Protein Degradation in Health and Disease

Ciba Foundation Symposium 75 (new series)

1980

Excerpta Medica Amsterdam.Oxford .New York

Contents

A L A N J . B A R R E T T I n t r oduc t i on : the classification o f proteinases 1 Discussion 9

H . K1RSCHKE, J . L A N G N E R , S. S I E M A N N , B . W I E D E R A N D E R S , S. A N S O R G E and

p. B O H L E Y Lysosomal cysteine proteinases 15 Discussion 31

A L A N J . B A R R E T T Cathepsin D : the lysosomal aspartic proteinase 37 Discussion 43

J A M E S T R A V I S , P . J . G I L E S , L . P O R C E L L I , C . F . R E I L L Y , R . B A U G H and J . P O W E R S

H u m a n leucocyte elastase and cathepsin G: structural and funct ional characteristics 51 Discussion 66

D A V I D E . W O O L L E Y H u m a n collagenases: comparative and immunolocal iza-t ion studies 69 Discussion 82

D A V I D J . E T H E R I N G T O N Proteinases in connective tissue breakdown 87 Discussion 100

M A R C O B A G G I O L I N I , J Ö R G S C H N Y D E R , U R S U L A B R E T Z , B E A T R I C E D E W A L D and

W A L T E R R U C H Cel lular mechanisms o f proteinase release f r o m in f l am­matory cells and the degradation o f extracellular proteins 105 Discussion 118

G . L . F R A N C I S , S . E . K N O W L E S and F . J . B A L L A R D Inact ivat ion o f cytosol enzymes by a liver membrane pro te in 123 Discussion 134

R O G E R τ . D E A N Lysosomes and prote in degradation 139 Discussion 145

ν

V I CO Ν "Γ Li Ν TS

J . B . L L O Y D Insights in to mechanisms o f intracel lular prote in turnover f r om studies on pinocytosis 151 Discussion 160

S A N T I A G O G R I S O L I A , E R W I N K N E C H T , J O S E H E R N A N D E Z - Y A G O and R U T H

W A L L A C E Turnover and degradation o f m i tochondr ia and their proteins 167 Discussion 185

B R I A N P O O L E , S H O J I O H K U M A and M I C H A E L W A R B U R T O N Prote in degra­dat ion in cells i n culture 189 Discussion 200

M . F . H O P G O O D and F . J . B A L L A R D Regulat ion o f pro te in breakdown in hepa-tocyte monolayers 205 Discussion 215

J O H N K A Y A possible role for neutral proteolysis in the degradation o f in t ra ­cellular proteins 219 Discussion 224

A L F R E D L . G O L D B E R G , N I N A P . S T R N A D and K . H . S R E E D H A R A S W A M Y Studies o f the A T P dependence o f prote in degradation in cells and cell extracts 227 Discussion 247

R. J O H N M A Y E R , S U S A N M . R U S S E L L , R O W L A N D J . B U R G E S S , C O L I N J . W I L D E and

N O R M A N P A S K I N Coord ina t i on o f prote in synthesis and degradation 253 Discussion 269

General discussion Sites o f prote in degradation 273

G L E N N E . M O R T I M O R E and C H A R L E S M . S C H W O R E R App l i ca t i on o f liver per­fusion as an in vitro model i n studies o f intracel lular prote in degradation 281 Discussion 298

D . J . M I L L W A R D , P . C . B A T E S , J . G . B R O W N , S . R . R O S O C H A C K I and M . J . RENN1E

Prote in degradation and the regulat ion o f prote in balance i n muscle 307 Discussion 323

CONTLNTS V I I

J . F R E D D I C E and C A R L O S D . W A L K E R Prote in degradation in metabolic and nu t r i t i ona l disorders 331 Discussion 345

H A N S F R I T Z Proteinase inh ib i tors i n severe in f lammatory processes (septic shock and experimental endotoxaemia) : biochemical, pathophysiological and therapeutic aspects 351 Discussion 375

J . - D . V A S S A L L I , A . G R A N E L L I - P I P E R N O and E . R E I C H Neutra l proteinases o f leucocytes and the i n f l ammato ry process 381 Discussion 392

Final general discussion 397

A L A N J . B A R R E T T Conc lud ing remarks 403

Index o f contr ibutors 407

Subject index 409

Proteinase inhibitors in severe inflammatory processes (septic shock and experimental endotoxaemia): biochemical, pathophysiological and therapeutic aspects

H A N S F R I T Z

Abteilung für Klinische Chemie und Klinische Biochemie in der Chirurgischen Klinik der Universität München, München

With contributions from J . W I T T E and Μ. J O C H U M (Chirurgische Klinik der Universität München, München), H . SCHIESSLER, U. SEEMÜLLER, S. KUPFER and G. SAMS-BORHAN

(Abteilung für Klinische Chemie und Klinische Biochemie in der Chirurgischen Klinik der Universität München, München), E. W Ä C H T E R (Institut für Physiologische Chemie, Physika­lische Biochemie und Zellbiologie der Universität München, München), W. S C H R A M M (Medizinische Klinik Innenstadt der Universität München, München), M . E U L I T Z (Institut für Hämatologie der Gesellschaft für Strahlen- und Umweltforschung, Abt. Immunologie, München), R . SCHERER (Max-Planck-Institut für Biochemie, Abteilung für Experimentelle Medizin, Martinsried bei München) and G. RUCKDESCHEL (Max-v.-Pettenkofer-Institut für Hygiene und Medizinische Mikrobiologie, München)

Abstract Plasma levels o f a n t i t h r o m b i n I I I , «2-macrog lobul in and inter -a-t ryps in i n h i b i t o r , as wel l as those o f various c l o t t ing , complement and other plasma factors, were s igni f i cant ly decreased in 18 patients suf fer ing f r o m hyperdynamic septic shock. A s imi lar statistically signif icant reduct ion o f the concentrat ions o f several plasma factors ( p r o t h r o m b i n and a n t i t h r o m b i n I I I , p lasminogen and «2-plasmin i n h i b i t o r , complement factor C3 and c l o t t ing factor X I I I ) was observed in exper imental endotoxaemia. I n this model the reduct ion in the plasma levels o f these factors was considerably d imin ished by the intravenous in ject ion o f a granulocyt ic elastase-cathepsin G i n h i b i t o r o f lower molecular weight f r o m soybeans. The results o f bo th studies indicate that consumpt ion o f plasma factors in the course o f Gram-negat ive sepsis proceeds not on ly via the classical routes (by ac t i va t ion o f the c lo t t ing , f i b r ino l y t i c and complement cascades by system-specific proteinases such as thrombok inase or the plasminogen act ivator ) bu t also to an appreciable degree by unspecific degradat ion o f plasma factors by neutra l proteinases such as elastase and cathepsin G . The endotox in-induced release o f b o t h sorts o f proteinases, the system-specific ones and the unspecif ic lysosomal proteinases f r o m leucocytes and other cells, is l ike ly to be main ly responsible for the c onsumpt i on o f a n t i t h r o m b i n I I I and c*2-macroglobulin via complex f o r m a t i o n ( f o l l owed by e l im ina t i on o f the complexes) and the increased turnover o f the in t e r -a - t r yps in i nh ib i t o r as observed in the c l in ica l study.

The therapeutic use o f an exogenous elastase-cathepsin G i n h i b i t o r in the exper imental mode l was s t imula ted by the observat ion that h u m a n mucous secretions conta in an acid-stable i n h i b i t o r o f the neutra l granulocyt ic proteinases, called H U S I - I o r ant i leucoprote inase. Th is i nh ib i t o r protects mucous membranes

351 © Excerpta Medica 1980 Protein degradation in health and disease (Ciba Foundation Symposium 75) ρ 351-379

352 Η. 1 RITZ ET A I .

and soluble proteins against proteo ly t ic attack by leucocytic proteinases released in the course o f a local i n f l ammato ry response. P re l iminary results indicate that HUS1- I , wh i ch is produced by the epithel ia l cells o f mucous membranes, does not belong to any k n o w n st ructura l type o f acid-stable proteinase i nh ib i t o r . The search for other candidates suitable for medicat ion in humans led to the discovery o f a potent elastase-cathepsin G inh ib i t o r , called eg l in, in the leech Hirudo medicinalis. This acid-stable i n h i b i t o r w i t h a molecular weight close to 8100 has an unusual s t ructura l proper ty in that the structure o f the molecule is not stabil ized by any d isulphide bridge.

The course o f a disease l ike septicaemia or septic shock is o f ten complicated by severe pathobiochemical processes in the c i rcu la t ion . These are, for example, disseminated intravascular coagulat ion (D IC ) caused by act ivat ion o f the c lo t t ing and f ibr ino ly t i c cascades, and anaphylactic responses induced by act ivat ion o f the complement system (Hami l t on et al 1978, Müller-Berghaus et al 1976, Garner et al 1974, McCabe 1973). These act ivat ion reactions may be triggered by endotoxins — that is, l ipopolysaccharides f r o m Gram-negative bacteria (Jeljaszewicz & Waldström 1978, Urbaschek et al 1975). Endotox ins can damage biological membranes and thus induce the release o f constituents, so-called mediators o f i n f l ammat i on , inc lud ing lysosomal enzymes, f r om various body cells (Weissmann 1974, Urbaschek et al 1975, M y r v o l d 1976, Mova t 1979). These enzymes normal ly exhibit their physiological func t ion , namely degrading phagocytosed mater ia l , inside the cell (K lebanof f & Clark 1978). I f released into the c i rcu la t ion , however, they may enhance the in f l ammatory response by several routes.

System-specific proteinases such as the plasminogen activator and thrombokinase activate the b lood systems (see Fig. 1) by proenzyme — enzyme conversion — that is, by specific proteolyt ic cleavages. These enzymes are responsible, therefore, for the 'classical' or specific consumpt ion o f factors o f these systems, inc luding the inhib i tors o f c lo t t ing ( A T I I I ) , ka l l ikre in ( C I I N A ) , f ibrinolysis (azPI) and complement ( C I I N A ) factors (cf. Fig. 2 and Table 1). E l im ina t i on o f both the proteinases and inhib i tors o f the b lood systems also proceeds specifically by the f o rmat ion o f enzyme- inh ib i t o r complexes that are phagocytosed by cells o f the ret iculoendothel ial system (RES) (Ohlsson 1974, Ohlsson & Laure l l 1976, Ohlsson 1978).

More recently it became evident f r om studies in vitro and in vivo that endotoxin- induced consumpt ion o f plasma proteins might also be due to a significant degree to unspecific degradation by leucocytic proteinases, especially an elastase (cf. Figs. 1 and 2) (Schmidt et al 1974, Haschen 1975, Egbr ing et al 1977, Egbr ing & Havemann 1978, Schiessler et al 1978a, Aasen

PROTEINASE INHIBITORS IN I N F L A M M A T I O N 353

Proteinases released from cells can cause:

Activation 1

Blood systems such as: clotting, fibrinolysis, complement,

kallikrein 1

S p e c i f i c

Degradation I

Blood system factors, other blood proteins, proteinase

inhibitors I

U n s p e c i f i c

c o n s u m p t i o n

F I G . 1. Effects of proteinases released from blood and tissue cells during inflammation on blood systems and plasma factors.

& Ohlsson 1978). I n such unspecific consumpt ion reactions — wh ich , i n contrast to the specific consumpt ion , are not l imited to the factors o f the b l o o d systems (cf. F ig . 1) — the biological act iv i ty o f the plasma proteins is irreversibly destroyed by proteo lyt ic degradation. The l iberated leucocytic proteinases are e l iminated, however, i n a specific manner by the f o rmat i on o f complexes w i th proteinase inh ib i to rs such as c ^ M , ct\A and cnAC (see Fig . 2 and Table 2) and the phagocytosis o f the complexes by the RES (Ohlsson & Laure l l 1976, Debanne et al 1976).

Consumption reactions

Specific activation Unspecific degradation

e.g. thrombokinase, plasminogen activator

e.g. PMN elastase, cathepsins G, B, D

1 Blood systems

I Blood proteins

e.g. AT II Plasma proteinase inhibitors 1 cx2P\, C1 INa e.g. αλ A, a i x AC, a 2M

1 ion (RES)

F I G . 2. Consumption of plasma factors during inflammation. System-specific proteinases such as thrombokinase activate the blood system factors which are inhibited thereafter by complex formation with antithrombin I I I (AT I I I ) , a2-plasmin inhibitor (a2PI) and CI inactivator (CI INA) , respectively. Proteinases liberated from blood and tissue cells like granulocytic (PMN) elastase can degrade plasma factors unspecifically before being inhibited by complex formation with cYj-antitrypsin (αϊA) , aj-antichymotrypsin (ajAC) and c*2-macroglobulin (c*2M), respecti­vely. The enzyme-inhibitor complexes are eliminated from the circulation by cells of the reticu­loendothelial system (RES).

354 Η. FRITZ ET ΑΙ .

TABLE 1

Plasma proteinase inhibitors of blood system factors (cf. Fig. 1). Theoretically highest possible plasma concentrations of these factors are given in parentheses (in /xmol/1). For literature references see Heimburger 1975, Harpel & Rosenberg 1975, Collen et al 1979

Inhibitor Abbreviation Concentration frmol/i)

Factors primarily inhibited in vivo

Antithrombin I I I AT I I I 4.0 Thrombin ( l ) a , factor Xa a

+ heparin + plasmin 6 , + kal l ikrein 6

(*2-Plasmin inhibitor c*2PI 0.9 Plasmin/ogen (2)

C I inactivator C I INA 2.4 CTs, CTr ( <1 ) , kallikrein (~ 1)

aReacting with A T I I I slowly without heparin but rapidly with heparin. b Only in the presence of heparin after consumption of c^PI or C I INA (Highsmith & Rosenberg 1974, Venneröd et al 1976), respectively.

The factual basis o f this concept is that the early phase o f experimental

septic shock is characterized by the release o f cell constituents inc lud ing

proteinases f r o m leucocytes, platelets, macrophages, mast cells and

endothelial cells o f the vessel walls, especially those o f the lung capillaries

(Weissmann 1974, Urbaschek et al 1975, M y r v o l d 1976, Starkey 1977, V a n e &

Ferreira 1978, Aasen & Ohlsson 1978, Mova t 1979). Because o f the presence

TABLE 2

Plasma inhibitors directed primarily against lysosomal proteinases of various body cells. For literature references see Heimburger 1975, Starkey & Barrett 1977 teM), Travis et al 1978 (a\A, a iAC ) , Ohlsson 1978 (crjA, a j A Q , Woolley et al 1976 (jojCI)

inhibitor Abbreviation Concentration Strong inhibition of (in vitro, (\kmol/l) probably also in vivo):

o!2-Macroglobulin cx2M 3.6 Neutral and acidic proteinases3

(complex elimination ty, ~ 10 min)

oc\-Antitrypsin (α-1-Pro­tease inhibitor)

αιΑ 52 Neutral proteinases from leuco­cytes6, pancreas and other tissues

α ι - Antichymotrypsin a jAC 6.4 Chymotrypsin, cathepsin G

β ι-Collagenase inhibitor ßiCl ~0 .4 C True collagenases (metallo enzymes)

Inter-a-trypsin inhibi tor d

I T I 2.8 Trypsin, chymotrypsin, acrosin

ae.g. granulocyte neutral proteinase (formerly collagenase, Ohlsson 1980), cathepsin G, elastases, collagenases, trypsin, chymotrypsin, plasma kallikrein; cathepsin Β and D, etc. be.g. elastase and neutral proteinase (cf. a ) . cEstimated value according to Woolley et al 1976. d The biological antagonists are not yet known, see text.

PROTEINASE INHIBITORS IN I N F L A M M A T I O N 355

o f potent proteinase inh ib i t o rs , however, direct measurement o f proteinase activities i n plasma or serum is not feasible. We have focused our ef forts, therefore, on indirect indicat ions such as the consumpt ion o f plasma proteinase inh ib i t o r s , the turnover o f selected plasma proteins, and the concentrations o f plasma factors expected to be either signif icantly elevated or d iminished dur ing septicaemia or septic shock. I n another approach we tested the therapeutic effect o f an exogenous inh ib i to r o f neutral leucocytic proteinases on the plasma levels o f various plasma proteins inc luding proteinase inh ib i to rs dur ing experimental endotoxaemia. The results obtained st imulated us to cont inue the search for potent inh ib i tors o f leucocytic proteinases as possible therapeutic agents preventing unspecific consumpt ion o f plasma factors dur ing septicaemia or septic shock.

HYPERDYNAMIC SEPTIC SHOCK: A CONTROLLED C L I N I C A L TR IAL

Criteria and methods

Criteria, The patients suf fer ing f r om hyperdynamic septic shock (n= 18, age 17-70 years) had to f u l f i l a l l o f the f o l l ow ing cr i ter ia: body temperature >38.5 °C, leucocytes > 1 5 000 or < 5 000/mm 3 , platelets < 1 3 0 000/mm 3 , positive b lood culture (twice) , positive evidence o f endotox in in serum (at least twice), cardiac index > 6 1/min/m 2 body surface, to ta l peripheral resistance o f the vessel system < 6 0 0 dyn χ s/cm 5 , mean arterial b lood pressure 80 -90 m m H g (10.7-12.0 kPa) , a l l c l in ical signs o f septic shock (Wi t te 1979).

Therapy. The f o l l ow ing therapeutic measures were taken: surgical cleansing o f the septic focus, vo lume subst i tut ion w i th human a lbumin under contro l o f the pu lmonary capi l lary wedge pressure, cont inuous low dose heparin therapy (200-400 U/h) , the infusion o f 100-400 ^g/min dopamine to mainta in an urine ou tput o f at least 80 -100 m l / h , adminis t ra t ion o f 30 mg/kg body weight o f methylprednisolone every 6 h up to 48 h (to prevent acute respiratory distress syndrome), medicat ion w i th antibiot ics according to the sensitivity patterns o f the microorganisms, and parenteral nu t r i t i on as well as mechanical vent i latory assistance (Wi t te 1979).

Methods. A l l parameters were fo l lowed up for four days after the patients ful f i l l ed al l the above cr i ter ia o f hyperdynamic septic shock. Haemodynamic measurements were made and b lood samples were taken at the beginning and 6, 12, 18, 24, 36, 48, 72 and 96 h thereafter. Further experimental details are given in the academic thesis o f J . Wi t t e (1979) and w i l l be published

356 Η. FRITZ ET A L

elsewhere. O f the various haematological , haemodynamic and biochemical data fo l lowed up dur ing the observation per iod, only those relevant to the topic o f this paper are reported here.

Activation of blood systems and concentration pattern of plasma factors

Acute-phase proteins. In f l ammatory processes in the organism are accompanied by increased serum or plasma levels o f the acute-phase proteins. Indeed, the serum concentrations o f C-reactive protein and ai-glycoprotein were signif icantly elevated dur ing the septic shock phase (Table 3). Probably because o f compet i t ion w i t h consumpt ion (cf. below), the mean plasma level o f fibrinogen, another acute-phase pro te in , was i n the upper part o f the standard range rather than above i t (Table 3). The high p roduc t i on rate o f these acute-phase proteins strongly indicates that the synthetic capacity o f the organism, especially o f the l iver, for the various plasma proteins was not yet restricted dur ing the hyperdynamic septic shock phase.

Clotting and fibrinolysis. Permanent act ivat ion o f the c lo t t ing cascade by system-specific proteinases throughout the observation per iod is indicated by the highly signi f icantly increased plasma concentrat ion o f fibrinopeptide A (Table 3), which is cleaved f r om f ibr inogen by the specific act ion o f t h r o m b i n (Gaffney 1977, Schramm et al 1980). S imi lar ly , act ivat ion o f plasmin and thus o f f ibrinolysis is responsible for the elevated plasma levels o f fibrin(ogen) split products (FSP, cf. Table 3) (Skansberg et al 1974). I n this case, however, a tendency to normal i za t ion was clearly visible towards the end o f the observation per iod.

Complement. Complement act ivat ion dur ing the shock phase proceeded pr imar i l y via the alternative pathway. This may be deduced f r om the clearly lower plasma level o f complement factor C3 than o f C4 (Table 3). This result is in agreement w i th our present knowledge o f the act ivat ion mechanism o f the complement cascade in endotoxaemia (Johnson et al 1976, Spragg & Austen 1977, Nicholson et al 1978, T u r k & Wi l l oughby 1978, Aasen et al 1980). O f special interest in this respect is the observation that the granulocyt ic elastase is capable o f cleaving C3 in such a way that direct act ivat ion o f the complement cascade via the alternative route could indeed result (Johnson et al 1976, Aasen et al 1980). On the other hand, there is g rowing evidence that endotoxins can trigger the release o f elastase f r om granulocytes (Aasen & Ohlsson 1978, Egbr ing et al 1977).

Liberated elastase could also be responsible, at least par t ly , for the highly

PROTEINASE INHIBITORS IN I N F L A M M A T I O N 357

TABLE 3

Plasma or serum levels of various biochemical parameters in septic shock. Standard values (norm) are given both as mean value and range (in parentheses). The values given at the beginning of the shock phase (0 h) and after 96 h are mean values x{n = 18) with the corresponding standard deviations ( ± SEM). Ρ, statistical significance (Student r-test)

Measured parameter Norm Oh 96 h Ρ

C-reactive protein (mg/dl) <1.2 13.5 ± 1.3 10.4 + 1.6 <0 .001 a

ai-Glycoprotein (mg/dl) 90(55-140) 139.3 ± 7.9 b 132.9 ± 15.2 b <0.001 a

Fibrinogen (mg/dl) 180-380 359.7 ± 23.4 293.5 ± 17.4 >0.05 a

Fibrinopeptide A (ng/ml) <3.0 13.1 ± 2.7 18.1 ± 4.5 <0.001 a

Factor X I I I (mg/dl) - 2 46.1 ± 4.9 b 52.9 ± 4.8 b <0.001 a

FSP'c fog/ml) < 10.0 21.4 ± 4.1 10.7 ± 1.7 <0.05 d

Complement C3 (mg/dl) 82(55-120) 70.1 ± 4.3 b 65.3 ± 6.7 b < 0 . 0 1 a

Complement C4 (mg/dl) 30(20-50) 82.2 ± 11.2b 74.5 ± 11.2 b >0.05 a

a For both 0 and 96 h. bExpressed as °7o o f the norm. cFibrin(ogen) split products. d For the 0 h value only.

signif icant (P <0 .001) consumpt ion o f factor XIII (45—55% reduct ion in the plasma concentrat ion) th roughout the shock phase. This view is supported by the results o f an imal studies showing that therapeutic admin is t ra t ion o f an elastase-cathepsin G inh ib i t o r in experimental septicaemia can prevent factor X I I I consumpt ion to a large degree (Schiessler et al 1978a).

Plasma proteinase inhibitors

Marked consumpt ion or turnover o f plasma proteinase inhib i tors dur ing the septic shock phase reflects extensive act ivat ion o f the b lood systems as well as the l iberat ion o f considerable amounts o f proteinases f r om various body cells, especially the granulocytes, in our patients.

Antithrombin I I I ( A T I I I ) is the most impor tant inh ib i t o r for mainta in ing the homeostasis o f the c lo t t ing system. To fu l f i l its physiological or pathophysiological func t i on , namely regulat ion o f the act iv i ty o f factor Xa and th r omb in to prevent thrombosis , A T I I I levels w i th in or close to the standard range are necessary; the risk o f thrombosis increases considerably w i th decreasing A T I I I plasma levels (Harpe l & Rosenberg 1975, Seegers 1978, Collen et al 1979).

In our patients the A T I I I concentrat ion was already reduced to 5 0 % o f the standard mean value at the beginning o f the observation period (Fig. 3). This indicates major consumpt ion o f A T I I I by the f o rmat i on o f complexes w i th the activated c lot t ing factors. W i t h the heparin medicat ion necessary to prevent thrombosis or disseminated intravascular coagulat ion, the consumption o f A T I I I is even further intensif ied, since the hepar in-AT I I I

358 Η. FRITZ ET A L

tim« [ h l - —

F I G . 3. Plasma levels of antithrombin I I I during hyperdynamic septic shock. The curves show mean values χ of all patients (n - 18, ), of the patients surviving the shock phase (n = 15,

), and of patients who died close to the end of the observation period (n = 3, ); standard deviations ( ± SEM) of the mean values χ are indicated for each test point. Ordinate, the percentage related to the standard level (estimated with pooled plasma from normal individuals). Abscissa, observation period. Antithrombin I I I concentration was estimated by LaurelPs electroimmunoassay (rocket technique). For further details see text.

complex also reacts w i t h other proteinases such as factor V i l a , I X a , X I a , X l l a and plasma ka l l i k re in (Col len et al 1979). Most remarkably , i n the patients surviv ing the shock phase ( f l =15 ) the A T I I I level increased cont inuously whereas it decreased dramatical ly i n those patients who died (n = 3) dur ing the shock phase (close to the end o f the observation per iod, cf. F ig . 3).

ai-Macroglobulin. The permanently low level o f a2-macroglobul in (aiM) throughout the septic shock phase is especially s t r ik ing (Fig. 4). Obv ious ly , considerable amounts o f neutral and acidic proteinases (cf. Table 2) are cont inuously l iberated in to the c i rculat ion f r om the various body cells, leading to major consumpt ion o f this most impor tan t proteinase-el iminating vehicle (Col len et al 1979, Ohlsson 1978, Starkey & Barrett 1977, Ohlsson & Laure l l 1976, Harpe l & Rosenberg 1975).

Considering the proteolyt ic potent ia l which might be l iberated by endotox in s t imulat ion f r om polymorphonuclear ( P M N ) leucocytes alone, for example approximate ly 1 g o f bo th elastase and neutral proteinase ( f o rmer ly collagenase) and 0.3 g o f cathepsin G per day (Laure l l 1975, Ohlsson 1980), the protective role o f aiM against unspecific proteolysis cannot be suf f ic ient ly emphasized. I n fact, onM seems to be chiefly responsible for the i n h i b i t i o n (cathepsin G, neutral proteinase) and e l iminat ion ( inc luding elastase, wh i ch is p r imar i l y inhib i ted by αι-antitrypsin but probably transferred to a2M for this purpose) o f these granulocyt ic proteinases (Ohlsson 1978, Ohlsson & Laure l l 1976).

PROTEINASE INHIBITORS IN I N F L A M M A T I O N 359

η = 16

ι 50-ΙΤ1160000 (normal 50-60)

η=10

Ε 3 30-Ε ΙΤΙ 30000 (normal 6 - 9 ) ^ ι ^ η = 1 ο

10J

0 24 48 72 96

timt [h]

F I G . 4. Plasma levels of a2-macroglobuiin (c^M) and serum levels of native inter-a-trypsin inhi­bitor (ITI160000) and acid-stable ITIiöoooo-derived ITI30000 during hyperdynamic septic shock. The curves represent mean values χ o f Λ = 18 teM) or η = 10 (ITI) patients; standard deviations ( ± SEM) of the mean values χ are indicated for each test point. Ordinate, the percentage of the standard level o f c*2M (estimated with pooled plasma from normal individuals) and the trypsin inhibitory activity o f I T I (substrate BzArgNHNp) per ml serum. Abscissa, observation period. C Q M was estimated by Mancini's radial immunodiffusion technique. I T I 1 6 0 000 and I T I 3 0 000 were determined according to Hochstrasser et al (1977a) by measuring the trypsin inhibitory activity of the supernatant of an acidified serum sample (ITI30000) and of the trypsin-treated precipitate after trypsin denaturation (ITIieoooo).

Inter-a-trypsin inhibitor. Recent observations indicate increased turnover o f the inter-a- tryps in inh ib i t o r ( I T I ) dur ing septicaemia (Hochstrasser et al 1977b). Indeed, the level o f native I T I ( I T I 1 6 0 0 0 0 ) was signif icantly reduced throughout the observation per iod whereas the concentrat ion o f the I T I -derived acid-stable i nh ib i t o r ( I T I 3 0 0 0 0 ) was signif icantly increased (Fig. 4). Studies in vitro showed that o f the various proteinases tested, granulocyt ic elastase (and cathepsin G ; K. Hochstrasser, personal communicat ion ) are capable o f l iberat ing I T I 3 0 0 0 0 most rap id ly f r o m ITI160000 (D ied et al 1979). This might mean that granulocyt ic proteinases are also responsible for the high turnover o f I T I i n septicaemia or septic shock.

Though the bio logical funct ion o f ITI160000 or ITI30000 is not yet clear,

ident i f icat ion o f the inh ib i t o r y active domain(s) as Kuni t z - or aprot in in-type inhib i tors is remarkable (Wächter & Hochstrasser 1979).

360 Η. FRITZ KT A I

Conclusions and outlook

The fo l l ow ing aspects o f the cl inical study should be especially emphasized. (i) The plasma or serum levels o f the measured parameters had already

reached pathological values when the first c l inical signs o f septic shock were established. This means that the search for other biochemical parameters suitable for ident i fy ing the onset o f septicaemia or septic shock has to be cont inued.

(ii) Consumpt ion o f plasma factors i n hyperdynamic septic shock and probably also in Gram-negative septicaemia is due to the proteolyt ic attack o f bo th system-specific and unspecific proteinases l iberated f r om leucocytes and other body cells. The large consumpt ion o f the protective potent ia l o f the plasma inhib i tors by the f o rmat i on o f complexes w i th these proteinases, fo l lowed by rapid e l iminat ion o f the complexes, is especially s t r ik ing .

( i i i ) The possibi l i ty o f reducing the consumpt ion o f plasma factors by administer ing suitable proteinase inh ib i tors should be envisaged. I n our op in ion such inhib i tors should p r imar i l y prevent the consumpt ion o f c ^ M , A T I I I and «2-plasmin inh ib i t o r (a2PI). App l i ca t i on at an early phase o f septicaemia seems to be most promis ing . Whereas a potent plasmin inh ib i t o r ( apro t in in , the effective agent in the drugs Trasylcl®, Antagosan® or Iniprol®) is already available for medical use, in the other two cases on ly subst i tut ion w i th A T I I I (Schramm 1977) or aiM concentrates is possible at present.

T H E A C I D - S T A B L E E L A S T A S E - C A T H E P S I N G I N H I B I T O R O F H U M A N M U C O U S

S E C R E T I O N S

Inhibitory properties

Our search for inhib i tors suitable for therapeutic use in humans was st imulated by the f ind ing that the acid-stable t ryps in-chymotryps in inh ib i t o r f r om human seminal plasma (HUSI - I ) also has strong a f f in i ty for granulocyt ic elastase and cathepsin G (Schiessler et al 1976a). For the corresponding enzyme- inh ib i to r complexes, K\ values close to 1 χ 1 0 " 9

moi/1 (human granulocyt ic elastase, bovine chymotryps in and trypsin) and 5 x 10~ 8 mol/1 (human granulocyt ic cathepsin G) were f ound .

Biochemical and structural features

H U S I - I isolated in our laboratory f r o m human seminal plasma consists o f

PROTEINASE INHIBITORS IN I N F L A M M A T I O N 361

several mul t ip le forms having the same inh ib i to ry characteristics and similar molecular weights o f approximate ly 11 000 (Schiessler et al 1976b). The heterogeneity o f H U S I - I preparations seems to be due part ly to proteolyt ic degradation in seminal plasma but also to a l l e lomorphism. This po l ymorph i sm complicates bo th the pur i f i ca t i on o f H U S I - I to homogeneity and the determinat ion o f the amino acid sequence. Basically, the H U S I - I molecule consists o f a single polypept ide chain o f about 100 amino acid residues which is cross-linked by six disulphide bridges. Pre l iminary results give the amino acid sequence shown below for positions 1-24 and 54-99 o f H U S I - I f o r m D (cf. Schiessler et al 1978b). I n these sequenced parts o f the H U S I - I molecule a s tructura l homology to either a Kazal-type or Kunitz- type inh ib i t o r cannot be recognized. This is surpris ing, in that inhib i tors o f similar molecular size and inh ib i t o ry properties f ound so far in animals and men are bui l t up o f either Kazal- or Kuni tz- type domains (cf. below).

1 10 Tyr -Va l -Asn-Thr-Pro-Asn-Pro-Arg-Asp-Arg-

20

Lys-Pro-Gly-Lys-Cys-Pro-Val-Thr-Tyr-Gly-

Gln-Cys-Leu-Met- and

60 -Met-Leu-Asn-Pro-Pro-Asn-Phe-

70 Cys-Glu-Met-Asp-Gly-Gln-Cys-Lys-Arg-Asp-

80 Leu-Lys-Cys-Ser-Met-Gly-Met-Cys-Gly-Lys-

90

Ser-Lys-Val-Glu-Pro-Val-Ala-Cys,Asp,Arg,

Gly ,Pro,Lys,Lys-Cys-Thr-Thr-Gly-Ser

The occurrence o f a new structural inh ib i t o r type, perhaps exclusively in humans, wou ld be especially interesting in relat ion to the evolut ion o f prote in proteinase inh ib i to rs . I n add i t i on , this wou ld be an indicat ion that H U S I - I had developed p r imar i l y for the inh ib i t i on o f granulocytic elastase and cathepsin G. This assumption is supported by considering the funct ional possibilities o f H U S I - I i n re lat ion to its d i s t r ibut i on w i th in the organism.

362 Η. F R I T Z E T A I .

Distribution, site of production and biological function

Inh ib i tors w i th HUSI - I - l i ke biochemical and/or immunolog ica l properties were f ound also in cervical mucus, bronchia l f l u i d , nasal secretion and tears (Schiessler 1976, Schiessler et al 1977a, 1978b). Clearly, H U S I - I occurs in human mucous f luids open to invading organisms and thus of ten exposed to large numbers o f leucocytes dur ing local in f l ammatory processes. The concentrat ion o f H U S I - I in these mucous f luids is normal l y far higher (up to 10 times) than the levels o f the transudated plasma inh ib i tors απ A and απ A C . Whereas the IT I -der ived acid-stable inh ib i t o r ITI30000 (cf. above) was also f ound in bronchia l fluid (K. Hochstrasser, personal communica t i on 1979), aiM is not usually present in human mucous secretions (Table 4). The major difference between H U S I - I and the transudated plasma inh ib i tors concerns its site o f p roduc t i on : H U S I - I is synthesized locally by the epithel ial cells o f the corresponding organs and is thus direct ly secreted in to the mucous fluids (Schiessler et al 1978b, Tegner & Ohlsson 1977, Schil l et al 1978).

F r om the local p roduc t i on , the relatively h igh level in mucous f luids, and the inh ib i t o ry spectrum o f H U S I - I (cf. Table 5), i t seems very l ike ly to us that H U S I - I has an impor tant defensive func t ion . I t presumably protects soluble proteins (such as immunog lobul ins ) and the mucous membranes against degradation by leucocytic proteinases l iberated dur ing local in f l ammat i on and thus helps the organism to prevent the in f l ammatory response f r o m intensi fy ing. Complexes o f H U S I - I w i t h leucocytic and/or bacter ial proteinases could indeed be demonstrated i n mucous fluids du r ing

T A B L E 4

Inhibitors in human mucous secretions

Inhibitor Mol. wt. (approx.)

Inhibition of leucocytic: Inhibitor Mol. wt. (approx.)

Elastase Cathepsin G Neutral proteinase0

Of plasma origin α ιΑΤ 50 000 + + ( + ) + onAC 70 000 - + -I T I 30 000 30 000 - - ? ( a 2 M ) b 725 000 + + + + +

Locally produced HUSI- I 11 000 + + + —

+ + , very strong; + , strong; ( + ), weak; - , no inhibit ion. aFormerly collagenase (Ohlsson 1980). b Only occasionally found during severe inflammation.

P R O T E I N A S E I N H I B I T O R S I N I N F L A M M A T I O N 363

T A B L E 5

Acid-stable inhibitors of granulocytic elastase and cathepsin G from various sources. For literature references see text

Inhibition of: Inhibitor Source Mol. wt.

(approx.) Elastase0 Cathepsin G° Chymo­trypsin0

Trypsin0

HUSI-I Human mucous 11 000 + + + + + + + secretions

DSI Dog submandibular 13 000 + + + + + + + glands

A A Soybeans 8000 + + + + + 4- + LB I Lima beans 9000 + + + + + + + Elastatinal Actinomycetes 500 ( + ) c - - -Elasnin Streptomyces 400 + d ( + ) ( + ) Chymostatin Actinomycetes 600 (- ) + + -Aprotinin Bovine organs 6500 ( + ) ( + ) + + +

0 not detected; + + , very strong; + , strong; ( + ), weak; ( - ) , very weak; - , no inhibit ion. a Human. bBovine. c Pancreatic: + dPancreatic:(+ )

i n f l ammat i on ; even free proteolyt ic act iv i ty may appear when the inh ib i to rs are to ta l ly consumed (Schiessler et al 1978b, Ohlsson 1978, K r u m m e et a l 1977). Clearly, H U S I - I is a funct iona l substitute for α-2-niacroglobulin, normal ly inaccessible to mucous fluids. Because o f the l imi ted amount o f H U S I - I available f r o m natura l sources — i t has so far been f ound only i n humans — we cannot envisage its therapeutic use for some time yet.

INHIBITORS FROM OTHER SOURCES AS POTENTIAL THERAPEUT iC AGENTS

Inhib i tors o f granulocyt ic elastase and/or cathepsin G were f ound in

various natura l sources; some o f them are listed in Table 5.

inhibitor from dog submandibular glands

The dog submandibular inh ib i t o r (DSI ) is present in these glands in exceptionally high concentrat ion. I t is a secretory prote in consisting o f a single polypeptide chain cross-l inked by six disulphide bridges (Fr i tz et al 1971). The D S I molecule is composed o f two independent inh ib i t o ry active domains (Fig. 5) bo th o f which are structural ly homologous to the so-called Kazal-type inh ib i tors (Hochstrasser et al 1975) — the pancreatic secretory tryps in inh ib i t o r (PSTI ) , the seminal acrosin inh ib i t o r (Tschesche et al 1976) and the ovomucoids and ovo inh ib i tors f r o m egg white (Laskowski Jr et al

364 Η. F R I T Z E T A L

F I G . 5. Covalent structure of the double-headed inhibitor DSI from dog submandibular glands. Domain I contains the trypsin-reactive site (Arg-Leu), domain I I the elastase- or chymotrypsin-reactive centre (Met-Asp).

1978). The Ν-terminal domain o f D S I contains the trypsin-directed reactive site (Arg-Leu) whereas the chymotryps in- or elastase-reactive centre (Met-Asp) is located in the C-terminal doma in . This structure implies that ternary complexes (e.g. a t ryps in-DSI -e las tase complex) may be f o rmed .

The double-headed DSI molecule probably developed f r om a Kazal-type domain (as represented for example by the single-headed PSTI ) by gene dupl icat ion and several suitable mutat ions . The dr i v ing force for the expansion o f the inh ib i t i on spectrum could have been the need for the f o rmat ion o f enzymes w i t h new biological funct ions in the course o f evo lut ion . I n this respect it is remarkable that the DSI seems to be especially adapted to the food requirements o f canines, as it also inh ib i ts pronase, Aspergillus oryzae proteinase and subt i l is in very effectively (Fr i tz et al 1971). I n contrast to D S I , H U S I - I does not inh ib i t bacterial and m o u l d proteinases.

Inhibitors from soybeans and lima beans

Other inh ib i tors which interact strongly w i th human granulocyt ic elastase and cathepsin G are inhib i tors A A f r om soybeans and L B I f r o m l ima beans (Schiessler et al 1977b). These inhib i tors are structural ly homologous to each other. They consist o f two domains w i th independent reactive sites against t ryps in and chymotryps in or elastase (Fig. 6), so that ternary complexes may

P R O T E I N A S E I N H I B I T O R S I N I N F L A M M A T I O N 365

! 30

F I G . 6. Covalent structure of inhibitor AA (Bowman-Birk inhibitor) from soybeans. The trypsin-reactive site (Lys-Ser) is located in the left loop, the elastase- or chymotrypsin-reactive centre (Leu-Ser) in the right loop.

be f o rmed ( Ikenaka et al 1974). Bo th are readily available in amounts suff ic ient for therapeutic studies. However , because agglutinins are present in the p lant extracts, they have to be thorough ly pur i f i ed .

A pro tin in

The basic t ryps in-ka i l ikre in inh ib i t o r f r om bovine organs, ap ro t in in , used as an a f f in i t y adsorbent for the pur i f i ca t ion o f granulocyt ic elastase and cathepsin G (Baugh & Travis 1976, Travis et al 1978), turned out to be an inh ib i t o r o f these enzymes, too , but w i th relatively low a f f in i ty (Lestienne & Bieth 1978, Starkey 1977). Whether inh ib i t i on o f the granulocyt ic proteinases plays a role in its therapeutic effectiveness, which is generally assumed to be due to the inh ib i t i on o f p lasmin and/or plasma ka l l i k r e in , has st i l l to be investigated.

Microbial peptides

Recently inhib i tors w i th more restricted specificity and considerably lower molecular weight but weaker a f f in i t y to the granulocyt ic proteinases have been isolated f r om bacteria (Table 5) (Umezawa 1976, Feinstein 1978, Ohno et al 1978). They are potent ia l candidates for studies designed to clari fy the pathophysio log ica l role o f ind iv idua l granulocyt ic proteinases in severe i n f l ammato r y processes.

366 Η . F R I T Z E T A L

Inhibitors from the leech Hirudo medicinalis

Prote in proteinase inh ib i tors isolated so far f r o m extracts o f the leech Hirudo medicinalis are listed in Table 6. Hirudin, the thrombin-speci f ic inh ib i t o r , turned out to be the effective pr inc ip le o f the leech, widely used former ly in medical therapy for ' b l ood d i l u t i o n ' . The biochemical , pharmacological and c lo t t ing - inh ib i to ry properties o f h i r u d i n have been studied in detail (Ma rkward t 1963, Badgy et al 1976); its amino acid sequence has been elucidated recently (Petersen et al 1976). I n our op in i on h i rud in is a most promis ing candidate for therapeutic app l i cat ion i n consumpt ion coagulopathy, especially i n acquired or hereditary A T I I I deficiency.

T A B L E 6

Proteinase inhibitors of the leech Hirudo medicinalis (for further details see text)

Inhibitor Mol. wt. (approx.)

Inhibition of: Inhibitor Mol. wt.

(approx.) Trypsin0 Plasmin0 Acrosinc Thrombin0 Chymo­trypsin0

Granulocytic proteinase^

Hirudin 7000 - _ - + + -Bdellin A 6000 + + + + + - -Bdellin Β 4000 + + + + + - -Eglin 8100 ( + ) - + + + +

+ + , very strong; + , strong; ( + ), weak; - , no inhibit ion. aBovine. bPorcine, human. cBoar, human. dBovine, human. e Human (elastase and cathepsin G).

The eglins. In the course o f isolat ion o f the bdellins f r om H. medicinalis (these are strong inhib i tors o f p lasmin or sperm acrosin, cf. Table 6) an ant ichymotryps in act iv i ty was separated. A more detailed investigation revealed that this was due to inh ib i tors o f another type, the eglins (Seemüller et al 1977). They turned out to be especially strong inhib i tors o f human granulocyt ic elastase and cathepsin G, w i th K\ values o f the corresponding complexes close to 1.5 χ 1 0 " 1 0 mol/1 (elastase) or 2.5 χ 10" 1 0 mol/1 (cathepsin G ). In add i t i on , subti l is in is also strongly inhib i ted by the eglins (K\ approx. 1.2 χ 10" 1 0 mol/1). The eglins are a mix ture o f iso inhibi tors w i th s imi lar biochemical and inh ib i to ry properties.

The amino acid sequence shown below was elucidated very recently for eglin c.

I t is most surprising that the eglins are extremely resistant to denaturat ion and proteolyt ic degradation, despite the lack o f any disulphide bond which

P R O T E I N A S E I N H I B I T O R S I N I N F L A M M A T I O N 367

1 10 Thr-Glu-Phe-Gly-Ser-Glu-Leu-Lys-Ser-Phe-

20 Pro-Glu-Va l -Va l -G ly -Lys-Thr-Va l -Asp-Gln-

30 A la -Arg -Glu-Tyr -Phe-Thr -Leu-His -Tyr -Pro -

40 Gln-Tyr-Asn-Val -Tyr-Phe-Leu-Pro-Glu-Gly-

50 Ser-Pro-Val -Thr-Leu-Asp-Leu-Arg-Tyr-Asn-

60 Arg-Va l -Arg-Va l -Phe-Tyr-Asn-Pro-Gly-Thr-

70 Asn-Val -Va l -Asn-His-Va l -Pro-His-Va l -Gly

could stabilize the structure o f the molecule. This new structural type o f prote in proteinase inh ib i t o r seems to be especially suitable, therefore, for therapeutic use in both local and generalized in f l ammat i on . I n this respect it is remarkable that the an t i - in f l ammatory ( 'anti-phlogist ic ' ) effect o f leech extracts has long been k n o w n ; it could be at least part ly due to the ant iproteo lyt ic effect o f the eglins.

It is evident f r om these results that quite dif ferent structural approaches (cf. H U S I - I , i nh ib i t o r A A and eglin c) may lead to a reactive site con fo rmat i on which fits perfectly in to the active site o f the same enzyme (e.g. o f granulocyt ic elastase).

INHIBITOR THERAPY IN EXPERIMENTAL ENDOTOXAEMIA

The endotoxaemia model

Experimental procedure, Endotoxaemia was induced in six anaesthetized dogs by cont inuous in fus ion o f Escherichia coli endotox in , 2 mg/kg body weight, in to the in fer ior vena cava over a period o f two hours. The six con t ro l dogs were subjected to the same procedure except that isotonic saline was infused instead o f the endotox in so lut ion . The data moni to red or collected just before endotox in (or isotonic saline) appl icat ion served as 100% values for each parameter. A l l values obtained dur ing the 14-hour experiment were expressed as percentages o f these start ing values. Further clearly defined experimental condit ions a l l ow ing a statistical treatment o f the results by the three factor ia l classif ication o f variance w i l l be published elsewhere ( M .

368 Η. F R I T Z E T A I

Jochum, J . Wi t t e , H . Schiessler, G. Ruckdeschel & H . Fr i t z , in preparat ion) .

T o determine the plasma levels o f c lo t t ing and f ibr inolysis factors we used amido ly t ic assays w i th chromogenic substrates: T o s G l y P r o A r g N H N p (Boehringer) for p r o t h r o m b i n , D - P h e P i p A r g N H N p (S-2238 K A B I ) for A T I I I , and D - V a l L e u L y s N H N p (S-2251 K A B I ) for plasminogen and a 2 P I . The biological act iv i ty o f factor X I I I was assayed according to the procedure ' Fak to r X I I I Schnelltest' f r om Behringwerke Marburg/Germany , whi le complement factor C3 was quant i f ied by the radia l immunod i f f u s i on technique.

Haematological data. I n contrast to the contro l group the endotox in-treated dogs showed a rap id and substantial decrease in c i rculat ing platelets and leucocytes (Figs. 7 and 8). I n add i t i on , a strong leucocytosis developed in the endotoxin group, producing leucocyte counts far higher than before endotox in adminis t ra t ion (Fig. 8).

Plasma factors. Compared to the cont ro l group the levels o f the selected plasma factors decreased substantially in the endotoxin-treated dogs up to six hours ( A T I I I , c ^ P I , p r o t h r o m b i n , complement C3) or 14 hours (plasminogen, factor X I I I ) f r om the start o f the experiment (Fig. 9). This was

25-

plati rltts

\ \

<^2<H Ε ο

^ 15H =!

10-

I

u π 6 timt (h]

10

F I G . 7. Behaviour pattern of circulating platelets in experimental endotoxaemia. The curves represent mean values χ of the control (full line, η = 6) and endotoxin groups (dashed line, η = 6); the standard deviations, ± SEM, are indicated for each test point. Ordinate, number o f platelets/μΐ plasma. Abscissa, observation period. The endotoxin infusion period is indicated as a thick line on the abscissa.

P R O T E I N A S E I N H I B I T O R S I N I N F L A M M A T I O N 369

F I G . 8. Behaviour pattern of circulating leucocytes in experimental endotoxaemia. The curves represent mean values χ of the control (full line, η = 6) and endotoxin groups (dashed line, η -6); the standard deviations, ± SEM, are indicated for each test point. Ordinate, number of leucocytes/μΐ plasma. Abscissa, observation period. The endotoxin infusion period is indicated as a thick line on the abscissa.

0-1 r — ι 1 1 1 J f — Ί 1 1 1

0 2 6 10 Κ [h] 0 2 6 K> 14 [h]

F I G . 9. Plasma levels of selected plasma factors during experimental endotoxaemia (left) and under inhibitor medication (right). The curves represent mean values χ of the control groups A (n = 6), the endotoxaemia group Β (η = 6) and the inhibitor-medicated endotoxaemia group (right panel, η = 4); standard deviations are given elsewhere (Witte 1979). Ordinate, the percentage of the starting level (cf. text). Abscissa, experimental period (the endotoxin infusion period is indicated by the thick line). For further details see text.

370 Η. F R I T Z E T Λ I

also true for the f ibr inogen concentrat ion (80% o f the start ing value at the end o f the experiment, not shown in the f igure) . The paral le l decrease i n the physiological antagonists p r o t h r o m b i n / A T I I I and plasminogen/a2PI is especially remarkable.

Establishment of the model. The statistical evaluation revealed that the differences i n the levels o f the plasma factors and b lood cell counts were statistically signif icant for the cont ro l and endotoxaemia g roup (P <0.05/0.001) , the various test times (P <0.04/0.001) , and the t ime course (P <0.05/0.001) . This result clearly shows the va l id i ty o f this experimental system as a model for endotoxaemia. I n fact, the substantial alterations observed in b lood cell counts and the levels o f the selected plasma proteins are characteristic o f endotoxin- induced D I C (McCabe 1973, Garner et al 1974, Urbaschek et al 1975, Müller-Berghaus et al 1976, Aasen et al 1978a, 1978b, 1980).

influence of inhibitor medication

Experimental procedure. I n a f irst approach the elastase-cathepsin G inh ib i t o r A A f r om soybeans (cf. Table 5), described or ig inal ly by B o w m a n and B i rk (B i rk 1976), was applied in the endotoxaemia model in the dog. The inh ib i t o r was isolated f r om a commercia l ly available soybean extract ( ' t ryps in inh ib i t o r f r om soybeans' f r om Serva, Heidelberg, no. 37 340) by repeated gel f i l t r a t i on chromatography on Sephadex G-75 in 2 % acetic acid so lut ion ( H . Schiessler, personal communica t ion 1978). I n the inh ib i t o r medicat ion g roup ( four dogs) infusions o f endotox in (cf. above) and inh ib i t o r were started simultaneously but the inh ib i to r infusion was cont inued over the to ta l experimental per iod (14 h) . Each o f the four animals received 3 -8 mg/kg body weight o f the pur i f i ed inh ib i to r w i th a specific trypsin inh ib i to ry act iv i ty (substrate B z A r g N H N p ) close to 3.2 I U / m g . Further experimental details w i l l be published elsewhere ( M . Jochum, J . Wi t t e , Η. Schiessler, G. Ruckdeschel & H . Fr i t z , in preparat ion) .

Results and outlook. The inh ib i to r medicat ion influenced the behaviour pattern o f neither the c irculat ing platelets nor the leucocytes when compared to the endotox in group. However, the endotoxin- induced decrease in plasma levels o f the selected factors was signi f icantly (P <0 .05/0 .01 , Student Mest ) reduced by inh ib i t o r treatment (Fig. 9) ; even the lowest dosage o f i nh ib i t o r used (3 mg/kg body weight) was effective. As may be deduced f r o m the i nh ib i t i on spectrum o f inh ib i t o r A A , degradation o f the plasma factors i n the

P R O T E I N A S E I N H I B I T O R S I N I N F L A M M A T I O N 371

course o f endotoxaemia is due chief ly to unspecific proteolysis rather than to act ivat ion o f the b l ood systems by system-specific proteinases, though quant i ta t i on o f the degree o f consumpt ion due to unspecific proteolysis is not yet feasible.

The results o f bo th the c l inical and experimental study indicate clearly that in generalized in f l ammato ry processes such as septicaemia or septic shock the regulatory inh ib i t o r system o f the organism may be overstressed. As shown for the first t ime in the endotoxaemia t r i a l , a suitable inh ib i t o r can signi f icantly prevent unspecific degradation o f plasma factors by leucocytic (and other?) proteinases and thus help the natural defence mechanisms to ma in ta in the physiological balance. I n view o f the key funct ion o f caM in the rap id e l iminat ion o f proteinases f r om the c i rculat ion, protect ion against early ouM consumpt ion could be the under ly ing explanation for the effectiveness o f such an exogenous inh ib i t o r .

Just as impor t an t , i t seems to us, is to minimize hypercoagulabi l i ty and hyperf ibr inolys is or D I C induced simultaneously by system-specific proteinases. For this purpose ap ro t i n in , a strong plasmin inh ib i t o r already used in medical therapy, and the thrombin-speci f ic inh ib i to r h i rud in (cf. Table 6) are potent ia l candidates, the latter especially i f A T I I I is consumed to such an extent that heparin medicat ion is o f reduced effectiveness. As the systemic appl icat ion o f suitable elastase-cathepsin G inhib i tors and h i r u d i n in humans is not possible u n t i l tox ico logical , pharmacological and cl inical tr ia ls have been sucessfully made, subst i tut ion w i th A T I I I (Schramm 1977, Schramm et al 1978) and aiM concentrates is the method o f choice at present.

ACKNOWLEDGEMENTS

This work has been supported by grants from the Sonderforschungsbereich 51-München (B/3, B/8, B/15, and B/29, B/30). We wish to thank Professor Dr K. Hochstrasser (München) for estimating the serum levels of ITIIÖOOOO and ITI30000, Professor Dr H . Graeff and Dr R. Hafter (München) for fibrinogen determinations, Priv.-Doz. Dr H.-K. Selbmann (München) for the statistical evaluations and Mrs B. Förg-Brey for skilful technical assistance. We are especially indebted to Professor Dr G. Heberer (director of the surgical clinic of the Universität München) for his generous support o f the clinical study. The careful reading of this manuscript by Mrs J . Whelan (London) and Dr E. Fink (München) is gratefully acknowledged.

References

Aasen AO, Ohlsson Κ 1978 Release of granulocyte elastase in lethal canine endotoxin shock. Hoppe Seyler's Ζ Physiol Chem 359:683-690

Aasen AO, Dale J , Ohlsson K, Gallimore MJ 1978a Effects of slow intravenous administra­tion of endotoxin on blood cells and coagulation in dogs. Eur Surg Res 10:194-205

Aasen AO, Ohlsson K, Larsbraaten M, Amundsen Ε 1978b Changes in plasminogen levels,

372 Η . F R I T Z E T A L

plasmin activity and activity of antiplasmins during endotoxin shock in dogs. Eur Surg Res 10:63-72

Aasen AO, Mellbye OJ, Ohlsson Κ 1980 Complement activation during subsequent stages of canine endotoxin shock. Clin Exp Immunol, in press

Badgy D, Barabas E, Graf L, Petersen TE, Magnusson S 1976 Hirudin. Methods Enzymol 45: 669-678

Baugh RI , Travis J 1976 Human leukocyte granule elastase: rapid isolation and characterization. Biochemistry 15:836-841

Birk Y 1976 Trypsin and chymotrypsin inhibitors from soybeans. Methods Enzymol 45:700-707 Collen D, Wiman B, Verstraete Μ (eds) 1979 The physiological inhibitors of blood coagulation

and fibrinolysis. Elsevier/North-Holland Biomedical Press, Amsterdam Debanne MT, Bell R, Dolovich J 1976 Characteristics of the macrophage uptake of proteinase-a-

macroglobulin complexes. Biochim Biophys Acta 428:466-475 Dietl T, Dobrinski W, Hochstrasser Κ 1979 Human inter-a-trypsin inhibitor - l imited proteo­

lysis by trypsin, plasmin, kallikrein and granulocytic elastase and inhibitory properties o f the cleavage products. Hoppe Seyler's Ζ Physiol Chem 360:1313-1318

Egbring R, Havemann Κ 1978 Possible role o f polymorphonuclear granulocyte proteases in blood coagulation. In : Havemann K, Janoff A (eds) Neutral proteases o f human polymorphonuclear leukocytes. Urban & Schwarzenberg, Baltimore/Munich, ρ 442-458

Egbring R, Schmidt W, Fuchs G, Havemann Κ 1977 Demonstration of granulocytic proteases in plasma of patients with acute leukemia and septicemia with coagulation defects. Blood 49:219-231

Feinstein G 1978 Studies on inhibitor complexes of granulocyte neutral proteases with endo­genous serum inhibitors and actinomycetes polypeptides. In : Havemann K, Janoff A (eds) Neutral proteases of human polymorphonuclear leukocytes. Urban & Schwarzenberg, Balti­more/Munich, ρ 179-190

Fritz H , Jaumann E, Meister R, Pasquay P, Hochstrasser K, Fink Ε 1971 Proteinase inhibitors from dog submandibular glands — isolation, amino acid composition, inhibition spectrum. In : Fritz H, Tschesche Η (eds) Proc Int Res Conf on Proteinase Inhibitors, Munich Nov 4-6. W de Gruyter, Berlin, ρ 257-270

Gaffney PI 1977 Structure of fibrinogen and degradation products of fibrinogen and f ibr in . Br Med Bull 33:245-252

Garner R, Chater BV, Brown DL 1974 The role of complement in endotoxin shock and disse­minated intravascular coagulation: experimental observations in the dog. Br J Haematol 28: 393-401

Hamilton PI, Stalker AL , Douglas AS 1978 Disseminated intravascular coagulation: a review. J Clin Pathol 31:609-619

Harpel PC, Rosenberg RD 1975 c^-Macroglobulin and antithrombin-heparin cofactor: modula­tors of haemostatic and inflammatory reactions. In : Spaet TH (ed) Progress in haemostasis and thrombosis. Grune & Stratton, New York, vol 3: 145-190

Haschen RJ 1975 Proteases in specific cell types. In: Dingle JT, Dean RT (eds) Lysosomes in biology and pathology. North-Holland, Amsterdam, vol 4:251-264

Heimburger Ν 1975 Proteinase inhibitors of human plasma — their properties and control functions. In: Reich Ε et al (eds) Proteases and biological control. Cold Spring Harbor Labo­ratory, Cold Spring Harbor, ρ 367-386

Highsmith RF, Rosenberg RD 1974 The inhibition of human plasmin by human antithrombin-heparin cofactor. J Biol Chem 249:4335-4338

Hochstrasser K, Bretzel G, Wächter Ε, Heindl S 1975 The amino acid sequence of the double-headed proteinase inhibitor from canine submandibular glands. Hoppe Seyler's Ζ Physiol Chem 356:1865-1877 (revised structure, see Schiessler et al 1978)

Hochstrasser K, Niebl J , Lempart Κ 1977a Über Abbauprodukte des Inter-a-trypsininhibitors im Serum, I I . Säurelösliche Derivate des Inter-a-trypsininhibitors. Hoppe Seyler's Ζ Physiol Chem 55:343-345

P R O T E I N A S E I N H I B I T O R S I N I N F L A M M A T I O N 373

Hochstrasser K, Niehl J , Feuth H , Lempart Κ 1977b Über Abbauprodukte des Inter-a-trypsin-inhibitors im Serum, I . Der Inter-a-trypsininhibitor als Prekursor des säurestabilen Ser-um-Trypsin-Plasmin-Inhibitors. Kl in Wochenschr 55:337-342

Ikenaka T , Odani S, Koide Τ 1974 Chemical structure and inhibitory activities of soybean pro­teinase inhibitors. In : Fritz Η et al (eds) Proteinase inhibitors. Springer, Berlin (Bayer Sym­posium V) ρ 325-343

Jeljaszewicz J , Waldström Τ (eds) 1978 Bacterial toxins and cell membranes. Academic Press, London

Johnson U, Ohlsson K, Olsson I 1976 Effects of granulocyte neutral proteases on complement components. Scand J Immunol 5:421-426

Klebanoff SJ, Clark RA (eds) 1978 The neutrophil: function and clinical disorders. North-Hol land, Amsterdam

Krumme D, Wallner O, Fritz Η 1977 Proteinases and proteinase inhibitors in human cervical mucus - selected properties in view of their clinical relevance. In: Insler V, Bettendorf G (eds) The uterine cervix in reproduction. Thieme, Stuttgart, ρ 92-99

Laskowski Jr M , Kato I , Kohr WJ 1978 Protein inhibitors of serine proteinases: convergent evolution, multiple domains and hypervariability of reactive sites. In: Choh Hao Li (ed) Ver­satility of proteins. Academic Press, New York, ρ 307-318

Laurell C-B 1975 Relation between structure and biologic function of the protease inhibitors in the extracellular fluids. In : Peeters Η (ed) Protides of the biological fluids. Pergamon, Oxford (22nd Colloquium) ρ 3-12

Lestienne P, Bieth JG 1978 The inhibition of human leukocyte elastase by basic pancreatic trypsin inhibitor. Arch Biochem Biophys 190:358-360

Markwardt F (ed) 1963 Blutgerinnungshemmende Wirkstoffe aus blutsaugenden Tieren. VEB Gustav Fischer, Jena

McCabe WR 1973 Serum complement levels in bacteremia due to gram-negative organisms. Ν Engl J Med 288:21-23

Movat H Z (ed) 1979 Inflammatory reactions. Curr Top Pathol 68 Müller-Berghaus G, Bohn E, Hobel W 1976 Activation of intravascular coagulation by endo­

toxin: the significance of granulocytes and platelets. Br J Haematol 33:213-220 Myrvold HE 1976 Experimental septic shock. Acta Chir Scand Suppl 470:1-18 Nicholson A, Fearon DT, Austen KF 1978 Complement. In: Vane & Ferreira 1978, ρ 425-463 Ohlsson Κ 1974 Interaction between endogenous proteases and plasma protease inhibitors in

vitro and in vivo. In : Fritz Η et al (eds) Proteinase inhibitors. Springer, Berlin (Bayer Sympo­sium V) ρ 96-105

Ohlsson Κ 1978 Interaction of granulocyte neutral proteases with alphaι-antitrypsin, alpha:-macroglobulin and alphaj-antichymotrypsin. In : Havemann K, Janoff A (eds) Neutral pro­teases of human polymorphonuclear leukocytes. Urban & Schwarzenberg, Baltimore/ Munich, ρ 167-177

Ohlsson Κ 1980 Polymorphonuclear leukocyte collagenase. In : Woolley DE, Evanson JM (eds) Collagenase in normal and pathological connective tissues. Wiley, Chichester, ρ 209-222

Ohlsson K, Laurell C-B 1976 The disappearance of enzyme-inhibitor complexes from the circu­lation of man. Clin Sei Mol Med 51:87-92

Ohno Η, Saheki Τ, Awaya J , Nakagawa A, Ömura S 1978 Isolation and characterization of elasnin, a new human granulocyte elastase inhibitor produced by a strain of Streptomyces. J Antibiot Tokyo 31:1116-1123

Petersen TE, Roberts HR, Sottrup-Jensen L, Magnusson S 1976 Primary structure of hirudin, a thrombin-specific inhibitor. In: Peeters Η (ed) Protides of the biological fluids. Pergamon, Oxford (23 r d colloquium) ρ 145-149

Schiessler Η 1976 Säurestabile Proteinase-Inhibitoren aus menschlichem Sperma und ihre Ziel­enzyme. In : Advances in andrology. Grosse Verlag, Berlin, vol 5:45-102

Schiessler H , Arnhold M , Ohlsson K, Fritz Η 1976a Inhibitors of acrosin and granulocyte protei­nases from human genital tract secretions. Hoppe Seyler's Ζ Physiol Chem 357:1251-1260

374 Η . F R I T Z E T A L

Schiessler Η, Fink Ε, Fritz Η 1976b Acid stable proteinase inhibitors from human seminal plasma. Methods Enzymol 45:847-859

Schiessler H , Ohlsson K, Fritz Η 1977a The acid-stable proteinase inhibitor (antileukoprotease) o f human cervical mucus. In : Insler V, Bettendorf G (eds) The uterine cervix in reproduction. Thieme, Stuttgart, ρ 84-89

Schiessler H , Ohlsson K, Olsson I , Arnhold M , Birk Y, Fritz Η 1977b Elastases from human and canine granulocytes: interaction with protease inhibitors of animal, plant, and microbial origin. Hoppe Seyler's Ζ Physiol Chem 358:53-58

Schiessler H , Kaplan O, Wartenberg S, Witte J 1978a Effect of a protease-inhibitor on the concentration of the fibrin-stabilizing factor (X I I I ) in the course of acute gram-negative sepsis. International Congress of Inflammation, Bologna/Italy Oct 31-Nov 3, Abstracts ρ 195

Schiessler H , Hochstrasser K, Ohlsson Κ 1978b Acid-stable inhibitors o f granulocyte neutral proteases in human mucous secretions: biochemistry and possible biological function. In : Havemann K, Janoff A (eds) Neutral proteases o f human polymorphonuclear leukocytes. Urban & Schwarzenberg, Baltimore/Munich, ρ 195-207

Schill W-B, Wallner Ο, Schiessler Η, Fritz Η 1978 Immuno fluorescent localization o f the acid-stable proteinase inhibitor (antileukoprotease) o f human cervical mucus. Experientia (Basel) 34:509-510

Schmidt W, Egbring R, Havemann Κ 1974 Effect o f elastase-like neutral protease from human granulocytes on isolated clotting factors. Thromb Res 6:315-326

Schramm W 1977 Discussion remarks. In : Marx R, Thies H A (eds) Klinische und ambulante A n ­wendung klassischer Antikoagulation. FK Schattauer, Stuttgart, ρ 139-141, 217-219

Schramm W, Gröschel G, Segerer W, Marx R 1978 Zur Optimierung der antithrombotischen Therapie bei Hämodialysen und der Hämofiltration. In : Marx R, Thies H A (eds) Niere — Blut­gerinnung und Hämostase. FK Schattauer, Stuttgart/New York, ρ 299-304

Schramm W, Drost W, Schmidt M , Borlinghaus P, Marx R 1980 Fibrinopeptid-A-Bestimmung: Methodik und Klinische Ergebnisse. In : Schimpf Κ (ed) Verhandlungsbericht über die 23. Tagung der Deutschen Arbeitsgemeinschaft für Blutgerinnungsforschung in Heidelberg im Februar 1979. FK Schattauer, Stuttgart/New York

Seegers W H 1978 Anti thrombin I I I . Theory and clinical applications. A m J Clin Pathol 69:367-374

Seemüller U, Meier M , Ohlsson K, Müller H-P, Fritz Η 1977 isolation and characterization of a low molecular weight inhibitor (of chymotrypsin and human granulocytic elastase and ca-thepsin G) from leeches. Hoppe Seyler's Ζ Physiol Chem 358:1 105-1117

Skansberg P, Cronberg S, Nilsson IM 1974 The occurrence and significance of fibrin/fibrinogen degradation products (FDP) in acute infections. Scand J Infect Dis 6:197-203

Spragg J , Austen KF 1977 Plasma factors: the Hageman-faetor-dependent pathways and the complement sequence. In: Hadden JW et al (eds) Immunopharmacology. Plenum Medical Book Co., New York, ρ 125-143

Starkey PM 1977 Elastase and cathepsin G; the serine proteinases of human neutrophil leuco­cytes and spleen. In: Barrett AJ (ed) Proteinases of mammalian cells and tissues. Elsevier/ North-Holland Biomedical Press, Amsterdam, ρ 57-89

Starkey PM, Barrett AJ 1977 «2-Macroglobulin, a physiological regulator of proteinase activity. In : Barrett AJ (ed) Proteinases of mammalian cells and tissues. Elsevier/North-Holland Bio­medical Press, Amsterdam, ρ 663-696

Tegner H, Ohlsson Κ 1977 Localization of a low molecular weight protease inhibitor to tracheal and maxillary sinus mucosa. Hoppe Seyler's Ζ Physiol Chem 358:425-429

Travis J , Baugh R, Giles PJ, Johnson D, Bowen J , Reilly CF 1978 Human leukocyte elastase and cathepsin G: isolation, characterization and interaction with plasma proteinase inhibitors. In : Havemann K, Janoff A (eds) Neutral proteases of human polymorphonuclear leukocytes. Urban & Schwarzenberg, Baltimore/Munich, ρ 118-128

Tschesche H , Kupfer S, Klauser R, Fink E, Fritz Η 1976 Structure, biochemistry and comparative aspects of mammalian seminal plasma acrosin inhibitors. In: Peeters Η (ed) Protides of the bio-

P R O T E I N A S E I N H I B I T O R S I N I N F L A M M A T I O N 375

logical fluids. Pergamon, Oxford (23 r d Colloquium) ρ 255-266 Turk JL, Willoughby DA 1978 Immunological and para-immunological aspects in inflammation.

In : Vane & Ferreira 1978, ρ 231-266 Umezawa Η 1976 Structures and activities of protease inhibitors of microbial origin. In : Methods

Enzymol 45:678-695 Urbaschek B, Urbaschek R, Neter Ε (eds) 1975 Gram-negative bacterial infections and mode of

endotoxin actions. Springer, Wien/New York Vane JR, Ferreira SH (eds) 1978 Inflammation. Springer, Berlin (Handb Exp Pharmacol, vol

50/1) Venneröd A M , Laake Κ, Solberg AK, Strömland S 1976 Inactivation and binding of human plas­

ma kallikrein by antithrombin I I I and heparin. Thromb Res 9:457-466 Wächter E, Hochstrasser Κ 1979 Kunitz-type proteinase inhibitors derived by limited proteolysis

of the inter-cv-trypsin inhibitor, I I I : Sequence of the two Kunitz-type domains inside the native inter-a-trypsin inhibitor, its biological aspects and also of its cleavage products. Hoppe Seyler's Ζ Physiol Chem 360:1305-1311

Weissmann G (ed) 1974 Mediators of inflammation. Plenum Press, New York Witte J 1979 Endotoxinämie und hyperdynamer septischer Schock: Pathobiochemie ausgewählter

Gerinnungs- und anderer Plasma-Proteinparameter. Academic Thesis, Medical Faculty of the University of Munich

Woolley DE, Roberts DR, Evanson JM 1976 Small molecular weight ß\ serum protein which spe­cifically inhibits human collagenases. Nature (Lond) 261:325-327