Activation of human plasma prekallikrein by Pseudomonas aeruginosa elastase in vitro

Human plasma prekallikrein, precursor of the bradykinin-generating enzyme, was activated in a purified system under a near physiological condition (pH 7.8, ionic strength I = 0.14, 37°C) by Pseudomonas aeruginosa elastase which is a tissue-destructive metalloproteinase. Compared with that, Pseudomonas aeruginosa alkaline proteinase poorly activated it with a rate as low as less than one-twentieth of that of elastate. The activation by elastase was blocked with a specific inhibitor of elastase, HONHCOCH(CH₂C₆H₅)CO-Ala-Gly-NH₂ (10 μM). Generation of kallikrein-like amidolytic activity was also observed in plasma deficient in Hageman factor by treatment with elastase, but was not in plasma deficient in prekallikrein. The kallikrein-like activity generated in Hageman factor deficient plasma as well as the generation process itself was indeed inhibited by antihuman prekallikrein goat antibody. These results suggest that the pathological activation of the kallikrein-kinin system might occur under certain clinical conditions in pseudomonal infections.     

Activation of hageman factor and prekallikrein and generation of kinin by various microbial proteinases

Activation of the Hageman factor-kallikrein-kinin system by serratial 56-kDa proteinase was previously demonstrated (Matsumoto, K., Yamamoto, T., Kamata, T., and Maeda, H. (1984) J. Biochem. (Tokyo) 96, 739-749; Kamata, R., Yamamoto, T., Matsumoto, K., and Maeda, H. (1985) Infect. Immun. 48, 747-753). To investigate whether the activation of the system is specific for 56-kDa proteinase or is found similarly with other microbial proteinases, 11 proteinases of microbial origins were studied; the 56-kDa proteinase was the control. For in vitro studies, activation of guinea pig Hageman factor and prekallikrein was examined in purified systems as well as in plasma as a zymogen source. Specific antibodies and inhibitors confirmed the activation steps of the cascade. In the in vivo study the enhancement of vascular permeability in guinea pig skin and its sensitivity to inhibitors of activated Hageman factor, plasma kallikrein, or a kininase were examined. The results from the in vivo experiments were consistent with those in vitro. Taking all the data together, we classified the 11 microbial proteinases into three groups as follows: 1) Serratia marcescens 56-, 60-, and 73-kDa proteinases, Pseudomonas aeruginosa alkaline proteinase and elastase, and Aspergillus melleus proteinase (this group activated Hageman factor but not prekallikrein); 2) Vibrio vulnificus proteinase, subtilisin from Bacillus subtilis, and thermolysin from Bacillus stearothermophilus (this group activated both Hageman factor and prekallikrein); 3) Streptomyces caespitosus proteinase and V8 proteinase from Staphylococcus aureus (this group activated neither Hageman factor nor prekallikrein, but generated kinin from high molecular weight kininogen directly).     

Activation of human Hageman factor by Pseudomonas aeruginosa elastase in the presence or absence of negatively charged substance in vitro.

Human Hageman factor, a plasma proteinase zymogen, was activated in vitro under a near physiological condition (pH 7.8, ionic strength I = 0.14, 37 degrees C) by Pseudomonas aeruginosa elastase, which is a zinc-dependent tissue destructive neutral proteinase. This activation was completely inhibited by a specific inhibitor of the elastase, HONHCOCH(CH2C6H5)CO-Ala-Gly-NH2, at a concentration as low as 10 microM. In this activation Hagemen factor was cleaved, in a limited fashion, liberating two fragments with apparent molecular masses of 40 and 30 kDa, respectively. The appearance of the latter seemed to correspond chronologically to the generation of activated Hageman factor. Kinetic parameters of the enzymatic activation were kcat = 5.8 x 10(-3) s-1, Km = 4.3 x 10(-7) M and kcat/Km = 1.4 x 10(4) M-1 x s-1. This Km value is close to the plasma concentration of Hageman factor. Another zinc-dependent proteinase, P. aeruginosa alkaline proteinase, showed a negligible Hageman factor activation. In the presence of a negatively charged soluble substance, dextran sulfate (0.3-3 micrograms/ml), the activation rate by the elastase increased several fold, with the kinetic parameters of kcat = 13.9 x 10(-3) s-1, Km = 1.6 x 10(-7) M and kcat/Km = 8.5 x 10(4) M-1 x s-1. These results suggested a participation of the Hageman factor-dependent system in the inflammatory response to pseudomonal infections, due to the initiation of the system by the bacterial elastase.     

Studies of C1 inactivator-plasma kallikrein complexes in purified systems and in plasma

An enzyme-linked immunosorbent assay (ELISA) has been developed for the quantification of C1 inactivator-kallikrein (C1In-K) complexes. The formation of complexes assayed by this method parallelled the inhibition of plasma kallikrein esterase activity by C1 inactivator in purified systems. C1In-K complexes were detected when a final concentration of 5.7 nM plasma kallikrein was added to plasma, equivalent to the activation of 1% of the plasma prekallikrein. Exogenous Hageman factor fragment added to plasma induced the rapid formation of C1In-K complexes, whereas there was an appreciable delay when the plasma contact system was activated by the addition of kaolin. In both systems, the rate of formation and final amount of complex generated were directly related to the concentration of Hageman factor fragment or of kaolin added, indicating that this proteolytic pathway is tightly regulated. C1In-K complexes were not generated by kaolin in plasma congenitally deficient in Hageman factor or prekallikrein or by kallikrein in hereditary angioedema plasma deficient in C1 inactivator, thus confirming the specificity of the assay. Sucrose gradient ultracentrifugation studies showed plasma C1In-K complexes to have a molecular weight consistent with a 1:1 molar complex. In contrast, the complex displayed an anomalously high molecular weight on gel filtration chromatography. These data demonstrate that a sensitive and specific probe has been developed for documenting plasma kallikrein activation.     

Contact activation of prekallikrein and plasminogen in plasma

Activation of the Hageman factor-prekallikrein system in the whole human blood plasma is studied as affected by organic silica (aerosils) with anionic and cationic properties. Positive- and negative-charged aerosils are shown to possess the same ability to activate prekallikrein. Activity of prekallikrein was manifested in hydrolysis of the chromogenic substrate--Benz-Pro-Phen-Arg-paranitroanilide . HCl, kininogen and protamine sulphate formed by kallikrein. The data permit supposing that optimal activation of the Hageman factor requires the polar (but not ionic) groups with hydrophilic properties on activating surfaces. Plasminogen under contact activation, in contrast to prekallikrein is activated only in the diluted plasma (pH 4.8), and not completely. Possible mechanisms of the contact activation and interaction of the Hageman factor, prekallikrein and high-molecular kininogen in this process are discussed.     

Activation of human plasma prekallikrein by Pseudomonas aeruginosa elastase. II. Kinetic analysis and identification of scissile bondof prekallikrein in the activation

Activation of human plasma prekallikrein by a bacterial metalloendopeptidase, Pseudomonas aeruginosa elastase, was reported (Shibuya et al. (1991) Biochim. Biophys. Acta 1097, 23–27). Details of the activation process were presently studied. The activation accompanied limited proteolysis of a peptide bond inside of a disulfide bridge of prekallikrein molecule. Amino acid sequencing analysis of the newly generated amino-terminal revealed that the cleavage site was Arg₃₇₁-Ile₃₇₂ bond which is the scissile bond in the activation of prekallikrein with trypsin-type proteinases. A pentapeptide substrate, 2-aminobenzoyl-Ser-Thr-Ile-Val-4-nitrobenzylamide, which contained the amino acid sequence identical to that around the scissile bond of prekallikrein was synthesized. Pseudomonal elastase, indeed, hydrolyzed the substrate at Arg-Ile bond with the kinetic parameters of K_m = 118 μM, k_cat = 1.56/s and k_cat/K_m = 1.33 · 10⁴/s M. These results indicated that the Arg₃₇₁-Ile₃₇₂ bond was sensitive not only to trypsin-type serine proteinases, but also a bacterial metalloproteinase. Kinetic analysis of the prekallikrein activation by psuedomonal elastase, however, revealed that the activation rate was show, though the K_m values was good enough to expect an occurence of this activation in vivo (K_m = 248 nM, k = 6.8 · 10^?4/s, and k_cat/K_m = 2.7 · 10³/s M. The activation rate of prekallikrein by pseudomonal elastase in Hageman factor deficient plasma was remarkably improved when the plasma was reconstituted with purified Hageman factor molecule. From the results, a biologuical significance of the proteinase cascade in the plasma kinin generation was also indicated. The present in vitro study might support the hypothesis that the Hageman factor/kallikrein-kinin system plays an important role in bacterial infection including the pseudomonal one.     

Autoactivatability of human Hageman factor (factor XII)

Purified Hageman factor was found to autodigest upon binding to a negatively charged surface such as kaolin. Assessment by incorporation of tritiated diisopropylfluorophosphate indicated that this cleavage was accompanied by activation and that the two known forms of activated Hageman factor result. Cleavage within a critical disulfide bridge generated activated Hageman factor, a two-chain enzyme of molecular weight 80,000 as well as the active Hageman factor fragment, a 28,000 molecular weight cleavage product. The autocleavage seen was dependent upon the percentage of activated Hageman factor in the starting material and was independent of HMW-kininogen. This result suggest that initiation of the intrinsic coagulation cascade may, in part, depend upon the autoactivatability of Hageman factor described herein. This observation may in turn, account for the ability of prekallikrein deficient plasma to gradually autoactivate as a function of the time of contact with initiating surfaces.     

Role of bacterial proteases in pseudomonal and serratial keratitis

Pseudomonas aeruginosa and Serratia marcescens can cause refractory keratitis resulting in corneal perforation and blindness. These bacteria produce various kinds of proteases. In addition to pseudomonal elastase (LasB) and alkaline protease, LasA protease and protease IV have recently been found to be more important virulence factors of P. aeruginosa . S. marcescens produces a cysteine protease in addition to metalloproteases. These bacterial proteases have a number of biological activities, such as degradation of tissue constituents and host defense-oriented proteins, as well as activation of zymogens (Hageman factor, prekallikrein and pro-matrix metalloproteinases) through limited proteolysis. In this article, the properties of these bacterial proteases are reviewed and the pathogenic roles of these proteases in pseudomonal keratitis are discussed.     

Pathogenesis of serratial infection: activation of the Hageman factor-prekallikrein cascade by serratial protease

A serratial protease with an apparent molecular weight of 56,000 (56K protease), which had been purified from the culture supernatant of a strain of Serratia marcescens isolated from a corneal lesion of a human eye [Matsumoto, K. et al. (1984) J. Bacteriol. 157, 225-232], greatly enhanced vascular permeability when injected into guinea pig skin. The 56K protease, which requires zinc ion for activity, was found to possess plasma kallikrein-like properties in vitro as judged by (i) preferential amidolysis of carbobenzoxy-Phe-Arg-4-methylcoumaryl-7-amide and Pro-Phe-Arg-4-methylcoumaryl-7-amide, which are known substrates for plasma kallikrein; (ii) release of kinin from high-molecular-weight kininogen; and (iii) prompt activation of Hageman factor followed by generation of kallikrein from plasma prekallikrein. These results suggest that the 56K protease enhances vascular permeability through activation of a Hageman factor-kallikrein-kinin pathway in vivo, and this molecular process appears to be a rational mechanism of enhancement of permeability and serratial pathogenesis.     

The significance of high molecular weight kininogen for contact activation of rat blood coagulation, in vitro.

The involvement of the high molecular weight rat kininogen in the activation of the rat contact system by kaolin-cephalin, kaolin, sulfatides and ellagic acid has been investigated, using a rat plasma congenitally devoid of this kininogen. Coagulation times induced by these activators were shorter in normal as well as in deficient rat plasma than in normal human plasma. Coagulation times were prolonged in deficient rat plasma, when the incubation times was three min or less. By kaolin or cephalin-kaolin, this prolongation disappeared when the incubation time reached ten min. The activation of plasma prekallikrein developed slowly in deficient plasma with all the triggers but reached control level after ten min of incubation. By kaolin-cephalin, the activation of Hageman factor was weak and slow in deficient plasma during the ten min of incubation. In rat, high molecular weight kininogen plays thus a role in the activation of the contact system by these triggers. But this role seems to be less important than in human plasma.     

Activation mechanism of human Hageman factor-plasma kallikrein-kinin system by Vibrio vulnificus metalloprotease.

Vivrio vulnificus, an opportunistic human pathogen, secretes a metalloprotease (VVP). The VVP inoculated into a guinea pig is known to generate bradykinin through activation of the Hageman factor-plasma kallikrein-kinin system. VVP was shown to possess the ability to activate the human system through the same mechanism as that clarified in the guinea pig system, namely, VVP converted both human zymogens (Hageman factor and plasma prekallikrein) to active enzymes (activated Hageman factor and plasma kallikrein), and the then generated kallikrein liberated bradykinin from high-molecular-weight kininogen. However, in the presence of plasma alpha 2-macroglobulin (alpha 2M), the VVP action was drastically decreased. This finding suggests that the human system might be activated only at the interstitial-tissue space which contains negligible amounts of alpha 2M or in the bloodstream of the individuals whose plasma alpha 2M level is extremely reduced.     

Use of chromogenic substrates in the clarification of disorders in the early stages of blood coagulation

The kallikrein specific chromogenic peptide substrates S-2302 (KABI) and Chromozym PK (Boehringer) were used in the first analysis of a familial defect in the early stage of clotting. Slight to extensive prolongation of the activated partial thromboplastin time was seen in the affected persons. Using dextransulfate for activation of plasma marked deficiency in kallikrein activity was found in 3 persons. Using factor XIIa (activated Hageman factor) for activation normal prekallikrein levels were found in 2 of them whereas factor XII levels, however, were below normal. The third had a prekallikrein deficiency presumably caused by oral contraceptives. In a fourth member of the family factor XII deficiency was found with normal kallikrein activity. The application of chromogenic peptide substrates for analysing the early stage of clotting has to take into account the special mechanisms of activation.     

Isolation and characterization of bovine plasma prekallikrein (Fletcher factor)

Prekallikrein (Fletcher factor) has been purified from bovine plasma approximately 25 000-fold with an overall yield of 14%. Purification steps included ammonium sulfate fractionation and column chromatography on heparin-agarose, DEAE-Sephadex, CM-Sephadex, benzamidine-agarose, and arginine methyl ester-agarose. The purified protein was homogeneous as judged by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and amino-terminal sequence analysis. Bovine plasma prekallikrein is a glycoprotein with a molecular weight of 82 000 as determined by sedimentation equilibrium centrifugation. It contains 12.9% carbohydrate, including 6.2% hexose, 4.5% N-acetylglucosamine, and 2.2% N-acetylneuraminic acid. Prekallikrein is a single polypeptide chain with an amino-terminal sequence of Gly-Cys-Leu-Thr-Gln-Leu-Tyr-His-Asn-Ile-Phe-Phe-Arg-Gly-Gly. This sequence is homologous to the amino-terminal sequence of human factor XI (plasma thromboplastin antecedent). Both prekallikrein and kallikrein require kaolin to correct Fletcher factor deficient plasma. Kallikrein, however, has a specific activity 3.5 times greater than prekallikrein. Prekallikrein does not correct plasma deficient in factor XII (Hageman factor), factor XI, or high molecular weight kininogen (Fitzgerald factor).     

Activation of prekallikrein from human blood serum by different activators]

It was shown that the activated Hageman factor and its active fragment convert a greater amount of prekallikrein into kallikrein than is observed under the effects of trypsin and kallikrein. The latter two enzymes convert from 30 to 60% of the Hageman factor-activated prekallikrein and its active fragment. A possible existence of two prekallikrein forms is discussed. A mechanism of interaction between individual components of the kininogenase system and their activators is discussed.     

High molecular weight kininogen-binding site of prekallikrein probed by monoclonal antibodies.

A panel of monoclonal antibodies against human prekallikrein was raised in mice and characterized with respect to the major antigenic epitopes. Of 18 antibodies, nine were directed against the light chain portion performing the proteolytic function of activated kallikrein, and nine recognized the heavy chain mediating the binding of prekallikrein to high molecular weight (H-)kininogen. Among the anti-heavy chain antibodies, one (PK6) interfered with the procoagulant activity of prekallikrein, and prolonged in a concentration-dependent manner the activated partial thromboplastin time of reconstituted prekallikrein-deficient plasma (Fletcher type). Antibody PK6 was subtyped IgG1,k and had an apparent Kass of 6.8 +/- 0.44.10(8) M-1 for prekallikrein. Functional analyses revealed that PK6 does not interfere with prekallikrein activation by activated Hageman factor (beta-F XIIa), and has no effect on the kininogenase function of activated kallikrein. Monoclonal antibody PK6 but none of the other anti-heavy chain antibodies completely prevented complex formation of prekallikrein with H-kininogen, and readily dissociated preformed complexes of prekallikrein and H-kininogen. Likewise, Fab' and F(ab')2 fragments of PK6 blocked H-kininogen binding to prekallikrein. A synthetic peptide of 31 amino acid residues encompassing the entire prekallikrein binding region of H-kininogen effectively competed with PK6 for prekallikrein binding indicating that the target epitope of PK6 is juxtaposed to, if not incorporated in the H-kininogen-binding site of prekallikrein. Extensive cross-reactivity of PK6 with another H-kininogen-binding protein of human plasma, i.e. factor XI, suggested that the structure of the target epitope of PK6 is well conserved among prekallikrein and factor XI, as would be expected for the kininogen-binding site shared by the two proteins. It is anticipated that monoclonal antibody PK6 will be an important tool for the precise mapping of the hitherto unknown kininogen-binding site of prekallikrein.     

The Hageman factor-dependent system in the vascular permeability reaction

The mechanism by which the Hageman factor-dependent system induces vascular permeability has been analyzed. The Mr-28,000 active fragment of guinea pig Hageman factor (beta-HFa), injected intradermally, induces an increase in local vascular permeability. Inhibition of vascular permeability resulted from pretreatment of the beta-HFa with immunopurified anti-Hageman factor F(ab')2 antibody at concentrations of 10(-6)-10(-7) M as well as by incubation with corn and pumpkin seed inhibitors of beta-HFa. To determine whether prekallikrein and kallikrein participated in the permeability induced by beta-HFa, circulating prekallikrein was depleted by intra-arterial injections of anti-prekallikrein F(ab')2 antibody. This resulted in about 80% diminution of the vascular permeability response to beta-HFa, without affecting the permeability reaction to bradykinin. Soybean trypsin inhibitor (10(-6) M), injected at the same cutaneous site as the beta-HFa, inhibited the vascular permeability response to beta-HFa by more than 90%. This concentration of soybean inhibitor blocked more than 90% of the activity of guinea pig plasma kallikrein, but did not inhibit the amidolytic capacity of beta-HFa. The permeability activity of beta-HFa (but not its amidolytic activity) was augmented 10-fold by simultaneous injection of a synthetic kinin potentiator, SQ 20,881 (Glu-Tyr-Pro-Arg-Pro-Gln-Ile-Pro-Pro-OH), and was almost completely inhibited by the simultaneous injection of a kinin-destroying enzyme, carboxypeptidase B. These results support the hypothesis that the greatest proportion of vascular permeability induced by beta-HFa is produced by the activation of prekallikrein followed by the release of kinin in the cutaneous tissue. These data offer the first in vivo evidence that the Hageman factor-dependent system by itself can induce inflammatory changes.     

Hageman factor substrates. Human plasma prekallikrein: mechanism of activation by Hageman factor and participation in hageman factor-dependent fibrinolysis.

Two molecular forms of prekallikrein can be isolated from pooled normal human plasma. Their approximate molecular weights by sodium dodecyl sulfate-gel electrophoresis are 88,000 and 85,000. The two bands observed are shown to represent prekallikrein by functional, immunochemical, and structural criteria. Both forms are cleaved by activated Hageman factor, they appear to share antigenic determinants, they are not interconvertible upon incubation with activated Hageman factor or kallikrein, and the ratio of kinin-generating, and plasminogen-activating activities of the preparations are independent of the relative proportion of each band. Activated Factor XII converts prekallikrein to kallikrein by limited proteolysis and two disulfide-linked chains designated kallikrein heavy chain (Mr = 52,000) and kallikrein light chains (Mr = 36,000 or 33,000) are formed. The active site is associated with the light chains as assessed by incorporation of [3H]diisopropyl fluorophosphate. No dissociable fragments were observed in the absence of reducing agents. However, kallikrein could digest prekallikrein to diminish its molecular weight by 10,000. In addition, two factors capable of activating plasminogen to plasmin have been isolated; one is identified as kallikrein. The second principle fractionates with Factor XI and is demonstrable in normal and prekallikrein-deficient plasma.     

Interaction of Hageman factor,prekallikrein activator and plasmin

When plasmin was incubated with active Hageman factor prepared from acetone-activated human plasma by adsorption and elution from pre-treated supercel, no change in prekallikrein activator (PKA) activity was found, even though by polyacrylamide gel electrophoresis and by supercel adsorption it was shown that the active Hageman factor had been converted to the 30, 000 molecular weight fragment. These data are in agreement with the concept that PKA is derived from Hageman factor, but do not support the concept that the conversion of plasminogen to plasmin is necessary for maximal generation of PKA activity in human plasma. Also reported is a radiochemical method for the measurement of PKA (active Hageman factor) activity which is 300 times more sensitive than the guinea pig ileum bioassay and 10 times more sensitive than the clotting tests.     

Human high molecular weight kininogen. Studies of structure-function relationships and of proteolysis of the molecule occurring during contact activation of plasma.

Human high Mr kininogen was purified from normal plasma in 35% yield. The purified high Mr kininogen appeared homogeneous on polyacrylamide gels in the presence of sodium dodecyl sulfate and mercaptoethanol and gave a single protein band with an apparent Mr = 110,000. Using sedimentation equilibrium techniques, the observed Mr was 108,000 +/- 2,000. Human plasma kallikrein cleaves high Mr kininogen to liberate kinin and give a kinin-free, two-chain, disulfide-linked molecule containing a heavy chain of apparent Mr = 65,000 and a light chain of apparent Mr = 44,000. The light chain is histidine-rich and exhibits a high affinity for negatively charged materials. The isolated alkylated light chain quantitatively retains the procoagulant activity of the single-chain parent molecule. 125I-Human high Mr kininogen undergoes cleavage in plasma during contact activation initiated by addition of kaolin. This cleavage, which liberates kinin and gives a two-chain, disulfide-linked molecule, is dependent upon the presence of prekallikrein and Factor XII (Hageman factor) in plasma. Addition of purified plasma kallikrein to normal plasma or to plasmas deficient in prekallikrein or Factor XII in the presence or absence of kaolin results in cleavage of high Mr kininogen and kinin formation.     

Hageman factor-dependent pathways: mechanism of initiation and bradykinin formation

The concentration of bradykinin in human plasma depends on its relative rates of formation and destruction. Bradykinin is destroyed by two enzymes: a plasma carboxypeptidase (anaphylatoxin inactivator) removes the COOH-terminal arginine to yield an inactive octapeptide, and a dipeptidase (identical to the angiotensin-converting enzyme) removes the COOH-terminal Phe-Arg to yield a fragment of seven amino acids that is further fragmented to an end product of five amino acids. Formation of bradykinin is initiated on binding of Hageman factor (HF) to certain negatively charged surfaces on which it autoactivates by an autodigestion mechanism. Initiation appears to depend on a trace of intrinsic activity present in HF that is at most 1/4000 that of activated HF (HFa); alternatively traces of circulating HFa could subserve the same function. HFa then converts coagulation factor XI to activated factor XI (XIa) and prekallikrein to kallikrein. Kallikrein then digests high-molecular-weight kininogen (HMW-kininogen) to form bradykinin. Prekallikrein and factor XI circulate bound to HMW-kininogen and surface binding of these complexes is mediated via this kininogen. In the absence of HMW-kininogen, activation of prekallikrein and factor XI is much diminished; thus HMW-kininogen has a cofactor function in kinin formation and coagulation. Once a trace of kallikrein is generated, a positive feedback reaction occurs in which kallikrein rapidly activates HF. This is much faster than the HF autoactivation rate; thus most HFa is formed by a kallikrein-dependent mechanism. HMW-kininogen is also therefore a cofactor for HF activation, but its effect on HF activation is indirect because it occurs via kallikrein formation. HFa can be further digested by kallikrein to form an active fragment (HFf), which is not surface bound and acts in the fluid phase. The activity of HFf on factor XI is minimal, but it is a potent prekallikrein activator and can therefore perpetuate fluid phase bradykinin formation until it is inactivated by the C1 inhibitor. In the absence of C1 inhibitor (hereditary angioedema) HFf may also interact with C1 and activate it enzymatically. The resultant augmented bradykinin formation and complement activation may account for the pathogenesis of the swelling characteristic of hereditary angioedema and the serologic changes observed during acute attacks.