Paris
et al.
J. ALLERGY
liquid chromatography (HPLC) [Abstract]. J ALLERGY CLIN IMMUNOL
1990;85:150.
25. Smestad-Paulsen B, Flo L, Nesje G, Wold JK. Allergens in pollen from mugwort (Arfemisia vulgaris L.). I. Partial characterization of allergen preparations from mugwort pollen with emphasis on the carbohydrate moiety. Int Arch Allergy Appl Immunol 1985;78:206-12. 26. Jones OT, Earnest JP, McNamee MG. Solubilization and reconstitution of membrane proteins. In: Findlay JBC, Evans WH, ed. Biological membranes: a practical approach. Oxford: IRL Press, 1987:139-77.
CLIN. IMMUNOL. DECEMBER 1991
27. Jaggi KS, Gangal SV. Purification and characterization of allergens from Xunthium srrumarium pollen. Mol Cell Biochem 1987;78:177-90. 28. Bush RK, Yunginger JW. Standardization of fungal allergens. Clin Rev Allergy 1987;5:3-21. 29. Barnes C, Pacheco F, Portnoy J. Carbohydrate and protein contribution to Alrernariaallergen activity [Abstract]. J ALLERGY CLIN IMMLJNOL
Acute consumption of Cl inhibitor patient with acquired Cl-inhibitor deficiency syndrome Bruce L. Zuraw, MD,* and Leonard C. Altman,
1990;85: 169.
in a
MD**
La Jolla, Calif., and Seattle, Wash. Acquired Cl-inhibitor (Cl INH) deficiency is usually found in association with an underlying disease that is believed to be responsible for increased Cl INH catabolism, ultimately leading to the development of Cl INH dejiciency. We report a remarkable patient with acquired Cl INH deficiency in whom a unique progression of complement- and contact-system abnormalities has been observed. S. G. suffers from recurrent episodes of angioedema and hypotension. Results of repeated complement studies were initially normal, and the patient was diagnosed as having idiopathic anaphylaxis. Two years later, the patient was found to develop acute consumption of CI INH with activation of the complement and contact systems during episodes of angioedema. The patient continued to have normal Cl INH levels and to have no evidence for complement- or contact-system activation between attacks of angioedema. One year later, her course evolved into a more typical course for acquired CI INH dejciency consisting of continuously low functional CI INH levels with evidence of activation of the complement and contact systems. S. G. provides a unique insight into the development of acquired Cl INH dejciency. (J ALLERGY CLINIMMUNOL 1991;88:908-18.) Key words: Cl inhibitor, acquired, dejiciency, angioedema, kallikrein, kininogen, autoantibody
From the *Department of Molecular and Experimental Medicine, The Scripps Research Institute, La Jolla, Calif., and **Division of Allergy and Infectious Diseases, Department of Medicine, University of Washington, Seattle, Wash. Supported in part by National Institutes of Health Grants A110386, RRO0833, and DE0 822944. Received for publication March 11, 1991. Revised July 17, 1991. Accepted for publication July 17, 1991. Reprint requests: Bruce L. &raw, MD, Department of Molecular and Experimental Medicine, The Scripps Research Institute, 10666 N. Toney Pines Rd., La Jolla, CA 92037. Publication No. 5164BCR from The Scripps Research Institute. *Dr. Bruce L. &raw is a recipient of the Burroughs Wellcome Developing Investigator Award in Immunopharmacology. l/1/32608 908
anaphylaxis,
complement,
Abbreviations
used
Cl INH: EACA: HMWK: PBS: RT: SDS-PAGE:
Cl inhibitor Epsilonaminocaproic acid High-molecular-weight kininogen Phosphate-buffered saline Roomtemperature Sodium dodecyl sulfate-polyacrylamidegel electrophoresis
The syndrome of acquired Cl INH deficiency was first describedin a patient with lymphoma, recurrent angioedema,and Cl INH deficiency. ’ Subsequently,
VOLUME NUMBER
Acute
88
consumption
of Cl INH
909
6
this syndrome has been recognized to occur in association with a variety of other disorders, including lymphoproliferative diseases, carcinomas, autoimmune diseases, and paraproteinemias.** 3 Unlike hereditary angioedema that results from a genetically determined decrease in the synthesis of functional C 1 INH, the genesis of C 1 INH deficiency in acquired Cl INH deficiency is an increased rate of Cl INH catabolism. The mechanism(s) of this increased Cl INH catabolism is not well defined. There is evidence ,that auto-anti-idiotypic4 or auto-antiCl INH5-’ antibodies may be responsible for Cl INH consumption in patients with these antibodies, whereas tumor-derived factors or antigen-antibody complexes may be responsible in other patients.’ We now describe a patient with acquired Cl INH deficiency whose presentation and course suggest ep-
isodes of acute intermittent consumption of Cl INH. This patient was initially observed with recurrent episodes of iangioedema and hypotension with normal Cl INH a.nd complement levels. Extensive studies eventually revealed that her attacks were characterized by profound C 1 INH deficiency accompanied by mas-
sive activation of the complement and contact systems . Bet ween attacks, however, these parameters were normal. Subsequently, her course has evolved
into a pattern more typical of acquired Cl INH deficiency with continuous consumption of Cl INH. This patient thus provides a unique insight into the development of the acquired Cl INH deficiency syndrome . MATERIAL AND METHODS Case report S. G. is a 50-year-old woman who was well until May 1985 when she began experiencing recurrent bouts of hypotension accompanied by abdominal discomfort, facial angioedema, and flushing of the head, neck, and upper anterior chest. Attacks resolved during 3 to 5 days. Cl INH and C4 levels were normal between and on two occasions during attacks. Tests for immune complexes, antinuclear antibodies, rheumatoid factor, and serum protein electrophoresis were all unremarkable. Urinary histamine and 5-hydroxyindole acelic acid were normal on two separate occasions, and extensive skin testing failed to reveal any IgE-mediated sensitivity. Past history was remarkable for multiple leiomyoma of the skin and a total abdominal hysterectomy in 1971, with the patient taking estrogen supplementation until the onset of her angioedema. There was no family history of angioedema or unexplained sudden death. A presumptive diagnosis of idiopathic anaphylaxis was made. S G. had repeated normal complement profiles in plasma samples drawn during and after episodes of angioedema. During an episode of swelling precipitated by trauma, she was found., for the first time, to have low total hemolytic complement, C4, Clq, and Cl INH levels.9 After resolution of the attack, repeated measurements of Cl INH, C4, and
Clq levels were entirely normal. Because of the severity of her hypotension and the difficulty in establishing venous access for fluid administration, a Hickman catheter was placed in December 1985 after the eighth episode of hypotension. S. G. also started receiving danazol, 600 mglday, in December 1985. The patient then became attack free, and remained free while the danazol was tapered to 50 mg/day. In January 1987, the danazol was stopped, and 4 days later, she had her first attack of hypotension and angioedemain 14 months. Danazol was then restarted; however, S. G. had two additional attacks in August and September 1987. Blood obtained during the attack in September 1987, while she was receiving 400 mg/day of danazol, again revealed decreased Cl INH, C4, and Clq levels. S. G. suffered four more attacks of angioedema in 1988, requiring hospitalization. With each attack, complement and Cl INH abnormalities were demonstrated; however, these abnormalities returned to normal between attacks. Because of the inability of danazol at 600 mg/day to prevent lifethreatening angioedema, EACA, at 16 gm/day, was added to her therapeutic regimen, beginning Nov. 26, 1988. Subsequent to the initiation of EACA , S G has not experienced any additional massive attacks of angioedema with hypotension, although she continues to have frequent (often twice per week) episodes of moderate swelling. These attacks usually consist of abdominal or back pain that lasts 1 to 3 days. She also experiences frequent itching and flushing with these episodes. S. G. was evaluated by Dr. Michael Frank, Duke University Medical Center, Durham, N.C., in July 1990. AntiCl INH autoantibodies of the IgG and IgA isotype were found during her evaluation, and the patient was administered three courses of intravenous immunoglobulin in an attempt to reduce autoantibody production. Subsequently, S. G. was hospitalized at the University of Washington for an infected indwelling venous catheter. During the course of this hospitalization, she developed deep venous thrombosis and was treated with heparin and then warfarin sodium (Coumadin; DuPont Pharmaceuticals, Wilmington, Del.). Because of the deep venous thrombosis, the EACA was stopped, and a prophylactic permanent tracheostomy was performed. With danazol, 600 mg/day only, S. G. has experienced angioedema attacks of increased severity. Plasma
samples
Blood was collected in plastic tubes containing protease inhibitors (aprotinin, 333 IU/ml; benzamidine, 0.1 mg/ml; and phenanthroline, 0.01 mmol/L). Plasma was obtained by centrifugation, immediately aliquoted, and stored at -70” C until analysis. All processing steps were carried out in pIastic containers. Antibodies Biotinylated horse antimouse IgG was purchased from Vector Laboratories (Burlingame, Calif.). Affinity-purified goat antihuman plasma prekallikrein that recognized both prekallikrein and plasma kallikrein and affinity-purified goat anti-Hageman factor was a gift from Dr. Charles Cochrane,
910
Zuraw and Altman
The ScrippsResearch Institute.Polyclonalgoatandmurine monoclonalantihumanCl INH antibodieswere prepared aspreviouslydescribed.”Goatantiserumwasraisedagainst the purified light chain of humanHMWK and purifiedby affinity chromatography on an HMWK-Sepharose4B column. AntihumanCls antiserawaspreparedby immunizing goatswith Cls and purifiedby affinity chromatography on a Cls-Sepharose 4B column.Goat antihumanplasminogen antiserumwaspreparedby immunizinggoatswith purified plasminogen.Rabbit antibody to plasmin-a,-antiplasmin neoantigenwas purchasedfrom CalbiochemCorp. (La Jolla, Calif.). Peroxidase-conjugated antirabbit IgG was purchasedfrom Tago, Inc. (Burlingame,Calif.). Peroxidase-conjugated goat antimouseIgG waspurchasedfrom CaltagLaboratories(SanFrancisco,Calif.).
J. ALLERGY
CLIN. IMMUNOL. DECEMBER 1991
pension of goat antiprekallikrein IgG-conjugated Sepharose 4B beadsfor 1 hourat RT. Theadsorbed proteins wereseparated by 7.5%SDS-PAGEandelectrophoretically transferredto an Immobilon-Pmembrane.The membrane wasblockedwith Blotto for 1 hour at RT andprobedwith 2.5 pi/ml of biotinylatedaffinity-purified goat antiprekall&rein IgG for 3 hours at RT. After this procedure,the membrane waswashedandincubatedwith ‘Z51-labeled streptavidin asdescribedabove. ELlSA for protease-inhibitor
complexes
Protease-inhibitor complexesweremeasured by doubleantibody sandwichELISA. ‘l-l3 To measurekallikrein-Cl INH complexes,ImmulonII (DynatechIndustries,Alexandria,Va.), flat bottom96-well plateswere coatedwith lmmunoblotting affinity-purified goat antiprekallikreinantibodiesat 2.5 kg/ml in PBSfor 16hoursat 4” C. Residualreactivesites Immunoblottingfor Cl INH wasperformedexactly as wereoccupiedby a 16-hourincubationat 4” C with a sopreviously described.”Fifteen microlitersof plasmawas lution containing5 mg/ml of bovine gammaglobulin, 1 incubatedwith 40 p.1of a 1: 1 suspension of goat anti-Cl mg/ml of bovine serumalbumin,and 1 mg/ml of gelatin INH IgG-conjugatedSepharose 4B beadsfor 1 hourat RT. in PBScontaining0.05%Tween20. After a washwith PBS Theadsorbed proteinswereseparated by 7.5% SDS-PAGE containing0.05%Tween20 (PBS-Tween),200pl of sample undernonreducingconditionsand then electrophoretically diluted in serumbuffer (0.01 mol/L of phosphatebuffer transferred(Mighty SmallTransphorUnit, HoeferScientific containing1 mg/ml of gelatin,0.82%NaCl, 0.1 mg/ml of Instruments,San Francisco, Calif.) to an Immobilon-P thimerosal,1 mg/ml of bovineserumalbumin,and0.05% membrane (Millipore, Bedford,Mass.).Themembrane was Tween 20) was addedto eachwell and incubatedfor 16 blockedwith Blotto (5% wt/vol dry nonfat milk in 0.01 hoursat 4” C. Wellswerewashedandthenincubatedwith mol/L of PBSbuffer, pH 7.4, containing0.01% antifoam monoclonalanti-Cl INH antibody at 2 pg/ml diluted in A, 1 mg/L of thimerosal,1 mmol/L of amidinophenylanti-immunoglobulin buffer (serumbuffer with 5 mg/ml of methanesuflonyl fluoride, and 0.2 gm/L of sodiumazide) bovineserumglobulinaddedandlackinggelatin)for 3 hours for 1houratRT andthenprobedwith 5 pi/ml of monoclonal at RT with shaking.After three additionalwashes,goat anti-Cl INH for 3 hoursat RT. After a wash,themembrane antimouseIgG horseradish-peroxidase conjugate(1: 4000 wasoverlaid with biotinylatedhorseantimouseIgG for 1 in anti-immunoglobulin buffer) was addedand incubated hour at RT, washedagain, and finally incubatedwith for 2 hoursat RT with shaking.After threefinal washes ‘ZSI-labeled streptavidin(40,000cpm/ml)for 5 minutes.The with PBS-Tweenand two washeswith PBS, substrate membranewasthen washed,dried, and examinedby au- solution, 2,2’-azino-di-(3-ethylbenzthiazoline-6-sulfonic toradiography.Native, modified, and protease-complexed acid), 1 mg/ml in a buffer containing94 mmol/L of Cl INH bandswere quantifiedby densitometry.The Cl NazHPO,,53 mmol/L of citric acid, pH 4.6, and0.005% INH bandsmigrateat the following apparentmolecular H,O, wasaddedandincubatedat RT with shakingfor 20 weights:native Cl INH, 110kd; modifiedCl INH, 94 kd; to 60 minutes.Optical densitywasreadwith a spectrophoCl&Cl INH andkallikrein-Cl INH complexes,198kd; and tometer(MR600, DynatechLaboratories). Clr-Cl INH complexes,213 kd.” The percentages of 94 Activated Hagemanfactor-Cl INH complexesandClskd Cl INH reportedhereinwere derivedby dividing the Cl INH complexesweremeasured exactly asaboveexcept densityof the 94 kd bandby the sumof the densitiesof all that the Immulonwells were coatedwith affinity-purified bandsin the samelaneandmultiplying by 100. goatanti-Hageman factor or affinity-purifiedgoat anti-Cls Immunoblottingfor HMWK andprekallikrein/kallikrein antibodies,respectively. Plasmin-antiplasmin complexes wasperformedin a similarmanneras describedabove. were measured asdescribedabovewith the following difFor immunoblottingof HMWK, 3 pl of plasmain Laemmli ferences:the ImmulonII wells werecoatedwith 2 p.l/ml samplebuffer was separated on 7.5% SDS-PAGE under of goat antiplasminIgG, the secondantibody was rabbit nonreducing conditions and then transferred to an antibody to plasmin-cw,-antiplasmin neoantigen,and the Immobilon-Pmembrane.The Immobilon-Pmembrane was secondantibodywasdetectedwith peroxidase-conjugated blockedwith Blotto for 1 hour at RT andprobedwith bioantirabbitIgG. tinylatedgoatanti-HMWK in Blotto for 3 hoursat RT. The membranewas then washed,incubatedwith 12SI-labeled streptavidin(40,000cpm/ml) for 5 minutes,washedagain, Other assays Activation of the classicpathway of complementwas air dried, andexposedto x-ray film. Amountsof native and for C4d cleavedHMWK werequantitatedby densitometry.Immu- determinedby performingrocket electrophoresis and C4 as previouslydescribed.‘*Clq and C4 antigenic noblottingfor prekallikrein/kallikin wasperformedby immunoadsorbing 15 pl of plasmawith 40 pl of a 1: 1 sus- levels were measuredby singleradial immunodiffusion.
VOLUME NUMBER
TABLE
Acute consumption
88 6
I. Complement
levels between
attacks
of angioedema Patient
Test
Cl IhlH c4 Cl¶
of Cl INH 911
Mean (mgldl)
No. of determinants
13 17 7
19 56 1.5
values Range (mgldl)
16-22 37-73 13-17
Normal range* (mg/dl)
IO-19 1O-40 12-25
*Normal values for laboratory of University of Washington. Total IgG 1e;vels were measured with a radial immunodiffusion kit (The Binding Site, Inc., San Diego, Calif.). Cl INH functional activity was measured by Clr immunodiffusion assay.”
RESULTS Complement- and contact-system during the early course
studies
S. G. was hospitalized seven times during 1985 for attacks of angioedema and hypotension. Repeated laboratory evaluation during and between the first six attacks revealed no evidence of complement abnormalities. During the seventh attack, precipitated by trauma, low levels of total hemolytic complement, C4, Clq, and Cl INH were finally detected. A trial of danazol therapy was begun in December 1985. S. G. then remained free of angioedema for 14 months until the danazol was stopped. Within 4 days after danazol was stopped, S. G. had an attack of angioedema that required hospitalization. She restarted danazol therapy but had two additional attacks of angioedema requiring hospitalization in 1987. During the attack in September 1987, she was again found to have complement abnormalities. While she was hospitalized, she had low antigenic and functional Cl INH levels as well as markedly depressed C4 and Clq levels. Between September 1987 and March 1988, S. G. continued to take danazol and was asymptomatic. C 1 INH, C lq and C4 levels were checked on multiple occasions during this period and were normal or elevated in each instance (Table I). S. G. was hospitalized for research studies in November 1987. In addition to normal levels of C4, Clq, and Cl INH, she was found to have a normal C4d/C4 rati.0. Studies of other patients with acquired C 1 INH deficiency have revealed that plasma C 1 INH circulates primarily in cleaved form. ‘OPreviously, we have demonstrated that cleavage of plasma HMWK correlates with the generation of 94 kd Cl INH.16 In contrast, imrnunoblotting studies in S. G. revealed that her Cl INH was in its native 110 kd form (Fig. 1, lane 3) and that her HMWK was not cleaved (Fig.
2, lane 3). Thus, S. G. demonstrated no evidence of C 1 INH deficiency, complement-system activation, or contact-system activation at this time. Activation of complement and contact system during an acute attack of angioedema
S. G. was hospitalized on March 15, 1988, with facial angioedema and an unobtainable blood pressure despite taking danazol, 600 mg/day. She was treated with a massive volume of intravenous fluids, parenteral corticosteroids , and diphenhydramine . Six blood specimens were collected during the first 36 hours of hospitalization for research studies. The Cl INH and complement values immediately before, during, and after this attack are presented in Table II. Cl INH, C4, and Clq levels were grossly abnormal during the attack but were normal both before and after the attack. Cleavage of Cl INH in the attack samples was assessedby immunoblotting (Fig. 1). The normal human plasma that contains native 110 kd Cl INH is illustrated in lane 1, and in vitro incubation of normal human plasma with dextran sulfate, which activates the plasma contact system and generates 94 kd Cl INH, is illustrated in lane 2. lo The November 1987 preattack sample of S. G. is illustrated in lane 3, and like normal human plasma, it contains only 110 kd Cl INH. By comparison, the attack samples (lanes 4 to 9) demonstrated considerable cleavage of the Cl INH to the 94 kd form. The Cl INH in the first sample (lane 4), obtained at the time of admission, is present almost entirely in the modified 94 kd form. Twelve hours later (lane 6), 16% of the Cl INH was native, and the percentage of native Cl INH maintained the trend toward normalization in the subsequent three attack samples taken at the intervals indicated (he 7, 54%; lane 8, 39%; and lane 9, 32%). The same samples were immunoblotted for HMWK to search for evidence of in vivo HMWK cleavage (Fig. 2). HMWK can be cleaved by enzymes, termed kininogenase, to release the vasoactive peptide, bra-
912
Zuraw and Altman
J. ALLERGY
CLIN. IMMUNOL. DECEMBER 1991
3456789
FIG. ‘I. Cl INH immunoblot of plasma samples of S. G. obtained during an attack of angioedema. Plasma samples were immunoadsorbed with goat anti-Cl INH-conjugated Sepharose beads, separated by SDS-PAGE, transferred, and probed with monoclonal anti-Cl INH antibody; lane 1, normal plasma control; lane 2, dextran sulfate-activated plasma (positive control); lane 3, preattack plasma of S. G.; lanes 4 to 9, angioedema attack plasma samples of S. G. obtained at the times indicated.
TABLE II. Complement Date 11/10/87
2122188 2/29/88 3/15/M
3/16/88 3122188 4125188
levels during Status
Preattack Preattack Preattack Attack Attack Attack Recovery
an episode
of angioedema
Cl INH” (mgldl)
20 16 16 4 9 22 22
(m:dl)
177 43 47 6 21 37 60
@Id/c4 (ratio)
cm (mgldl)
Cl.0 NT
44 NT
NT
NT
8.7 3.9 NT NT
0.9 1.5 15 NT
NT, Not tested. *Normal values for Scripps Clinic: CI INH, 17 to 24 mgidl; C#, 44 to 74 mgidl; C4diC4, c1.2; Clq, 5 to 8 mgidi.
dykinin, which is believed to be one of the mediators of angi~e~ in Cl INH-deficient patients.17. ” Intact HMWK is demonstrated under nonreducing conditions at an apparent molecular weight of 150 kd. Kininogenases cleave HMWK into a 107 kd fragment, and further cleavage results in a 95 kd fragment. That normal plasma contains uncleaved 150 kd HMWK is illustrated in lane 1. Dextran-activated plasma, which contains the cleaved 107 and 95 kd fragments of HMWK, is illustrated in lane 2. The patient’s HMWK was totally cleaved into the 95 kd fragment in the first three attack samples (lanes 4 to 6). The percentage of uncleaved HMWK was 4% 14 hours after admission (lane 7), 36% at 30 hours ftane 8), and 11% 36 hours after admission (lane 9).
The major kininogenase in plasma is kallikrein of the contact system. K~li~ein normally circulates as an inactive zymogen, called prekallikrein. Since activation of the contact system can efficiently cleave Cl INH to its modified 94 kd form, we sought evidence of direct activation of plasma prek~li~ein to kallikrein during this patient’s angioedema attack. Plasma samples were immunoblotted under nonreducing conditions with an antibody that recognizes both prekallikrein and kallikrein (Fig. 3). Normal plasma demonstrates only 80 kd prekallikrein (lane 1). When the contact system was activated in vitro with dextran sulfate, a loss of the 80 kd band with appearance of two new bands of higher molecular weights were observed (lane 2). One band is a 198
VOLUME NUMBER
Acute
88 6
consumption
of Cl INH
913
FIG. 2. HMWK immunoblot of plasma samples of S. G. obtained during an attack of angioedema. Plasma samples were separated by SDS-PAGE, transferred, and probed with biotinylated goat anti-HMWK light-chain-specific antibody. Lanes are identified in Fig. 1.
FIG. 3. Prekallikrein/kallikrein immunoblot of plasma samples of S. G. obtained during an attack of angioedema. Plasma samples were immunoadsorbed with goat anti-prekallikrein-conjugated Sepharose beads, separated by SDS-PAGE under nonreducing conditions, transferred, and probed with biotinylated goat anti-prekallikrein antibody. Lanes are identified in Fig. 1.
kd band th.at is a kallikrein-Cl INH complex, ‘O and the other band is a 145 kd band that may be antithrombin III-kallikrein complex; however, this is not proven. Antithrombin III can inhibit kallikrein, and the resulting complex would have a predicted molecular weight of 145 kd. I9 The patient’s preattack plasma (lane 3) reveals a prekallikrein band at 80 kd, identical to the normal plasma control. In the first two attack plasmas (lanes 4 to 5), there is a prominent 145 kd
band with a faint 145 kd band also visible in the plasma obtained 6 hours after admission (lane 6). No 198 kd kallikrein-Cl INH complex was observed in these samples; however, there was essentially no functional 110 kd Cl INH in the early attack samples (Fig. 1). To determine whether the 80 kd band observed in the attack samples represented prekallikrein or kallikrein, we repeated the prekallikrein immunoblot under
914
Zuraw
and Altman
&ms
J. ALLERGY
after admission:
0 2
6
CLIN. IMMUNOL. DECEMBER 1991
14 30 36
FIG. 4. Prekallikrein immunoblot of plasma samples of S. G. obtained during an attack of angioedema. lmmunoblot was performed in a similar fashion as described in Fig. 3 except that the SDS-PAGE was performed under reducing conditions. Lanes are identified in Fig. 1.
reducing conditions (Fig. 4). Under these conditions, prekallikrein will still blot at 80 kd; however, kallikrein will be separated into free, heavy (43 kd) and light (37 kd) chains, which are not detected in this assay. The 80 kd prekallikrein band in normal plasma is illustrated in lane 1. Normal plasma incubated in vitro with dextran sulfate is illustrated in lane 2. The free, heavy and light chains are not observed; however, the decrease in intact prekallikrein is readily observed as a reduction in the intensity of the 80 kd band. The patient’s preattack sample (lane 3) reveals a strong prekallikrein band, 80% that of the normal control. In contrast, there was a marked reduction in the intensity of the 80 kd band in the first three attack plasmas (lanes 4 to 6), measuring 47%, 42%, and 38% of the normal control, respectively. This finding agrees with the results of the nonreduced prekallikrein/kallikrein immmioblot, and we interpret it to indicate that there was activation of prekallikrein to kallikrein. In the attack plasmas obtained at 14, 30, and 36 hours after admission (lanes 7 to 9), the amount of prekallikrein increased from 56% to 71% of the normal control. To exclude the possibility that this was due to dilution artifact from angioedema, we measured total IgG levels in the preattack and attack samples. There was an increased IgG in the admission sample (presumably because of hemoconcentration) but no other differences between the attack and preattack samples. S. G. suffered three more attacks requiring hospitalization during 1988, although all attacks were less severe than the attack of March 1988. Repeat complement- and contact-system studies were obtained during the last of these attacks in November 1988. The results of studies on plasma samples collected
during this attack are summarized in Table III. The first sample, dated Nov. 24, 1988, is from the time of admission, and samples were collected daily thereafter. The Cl INH level at the time of hospitalization was normal (217 mg/dl), but before admission, the level had been supranormal(>400 mg/ dl) because of high-dose danazol therapy. The C4d/C4 ratio was markedly elevated in the first attack sample but rapidly fell toward normal and demonstrated only modest activation of the classic complement pathway in the final attack sample. Immunoblotting for Cl INH and MHWK again demonstrated marked cleavage of both proteins in the first attack sample, with decreased cleavage in the later plasmas as the attack subsided. Complexes between Cl INH and its target proteases, ClS, kallikrein, and activated Hageman factor, were measured in these attack samples by proteaseinhibitor-complex ELISA’1-‘3 (Table III). There were increased levels of complexes on Nov. 24, 1988, and these levels rapidly decreased with recovery. Because plasmin can also cleave HMWK, and perhaps Cl INH, we also measured plasmin-anti-plasmin levels, and these levels did not significantly change during the attack. Therefore, as in the previous attack, there was evidence of acute Cl INH consumption and activation of the complement and contact systems during the attack. Continuous activation and contact systems receiving EACA
of the complement while patient was
Because of an inability to control the patient’s attacks despite long-term high-dose danazol therapy, treatment with EACA was added in addition to the danazol after the attack described above. After this
VOLUME NUMBER
Acute consumption
BE: 6
TABLE Ill. Complement system, attack in November 1988 --
contact
system,
and fibrinolytic
(ratio)
% 94 kd Cl INH
% Cleaved HMWK
Cl.0 7.44 4.35 3.22 1.97 NT
-6 70 50 44 43 NT
et al.
J. ALLERGY
liquid chromatography (HPLC) [Abstract]. J ALLERGY CLIN IMMUNOL
1990;85:150.
25. Smestad-Paulsen B, Flo L, Nesje G, Wold JK. Allergens in pollen from mugwort (Arfemisia vulgaris L.). I. Partial characterization of allergen preparations from mugwort pollen with emphasis on the carbohydrate moiety. Int Arch Allergy Appl Immunol 1985;78:206-12. 26. Jones OT, Earnest JP, McNamee MG. Solubilization and reconstitution of membrane proteins. In: Findlay JBC, Evans WH, ed. Biological membranes: a practical approach. Oxford: IRL Press, 1987:139-77.
CLIN. IMMUNOL. DECEMBER 1991
27. Jaggi KS, Gangal SV. Purification and characterization of allergens from Xunthium srrumarium pollen. Mol Cell Biochem 1987;78:177-90. 28. Bush RK, Yunginger JW. Standardization of fungal allergens. Clin Rev Allergy 1987;5:3-21. 29. Barnes C, Pacheco F, Portnoy J. Carbohydrate and protein contribution to Alrernariaallergen activity [Abstract]. J ALLERGY CLIN IMMLJNOL
Acute consumption of Cl inhibitor patient with acquired Cl-inhibitor deficiency syndrome Bruce L. Zuraw, MD,* and Leonard C. Altman,
1990;85: 169.
in a
MD**
La Jolla, Calif., and Seattle, Wash. Acquired Cl-inhibitor (Cl INH) deficiency is usually found in association with an underlying disease that is believed to be responsible for increased Cl INH catabolism, ultimately leading to the development of Cl INH dejiciency. We report a remarkable patient with acquired Cl INH deficiency in whom a unique progression of complement- and contact-system abnormalities has been observed. S. G. suffers from recurrent episodes of angioedema and hypotension. Results of repeated complement studies were initially normal, and the patient was diagnosed as having idiopathic anaphylaxis. Two years later, the patient was found to develop acute consumption of CI INH with activation of the complement and contact systems during episodes of angioedema. The patient continued to have normal Cl INH levels and to have no evidence for complement- or contact-system activation between attacks of angioedema. One year later, her course evolved into a more typical course for acquired CI INH dejciency consisting of continuously low functional CI INH levels with evidence of activation of the complement and contact systems. S. G. provides a unique insight into the development of acquired Cl INH dejciency. (J ALLERGY CLINIMMUNOL 1991;88:908-18.) Key words: Cl inhibitor, acquired, dejiciency, angioedema, kallikrein, kininogen, autoantibody
From the *Department of Molecular and Experimental Medicine, The Scripps Research Institute, La Jolla, Calif., and **Division of Allergy and Infectious Diseases, Department of Medicine, University of Washington, Seattle, Wash. Supported in part by National Institutes of Health Grants A110386, RRO0833, and DE0 822944. Received for publication March 11, 1991. Revised July 17, 1991. Accepted for publication July 17, 1991. Reprint requests: Bruce L. &raw, MD, Department of Molecular and Experimental Medicine, The Scripps Research Institute, 10666 N. Toney Pines Rd., La Jolla, CA 92037. Publication No. 5164BCR from The Scripps Research Institute. *Dr. Bruce L. &raw is a recipient of the Burroughs Wellcome Developing Investigator Award in Immunopharmacology. l/1/32608 908
anaphylaxis,
complement,
Abbreviations
used
Cl INH: EACA: HMWK: PBS: RT: SDS-PAGE:
Cl inhibitor Epsilonaminocaproic acid High-molecular-weight kininogen Phosphate-buffered saline Roomtemperature Sodium dodecyl sulfate-polyacrylamidegel electrophoresis
The syndrome of acquired Cl INH deficiency was first describedin a patient with lymphoma, recurrent angioedema,and Cl INH deficiency. ’ Subsequently,
VOLUME NUMBER
Acute
88
consumption
of Cl INH
909
6
this syndrome has been recognized to occur in association with a variety of other disorders, including lymphoproliferative diseases, carcinomas, autoimmune diseases, and paraproteinemias.** 3 Unlike hereditary angioedema that results from a genetically determined decrease in the synthesis of functional C 1 INH, the genesis of C 1 INH deficiency in acquired Cl INH deficiency is an increased rate of Cl INH catabolism. The mechanism(s) of this increased Cl INH catabolism is not well defined. There is evidence ,that auto-anti-idiotypic4 or auto-antiCl INH5-’ antibodies may be responsible for Cl INH consumption in patients with these antibodies, whereas tumor-derived factors or antigen-antibody complexes may be responsible in other patients.’ We now describe a patient with acquired Cl INH deficiency whose presentation and course suggest ep-
isodes of acute intermittent consumption of Cl INH. This patient was initially observed with recurrent episodes of iangioedema and hypotension with normal Cl INH a.nd complement levels. Extensive studies eventually revealed that her attacks were characterized by profound C 1 INH deficiency accompanied by mas-
sive activation of the complement and contact systems . Bet ween attacks, however, these parameters were normal. Subsequently, her course has evolved
into a pattern more typical of acquired Cl INH deficiency with continuous consumption of Cl INH. This patient thus provides a unique insight into the development of the acquired Cl INH deficiency syndrome . MATERIAL AND METHODS Case report S. G. is a 50-year-old woman who was well until May 1985 when she began experiencing recurrent bouts of hypotension accompanied by abdominal discomfort, facial angioedema, and flushing of the head, neck, and upper anterior chest. Attacks resolved during 3 to 5 days. Cl INH and C4 levels were normal between and on two occasions during attacks. Tests for immune complexes, antinuclear antibodies, rheumatoid factor, and serum protein electrophoresis were all unremarkable. Urinary histamine and 5-hydroxyindole acelic acid were normal on two separate occasions, and extensive skin testing failed to reveal any IgE-mediated sensitivity. Past history was remarkable for multiple leiomyoma of the skin and a total abdominal hysterectomy in 1971, with the patient taking estrogen supplementation until the onset of her angioedema. There was no family history of angioedema or unexplained sudden death. A presumptive diagnosis of idiopathic anaphylaxis was made. S G. had repeated normal complement profiles in plasma samples drawn during and after episodes of angioedema. During an episode of swelling precipitated by trauma, she was found., for the first time, to have low total hemolytic complement, C4, Clq, and Cl INH levels.9 After resolution of the attack, repeated measurements of Cl INH, C4, and
Clq levels were entirely normal. Because of the severity of her hypotension and the difficulty in establishing venous access for fluid administration, a Hickman catheter was placed in December 1985 after the eighth episode of hypotension. S. G. also started receiving danazol, 600 mglday, in December 1985. The patient then became attack free, and remained free while the danazol was tapered to 50 mg/day. In January 1987, the danazol was stopped, and 4 days later, she had her first attack of hypotension and angioedemain 14 months. Danazol was then restarted; however, S. G. had two additional attacks in August and September 1987. Blood obtained during the attack in September 1987, while she was receiving 400 mg/day of danazol, again revealed decreased Cl INH, C4, and Clq levels. S. G. suffered four more attacks of angioedema in 1988, requiring hospitalization. With each attack, complement and Cl INH abnormalities were demonstrated; however, these abnormalities returned to normal between attacks. Because of the inability of danazol at 600 mg/day to prevent lifethreatening angioedema, EACA, at 16 gm/day, was added to her therapeutic regimen, beginning Nov. 26, 1988. Subsequent to the initiation of EACA , S G has not experienced any additional massive attacks of angioedema with hypotension, although she continues to have frequent (often twice per week) episodes of moderate swelling. These attacks usually consist of abdominal or back pain that lasts 1 to 3 days. She also experiences frequent itching and flushing with these episodes. S. G. was evaluated by Dr. Michael Frank, Duke University Medical Center, Durham, N.C., in July 1990. AntiCl INH autoantibodies of the IgG and IgA isotype were found during her evaluation, and the patient was administered three courses of intravenous immunoglobulin in an attempt to reduce autoantibody production. Subsequently, S. G. was hospitalized at the University of Washington for an infected indwelling venous catheter. During the course of this hospitalization, she developed deep venous thrombosis and was treated with heparin and then warfarin sodium (Coumadin; DuPont Pharmaceuticals, Wilmington, Del.). Because of the deep venous thrombosis, the EACA was stopped, and a prophylactic permanent tracheostomy was performed. With danazol, 600 mg/day only, S. G. has experienced angioedema attacks of increased severity. Plasma
samples
Blood was collected in plastic tubes containing protease inhibitors (aprotinin, 333 IU/ml; benzamidine, 0.1 mg/ml; and phenanthroline, 0.01 mmol/L). Plasma was obtained by centrifugation, immediately aliquoted, and stored at -70” C until analysis. All processing steps were carried out in pIastic containers. Antibodies Biotinylated horse antimouse IgG was purchased from Vector Laboratories (Burlingame, Calif.). Affinity-purified goat antihuman plasma prekallikrein that recognized both prekallikrein and plasma kallikrein and affinity-purified goat anti-Hageman factor was a gift from Dr. Charles Cochrane,
910
Zuraw and Altman
The ScrippsResearch Institute.Polyclonalgoatandmurine monoclonalantihumanCl INH antibodieswere prepared aspreviouslydescribed.”Goatantiserumwasraisedagainst the purified light chain of humanHMWK and purifiedby affinity chromatography on an HMWK-Sepharose4B column. AntihumanCls antiserawaspreparedby immunizing goatswith Cls and purifiedby affinity chromatography on a Cls-Sepharose 4B column.Goat antihumanplasminogen antiserumwaspreparedby immunizinggoatswith purified plasminogen.Rabbit antibody to plasmin-a,-antiplasmin neoantigenwas purchasedfrom CalbiochemCorp. (La Jolla, Calif.). Peroxidase-conjugated antirabbit IgG was purchasedfrom Tago, Inc. (Burlingame,Calif.). Peroxidase-conjugated goat antimouseIgG waspurchasedfrom CaltagLaboratories(SanFrancisco,Calif.).
J. ALLERGY
CLIN. IMMUNOL. DECEMBER 1991
pension of goat antiprekallikrein IgG-conjugated Sepharose 4B beadsfor 1 hourat RT. Theadsorbed proteins wereseparated by 7.5%SDS-PAGEandelectrophoretically transferredto an Immobilon-Pmembrane.The membrane wasblockedwith Blotto for 1 hour at RT andprobedwith 2.5 pi/ml of biotinylatedaffinity-purified goat antiprekall&rein IgG for 3 hours at RT. After this procedure,the membrane waswashedandincubatedwith ‘Z51-labeled streptavidin asdescribedabove. ELlSA for protease-inhibitor
complexes
Protease-inhibitor complexesweremeasured by doubleantibody sandwichELISA. ‘l-l3 To measurekallikrein-Cl INH complexes,ImmulonII (DynatechIndustries,Alexandria,Va.), flat bottom96-well plateswere coatedwith lmmunoblotting affinity-purified goat antiprekallikreinantibodiesat 2.5 kg/ml in PBSfor 16hoursat 4” C. Residualreactivesites Immunoblottingfor Cl INH wasperformedexactly as wereoccupiedby a 16-hourincubationat 4” C with a sopreviously described.”Fifteen microlitersof plasmawas lution containing5 mg/ml of bovine gammaglobulin, 1 incubatedwith 40 p.1of a 1: 1 suspension of goat anti-Cl mg/ml of bovine serumalbumin,and 1 mg/ml of gelatin INH IgG-conjugatedSepharose 4B beadsfor 1 hourat RT. in PBScontaining0.05%Tween20. After a washwith PBS Theadsorbed proteinswereseparated by 7.5% SDS-PAGE containing0.05%Tween20 (PBS-Tween),200pl of sample undernonreducingconditionsand then electrophoretically diluted in serumbuffer (0.01 mol/L of phosphatebuffer transferred(Mighty SmallTransphorUnit, HoeferScientific containing1 mg/ml of gelatin,0.82%NaCl, 0.1 mg/ml of Instruments,San Francisco, Calif.) to an Immobilon-P thimerosal,1 mg/ml of bovineserumalbumin,and0.05% membrane (Millipore, Bedford,Mass.).Themembrane was Tween 20) was addedto eachwell and incubatedfor 16 blockedwith Blotto (5% wt/vol dry nonfat milk in 0.01 hoursat 4” C. Wellswerewashedandthenincubatedwith mol/L of PBSbuffer, pH 7.4, containing0.01% antifoam monoclonalanti-Cl INH antibody at 2 pg/ml diluted in A, 1 mg/L of thimerosal,1 mmol/L of amidinophenylanti-immunoglobulin buffer (serumbuffer with 5 mg/ml of methanesuflonyl fluoride, and 0.2 gm/L of sodiumazide) bovineserumglobulinaddedandlackinggelatin)for 3 hours for 1houratRT andthenprobedwith 5 pi/ml of monoclonal at RT with shaking.After three additionalwashes,goat anti-Cl INH for 3 hoursat RT. After a wash,themembrane antimouseIgG horseradish-peroxidase conjugate(1: 4000 wasoverlaid with biotinylatedhorseantimouseIgG for 1 in anti-immunoglobulin buffer) was addedand incubated hour at RT, washedagain, and finally incubatedwith for 2 hoursat RT with shaking.After threefinal washes ‘ZSI-labeled streptavidin(40,000cpm/ml)for 5 minutes.The with PBS-Tweenand two washeswith PBS, substrate membranewasthen washed,dried, and examinedby au- solution, 2,2’-azino-di-(3-ethylbenzthiazoline-6-sulfonic toradiography.Native, modified, and protease-complexed acid), 1 mg/ml in a buffer containing94 mmol/L of Cl INH bandswere quantifiedby densitometry.The Cl NazHPO,,53 mmol/L of citric acid, pH 4.6, and0.005% INH bandsmigrateat the following apparentmolecular H,O, wasaddedandincubatedat RT with shakingfor 20 weights:native Cl INH, 110kd; modifiedCl INH, 94 kd; to 60 minutes.Optical densitywasreadwith a spectrophoCl&Cl INH andkallikrein-Cl INH complexes,198kd; and tometer(MR600, DynatechLaboratories). Clr-Cl INH complexes,213 kd.” The percentages of 94 Activated Hagemanfactor-Cl INH complexesandClskd Cl INH reportedhereinwere derivedby dividing the Cl INH complexesweremeasured exactly asaboveexcept densityof the 94 kd bandby the sumof the densitiesof all that the Immulonwells were coatedwith affinity-purified bandsin the samelaneandmultiplying by 100. goatanti-Hageman factor or affinity-purifiedgoat anti-Cls Immunoblottingfor HMWK andprekallikrein/kallikrein antibodies,respectively. Plasmin-antiplasmin complexes wasperformedin a similarmanneras describedabove. were measured asdescribedabovewith the following difFor immunoblottingof HMWK, 3 pl of plasmain Laemmli ferences:the ImmulonII wells werecoatedwith 2 p.l/ml samplebuffer was separated on 7.5% SDS-PAGE under of goat antiplasminIgG, the secondantibody was rabbit nonreducing conditions and then transferred to an antibody to plasmin-cw,-antiplasmin neoantigen,and the Immobilon-Pmembrane.The Immobilon-Pmembrane was secondantibodywasdetectedwith peroxidase-conjugated blockedwith Blotto for 1 hour at RT andprobedwith bioantirabbitIgG. tinylatedgoatanti-HMWK in Blotto for 3 hoursat RT. The membranewas then washed,incubatedwith 12SI-labeled streptavidin(40,000cpm/ml) for 5 minutes,washedagain, Other assays Activation of the classicpathway of complementwas air dried, andexposedto x-ray film. Amountsof native and for C4d cleavedHMWK werequantitatedby densitometry.Immu- determinedby performingrocket electrophoresis and C4 as previouslydescribed.‘*Clq and C4 antigenic noblottingfor prekallikrein/kallikin wasperformedby immunoadsorbing 15 pl of plasmawith 40 pl of a 1: 1 sus- levels were measuredby singleradial immunodiffusion.
VOLUME NUMBER
TABLE
Acute consumption
88 6
I. Complement
levels between
attacks
of angioedema Patient
Test
Cl IhlH c4 Cl¶
of Cl INH 911
Mean (mgldl)
No. of determinants
13 17 7
19 56 1.5
values Range (mgldl)
16-22 37-73 13-17
Normal range* (mg/dl)
IO-19 1O-40 12-25
*Normal values for laboratory of University of Washington. Total IgG 1e;vels were measured with a radial immunodiffusion kit (The Binding Site, Inc., San Diego, Calif.). Cl INH functional activity was measured by Clr immunodiffusion assay.”
RESULTS Complement- and contact-system during the early course
studies
S. G. was hospitalized seven times during 1985 for attacks of angioedema and hypotension. Repeated laboratory evaluation during and between the first six attacks revealed no evidence of complement abnormalities. During the seventh attack, precipitated by trauma, low levels of total hemolytic complement, C4, Clq, and Cl INH were finally detected. A trial of danazol therapy was begun in December 1985. S. G. then remained free of angioedema for 14 months until the danazol was stopped. Within 4 days after danazol was stopped, S. G. had an attack of angioedema that required hospitalization. She restarted danazol therapy but had two additional attacks of angioedema requiring hospitalization in 1987. During the attack in September 1987, she was again found to have complement abnormalities. While she was hospitalized, she had low antigenic and functional Cl INH levels as well as markedly depressed C4 and Clq levels. Between September 1987 and March 1988, S. G. continued to take danazol and was asymptomatic. C 1 INH, C lq and C4 levels were checked on multiple occasions during this period and were normal or elevated in each instance (Table I). S. G. was hospitalized for research studies in November 1987. In addition to normal levels of C4, Clq, and Cl INH, she was found to have a normal C4d/C4 rati.0. Studies of other patients with acquired C 1 INH deficiency have revealed that plasma C 1 INH circulates primarily in cleaved form. ‘OPreviously, we have demonstrated that cleavage of plasma HMWK correlates with the generation of 94 kd Cl INH.16 In contrast, imrnunoblotting studies in S. G. revealed that her Cl INH was in its native 110 kd form (Fig. 1, lane 3) and that her HMWK was not cleaved (Fig.
2, lane 3). Thus, S. G. demonstrated no evidence of C 1 INH deficiency, complement-system activation, or contact-system activation at this time. Activation of complement and contact system during an acute attack of angioedema
S. G. was hospitalized on March 15, 1988, with facial angioedema and an unobtainable blood pressure despite taking danazol, 600 mg/day. She was treated with a massive volume of intravenous fluids, parenteral corticosteroids , and diphenhydramine . Six blood specimens were collected during the first 36 hours of hospitalization for research studies. The Cl INH and complement values immediately before, during, and after this attack are presented in Table II. Cl INH, C4, and Clq levels were grossly abnormal during the attack but were normal both before and after the attack. Cleavage of Cl INH in the attack samples was assessedby immunoblotting (Fig. 1). The normal human plasma that contains native 110 kd Cl INH is illustrated in lane 1, and in vitro incubation of normal human plasma with dextran sulfate, which activates the plasma contact system and generates 94 kd Cl INH, is illustrated in lane 2. lo The November 1987 preattack sample of S. G. is illustrated in lane 3, and like normal human plasma, it contains only 110 kd Cl INH. By comparison, the attack samples (lanes 4 to 9) demonstrated considerable cleavage of the Cl INH to the 94 kd form. The Cl INH in the first sample (lane 4), obtained at the time of admission, is present almost entirely in the modified 94 kd form. Twelve hours later (lane 6), 16% of the Cl INH was native, and the percentage of native Cl INH maintained the trend toward normalization in the subsequent three attack samples taken at the intervals indicated (he 7, 54%; lane 8, 39%; and lane 9, 32%). The same samples were immunoblotted for HMWK to search for evidence of in vivo HMWK cleavage (Fig. 2). HMWK can be cleaved by enzymes, termed kininogenase, to release the vasoactive peptide, bra-
912
Zuraw and Altman
J. ALLERGY
CLIN. IMMUNOL. DECEMBER 1991
3456789
FIG. ‘I. Cl INH immunoblot of plasma samples of S. G. obtained during an attack of angioedema. Plasma samples were immunoadsorbed with goat anti-Cl INH-conjugated Sepharose beads, separated by SDS-PAGE, transferred, and probed with monoclonal anti-Cl INH antibody; lane 1, normal plasma control; lane 2, dextran sulfate-activated plasma (positive control); lane 3, preattack plasma of S. G.; lanes 4 to 9, angioedema attack plasma samples of S. G. obtained at the times indicated.
TABLE II. Complement Date 11/10/87
2122188 2/29/88 3/15/M
3/16/88 3122188 4125188
levels during Status
Preattack Preattack Preattack Attack Attack Attack Recovery
an episode
of angioedema
Cl INH” (mgldl)
20 16 16 4 9 22 22
(m:dl)
177 43 47 6 21 37 60
@Id/c4 (ratio)
cm (mgldl)
Cl.0 NT
44 NT
NT
NT
8.7 3.9 NT NT
0.9 1.5 15 NT
NT, Not tested. *Normal values for Scripps Clinic: CI INH, 17 to 24 mgidl; C#, 44 to 74 mgidl; C4diC4, c1.2; Clq, 5 to 8 mgidi.
dykinin, which is believed to be one of the mediators of angi~e~ in Cl INH-deficient patients.17. ” Intact HMWK is demonstrated under nonreducing conditions at an apparent molecular weight of 150 kd. Kininogenases cleave HMWK into a 107 kd fragment, and further cleavage results in a 95 kd fragment. That normal plasma contains uncleaved 150 kd HMWK is illustrated in lane 1. Dextran-activated plasma, which contains the cleaved 107 and 95 kd fragments of HMWK, is illustrated in lane 2. The patient’s HMWK was totally cleaved into the 95 kd fragment in the first three attack samples (lanes 4 to 6). The percentage of uncleaved HMWK was 4% 14 hours after admission (lane 7), 36% at 30 hours ftane 8), and 11% 36 hours after admission (lane 9).
The major kininogenase in plasma is kallikrein of the contact system. K~li~ein normally circulates as an inactive zymogen, called prekallikrein. Since activation of the contact system can efficiently cleave Cl INH to its modified 94 kd form, we sought evidence of direct activation of plasma prek~li~ein to kallikrein during this patient’s angioedema attack. Plasma samples were immunoblotted under nonreducing conditions with an antibody that recognizes both prekallikrein and kallikrein (Fig. 3). Normal plasma demonstrates only 80 kd prekallikrein (lane 1). When the contact system was activated in vitro with dextran sulfate, a loss of the 80 kd band with appearance of two new bands of higher molecular weights were observed (lane 2). One band is a 198
VOLUME NUMBER
Acute
88 6
consumption
of Cl INH
913
FIG. 2. HMWK immunoblot of plasma samples of S. G. obtained during an attack of angioedema. Plasma samples were separated by SDS-PAGE, transferred, and probed with biotinylated goat anti-HMWK light-chain-specific antibody. Lanes are identified in Fig. 1.
FIG. 3. Prekallikrein/kallikrein immunoblot of plasma samples of S. G. obtained during an attack of angioedema. Plasma samples were immunoadsorbed with goat anti-prekallikrein-conjugated Sepharose beads, separated by SDS-PAGE under nonreducing conditions, transferred, and probed with biotinylated goat anti-prekallikrein antibody. Lanes are identified in Fig. 1.
kd band th.at is a kallikrein-Cl INH complex, ‘O and the other band is a 145 kd band that may be antithrombin III-kallikrein complex; however, this is not proven. Antithrombin III can inhibit kallikrein, and the resulting complex would have a predicted molecular weight of 145 kd. I9 The patient’s preattack plasma (lane 3) reveals a prekallikrein band at 80 kd, identical to the normal plasma control. In the first two attack plasmas (lanes 4 to 5), there is a prominent 145 kd
band with a faint 145 kd band also visible in the plasma obtained 6 hours after admission (lane 6). No 198 kd kallikrein-Cl INH complex was observed in these samples; however, there was essentially no functional 110 kd Cl INH in the early attack samples (Fig. 1). To determine whether the 80 kd band observed in the attack samples represented prekallikrein or kallikrein, we repeated the prekallikrein immunoblot under
914
Zuraw
and Altman
&ms
J. ALLERGY
after admission:
0 2
6
CLIN. IMMUNOL. DECEMBER 1991
14 30 36
FIG. 4. Prekallikrein immunoblot of plasma samples of S. G. obtained during an attack of angioedema. lmmunoblot was performed in a similar fashion as described in Fig. 3 except that the SDS-PAGE was performed under reducing conditions. Lanes are identified in Fig. 1.
reducing conditions (Fig. 4). Under these conditions, prekallikrein will still blot at 80 kd; however, kallikrein will be separated into free, heavy (43 kd) and light (37 kd) chains, which are not detected in this assay. The 80 kd prekallikrein band in normal plasma is illustrated in lane 1. Normal plasma incubated in vitro with dextran sulfate is illustrated in lane 2. The free, heavy and light chains are not observed; however, the decrease in intact prekallikrein is readily observed as a reduction in the intensity of the 80 kd band. The patient’s preattack sample (lane 3) reveals a strong prekallikrein band, 80% that of the normal control. In contrast, there was a marked reduction in the intensity of the 80 kd band in the first three attack plasmas (lanes 4 to 6), measuring 47%, 42%, and 38% of the normal control, respectively. This finding agrees with the results of the nonreduced prekallikrein/kallikrein immmioblot, and we interpret it to indicate that there was activation of prekallikrein to kallikrein. In the attack plasmas obtained at 14, 30, and 36 hours after admission (lanes 7 to 9), the amount of prekallikrein increased from 56% to 71% of the normal control. To exclude the possibility that this was due to dilution artifact from angioedema, we measured total IgG levels in the preattack and attack samples. There was an increased IgG in the admission sample (presumably because of hemoconcentration) but no other differences between the attack and preattack samples. S. G. suffered three more attacks requiring hospitalization during 1988, although all attacks were less severe than the attack of March 1988. Repeat complement- and contact-system studies were obtained during the last of these attacks in November 1988. The results of studies on plasma samples collected
during this attack are summarized in Table III. The first sample, dated Nov. 24, 1988, is from the time of admission, and samples were collected daily thereafter. The Cl INH level at the time of hospitalization was normal (217 mg/dl), but before admission, the level had been supranormal(>400 mg/ dl) because of high-dose danazol therapy. The C4d/C4 ratio was markedly elevated in the first attack sample but rapidly fell toward normal and demonstrated only modest activation of the classic complement pathway in the final attack sample. Immunoblotting for Cl INH and MHWK again demonstrated marked cleavage of both proteins in the first attack sample, with decreased cleavage in the later plasmas as the attack subsided. Complexes between Cl INH and its target proteases, ClS, kallikrein, and activated Hageman factor, were measured in these attack samples by proteaseinhibitor-complex ELISA’1-‘3 (Table III). There were increased levels of complexes on Nov. 24, 1988, and these levels rapidly decreased with recovery. Because plasmin can also cleave HMWK, and perhaps Cl INH, we also measured plasmin-anti-plasmin levels, and these levels did not significantly change during the attack. Therefore, as in the previous attack, there was evidence of acute Cl INH consumption and activation of the complement and contact systems during the attack. Continuous activation and contact systems receiving EACA
of the complement while patient was
Because of an inability to control the patient’s attacks despite long-term high-dose danazol therapy, treatment with EACA was added in addition to the danazol after the attack described above. After this
VOLUME NUMBER
Acute consumption
BE: 6
TABLE Ill. Complement system, attack in November 1988 --
contact
system,
and fibrinolytic
(ratio)
% 94 kd Cl INH
% Cleaved HMWK
Cl.0 7.44 4.35 3.22 1.97 NT
-6 70 50 44 43 NT
