Chapter 19 The kallikrein-kinin-kininogen system

Chapter 19

The Kallikrein-Kinin-Kininogen System DONALD H. MILLER and HARRY S. MARGOLIUS

363

Introduction Kinins and Kininogens

364

The Plasma System Tissue Kallikrein Kinins Involvement in Physiological and Pathophysiological Processes Pain and Hyperalgesia Summary

366 370 371 377 378 381

INTRODUCTION The chief interest in the kallikrein-kinin systems over the years has been due to their actions in the circulatory system. These actions include participation in clotting reactions, and a potent vasodilating capability long suspected of being important to the regulation of blood pressure. But with the recent availability of kinin receptor antagonists as experimental tools and as potential therapeutic agents, attention has increasingly turned to the possible involvement of the system in pain transmission, allergy, inflammation, and asthma. The activities of the kallikrein-kinin systems are expressed after the proteolytic actions of a kallikre in enzyme on kininogens, substrates that exist in both high

Principles of Medical Biology, Volume 8B Molecular and Cellular Pharmacology, pages 363-384. Copyright 9 1997 by JAI Press Inc. All rights of reproduction in any form reserved. ISBN: 1-55938-813-7 363

364

DONALD H. MILLER and HARRY S. MARGOLIUS

molecular weight and low molecular weight forms. Plasma kallikrein is a serine protease that circulates in blood as prekallikrein, an inactive precursor, and participates in the surface-mediated defense system via cleavage of both factor XII and high molecular weight kininogen (HK). Tissue kallikrein, a serine protease structurally unrelated to plasma kallikrein, exerts its actions via the generation of kinins from low molecular weight kininogen (LK) in various tissues and/or their secretions. Figure 1 outlines these relationships as they exist in humans. We will describe each of these components in turn, with some discussion of the biochemistry and molecular biology of each, then take up their roles in human physiology and pathophysiology. The systems have been most thoroughly characterized in mammals, though evidence exists for the presence of kallikrein-kinin system components in lower animals. Much that we know has come from the study of the system in laboratory rats, but we will here restrict our discussion primarily to what is known of the system in man. Several exhaustive reviews have appeared that more comprehensively cover the subject (Bhoola et al., 1992; Hall, 1992). KININS AND

KININOGENS

Before taking up the actions of the kallikreins, it is helpful to understand the structure of their substrates and products, since, as shown in Figure 1, these are common to both. Both HK and LK derive from transcription of a single gene. Alternative mRNA splicing events determine which protein is produced, and Figure 2 shows the relationship between the two. HK (Fitzgerald or Williams factor) has a molecular weight of 120 kD and circulates in plasma at a concentration of 70-90 lag/ml. LK has a molecular weight of 68 kD and is found in plasma at 160 ~tg/ml. Kininogen is also found in interstitial fluid, and is thought to perfuse virtually every tissue in the body, a fact that may be of functional importance since tissue kallikrein is often found to be associated with cell surfaces. The kininogens are multifunctional proteins, and it has been possible to divide their structures into functional domains as seen in Figure 2 (Mtiller-Esterl et al., 1986). Domains 2 and 3 are sequences identified as having inhibitory activities toward enzymes known as cysteine proteases. Domain 1 has significant sequence homology to domains 2 and 3, but does not contain the cysteine protease inhibitor region. Domain 4 is the kinin domain, i.e., that part of the kininogen molecule that contains the sequence of the bioactive kinin peptide, either bradykinin or lysbradykinin (kallidin). Domain 5 of the HK molecule contains the region responsible for binding to surfaces in contact-mediated reactions, and domain 6 binds either prekallikrein or factor XI in the circulation. These functions and their interrelationships will be discussed later. Plasma kallikrein acts on HK to release bradykinin (BK), a nonapeptide whose sequence is shown in Figure 1. Tissue kallikrein can act on either HK or LK in vitro to release lys-bradykinin, but is thought to use chiefly, if not exclusively, LK in

K gene Transcript Tissue Kallikrein - ~

~ (i)

.?~o-p q,

R, r

' 9 0 ,,

Plasma Kallikrein

S

LK

HK'

Lys-BK

BK

.~ Surface Contact Reactions *- Cysteine Protease Inhibition

~

r

N/

G

/ !

"11 I

I

r I

i.?

,4"

jl

,i

. . . .

q~,,pam

9

9

,

,~

o

,;

:,

i.~

9 9'

-.

~

,

to

9 ,'

;-.~.

9

,,

B2 Receptor

oI ,,,

--_~.'~

Figure 1. Kallikreins, their kininogen substrates, and their active products. Transcription of the kininogen (K) gene is followed by alternative mRNA splicing reactions (~) that lead to the synthesis of either low molecular weight kininogen (LK) or high molecular weight kininogen (HK). Proteolytic cleavage of either kininogen by kallikrein releases the bioactive kinin peptides iys-bradykinin (Lys-BK) or bradykinin (BK). These peptides, whose amino acid sequences are shown, have random conformations in solution. Either may react with the B2 receptor, which is a protein typical of membrane receptors with seven transmembrane domains. 365

366

DONALD H. MILLER and HARRY S. MARGOLIUS

Cysteine Protease Inhibition

Cysteine Protease Inhibition I

D1

D2

Bk I

D3

Prekallikrein and Factor XI Binding

Surface Binding

I

D4

I

D5

I

D6

R

! I

I

i

HK

Figure 2.

Kininogen structure: The amino acid sequences of domains 1-4 (D1-D4) of HK and LK constitute the heavy chain of the molecule, and are identical, ending at residue 383. The sequences of the light chains differ from each other, with the light chain of LK terminating at residue 409, and that of HK at residue 626. When the kinin sequence is excised, the flanking sequences remain bound by a disulfide bridge, giving rise to the terms "heavy chain" and "light chain." Functions associated with the domains are described in the text.

vivo. Production of kinin does not result in fragmentation of the parent molecule, as the heavy chain (N-terminal to the kinin sequence, and common to both HK and LK) remains attached to the light chain (C-terminal to the kinin, and different for HK and LK) by one or more disulfide bridges. Bradykinin is the prototype kinin, and has been the most studied, though as far as is known, both kinins have virtually identical actions. These actions are carried out through interaction with kinin receptors, and both the actions and the receptors are discussed in greater detail below.

The Plasma System

Plasma kallikrein circulates as prekallikrein (Fletcher factor), an inactive precursor of molecular weight 88 kD. The gene sequence of the mature enzyme in the rat is known, and consists of 619 amino acids (Seidah et al., 1989). As mentioned above, the active enzyme releases bradykinin from an internal sequence of HK. These molecules are involved in a number of physiological processes.

Surface-Mediated Defense Reactions The first step in the intrinsic blood clotting pathway is traditionally thought to be the surface-mediated activation of factor XII to factor XIIa (Hageman factor), a protease. Factor XIIa converts factor XI to factor XIa, and also converts prekal-

The Kallikrein-Kinin-Kininogen System

367

likrein to active plasma kallikrein. Plasma kallikrein in turn proteolytically activates still more factor XII (Figure 3). The positive feedback evident at this step results in rapid activation of the surface-mediated clotting cascade. The active kallikrein also acts on HK to produce bradykinin, which may then participate in the inflammatory response to the injury. Because factor XIIa, plasma kallikrein, and factor XIa have all been shown to convert plasminogen to plasmin, these reactions may also have a function in fibrinolysis, though the rates of this conversion seem too slow to be of major physiological significance (Kaplan et al., 1987; Wachtfogel et al., 1993). The reactions shown in Figure 3 do not proceed without regulation, and C1 inhibitor and t~2-macroglobulin are the most important physiological inhibitors of plasma kallikrein. This is dramatically apparent in hereditary angioedema, a disease where the primary defect is a deficiency of C1 inhibitor. Unregulated plasma kallikrein activity gives rise to episodes of excessive bradykinin production, leading to localized edema in the gastrointestinal tract (abdominal pain and vomiting), the skin and subcutaneous tissue (blistering), and the upper respiratory tract (wheezing and dyspnea). HK is an essential cofactor for these activating reactions. HK is bound (via domain 6) to both prekallikrein and factor XI in the blood. This complex, either before, or perhaps more importantly even after, cleavage of the kinin sequence, is capable of binding to a surface (i.e., a site of vascular injury), where the activation by factor XIIa takes place. The surface-binding is mediated by a histidine-rich domain in HK (domain 5), and may be augmented after this domain has been exposed via the bradykinin cleavage. Binding of the HK-factor XI complex to the site also occurs via the same mechanism, resulting in activation to factor XIa. Thus, HK binding to an appropriate surface aligns prekallikrein, factor XII, and factor XI with each other and facilitates their interactions (Kaplan et al., 1987). Genetic deficiencies of these proteins have been described, such as lack of factor XII (Hageman factor), prekallikrein (Fletcher factor), or HK (Williams factor). While these conditions are detectable in laboratory clotting time tests, the individuals bearing these traits appear to live normal lives. This has led to the hypothesis that these contact-activated reactions are less important to the clotting cascade than previously thought, and may be more important to processes other than hemostasis.

An Alternative View The discovery of apparently healthy individuals with genetically determined deficiencies in HK has always been puzzling. A deficiency in factor XI, or any factor subsequent to factor XI in the clotting cascade, leads to significant bleeding disorders. But deficiencies in factor XII, prekallikrein, or HK, while detectable in laboratory clotting tests, do not lead to clinically significant bleeding abnormalities. This has led to a proposed revision of the model of coagulation that postulates that thrombin is the primary activator of factor XI, and not factor XIIa or kallikrein (Gailani et al., 1991).

368

D O N A L D H. MILLER and HARRY S. MARGOLIUS

> o

@~.o-/~.Ag O

9

+

Ka

X l l a -<

XII

XIa

Further Clotting Reactions

Figure 3.

Surface-mediated defense reactions involving plasma prekallikrein (PK), active plasma kallikrein (Ka), high molecular weight kininogen (HK), clotting factors Xll (XlI) and XI (XI), and activated clotting factors Xll (Xlla) and XI (Xla). The HK-PK and HK-Xl complexes are shown bound to a surface via domain 5 (D5) in HK. This binding facilitates the activation of prekallikrein and factor Xl by factor Xlla.

Factor XII, prekallikrein, and HK, on the other hand, have been proposed to be components of a system that regulates the delivery of bradykinin to cells. Specific binding sites for kininogens have been found on platelets, granulocytes, and

The Kallikrein-Kinin-KininogenSystem

369

endothelial cells, and these sites are almost certainly saturated in vivo. Kinins are thought to act in an autocrine or paracrine fashion, so that concentrating kininogen on cell surfaces may constitute a mechanism for insuring local delivery of high concentrations of the peptide to neighboring kinin receptors. Specific domains, as discussed above, on the heavy chains of LK and HK, and also on the light chain of HK, are known to be responsible for the binding, so the structure of the entire kininogen molecule plays a role in determining the biological activity. This view holds that unregulated activation of the contact system results in hypotension due to bradykinin release. In patients with bacterial sepsis, for example, the degree of hypotension correlates with the level of activation of the contact system. It has thus been suggested that agents that alter the binding of kininogen to cells may have modulatory effects on blood pressure (Schmaier, 1992). However, this interesting idea remains a hypothesis, since there is as yet no experimental evidence that cell-bound kininogen is the source of physiologically active kinin. The challenge to understand the autocrine nature of kinin action persists, and the solution to this problem will be important to understanding the role of the kallikrein-kinin system in blood pressure regulation, as well as other physiological processes discussed below.

Inflammation Kinins are importantly involved in the inflammatory response, as discussed in greater detail below. Generation of kinins by tissue kallikrein leads to an increase in capillary permeability, and the subsequent leakage of plasma into an injury site would be expected to bring plasma kallikrein to the area, which could contribute to the generation of still more kinin. Plasma kallikrein also has chemotactic activity for neutrophils, and this too may augment the response. Enzymatic activity is necessary for this, since prekallikrein, or kallikrein inactivated with an inhibitor, does not cause chemotaxis.

Cysteine Protease Inhibition Kininogens in the plasma also have another function. The heavy chains of both LK and HK are identical (see Figure 2), and domains 2 and 3 are highly homologous to the family of cysteine protease inhibitors. LK, in fact, was formerly known as 2+ ~l-cysteine protease inhibitor, an inhibitor of Ca -activated cysteine proteases (calpains), cathepsins B, H, and L, ficin and papain. This may have regulatory importance because platelet-derived calpains are capable of releasing kinins from kininogen during an inflammatory response. In so doing, therefore, they simultaneously create an inhibitor of their activity (the remainder of the kininogen molecule containing domains 2 and 3), providing negative feedback to this process.

370

DONALD H. MILLER and HARRY S. MARGOLIUS

Bacterial Septicemia Components of the plasma kallikrein-kinin system have been implicated in endotoxemia. In patients with hypotension due to gram-negative bacterial sepsis, for example, levels of functional factor XII and prekallikrein were significantly reduced compared to controls, implying activation had occurred. In following up this observation, an experimental model of bacterial sepsis in pigs was established, and inhibition of contact-activated proteases (such as factor XIIa, plasma kallikrein, factor XIa, and thrombin) significantly improved survival rate early in the infection, though the accompanying hypotension was not relieved (Colman, 1989). In another study of experimental sepsis in sheep, treatment with aprotinin, an inhibitor of plasma kallikrein, as well as of other serine proteases, effectively prevented the hypotension and the adverse effects on hemodynamics and renal function (Cumming et al., 1992). The interesting possibility that inhibition of the contact system as an adjunct therapy in infectious disease remains to be explored. Tissue Kallikrein

Tissue kallikrein in humans, coded for by the KLK1 gene, is a member of a multigene family that specifies three closely related enzymes. In addition to tissue kallikrein, these are prostate specific antigen (PSA) and hGK- 1. PSA is not thought to release kinins from kininogen, but is believed to be involved in the liquefaction of seminal fluid clots via proteolytic attack on semenogelin. Also, increased serum PSA levels are a marker for prostatic cancer. The function of hGK-1 has not been determined, but the amino acid sequence at its active site suggests that kininogen does not serve as a substrate for it either. The actions of tissue kallikrein, on the other hand, are thought to derive from its ability to process kininogen and release kinins. Either HK or LK can serve as substrates in vitro (though LK is a better substrate), but whether both serve as in vivo substrates is not known. In man, the enzyme consists of a single-chain polypeptide of 238 amino acids. The protein is acidic, glycosylated, and belongs to the trypsin superfamily of serine proteases. It appears as a proenzyme in most tissues and secretions, the exception being the submaxillary gland and saliva, where no proenzyme form is found. Large amounts of kidney-synthesized protein are secreted into urine, and urinary levels have historically been taken to be indicators of renal kallikrein-kinin system activity, although this assumption may not be strictly true.

Localization and Function Tissue kallikrein has been localized to various tissues by biochemical and immunohistochemical techniques. These tissues include the kidney, salivary glands, pancreas, intestine, pituitary gland, brain, vasculature, and serum. Ideally

The Kallikrein-Kinin-Kininogen System

3 71

a function could be ascribed for the enzyme in each of these locations, but one of the enduring problems concerning the kallikrein-kinin field is that accurately defining physiological roles has been difficult. Indeed, it has been suggested that the function of kallikrein may be tissue specific. Apart from its processing of kininogen to kinin, with the various activities associated with kinins (see below), other processing roles have been hypothesized, such as activation of the proenzyme of prolactin in the anterior pituitary, prorenin in the kidney, and proinsulin in the pancreas. Whether these activities, demonstrable in vitro, have physiological significance is not known. In any case these hypotheses derive from experiments with rats. Since the kallikrein gene family in the rat consists of, not three, but 20 closely related genes, many of whose proteins have unknown activities, it is unclear how pertinent these observations are with respect to humans. Whether human tissue kallikrein will be found to have important roles in the processing of zymogens or hormones remains to be established, but there is general agreement that a primary function of this enzyme is to generate kinin peptides.

Kinins Structure The prototype kinin is bradykinin, a nine amino acid peptide that is formed in man by the action of plasma kallikrein on kininogen (probably HK). Lysylbradykinin, also known as kallidin, is formed by the action of tissue kallikrein on kininogen (probably LK). Hyp3-kinins have been found in human urine, probably arising from a posttranslational modification in the kininogen molecule, but it is not known if this modification has any functional significance. Kinin-like peptides with sequences highly similar to that of bradykinin have also been shown to exist in birds, fish, amphibians, reptiles, wasps and hornets, suggesting that this peptide has been importantly involved in living systems for a phylogenetically long time (Bhoola et al., 1992). Activities

The in vitro activities of bradykinin and kallidin are virtually identical. Many physiological functions can be shown to be affected, including effects on blood pressure, smooth muscle contraction, local blood flow, and electrolyte transport. Some of these effects are direct, whereas others are indirect, involving the actions of other cellular mediators.

Blood pressure. Systemically administered kinins are potent but transient vasodilators. Blood pressure can be dramatically lowered by kinins, but the effects of a single dose wear off in minutes due to the very short half-life of kinins in plasma (-15 seconds). Chronic administration of kinins does not appreciably

372

D O N A L D H. MILLER and HARRY S. MARGOLIUS

extend the duration of the hypotensive effect because the tissue becomes rapidly desensitized. Kinins administered into the brain ventricular system, have the opposite effect, i.e., a pressor effect. The in vivo significance of these actions is not entirely clear, and it is thought that peripherally synthesized kinins probably have their greatest effects on local vascular beds in an autocrine or paracrine mode of action, rather than on systemic blood pressure as circulating hormones.

Smooth muscle contraction. Kinins can either contract or relax isolated smooth muscle preparations. Uterine muscle and ileum from the guinea pig are contracted, whereas duodenum from the rat is relaxed. Airway smooth muscle is biphasic, being relaxed at low concentrations, but contracted at high concentrations. Isolated vascular smooth muscle preparations respond variably, with artery preparations generally being relaxed, and vein preparations being contracted. Many of these effects are due to the kinin-induced release of other mediators. Relaxation responses, such as that of the isolated human basilar artery, are dependent upon the release of eicosanoids, and the dilation of certain vascular beds is dependent upon the generation of nitric oxide ("endothelium-derived relaxing factor" or EDRF) in the endothelium. This last endothelium-dependent response is currently creating excitement in the field. A role for kinins in the regulation of renal and coronary blood flow is becoming established (see below), and the discovery of the role of nitric oxide is increasing our understanding of these phenomena and of the kallikrein-kinin system. Increase in capillary permeability. Application of kinins to a vascular endothelial cell layer results in contraction of the cells and rupture of intercellular junctions. Plasma can then flow through the cell layer to the surrounding tissue, causing localized edema and delivering plasma proteins to the site. This effect is important to the role of kinins in inflammation, and is responsible for the manifestations of hereditary angioedema discused earlier.

Local blood flow. There is now much evidence suggesting an important role for kinins in the regulation of blood flow in the heart and the kidney. In the kidney, kinins decrease both afferent and efferent arteriolar resistance, though the actions on renal blood flow probably occur in conjunction with the actions of other regulators as well, such as angiotensin II, endothelin, atrial natriuretic peptide, and vasopressin. A finely regulated balance between the actions of kinins and these and other regulators is thought to be a major determinant of renal blood flow. Kinins also dilate coronary arteries, resulting in improved coronary circulation. Indeed, experiments in animals have demonstrated a cardioprotective effect of kinins in response to local ischemic events, as will be discussed in greater detail below.

The Kallikrein-Kinin-Kininogen System

373

Electrolyte transport. Kinins have been shown to influence the transport of ions across epithelia in a variety of experimental settings. An augmented secretion of chloride from intestinal or airway epithelia occurs in response to kinin administration, with a concomitant passive movement of sodium and water in the same direction. In the guinea pig gall bladder, however, kinins stimulate the movement of bicarbonate, and in human sweat glands, the movement of sodium. How these actions relate to other systems involved in the regulation of ion and water movements is unclear, but it seems safe to infer that the kallikrein-kinin system has a function in salt and water homeostasis. Release of other mediators. Perhaps one of the most important actions of kinins is that on various cell types to bring about the release of an array of other biologically active substances. Indeed, of the above kinin actions, only those on electrolyte transport have been shown to be partially due to direct effects. Even there, kinin-induced prostaglandin formation is important to the total response. On various cell types under various conditions kinins cause production and/or release of eicosanoids, nitric oxide, substance P, neurokinin A, cytokines, histamine, and serotonin. Some of the physiological consequences of these complex interactions will be considered in detail below, but it should already be clear that a particular biological effect as a direct response to kinin is difficult to discern when multiple responses are initiated.

Kinin Receptors To date, two kinin receptor types have been identified. The B. receptor is defined . . . . by its affimty for des[Arg ]-bradyklnln, a metabohte of bradykmm. Bradykmln itself has little effect at B 1 receptors, and instead exerts its actions through B 2 receptors. The B 1 receptor is evident only under special circumstances, such as tissue inflammation or trauma. In rabbits, for example, intravenous injection of E. coli endotoxin led to a hypotensive response to des[Arg9]-bradykinin. Control animals not injected with the toxin failed to respond. One report has suggested that interleukin- 1 may be the signal for B 1 expression. The only demonstration of a B 1 receptor in human tissue was its occurrence in a human fibroblast cell line. The role of the B 1 receptor in pathophysiology is not clear. The possibility that it may be involved in the response to tissue trauma is intriguing, and deserves further study. The B 2 receptor, however, mediates all the kinin activities discussed above, and there is a great deal more known about its structure and mechanism of action. The human B 2 receptor gene has been cloned (Hess et al., 1992), and its sequence suggests it to be a member of the G-protein-coupled family of receptors, proteins 9

9

.

.

.

I

374

D O N A L D H. MILLER and HARRY S. MARGOLIUS

predicted to cross the membrane seven times. The synthesis of effective antagonists to the B 2 receptor has led to new insights regarding the physiological role of kinins, some of which will be discussed below. The possible existence of B 3 receptors, or B 2 receptor subtypes, are questions that have not yet been resolved. On the basis of low affinities of B 2 antagonists for kinin receptors in a guinea pig trachea preparation, and the failure of these antagonists to displace 3H-bradykinin in binding experiments, a B 3 kinin receptor was proposed (Farmer et al., 1989). It has been suggested, however, that these differences may be due to species differences, so a third, distinct kinin receptor cannot yet be considered to be established (Hall, 1992), and the question is still open. Differences in agonist and antagonist binding in diverse tissues within the same species have led to proposals that B 2 receptor subtypes exist, and in at least one case different functional responses were observed when bradykinin was applied to the apical side of canine tracheal epithelial cells versus application to the serosal side (Denning et al., 1991). The evidence is suggestive of different subtypes, but not yet conclusive. The issue of distinct receptors, or distinct receptor subtypes, will not be resolved until different molecular entities are demonstrated.

B2 Kinin Receptor Antagonists The first effective and specific B 2 receptor antagonists were discovered in 1985, when it was found that substitution of D-Phe for Pro at position seven of the bradykinin molecule inhibited bradykinin-induced responses in the rat and the guinea pig (Vavrek et al., 1985). A series of analogs designed around this modification was developed, and for the first time it became possible to design experiments that directly tested for in vivo kinin actions. Since that time, further developments have led to antagonists with significantly greater potency and longer in vivo half-lives (Hock et al., 1991; Wirth et al., 1991 ; Cheronis et al., 1992). Development of still more effective antagonists continues. Peptide analogs suffer from their inherent in vivo instability, whether taken orally or administered parenterally. This makes them difficult to deliver to the required site of action in high enough concentrations, and with therapeutically useful half-lives. Non-peptide mimetics, such as have been developed for angiotensin II and tachykinin receptors, may overcome this problem, but only a few reports have yet appeared (Calixto et al., 1991; Salvino et al., 1993). Creation of compounds with efficacy in man is the goal, since such drugs would have exciting therapeutic potential in asthma, rhinitis, endotoxic shock, and certain inflammatory disorders, to name just a few (see below).

Signal Transduction As mentioned above, the B 2 receptor is coupled to G-proteins. There is evidence that the actions of both G o and Go~i2 are activated by kinins in rat dorsal root

The Kallikrein-Kinin-Kininogen System

375

ganglia cells (Ewald et al. ' 1989), and G q proteins in NG108 cells (Gutowski et al., 1991). Kinin receptors, therefore, are coupled to multiple G-proteins, and the coupling is cell-type specific. Two second messenger pathways activated by kinin receptors have been characterized, using a variety of experimental systems (Figure 4). One leads to an increase in intracellular calcium the result of G-protein-mediated activation of phospholipase C. Inositol trisphosphate, which mobilizes intracellular calcium stores, and diacylglycerol, which activates protein kinase C, are both generated by this enzyme. The second pathway results from activation of phospholipase A 2, and results in the metabolism of arachidonic acid and the production of various eicosanoids. Direct coupling of the receptor to phospholipase A 2 via a G-protein has not been demonstrated. Whether the same B 2 receptor, or different B 2 receptor subtypes, participate in these reactions is not known. How many of the possible combinations of kinin receptor or receptor subtype-G-protein-effectors exist, and how they relate to the tissue specific actions of kinins, is also not known, but this is an area of intense interest and research. Both cAMP and cGMP are elevated by kinins in some cell types, but these are probably indirect actions. The activation of adenylyl cyclase occurs as a result of the production of eicosanoids, and the activation of guanylyl cyclase as a result of the production of nitric oxide. In vascular endothelium, for example, kinins stimulate the formation of nitric oxide, which in turn leads to the activation of guanylyl cyclase, elevation of cGMP, and smooth muscle relaxation. Kinins, therefore, are capable of initiating cellular responses through a variety of signaling systems. Much of the experimental data has been obtained using isolated systems, however, or cultures of cells such as vascular smooth muscle, fibroblasts, or neuroblastoma-glioma hybrids. The relevance of these data to the in vivo roles of kinins, and how they may be integrated into other cellular activities, remains an active area of research. Kininases

Various peptidases have been shown to degrade kinins in vivo (Erd6s, 1990). While not specific for kinins, these enzymes can be considered to contribute to the regulation of kinin action by terminating their activities. Angiotensin converting enzyme (ACE), also known as kininase II, is one of these. Action of this enzyme on kinins results in cleavage of the terminal Phe-Arg dipeptide, and complete loss of kinin activity. ACE is a circulating plasma enzyme, but is also found in the vascular endothelium. It is found in high concentrations in lung tissue, and in especially high concentrations on the brush border of renal proximal tubules. It is now clear that the therapeutic action of the ACE inhibitors is due, not only to inhibition of angiotensin II synthesis, but also to the inhibition of kinin degradation, with the presumed result of elevating local kinin concentrations. This has been

.~I

m "U

~Eicosanoids ,

-_

B2RF'cgM~oR . . . . . . .

.

.

.

.

.

.

.....

____

,

,

:

........

----~-

....

,,,,,

....

,

Ca2+ PKC cAMP

Blood Pressure/Local Blood Flow (vasodilation) C ~

ElectrolyteTransport (epithelialcells)

~ "~Y2..-

~

~ "~'..":i'..-..

Capillary Permeability (endothelialcells)

Mediator Release (various cell types) NO, histamine, C G R P , substance P, serotonin, eicosanoids, etc.

Figure 4. Kinin signaling through the B2 receptor: Occupation of the receptor by a kinin may activate a G protein-mediated pathway that causes hydrolysis of phosphatidylinositol-4-5-bisphosphate (PIP2) due to the action of the phospholipase C (PLC). Inositol trisphosphate (IP3) releases calcium from internal stores, and diacylglycerol (DAG) activates protein kinase C (PKC). The B2 receptor may also activate phospholipase A2 (PLA2), through incompletely understood mechanisms that may be indirect. PLA2 hydrolyses arachidonic acid (AA) to release eicosanoids, which act through an eicosanoid receptor (ER) to activate adenylyl cyclase (AC) and produce cAMP. cAMP, calcium, and PKC all participate in further signaling pathways to ultimately produce biological responses. 376

The Kallikrein-Kinin-KininogenSystem

377

particularly evident in studies of kinin function in cardiac tissue, as discussed below in the section on cardioprotection (Linz et al., 1992). Neutral endopeptidase 24.11 (NEP) has been implicated as a major contributor to kinin degradation in the kidney. This enzyme also cleaves the C-terminal Phe-Arg from kinin, and is located in the brush border of kidney proximal tubules in approximately the same concentration as ACE. Using specific inhibitors for either NEP or ACE, however, revealed that NEP is responsible for 68% of the total kininase activity in the rat kidney, while ACE is responsible for only 23% of the total. Treatment of rats with the NEP inhibitor phosphoramidon significantly raised both urinary volume and urinary sodium excretion, suggesting that endogenously formed kinins may be involved in the regulation of these phenomena, and that modulation of NEP is potentially a mechanism for regulating kinin activity. At least two other peptidases, carboxypeptidase N and carboxypeptidase M, have also been suggested to have in vivo kininase activities. Carboxypeptidase N is a circulating enzyme, and carboxypeptidase M is found in membranes of various tissues, including placenta, kidney, lung, and pulmonary endothelial cells. Both enzymes cleave the C-terminal Arg from the kinin, but the physiological roles of these molecules with regard to regulation of kinin activities is less well defined than those of ACE and NEP.

Involvement in Physiological and Pathophysiological Processes In the previous discussion, we have delineated the various parts of the kallikreinkinin system, and considered their properties and actions. Explaining how these actions are intergrated into a physiological context is less straightforward. Reasonable hypotheses exist to explain how and why the system functions in the maintenance of normal activities of cells and tissues, but they remain hypotheses; better established is the participation of kinins as pathophysiological mediators in various disease states.

Inflammation The case for the participation of kinins in the inflammatory process is persuasive (Proud et al., 1988). Activation of the surface-mediated defense system after tissue injury would result in the formation of bradykinin, and activation of tissue kallikrein would lead to lys-bradykinin formation. The kinins could then (1) activate C-fibers to produce pain (see below), (2) increase vascular permeability, leading to extravasation of plasma, with swelling as a result, (3) cause vasodilation of the capillaries, leading to increased blood flow (i.e., erythema and warmth), and constriction of venous vessels, leading to further edema, and (4) recruit leukocytes to the area, some of which can be activated by kinins to release other inflammatory mediators (e.g., cytokines, histamine). This scheme is supported by the finding of kallikrein, kininogen, and kinins in inflammatory exudates, and by the appearance of all the symptoms of inflamma-

378

DONALD H. MILLER and HARRY S. MARGOLIUS

tion, subsequent to subcutaneous injection of kinins in human volunteers. Furthermore, in some, but not all, animal models of inflammation the use of B 2 receptor antagonists can block the response. The injection of carrageenin or urate crystals into the rat paw, for example, causes extravasation of plasma into the area and edema. Prior treatment of the animals with B 2 antagonists prevents the edema, implying that endogenously produced kinins were active in the response. Injection of zymosan also causes edema, but in this case the B 2 antagonists do not prevent the response (Damas et al., 1992). There is evidence that B 1 receptors may also be involved, at least in chronic inflammation. The B 1 antagonist [des-Arg9LeuS]-bradykinin was much more effective in reducing a hyperalgesic state induced by persistent inflammation in the knee joint of the rat than was a B 2 antagonist (Dray, et al. 1993b). The notion is appealing because B 1 receptors only appear after tissue trauma or injury. The entry of plasma into the area would introduce kininase I, thereby creating the potential for des-Arg9-bradykinin production, and interaction with the B l receptors. Pain and Hyperalgesia

It has long been realized that kinin concentrations are increased in injured tissue to levels capable of producing pain, and that exogenous administration of kinins potently induces a nociceptive response. But not until the last few years has it been possible to convincingly demonstrate that endogenous kinins are involved in pain and hyperalgesia (Steranka et al., 1988). Kinin receptors in the guinea pig have been localized by autoradiography to sensory neurons thought to be involved in nociceptive pathways, providing support for such a role. Bradykinin causes depolarization of sensory neurons and nociceptive fibers, and therefore direct excitation, but also elicits production of eicosanoids, leading to sensitization of nociceptors and hyperalgesia (Dray et al., 1993b). Several animal pain models have confirmed these effects, and highlighted the potential therapeutic utility of kinin receptor antagonists. The pain response induced in rats by the injection of acetic acid can be blocked by antagonists of B 2 receptors, suggesting that endogenous kinin production is involved. In a model of hyperalgesia, urate crystals are injected into the paw of a rat, then pressure is applied by a device that measures the force exerted until the rat pulls the paw away. Urate crystals cause significant hyperalgesia, since the amount of force required before the rat first perceives pain is reduced, but concomitant injection of a B 2 receptor antagonist prevents this response (Steranka et al., 1988).

Endotoxic Shock Since kinins are such potent vasodilators, involvement in the marked fall in blood pressure during the early stages of endotoxic shock has often been postulated.

The

Kallikrein-Kinin-KininogenSystem

3 79

Experiments in several laboratories employing animal models of endotoxic shock have shown that treatment with kinin receptor antagonists prior to the induction of the shock essentially inhibits the hypotensive response (Wilson et al., 1989; Whalley et al., 1992). Blockade of kinin actions may therefore be of use in the prevention or treatment of endotoxic shock in human patients.

Diabetes Mellitus Experiments with animal models of diabetes mellitus have shown that animals made moderately hyperglycemic have increased levels of urinary kallikrein, and increased rates of renal kallikrein synthesis. These animals also show glomerular hyperfiltration, but when treated with aprotinin or a B 2 antagonist, renal blood flow and glomerular filtration rate are returned to normal. These data have led to the hypothesis that kinin action is involved in the increased renal blood flow due to effects on the vasculature. Kinins also cause rapid growth of some cell types, and therefore may also be involved in promoting the characteristic glomerular pathology associated with diabetes. Increased urinary kallikrein excretion has been observed in human patients with diabetes mellitus, and the amounts of kallikrein excreted correlated positively with glomerular filtration rate. Kallikrein excretion also correlated with distal tubular sodium absorption. Kinin effects on both blood flow and epithelial ion transport are therefore suspected, but no studies with humans have been conducted to test this.

Hypertension The fact that populations of essential hypertension patients excrete less kallikrein in their urine than do populations of normotensives has been recognized for a long time (Margolius, 1995). The possibility that subnormal kinin-mediated renal vasodilation might therefore contribute to the hypertension has spurred many investigations, including corroborating data that urinary kininogen and kinin are also reduced, and total renal kininase activity is increased in essential hypertension. Epidemiological studies have revealed a familial aggregation of kallikrein excretion, and families with the highest levels of urinary kallikrein had lower blood pressures than families with low levels of urinary kallikxein. Another study concluded that a high urinary kallikrein excretion rate represented a genotype associated with a reduced risk of essential hypertension (Berry et al., 1989). Still more convincing evidence has come from the use of the B 2 receptor antagonist Hoe- 140 in animal models. Chronic administration to rats of deoxycorticosterone, a steroid with salt-retaining actions, does not significantly affect their blood pressure unless the antagonist is co-administered. In that case, blood pressure rises continuously for at least six weeks. In rats not given the steroid, however, Hoe- 140 has no effect on blood pressure (Madeddu et al., 1993). This and other experiments

380

D O N A L D H. MILLER and HARRY S. MARGOLIUS

support the notion that the kallikrein-kinin system functions to defend against pathological increases in blood pressure, but may be relatively inactive in the absence of pressure-raising stimuli. Research with animal models of hypertension continues.

Asthma The effects of bradykinin in airway tissues distinguish it as a mediator in the genesis and maintenance of asthma (Farmer, 1991; Barnes, 1992). Bradykinin acts on airway smooth muscle to bring about bronchoconstriction, and at the same time releases other mediators from epithelia and endothelia that are also bronchoconstrictors (e.g., substance P, neurokinin A, eicosanoids). Bradykinin also activates nociceptive nerve fibers, bringing about the tightness in the chest and coughing associated with an asthma attack. Approximately 10% of patients treated with angiotensin converting enzyme inhibitors (ACE inhibitors) develop a cough as a side effect, and it has been suggested that the presumptive increase in kinins is the cause. Inhalation of bradykinin by asthmatic subjects brings on symptoms of an asthma attack, but there is no effect on non-asthmatic volunteers. Inhalation of bradykinin by guinea pigs or sheep causes bronchoconstriction, and prior treatment with B 2 antagonists prevents this. As in hypertension, activation of the kallikrein-kinin system is clearly implicated in asthma, but its contribution to the total symptomatology is unknown.

Rhinitis Kinins appear in lavage fluid from the nasal cavity of volunteers with rhinitis caused by either allergen challenge or rhinovirus infection. No kinin could be demonstrated in control subjects, however, or in subjects who were asymptomatic. When bradykinin was administered intranasally, symptoms typical of rhinitis resulted, such as sore throat, nasal obstruction, and rhinorrhea (Proud et al., 1988). The evidence that kinins are involved in symptoms associated with rhinitis is convincing, but as in asthma, other mediators such as histamine and various eicosanoids are also present. Given that kinins potently release these and other mediators from mast cells and tissue, it is tempting to think that kinin production may occupy a position near the beginning of such mediator cascades, and therefore be a prime target for therapeutic intervention with kinin receptor antagonists, but this has not been definitively shown.

Cardiop ro tection ACE inhibitors are known to have protective effects on cardiac tissue when cardiac function is threatened. A significant component of this protection can now

The Kallikrein-Kinin-KininogenSystem

381

be ascribed to the elevation of endogenous kinin levels. Reperfusion injuries subsequent to ischemia in isolated rat hearts can be successfully treated with an ACE inhibitor, and also by administration of low doses of bradykinin. The beneficial effects of either treatment, however, could be blocked by pretreatment with a B 2 receptor antagonist (Linz et al., 1992). Likewise, ligation of the left descending coronary artery in dogs produced an infarction, the size of which was reduced by half by perfusing the ACE inhibitor into the main stream of the left coronary artery for 30 min before and during the occlusion. When however, a B 2 antagonist was perfused with the ACE inhibitor, no reduction in infarct size occurred. Left ventricular hypertrophy induced in rats by banding the aorta can also be prevented. Chronic treatment of rats with a dose of an ACE inhibitor that had no effect on blood pressure during the six weeks of the experiment prevented the development of hypertrophy, but again, when the B 2 antagonist was administered along with the ACE inhibitor, left ventricular hypertrophy was indistinguishable from controls. The effects of the B 2 antagonist in all three situations suggest that endogenously formed kinins were at least partially responsible for the protective effects of the ACE inhibitors (Linz et al., 1992).

Other Diseases Indications of excessive activity of the kallikrein-kinin system in other conditions exist. For example, there is evidence that kinin actions are involved in trauma-or ischemia-induced cerebroarteriolar abnormalities, as well as in the pathophysiology of parenchymal and subarachnoid hemorrhage (Francel, 1992). Involvement in carcinoid flushing, postgastrectomy dumping syndrome, and inflammation of the intestinal mucosa has also been noted. In none of these, or in any of the above, can we say with certainty whether anomalous activities of the kallikrein-kinin system cause, or are the result of the disease. What is clear is that activation of the system at least contributes to the signs and symptoms of these pathophysiological conditions.

SUMMARY Elements of the kallikrein-kinin system are involved in such diverse activities as surface-mediated defense reactions, regulation of blood flow and blood pressure, smooth muscle contractility, nociception, electrolyte transport, and mediator release. Such a variety of actions suggests the potential for wide-ranging biological significance, and we have reviewed the involvement of the system in various physiological and pathophysiological situations. It is worth reiterating, however, that kinin actions are unlikely to be the sole regulatory influences in such conditions. The contractile state of the renal microvasculature, for example, is the result of the vasoconstrictor actions of angiotensin It, endothelin, and vasopressin, and

382

DONALD H. MILLER and HARRY S. MARGOLIUS

the vasodilator actions of atrial natriuretic peptide (ANP) and kinins, to name just the peptidergic controls we know about. Add to these influences the contributions from adrenergic and cholinergic systems, plus the notion that regulation of p r e g l o m e r u l a r vessels is different from that o f p o s t g l o m e r u l a r vessels, and it is clear that even a thorough understanding of kinin actions and their regulation would lead to only an incomplete understanding of the regulation of renal blood flow and glomerular filtration rate (Carmines et al., 1990). Similarly, the mediation of pain, and the s y m p t o m s of asthma and inflammation are doubtless due in part to kinin actions, but are also due to actions of substance P, neurokinin A, prostaglandins, leukotrienes, and cytokines (Dray et al., 1993a). The fact that bradykinin elicits the production and/or release of all of these underlines its importance as a plausible initiator of these conditions, and hence the therapeutic potential of kinin receptor antagonists, but also emphasizes that kinin actions often occur as only a single c o m p o n e n t of a complex interplay of many influences. Nevertheless, the impending development of specific antagonists with high affinity for kinin receptors in man is noteworthy. Such a development may signal the introduction of significant, new therapeutic agents for the treatment of human disease.

REFERENCES Barnes, P.J. (1992). Bradykinin and asthma. Thorax 47, 979-983. Berry, T.D., Hasstedt, S.J., Hunt, S.C., Wu, L.L., Smith, J.B., Ash, K.O., Kuida, H., & Williams, R.R. (1989). A gene for high urinary kallikrein may protect against hypertension in Utah kindreds. Hypertension 13, 3-8. Bhoola, K.D., Figueroa, C.D., & Worthy, K. (1992). Bioregulation of kinins: kallikreins, kininogens, and kininases. Pharmacol. Revs. 44, 1-80. Calixto, J.B., Yunes, R.A., Rai, G.A., & Madeiros, Y.S. (1991). Nonpeptide bradykinin antagonist. In: Bradykinin Antagonists. Basic and Clinical Research (Burch, R.M. Eds.), pp. 97-129. Marcel Dekker, Inc., New York. Carmines, P.K., & Fleming, J.T. (1990). Control of the renal microvasculature by vasoactive peptides. FASEB J. 4, 3300-3309. Cheronis, J.C., Whalley, E.T., Nguyen, K.T., Eubanks, S.R., Allen, L.G., Duggan, M.J., Loy, S.D., Bonham, K.A., & Blodgett, J.K. (1992). A new class of bradykinin antagonists: synthesis and in vitro activity of bissuccinimidoalkane peptide dimers. J. Med. Chem. 35, 1563-1572. Colman, R.W. (1989). Contact systems in infectious disease. Revs. Infect. Dis. 11, suppl 4, $689-$699. Cumming, A.D., & Nimmo, G.R. (1992). Hemodynamic, renal, and hormonal actions of aprotinin in an ovine model of septic shock. Crit. Care Med. 20, 1134-1139. Damas, J., & Remacle-Volon, G. (I 992). Influence of a long-acting bradykinin antagonist, Hoe 140, on some acute inflammatoryreactions in the rat. Eur. J. Pharmacol. 211, 81-86. Denning, G.M., & Welsh, M.J. (1991). Polarized distribution of bradykinin receptors on airway epithelial cells and independent coupling to second messenger pathways. J. Biol. Chem. 266, 12932-12938. Dray, A., & Bevan, S. (1993a). Inflammation and hyperalgesia: highlighting the team effort. Tr. Pharmacol. Sci. 14, 287-290. Dray, A., & Perkins, M. (1993b). Bradykinin and inflammatory pain. Tr. Neurosci. 16, 99-104. Erd6s, E.G. (1990). Some old and some new ideas on kinin metabolism. J. Card. Pharmacol. 15(Suppl. 6), $20-$24.

The Kallikrein-Kinin-KininogenSystem

383

Ewald, D.A., Pang, I.-H., Sternweis, P.C., & Miller, R.J. (1989). Differential G protein-mediated coupling of neurotransmitter receptors to Ca 2§ channels in rat dorsal root ganglion neurons in vitro. Neuron 2, 1185-1193. Farmer, S.G. (199 I). Role of kinins in airway diseases. Immunopharmacology 22, 1-20. Farmer, S.G., Burch, R.M., Meeker, S.A., & Wilkins, D.E. (1989). Evidence for a pulmonary B 3 bradykinin receptor. Molec. Pharmacol. 36, 1-8. Francel, P.C. (1992). Bradykinin and neuronal injury. J. Neurotrauma, 9 (Suppl. 1), $27-$45. Gailani, D., & Broze, G.J., Jr. (1991). Factor XI activation in a revised model of blood coagulation. Science, 253,909-912. Gutowski, S., Smrcka, A., Nowak, L., Wu, D., Simon, M., & Sternweis, P.C. (1991). Antibodies to the o~ subfamily of guanine nucleotide-binding regulatory protein ~ subunits attenuate activation o~ phosphatidylinositol 4,5-bisphosphate hydrolysis by hormones. J. Biol. Chem. 266, 20519-20524. Hall, J.M. (1992). Bradykinin receptors: Pharmacological properties and biological roles. Pharmac. Ther. 56, 131-190. Hess, J.F., Borkowski, J.A., Young, G.S., Strader, C.D., & Ransom, R.W. (1992). Cloning and pharmacological characterization of a human bradykinin (BK-2) receptor. Biochem. Biophys. Res. Comm. 184, 260-268. Hock, F.J., Wirth, K., Albus, U., Linz, W., Gerhards, H.J., Wiemer, G., Henke, S., Breipohl, G., KOnig, W., Knolle, J., & Sch61kens, B. A. (1991). Hoe 140 a new potent and long acting bradykinin-antagonist: in vitro studies. Br. J. Pharmacol. 102, 769-773. Kaplan, A.P., & Silverberg, M. (1987). The coagulation-kinin pathway of human plasma. Blood 70, 1-15. Linz, W., & Sch61kens, B.A. (1992). Role of bradykinin in the cardiac effects of angiotensin-converting enzyme inhibitors. J. Card. Pharmacol. 20 (Suppl. 9), $83-$90. Madeddu, P., Anania, V., Parpaglia, P.P., Demontis, M.P., Varoni, V., Fattaccio, M.C., & Glorioso, N. (1993). Chronic kinin receptor blockade induces hypertension in deoxycorticosterone-treated rats. Br. J. Pharmacol. 108, 651-657. Margolius, H.S. (1995). Kallikreins and kinins. Some unanswered questions about system characteristics and roles in human disease. Hypertension 26, 221-229. Mtiller-Esterl, W., Iwanaga, S., & Nakanishi, S. (1986). Kininogens revisited. Tr. Biochem. Sci. 11, 336-339. Proud, D., & Kaplan, A.P. (1988). Kinin formation: mechanisms and role in inflammatory disorders. Ann. Rev. Immunol. 6, 49-83. Salvino, J.M., Seoane, P.R., Douty, B.D., Awad, M.M.A., Dolle, R.E., Houck, W.T., Faunce, D.M., & Sawutz, D.G. (1993). Design of Potent Non-Peptide Competitive Antagonists of the Human Bradykinin-B2 Receptor. J. Med. Chem. 36, 2583-2584. Schmaier, A.H. (1992). Modulation of the cell membrane expression of the kininogens regulates the rate of bradykinin delivery to cells. Trends Cardiovasc. Med. 2, 108-114. Seidah, N.G., Ladenheim, R., Mbikay, M., Hamelin, J., Lutfalla, G., Rougeon, F., Lazure, C., & Chr6tien, M. (1989). The cDNA structure of rat plasma kallikrein. DNA 8, 563-574. Steranka, L.R., Manning, D.C., DeHaas, C.J., Ferkany, J.W., Borosky, S.A., Connor, J.R., Vavrek, R.J., Stewart, J.M., & Snyder, S.H. (1988). Bradykinin as a pain mediator: Receptors are localized to sensory neurons, and antagonists have analgesic actions. Proc. Natl. Acad. Sci. USA 85, 3245-3249. Vavrek, R.J., & Stewart, J.M. (1985). Competitive antagonists of bradykinin. Peptides 6, 161-164. Wachtfogel, Y.T., Cadena, R.A.D., & Colman, R.W. (1993). Structural biology, cellular interactions and pathophysiology of the contact system. Thrombosis Research 72, 1-21. Whalley, E.T., Loy, S.D., Modiferri, D., Blodgett, J.K., & Cheronis, J.C. (1992). A novel potent bis(succinimido)hexane peptide heterodimer antagonist (CP-0364) of bradykinin BK 1 and BK 2 receptors: in vitro and in vivo studies. Br. J. Pharmac. 107, 257.

384

D O N A L D H. MILLER and HARRY S. MARGOLIUS

Wilson, D.D., Garavilla, L.d., Kuhn, W., Togo, J., Burch, R.M., & Steranka, L.R. (1989). D-Arg-[Hyp3-D-Phe7]-bradykinin, a bradykinin antagonist, reduces mortality in a rat model of endotoxic shock. Circ. Shock 27, 93-101. Wirth, K., Hock, F.J., Albus, U., Linz, W., Alpermann, H.G., Anagnostopoulos, H., Henke, S., Breipohl, G., K6nig, W., Knolle, J., & Sch/31kens, B.A. (1991). Hoe 140 a new potent and long acting bradykinin-antagonist: in vivo studies. Br. J. Pharmacol. 102, 774-777.

RECOMMENDED READING Bhoola, K.D., Figueroa, C.D., & Worthy, K. (1992). Bioregulation of kinins: kallikreins, kininogens, and kininases. Pharmacol. Revs. 44, 1-80.