- Email: [email protected]
Contact-system activation in children with vasculitis
MECHANISMS OF DISEASE
Mechanisms of disease
Contact-system activation in children with vasculitis
Robin Kahn, Heiko Herwald, Werner Müller-Esterl, Roland Schmitt, Ann-Christine Sjögren, Lennart Truedsson, Diana Karpman
Summary Background The contact system triggers the kallikrein-kinin cascade, liberating bradykinin from high-molecular-weight kininogen. Effectors of the contact system have proinflammatory and vasoactive properties. Vasculitis is a condition characterised by inflammation around vessel walls, leading to secondary tissue damage for which the underlying molecular mechanisms are poorly understood. Our aim was to investigate contact-system activation in children with vasculitis. Methods We compared 17 children, aged 4–19 years, with vasculitis, engaging the skin, joints, intestines, or kidneys, with 21 controls, aged 2–18 years. We analysed proteolysis of high-molecular-weight kininogen by immunoblotting. Plasma bradykinin concentrations were quantified by ELISA. Kidney and skin biopsies were stained in situ for kinins. Concentrations of heparin binding protein (HBP) were quantified by ELISA. Findings We noted extensive proteolysis of high-molecularweight kininogen in the plasma of 13 of 17 patients, but in only one of 21 controls (p<0·0001). Bradykinin concentrations were higher in the patients’ plasma (median 320 ng/L, range <1–19 680) than in plasma from controls (11 ng/L, <1–304; p=0·0004). Patients had local release of kinins at sites of inflammation in kidney and skin biopsies. HBP values were raised in patients (17·4 g/L, 5·4–237·6) compared with controls (6 g/L, 2·5–43·4; p=0·008). Interpretation Activation of the contact system could play a part in the pathogenesis of vasculitis, and explain the inflammation, pain, vasodilatation, and oedema seen in patients.
Lancet 2002; 360: 535–41
Departments of Paediatrics (R Kahn, R Schmitt MD, A-C Sjögren BSc, D Karpman MD), Cell and Molecular Biology (H Herwald PhD), and Laboratory Medicine (L Truedsson MD), Lund University, Lund, Sweden; and Institute of Biochemistry II, University Hospital, Frankfurt, Germany (Prof W Müller-Esterl PhD) Correspondence to: Dr Diana Karpman (e-mail: [email protected])
THE LANCET • Vol 360 • August 17, 2002 • www.thelancet.com
Introduction The contact system of human plasma consists of four components, plasma prekallikrein, high-molecular-weight kininogen, factor XI, and factor XII.1 Activation of the system is triggered by conversion of small quantities of factor XII to factor XIIa by endogenous or exogenous stimuli, and amplified by the mutual activation of factor XII and prekallikrein, resulting in a high local concentration of active plasma kallikrein, PKa, which is tethered to endothelial, subendothelial, and granulocyte surfaces via high-molecular-weight kininogen.1,2 Activated PKa rapidly cleaves high-molecular-weight kininogen into heavy and light chains, and liberates the terminal effector peptide, bradykinin, which binds to neighbouring kinin receptors of the B2 subtype and triggers various signalling cascades.1 Degradation of bradykinin by carboxypeptidase N or angiotensin converting enzyme (ACE) leads to generation of desArg9-bradykinin, which binds B1 kinin receptors.3 The B2 kinin receptor is constitutively expressed, whereas the B1 receptor is induced during inflammatory conditions by stimuli such as interleukin 1.4 Activity of the contact-system cascade is terminated in vivo by irreversible inhibiton of PKa by C1 inhibitor and ␣2-macroglobulin.1 The contact system can be triggered on cell membranes or on negatively charged artificial surfaces, such as kaolin and dextran sulphate.5 Activation of the contact system arises in several clinical conditions, including hereditary angio-oedema,6 bacteraemia, and sepsis,7,8 ulcerative colitis,9 Alzheimer’s disease,10 and during cardiopulmonary bypass,11 but has not been described in other conditions that involve systemic inflammation. The components of the contact system have multiple physiological effects. Plasma kallikrein is chemotactic for neutrophils12 and induces the secretion of neutrophil elastase.13 Bradykinin is a potent inflammatory mediator that reduces local blood pressure, causes capillary leakage and oedema, and induces pain.1 Bradykinin might stimulate endothelial cells to form superoxide, nitric oxide, and prostacyclin,14 which inhibits platelet aggregation. When vessels are damaged, bradykinin stimulates intimal hypertrophy and the proliferation of smooth muscle.1 Vasculitis is characterised by inflammation with neutrophil influx in and around vessel walls, leading to perturbed vessel patency and tissue damage. Injured vessels can leak plasma and blood cells, sometimes resulting in formation of thrombi. Various organs can be involved, including the kidney, intestine, skin, and joints. Patients with vasculitis exhibit symptoms such as abdominal pain purpura, joint pain and swelling, as well as oedema and other manifestations of renal dysfunction.15 The pathogenetic mechanisms of vasculitis have not been completely elucidated. Damaged endothelial cells and neutrophil-induced tissue injury are thought to be the underlying cause of pathology.16 Neutrophil components, such as proteinase 3 and heparin-binding protein, mediate 535
For personal use. Only reproduce with permission from The Lancet Publishing Group.
MECHANISMS OF DISEASE
GLOSSARY CARBOXYPEPTIDASE N
Carboxypeptidase N (arginine carboxypeptidase, kininase I) is an enzyme synthesised by the liver and secreted into the plasma. It is a kininase, which inactivates bradykinin and kallidin by removing carboxyterminal arginine from these peptides. Carboxypeptidase N also hydrolyses aminoacids of other peptides, including complement anaphylatoxins. CONTACT SYSTEM
The contact (kallikrein/kinin) system is composed of three serine proteinases: factor XII, factor XI, plasma prekallikrein (kallikrein in active form), and a non-enzymatic component, high-molecular-weight kininogen, which are synthesised by the liver and secreted into the plasma. Upon activation, the contact system is involved in the regulation of different procoagulant, profibrinolytic, antiadhesive, and proinflammatory processes, such as activation of the intrinsic pathway of coagulation, blood-pressure regulation, inhibition of thrombin activation of cells, fibrinolysis, and generation of kinins. ROCKET IMMUNOELECTROPHORESIS
A quantitative electrophoretic method for measurement of antigen, in which the antigen is placed on agar containing specific antiserum to the antigen being assayed. Antigens precipitate in rocket shape due to the electric current, and the area under the rocket is proportional to the antigen concentration.
inflammation.17,18 Certain patients with systemic vasculitis have serum antibodies directed against cytoplasmic components of neutrophils (antineutrophil cytoplasmic antibodies [ANCA]), such as proteinase 3 and myeloperoxidase. These antibodies are used in the diagnosis and management of vasculitis, since they can function as markers of disease activity.16 Theoretically, activation of the contact system could explain some of the characteristics of vasculitis— eg, inflammation, oedema, pain, vasodilatation, and local bleeding. Our aim was to assess systemic and local activation of the contact system in patients with vasculitis. We postulate that inappropriate activation of the contact system could contribute to the pathogenesis of this condition.
Methods Participants We included all children with acute and chronic forms of vasculitis, treated at the Department of Paediatrics, section of Paediatric Nephrology, University Hospital, Lund, Sweden, between June, 1999, and September, 2001. We defined subtypes of vasculitis in accord with the Chapel Hill nomenclature.19 We obtained control plasma samples from consecutive patients seen at the outpatient clinic of the same hospital between June and July, 1999, for follow-up and treatment of various diseases. Control serum samples were obtained from individuals seen at the outpatient clinic for follow up 2 months to 10 years after diarrhoea-associated haemolytic uraemic syndrome. All controls were asymptomatic with normal renal function at time of sampling. The ethics committee of Lund University approved the study, and all samples were taken with the informed consent of participants (older than age 15 years) or their parents (when younger than age 15 years). Protocol We obtained blood samples during or after acute episodes of vasculitis, or at follow up for chronic vasculitis. Blood samples were also available from some patients 1–2 years after recovery from the acute episode. Venous blood from patients and controls was collected in
536
5 mL vacutainer tubes, containing 0·5 mL 0·129 mol/L sodium citrate for plasma, and in 4 mL vacutainer Hemogard SST tubes for serum (plasma and serum tubes from Becton Dickinson, Plymouth, UK). We centrifuged the samples immediately after collection at 2000 g for 10 min. Platelet-poor plasma was removed in plastic pipettes and frozen at –80ºC until assayed. Serum was frozen at –20ºC until assayed. High-molecular-weight kininogen is a single-chain protein, consisting of six domains; domains D1–D3 represent the heavy chain region, domain D4 the bradykinin segment, and domains D5H and D6H the light chain region.1 Domains D1–D4 of low-molecular-weight kininogen are identical to those of high-molecular-weight kininogen, whereas the light chain region of lowmolecular-weight kininogen is composed of a single domain, D5L, which differs from D5H.1 We used domainspecific antibodies to D2, D5H, and D6H to discriminate between the various kininogen forms and fragments in immunoblots. We incubated citrated plasma from two healthy adult volunteers with dextran sulphate (molecular mass 5000, final concentration 75 mol/L; Sigma, St Louis, MO, USA) in the presence of 2 mmol/L zinc chloride (Sigma) at 37ºC for 1 h, to induce contactphase activation and proteolysis of high-molecular-weight kininogen.5 In separate tubes, we treated the same plasma samples with 2 mmol/L zinc chloride without dextran sulphate at 37ºC for 1h. We diluted plasma samples 1/100 in 4% (w/v) SDS (Bio Rad, Hercules, CA, USA) in Tris buffer, pH 6·8 (ICN Biomedicals, Aurora, OH, USA), containing 10% (v/v) 2-mercaptoethanol (Kebo Lab, Spånga, Sweden), and boiled them for 5 min. We ran 10 L per sample on a 10% SDS-PAGE gel in an electrophoresis chamber (Bio Rad). The gels were electroblotted for 1 h onto 0·45 m Protran nitrocellulose membranes (Schleicher and Schuell, Dassel, Germany) with a semidry electroblotter (Ancos, Hoejby Denmark). We blocked the membranes at 37ºC for 45 min with 5% dry milk powder and 0·05% Tween 20 in PBS (pH 7·4, 0·01 mol/L phosphate buffer; Life Technologies, Uppsala, Sweden). Immunoblots were done with the following primary antibodies: monoclonal antibody HKH8 (from mouse) directed to common domain D2 of high-molecularweight and low-molecular-weight kininogen20 was used at 1/1000; polyclonal antibody to HKH20 (␣-HKH20; from rabbit) directed to high-molecular-weight kininogen domain D5H21 was used at 1/500; monoclonal antibody HKL22 directed to high-molecular-weight kininogen domain D6H was used at 1/500;20 and antiserum AS88 (from sheep) against purified human high-molecularweight kininogen, which crossreacts with low-molecularweight kininogen was used at 1/5000.22 Secondary antibodies were applied as appropriate: goat antibody against mouse/horseradish peroxidase (HRP, 1/3000), goat antibody against rabbit/HRP (1/3000, both antibodies from Dako Glostrup, Denmark), and donkey antibody against sheep/HRP (1/3000, ICN Biomedicals). We assayed proteolytic processing of high-molecularweight kininogen by immunoblotting as previously described.20 Citrated plasma samples from patients and controls were separated by SDS-PAGE and electrotransferred as described above. They were further incubated for 45 min with antiserum AS88. After washing in PBS/Tween, containing 0·35 mol/L sodium chloride, we incubated the membranes for 45 min with donkey antibody against sheep immunoglubulin/HRP (1/3000), and detected the signal by chemiluminescence. As a positive control for high-molecular-weight kininogen
THE LANCET • Vol 360 • August 17, 2002 • www.thelancet.com
For personal use. Only reproduce with permission from The Lancet Publishing Group.
MECHANISMS OF DISEASE
proteolysis, we added dextran sulphate to normal human plasma from an adult. Plasma was incubated with dextran sulphate in the presence of 2 mmol/L zinc at a final concentration of 75 mol/L at 37ºC for 1h. We detected bradykinin with the MARKIT-M Bradykinin ELISA kit (Dainippon Pharmaceutical, Osaka, Japan) as per the manufacturers instructions and as previously described.23 This analysis requires 500 L of plasma and could, therefore, not be tested in all participants because of lack of sample. The antibody against bradykinin used in this kit is specific for free bradykinin and does not cross react with high-molecularweight or low-molecular-weight kininogen. ACE cleaves and inactivates bradykinin.24 We therefore investigated if increased concentrations of bradykinin in patients with vasculitis were due to reduced ACE activity. We analysed ACE activity in sera from patients, and from controls who had recovered from haemolytic uraemic syndrome (since serum was not available from the other paediatric controls). A spectrophotometric assay based on hydrolysis of the substrate furanacryloyl-Lphenylalanylglycylglycine was used as previously described.25 The substrate was obtained from Bachem (Bubendorf, Switzerland). C1 inhibitor is the major inhibitor of the contact system. Deficiency or dysfunction of C1 inhibitor leads to contact system activation and also to reduced concentrations of C4. For this reason, we measured serum concentrations of C1 inhibitor and C4 in patients by rocket immunoelectrophoresis and turbidimetry, respectively. We extracted DNA from the blood of some patients with the salting-out procedure.26 Two DNA fragments were amplified, a 656 bp fragment of exon 8 of the C1 inhibitor gene, using primers E1 (biotinylated) and E2,27 and a 480 bp fragment, using primers E1 and E2b (5´GCTGACAGAGGACCCAGATC3´). We did the PCR reaction with a Minicycler (MJ Research, MA, Patient number
Sex
1 2 3 4
F F F F
5
F
6
M
7 8 9 10 11 12 13
F M F F F M M
14 15
F F
16 17
M F
Diagnosis
SLE Acute HSP* Acute HSP C-ANCA positive vasculitis†‡ C-ANCA positive vasculitis‡ Microscopic polyangitis Acute HSP§ Chronic HSP Acute HSP§ Status post HSP¶ Acute HSP§ Acute HSP Acute leukocytoclastic vasculitis Acute HSP Wegener’s granulomatosis|| Acute HSP Chronic HSP
USA) with an annealing temperature of 60ºC. The reaction mixture (100 L volume) contained 1 g of genomic DNA, 1 mol/L of each primer, 200 mol/L dNTP, and 0·03 units/L of AmpliTaq Gold polymerase and buffer (Perkin-Elmer, Norwalk, CT, USA). To separate the DNA strands, immobilisation of the PCR products was done, according to a protocol from Dynal AS (Oslo, Norway). We used Sequenase (version 2.0, USB, Cleveland, OH, USA) for the sequencing reaction, and then loaded samples onto a sequencing gel. After drying, we visualised the bands by autoradiography. Heparin binding protein (HBP), also known as azurocidin, is a glycoprotein secreted from azurophilic granules and secretory vesicles of neutrophils.28 HBP functions as an inflammatory mediator.18 We measured the concentrations of the protein in plasma of patients and controls by ELISA as previously described.28 We obtained renal cortical and skin biopsies for medical purposes from the children with vasculitis. Renal cortical and skin biopsy material was also available from two adults biopsied for clear-cell carcinoma (kidney biopsy) and breast cancer (skin biopsy). We used these samples as controls, since they contained areas of histologically normal tissue adjacent to cancerous areas. We took biopsies during acute episodes of vasculitis, and examined them for inflammatory and vascular changes. We also examined renal specimens for glomerular, tubular, and interstitial changes. Samples were tested for the presence of kinin peptides with a rabbit antibody that reacts with bradykinin, lysylbradykinin (kallidin), and desArg9-bradykinin (Biogenesis, Poole, UK), and the streptavidin-HRPconjugate method as previously described.29 We incubated samples at room temperature overnight with antibody against kinin diluted 1/400 in Tris buffer (containing 2·5% [w/v] of bovine serum albumin) for kidney samples, and at 1/800 or 1/1600 for skin samples. Positive samples stained brown. As a control, the primary
Age Clinical manifestation (years) R F A GI
GN
ESR
Laboratory findings CRP
WBC
SA
LFT
Cr
UB
UP
ANCA
19 11 6 11
+ + + +
– + – –
– + + +
– – – –
– – + –
↑ ·· ·· ··
·· ·· ·· ··
·· ·· ·· ··
·· ↓ ↓ ··
·· ·· ·· ··
·· ·· ·· v
– – + +
+ – + –
NA – – +PR-3
14
+
–
+
–
–
↑
··
··
··
··
↑
–
–
+PR-3
13
+
–
+
+
+
↑
↑
↑
↓
↑
↑
+
+
–
4 16 6 18 10 12 12
+ + + – + + +
– – – – – – –
+ + + – + + +
–+ – – + + +
– + – + – – –
·· ·· ·· ·· ·· ·· ↑
·· ·· ↑ ·· ·· ↑ ··
·· ·· ·· ·· ↑ ↑ ··
·· ·· ↓ ·· ·· ·· ··
·· ·· ·· ·· ·· ·· ··
·· ·· ·· ↑ ↑ ·· ↑
·· + – ·· + + –
·· + – ·· + + –
– – NA – +elastase NA –
16 12
+ –
– +
+ +
– –
+ +
·· ↑
↑ ↑
·· ↑
↓ ↓
·· ··
·· ↑
– +
+ +
– +PR-3 and MPO
15 14
+ +
+ –
– –
+ –
+ –
↑ ··
·· ··
·· ··
↓ ··
·· ··
↑ ··
+ +
+ –
– +MPO
NA=not available. F=female. M=male. R=rash. F=fever. A=arthralgia or arthritis. GI=gastrointestinal pain or bleeding. GN=glomerulonephritis verified by renal biopsy. ESR=erythrocyte sedimentation rate. CRP=C-reactive protein. WBC=white blood cell count. SA=serum albumin. LFT=liver function tests. Cr=creatinine. UB=urine blood. UP=urine protein. PR-3=proteinase 3. MPO=myeloperoxidase. SLE=systemic lupus erythematosus. HSP=Henoch Schönlein purpura. ··=Normal laboratory results. *Chronic purpura and infectious episodes after HSP. †Patient had raised circulating immune complex concentration. ‡Patients are siblings. §Patients had mild transient symptoms and recovered within days. ¶Patient had meningococcal sepsis several years after HSP and 1 year before sampling. ||Patient had hearing loss, cough, and fatigue in addition to other symptoms.
Table 1: Clinical and laboratory data of children with vasculitis at time of sampling
THE LANCET • Vol 360 • August 17, 2002 • www.thelancet.com
537
For personal use. Only reproduce with permission from The Lancet Publishing Group.
MECHANISMS OF DISEASE
antibody against kinin was excluded from the protocol. Antibody specificity assay was tested by pretreating the antibody (1/800 and 1/1600) with bradykinin (50 g/L, Dainippon) or des-arg9-bradykinin (10 and 100 mg/L, Biogenesis) for 1h at 37ºC. These combinations were then separately added to skin sections from two patients. The antibody against kinin did not cross-react with highmolecular-weight kininogen on immunoblotting (data not shown). Statistical analysis We assessed differences between patients with vasculitis and controls with respect to breakdown of highmolecular-weight kininogen by Fisher’s exact test. Differences between these groups with respect to bradykinin or HBP concentrations and ACE activity were tested with the Mann Whitney U test. We used the Sign test to ascertain differences between patients during the acute phase of disease and convalescence with respect to breakdown of high-molecular-weight kininogen. We judged a p value of 0·05 or less significant. We used GraphPad Instat (version 2.0) for statistical analyses. Role of the funding source The sponsors of the study had no role in study design, data collection, data analysis, data interpretation, writing of the paper, or in the decision to submit the report. AS88 HK HK*
116 kDa
66 kDa 58 kDa
No rm al pl as Pl de as ma xt ma ra w n i su th lfa te
45 kDa
␣-HKH20 HKL22
HKH8 HK
HK
LK and H63 L58
L58 L45
Results We enrolled 17 patients, five boys and 12 girls, aged 4–19 years (median 12·5). Table 1 shows a summary of their clinical and laboratory data at the time of sampling. The children had various symptoms, including purpura, vasculitic rash, arthritis, arthralgia, abdominal pain, haematuria, and oedema. One child with Wegener’s granulomatosis had, in addition, upper-respiratory-tract symptoms. Two children had chronic Henoch-Schönlein purpura manifested as constant rash and haematuria for more than 2 years. None of the patients were receiving treatment with immunosuppressives or ACE inhibitors at the time of sampling. We took blood samples during (n=14) or after (n=1) acute vasculitis, or at follow up for chronic vasculitis (n=2). We obtained control plasma samples from 21 children, nine boys and 12 girls, aged 2–18 years (median 9). The children were receiving outpatient treatment for diabetes mellitus (n=4), pyelonephritis (n=6), neuroblastoma (n=1), thrombotic thrombocytopenic purpura during remission (n=2), nephrotic syndrome (n=1), anal atresia (n=1), epilepsy (n=3), pigmented purpura (n=1), migraine (n=1), and psychiatric disorder (n=1). Control serum samples were obtained from 14 children, six boys and eight girls, aged 2–18 years (median 5) who had had diarrhoeaassociated haemolytic uraemic syndrome. Figure 1 shows the products produced after proteolysis of high-molecular-weight kininogen. High-molecularweight kininogen was degraded in the plasma of 13 of 17 patients but in only one of 21 controls (Fisher’s exact test p<0·0001; figure 2 and table 2). This control had migraine. Convalescent plasma was available for eight patients (numbers 3, 4, 6, 9, 12, 13, 14, and 15) 1–2 years after the acute episode of vasculitis. Of these eight patients, high-molecular-weight kininogen was degraded in five during the acute phase of disease, but normalised after recovery; was not degraded in two during the acute phase or at convalescence; and was weakly positive in one (patient 13) at convalescence (data not shown). Differences between patients during the acute phase of disease and after recovery were non-significant (Sign test p=0·22). Concentrations of bradykinin were raised in patients (median 320 ng/L, range <1–19 680, IQR 17–655·5; table 2) compared with controls (11 ng/L, <1–304, IQR <1–174, Mann Whitney U test p=0·0004). The control with migraine had a bradykinin concentration of 11 ng/L. Two controls could not be tested because of insufficient plasma. Patients with vasculitis
kDa 116
NH2
D1
D2
D3
D4
D5
D6
Figure 1: Immunoblot of proteolysis products of high-molecularweight kinogen (HK) Western blot of normal plasma and plasma stimulated with dextran sulphate, leading to consumption of all intact HK (upper). The antiserum AS88 against kininogen revealed two major bands of 63–68 kDa and 58 kDa, a minor band of 45 kDa, and a weak band of 90 kDa (HK*=terminally truncated HK). Immunoblot of HK and its fragments with domain-specific antibodies to discriminate between the various forms and degradation fragments (lower). Normal plasma was incubated with dextran sulphate, separated by SDS-PAGE, and immunoblotted with antibodies against domain 2 of the HK and low-molecular-weight kinogen (LK) heavy chain (HKH8) or domains 5 and 6 of the HK light chain (a-HKH20 and HKL22, respectively). H63=common heavy chain of 63 kDa. L58/L45=HK light chain forms of 58 kDa and 45 kDa, respectively. The 45 kDa band is a truncated form of HK light chain. HK domains are depicted at bottom of figure.
538
66 58 45
COOH DXS 1
2
3 4
5
6
7
8 9 10 11 12 13 14 15 16 17
Controls 116 66 58 45 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38
Figure 2: Immunoblot of plasma samples from patients (upper) and controls (lower) with antiserum against kininogen AS88 DXS=dextran sulphate. Kininogen and their fragments as per figure 1.
THE LANCET • Vol 360 • August 17, 2002 • www.thelancet.com
For personal use. Only reproduce with permission from The Lancet Publishing Group.
MECHANISMS OF DISEASE
Patient number
HK proteolysis
Bradykinin (ng/L)
HBP (g/L)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
+ +* +* +* +* – – +* + +* + –‡ –* + + + +
620 156* 550* 340* 472* 189 <1† 256* 184 299* 81 NA 134* 19 680 1488 3360 691
13·9 7·8 7·2 5·4 6·5 237·6 8·9 25·3 221·8 16·7 18·2 NA 39·6 107·2 26·7 6·1 51·4
NA=not available. *Two or more samples with similar results taken at different time points available. †Below the detection limit. ‡Plasma sufficed only for analysis of HK proteolysis.
Table 2: Proteolysis of high-molecular-weight-kininogen (HK) and bradykinin and HBP concentrations in patients
ACE activity in the sera of 14 patients (excluding patients 6, 9, and 17) was in the range 16–61 U/mL (median 38, IQR 27–51). Activity in sera of 14 controls was in the range 27–66 U/mL (48, 43–57, Mann Whitney U test p=0.053). Thus, we cannot rule out the possibility that increased concentrations of bradykinin were associated with reduced degradation due to reduced ACE activity. Inability to inhibit PKa because of a mutated C1 inhibitor protein or the presence of neutralising autoantibodies to native C1 inhibitor causes angio-oedema in patients with excessively high concentrations of bradykinin. We therefore investigated the C1 inhibitor and C4 concentrations in 15 patients (not in patients 5 and 9). The concentrations of C1 inhibitor were 92–199% (median 125; % standardised by concentrations obtained from 100 blood donors); four patients had values above the applied reference interval of 72–153%. The C4 concentrations were 0·07–0·44 g/L (0·24; reference interval 0·1–0·4 g/L). Patient 6 had reduced concentrations of C4 during the early stage of disease, which later normalised. Patient 2 had raised C4; all other patients had normal C4 concentrations. DNA fragments (from patients 2, 6, 8, and 13) of exon 8 of the C1 inhibitor gene, corresponding to positions 400–476 of the aminoacid sequence, were read from the gel. No mutations were present in this sequence, which contains the reactive site. Since concentrations of C1 inhibitor were normal and no mutation in the reactive site was found, it is unlikely that the patients have a C1 inhibitor dysfunction. We measured concentrations of HBP in the plasma of patients, as a marker of neutrophil activation and granular shedding.28 Plasma samples from 16 children with vasculitis (not patient 12) were compared with those of 17 controls. Patients’ HBP concentrations were in the range 5·4–237·6 g/L (median 17·4, IQR 7·5–45·5; table 2). Controls had concentrations in the range 2·5–43·4 g/L (6·0, 4·4–15·7, Mann Whitney U test p=0·008). These results indicate neutrophil activation in patients with vasculitis. Kidney biopsies from seven patients (3, 6, 8, 10, 14, 15, 16) showed diffuse mesangioproliferative glomerulonephritis with formation of crescents. Skin biopsies from patients 5, 13, 16, and 17 showed swollen endothelial cells with infiltrates of granulocytes, including eosinophils, in and around vessel walls and in the interstitium. Extravasation of red blood cells was noted.
THE LANCET • Vol 360 • August 17, 2002 • www.thelancet.com
Figure 3: In-situ immunohistochemistry of kidney and skin biopsies labelled with antibody against kinin A=kidney biopsy from patient 3. Prominent brown staining of glomeruli and tubuli, demonstrating the presence of kinin peptides is noted. Arrow points to glomerular capillary staining. B=control kidney biopsy, showing no staining. C=skin biopsy from patient 13 with acute leukocytoclastic vasculitis. Arrow shows staining in areas of inflammatory influx. D=control skin biopsy, showing weak diffuse staining of the epidermis.
Tissues from patients with vasculitis were examined to ascertain whether or not proteolytic products of the contact system were present in the perivascular and inflammatory areas. Sections taken from patients stained more intensively with antibodies against kinin than did sections from two controls. Kidney biopsies from patients 3, 8, 15, and 16 were examined with the antibody against kinin and had pronounced staining in glomeruli, in which a specific staining of the luminal side of the capillaries and of Bowman’s capsule was noted (figure 3, A). We noted granular staining in the proximal and distal tubuli. In the four skin biopsies, intense staining was observed at the site of inflammatory influx (figure 3, C). In sections from patients and controls, a diffuse staining was seen in the epidermis (figure 3, D). Preincubation of the primary antibody (1/800 and 1/1600) with bradykinin partially inhibited the reaction in two skin biopsies. Figure 4 shows the specificity of the assay for bradykinin. We noted similar results when the primary antibody was preincubated with desArg9-bradykinin, indicating that the antibody does not discriminate between these two active kinin peptides. There was no staining when the primary antibody was omitted.
Figure 4: Specificty of in-situ immunohistochemistry of skin biopsy from patient 16 A=section labeled with antibody against kinin (1/1600). B=antibody (1/1600) preincubated with bradykinin 50 g/L for 1 h at 37ºC before labelling to inhibit specific staining (brown labelling). Arrows point to areas of inflammatory infiltrate with specific granular staining in section A, which diminished in section B. Several sections were taken, 4 m apart. The sections shown in panel A and B show the same area and some of the same cells.
539
For personal use. Only reproduce with permission from The Lancet Publishing Group.
MECHANISMS OF DISEASE
Discussion Our results indicate that, in children with vasculitis, high-molecular-weight kininogen is extensively proteolysed and bradykinin concentrations are greatly raised. Furthermore, kinin peptide is present in inflamed tissues. Bradykinin is known to have potent proinflammatory properties;1 our findings show that it is present at the site of inflammation during systemic vasculitis, suggesting that bradykinin could contribute to the inflammatory state in vasculitis. The contact system is activated in other inflammatory and infectious conditions.7–9 Reduced concentrations of high-molecular-weight kininogen and prekallikrein have been shown in patients with systemic inflammatory response syndrome,7 in which inflammation of multiple organs is a prominent feature. Since bradykinin is raised in a range of inflammatory diseases, its release is probably a distal event in the inflammatory cascade and could be triggered by various different mechanisms. The disease spectrum in children with vasculitis differs from that of adults—eg, children generally have milder forms of vasculitis, such as Henoch Schönlein purpura, whereas adults tend to have severe and chronic ANCAassociated disease. We did not address the possibility of contact-system activation in adults, however, our results indicate that the contact system was often activated in children with chronic symptoms. Furthermore, some patients who had undergone acute Henoch Schönlein purpura and later developed chronic symptoms such as purpura or severe infection continued to have contactsystem activation even after the initial acute event. This finding could suggest that the inflammatory process was ongoing in these patients. We also noted increased breakdown of high-molecular-weight kininogen in one paediatric control with migraine. Although activation of the contact system has not been previously associated with this vascular disorder, intra-arterial administration of bradykinin induces pain in these patients.30 Studies in additional patients with migraine are required to investigate the association between contact-system activation and pain in this condition. Results of in-vitro studies have shown that highmolecular-weight kininogen can be activated on the surface of endothelial cells and granulocytes.1 These cells are involved in the pathological process during vasculitis, and either or both may secrete agents that trigger the disease. Since antibodies directed to cytoplasmic components of neutrophils (ANCA) are seen in certain subgroups of vasculitides, it has been postulated that these cytoplasmic components, or the antibodies themselves, could be essential for the development of vasculitis.15 HBP is a glycoprotein localised in neutrophil azurophilic granules and secretory vesicles, which is a chemoattractant for monocytes and T cells 18 and has known antimicrobial properties. This glycoprotein induces vascular permeability.31 Increased concentrations of HBP have previously been measured in the sera of patients with vasculitis.32 We also noted raised plasma concentrations of HBP in our patients, indicating neutrophil activation and degranulation. Bradykinin was increased in plasma and tissue of patients. Bradykinin can be produced on the surface of various cell types, such as endothelial cells and neutrophils.1 Previous studies have shown that bradykinin can be generated on the luminal side of endothelial cells, which could explain its effect on vascular permeability.33 Cell surfaces that do not produce bradykinin under normal circumstances, such
540
as glomerular endothelial cells and cells in the Bowman’s capsule, also stained positively for kinin peptide in our study. We postulate that there is proteolysis of high-molecular-weight kininogen, generating release of bradykinin, independently in the plasma and in the tissues of patients. Results of investigations have indicated an association between ACE polymorphisms and certain types of vasculitis, such as systemic lupus erythematosus34 and Henoch Schönlein purpura.35 However, these results have been challenged.36 We noted that differences in ACE activity between patients and controls were not significant, though almost. Whether or not reduced ACE activity in vasculitis patients contributes to raised concentrations of bradykinin, therefore, remains unclear and should be addressed in future studies. Our findings show that the contact system is activated in children with vasculitis. Irrespective of whether activation is a primary or secondary event, we suggest that the production of bradykinin could play an important part in the pathogenesis of vasculitis. Bradykinin B1 or bradykinin B2 receptor antagonists, or both,4 could, thus, be useful in reducing inflammatory symptoms in patients with vasculitis. Contributors R Kahn was responsible for planning, did experiments, data analysis, and wrote the report. H Herwald participated in the experimental planning, did experiments, and contributed to the interpretation of data and write-up. W Müller-Esterl helped with study design, data interpretation, and write-up. R Schmitt assisted with experimental planning, did experiments, and interpreted data. A-C Sjögren participated in experimental planning and work as well as data interpretation. L Truedsson assisted with complement analysis, PCR, and data interpretation. D Karpman contributed to the experimental conception, design, work, and patient care, as well as to data interpretation and write-up.
Conflict of interest statement D Karpman has received grants from Centeon Pharma, and travel grants from Roche Pharmaceuticals and Janssen-Cilag Pharmaceuticals. H Herwald was employed by Ferring. W Müller-Esterl is a consultant for Aventis Pharma Deutschland. R Kahn, R Schmitt, L Truedsson, and A-C Sjögren have no conflicts of interest.
Acknowledgments We thank Monica Heidenholm, Christina Möller, and Birgitta Gullstrand for excellent technical assistance, and Vibeke Horstmann for statistical assistance. Some of the work was presented at the 9th International Vasculitis/ANCA Workshop, April 12–15, 2000, Groningen, The Netherlands, and at the 12th Congress of the International Pediatric Nephrology Association, Seattle, Washington, USA, Sept 1–5, 2001. The work was supported by grants from the Swedish Research Council (06X–14008 to DK, 06X–13413 to HH, and 12631 to LT); the Knut and Alice Wallenberg Foundation, the Swedish Renal Foundation, the Anna–Lisa and Sven-Eric Lundgren Foundation for Medical Research, the Greta and Johan Kock Foundation, and the Swedish Society of Medicine (all to DK); the Crafoord foundation, the Royal Physiographic Society in Lund, and the Alfred Österlund Foundation (DK and HH); the Lars Hiertas Minne Foundation, the Tore Nilson Foundation, and the Åke Wiberg Foundation (HH); the Deutsche Forschungsgemeinschaft and the Fonds der Chemischen Industrie (WME); and the King Gustaf V 80th Birthday Fund (LT).
References 1
2
Colman RW, Schmaier AH. Contact system: a vascular biology modulator with anticoagulant, profibrinolytic, antiadhesive, and proinflammatory attributes. Blood 1997; 90: 3819–43. Gustafson EJ, Schmaier AH, Wachtfogel YT, Kaufman N, Kucich U, Colman RW. Human neutrophils contain and bind high molecular weight kininogen. J Clin Invest 1989; 84: 28–35.
THE LANCET • Vol 360 • August 17, 2002 • www.thelancet.com
For personal use. Only reproduce with permission from The Lancet Publishing Group.
MECHANISMS OF DISEASE
3 4 5
6
7
8
9
10
11
12
13
14 15 16
17
18 19
20
Mahabeer R, Bhoola KD. Kallikrein and kinin receptor genes. Pharmacol Ther 2000; 88: 77–89. Regoli D, Nsa Allogho S, Rizzi A, Gobeil FJ. Bradykinin receptors and their antagonists. Eur J Pharmacol 1998; 348: 1–10. van der Graaf F, Keus FJ, Vlooswijk RA, Bouma BN. The contact activation mechanism in human plasma: activation induced by dextran sulfate. Blood 1982; 59: 1225–33. Schapira M, Silver LD, Scott CF, et al. Prekallikrein activation and high-molecular-weight kininogen consumption in hereditary angioedema. N Engl J Med 1983; 308: 1050–53. Pixley RA, Colman RW. The kallikrein-kinin system in sepsis syndrome. In: Farmer SG, ed. Handbook of immunopharmacology: the kinin system. New York: Academic Press, 1997: 173–86. Mattsson E, Herwald H, Cramer H, Persson K, Sjobring U, Bjorck L. Staphylococcus aureus induces release of bradykinin in human plasma. Infect Immun 2001; 69: 3877–82. Stadnicki A, Gonciarz M, Niewiarowski TJ, et al. Activation of plasma contact and coagulation systems and neutrophils in the active phase of ulcerative colitis. Dig Dis Sci 1997; 42: 2356–66. Bergamaschini L, Parnetti L, Pareyson D, Canziani S, Cugno M, Agostoni A. Activation of the contact system in cerebrospinal fluid of patients with Alzheimer disease. Alzheimer Dis Assoc Disord 1998; 12: 102–08. Wachtfogel YT, Harpel PC, Edmunds LH Jr, Colman RW. Formation of C1s-C1-inhibitor, kallikrein-C1-inhibitor, and plasminalpha 2-plasmin-inhibitor complexes during cardiopulmonary bypass. Blood 1989; 73: 468–71. Kaplan AP, Kay AB, Austen KF. A prealbumin activator of prekallikrein, 3: appearance of chemotactic activity for human neutrophils by the conversion of human prekallikrein to kallikrein. J Exp Med 1972; 135: 81–97. Wachtfogel YT, Kucich U, James HL, et al. Human plasma kallikrein releases neutrophil elastase during blood coagulation. J Clin Invest 1983; 72: 1672–77. Bhoola KD, Figueroa CD, Worthy K. Bioregulation of kinins: kallikreins, kininogens, and kininases. Pharmacol Rev 1992; 44: 1–80. Jennette JC, Falk RJ. Antineutrophil cytoplasmic autoantibodies and associated diseases: a review. Am J Kidney Dis 1990; 15: 517–29. Savige J, Davies D, Falk RJ, Jennette JC, Wiik A. Antineutrophil cytoplasmic antibodies and associated diseases: a review of the clinical and laboratory features. Kidney Int 2000; 57: 846–62. Bank U, Ansorge S. More than destructive: neutrophil-derived serine proteases in cytokine bioactivity control. J Leukoc Biol 2001; 69: 197–206. Pereira HA. CAP37, a neutrophil–derived multifunctional inflammatory mediator. J Leukoc Biol 1995; 57: 805–12. Jennette JC, Falk RJ, Andrassy K, et al. Nomenclature of systemic vasculitides: proposal of an international consensus conference. Arthritis Rheum 1994; 37: 187–92. Kaufmann J, Haasemann M, Modrow S, Müller-Esterl W. Structural dissection of the multidomain kininogens: fine mapping of the target epitopes of antibodies interfering with their functional properties. J Biol Chem 1993; 268: 9079–91.
THE LANCET • Vol 360 • August 17, 2002 • www.thelancet.com
21 Hasan AHK, Cines DB, Herwald H, Schmaier AH, Müller-Esterl W. Mapping the cell binding site on high molecular weight kininogen domain 5. J Biol Chem 1995; 270: 19256–61. 22 Müller-Esterl W, Johnson D, Salvesen G, Barrett AA. Human kininogens. Methods Enzymol 1988; 163: 240–56. 23 Herwald H, Collin M, Müller-Esterl W, Björck L. Streptococcal cysteine proteinase releases kinins: a novel virulence mechanism. J Exp Med 1996; 184: 665–73. 24 Yang HY, Erdos EG, Levin Y. A dipeptidyl carboxypeptidase that converts angiotensin I and inactivates bradykinin. Biochim Biophys Acta 1970; 214: 374-76. 25 Buttery JE, Stuart S. Assessment and optimization of kinetic methods for angiotensin-converting enzyme in plasma. Clin Chem 1993; 39: 312–16. 26 Miller SA, Dykes DD, Polesky HF. A simple salting out procedure for extracting DNA from human nucleated cells. Nucleic Acids Res 1988; 16: 1215. 27 Verpy E, Couture-Tosi E, Eldering E, et al. Crucial residues in the carboxy-terminal end of C1 inhibitor revealed by pathogenic mutants impaired in secretion or function. J Clin Invest 1995; 95: 350–59. 28 Tapper H, Karlsson A, Mörgelin M, Flodgaard H, Herwald H. Secretion of heparin-binding protein from human neutrophils is determined by its localization in azurophilic granules and secretory vesicles. Blood 2002; 99: 1785–93. 29 Bjartell A, Laine S, Pettersson K, Nilsson E, Lovgren T, Lilja H. Time-resolved fluorescence in immunocytochemical detection of prostate-specific antigen in prostatic tissue sections. Histochem J 1999; 31: 45–52. 30 Volpe AR, Fontecchio G, Carmignani M. Regulatory role of bradykinin in the coronary and cerebral circulations and in systemic hemodynamics. Immunopharmacology 1999; 44: 87–92. 31 Gautam N, Olofsson AM, Herwald H, et al. Heparin-binding protein (HBP/CAP37): a missing link in neutrophil-evoked alteration of vascular permeability. Nat Med 2001; 7: 1123–27. 32 Zhao MH, Lockwood CM. Azurocidin is a novel antigen for antineutrophil cytoplasmic autoantibodies (ANCA) in systemic vasculitis. Clin Exp Immunol 1996; 103: 397–402. 33 Reddigari S, Silverberg M, Kaplan AP. Assembly of the human plasma kinin-forming cascade along the surface of vascular endothelial cells. Int Arch Allergy Immunol 1995; 107: 93–94. 34 Pullmann R Jr, Lukac J, Skerenova M, et al. Association between systemic lupus erythematosus and insertion/deletion polymorphism of the angiotensin converting enzyme (ACE) gene. Clin Exp Rheumatol 1999; 17: 593–96. 35 Yoshioka T, Xu YX, Yoshida H, Shiraga H, Muraki T, Ito K. Deletion polymorphism of the angiotensin converting enzyme gene predicts persistent proteinuria in Henoch-Schönlein purpura nephritis. Arch Dis Child 1998; 79: 394–99. 36 Dudley J, Afifi E, Gardner A, Tizard EJ, McGraw ME. Polymorphism of the ACE gene in Henoch-Schönlein purpura nephritis. Pediatr Nephrol 2000; 14: 218–20.
541
For personal use. Only reproduce with permission from The Lancet Publishing Group.
Mechanisms of disease
Contact-system activation in children with vasculitis
Robin Kahn, Heiko Herwald, Werner Müller-Esterl, Roland Schmitt, Ann-Christine Sjögren, Lennart Truedsson, Diana Karpman
Summary Background The contact system triggers the kallikrein-kinin cascade, liberating bradykinin from high-molecular-weight kininogen. Effectors of the contact system have proinflammatory and vasoactive properties. Vasculitis is a condition characterised by inflammation around vessel walls, leading to secondary tissue damage for which the underlying molecular mechanisms are poorly understood. Our aim was to investigate contact-system activation in children with vasculitis. Methods We compared 17 children, aged 4–19 years, with vasculitis, engaging the skin, joints, intestines, or kidneys, with 21 controls, aged 2–18 years. We analysed proteolysis of high-molecular-weight kininogen by immunoblotting. Plasma bradykinin concentrations were quantified by ELISA. Kidney and skin biopsies were stained in situ for kinins. Concentrations of heparin binding protein (HBP) were quantified by ELISA. Findings We noted extensive proteolysis of high-molecularweight kininogen in the plasma of 13 of 17 patients, but in only one of 21 controls (p<0·0001). Bradykinin concentrations were higher in the patients’ plasma (median 320 ng/L, range <1–19 680) than in plasma from controls (11 ng/L, <1–304; p=0·0004). Patients had local release of kinins at sites of inflammation in kidney and skin biopsies. HBP values were raised in patients (17·4 g/L, 5·4–237·6) compared with controls (6 g/L, 2·5–43·4; p=0·008). Interpretation Activation of the contact system could play a part in the pathogenesis of vasculitis, and explain the inflammation, pain, vasodilatation, and oedema seen in patients.
Lancet 2002; 360: 535–41
Departments of Paediatrics (R Kahn, R Schmitt MD, A-C Sjögren BSc, D Karpman MD), Cell and Molecular Biology (H Herwald PhD), and Laboratory Medicine (L Truedsson MD), Lund University, Lund, Sweden; and Institute of Biochemistry II, University Hospital, Frankfurt, Germany (Prof W Müller-Esterl PhD) Correspondence to: Dr Diana Karpman (e-mail: [email protected])
THE LANCET • Vol 360 • August 17, 2002 • www.thelancet.com
Introduction The contact system of human plasma consists of four components, plasma prekallikrein, high-molecular-weight kininogen, factor XI, and factor XII.1 Activation of the system is triggered by conversion of small quantities of factor XII to factor XIIa by endogenous or exogenous stimuli, and amplified by the mutual activation of factor XII and prekallikrein, resulting in a high local concentration of active plasma kallikrein, PKa, which is tethered to endothelial, subendothelial, and granulocyte surfaces via high-molecular-weight kininogen.1,2 Activated PKa rapidly cleaves high-molecular-weight kininogen into heavy and light chains, and liberates the terminal effector peptide, bradykinin, which binds to neighbouring kinin receptors of the B2 subtype and triggers various signalling cascades.1 Degradation of bradykinin by carboxypeptidase N or angiotensin converting enzyme (ACE) leads to generation of desArg9-bradykinin, which binds B1 kinin receptors.3 The B2 kinin receptor is constitutively expressed, whereas the B1 receptor is induced during inflammatory conditions by stimuli such as interleukin 1.4 Activity of the contact-system cascade is terminated in vivo by irreversible inhibiton of PKa by C1 inhibitor and ␣2-macroglobulin.1 The contact system can be triggered on cell membranes or on negatively charged artificial surfaces, such as kaolin and dextran sulphate.5 Activation of the contact system arises in several clinical conditions, including hereditary angio-oedema,6 bacteraemia, and sepsis,7,8 ulcerative colitis,9 Alzheimer’s disease,10 and during cardiopulmonary bypass,11 but has not been described in other conditions that involve systemic inflammation. The components of the contact system have multiple physiological effects. Plasma kallikrein is chemotactic for neutrophils12 and induces the secretion of neutrophil elastase.13 Bradykinin is a potent inflammatory mediator that reduces local blood pressure, causes capillary leakage and oedema, and induces pain.1 Bradykinin might stimulate endothelial cells to form superoxide, nitric oxide, and prostacyclin,14 which inhibits platelet aggregation. When vessels are damaged, bradykinin stimulates intimal hypertrophy and the proliferation of smooth muscle.1 Vasculitis is characterised by inflammation with neutrophil influx in and around vessel walls, leading to perturbed vessel patency and tissue damage. Injured vessels can leak plasma and blood cells, sometimes resulting in formation of thrombi. Various organs can be involved, including the kidney, intestine, skin, and joints. Patients with vasculitis exhibit symptoms such as abdominal pain purpura, joint pain and swelling, as well as oedema and other manifestations of renal dysfunction.15 The pathogenetic mechanisms of vasculitis have not been completely elucidated. Damaged endothelial cells and neutrophil-induced tissue injury are thought to be the underlying cause of pathology.16 Neutrophil components, such as proteinase 3 and heparin-binding protein, mediate 535
For personal use. Only reproduce with permission from The Lancet Publishing Group.
MECHANISMS OF DISEASE
GLOSSARY CARBOXYPEPTIDASE N
Carboxypeptidase N (arginine carboxypeptidase, kininase I) is an enzyme synthesised by the liver and secreted into the plasma. It is a kininase, which inactivates bradykinin and kallidin by removing carboxyterminal arginine from these peptides. Carboxypeptidase N also hydrolyses aminoacids of other peptides, including complement anaphylatoxins. CONTACT SYSTEM
The contact (kallikrein/kinin) system is composed of three serine proteinases: factor XII, factor XI, plasma prekallikrein (kallikrein in active form), and a non-enzymatic component, high-molecular-weight kininogen, which are synthesised by the liver and secreted into the plasma. Upon activation, the contact system is involved in the regulation of different procoagulant, profibrinolytic, antiadhesive, and proinflammatory processes, such as activation of the intrinsic pathway of coagulation, blood-pressure regulation, inhibition of thrombin activation of cells, fibrinolysis, and generation of kinins. ROCKET IMMUNOELECTROPHORESIS
A quantitative electrophoretic method for measurement of antigen, in which the antigen is placed on agar containing specific antiserum to the antigen being assayed. Antigens precipitate in rocket shape due to the electric current, and the area under the rocket is proportional to the antigen concentration.
inflammation.17,18 Certain patients with systemic vasculitis have serum antibodies directed against cytoplasmic components of neutrophils (antineutrophil cytoplasmic antibodies [ANCA]), such as proteinase 3 and myeloperoxidase. These antibodies are used in the diagnosis and management of vasculitis, since they can function as markers of disease activity.16 Theoretically, activation of the contact system could explain some of the characteristics of vasculitis— eg, inflammation, oedema, pain, vasodilatation, and local bleeding. Our aim was to assess systemic and local activation of the contact system in patients with vasculitis. We postulate that inappropriate activation of the contact system could contribute to the pathogenesis of this condition.
Methods Participants We included all children with acute and chronic forms of vasculitis, treated at the Department of Paediatrics, section of Paediatric Nephrology, University Hospital, Lund, Sweden, between June, 1999, and September, 2001. We defined subtypes of vasculitis in accord with the Chapel Hill nomenclature.19 We obtained control plasma samples from consecutive patients seen at the outpatient clinic of the same hospital between June and July, 1999, for follow-up and treatment of various diseases. Control serum samples were obtained from individuals seen at the outpatient clinic for follow up 2 months to 10 years after diarrhoea-associated haemolytic uraemic syndrome. All controls were asymptomatic with normal renal function at time of sampling. The ethics committee of Lund University approved the study, and all samples were taken with the informed consent of participants (older than age 15 years) or their parents (when younger than age 15 years). Protocol We obtained blood samples during or after acute episodes of vasculitis, or at follow up for chronic vasculitis. Blood samples were also available from some patients 1–2 years after recovery from the acute episode. Venous blood from patients and controls was collected in
536
5 mL vacutainer tubes, containing 0·5 mL 0·129 mol/L sodium citrate for plasma, and in 4 mL vacutainer Hemogard SST tubes for serum (plasma and serum tubes from Becton Dickinson, Plymouth, UK). We centrifuged the samples immediately after collection at 2000 g for 10 min. Platelet-poor plasma was removed in plastic pipettes and frozen at –80ºC until assayed. Serum was frozen at –20ºC until assayed. High-molecular-weight kininogen is a single-chain protein, consisting of six domains; domains D1–D3 represent the heavy chain region, domain D4 the bradykinin segment, and domains D5H and D6H the light chain region.1 Domains D1–D4 of low-molecular-weight kininogen are identical to those of high-molecular-weight kininogen, whereas the light chain region of lowmolecular-weight kininogen is composed of a single domain, D5L, which differs from D5H.1 We used domainspecific antibodies to D2, D5H, and D6H to discriminate between the various kininogen forms and fragments in immunoblots. We incubated citrated plasma from two healthy adult volunteers with dextran sulphate (molecular mass 5000, final concentration 75 mol/L; Sigma, St Louis, MO, USA) in the presence of 2 mmol/L zinc chloride (Sigma) at 37ºC for 1 h, to induce contactphase activation and proteolysis of high-molecular-weight kininogen.5 In separate tubes, we treated the same plasma samples with 2 mmol/L zinc chloride without dextran sulphate at 37ºC for 1h. We diluted plasma samples 1/100 in 4% (w/v) SDS (Bio Rad, Hercules, CA, USA) in Tris buffer, pH 6·8 (ICN Biomedicals, Aurora, OH, USA), containing 10% (v/v) 2-mercaptoethanol (Kebo Lab, Spånga, Sweden), and boiled them for 5 min. We ran 10 L per sample on a 10% SDS-PAGE gel in an electrophoresis chamber (Bio Rad). The gels were electroblotted for 1 h onto 0·45 m Protran nitrocellulose membranes (Schleicher and Schuell, Dassel, Germany) with a semidry electroblotter (Ancos, Hoejby Denmark). We blocked the membranes at 37ºC for 45 min with 5% dry milk powder and 0·05% Tween 20 in PBS (pH 7·4, 0·01 mol/L phosphate buffer; Life Technologies, Uppsala, Sweden). Immunoblots were done with the following primary antibodies: monoclonal antibody HKH8 (from mouse) directed to common domain D2 of high-molecularweight and low-molecular-weight kininogen20 was used at 1/1000; polyclonal antibody to HKH20 (␣-HKH20; from rabbit) directed to high-molecular-weight kininogen domain D5H21 was used at 1/500; monoclonal antibody HKL22 directed to high-molecular-weight kininogen domain D6H was used at 1/500;20 and antiserum AS88 (from sheep) against purified human high-molecularweight kininogen, which crossreacts with low-molecularweight kininogen was used at 1/5000.22 Secondary antibodies were applied as appropriate: goat antibody against mouse/horseradish peroxidase (HRP, 1/3000), goat antibody against rabbit/HRP (1/3000, both antibodies from Dako Glostrup, Denmark), and donkey antibody against sheep/HRP (1/3000, ICN Biomedicals). We assayed proteolytic processing of high-molecularweight kininogen by immunoblotting as previously described.20 Citrated plasma samples from patients and controls were separated by SDS-PAGE and electrotransferred as described above. They were further incubated for 45 min with antiserum AS88. After washing in PBS/Tween, containing 0·35 mol/L sodium chloride, we incubated the membranes for 45 min with donkey antibody against sheep immunoglubulin/HRP (1/3000), and detected the signal by chemiluminescence. As a positive control for high-molecular-weight kininogen
THE LANCET • Vol 360 • August 17, 2002 • www.thelancet.com
For personal use. Only reproduce with permission from The Lancet Publishing Group.
MECHANISMS OF DISEASE
proteolysis, we added dextran sulphate to normal human plasma from an adult. Plasma was incubated with dextran sulphate in the presence of 2 mmol/L zinc at a final concentration of 75 mol/L at 37ºC for 1h. We detected bradykinin with the MARKIT-M Bradykinin ELISA kit (Dainippon Pharmaceutical, Osaka, Japan) as per the manufacturers instructions and as previously described.23 This analysis requires 500 L of plasma and could, therefore, not be tested in all participants because of lack of sample. The antibody against bradykinin used in this kit is specific for free bradykinin and does not cross react with high-molecularweight or low-molecular-weight kininogen. ACE cleaves and inactivates bradykinin.24 We therefore investigated if increased concentrations of bradykinin in patients with vasculitis were due to reduced ACE activity. We analysed ACE activity in sera from patients, and from controls who had recovered from haemolytic uraemic syndrome (since serum was not available from the other paediatric controls). A spectrophotometric assay based on hydrolysis of the substrate furanacryloyl-Lphenylalanylglycylglycine was used as previously described.25 The substrate was obtained from Bachem (Bubendorf, Switzerland). C1 inhibitor is the major inhibitor of the contact system. Deficiency or dysfunction of C1 inhibitor leads to contact system activation and also to reduced concentrations of C4. For this reason, we measured serum concentrations of C1 inhibitor and C4 in patients by rocket immunoelectrophoresis and turbidimetry, respectively. We extracted DNA from the blood of some patients with the salting-out procedure.26 Two DNA fragments were amplified, a 656 bp fragment of exon 8 of the C1 inhibitor gene, using primers E1 (biotinylated) and E2,27 and a 480 bp fragment, using primers E1 and E2b (5´GCTGACAGAGGACCCAGATC3´). We did the PCR reaction with a Minicycler (MJ Research, MA, Patient number
Sex
1 2 3 4
F F F F
5
F
6
M
7 8 9 10 11 12 13
F M F F F M M
14 15
F F
16 17
M F
Diagnosis
SLE Acute HSP* Acute HSP C-ANCA positive vasculitis†‡ C-ANCA positive vasculitis‡ Microscopic polyangitis Acute HSP§ Chronic HSP Acute HSP§ Status post HSP¶ Acute HSP§ Acute HSP Acute leukocytoclastic vasculitis Acute HSP Wegener’s granulomatosis|| Acute HSP Chronic HSP
USA) with an annealing temperature of 60ºC. The reaction mixture (100 L volume) contained 1 g of genomic DNA, 1 mol/L of each primer, 200 mol/L dNTP, and 0·03 units/L of AmpliTaq Gold polymerase and buffer (Perkin-Elmer, Norwalk, CT, USA). To separate the DNA strands, immobilisation of the PCR products was done, according to a protocol from Dynal AS (Oslo, Norway). We used Sequenase (version 2.0, USB, Cleveland, OH, USA) for the sequencing reaction, and then loaded samples onto a sequencing gel. After drying, we visualised the bands by autoradiography. Heparin binding protein (HBP), also known as azurocidin, is a glycoprotein secreted from azurophilic granules and secretory vesicles of neutrophils.28 HBP functions as an inflammatory mediator.18 We measured the concentrations of the protein in plasma of patients and controls by ELISA as previously described.28 We obtained renal cortical and skin biopsies for medical purposes from the children with vasculitis. Renal cortical and skin biopsy material was also available from two adults biopsied for clear-cell carcinoma (kidney biopsy) and breast cancer (skin biopsy). We used these samples as controls, since they contained areas of histologically normal tissue adjacent to cancerous areas. We took biopsies during acute episodes of vasculitis, and examined them for inflammatory and vascular changes. We also examined renal specimens for glomerular, tubular, and interstitial changes. Samples were tested for the presence of kinin peptides with a rabbit antibody that reacts with bradykinin, lysylbradykinin (kallidin), and desArg9-bradykinin (Biogenesis, Poole, UK), and the streptavidin-HRPconjugate method as previously described.29 We incubated samples at room temperature overnight with antibody against kinin diluted 1/400 in Tris buffer (containing 2·5% [w/v] of bovine serum albumin) for kidney samples, and at 1/800 or 1/1600 for skin samples. Positive samples stained brown. As a control, the primary
Age Clinical manifestation (years) R F A GI
GN
ESR
Laboratory findings CRP
WBC
SA
LFT
Cr
UB
UP
ANCA
19 11 6 11
+ + + +
– + – –
– + + +
– – – –
– – + –
↑ ·· ·· ··
·· ·· ·· ··
·· ·· ·· ··
·· ↓ ↓ ··
·· ·· ·· ··
·· ·· ·· v
– – + +
+ – + –
NA – – +PR-3
14
+
–
+
–
–
↑
··
··
··
··
↑
–
–
+PR-3
13
+
–
+
+
+
↑
↑
↑
↓
↑
↑
+
+
–
4 16 6 18 10 12 12
+ + + – + + +
– – – – – – –
+ + + – + + +
–+ – – + + +
– + – + – – –
·· ·· ·· ·· ·· ·· ↑
·· ·· ↑ ·· ·· ↑ ··
·· ·· ·· ·· ↑ ↑ ··
·· ·· ↓ ·· ·· ·· ··
·· ·· ·· ·· ·· ·· ··
·· ·· ·· ↑ ↑ ·· ↑
·· + – ·· + + –
·· + – ·· + + –
– – NA – +elastase NA –
16 12
+ –
– +
+ +
– –
+ +
·· ↑
↑ ↑
·· ↑
↓ ↓
·· ··
·· ↑
– +
+ +
– +PR-3 and MPO
15 14
+ +
+ –
– –
+ –
+ –
↑ ··
·· ··
·· ··
↓ ··
·· ··
↑ ··
+ +
+ –
– +MPO
NA=not available. F=female. M=male. R=rash. F=fever. A=arthralgia or arthritis. GI=gastrointestinal pain or bleeding. GN=glomerulonephritis verified by renal biopsy. ESR=erythrocyte sedimentation rate. CRP=C-reactive protein. WBC=white blood cell count. SA=serum albumin. LFT=liver function tests. Cr=creatinine. UB=urine blood. UP=urine protein. PR-3=proteinase 3. MPO=myeloperoxidase. SLE=systemic lupus erythematosus. HSP=Henoch Schönlein purpura. ··=Normal laboratory results. *Chronic purpura and infectious episodes after HSP. †Patient had raised circulating immune complex concentration. ‡Patients are siblings. §Patients had mild transient symptoms and recovered within days. ¶Patient had meningococcal sepsis several years after HSP and 1 year before sampling. ||Patient had hearing loss, cough, and fatigue in addition to other symptoms.
Table 1: Clinical and laboratory data of children with vasculitis at time of sampling
THE LANCET • Vol 360 • August 17, 2002 • www.thelancet.com
537
For personal use. Only reproduce with permission from The Lancet Publishing Group.
MECHANISMS OF DISEASE
antibody against kinin was excluded from the protocol. Antibody specificity assay was tested by pretreating the antibody (1/800 and 1/1600) with bradykinin (50 g/L, Dainippon) or des-arg9-bradykinin (10 and 100 mg/L, Biogenesis) for 1h at 37ºC. These combinations were then separately added to skin sections from two patients. The antibody against kinin did not cross-react with highmolecular-weight kininogen on immunoblotting (data not shown). Statistical analysis We assessed differences between patients with vasculitis and controls with respect to breakdown of highmolecular-weight kininogen by Fisher’s exact test. Differences between these groups with respect to bradykinin or HBP concentrations and ACE activity were tested with the Mann Whitney U test. We used the Sign test to ascertain differences between patients during the acute phase of disease and convalescence with respect to breakdown of high-molecular-weight kininogen. We judged a p value of 0·05 or less significant. We used GraphPad Instat (version 2.0) for statistical analyses. Role of the funding source The sponsors of the study had no role in study design, data collection, data analysis, data interpretation, writing of the paper, or in the decision to submit the report. AS88 HK HK*
116 kDa
66 kDa 58 kDa
No rm al pl as Pl de as ma xt ma ra w n i su th lfa te
45 kDa
␣-HKH20 HKL22
HKH8 HK
HK
LK and H63 L58
L58 L45
Results We enrolled 17 patients, five boys and 12 girls, aged 4–19 years (median 12·5). Table 1 shows a summary of their clinical and laboratory data at the time of sampling. The children had various symptoms, including purpura, vasculitic rash, arthritis, arthralgia, abdominal pain, haematuria, and oedema. One child with Wegener’s granulomatosis had, in addition, upper-respiratory-tract symptoms. Two children had chronic Henoch-Schönlein purpura manifested as constant rash and haematuria for more than 2 years. None of the patients were receiving treatment with immunosuppressives or ACE inhibitors at the time of sampling. We took blood samples during (n=14) or after (n=1) acute vasculitis, or at follow up for chronic vasculitis (n=2). We obtained control plasma samples from 21 children, nine boys and 12 girls, aged 2–18 years (median 9). The children were receiving outpatient treatment for diabetes mellitus (n=4), pyelonephritis (n=6), neuroblastoma (n=1), thrombotic thrombocytopenic purpura during remission (n=2), nephrotic syndrome (n=1), anal atresia (n=1), epilepsy (n=3), pigmented purpura (n=1), migraine (n=1), and psychiatric disorder (n=1). Control serum samples were obtained from 14 children, six boys and eight girls, aged 2–18 years (median 5) who had had diarrhoeaassociated haemolytic uraemic syndrome. Figure 1 shows the products produced after proteolysis of high-molecular-weight kininogen. High-molecularweight kininogen was degraded in the plasma of 13 of 17 patients but in only one of 21 controls (Fisher’s exact test p<0·0001; figure 2 and table 2). This control had migraine. Convalescent plasma was available for eight patients (numbers 3, 4, 6, 9, 12, 13, 14, and 15) 1–2 years after the acute episode of vasculitis. Of these eight patients, high-molecular-weight kininogen was degraded in five during the acute phase of disease, but normalised after recovery; was not degraded in two during the acute phase or at convalescence; and was weakly positive in one (patient 13) at convalescence (data not shown). Differences between patients during the acute phase of disease and after recovery were non-significant (Sign test p=0·22). Concentrations of bradykinin were raised in patients (median 320 ng/L, range <1–19 680, IQR 17–655·5; table 2) compared with controls (11 ng/L, <1–304, IQR <1–174, Mann Whitney U test p=0·0004). The control with migraine had a bradykinin concentration of 11 ng/L. Two controls could not be tested because of insufficient plasma. Patients with vasculitis
kDa 116
NH2
D1
D2
D3
D4
D5
D6
Figure 1: Immunoblot of proteolysis products of high-molecularweight kinogen (HK) Western blot of normal plasma and plasma stimulated with dextran sulphate, leading to consumption of all intact HK (upper). The antiserum AS88 against kininogen revealed two major bands of 63–68 kDa and 58 kDa, a minor band of 45 kDa, and a weak band of 90 kDa (HK*=terminally truncated HK). Immunoblot of HK and its fragments with domain-specific antibodies to discriminate between the various forms and degradation fragments (lower). Normal plasma was incubated with dextran sulphate, separated by SDS-PAGE, and immunoblotted with antibodies against domain 2 of the HK and low-molecular-weight kinogen (LK) heavy chain (HKH8) or domains 5 and 6 of the HK light chain (a-HKH20 and HKL22, respectively). H63=common heavy chain of 63 kDa. L58/L45=HK light chain forms of 58 kDa and 45 kDa, respectively. The 45 kDa band is a truncated form of HK light chain. HK domains are depicted at bottom of figure.
538
66 58 45
COOH DXS 1
2
3 4
5
6
7
8 9 10 11 12 13 14 15 16 17
Controls 116 66 58 45 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38
Figure 2: Immunoblot of plasma samples from patients (upper) and controls (lower) with antiserum against kininogen AS88 DXS=dextran sulphate. Kininogen and their fragments as per figure 1.
THE LANCET • Vol 360 • August 17, 2002 • www.thelancet.com
For personal use. Only reproduce with permission from The Lancet Publishing Group.
MECHANISMS OF DISEASE
Patient number
HK proteolysis
Bradykinin (ng/L)
HBP (g/L)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
+ +* +* +* +* – – +* + +* + –‡ –* + + + +
620 156* 550* 340* 472* 189 <1† 256* 184 299* 81 NA 134* 19 680 1488 3360 691
13·9 7·8 7·2 5·4 6·5 237·6 8·9 25·3 221·8 16·7 18·2 NA 39·6 107·2 26·7 6·1 51·4
NA=not available. *Two or more samples with similar results taken at different time points available. †Below the detection limit. ‡Plasma sufficed only for analysis of HK proteolysis.
Table 2: Proteolysis of high-molecular-weight-kininogen (HK) and bradykinin and HBP concentrations in patients
ACE activity in the sera of 14 patients (excluding patients 6, 9, and 17) was in the range 16–61 U/mL (median 38, IQR 27–51). Activity in sera of 14 controls was in the range 27–66 U/mL (48, 43–57, Mann Whitney U test p=0.053). Thus, we cannot rule out the possibility that increased concentrations of bradykinin were associated with reduced degradation due to reduced ACE activity. Inability to inhibit PKa because of a mutated C1 inhibitor protein or the presence of neutralising autoantibodies to native C1 inhibitor causes angio-oedema in patients with excessively high concentrations of bradykinin. We therefore investigated the C1 inhibitor and C4 concentrations in 15 patients (not in patients 5 and 9). The concentrations of C1 inhibitor were 92–199% (median 125; % standardised by concentrations obtained from 100 blood donors); four patients had values above the applied reference interval of 72–153%. The C4 concentrations were 0·07–0·44 g/L (0·24; reference interval 0·1–0·4 g/L). Patient 6 had reduced concentrations of C4 during the early stage of disease, which later normalised. Patient 2 had raised C4; all other patients had normal C4 concentrations. DNA fragments (from patients 2, 6, 8, and 13) of exon 8 of the C1 inhibitor gene, corresponding to positions 400–476 of the aminoacid sequence, were read from the gel. No mutations were present in this sequence, which contains the reactive site. Since concentrations of C1 inhibitor were normal and no mutation in the reactive site was found, it is unlikely that the patients have a C1 inhibitor dysfunction. We measured concentrations of HBP in the plasma of patients, as a marker of neutrophil activation and granular shedding.28 Plasma samples from 16 children with vasculitis (not patient 12) were compared with those of 17 controls. Patients’ HBP concentrations were in the range 5·4–237·6 g/L (median 17·4, IQR 7·5–45·5; table 2). Controls had concentrations in the range 2·5–43·4 g/L (6·0, 4·4–15·7, Mann Whitney U test p=0·008). These results indicate neutrophil activation in patients with vasculitis. Kidney biopsies from seven patients (3, 6, 8, 10, 14, 15, 16) showed diffuse mesangioproliferative glomerulonephritis with formation of crescents. Skin biopsies from patients 5, 13, 16, and 17 showed swollen endothelial cells with infiltrates of granulocytes, including eosinophils, in and around vessel walls and in the interstitium. Extravasation of red blood cells was noted.
THE LANCET • Vol 360 • August 17, 2002 • www.thelancet.com
Figure 3: In-situ immunohistochemistry of kidney and skin biopsies labelled with antibody against kinin A=kidney biopsy from patient 3. Prominent brown staining of glomeruli and tubuli, demonstrating the presence of kinin peptides is noted. Arrow points to glomerular capillary staining. B=control kidney biopsy, showing no staining. C=skin biopsy from patient 13 with acute leukocytoclastic vasculitis. Arrow shows staining in areas of inflammatory influx. D=control skin biopsy, showing weak diffuse staining of the epidermis.
Tissues from patients with vasculitis were examined to ascertain whether or not proteolytic products of the contact system were present in the perivascular and inflammatory areas. Sections taken from patients stained more intensively with antibodies against kinin than did sections from two controls. Kidney biopsies from patients 3, 8, 15, and 16 were examined with the antibody against kinin and had pronounced staining in glomeruli, in which a specific staining of the luminal side of the capillaries and of Bowman’s capsule was noted (figure 3, A). We noted granular staining in the proximal and distal tubuli. In the four skin biopsies, intense staining was observed at the site of inflammatory influx (figure 3, C). In sections from patients and controls, a diffuse staining was seen in the epidermis (figure 3, D). Preincubation of the primary antibody (1/800 and 1/1600) with bradykinin partially inhibited the reaction in two skin biopsies. Figure 4 shows the specificity of the assay for bradykinin. We noted similar results when the primary antibody was preincubated with desArg9-bradykinin, indicating that the antibody does not discriminate between these two active kinin peptides. There was no staining when the primary antibody was omitted.
Figure 4: Specificty of in-situ immunohistochemistry of skin biopsy from patient 16 A=section labeled with antibody against kinin (1/1600). B=antibody (1/1600) preincubated with bradykinin 50 g/L for 1 h at 37ºC before labelling to inhibit specific staining (brown labelling). Arrows point to areas of inflammatory infiltrate with specific granular staining in section A, which diminished in section B. Several sections were taken, 4 m apart. The sections shown in panel A and B show the same area and some of the same cells.
539
For personal use. Only reproduce with permission from The Lancet Publishing Group.
MECHANISMS OF DISEASE
Discussion Our results indicate that, in children with vasculitis, high-molecular-weight kininogen is extensively proteolysed and bradykinin concentrations are greatly raised. Furthermore, kinin peptide is present in inflamed tissues. Bradykinin is known to have potent proinflammatory properties;1 our findings show that it is present at the site of inflammation during systemic vasculitis, suggesting that bradykinin could contribute to the inflammatory state in vasculitis. The contact system is activated in other inflammatory and infectious conditions.7–9 Reduced concentrations of high-molecular-weight kininogen and prekallikrein have been shown in patients with systemic inflammatory response syndrome,7 in which inflammation of multiple organs is a prominent feature. Since bradykinin is raised in a range of inflammatory diseases, its release is probably a distal event in the inflammatory cascade and could be triggered by various different mechanisms. The disease spectrum in children with vasculitis differs from that of adults—eg, children generally have milder forms of vasculitis, such as Henoch Schönlein purpura, whereas adults tend to have severe and chronic ANCAassociated disease. We did not address the possibility of contact-system activation in adults, however, our results indicate that the contact system was often activated in children with chronic symptoms. Furthermore, some patients who had undergone acute Henoch Schönlein purpura and later developed chronic symptoms such as purpura or severe infection continued to have contactsystem activation even after the initial acute event. This finding could suggest that the inflammatory process was ongoing in these patients. We also noted increased breakdown of high-molecular-weight kininogen in one paediatric control with migraine. Although activation of the contact system has not been previously associated with this vascular disorder, intra-arterial administration of bradykinin induces pain in these patients.30 Studies in additional patients with migraine are required to investigate the association between contact-system activation and pain in this condition. Results of in-vitro studies have shown that highmolecular-weight kininogen can be activated on the surface of endothelial cells and granulocytes.1 These cells are involved in the pathological process during vasculitis, and either or both may secrete agents that trigger the disease. Since antibodies directed to cytoplasmic components of neutrophils (ANCA) are seen in certain subgroups of vasculitides, it has been postulated that these cytoplasmic components, or the antibodies themselves, could be essential for the development of vasculitis.15 HBP is a glycoprotein localised in neutrophil azurophilic granules and secretory vesicles, which is a chemoattractant for monocytes and T cells 18 and has known antimicrobial properties. This glycoprotein induces vascular permeability.31 Increased concentrations of HBP have previously been measured in the sera of patients with vasculitis.32 We also noted raised plasma concentrations of HBP in our patients, indicating neutrophil activation and degranulation. Bradykinin was increased in plasma and tissue of patients. Bradykinin can be produced on the surface of various cell types, such as endothelial cells and neutrophils.1 Previous studies have shown that bradykinin can be generated on the luminal side of endothelial cells, which could explain its effect on vascular permeability.33 Cell surfaces that do not produce bradykinin under normal circumstances, such
540
as glomerular endothelial cells and cells in the Bowman’s capsule, also stained positively for kinin peptide in our study. We postulate that there is proteolysis of high-molecular-weight kininogen, generating release of bradykinin, independently in the plasma and in the tissues of patients. Results of investigations have indicated an association between ACE polymorphisms and certain types of vasculitis, such as systemic lupus erythematosus34 and Henoch Schönlein purpura.35 However, these results have been challenged.36 We noted that differences in ACE activity between patients and controls were not significant, though almost. Whether or not reduced ACE activity in vasculitis patients contributes to raised concentrations of bradykinin, therefore, remains unclear and should be addressed in future studies. Our findings show that the contact system is activated in children with vasculitis. Irrespective of whether activation is a primary or secondary event, we suggest that the production of bradykinin could play an important part in the pathogenesis of vasculitis. Bradykinin B1 or bradykinin B2 receptor antagonists, or both,4 could, thus, be useful in reducing inflammatory symptoms in patients with vasculitis. Contributors R Kahn was responsible for planning, did experiments, data analysis, and wrote the report. H Herwald participated in the experimental planning, did experiments, and contributed to the interpretation of data and write-up. W Müller-Esterl helped with study design, data interpretation, and write-up. R Schmitt assisted with experimental planning, did experiments, and interpreted data. A-C Sjögren participated in experimental planning and work as well as data interpretation. L Truedsson assisted with complement analysis, PCR, and data interpretation. D Karpman contributed to the experimental conception, design, work, and patient care, as well as to data interpretation and write-up.
Conflict of interest statement D Karpman has received grants from Centeon Pharma, and travel grants from Roche Pharmaceuticals and Janssen-Cilag Pharmaceuticals. H Herwald was employed by Ferring. W Müller-Esterl is a consultant for Aventis Pharma Deutschland. R Kahn, R Schmitt, L Truedsson, and A-C Sjögren have no conflicts of interest.
Acknowledgments We thank Monica Heidenholm, Christina Möller, and Birgitta Gullstrand for excellent technical assistance, and Vibeke Horstmann for statistical assistance. Some of the work was presented at the 9th International Vasculitis/ANCA Workshop, April 12–15, 2000, Groningen, The Netherlands, and at the 12th Congress of the International Pediatric Nephrology Association, Seattle, Washington, USA, Sept 1–5, 2001. The work was supported by grants from the Swedish Research Council (06X–14008 to DK, 06X–13413 to HH, and 12631 to LT); the Knut and Alice Wallenberg Foundation, the Swedish Renal Foundation, the Anna–Lisa and Sven-Eric Lundgren Foundation for Medical Research, the Greta and Johan Kock Foundation, and the Swedish Society of Medicine (all to DK); the Crafoord foundation, the Royal Physiographic Society in Lund, and the Alfred Österlund Foundation (DK and HH); the Lars Hiertas Minne Foundation, the Tore Nilson Foundation, and the Åke Wiberg Foundation (HH); the Deutsche Forschungsgemeinschaft and the Fonds der Chemischen Industrie (WME); and the King Gustaf V 80th Birthday Fund (LT).
References 1
2
Colman RW, Schmaier AH. Contact system: a vascular biology modulator with anticoagulant, profibrinolytic, antiadhesive, and proinflammatory attributes. Blood 1997; 90: 3819–43. Gustafson EJ, Schmaier AH, Wachtfogel YT, Kaufman N, Kucich U, Colman RW. Human neutrophils contain and bind high molecular weight kininogen. J Clin Invest 1989; 84: 28–35.
THE LANCET • Vol 360 • August 17, 2002 • www.thelancet.com
For personal use. Only reproduce with permission from The Lancet Publishing Group.
MECHANISMS OF DISEASE
3 4 5
6
7
8
9
10
11
12
13
14 15 16
17
18 19
20
Mahabeer R, Bhoola KD. Kallikrein and kinin receptor genes. Pharmacol Ther 2000; 88: 77–89. Regoli D, Nsa Allogho S, Rizzi A, Gobeil FJ. Bradykinin receptors and their antagonists. Eur J Pharmacol 1998; 348: 1–10. van der Graaf F, Keus FJ, Vlooswijk RA, Bouma BN. The contact activation mechanism in human plasma: activation induced by dextran sulfate. Blood 1982; 59: 1225–33. Schapira M, Silver LD, Scott CF, et al. Prekallikrein activation and high-molecular-weight kininogen consumption in hereditary angioedema. N Engl J Med 1983; 308: 1050–53. Pixley RA, Colman RW. The kallikrein-kinin system in sepsis syndrome. In: Farmer SG, ed. Handbook of immunopharmacology: the kinin system. New York: Academic Press, 1997: 173–86. Mattsson E, Herwald H, Cramer H, Persson K, Sjobring U, Bjorck L. Staphylococcus aureus induces release of bradykinin in human plasma. Infect Immun 2001; 69: 3877–82. Stadnicki A, Gonciarz M, Niewiarowski TJ, et al. Activation of plasma contact and coagulation systems and neutrophils in the active phase of ulcerative colitis. Dig Dis Sci 1997; 42: 2356–66. Bergamaschini L, Parnetti L, Pareyson D, Canziani S, Cugno M, Agostoni A. Activation of the contact system in cerebrospinal fluid of patients with Alzheimer disease. Alzheimer Dis Assoc Disord 1998; 12: 102–08. Wachtfogel YT, Harpel PC, Edmunds LH Jr, Colman RW. Formation of C1s-C1-inhibitor, kallikrein-C1-inhibitor, and plasminalpha 2-plasmin-inhibitor complexes during cardiopulmonary bypass. Blood 1989; 73: 468–71. Kaplan AP, Kay AB, Austen KF. A prealbumin activator of prekallikrein, 3: appearance of chemotactic activity for human neutrophils by the conversion of human prekallikrein to kallikrein. J Exp Med 1972; 135: 81–97. Wachtfogel YT, Kucich U, James HL, et al. Human plasma kallikrein releases neutrophil elastase during blood coagulation. J Clin Invest 1983; 72: 1672–77. Bhoola KD, Figueroa CD, Worthy K. Bioregulation of kinins: kallikreins, kininogens, and kininases. Pharmacol Rev 1992; 44: 1–80. Jennette JC, Falk RJ. Antineutrophil cytoplasmic autoantibodies and associated diseases: a review. Am J Kidney Dis 1990; 15: 517–29. Savige J, Davies D, Falk RJ, Jennette JC, Wiik A. Antineutrophil cytoplasmic antibodies and associated diseases: a review of the clinical and laboratory features. Kidney Int 2000; 57: 846–62. Bank U, Ansorge S. More than destructive: neutrophil-derived serine proteases in cytokine bioactivity control. J Leukoc Biol 2001; 69: 197–206. Pereira HA. CAP37, a neutrophil–derived multifunctional inflammatory mediator. J Leukoc Biol 1995; 57: 805–12. Jennette JC, Falk RJ, Andrassy K, et al. Nomenclature of systemic vasculitides: proposal of an international consensus conference. Arthritis Rheum 1994; 37: 187–92. Kaufmann J, Haasemann M, Modrow S, Müller-Esterl W. Structural dissection of the multidomain kininogens: fine mapping of the target epitopes of antibodies interfering with their functional properties. J Biol Chem 1993; 268: 9079–91.
THE LANCET • Vol 360 • August 17, 2002 • www.thelancet.com
21 Hasan AHK, Cines DB, Herwald H, Schmaier AH, Müller-Esterl W. Mapping the cell binding site on high molecular weight kininogen domain 5. J Biol Chem 1995; 270: 19256–61. 22 Müller-Esterl W, Johnson D, Salvesen G, Barrett AA. Human kininogens. Methods Enzymol 1988; 163: 240–56. 23 Herwald H, Collin M, Müller-Esterl W, Björck L. Streptococcal cysteine proteinase releases kinins: a novel virulence mechanism. J Exp Med 1996; 184: 665–73. 24 Yang HY, Erdos EG, Levin Y. A dipeptidyl carboxypeptidase that converts angiotensin I and inactivates bradykinin. Biochim Biophys Acta 1970; 214: 374-76. 25 Buttery JE, Stuart S. Assessment and optimization of kinetic methods for angiotensin-converting enzyme in plasma. Clin Chem 1993; 39: 312–16. 26 Miller SA, Dykes DD, Polesky HF. A simple salting out procedure for extracting DNA from human nucleated cells. Nucleic Acids Res 1988; 16: 1215. 27 Verpy E, Couture-Tosi E, Eldering E, et al. Crucial residues in the carboxy-terminal end of C1 inhibitor revealed by pathogenic mutants impaired in secretion or function. J Clin Invest 1995; 95: 350–59. 28 Tapper H, Karlsson A, Mörgelin M, Flodgaard H, Herwald H. Secretion of heparin-binding protein from human neutrophils is determined by its localization in azurophilic granules and secretory vesicles. Blood 2002; 99: 1785–93. 29 Bjartell A, Laine S, Pettersson K, Nilsson E, Lovgren T, Lilja H. Time-resolved fluorescence in immunocytochemical detection of prostate-specific antigen in prostatic tissue sections. Histochem J 1999; 31: 45–52. 30 Volpe AR, Fontecchio G, Carmignani M. Regulatory role of bradykinin in the coronary and cerebral circulations and in systemic hemodynamics. Immunopharmacology 1999; 44: 87–92. 31 Gautam N, Olofsson AM, Herwald H, et al. Heparin-binding protein (HBP/CAP37): a missing link in neutrophil-evoked alteration of vascular permeability. Nat Med 2001; 7: 1123–27. 32 Zhao MH, Lockwood CM. Azurocidin is a novel antigen for antineutrophil cytoplasmic autoantibodies (ANCA) in systemic vasculitis. Clin Exp Immunol 1996; 103: 397–402. 33 Reddigari S, Silverberg M, Kaplan AP. Assembly of the human plasma kinin-forming cascade along the surface of vascular endothelial cells. Int Arch Allergy Immunol 1995; 107: 93–94. 34 Pullmann R Jr, Lukac J, Skerenova M, et al. Association between systemic lupus erythematosus and insertion/deletion polymorphism of the angiotensin converting enzyme (ACE) gene. Clin Exp Rheumatol 1999; 17: 593–96. 35 Yoshioka T, Xu YX, Yoshida H, Shiraga H, Muraki T, Ito K. Deletion polymorphism of the angiotensin converting enzyme gene predicts persistent proteinuria in Henoch-Schönlein purpura nephritis. Arch Dis Child 1998; 79: 394–99. 36 Dudley J, Afifi E, Gardner A, Tizard EJ, McGraw ME. Polymorphism of the ACE gene in Henoch-Schönlein purpura nephritis. Pediatr Nephrol 2000; 14: 218–20.
541
For personal use. Only reproduce with permission from The Lancet Publishing Group.