Functional and structural properties of C1 inhibitor

C1 inhibitor (C1-INH) is a serine protease inhibitor (serpins) that inactivates several different proteases: C1r, C1s, and MASPs in the complement system, Factor XII and kallikrein in the contact system, Factor XI and thrombin in the coagulation system, tPA and plasmin in the fibrinolytic system [1–14]. C1-INH binds and blocks the activity of its target proteases by the suicide mechanism, which is typical of the serpins [15]. Proteins of this family share an amino acid sequence homology, which confers a similar three-dimensional structure to the C-terminal part (serpin domain) characterized by three β-sheets and an exposed mobile reactive loop. The P1–P1′ amino acid bond in this loop behaves as a pseudo-substrate for the proteases that cleave the peptide bond and covalently bind the amino acid residue. After binding, C1-INH acts as a mousetrap swinging the inactivated protease from the upper to the lower pole with insertion of the reactive loop as a fourth strand in the β-sheet [16]. Mutations causing substitutions of an amino acid involved in this conformational change almost invariably result in loss of inhibitory activity.

The N-terminal end of C1-INH (non-serpin domain) has no homologies with other serpins and is not essential for protease-inhibitor complex formation [17]. It is heavily glycosylated with 3 N-linked and 7 O-linked carbohydrates and contains two disulfide bridges connecting the N-terminal domain to the serpin domain [18]. Mutations abolishing the disulfide bridges lead to a conformational change that eliminates the metastable properties of the serpin domain and allows multimerization [19]. In addition, recent studies brought important evidence that C1-INH has a novel biological role as an anti-inflammatory protein. This role does not depend on the presence of an intact reactive site, but on the carbohydrates linked to the amino acid residues of the non-serpin domain in the amino-terminal end of the molecule. Owing to this moiety, C1-INH is able to bind lipopolysaccharides (LPS) and E-selectin. Hence, via direct binding to LPS, C1-INH protects mice from lethal Gram-negative endotoxemia and inhibits LPS-triggered macrophage expression of TNF-alpha mRNA [20]. With this binding, C1-INH prevents the endothelial cell response to LPS, which causes altered cell morphology, intercellular gap formation, increased transendothelial permeability and eventually leads to the capillary leak syndrome that complicates Gram-negative sepsis and septic shock [21]. The capacity to bind selectins is mediated by the Sialyl Lewis epitope, a fucose-containing tetrasaccharide known to bind all three selectins [22], which is present in C1-INH [23]. By binding E-selectin present in endothelial cells, C1-INH can concentrate at sites of inflammation and can also regulate leukocyte adhesion and transmigration across the endothelial surface [24]. By combining protease-inhibitory functions and these newly described anti-inflammatory properties, C1-INH could play a major role in several pathological conditions: future studies will certainly delineate the actual in vivo relevance for human diseases.

In order to have a complete scenario of the potential consequences of changes in C1-INH structure and/or function, we must introduce the concept of conformational diseases and serpinopathies [25, 26]. These terms identify a growing number of diseases arising from the same general mechanism of abnormal unfolding and then aggregation of an underlying protein: the new conformation of the mutant protein, rather than the reduction of its specific function, is a fundamental pathological event. Alpha1 antitrypsin, the archetypal serpin, provided the first example of this disease mechanism. Its Z mutant undergoes polymerization and aggregation, and, as a consequence, it is not secreted and accumulated in the endoplasmic reticulum of the hepatocyte. Much of it is degraded, but the remainder aggregates to form insoluble intracellular inclusions. These inclusions are associated with hepatocellular damage, and 10% of newborn Z homozygotes develop liver disease, which often leads to fatal childhood cirrhosis [27, 28]. Another single monoacid substitution of alpha1 antitrypsin (alpha 1-antitrypsin Pittsburgh) modifies its protease specificity from elastase to thrombin and causes bleeding disorders [29, 30]. Conformational changes in antithrombin leading to polymerization have been shown to cause thrombosis [31]. Likewise, inappropriate folding of serpin is now investigated as the primary defect leading to degenerative encephalopathies and dementia [32]. The list of diseases in this group and the underlying responsible serpins are still undefined and will probably be growing in the near future.

Extrapolating from the above, one could predict several pathological consequences deriving from C1-INH defects and, on the other side, several potential applications of C1-INH as a therapeutic agent.

C1 inhibitor deficiency and angioedema

In 1963, Donaldson and Evans discovered that subjects with hereditary angioedema (HAE) genetically lacked C1-INH [33]. Angioedema is a self-limiting, transient, localized edema due to a reversible increase in vascular permeability. HAE patients present with functional plasma levels of C1-INH ranging from less than 10% up to 35% compared to normal subjects. This defect is associated to mutations in one of the two alleles of C1-INH gene which is located in chromosome 11q11.2–q13 [18, 34, 35]. More than 150 mutations in this gene, associated with HAE, have been described so far, and a register listing all published mutations is currently on the web (http://hae.biomembrane.hu/) [36]. The clinical phenotype of HAE, i.e. angioedema recurrence, is transmitted as an autosomal dominant trait [37]. The mutated allele and C1-INH deficiency segregate into families according to Mendelian law [38, 39]. From the above, one could conclude that HAE is a monogenic disease, but attempts to correlate specific mutations with phenotypic aspects such as age at onset, frequency or severity of symptoms were unsuccessful. In front of consistently similar C1-INH functional levels within the same patients throughout life and also among different patients, the clinical phenotype may be extremely varied. Frequency and severity of angioedema symptoms vary from patient to patient as well as within the same patient from time to time. This variability clearly does not segregate with the transmission of C1-INH deficiency. Therefore, in HAE, as in the majority of genetic diseases, mutations cannot predict the phenotype of a patient, indicating that genetic factors and the environment might be important modulating factors [40].

This conclusion does not lessen the importance of mutation screening in HAE families. Besides the obvious value of the genetic diagnosis, these experiments could provide new insight into the understanding of conformational diseases and serpinopathies. Moreover, identifying mutations in C1-INH gene may still shed some light on the clinical variability of disease expression. Clinical phenotypes primarily dependent on protein conformation more than on protein function have low penetrance and probably need a cofactor to become expressed. As we can learn from the molecular studies on the Z mutant of alpha1 antitrypsin, temperature and rate of protein synthesis influence its polymerization, and the pro-inflammatory effect of the polymers is exacerbated by the presence of endogenous mediators, such as inflammatory cytokines or external factors such as cigarette smoke [16, 27]. Therefore, we can expect that a clinical phenotype arising from the conformational change induced by specific mutations will present a wide range of variation depending on the individual history of exposure to infections, environmental factors, genetic background, etc. Hence, a better understanding of the size and impact of conformational diseases will come from combining careful analysis of specific genotype/phenotype correlates with the understanding of the structure/function correlates of specific mutations. C1-INH offers a valuable opportunity to reach this goal: there is the register of mutations already mentioned, and there is a register of HAE patients which already contains more than 1,000 entries with detailed clinical information (http://www.haeregister.org). Physicians, scientists, and also patients who are working on these registers in Europe are associated in an active and expanding network [41]. These resources represent the appropriate background to exploit molecular studies on C1-INH mutants to understand clinical variability.

Identification of candidate factors governing the variability of angioedema recurrence in HAE patients is a main target for development of new therapeutic approaches. The episodic appearance of angioedema in C1-INH-deficient patients has been associated to further derangement of the defective control of activation of the contact/complement system, a direct result of the genetic defect [42–47]. The relative importance of these two systems in the pathogenesis of angioedema has been largely debated as detailed in recent reviews on this topic [48–50]. The controversial point has been the final mediator of the increased vascular permeability: a C2-derived peptide, in the “complementcentric theory” and bradykinin, in the “contactcentric theory”. Nowadays the contribution of the contact system and bradykinin is supported by a large mass of data coming from in vitro studies, analysis of patients, and analysis of C1-INH-deficient mice. We know now that contact system is activated during attacks [45, 46]. When depleted of enzymes inducing contact activation, C1-INH-deficient plasma loses its capacity to generate vascular permeability-enhancing activity [51]. During angioedema attacks, C1-INH-deficient patients have increased plasma levels of bradykinin, which is generated in the affected site [52, 53]. A C1-INH mutation, which selectively blocks the complement inhibitory activity but not the activity of C1-INH on the contact system, causes depletion of classical pathway complement components but does not result in angioedema [54]. C1-INH knockout mice, both homozygous and heterozygous, do not have obvious angioedema episodes but present an ongoing increased vascular permeability, which depends on C1-INH deficiency being corrected by intravenous C1-INH [55, 56]. Such an increased vascular permeability depends on contact system activation and is mediated via bradykinin receptor 2 (BK2R). These data lead to the conclusion that in C1-INH deficiency, angioedema takes place in sites where BK2R is stimulated by bradykinin released upon kallikrein generation. Highly specific and potent inhibitors of kallikrein and BK2R are now under investigation for their ability to revert angioedema in HAE patients [57–59]. Data from these studies should rapidly confirm, in a clinical setting, the appropriateness of the pathogenic view, which has been extrapolated from the laboratory.

As mentioned, the clinical picture of C1-INH deficiency is characterized by angioedema, i.e. bouts of local increase in vascular permeability. Depending on the affected site, patients suffer from disfiguring subcutaneous edema, abdominal pain, vomiting and/or diarrhoea for edema of the gastrointestinal mucosa, dysphagia, and dysphonia up to asphyxia for edema of the pharynx and larynx [60]. The episodes last from 2 to 5 days and usually render patients incapable of pursuing their normal activities. Half the patients become symptomatic during the first decade of life and less then 10% after the second. One third of untreated patients have more than one attack per month, while in another third, the disease tends to be mild with a few attacks per year. However, even patients with the mild form may experience life-threatening laryngeal edema [61]. Attacks can be spontaneous, but microtrauma and psychological stress are frequently recognized as triggering factors. Due to the disabling consequences of attacks and to the risk of fatality, treatment is a crucial issue in HAE. The classical approach comprises: (a) long-term prophylaxis, for patients with frequent severe attacks, (b) short-term prophylaxis, for patients exposed to potential triggers of laryngeal attacks (e.g. medical manipulation traumatizing the oropharynx), (c) treatment to revert an ongoing attack. Detailed reviews on the clinical approach to HAE and its treatment have recently been published, and therefore we will not further review this topic [41, 62–64].

Three new compounds are now being tested in clinical trials to assess their capacity to revert acute attacks in HAE patients. The first one, Dx88 (Dyax Corp., Cambridge, MA), is a novel recombinant protein with potent and specific inhibitory activity for plasma kallikrein. One randomized placebo-controlled and two open label studies have been performed so far with a total of 230 doses administered intravenously to more than 130 patients [59]. The second is a recombinant human C1-INH isolated from the milk of transgenic rabbits (Pharming Technologies BV, Leiden). Nine acute attacks have been treated in a phase II open label study; a randomized, controlled study is ongoing [65]. The third is Icatibant (Jerini AG, Berlin), a potent, specific bradykinin B2 receptor antagonist. Twenty attacks have been treated so far in an open label study, 12 by intravenous infusion and 8 by subcutaneous injection [66]. Randomized, controlled studies are ongoing with the subcutaneous formulation. For all three compounds, the available results appear to be favorable.

Ongoing clinical trials for treatment of acute attacks in HAE patients are not restricted to new compounds. Plasma-derived C1-INH represents an established treatment for acute attacks in HAE patients [67–70] and has been approved for this purpose in most European countries for 20 years or more. However, this product is not available in the United States. For this reason, ZLB Behring (Marburg) and Lev Pharmaceutical Inc. (New York, NY) have just started recruiting HAE patients for clinical trials with their plasma-derived C1-INHs.

C1 inhibitor and other diseases

Based on the wide range of biological activities of C1-INH, one could expect several clinical abnormalities to result from its deficiency. Instead, patients lacking C1-INH appear perfectly normal except for angioedema; also C1-INH K.O. mice, both heterozygous and homozygous, do not present obvious phenotypic abnormalities [55]. Nevertheless, there are several pathological conditions in which C1-INH has been reported to either play a pathogenic role or be a potential therapeutic tool. These conditions will be briefly analyzed below.

Ischemia–reperfusion injury

This term refers to the pathological events that follow restoration of blood flow to ischemic tissues. Ischemia rapidly damages metabolically active tissues, and reperfusion paradoxically initiates a cascade of pathologies that leads to additional cell or tissue injury. It can occur spontaneously during a pathological event, but its clinical relevance has grown tremendously with expansion of techniques aimed at restoring patency of occluded arteries and of organ transplantation. Complement activation and neutrophil stimulation are two major components in events leading to tissue injury [71]. During reperfusion, complement can be activated by exposure to intracellular components such as mitochondrial membranes or intermediate filaments. Two elements of the activated complement contribute directly or indirectly to damage: anaphylatoxins (C3a and C5a) and the membrane attack complex [72]. The use of specific inhibitors may find wide clinical application because there are no effective drug therapies currently available to treat I/R injuries [73]. In a model of I/R, organs of transgenic animals overexpressing C1-INH are protected from endothelial cell damage [74]. C1-INH has been tested as potential therapy for I/R injury either in models such as the ischemic muscle [75, 76] or in specific diseases.

I/R injury of the myocardium that follows coronary artery occlusion has been investigated in animal studies, in which it has been shown that C1-INH has the capacity to reduce the myocardial ischemia–reperfusion injury [77–80]. The effect of intravenous C1-INH following reperfusion therapy for acute myocardial infarction has also been tested in pilot human studies. Three patients received C1-INH as rescue therapy for severe reperfusion injury after coronary surgery, with rapid hemodynamic stabilization [81]. In one study, specifically designed to assess the effects of intravenous C1-INH following reperfusion therapy in 22 patients with acute myocardial infarction, this treatment was proved to be safe and effective in reducing complement activation [82]. Due to the small size of the population and the absence of a control group, it was not possible to reach a statistically significant conclusion on clinical effectiveness.

Brain damage following ischemia has recently become a major area of investigation with the possibility to exploit these studies for treatment of stroke, one of the major causes of death and disability worldwide, and for which there is still no effective therapeutic tool. Different animal models of ischemic brain have been studied. Important evidence for a neuroprotective activity of C1-INH has been provided by the group of Bergamaschini using a murine model of transient ischemia induced by introducing a nylon microfilament into the middle cerebral artery [83–85]. Analogous positive conclusions were also reached with other animal models of brain ischemia and of brain damage deriving from hypothermic circulatory arrest [86–88]. We expect that these results will be confirmed in controlled human studies.

I/R injury mediated by complement activation also occurs during organ transplantation. Such an activation and related pathological consequences can be prevented by C1-INH. C1-INH has been circulated into the organ before implant [89, 90] or added to the preserving solution that protects the organ during transport [91]. Exposure of organs to be transplanted to high concentrations of C1-INH significantly reduces such complications due to the capacity of C1-INH to bind endothelial cells maintaining intact functional capacity [92].

Septic shock

Severe sepsis and septic shock are inflammatory diseases triggered by bacterial infection. Despite antibiotics and supportive care, mortality remains extremely high varying from 20 to 50%. Complement and contact systems are activated in sepsis and probably involved in its pathogenesis [93, 94]. Specifically in plasma from septic patients, there is a relative deficiency of C1-INH due to its proteolytic inactivation by neutrophil elastase. In addition to these studies, the rationale for supplying C1-INH to patients with sepsis is further supported by animal models of this condition showing the protective effect of this protein [95–98]. When these studies have been transferred to humans, the positive effect of C1-INH has been confirmed either in anecdotal reports or in pilot studies [99–101]. However, none of these studies has been designed with appropriate control groups to reach a definitive and statistically proven conclusion.

Other diseases

In addition to the above conditions, diverse reports suggest additional clinical indications where C1-INH infusion could be beneficial. Experimental pancreatitis can be prevented by conditioning animals with human C1-INH. These findings have been confirmed in several different animal models [102–106] but are difficult to transfer to humans where treatment cannot obviously anticipate the development of the disease. Only few experiences have been reported in humans and in very specific situations. One is the report of two children who improved upon C1-INH after developing an acute pancreatitis following allogeneic hematopoietic stem cell transplantation [107]. The other is a pilot study performed by our group, in which we demonstrated that administration of C1-INH reduces hyperamylasemia consequent to endoscopic cholangio-pancreatography, a procedure that can be complicated by overt clinical pancreatitis [108].

Reduction in C1-INH plasma levels and complement activation have been involved in the pathogenesis of reactions to radiographic contrast media (RCM). The pathogenic mechanism causing this reaction is still not defined. Signs of activation of complement and contact systems have been detected during RCM infusion. These findings, along with the evidence for reduced plasma levels of protease inhibitors, and specifically of C1-INH, led to the hypothesis that reduced control of these systems could underlie RCM reactions [109–115]. This hypothesis has never been definitively proven, and supporting data have been hampered, at least partially, by hemodilution accompanying infusion of RCM with elevated osmotic activity that alters plasma protein concentration [116].

A final field of intervention for C1-INH is vascular leakage. It may occur spontaneously as in idiopathic capillary leak syndrome (Clarkson disease), a lethal disease characterized by recurrent fluid extravasations leading to hypovolemic shock underlain by marked hemoconcentration (hematocrit as high as 80%) [117]. However, a clinical picture dominated by fluid extravasation can complicate diseases such as septic shock or burns, or specific therapeutic interventions such as treatment with recombinant interleukin 2, bone marrow transplantation, and cardiopulmonary bypass. A role for C1-INH in these conditions is supported by evidence of activation of its target proteases in animal models and by anecdotal experience in humans [118–130].

In conclusion, we can see a broad area of intervention for C1-INH as a therapeutic tool. Some of these potential applications were identified long time ago, but we have not yet seen any controlled study supporting anecdotal reports or pilot experiences. In the last few years, a recombinant C1-INH, produced in transgenic animals, has been proposed for treatment of acute attacks in patients with HAE, and clinical trials with this product are, at present, in progress in North America and Europe [65]. We can expect that the introduction of this new product, along with the existing plasma derivative, will renew interest in exploiting C1-INH as a therapeutic agent in conditions other than genetic deficiency.