,000 and 45,000 daltons chains. The liberation of an 8,000 daltons fragment was identified when the 56,000 daltons chain isolated by SP-Sephadex C-50 chromatography of reduced and alkylated KFP-I was cleaved by plasma kallikrein into the 45,000 daltons chain. Although the antiserum against HMW kininogen cross-reacted with low molecular weight (LMW) kininogen, the antiserum against the 45,000 daltons chain was specific for HMW kininogen. These results suggest that the antigenic determinant groups common to HMW and LMW kininogens are located in the 62,000 daltons heavy chain, while those specific for HMW kininogen are located in the 45,000 daltons light chain, which is known to retain blood coagulation activity.     

Purification and characterization of rat low molecular weight kininogen

Low molecular weight (LMW) kininogen was isolated from pooled rat plasma by chromatography on DEAE-Sephadex A-50, CM-Sephadex C-50, Blue-Sepharose CL-6B, and Sephadex G-100. It was shown to be homogeneous by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and immunoelectrophoresis. The molecular weight of rat LMW kininogen was determined to be 72,000 by SDS-PAGE. The LMW kininogen contained 83.5% protein, 4.0% hexose, 5.5% hexosamine, and 2.7% sialic acid. Kinin liberated from LMW kininogen by trypsin treatment was identified as an Ile-Ser-bradykinin(T-kinin) by analysis involving ion exchange column chromatography on CM-Sephadex C-25 and high performance liquid chromatography on a reverse-phase column (ODS-120T). LMW kininogen formed kinin with rat submaxillary gland kallikrein, but the kinin liberated was only 14% of the total kinin content, that is, that released by trypsin. In order to determine the immunochemical properties of LMW kininogen, specific antiserum was prepared in rabbits. The antiserum cross-reacted with high molecular weight (HMW) kininogen, but spur formation was observed between the LMW and HMW kininogens. The kininogen level in rat plasma was estimated to be 433 microgram/ml by a quantitative single radial immunodiffusion test.     

Release of a new vascular permeability enhancing peptide from kininogens by human neutrophil elastase

Stimulated neutrophils produced vascular permeability enhancing (VPE) activity in the presence of high molecular weight kininogen (HMWK), which was inhibited mainly by a neutrophil elastase (NE) inhibitor or a bradykinin (BK) B(2)-receptor antagonist. NE (>3 nM) generated VPE activity from kininogens at normal plasma concentrations with the smaller protein being several fold more responsive than the larger protein, through releasing a new VPE peptide (E-kinin), SLMKRPPGFSPFRSSRI. Synthetic E-kinin, SLMKRPPGFSPFRSS and SLMKRPPGFSPFR had VPE and blood pressure lowering activities, which were comparable to the activities of BK and completely inhibited by B(2)-receptor antagonists. Interestingly, E-kinin and SLMKRPPGFSPFRSS did not induce smooth muscle contraction. These results suggest that E-kinin formed in vivo may be processed at the carboxy-terminus to give a peptide that can bind to the B(2)-receptor. The molecular mechanism for neutrophil-associated VPE may be explained by excision of E-kinin from kininogens by NE, followed by further processing of the peptide.     

Differing expression patterns and evolution of the rat kininogen gene family

The present investigation using molecular cloning and sequence analysis concerns the examination of the molecular basis for different expression patterns of two types of the rat kininogen genes. We show that the low molecular weight and high molecular weight forms of K kininogens are produced from a single gene through alternative usage of two 3'-coding regions, whereas only the low molecular weight forms of T kininogens are generated as a result of several mutational changes in the high molecular weight-specifying regions of both T-I and T-II kininogen genes. The mutational changes include a nucleotide substitution at the polyadenylation/processing signal site, nucleotide deletions resulting in the frame-shift mutation, and an insertion of the type 2 Alu-equivalent sequence. Because kininogens represent a multifunctional protein comprising the proteinase-inhibitory activity, the kinin moiety, and the clotting activity, these results present evidence indicating the molecular basis for the disappearance of a part of the gene functions. We also show that the K and T kininogen genes as well as the two T kininogen genes are extremely homologous, excluding and including the above mutational changes, respectively. These structural relationships allow us to envisage evolutionary processes for the generation of the rat kininogen gene family, particularly for the disappearance of a part of the gene functions.     

Determination of the bifunctional properties of high molecular weight kininogen by studies with monoclonal antibodies directed to each of its chains

In normal human plasma two forms of kininogen exist, low molecular weight kininogen (LMWK) and high molecular weight kininogen (HMWK). When these proteins are cleaved they are found to have a common heavy chain and bradykinin, but each has a unique light chain. Monoclonal antibodies to the heavy and light chains of HMWK have been developed, and the effects of each on the function of this protein are defined. Initial studies showed that an antibody, C11C1, completely neutralized the coagulant activity of plasma HMWK whereas another antibody, 2B5, did not. On a competitive enzyme-linked immunosorbent assay (CELISA) the C11C1 antibody was consumed by kininogen antigen in normal plasma but not by kininogen antigen in HMWK-deficient plasma. On immunoblot, the C11C1 antibody recognized one kininogen protein in normal plasma and did not recognize any kininogen antigen in HMWK-deficient plasma. These combined studies indicated that the C11C1 antibody was directed to an epitope on the unique 46-kDa light chain of HMWK. In contrast, the 2B5 antibody on a CELISA was consumed by kininogen antigen in both normal plasma and HMWK-deficient plasma but not by total kininogen-deficient plasma. On immunoblot, the 2B5 antibody recognized both kininogens in normal plasma but only LMWK in HMWK-deficient plasma. These combined studies indicated that the 2B5 antibody was directed to the common 64-kDa heavy chain of the plasma kininogens. Utilizing direct binding studies or competition kinetic experiments, the 2B5 and C11C1 antibodies bound with high affinity (1.71 and 0.77 nM, respectively) to their antigenic determinants on the HMWK molecule. The 2B5 antibody did neutralize the ability of HMWK to inhibit platelet calpain. These studies with monoclonal antibodies directed to each of the HMWK chains indicate that HMWK is a bifunctional molecule that can serve as a cofactor for serine zymogen activation and an inhibitor of cysteine proteases.     

High molecular mass kininogen inhibits metalloproteinases of Bothrops jararaca snake venom

High molecular mass kininogen (HK) purified from Bothrops jararaca (Bj) plasma was tested on activities of the Bj venom in vivo and in vitro. Results showed that, when incubated with BjHK, the Bj venom presented inhibition on hemorrhagic, edema forming, myotoxic, and coagulant activities. It is well known that metalloproteinases are directly or indirectly involved in these activities. Similarly, human HK inhibits the hemorrhagic effect of the Bj venom as well as hemorrhagic and enzymatic effects of jararhagin, a hemorrhagic metalloproteinase isolated from Bj venom. Complex between HK and jararhagin was not detected by gel filtration. Nevertheless, the inhibitory effect of the hemorrhagic activity of the venom was only partial when HK was pre-incubated with 0.4mM ZnCl(2) or with 0.45mM CaCl(2). These data suggest that the inhibitory effect depends, at least partially, on the competition for ions between kininogen and metalloproteinases of the venom.     

Heparin affects the interaction of kininogen on endothelial cells

In the plasma kallikrein-kinin system, it has been shown that when plasma prekallikrein (PK) and high molecular weight kininogen (HK) assemble on endothelial cells, plasma kallikrein (huPK) becomes available to cleave HK, releasing bradykinin, a potent mediator of the inflammatory response. Because the formation of soluble glycosaminoglycans occurs concomitantly during the inflammatory processes, the effect of these polysaccharides on the interaction of HK on the cell surface or extracellular matrix (ECM) of two endothelial cell lines (ECV304 and RAEC) was investigated. In the presence of Zn⁺², HK binding to the surface or ECM of RAEC was abolished by heparin; reduced by heparan sulfate, keratan sulfate, chondroitin 4-sulfate or dermatan sulfate; and not affected by chondroitin 6-sulfate. By contrast, only heparin reduced HK binding to the ECV304 cell surface or ECM. Using heparin-correlated molecules such as low molecular weight dextran sulfate, low molecular weight heparin and N-desulfated heparin, we suggest that these effects were mainly dependent on the charge density and on the N-sulfated glucosamine present in heparin. Surprisingly, PK binding to cell- or ECM-bound-HK and PK activation was not modified by heparin. However, the hydrolysis of HK by huPK, releasing BK in the fluid phase, was augmented by this glycosaminoglycan in the presence of Zn²⁺. Thus, a functional dichotomy exists in which soluble glycosaminoglycans may possibly either increase or decrease the formation of BK. In conclusion, glycosaminoglycans that accumulated in inflammatory fluids or used as a therapeutic drug (e.g., heparin) could act as pro- or anti-inflammatory mediators depending on different factors within the cell environment.