A mechanism for hereditary angioedema caused by a methionine-379–to–lysine substitution in kininogens

Patients with the disorder hereditary angioedema (HAE) experience recurring bouts of soft tissue swelling primarily involving the oropharyngeal mucosa, subcutaneous tissues of the face and hands, genitals, and gastrointestinal tract.^1-4 In most cases, angioedema is due to excessive formation of the vasoactive peptide bradykinin because of increased activity of the plasma kallikrein-kinin system (KKS).^1,3,5 The KKS is composed of the zymogens prekallikrein (PK) and factor XII (FXII) and the cofactor/substrate high–molecular weight kininogen (HK).^6-8 PK and FXII reciprocally convert each other to the proteases plasma kallikrein (PKa) and FXIIa.^9-11 PKa then cleaves HK after Lys³⁸⁰ and Arg³⁸⁹ to release bradykinin.^6-9 The physiologic effects of bradykinin are mediated primarily through specific bradykinin receptors expressed in many tissues.^12,13 Basal bradykinin generation likely contributes to setting blood vessel permeability and tone,^14,15 whereas higher concentrations at injury sites promote edema and pain sensation.^6-8 

PKa and FXIIa activity is regulated by the serpin C1-inhibitor (C1-INH).^1-3,16 In most patients with HAE, KKS dysregulation is caused by reduced plasma C1-INH activity (5%-30% of normal activity).^1-3,17 However, at least 10% of patients with HAE have normal plasma C1-INH activity (referred to as HAEnC1).^3,18-21 Several mutations in genes unrelated to C1-INH have been identified in such individuals. A Thr³²⁸Lys/Arg substitution in FXII and a Lys³³⁰Glu replacement in plasminogen are the most common,^18-24 and mechanisms have been proposed for how they may contribute to disease.^10,25-28 In 2019, Bork et al described a mutation in the Kng1 gene in several members of a family.²⁹ The mutation segregated with signs and symptoms of HAE in a manner indicating a high probability that it is causative. Kng1 encodes messenger RNAs for HK and the related plasma protein low–molecular weight kininogen (LK). The mutation causes Met³⁷⁹ in both kininogens to be replaced with Lys. The substitution is adjacent to Lys³⁸⁰ at one of the sites cleaved by PKa to release bradykinin from HK, suggesting that it may alter bradykinin release from HK by PKa. Alternatively, the basic Lys³⁷⁹ residue may create a novel cleavage site for PKa or other trypsin-like proteases that would result in release of the decapeptide Lys-bradykinin (kallidin), which also interacts with bradykinin receptors.^13,30 We introduced Lys³⁷⁹ into human HK and LK, and studied the effects on kinin formation in purified protein and plasma systems.

Hemagglutinin (HA) epitope tag antibody agarose conjugate (Fisher Scientific, Atlanta, GA); human plasma–derived FIXa, FXa, FXIa, FXII, FXIIa, PK, and PKa (Enzyme Research Laboratories, South Bend, IN); α-thrombin and plasmin (Prolytix, Essex Junction, VT); S2302 (H-d-prolyl-l-phenylalanyl-l-arginine-p-nitroanilide dihydrochloride; DiaPharma, West Chester, OH); silica nanospheres (100 mm; nanoComposix, San Diego, CA); bradykinin enzyme-linked immunosorbent assay (ELISA) kit (Enzo Life Sciences, Farmingdale, NY); bradykinin standard (Eurofins DiscoveRx, Fremont, CA); Lys-bradykinin standard (Peptide Institute, Osaka, Japan). Captopril (Sigma-Aldrich, St Louis, MO); PKa inhibitor KV999272 (formerly VA999272) has been described previously³¹; tissue plasminogen activator (tPA; Alteplase, Genentech, South San Francisco, CA; normal human plasma (Precision BioLogic, Dartmouth, NS, Canada); human plasmas lacking HK and LK (George King Biomed); and PTT-A reagent (Diagnostica Stago, Asnieres-sur-Seine, France) were used in this study and sourced as indicated.

The amino acid numbering system used in this manuscript for human HK, LK, FXII, and plasminogen assigns the number 1 to the initiator Met on the signal peptide, according to current international standards. The table in supplemental Figure 1, available on the Blood website, lists corresponding amino acid numbering based on assigning the number 1 to the first amino acid of the mature protein, to facilitate comparing results to older literature.

Complementary DNAs (cDNA) encoding wild-type human HK and LK (HK-Met³⁷⁹ and LK-Met³⁷⁹, respectively) in the expression vector pJVCMV were modified by adding a HA tag (Tyr-Pro-Tyr-Asp-Val-Pro-Asp-Tyr-Ala) to the C-termini.²⁷ Site-directed mutagenesis was performed to alter the sequence encoding Met³⁷⁹ in wild-type HK and LK (HK-Met³⁷⁹ and LK-Met³⁷⁹) cDNAs to Lys to create the variants HK-Lys³⁷⁹ and LK-Lys³⁷⁹, respectively. All cDNAs underwent sequencing to verify the proper sequence. Kininogens were expressed in HEK293 cells and purified by chromatography on anti-HA agarose, as previously reported.²⁷ 

Human Glu-plasminogen was expressed in Expi293 cells (Fisher Scientific), concentrated from conditioned media on Lys-sepharose, and purified by ion exchange and size exclusion chromatography, as previously described.²⁷ 

Assays were performed in polyethylene glycol (PEG)-20000–coated microtiter plates. Assays for reciprocal activation of FXII (25 nM) and PK (60 nM) with or without HK (60 nM) were performed in 100 μL standard buffer (20 mM N-2-hydroxyethylpiperazine-N9-2-ethanesulfonic acid [pH 7.4], 100 mM NaCl, 0.1% PEG-8000, and 10 mM ZnCl₂) containing 200 mM S-2302. In some reactions, 25 μg/mL silica nanospheres were added. Changes in optical density at 405 nm at 37°C were continuously monitored.

For purified protein assays, HK-Met³⁷⁹, LK-Met³⁷⁹, HK-Lys³⁷⁹, and LK-Lys³⁷⁹ (200 nM) were incubated with PKa (2 nM for reactions containing HK; 50 nM for LK), plasmin (40 nM), thrombin (1 μM), FIXa (20 nM), FXa (20 nM), FXIa, (30 nM), or FXIIa (375 nM) in 150 μL reaction buffer (tris(hydroxymethyl)aminomethane [Tris]-buffered saline: 20 mM Tris-HCl, 136 mM NaCl; pH7.4) at 37°C. At various time points, 10 μL volumes were transferred into 90 μL ice-cold ethanol and placed at −80°C overnight. Samples underwent clarification by centrifugation at 10 000g for 1 hour. For all reactions, kinins were measured by ELISA (Enzo Life Sciences, Inc). For bradykinin release experiments, vehicle refers to 20 mM Tris-HCl and 136 mM NaCl, at pH7.4. Bradykinin and Lys-bradykinin were used to create standard curves.

For plasma assays, normal plasma was supplemented with captopril (500 μM), KV999272 (10 μM), and plasminogen (600 nM). After a 5-minute incubation, HK or LK (wild-type or variant), were added to 100 nM. Reactions (40 μL volumes) were started with tPA (125 nM) at 37°C. At various times, 3-μL reaction volumes were transferred to 12 μL ice-cold ethanol. Samples were clarified by centrifugation at 2000g for 10 minutes. Kinin release was determined as described earlier.

Purified kininogen samples were prepared in the same manner as for the kinin ELISA. For all analytes, ionization was achieved using matrix-assisted laser desorption ionization (MALDI). Analytes were deposited with matrix on MALDI targets. Matrix, 2,5-dihydroxybenzoic acid (Sigma), was purified by recrystallization from hot water. The matrix solution was prepared with recrystallized 2,5-dihydroxybenzoic acid (15 mg/dL) in a mixture of water, acetonitrile, and formic acid at a ratio of 1:1:0.1. Matrix solution (1 μL) was spotted onto the MALDI target with 1 μL analyte sample and allowed to dry at room temperature. Samples were analyzed in a Bruker AutoFlex (Billerica, MA) mass spectrometer. Mass spectrometry data were collected in positive-ion reflectron mode, with a mass-to-charge (m/z) range of 600 to 6000. Spectra are from an average of at least 100 laser shots per sample spot. Mass spectra were analyzed with Compass 1.4 Flex Series software (Bruker Daltonics).

To determine the nature of an apparent 16-dalton increase in molecular mass for bradykinin and Lys-bradykinin observed in some experiments, tandem mass spectrometry measurements were collected on a Bruker SolariX XR 12 T Fourier Transform Ion Cyclotron Resonance mass spectrometer equipped with a dual electron spray ionization (ESI)/MALDI source (Bremen, Germany). Mass spectra were collected in positive mode between 98 and 2000 m/z with 2 million data points and a 0.5592-second transient time. Time of flight was set at 0.8 milliseconds; 24 to 48 scans were averaged per spectrum. Tandem mass spectrometry was performed using collision-induced dissociation with collision energies ranging from 30 to 50 V.

Mass calibration was performed using sodium trifluoro acetate (1 mg/mL in 50:50::MeOH:H₂O) using ESI. Peptide spectra were acquired using MALDI and ESI. For the MALDI experiments, the laser power was set to a range from 50% to 60%, the laser frequency was 200 Hz, and between 5 and 350 laser shots were used, depending on the amount of signal. For the ESI experiments, samples were infused at a rate of 2 μL/min. The capillary inlet voltage was set to 4500 V with a −500 V end-plate offset. The nebulizer gas was set at 0.5 L/min, and the drying gas was set to 4 L/min at 180°C.

HK and LK are encoded by alternatively spliced messenger RNAs from the Kng1 gene.^32-35 Both proteins have D1, D2, D3, and D4 domains (Figure 1A). D4 contains the bradykinin (Figure 1B, top) and Lys-bradykinin (Figure 1B, middle) sequences. The HK D5 domain mediates surface binding,³⁶ and the HK D6 domain binds to PK or FXI,³⁷ allowing HK to function as a cofactor during surface-induced contact activation (Figure 1A). LK lacks D6, and its shorter D5 domain probably does not support surface binding. PKa cleaves wild-type HK after Lys³⁸⁰ and Arg³⁸⁹ (Figure 1B, top), releasing bradykinin. If a kininogen amino acid numbering system that assigns the number 1 to the first residue of the mature kininogens found in plasma is used, then these residues are designated Lys³⁶² and Arg³⁷¹, respectively (supplemental Figure 1). PKa cleaves wild-type LK at the same sites, although LK is a poor PKa substrate, probably because it lacks the PKa–binding site on the D6 domain. LK is thought to be primarily a substrate for tissue kallikreins, which cleave it after Met³⁷⁹ and Arg³⁸⁹ to release Lys-bradykinin.^38,39 The position of the Met³⁷⁹Lys substitution in variant HK and LK is shown in the last row of Figure 1B. The residue corresponding to amino acid 379 is Met in most mammalian kininogens and is a hydrophobic amino acid in nearly all (supplemental Figure 2).

Recombinant wild-type (HK-Met³⁷⁹ and LK-Met³⁷⁹) and variant (HK-Lys³⁷⁹ and LK-Lys³⁷⁹) kininogens were expressed in HEK293 cells and purified from conditioned media (Figure 1C). When PK and FXII are mixed in solution in the absence of a surface they reciprocally convert each other to the proteases PKa and FXIIa (Figure 2A).^9-11 In such reactions, HK-Met³⁷⁹ enhanced the rate of reciprocal activation modestly, whereas HK-Lys³⁷⁹ had little effect. This indicates that the Lys³⁷⁹ substitution is unlikely to cause HAE by increasing basal reciprocal activation of PK and FXII. When silicate nanospheres were added to reactions to induce contact activation, HK-Met³⁷⁹ and HK-Lys³⁷⁹ enhanced reciprocal activation comparably, indicating that the cofactor activity of the variant is intact. (Figure 2B).

PKa-mediated kinin release from HK was measured by ELISA (Figure 3A). The rate of release was comparable for HK-Met³⁷⁹ and HK-Lys³⁷⁹, indicating that the Lys³⁷⁹ substitution does not alter PKa-mediated HK cleavage. A higher PKa concentration is required to release detectable bradykinin from LK.²⁷ In such reactions, the rate of kinin release was slightly greater with wild-type LK-Met³⁷⁹ than with HK-Lys³⁷⁹ (Figure 3B). These findings suggest that PKa is not responsible for excess kinin generation in patients carrying the kininogen Lys³⁷⁹ substitution.

The Lys³⁷⁹ substitution could create a new cleavage site for trypsin-like serine proteases. Several plasma serine proteases involved in coagulation were tested for their capacities to release kinins from recombinant HKs (Figure 3C) and LKs (Figure 3D). In these reactions, the kininogen concentration was held constant, but protease concentrations varied. FXIa, a homolog of PKa, cleaves HK-Met³⁷⁹ to release bradykinin, although at a slower rate than does PKa.^27,39,40 Kinin release from HK-Lys³⁷⁹ by FXIa (30 nM) was slower than from HK-Met³⁷⁹ (Figure 3C). The KKS protease FXIIa (375 nM) did not release kinins from either HK-Met³⁷⁹ or HK-Lys³⁷⁹ (Figure 3C). The proteases thrombin (1000 nM), FXa (20 nM), and FIXa (20 nM) cleave HK-Lys³⁷⁹ but not HK-Met³⁷⁹, (supplemental Figure 3); however, this did not translate into kinin release (Figure 3C).

Several groups reported that the fibrinolytic serine protease plasmin cleaves HK.^27,41-43 Plasmin-catalyzed kinin generation was greater from both HK-Lys³⁷⁹ and LK-Lys³⁷⁹ than from their wild-type counterparts (Figures 3C-D). The initial rate of kinin release from HK-Lys³⁷⁹ was 3.3-fold greater than from HK-Met³⁷⁹, but this does not tell the whole story. Previously, we observed that plasmin-catalyzed bradykinin release from HK appears to be stoichiometric, with the maximum release mirroring the plasmin concentration.²⁷ This may reflect the formation of a tight interaction between HK and the plasmin kringle 5 domain.⁴⁴ Results for HK-Met³⁷⁹ in Figure 3E are consistent with earlier results, but HK-Lys³⁷⁹ cleavage by plasmin does not follow this pattern, with kinin generation exceeding plasmin concentration (Figure 3E). The initial rate of kinin release from LK-Lys³⁷⁹ catalyzed by plasmin was ∼13-fold faster than with LK-Met³⁷⁹ (Figure 3F). These results suggest that the Lys³⁷⁹ substitution either alters cleavage of kininogens by plasmin after residue Lys³⁸⁰ (increasing bradykinin production) or creates a novel cleavage site for plasmin that would result in Lys-bradykinin release.

The ELISA we used to measure kinin release does not discriminate between bradykinin and Lys-bradykinin and is equally sensitive to both (supplemental Figure 4). It may recognize other kinins as well. We used mass spectrometry to characterize the kinins released by PKa and plasmin from recombinant kininogens. Previously, we showed that incubating plasma-derived HK with PKa or plasmin generated a peak with the same m/z ratio as a bradykinin standard (m/z = 1060-1061).²⁷ With recombinant HK-Met³⁷⁹ and HK-Lys³⁷⁹, peaks obtained (m/z = 1076.4 and m/z = 1077.2, respectively; Figure 4) were ∼16 m/z units larger than with the bradykinin standard (m/z = 1060.5). Using tandem mass spectrometry (supplemental Figure 4), we determined that the peak represented bradykinin with replacement of a hydrogen atom with a hydroxyl group on the third amino acid in the bradykinin sequence (Pro³⁸³; Figure 1B). Modifications of prolines and other amino acids in this manner are relatively common and, in the case of the recombinant kininogens, likely occurred in cell culture or during protein purification.

Consistent with our previous results,²⁷ plasmin released bradykinin from wild-type HK-Met³⁷⁹ and LK-Met³⁷⁹. Again, the major peaks (m/z = 1077.2 and m/z = 1076.9, respectively; Figure 4) are consistent with the addition of a hydroxyl group on Pro³⁸³. For LK-Met³⁷⁹, there is also a smaller peak at m/z = 1060.9 matching the bradykinin standard, consistent with partial oxidation at Pro³⁸³. In contrast, plasmin cleavage of HK-Lys³⁷⁹ and LK-Lys³⁷⁹ generated major peaks with m/z ratios of 1205.5 and 1204.8, respectively, representing Lys-bradykinin (m/z = 1189) with a hydroxyl group on Pro³⁸³. For LK-Lys³⁷⁹ incubated with plasmin there is a second peak (m/z = 1189.8) matching the Lys-bradykinin standard, again consistent with partial oxidation at Pro³⁸³. These findings indicate that the Lys³⁷⁹ substitution creates a novel cleavage site for plasmin that is preferred over cleavage after Lys³⁸⁰ in HK and LK, and that the kinin produced as a result is Lys-bradykinin.

In the original report by Bork et al,²⁹ individuals who were symptomatic were heterozygotes for the Kng1 mutation causing the Lys³⁷⁹ substitutions in kininogens. Their plasmas, therefore, likely contain mixtures of wild-type and variant kininogens. We supplemented normal human plasma with wild-type or variant HK or LK to final concentrations of 100 nM, leaving endogenous plasma HK and LK in excess of the recombinant proteins. The amount of recombinant kininogen was chosen to avoid abnormally high total plasma kininogen concentrations. The plasmas were also supplemented with recombinant Glu-plasminogen.²⁷ Compared with the endogenous Glu-plasminogen in plasma, the recombinant protein is in an open conformation that is activated rapidly by tPA,^45,46 facilitating the generation of robust kinin signals in the plasma system. As shown in Figure 5, kinin generation is greater in plasma supplemented with HK-Lys³⁷⁹ (left panel) or LK-Lys³⁷⁹ (right panel) than if supplemented with HK-Met³⁷⁹ or LK-Met³⁷⁹. Signals for HK-Met³⁷⁹ or LK-Met³⁷⁹ were roughly similar. However, given the differences in normal plasma HK and LK concentrations, (640 nM and 2.4 μM, respectively) the results suggest that LK could be the major source of kinins during attacks of HAE.

In this study, replacing Met³⁷⁹ in HK and LK with Lys renders both proteins more susceptible to plasmin-mediated kinin release by creating a novel cleavage site at the N-terminus of the kinin sequence. Plasmin preferentially cleaves the variant kininogens after Lys³⁷⁹ and Arg³⁸⁹ rather than at the sites normally used by PKa for bradykinin release (after Lys³⁸⁰ and Arg³⁸⁹). Consequently, plasmin cleavage of HK-Lys³⁷⁹ and LK-Lys³⁷⁹ generates Lys-bradykinin (kallidin) rather than bradykinin. In plasma in which fibrinolysis is induced by addition of tPA, kinin release was substantially greater in plasma containing HK-Lys³⁷⁹ or LK-Lys³⁷⁹ than in plasma containing wild-type HK-Met³⁷⁹ or LK-Met³⁷⁹, consistent with a role for the Lys³⁷⁹ variants in angioedema. Lys-bradykinin and bradykinin both bind to the B1 and B2 bradykinin receptors and appear to have equipotent effects on cell Ca²⁺ signaling.^13,30,47 Because Lys-bradykinin is readily converted to bradykinin by aminopeptidase B, it is also possible that in carriers of the Lys³⁷⁹ substitution, angioedema may be partly, or largely, driven by bradykinin derived from Lys-bradykinin. To our knowledge, our study is the first to suggest that a form of HAE is due to generation of a kinin other than bradykinin.

Soft tissue swelling in HAE is typically caused by episodic increases in bradykinin generation, distinguishing it from more common histamine-induced swelling during allergic reactions.^1,3-5 A defect in regulating PK and FXII activation and PKa and FXIIa activity caused by reduced plasma C1-INH activity accounts for most cases of HAE.^1-3,16,17 However, at least 10% of patients with clinical presentations consistent with HAE have normal C1-INH levels (HAEnC1).^18-21 Single amino acid substitutions in 6 proteins have been linked to HAEnC1. The first discovered, and most common, are Lys or Arg substitutions for Thr³²⁸ in FXII.^10,19-22 The introduction of a basic amino acid creates a novel cleavage site for proteases such as plasmin and thrombin, allowing for the removal of the FXII heavy chain region.^10,25,26 The resulting truncated FXII is activated by PKa more rapidly than full-length FXII, accelerating KKS activation and overwhelming the capacity of C1-INH to control kinin generation. A Thr¹⁴⁴Ser substitution in heparan sulfate 3-O-sulfotransferase 6 may alter glycosylation of proteoglycans on vascular endothelial cell surfaces.⁴⁸ Bork et al postulated that the mutation causes loss of enzyme function, resulting in changes to key HK binding partners that render the protein more susceptible to cleavage by PKa.⁴⁸ As with C1-INH deficiency, the proposed mechanisms for HAE in patients with FXII Lys/Arg³²⁸ and 3-O-sulfotransferase 6 Thr¹⁴⁴Ser involve increased bradykinin production. Two other mutations identified in patients with HAEnC1 may alter vascular endothelium responses to factors that influence vessel permeability. Ala¹¹⁹Ser in the soluble protein angiopoietin-1 causes loss of function that may confer increased endothelial cell sensitivity to permeability-enhancing factors, including bradykinin and vascular endothelial growth factor.^49,50 Similarly, an Arg²¹⁷Ser replacement in the membrane protein myoferlin may lead to increased cell permeability by redistributing vascular endothelial growth factor receptor 2 from the cytoplasm to the cell membrane.⁵¹ 

Interactions between the fibrinolytic system and the KKS were postulated more than 50 years ago. Plasmin converts FXII to FXIIa by proteolysis of the Arg³⁷²-Val³⁷³ peptide bond.^25,27,52,53 Furthermore, FXII truncation by plasmin may contribute to inflammatory disorders including angioedema.^11,53,54 There are several reports describing HK cleavage by plasmin, although it is not clear that this results in kinin production in all situations.^41-43 Recently, we conducted a head-to-head comparison of the capacities of PKa and plasmin to release bradykinin from HK and LK.²⁷ PKa released bradykinin 50- to 100-fold faster from HK than from LK, whereas plasmin released bradykinin threefold to fourfold faster from LK than from HK. The rate of bradykinin release from HK was at least 50-fold greater for PKa than for plasmin, but the enzymes had similar capacities to release bradykinin from LK.

Several lines of evidence implicate plasmin-catalyzed reactions in angioedema. Observations of patients with congenital bleeding disorders strongly indicate that fibrinolytic activity is particularly robust in the mouth and nasopharynx.⁵⁵ Plasma bradykinin levels rise after infusions of the fibrinolytic activator tPA during treatment for thrombotic events,⁵⁶ and in some of these individuals, angioedema develops that is typically confined to the mouth (oral-lingual edema).^57,58 Similarly, angiotensin-converting enzyme (ACE) inhibitors can induce oral-lingual angioedema by interfering with kinin degradation.^59,60 The predilection for oral-lingual edema with tPA or ACE inhibitor treatment, and in some patients with HAE, may reflect intrinsically higher fibrinolytic activity in the oropharynx. Fibrinolytic inhibitors such as tranexamic acid have been used successfully to treat ACE inhibitor–induced angioedema.^61-64 They are also used to treat patients with HAEnC1 and some with HAE due to low C1-INH.^65,66 It is not clear why antifibrinolytics seem to be effective in patients with PKa-mediated HAE.

In 2018 Bork et al and Dewald et al described a mutation in the PLG gene encoding plasminogen in some patients with HAEnC1 that causes Lys³³⁰ to be replaced with glutamic acid.^23,24 The substitution creates a consensus binding site for side chains of basic amino acids such as Lys and Arg. It is notable that 80% of patients with plasminogen-Glu³³⁰ experience tongue swelling, and in many, this is the only manifestation of HAE.^16-19,23,67 We showed that replacing Lys³³⁰ with glutamic acid results in a striking increase in the rate of plasmin-catalyzed bradykinin release from both HK and LK, probably because the plasmin variant binds basic amino acids on kininogens differently than does wild-type plasmin.²⁷ Plasmin-Glu³³⁰ catalyzes bradykinin release comparably from HK and LK. The plasma concentration of LK is twofold to fourfold higher than that of HK, and plasma-based experiments support the conclusion that LK may be the major kinin source in carriers of plasminogen-Glu³³⁰.²⁷ 

As with plasminogen-Glu³³⁰, the results presented here for kininogens with the Lys³⁷⁹ substitution suggest that angioedema is the result of a mechanism that involves plasmin-mediated cleavage of both HK and LK. However, our work does not rule out the possibility that other proteases cleave the variant kininogens to release kinins. Because trypsin-like enzymes primarily cleave proteins after basic amino acids, introducing Lys³⁷⁹ might have created a cleavage site for proteases that were not identified in our limited survey. Furthermore, LK is primarily a substrate for the tissue kallikrein hK1 (encoded by the KLK1 gene) rather than for PKa (encoded by the KLKB1 gene), and the effects of the Lys³⁷⁹ replacement on LK cleavage by hK1 and other tissue kallikreins are unknown. The data shown here, and prior work with plasminogen-Glu³³⁰, support the conclusion that some forms of HAE do not require, and may not involve, PKa and FXIIa. This has implications for treating patients with HAE. Drugs specifically targeting PKa or FXIIa may be less effective for carriers of kininogen Lys³⁷⁹ or plasminogen Glu³³⁰ than for patients with other forms of HAE. Agents that block the B2 bradykinin receptor or inhibit fibrinolysis may be more appropriate for patients with angioedema that is primarily plasmin driven. Controlled clinical trials are needed to determine how genetic information in patients with HAEnC1 can be best used to guide selection of treatments.

This work was supported by awards HL140025 from the National Institutes of Health, National Heart, Lung, and Blood Institute (D.G.); APP1129592 from the National Health and Medical Research Council Australia (R.H.P.L.); and FL18010019 from the Australian Research Council (R.H.P.L.). Protein and peptide tandem mass spectrometry measurements were supported by the National Institutes of Health, Office of the Director under grant S10 OD025118 (principal investigator: Jon Amster, University of Georgia).

Contribution: S.K.D. designed and performed assays for kininogen cleavage and bradykinin generation, performed expression and purification of proteins, and contributed to writing the manuscript; S.K., D.R.P., and E.T.R. designed and performed mass spectrometry analyses; M.-f.S. developed the system for expressing recombinant HK; M.L. studied surface-dependent KKS activation; T.Z.H. performed assays for protease activation and bradykinin generation; R.H.P.L. developed the system for expressing recombinant Glu-plasminogen; E.P.F. helped design experiments involving the kallikrein inhibitor KV999272 and contributed to writing the manuscript; and D.G. oversaw the project and writing of the manuscript.

Conflict-of-interest disclosure: D.G. receives consultation fees from pharmaceutical companies with an interest in inhibition of contact activation and the KKS for therapeutic purposes. E.P.F. is an employee of Kalvista Pharmaceuticals, Inc. The remaining authors declare no competing financial interests.

Correspondence: David Gailani, Pathology, Microbiology, and Immunology, Vanderbilt University Medical Center, Room 4918, The Vanderbilt Clinic, 1301 Medical Center Dr, Nashville, TN 37232; email: dave.gailani@vanderbilt.edu.