the role of complement in acute antibody ...skin allografts could not be destroyed by injection of...

THE ROLE OF COMPLEMENT IN ACUTE ANTIBODY-MEDIATED REJECTION

OF SKIN GRAFTS IN THE MOUSE

J. H. M. BERDEN

PROMOTOR : D r . R . A . P . Koene

CO-REFERENT : D r . P . J . A . C a p e l



PROEFSCHRIFT

TER VERKRIJGING VAN DE GRAAD VAN DOCTOR IN DE

GENEESKUNDE AAN DE KATHOLIEKE UNIVERSITEIT TE

NIJMEGEN, OP GEZAG VAN DE RECTOR MAGNIFICUS

PROF. DR. P. G. A. B. WIJDEVELD VOLGENS BESLUIT VAN

HET COLLEGE VAN DECANEN IN HET OPENBAAR TE

VERDEDIGEN OP 27 APRIL 1979 DES NAMIDDAGS TE

4 UUR

door

JOHANNES HENRICUS MARIE BERDEN

GEBOREN TE BLERICK

KRIPS REPRO - MEPPEL

Dit proefschrift werd bewerkt op de afdeling nierziekten (hoofd:

Prof.Dr. P.G.A.B. Wijdeveld) van de Universiteitskliniek voor

Inwendige Ziekten (directeur Prof.Dr. C.L.H. Majoor) van het St.

Radboud Ziekenhuis te Nijmegen.

Deze onderzoekingen werden gesteund door de Nier Stichting Neder

land en door de Stichting voor Medisch Wetenschappelijk Onderzoek

FUNGO. Het proefschrift werd uitgegeven met steun van de Nier

Stichting Nederland.

/¡ом. 9'Ш&.) iwf- ги, Ъщ.

Й-еЬи. IUAU<W O-UjdMJ.

CONTENTS

CHAPTER I

INTRODUCTION AND PROBLEM STATEMENT 9

References 14

CHAPTER II

THE COMPLEMENT SYSTEM 17

2.1. Introduction ig

2.2. Nomenclature of the complement system 20

2.3. Immunochemstry of complement 21

2.3.1. The classical activation unit 21

2.3.2. The alternative activation unit 24

2.3.3. The terminal attack unit 28

2.4. Immunobiology of complement 29

2.4.1. Alteration of the target complex 30

- Immune adherence 30

- Cytolysis 31

- Virus neutralization 32

- Solubilization of immune conplexes 32

2.4.2. Alteration of the micro-environment 33

- Permeability factors 34

- Chemotactic factors 35

2.4.3. Systemic effects of complement 35

- Mobilization of leukocytes 35

- Interrelations between the complement-, the clotting-,

the kinine generating- and the fibrinolytic system 36

2.5. Genetics of the complement system 39

2.5.1. Complement deficiency in man 39

2.5.2. Complement deficiency in animals 41

- C4 deficiency in guinea pigs 41

- C5 deficiency in mice 42

- C6 deficiency in rabbits 43

б

2.5.3. Relation between complément and major histocompatibility

complex 44

2.6. References 46

CHAPTER III

65

66

68

68

69

PARTICIPATION OF COMPLEMENT IN GRAFT REJECTION

A REVIEW OF THE LITERATURE

3.1. Introduction

3.2. Immunofluorescence studies of grafts

3.3. Prevention of complement fixation

3.4. Measurement of complement activity during graft rejection

3.5. Influence of complement depletion or congenital complement

deficiency of the host on allograft survival 75

3.6. Influence of complement depletion or congenital complement

deficiency of the host on xenograft survival 77

3 . 7 . C o n c l u s i o n 77

3 . 8 . R e f e r e n c e s 79

CHAPTER IV

ROLE OF ANTISERUM AND COMPLEMENT IN THE ACUTE ANTIBODY-MEDIATED

REJECTION OF MOUSE SKIN ALLOGRAFTS IN STRAIN COMBINATIONS WITH

INCREASING HISTOCOMPATIBILITY 89

Summary 91

Introduction 91

Materials and methods 92

Results 94

Discussion 96

References 98

CHAPTER V

THE ROLE OF COMPLEMENT FACTORS IN ACUTE ANTIBODY-MEDIATED REJECTION

OF MOUSE SKIN ALLOGRAFTS 99

Summary 100

Introduction 100


Results 102

Discussion Ю З

References 104

CHAPTER VI

A SENSITIVE HAEMOLYTIC ASSAY OF MOUSE COMPLEMENT 105

Summary 107

Introduction 107


Results 110

Discussion 115

References 116

CHAPTER VII

COMPLEMENT DEPENDENT AND -INDEPENDENT MECHANISMS IN ACUTE ANTIBODY-

MEDIATED REJECTION OF SKIN XENOGRAFTS IN THE MOUSE 119

Introduction 120


Results I2 4

Discussion 133

Summary 136

References 137

CHAPTER VIII

SUMMARY 141

8

Chapter I

INTRODUCTION AND PROBLEM STATEMENT

9

Replacement of kidneys v/hich have irreversibly lost their

function is now a generally accepted form of treatment. The main

technical and surgical problems in kidney transplantation have

been solved. However, five year survival rates of transplanted

kidneys in centres participating in the Eurotransplant Organiza

tion are still not higher than 39.6% if there is a four antigen

mismatch and 56.6% if there are no HLA-A and В mismatches (Euro-

transplant Annual Report, 1977). The majority of losses is caused

by rejection of the graft.

After transplantation a dual response can occur which may lead

to rejection. The first type of response is mediated by specifi

cally activated cytotoxic Τ cells. It most often occurs within

a few days after transplantation, and leads to a so-called

cellular rejection of acute onset. In many instances this type of

rejection can be treated with success by increasing the dose of

immunosuppressive drugs. The second type of rejection is mediated

by antibodies specifically directed against the donor antigens.

These antibodies, if complexed with the antigens toward which

they are directed,can trigger a number of effector systems, which

can cause the eventual damage of the graft. One of the important

inflammatory effector systems triggered by antigen-antibody com

plexes is the complement sequence. The humoral type of rejection

has in most instances an insidious onset and responds very poorly

to immunosuppressive treatment. It therefore accounts for a con

siderable amount of graft losses. There exists a host of informa

tion concerning the pathogenic significance of humoral anti-donor

responses for the ultimate fate of the graft (reviewed by Carpen

ter et al, 1976). Anti-donor antibodies in the complement dependent

cytotoxicity assay or a positive antibody dependent lymphocyto-

toxicity were found during rejection episodes in almost 100% of

the cases (Stiller et al, 1976; Gailiunas et al, 1978). Histologic

examinations of grafts removed after chronic rejection show

intimai proliferation and media thickening (Kincaid-Smith, 196 7;

Busch et al, 1969), associated with deposition of IgG on the

vascular endothelium and in the tunica media (Busch et al, 1971).

10

The development of these lesions is associated with circulating

anti-donor antibodies (Jeannet et al, 19 70). The anti-donor

antibodies are not always detectable in the circulation, but this

can not serve as an argument against humorally mediated rejection,

because the antibodies formed can to a great extent be absorbed

by the graft. In these cases they are only detectable in the

circulation after removal of the grafted kidney (Milgrom, 1977) .

The most ideal approach to improve the results of human renal

allografting is the prevention of the above mentioned rejection

responses in the recipient by complete matching for the antigens

coded for by the major histocompatibility complex. This has been

done most successfully in the transplantation of kidneys from

living, МНС-identical sibling donors. In such donor-recipient

combinations graft survival rates between 90 - 100% have been

achieved. However, in cadaveric kidney transplantation complete

compatibility between donor and recipient is very rare because

of the great polymorphism of the МНС-system and the presence of

hitherto unidentified transplantation loci. This problem of

graft rejection calls for a continuing search for better and less

toxic forms of immunosuppression. The recently discovered strong

immunosuppressive activity of cyclosporin A might open new possi

bilities in this area (Borei et al, 1977; Green and Allison,

1978; Calne et al, 1978a> 1978b). In the humoral rejection process

the destructive effects of formed anti-donor antibodies might be

prevented by manipulating the host's complement system. There

exists a great deal of evidence in the literature that complement

participates in humoral graft rejection, especially in human

renal allografting. The literature concerning this aspect is dis

cussed in one of the subsequent sections (chapter III).

Before complement inhibition can become applicable we must

have a better understanding of the fundamental mechanisms that

are operative in this humoral rejection process. In 19 73 an ex

perimental animal model was developed in our laboratory in which

antibody-mediated destruction could be reproducibly elicited

(Koene et al, 1973). The model was based on previous studies in

which it was shown that mouse alloantisera are devoid of cyto

toxic activity versus donor lymphoid cells if fresh mouse serum

11

is used as a complement source. Cytotoxicity becomes only apparent

in the presence of guinea pig complement, and is highest with

rabbit complement (Koene and McKenzie, 1973). Similarly, mouse

skin allografts could not be destroyed by injection of anti-donor

antibodies. Acute antibody-mediated rejection (AAR) of established

skin grafts in immunosuppressed mice could, however, be induced

by the administration of anti-donor serum together with rabbit

complement. Administration of exogenous complement was not in all

donor-recipient combinations a prerequisite to induce antibody-

mediated destruction. Baldamus et al (1973) described a xeno

geneic skin graft model, in which rat skin grafted onto immuno

suppressed mice was destroyed by passive administration of mouse

anti rat serum alone. The mouse endogenous complement was suffi

ciently activated in this model, which was most likely a conse

quence of the greater antigenic disparity. We, therefore, decided

to investigate the influence of antigenic disparity on AAR in a

number of graft experiments in strain combinations with increasing

histoincompatibility. Because of the earlier mentioned differences

in the cytolytic efficiency of different complement sources in

lymphocytotoxicity assays their ability to mediate AAR in vivo

was compared to their cytotoxicity in vitro with the same anti

serum (chapter IV).

From earlier experiments it was known that, if rabbit comple

ment was inactivated prior to its administration along with anti

serum, AAR did not occur (Gerlag, 1975) . This effect has now been

studied in more detail and we have especially focused our attention

on the question which complement factors were necessary for the

induction of AAR. This was done by depleting the complement source

in vitro of distinct complement factors prior to its administration

and by the use of complement factor deficient sera. This approach

can provide information at which level the complement sequence

has to be inhibited in order to prevent AAR (chapter V).

Although manipulation of an administered exogenous complement

species can give fundamental insight into the mechanisms of AAR,

it is indispensable to complete the information thus gained by

studies of the endogenous complement system. Especially its con-

12

sumption during AAR, and on the other hand the influence of in

hibition of the complement sequence on AAR are important in this

regard. For this purpose we used the xenogeneic skin graft model

in the mouse, because in this system AAR is mediated by the en

dogenous complement of the mouse (chapter VII).

Before such a study was possible we had to develop a sensitive

assay for the accurate measurements of mouse complement levels,

since there was no method reported in the literature that had the

desired sensitivity (chapter VI) .

Prior to the presentation of the experimental data a review is

given concerning the immunochemical sequences and immunobiological

effects of the complement system. Special reference is given to

certain complement deficiencies in animals since we used sera of

these animals in a number of experiments (chapter II).

13

REFERENCES

B a l d a m u s , C A . , M c K e n z i e , I . F . С , Winn, H . J . , a n d R u s s e l l , P . S .

( 1 9 7 3 ) : A c u t e d e s t r u c t i o n by h u m o r a l a n t i b o d y o f r a t s k i n g r a f t e d

o n t o m i c e .

J . I m m u n o l . 110, 1 5 3 2 .

B o r e i , J . F . , F e u r e r , C , Magnée , С , a n d S t ä h e l i n , H. ( 1 9 7 7 ) :

E f f e c t s o f t h e new a n t i - l y m p h o c y t i c p e p t i d e c y c l o s p o r i n A i n

a n i m a l s .

Immunology 32, 1017.

Busch, G.J., Braun, W.E., Carpenter, C.B., Corson, J.M., Galvanek,

E.R., Reynolds, E.S., Merrill, J.P., and Dammin, G.J. (1969):

Intravascular coagulation in human renal allograft rejection.

Transplant. Proc. 1, 267.

Busch, G.J., Reynolds, E.S., Galvanek, E.G., Braun, W.E., and

Dammin, G.J. (1971): Human renal allografts. The role of vascular

injury in early graft failure.

Medicine (Baltimore) 50, 29.

Calne, R.Y., Rolles, К., White, D.J.G., Smith, D.P., and Herbert-

son, B.M. (1978a): Prolonged survival of pig orthotopic heart

grafts treated with cyclosporin A.

Lancet 1, 1183.

Calne, R.Y., Thiru, S., McMaster, P., Craddock, G.N., White, D.J.G.,

Evans, D.B., Dunn, D.C., Pentlow, B.D., and Rolles, К. (1978b):

Cyclosporin A in patients receiving renal allografts from cadaver

donors.

Lancet 2, 1323.

Carpenter, C.B., d'Apice, A.J.F., and Abbas, A.K. (1976): The role

of antibody in the rejection and enhancement of organ allografts.

Adv. Immunol. 22, 1.

14

Gailiunas, P., Suthanthiran, M., Person, Α., Strom, Τ.Β., Carpen

ter, C.B., and Garovoy, M.R. (1978): Posttransplant immunologic

monitoring of the renal allograft recipient.


Gerlag, P.G.G. (1975): Hyperacute afstoting van huidtransplantaten

bij de muis.

Thesis Katholieke Universiteit Nijmegen, pag. 61.

Green, C.J., and Allison, A.C. (1978): Extensive prolongation of

rabbit kidney allograft survival after short-term cyclosporin A

treatment.

Lancet 1, 1182.

Jeannet, M., Finn, V.W., Flax, M.H., Winn, H.J., and Russell, P.S.

(1970): Humoral antibodies in renal allotransplantation in man.

New Engl. J. Med. 282, 111.

Kincaid-Smith, P. (1967): Histological diagnosis of rejection in

renal homografts in man.

Lancet 2, 849.

Koene, R.A.P., Gerlag, P.G.G., Hagemann, J.H.F.M., van Haelst,

U.J.G., and Wijdeveld, P.G.А.В. (1973): Hyperacute rejection of

skin allografts in the mouse by the administration of alloantibody

and complement.

J. Immunol. Ill, 520.

Koene, R.A.P., and McKenzie, I.F.С. (1973): A comparison of the

cytolytic action of guinea pig and rabbit complement on sensitized

nucleated mouse cells.


Milgrom, F. (1977): Role of humoral antibodies in transplantation.


Stiller, C.R. (1976): Anti-donor immune responses in prediction

of transplant rejection.

New Engl. J. Med. 234, 978.

15

Chapter II

THE COMPLEMENT SYSTEM

17

2.1. INTRODUCTION

Complement is one of the major effector systems in the inflan-

matory and immunologic response. It consists of a number of

proteins, present in serum in non-active precursor forms (Table

1). Most of these obtain enzyme-like activity after activation,

leading to a sequential interaction of the individual complement

components. After activation by immunoglobulins, fungal or bac

terial products, multimolecular assemblages are formed, prefer

entially on the surface of biological membranes. This can lead

to the alteration of the target complex, and, depending on the

cell type involved, specialized cell functions are initiated

such as histamine release, enhancement of phagocytosis, con

traction of smooth muscles, aggregation and fusion of platelets.

Alternatively, lysis of the cell membrane may occur with conse

quent death of the target cell. Activated complement components

released in the fluid phase can also induce alterations of the

micro-environment such as Increased vascular permeability, and

Chemotaxis. Furthermore, the complement system shows interactions

with other plasma protein systems, e.g. the clotting sequence,

the fibrinolytic system, and the kinm-generating system.

The complement system is most conveniently divided into four

parts. This is schematically illustrated in figure 1. There are

two pathways for activation: the classical and alternative path

way which both lead to the generation of a C3-convertase. The

amplification loop is a positive feed-back cycle, in which

generation of C3b promotes the activation of the alternative

pathway. The terminal "attack" unit is a kind of final common

pathway, which generates Chemotaxis and induces cell lysis.

Spontaneous and generalized complement activation, with its

potentially dangerous effects, is prevented by the rapid intrin

sic decay of the formed enzymes and by the existence of several

control proteins. Table 1 summarizes some characteristics of the

various proteins involved in the complement system. The inrnmno-

chemistry and immunobiology of the various complement components

will be discussed in more detail in the subsequent sections.

18

TABLE I. Physiooaherrtcal aharaoteristzos of proteins of the oorrplement system

Approximate Electro- Serum Cleavage

molecular phoretic concentration fragments

weight(daltons) nobility (ug/ml)

Classiaal pathway

Clq

Clr

Cls

C4

C2

Alternative pathuay

Properdin

Factor D (C3PAse)b

390,000

188,000

86,000

209,000

117,000

223,000

25,000

Factor В (C3PA,GBG)100,000

C3 (BiO 185,000

Ύ2

6

аг

ei

ß2

Y2

«2

ß2

Bi

190

100

120

430

30

25

trace

240

1300

C4a,C4b,C4c,C4d

C2a,C2b(?C-kinin)

Bb(C3A,GGG),

Ba(GAG)

CSajCSb.CScißjA)

C3d(a2D)

Initiating factor IF

C3 nephritic factor

C3NeF

Attack sequence

C5

C6

C7

C8

C9

Control proteins

cT. INH

3iH

C3b.INA(KAF)

AI

170,000

160,000

206,000

95,000

120,000

153,000

79,000

105,000

150,000

100,000

310,000

6

Y

ei

ß2

ß2

<n a

<*2

e ß2

02

?

•p

75

60

55

40

10

180

600

25

7

C5a,C5b

a) Modified, according to Fearon and Austen, 1976.

b) Formerly used synonyms given between parentheses.

19

classical activation , unit r\ \ CA >

\ C2 J ca \ 1

/сзь J / в / 1 D / /

\ / Сэ- convertase

\

С 5-convertase

Ï

/ C 6

C7 ce с э У

terminal attack unit

alternative , activation | unit

\

J amplification loop

Figure 1. Schematic representation of the four functional parts

of the complement system.

2.2. NOMENCLATURE OF THE COMPLEMENT SYSTEM

In 1968 a standard nomenclature for the classical complement

sequence was chosen (Bull. WHO, 1968) to cope with the existing

confusion in terminology (see synonyms table 1). The complement

factors are notated by numbers in the same sequence as they

react in immune hemolysis. Each number is preceded by the symbol

C. For historical reasons an exception has been made for C4,

giving the sequence: C1-C4-C2-C3-C5-C6-C7-C8-C9. The first com

ponent CI comprises 3, possibly 4, distinct protein subcomponents

named Clq-Clr-Cls (and probably Clt). If a component becomes

activated, the activated state is denoted by a vertical bar over

the number for example: cTs. The postscript "i" represents the

inactive state of a previously active component (e.g. C2i) .

If a component is cleaved the postscript "a" is used for the

20

smaller fragment, and the postscript "b" for the bigger one

(C3a-C3b). C3b and C4b further cleave into two fragments each,

which are denoted as C3c-C3d and C4c-C4d. The factors involved

in the alternative pathway are notated in capitals (IF, B, D and

P). The antigen-antibody complex that initiates the classical

pathway is denoted as SA, while the activating substance for the

alternative pathway is represented by S. The control proteins of

the complement sequence are generally abbreviated as Cl.INH

(CI esterase inhibitor); C3b.INA (C3b inactivator); ßlH; and AI

(anaphylatoxin inactivator).

2.3. IMTIUNOCHEMISTRY OF COMPLEMENT

2.3.1. The classical activation unit (fig. 2)

Activation of the factors of this unit leads to the formation

of a C3 convertase. The activation is initiated by binding of the

CI molecule to antigen-antibody complexes, or to non-specifically

aggregated immunoglobulins (IgG and IgM). CI consists of 3 sub-

units Clq, Clr, and Cls presumably in equimolar ratio in the

presence of one molecule Ca which is necessary for the struc

tural integrity of the CI molecule (Lepow et al, 1963) . Recently

a fourth subunit of CI has been described, called Clt (Assimeh

and Painter, 1975). This subunit is not necessary for the hemo

lytic activity of CI, and its precise function is as yet unknown.

The process is started by binding of Clq to the immunoglobulin.

The binding site for Clq resides in the IgG molecule in the

second constant domain (CH2) of the Fc region (Ellerson et al,

1972) , and in the fourth constant domain (CH4) of the IgM mole

cule (Hurst et al, 1975). For the activation of Clq at least

2 IgG antibodies are necessary, which moreover should be located

in a certain spatial relationship, while in the case of IgM one

molecule is sufficient (Borsos and Rapp, 1965). The Clq subunit

consists of 6 identical non-covalently linked subunits, with an

N-terminal collagen-like region and a C-terminal globular segment

(Reíd and Porter, 1976). The six subunits are joined at their

21

SA

^-Clqrs

SA C i

= = inhibition by Cî INH

• S.CÎbiô classical Сз convertase

C4c-C4d

Figure 2. The classical activation unit

N-terminal regions to form a stem fibril from which branch the

individual C-globular segments (fig. 3). Each of these six glo

bular portions can bind to one IgGl, IgG2, IgG3 or IgM antibody.

Binding of at least two globular portions of the Clq molecule

leads to steric changes in the Clr subunit (Reid and Porter,

1976), which results in the formation of cTr, that has an esterase

activity (Naff and Ratnoff, 1968). The natural substrate of cTr

is the Cls subunit, that obtains esterase activity after cleavage.

cTs and cTr are susceptible to the naturally occurring cT in

hibitor (CT.INH) (Pensky et al, 1961). cT.INH stoichiometrically

inhibits the enzymatic function of cTr and cTs (Gigli et al, 1968) .

Only if Clr is formed in such amounts that it locally exceeds the

inhibitory capacity of cT.INH, the next activation step follows.

The substrates of CTs are the complement factors C4 and C2. C4 is

cleaved into two fragments by Cls, a larger fragment C4b which

binds to cell membranes and immune complexes, (Müller-Eberhard

and Biro, 1963) and a smaller fragment C4a that is released into

the fluid-phase. This latter fragment has weak smooth muscle

contractile properties (Schreiber and Müller-Eberhard, 1974).

Bound C4b has also an affinity to immune adherence receptors,

22

present on various cells, to which C3b is preferentially bound

(Cooper, 1969). The binding of C4b to the cell membrane is very

labile; most molecules dissociate and are released into the

fluid phase, where they are rapidly cleaved by C4b-inactivator

into C4c and C4d. This leads to loss of both hemolytic

activity and the capacity to bind to immune complexes or cell

antibody binding site

globular portion (C terminal)

end fibril j „„(аде,,.,.!« J portion

stem fibril Μ N terminal)

Figure 3. Ultrastructural composition of human Clq (after Reid

and Porter, 1976) — , portions of the moleoule pointing

towards the reader; , portions of the moleoule

pointing away from the reader.

membranes. This C4b inactivator is very probably identical with

C3b.INA (Cooper, 1975). The next step after cleavage of C4 is

the splitting of C2 by cTs in C2a and one or more minor fragments

(Polley and Müller-Eberhard, 1968). Before C2 is cleaved, it is

loosely bound to the C4b molecule, which requires the presence

of Mg (Sitomer et al, 1966). The presence of C4b on the cTs

molecule enhances the splitting of C2 by Cls, by an allosteric

modification of cTs or C2 (Gigli and Austen, 1969). The C2a

molecule forms on the cell membrane a bimolecular complex with

C4b, yielding the classical C3 convertase C4b2a, which cleaves

and activates C3 (Müller-Eberhard et al, 1967) . For this C3

23

Splitting activity Cis is not required. Within the C4b2a complex

C4b is responsible for the binding to the target cell membrane

(Müller-Eberhard et al, 1967) while the enzymatic properties

reside in the C2a fragment. After cleavage of C3 and assembly of

C3b into the complex, forming the classical C5 convertase C4b2a3b,

C2a also carries out the enzymatic cleavage of C5 (Cooper, 1971;

Shin et al, 1971). Once formed C3 convertase has a relatively

short half-life (5 min. at 37 C), because C2a easily dissociates

into the fluid phase, where it is readily inactivated into C2i.

The C2i molecule is not re-usable for formation of a C4b2a com

plex in contrast to C14b that again can react with a native C2

molecule to restore C4b2a.

2.3.2. The alternative activation unit (fig. 4)

Activation of the alternative pathway also yields C3 conver

tase activity, but without the use of the classical activation

factors CI, C4 and C2. Activation of the alternative pathway can

be initiated by IgA immunoglobulins, polysaccharides (zymosan,

inulin, endotoxins of gram negative bacteria), dextran sulphate,

T-cell independent antigens, B-cell mitogens (pokeweed), and other

substances (Götze and Müller-Eberhard, 1977). No specific chemi

cal structure responsible for this activation has hitherto been

recognized. However, all activating substances have a polymeric

structure, consist of repeating subunits and contain poly

saccharides. On the surface of the activating substance, a very

labile C3 convertase is formed which is an Mg dependent complex

of the initiating factor (IF), factor B, factor D, and native C3.

The existence of the IF has been discovered after recognition of

a certain protein present in sera of some patients with hypo-

complementaemia and membranoproliferative glomerulonephritis

(Spitzer et al, 1969) or partial lipodystrophy (Sissons et al,

1976). This protein, called C3 nephritic factor (C3MeF), generates,

in the presence of factor B, factor D, C3, and tig , a stable

soluble C3 convertase (Schreiber et al, 1975). Antisera prepared

against C3NeF, remove a factor from normal serum, after which

activation of the alternative pathway with for instance zymosan

24

52-СЭЬРВЬ

Ρ stabilized Сэ convertase

C 3 № F stabilized Сз convertase

C3b dependent Сз convertose (labile)

C5 convertase (labile)

Ρ stabilized CS convertase

Figure 4. Activation sequence of the alternative pathway (Modified, after Götze and Müller-Eberhard, 1977),

сап no longer be induced (Schreiber et al, 1976a). The factor

removed by anti C3NeF antisera was called initiating factor (IF).

However, this IF is a distinct protein, not identical with C3NeF.

The nature of interactions between the activating substances and

IF are completely unknown. However, the initiation is started

by the formation of an initial C3-convertase. After formation of

this convertase on the surface of the activating substance, the

enzyme deposits formed C3b in its vicinity. The formation of C3b

is a prerequisite for the further activation of the alternative

pathway. Soluble or bound C3b has an affinity to factor B, with

which it forms, in the presence of Mg , a bimolecular complex

(Medicus et al, 1976a). This complexed factor В becomes susceptible

to factor D (Vogt et al, 1976) which cleaves factor В in two

fragments Bb (retained in the complex) and Ba (dissociated into

the fluid phase). Complex СЗЬВЬ is the alternative pathway C3

convertase (Müller-Eberhard and Götze, 1972). The formation of

this alternative C3 convertase is further enhanced by C3b (the

so-called amplification loop). The СЗЬВЬ enzyme has when present

in the fluid phase only C3 splitting properties (Medicus et al,

25

1976b) but in the solid phase it also has a slight C5 splitting

activity (Medicus et al, 1976a), which is greatly enhanced when

more than one molecule C3b is present in the СЗЬВЬ complex

(S2-C3b(n)Bb) (Medicus et al, 1976c). This is quite similar to the

classical pathway where the C3 convertase (C4b2a) obtains C5

splitting activity after incorporation of a C3b molecule (C4b2a3b).

Both C3b(n)Bb and СЗЬВЬ have properdin (P) activating activity.

After binding,? undergoes conformational changes by which acti

vated properdin (P) is formed (Götze et al, 1976). The presence

of Ρ stabilizes the C3b(n)Bb complex and prolongs its half-life

by inhibiting the spontaneous dissociation of Bb from the complex

(Fearon and Austen, 1975; Medicus et al, 1976c). Furthermore Ρ

indirectly makes the СЗЬВЬ complex less susceptible to C3b in-

activator (C3b.INA) because C3b.INA cannot cleave C3b into C3c

and C3d as long as C3b is complexed with Bb (Fearon et al, 1976).

In contrast to earlier ideas properdin itself does not play a

central role in the alternative pathway. It is activated rather

late in the sequence, and its sole function seems to be to

stabilize the С3b(n)Bb complex.

As can be seen in figure 4 there exists also another molecular

sequence in the alternative pathway. After formation of S2-C3b

activated fluid phase properdin (P) can bind to the C3b molecule.

In the presence of factor В and D this enzyme obtains C3 splitting

activity which subsequently leads to the formation of a Ρ stabi

lized C5 convertase. (Götze and Müller-Eberhard, 1974). The formed

СЗЬР is susceptible to C3b.INA which splits the complex in S2-

C3d and СЗсР. The latter complex can dissociate with recovery of

P. However, this sequence seems to be less important and there are

even serious doubts whether it exists under physiological condi

tions (Götze and Müller-Eberhard, 1977).

As already mentioned sera of some patients with membrano-

proliferative glomerulonephritis contain a protein called C3NeF.

The presence of C3NeF is accompanied by depressed serum concen

trations of C3, B, and P, but relatively normal concentrations CI,

C4 and C2 (Gewürz et al, 1968; McLean and Michael, 1973) . This

proves that only activation of the alternative pathway has taken

place. If C3NeF containing serum is added to normal serum C3

26

activation will occur. C3NeF has tvo functions, first it induces

the formation of a C3 convertase (Schreiber et al, 1975), second

it stabilizes the C3 convertase so that it becoires insusceptible

to the control proteins of the alternative pathway C3b.INA and

B1H globulin (Daha et al, 1976).

Recently it has been reported that C3NeF is an antibody pro

bably directed against alternative C3 convertase СЗЬВЬ (Davis et

al, 1977). However, this report is in contrast with the ideas of

the group of Müller-Eberhard, who think that C3NeF represents a

permanently active form of IF (Schreiber et al, 1976a, 1976b).

Further elucidation of the structure, function and reaction pro

ducts of IF is necessary for a better delineation of the role of

C3Ner m disease, (see addendum: page 45).

Finally e few remarks should be made regarding the control

systems in the alternative pathway. An important control mechanism

is the lability of the formed enzymes. By their rapid decay a

prolonged and sustained action is prevented. The intrinsic decay

of the СЗЬВЬ complex is the result of the rapid dissociation of

Bb out of the complex. There also exists an extrinsic decay due

to C3b.INA and B1H (fig. 5). C3b.INA was first recognized by

GENERATION STABILIZATION CONTROL

D ι

Сзь.в-^Сзь Bb- 1 -Bo

- W СЗЬ Ρ Bb -

Intrinsic decay

η—Ç*C3b-Βι

CîNeF ^ - » С з ь В ь C3NeFH

Ρ IH extrinsic decoy

СЗЬ INA inactivatran

-!*СЭс »СЗй

- ^ C 3 b P - > C 3 c P . C 3 d

Escape

Сзь Bb СЭМеГ

Figure S. Stabilization and control of alternatzve C3 aonvertase

(after Weiler et al, 1976).

27

Tamura and Nelson (1967), who found that it blocked C3b in immune

adherence and hemolysis. Later on it was found that СЗЬ.ІЫА

cleaved C3b into two fragments C3c and C3d (Ruddy and Austen,

1971). The action of СЗЬ.ІМЛ leads to four effects: destruction

of the C5 convertase activity of C4b2a3b, inhibition of binding

of factor В to C3b, inhibition of immune adherence, and of phago

cytosis. It is the major control protein of the amplification

loop, which can be derived from the observation that a patient

with a genetic deficiency of C3b.lNA had very low levels of C3

and factor В (Alper et al, 1972a). As already mentioned C3b

cannot be cleaved by СЗЬ.ІМА if it is complexed with Bb. The in

trinsic decay of СЗЬВЬ is inhibited by P. The stabilizing effect

of Ρ is antagonized by the second control protein BiH (Weiler et

al, 1976), whereafter Bb can dissociate again out of СЗЬВЬ.

Subsequently C3b can be cleaved by C3b.INA (fig. 5). If the com

plex is stabilized by C3NeF it is much more (approximately 200

fold) resistant to BiH and thus escapes from the control mecha

nisms. This results in sustained C3 cleavage (Daha et al, 1976).

2.3.3. The terminal attack unit (fig. 6)

This phase of the complement sequence is less well understood

than the preceding steps. After formation of the classical or al

ternative C3 convertase, C3 is cleaved into a major fragment C3b

which can be incorporated into the C3 convertase, and a minor

fragment C3a which evokes important immunobiological effects that

will be discussed in the following section. After incorporation

of C3b into the C3 convertase the specificity of the enzyme is

changed, such that it now has C5 splitting activity (Shin et al,

1971; Daha et al, 1976). C5 is cleaved into a major fragment C5b

which interacts physicochemically with C6 and C7 and a minor

fragment C5a that has a variety of immunobiological properties

that will be discussed later. The binding of C6 to C5b must occur

immediately after the generation of C5b, otherwise C5b will decay

into cytolytically inactive C5i (Nilsson and Müller-Eberhard,

1967a). Binding of C6 and later on also C7 can occur both on the

28

classical CS altematn* Cs convertase convertase

С4Ь 2o ЗЬ Cs Сзь п РВь 1 >4 1

^ C 5 a

CSb—»CSI

^ С б С 7

C 5 b 6 7 — » С 5 І 6 7

^ - С

С5Ь-в

^ 6 С 9

• Csb^g

cell lysis

Figure 6. Activation sequence of the terminal attack unit.

membrane and in the fluid phase. The formation of the trinolecular

complex C5b67 is non-enzymatic, v/hich implies that no cleavage

occurs and that its formation is the result of non-covalent assem

bly. The C5b67 complex binds to membranes via its C5b part. If the

fluid phase C5b67 does not encounter a membrane, the binding site

decays and cytolytically inactive С5І67 is formed. The fluid phase

C5b67 complex can also bind via C5b to membranes not sensitized

with antibody, thereby extending the cytolytic effects of comple

ment to "innocent bystander" cells. This mechanism is known as

"reactive lysis" (Thompson and Lachmann, 1970). After formation

of the trimolecular complex C5b67 one molecule C8 becomes attached

whereafter six C9 molecules can bind to C8 to form a decamolecular

complex (Kolb and Müller-Eberhard, 1973). This complex causes

cytolysis by an impairment of osmotic regulation of the cell.

2.4. IMMUNOBIOLOGY OF COMPLEMENT

The biological activities that are generated during the comple

ment cascade are mainly a consequence of activation of C3 and the

terminal attack sequence. These activities can be divided into

three types :

29

1. alteration of the target complex (e.g. immune adherence, cyto-

lysis)

2. alteration of the nicro-environment (anaphylaxis, leucotaxis)

3. systemic effects (interaction with coagulation system etc.).

2.4.1. Alteration of the target complex

Immune adherence

One of the most important biological activities generated

during complement activation is the immune adherence phenomenon.

Immune adherence is the binding of C3b coated immune complexes,

cells, or particles to cells bearing a C3b receptor on their sur

face. Such C3b receptors are present on human polymorphonuclear

leukocytes, macrophages, erythrocytes, В lymphocytes, monocytes

and renal glomerular basement membrane. Presumably an intact C3b

molecule is required, because it was shown that the phenomenon

is inhibited by C3b.INA. (Tamura and Nelson, 1967; Gigli and

Nelson, 1968). Bound C4b shows also an affinity to the same re

ceptors (Cooper, 1969). Binding to the membrane receptor acts as

a signal to start certain cellular processes, and the resulting

biological effects depend on the cell type involved. Immune ad

herence to phagocytic cells results in an enhancement of phago

cytosis (Gigli and Nelson, 1968). The biological importance of

this event can be derived from the observation that in patients

who are genetically deficient in C3 repeated and severe bacterial

infections can occur (Alper and Hosen, 1974). If immune adherence

with non-phagocytosable particles takes place extracellular

release of lysozymes occurs, which can contribute to inflammatory

damage (glomerulonephritis, graft-rejection) (Henson, 1972).

Immune adherence to В lymphocytes seems to be important for antibody

production, because C3 depletion leads to a delayed antibody pro

duction towards thyrrus-dependent antigens but not to thymus-

independent antigens (Pepys, 1974). It seems therefore very likely

that C3 and C3 receptors play a role in the T-B cell cooperation.

There exists also evidence indicating that C3 plays a role in

В cell proliferation. It has been even suggested that the Τ cell

30

independency of certain antigens is caused by the fact that they

can initiate the alternative pathway leading to the formation of

C3b on the surface of these antigens. This can result in trigger

ing of the В cell without Τ cell help (Dukor et al, 1974) . How

ever the fact that in a patient with a genetic deficiency of C3

a normal response to antigens was found, strongly suggests that

C3 is not an absolute requirement for antibody production (Alper

et al, 1972b). Immune adherence to macrophages activates the

macrophage and induces the release of several lysosomal enzymes,

like hydrolases and proteinases. One of these enzymes has C3

splitting activity (Schorlemmer and Allison, 1976) . Since macro

phages synthesize factor C3 (Lai A Fat and van Fürth, 1975) and

factor В (Bentley et al, 1976), and activated macrophages even

release C3b (Allison, 1978), СЗЬВЬ convertase can be formed in

the vicinity of the macrophage. This convertase enhances further

C3b generation which can lead to subsequent activation of other

macrophages, so that serial activation can take place. These in

vitro findings have probably immunopathological significance in

the production of chronic inflammatory responses evoked by bac

terial or fungal cell walls, or other polyanions. They can induce

prolonged macrophage activation, by generation of C3b via the

alternative pathway (Schorlemmer et al, 1977). Furthermore C3b

activated macrophages produce C3a (Ferluga et al, 1978) which has

cytolytic properties against tumor cells (Ferluga et al, 1976).

Immune adherence of C3b to human monocytes gives a marked increase

in the generation of tissue thromboplastin by the monocyte

(Prydz et al, 1977). C3b receptors have not been found on human

platelets, but in experimental animal studies immune adherence

to platelets can lead to secretion of coagulation active and vaso

active intracellular products (Henson and Cochrane, 1971) and to

aggregation of platelets.

Cytolysis

Generation of the membrane attack unit with formation of the

multimolecular C5-C9 complex on the membrane induces cytolysis

31

of cells and gram negative bacteria. Kinetic studies revealed

that cytolysis is a one hit phenomenon, or in other words that

one C5-C9 complex on a membrane is sufficient for cytolysis

(Kitamura et al, 1976). This makes it very unlikely that cyto

lysis is caused by enzymatic attack, because this process is a

multi-hit process. Work by the group of Mayer (Mayer, 1972; 1977;

Hammer et al, 1975; 1977) indicates that the lesion produced is

a transmembrane channe], which is formed by a hydrophobic outer

ring and an hydrophilic inner core. During the formation of the

C5-C9 complex conformational changes of the C8 and C9 molecule

lead to the exposure of hydrophobic polypeptide regions from the

interior of the molecule. If this occurs in the immediate vici

nity of the lipid bilayer of the membrane, insertion of the mole

cule will occur with formation of the transmembrane channel.

Through this transmembrane channel salt and water will flow into

the cell because of the Donnan effect. This causes swelling which

leads to bursting of the cell membrane and release of macro-

molecules .

Virus-neutralisation

Under certain conditions antisera against viruses have little

or no ability to neutralize viruses, unless CI, C4, C2 and C3 are

present, presumably to cover the surface of the virus (Daniels

et al, 1970) .

Solubilization of immune complexes

Immune complexes can interact with the complement system at

several levels. Clq has 6 binding sites for immunoglobulins, while

C4b and C3b can bind to immune complexes.

Recently, Miller and Nussenzweig (1975) described a new function

of the complement system that consists of solubilization of pre

cipitated Ag-Ab complexes. This solubilization is dependent on

C3, which is the effector molecule in this release reaction

(Takahashi et al, 1977). The complex release activity (CRA) is

generated via the alternative pathway, and this can occur in the

32

absence of the classical activation unit (Miller and Nussenzweig,

1975). However, the presence of C2, C4 and Ca leads to a more

efficient release (Czop and Nussenzweig, 1976) . The late compo

nents (C5-C9) play no role in CRA since it was found unaltered

with C5 or C6 deficient sera. The reaction leading to CRA is

probably started by binding of Clq to the Ag-Ab complex, and sub

sequently some C3b molecules are deposited on the immune complexes

via the classical C3 convertase. Via the amplification loop C3b

dependent alternative C3 convertase is formed on the immune com

plexes which greatly enhances further deposition of C3b. The C3b

molecules somehow interact with the complexes and lead to dis

ruption of the lattice formation. After dissociation of antigen

from antibody the C3 molecules prevent reassociation and cross-

linking with antigen (Takahashi et al, 1976). Although this re

lease activity was studied in vitro, it has probably relevance

in vivo, where by this mechanism complement can prevent deposi

tion of immune complexes in tissues. It could explain the finding

that in patients with a congenital deficiency of the complement

system, immune complex mediated diseases are frequently found

(see 2.5.1.) .

2.4.2. Alteration of the micro-environment

Alterations of the micro-environment may be brought about

primarily by complement-mediators formed during complement acti

vation and released in the fluid phase (C3a, C5a). They increase

vascular permeability, contract smooth muscle and attract phago

cytic cells. Alternatively, alterations are induced by secondary

mechanisms, such as release of lysozymes from neutrophils after

immune-adherence with non-phagocytosable particles or surfaces,

or exposure of platelets to vascular basement membrane after

complement mediated damage of endothelial cells. Both primary and

secondary mediators are important in inflammatory processes

associated with immune reactions.

33

Permeability factors

The most important factors that lead to an increased vascular

permeability, generated during complement activation are the split

products of C3 and C5: C3a and C5a. These effects are mediated by

histamine liberated from mast-cells and basophils. Moreover, these

factors have smooth-muscle contractile activity, which is inde

pendent of histamine release (Vallota and Müller-Eberhard, 1973).

The effects of C3a and C5a on smooth muscle and vascular permea

bility promote sludging and margination of leukocytes, thereby

producing optimal conditions for Chemotaxis, which is an addition

al activity of C3a and C5a. The vasoactive properties of C3a and

C5a are susceptible to rapid enzymatic inactivation by anaphyla-

toxin inactivator (A.I.) which cleaves arginine (СЗа), or arginine

or lysine (C5a) from the molecule (Bökisch and Müller-Eberhard,

1970) .

As already mentioned, it has been shown in animal experimental

studies that vasoactive peptides can be released by thrombocytes

after binding of C3b to their C3b receptor.

In sera from patients with heriditary angioneurotic edema

(HANE), which is caused by an autosomal dominant deficiency of

Cl.INH, a complement factor is found which is called "C-kinin".

This protein has vasoactive properties, and is responsible for the

clinical manifestations in HANE patients (Donaldson et al, 1969).

During an attack pathologically increased CI activity can be found

in sera from these patients, accompanied by very low levels of

C4 and C2. (Lepow, 1971). The high cT activity is due to the lack

of its inhibitor Cl.INH, which leads to an increased cleavage of

the substrates of cT i.e. C4 and C2. Intradermal injection of cT

causes a local increase in vascular permeability in normal indi

viduals, an extensive area of angioedema in HANE patients, while

no reaction is seen in C2 deficient patients (Klemperer et al,

1968) . A peptide having the same vasoactive properties has been

isolated from incubation mixtures of CI, C4 and C2. It is very

likely that the polypeptide, (called "C-kinin") responsible for

the clinical features of HANE-patients, is liberated from C2 by

the action of cTs in the presence of C4.

34

Chemotaatia factors

Three products generated during the complement sequence attract

leucocytes by their chemotactic properties: C3a, C5a and probably

fluid phase C5b67 (Ward et al, 1965; Ward, 1972). The most potent

among these three is C5a, C3a has a much lower activity. Although

chemotactic activity of C5b67 has been described in in vitro ex

periments, serious doubts exist whether this effect is of signi

ficance in vivo. Thus, C6 or C7 deficient sera have unimpaired

chemotactic properties (Snijderman et al, 1969; Stecher and Sorkin,

1969; Hannema et al, 1975; Lim et al, 1976). The chemotactic

factors C3a and C5a are not only formed during complement acti

vation but are also liberated through the action of plasmin, tryp

sin and tissue- and neutrophil proteinases (Ward, 1972). Chemo

tactic factors do not only attract neutrophils but also eosino

phils and monocytes. After exposure to chemotactic molecules,

the cell becomes unresponsive to further attraction. This keeps

the cell at the inflammatory focus, so that other cell functions

can come to expression. One of the properties of neutrophils is

to release a proteinase that can form C5a, so that vascular

permeability will increase even more and further accumulation of

leucocytes and monocytes will occur (Snijderman et al, 1971a).

The chemotactic properties are abolished by a serum factor called

chemotactic-factor inactivator. This inactivator is distinct

from anaphylatoxin inactivator, that abolishes the vasoactive

activities of C3a and C5a (Till and Ward, 1975) .

2.4.3. Systemic effects of complement

МоЪгЪъгаЬгоп of leucocytes

During cleavage of C3 a factor is formed (called leucocyte

mobilizing factor) which mobilates leucocytes from the bone-

marrow reserves into the circulation (Rother, 1972). The relevance

of this factor in vivo can be derived from the fact that in pa

tients with a genetic deficiency of C3, leucocytosis does not

occur during bacterial infections (Alper and Rosen, 1974).

35

Inter-relations between the complement-, the alotting-3 the кгпгп

generating-, and the fibrinolytic system

Activation of the clotting system can directly lead to comple

ment activation. The mechanisms involved will be discussed in this

section. Although direct activation in the onposite direction does

not exist, certain effects produced by complement activation

(release of platelet factors due to immune adherence, lysis of

platelets, release of coagulatory enzymes from leucocytes, etc.)

can secondarily induce activation of the coagulation sequence.

A central place in the relation between these enzymatic cascades

is the Hageman factor (F XII), as can be seen in figure 7.

FURTHER ACTIVATION CLOTTING SEQUENCE m

ACTIVATION KININ SEQUENCE

S a -

XI

/ activation

Activated Hageman

(actor Ш a

plasminogen proactivator

^ plasminogen activator

plasminogen •

• positive feed back inhibition by CÎ.INH

plasmin

prekallikrem

* СЗЬ . C3a

Figure 7. Inter-relations between the clotting-, kinin generating-,

fibrinolytic-, and complement system.

36

Activation of Hageman factor leads to four different functions

(Kaplan and Austen, 1975):

a) Initiation of the intrinsic pathway of coagulation, by acti

vating factor XI, which induces further activation of the

clotting sequence.

b) Initiation of fibrinolysis by formation of plasmin, the fibri

nolytic enzyme. Activated factor XII activates plasminogen-

proactivator to plasminogen activator which converts plasmino

gen to plasmin. Plasmin cleaves fibrin and fibrinogen into

smaller fragments.

c) Activation of the kinin-system, by liberation of kallikrein

from prekallikrein. The formed kallikrein cleaves kinins from

kininogen. The formation of kallikrein by activated factor XII

is enhanced by plasmin, which splits Xlla into smaller frag

ments (Xll-f), that have a much greater capacity than Xlla to

form kallikrein out of prekallikrein. Furthermore plasmin can

directly, without F XII, convert prekallikrein to kallikrein.

d) Activation of the complement system. Activated Hageman factor

can also via plasmin, trigger the complement sequence at two

different levels. First plasmin has the capacity of cleaving

CI into its active form cT; second it can split C3 in the fluid

phase into its biologically active fragments C3a and C3b.

The membrane attack unit is not activated by Hageman-factor.

Activation of Hageman factor can occur by contact with bio

logical materials presenting dense negative charges. These inter

act with positively charged aminoacids in the molecule (e.g.

arginine) leading to a rearrangement into the molecule, with

uncovering of the enzymatic site. Collagen, vascular basement

membrane, uric acid, pyrophosphate and a number of insoluble

materials (glass, kaolin, etc.) belong to these activating sub

stances. It has been found that antigen-antibody complexes and

immune aggregates can activate F XII after binding of Clq (which

exhibits a chemical composition similar to that of collagen)

(Austen, 1974). Furthermore, complement-mediated immunologic tissue

injury is very often primarily localized at the surface of blood

vessels. The alterations of the vascular basement membrane induced

37

by this injury can result in activation of Hagenan factor. There

is however as yet no experimental evidence that immune complexes

are able to activate F XII directly.

At several levels inhibition by Cl.INH can occur (see fig. 7),

so that all effects of activated factor XII can become blocked.

Deficiency of Cl.INH, as in patients with heriditary angioneurotic

edema, leads to unappropriate activation of factor XII, resulting

in kinin generation and complement activation, which are res

ponsible for the clinical symptoms.

Activation of factor XII leads to a number of biological

activities. Plasminogen activator and kallikrein both have chemo-

tactic properties (Kaplan et al, 1972; 1973), cleavage products

of fibrinogen and fibrin have also chemotactic properties, and

can enhance vascular permeability (Marder, 1971; Barnhart et al,

1971), while the formed kinins are potent vasoactive peptides.

So far we have presented some data about the influence of the

coagulation system on complement. There exists very little evi

dence that there are also direct interactions in the opposite

direction. Deficiencies of virtually all classical complement

pathway factors have been reported in man, but defects in coa

gulation were not mentioned in these reports (Brown, 1975).

However, in rabbits congenital deficient in C6 (Zimmerman et al,

1971), and in guinea pigs deficient in C4 (Dodds et al, 1977) a

decrease in coagulation activity was found that normalized after

reconstitution of the deficient factor. The fact that coagulation

is normal in humans with C6 deficiency (Heusinkveld et al, 1974) ,

while C6 deficient rabbits have a clotting defect is probably

the result of a peculiar property of the rabbit platelet, in that

it requires all complement factors for the release of initiating

clotting factors (Rosen, 1975), and for aggregation (Fong and

Good, 1975). As already mentioned earlier, in many experimental

animals (including the mouse), but not in man, an indirect rela

tion exists between complement and the coagulation system via the

platelets, because of their C3b receptors. Binding of C3b to these

receptors leads to platelet aggregation with subsequent release

of platelet procoagulants (Henson and Cochrane, 1971). If the

complement cascade is further activated up to C9 on the surface

38

of the platelet, platelet lysis does even occur. This is accom

panied by a substantial increase in the release of platelet

factors (Henson, 1970) . However, even this link between comple

ment and coagulation is incomplete, because it does not lead to

a direct triggering of the coagulation system. It only can

accelerate an already initiated clotting sequence. Probably the

most important interaction between complement and coagulation

takes place on the endothelium of small vessels via a number of

biological activities generated during complement activation.

Generation of C3a and C5a causes increased vascular permeability

by releasing vasoactive amines from basophils and mast cells,

and by inducing release of serotonin from platelets via platelet

aggregation factor liberated fron basophils (Benveniste, 1974).

Neutrophils and monocytes marginate at the sites of altered vas

cular permeability, leading to endothelial cell damage (Stewart

et al, 1974), release of lysosomal enzymes (Janoff, 1970), and

release of a leucocyte procoagulant which can initiate coagula

tion (Kociba et al, 1972; Prydz et al, 1977). Local coagulation

is further enhanced by platelet aggregation on exposed subendo-

thelial surfaces (Baumgartner, 1974), and by Hageman factor

activation by the altered vascular basement membrane.

2.5. GENETICS OF THE COMPLEMENT SYSTEM

2.5.1. Complement deficiency in man

Genetic defects have now been described for almost all the

complement components in man (see table II). In most cases the

defects are transmitted as autosomal recessive traits. Some of

these deficiencies predispose to certain diseases, which are

listed in table II. As mentioned previously deficiency of Cl.INH

is associated with heriditary angioneurotic edema. (Donaldson and

Evans, 1963). The disease consists of recurrent attacks of non

inflammatory swelling localized in the subcutaneous tissues,

intestinal wall, airways and lungs. Laryngeal edema may often be

fatal (Donaldson and Rosen, 1966) . The deficiency is most pro

bably transmitted as an autosomal dominant trait. Attacks are

39

о TABLE il. Deficienaies of the complement system in man*

Deficient factor Clinical findings

Clq

Clr

Cls

C4

C2

C3

C5

C5 dysfunction

C6

C7

С

combined immunodeficiency state

SLE, glomerulonephritis, reumatoid disease

SLE

SLE

none, discoid LE, SLE, anaphylactoid purpura, derraa-

tomyositis, glomerulonephritis

infections

SLE

Leiner's syndrome (eczema and secondary infections)

none, infections

none, Raynaud

gonococcal infection, SLE

Deficient control protein

Cl.INH

С3b.IMA

heriditary angioneurotic edema, SLE

infections

* Based on Day and Good, 1976; Alper and Rosen, 1976, Agnello, 1978

evoked by an unappropriate activation of Cl, possibly via plasmin

generated after activation of Hageman factor. One patient has

been described with a deficiency of another control protein,

C3b.INA (Alper et al, 1972a). In the serum of this patient very

low C3 levels have been found. Clinically the deficiency is accom

panied by a life long history of severe infections.

2.5.2. Complement deficiency in animals

In animals also a number of complement deficiencies (C4, C5

and C6) have been found. Because we made use of these animals in

our experiments (see chapter 5 and 7) the immunobiological effects

that are found in association with these complement deficiencies

will be described here in more detail.

C4 deficienoy in guinea pigs

In 1970 Ellman et al described a strain of guinea pigs that

lacked C4. This deficiency is transmitted as an autosomal reces

sive trait (Ellman et al, 1970). The serum of these animals lacks

hemolytic complement activity, which can be restored by the

addition of normal guinea pig serum or of purified guinea pig C4.

Serum concentrations of the other complement components in C4-

deficient guinea pigs are normal except for C2 (50 percent of

normal value). The alternative pathway of complement is entirely

normal (Frank et al, 1971). A number of immunological mechanisms

have been studied in these animals (Ellman et al, 1971). Passive

cutaneous anaphylaxis, contact and delayed hypersensitivity, and

the cellular exsudative response to foreign body were normal.

Somewhat striking was the finding that the direct and reverse

passive Arthus reaction are unimpaired, while in C6 deficiency

(Biro, 1966) and in experimental C3 depletion (Cochrane et al,

1970) these reactions are diminished. This difference is the

result of the fact that guinea pig antibodies are able to acti

vate the alternative pathway, so that via this bypass mechanism

the late acting complement components can become activated (Sand-

berg and Osier, 1971). C4 deficient guinea pigs show furthermore

41

a decrease in the clearance of autologous erythrocytes coated

with antibodies. Again this can be restored by the addition of

normal guinea pig serum. Clearance of bacteria v/hich is dependent

of alternative pathway activation is normal in C4 deficient guinea

pigs. Induction of Chemotaxis in C4 deficient guinea pigs is

normal, but the kinetics of the production of chemotactic factors

shows a latency period which is characteristic for alternative

pathway activation (Clarck et al, 1973).

CS deficiency in mice

Genetically defined complement deficiency in mice was found in

many inbred mouse strains. The locus coding for the presence or

absence of a β globulin was originally designated He' (Erickson

et al, 1964) or MuBl (Cinader et al, 1964) . Later on it was found

that the product of this locus was C5 (Nilsson and Müller-Eberhard,

1967b). This deficiency is transmitted as an autosomal recessive

trait. The serum of these mice completely lacks hemolytic comple

ment activity. In general C5 deficient mice are as healthy as

normal mice, although there are some immunobiological deficiencies.

C5 deficient mice are more susceptible to infections with Coryne-

bacterium Kutscher! (Caren and Rosenberg, 1966). The clearance of

erythrocytes coated with antibodies and bacteria (S. typhimurium

and E.coli) was equal to normal mice (Stiffel et al, 1964) .

The finding that Chemotaxis for leucocytes was not produced by

C5 deficient serum after incubation with immune complexes or

zymosan, has lead to the identification of the central function

of C5a in Chemotaxis (Ward et al, 1965). This defect was also

found to exist in vivo. Injection of endotoxin into the peritoneal

cavity of C5 normal mice resulted in the generation of a factor

that was chemotactic for mouse polymorphonuclear leucocytes and

was followed by the accumulation of these cells in the peritoneal

fluid. By contrast the same procedure carried out in C5 deficient

mice did not reveal chemotactic activity tested in vitro nor

accumulation of polymorphonuclear leucocytes (Snijderman et al,

1971b). There exist two brief reports in the literature concerning

the Arthus phenomenon in C5 deficient mice, and it was found to

42

be unimpaired in comparison to normal mice (Linscott and Cochrane,

1964; Crisler and Frank, 1965). With immunofluorescence studies

C3 was found in Arthus lesions in C5 deficient mice (Linscott and

Cochrane, 1964). From these findings it might be concluded that

activation up till C3 is sufficient to induce the Arthusphenomenon,

however in C6 deficient rabbits an impaired Arthus reaction has

been found. Since nephrotoxic serum nephritis (Lindberg and Rosen

berg, 1968; Unanue et al, 1967) and autoimmune disease occur in

C5 deficient mice (Howie and Helyer, 1965) it is likely that these

immunologic inflammatory reactions can occur without activation

of the terminal complement attack sequence. Cellular immune res

ponses are normal in C5 deficient mice since normal delayed hyper

sensitivity reactions and skin allograft rejection have been ob

served (Caren and Rosenberg, 1965; Crisler and Frank, 1965) .

Св deficiency in rabbits

In 1961 Rother discovered an inborn defect in the complement

system in certain rabbits, which turned out later to be caused by

a deficiency of C6 (Rother and Rother, 1961; Rother et al, 1966a).

The characteristic is inherited as a single autosomal recessive

trait. The serum of these animals lacks hemolytic complement and

bactericidal activity which are both restored after addition of

C6 (Rubin et al, 1967; Rother et al, 1964a). Passive Arthus

reaction in C6 deficient rabbits was macroscopically absent or

incomplete (Rother et al, 1964b). However, on histologic examina

tion, it was found that the leucocyte infiltration was only slight

ly less than in normal rabbits. The most striking difference was

the complete absence of hemorrhagic necrosis (Biro, 1966), that

is considered to be a hallmark of the Arthus reaction. As already

mentioned these observations are in apparent contrast to those of

unimpaired Arthus reactions in C5 deficient mice (Linscott and

Cochrane, 1964; Crisler and Frank, 1965). However, these latter

studies only mention leucocyte infiltration and fail to give in

formation on the presence of hemorrhagic necrosis. If one assumes

that necrosis was also absent in the C5 deficient mice, the fin

dings can be explained by the hypothesis that, for leucocyte

43

accumulations, activation up to C3 suffices, while activation of

the late complement components is necessary for the development

of hemorrhagic necrosis. Clearance of bacteria, and phagocytosis

of antibody sensitized erythrocytes was unimpaired in C6 deficient

rabbits (Rother and Rother, 1965; Rother et al, 1966a), which

evidently shows that complement-dependent functions up to C5 are

normal. This is also supported by the finding that anaphylatoxin

generation and histamine release were the same for normal and C6

deficient rabbit serum (Giertz et al, 1964) . Nephrotoxic serum

nephritis could be induced in C6 deficient rabbits and had the

same characteristics as in normal rabbits (Rother et al, 1966b;

1967). Reports regarding the cellular immune reactivity in C6

deficient rabbits are contradictory. Delayed hypersensitivity was

found to be absent in one study (Volk et al, 1964), whereas it

was completely normal in another (Biro, 1966). Equally contro

versial findings have been reported on allograft rejection, that

was impaired in a study by Volk et al, 1964 but found to be normal

in another study (Biro, 1966) . As already mentioned slight dis

turbances in the clotting system of C6 deficient rabbits have been

found.

2.5.3. Relation between complement and the major histocompatibili

ty complex

In man the inheritance of a number of isolated complement

deficiencies is linked to the HLA-system (reviewed by Jersild et

al, 1976; and Rittner, 1976). This linkage has been shown for C2,

C4 and probably C8 deficiency. C2 deficiency which is the most

frequently occurring complement deficiency, even shows linkage

with two particular haplotypes (AW25, B19, DW2 and AIO, B18, DW2).

A number of other complement deficiencies (Clr, cT.INH, C6 and

probably C5) does not segregate with HLA. Furthermore the electro-

phoretic polymorphism that has been found for factor В is linked

to HLA. By contrast a comparable polymorphism of factor C3, shows

no linkage to HLA.

Also in the mouse a correlation has been found between H-2

haplotype and hemolytic complement activity (Demant et al, 1973).

44

The genes controlling complement activity are mapped in the S-

region of the H-2 complex. Initially it was found that some mouse

strains had high concentrations of a certain serum protein (called

Ss-protein) and others showed consistently lower levels. This

characteristic was controlled by the S-region in the H-2 complex.

In 1975 Lachmann et al demonstrated that the product of this S-

region was complement factor C4. Furthermore, there is evidence

that at least one gene within the H-2 complex influences C3

levels, but also one or more genes are involved that are not

coded for by H-2 (Ferreira and Nussenzweig, 1975). Recently it

was shown that not only variation in levels of C4 but also CI and

C2 levels are governed by the S-region, since mice with the Ss

high genotype have higher functional levels of CI, C4 and C2 than

strains of the Ss low genotype (Goldman and Goldnan, 1976). The

only known complement deficiency in mice, C5, is coded for by a

single gene which is not located within the H-2 complex (Herzen

berg et al, 1963). A further relationship between H-2 and comple

ment is found in its influence on C3b receptors on lymphocytes.

Complement can trigger a number of cellular responses via the C3b

receptor present on various cells. The density of these C3b

receptors on the lymphocyte membrane is controlled by two genes,

and one of these is linked to H-2 (Gelfand et al, 1974).

In conclusion we can say that there is growing evidence that

especially the early components of complement both in man and in

mice are controlled and influenced by genes of the major histo

compatibility complex (MHC). Within the MHC also genes (SD and

la) are located that code for several responder and stimulator

characteristics of the immune response. Thus, there seems to be

a functional relationship between immune recognition, mediation

and response which is governed by a set of closely linked genes

on the chromosomal region of the MHC.

Addendum (to page 27): Recently it was convincingly demonstrated

that C3NeF is an autoantibody against СЗЬВЬ (Daha, M.R., Austen,

K.F., and Fearon, D.T., J. Immunol. (1978) 120, 1389).

45

REFERENCES

Agnello, V. (1978): Conplement deficiency states.

Medicine (Baltirpore) 57, 1.

Allison, A.C. (1978): flechanisir.s by v.'hich activated macrophages

inhibit lymphocyte responses.

Immunol. Rev. 40, 3.

Alper, C.A., Чозеп, F.S., and Lachman, P.J. (1972a): Inactivator

of the third component of complenent as an inhibitor in the

properdin pathway.

Proc. Natl. Acad. Sci. USA 69, 2910.

Alper, C.A., Colten, H.R., Rosen, F.S., Rabson, A.S., Macnab,

G.M., and Gear, J.S.S. (1972b): Homozygous deficiency of C3 in a

patient with repeated infections.

Lancet 2, 1179.

Alper, C.A., and Rosen, F.S. (1974): The role of complement in

vivo as revealed by genetic defects.

In: Progress in Immunology II

Eds.: Brent, L. and Kolborov, J.

Elsevier, Amsterdam, vol. 1, p. 201.

Alper, C.A., and Rosen, F.S. (1976): Genetics of the complement

system.

Adv. Hum. Genet. 7, 141.

Assimeh, S.N., and Painter, R.H. (1975): The identification of a

previously unrecognized subcomponent of the first component of

complement.

J. Immunol. 115, 482.

46

Austen, K.F. (1974): Hageman-factor dependent coagulation» fibri

nolysis, and kinin generation.


Barnhart, M.I., Sulisz, L.,and Bluhm, G.B. (1971): Role for

fibrinogen and its dérivâtes in acute inflammation.

In: Immunopathology of inflammation.

Eds.: Forscher, В.К., and Houck, J.C.

Excerpta Iledica Amsterdam, p. 59.

Baumgartner, H.R. (1974): Morphometric quantitation of adherence

of platelets to an artificial surface and components of connective

tissue.

Thromb. Diath. Haem. supnl. 60, 39.

Bentley, С , Bitter-Suermann, D., Hadding, U., and Brade, V.

(1976): In vitro synthesis of factor В of the alternative path

way of complement activation by mouse peritoneal macrophages.

Eur. J. Immunol. 6, 39 3.

Benveniste, J. (1974): Platelet activating factor.

Nature 249, 581.

Biro, C.E. (1966): The role of the 6th component of complement

in some types of hypersensitivity.

Immunology 10, 56 3.

Bökisch, V.A., and Müller-Eberhard, H.J. (1970): Anaphylatoxin

inactivator of human plasma; its isolation and characterization

as a carboxypeptidase.

J. Clin. Invest. ÍS, 2427.

Borsos, T., and Rapp, H.J. (1965): Complement fixation of cell

surfaces by 19S and 7S antibodies.

Science 150, 505.

Brown, D.L. (1975): Complement and coagulation.

Brit. J. Haem. 30, 377.

47

Caren, L.D., and Rosenberg, L.T. (1965): Complenent m sUn-

grafting in mice.

Immunology 9, 3 59.

Caren, L.D., and Rosenberg, L.T. (1966): The role of complement

in resistance to endogenous and exogenous infection v/ith a common

mouse pathogen, Corynebactenum Kutschen.

J. Exp. Med. 124, 689.

Cinader, В., Dubiski, S., and Wardlaw, A.C. (1964):

Distribution, inheritance, and properties of an antigen MuBl,

and its relation to hemolytic complement.

J. Exp. Med. 120, 897.

Clark, R.A., Frank, M.M., and Kimball, H.P. (1973): Generation of

chemotactic factors in guinea pig serum via the activation of the

classical and alternate complement pathways.

Clin. Immunol. Immunopath. 1, 414.

Cochrane, C.G., Müller-Eberhard, H.J., and Aiken, B.S. (1970):

Depletion of plasma complement in vivo by a protein of cobra

venom: its effect on various immunologic reactions.

J. Immunol. 10 5, 55.

Cooper, N.P. (1969): Inmune adherence by the fourth component of

complement.

Science 165, 396.

Cooper, N.R. (1971): Enzymes of the complement system.

In: Progress in Immunology.

Ed.: D.В. Amos

Academic Press New York, p. 567.

Cooper, N.R. (1975): Isolation and analysis of the mechanism of

action of an mactivator of C4b m normal human serum.

J. Exp. Med. 141, 890.

Cnsler, C , and Trank, M.M. (1965): Skin graft rejection and the

Arthus phenomenon m mice deficient m the third component of

48

complement.

Fed. Proc. 24, 447.

Czop, J., and Nussenzweig, V. (1976): Studies on the mechanism

of solubilization of immune precipitates by serum.

J. Exp. Med. 14 2, 615.

Daha, M.R., Fearon, D.T., and Austen, K.F. (1976): C3 nephritic

factor: stabilization of fluid-phase and cell-bound alternative

pathway convertase.

J. Immunol. 116, 1.

Daniels, CA., Borsos, T., Uapp, H.J. (1970): Neutralization of

sensitized virus by purified components of complement.

Proc. Natl. Acad. Sci. 65, 528.

Davis, Α., Ziegler, J.В., Gelfand, E.W., Rosen, F.S., and Лірег,

CA. (1977): Heterogeneity of nephritic factor and its identifi

cation as an immunoglobulin.

Proc. Natl. Acad. Sci. U.S.A. 74, 3980.

Day, N.K., and Good, R.A. (1975): Deficiencies of the complement

system in man.

Birth Defects 9, 306.

Dêmant, P., Capkova, J., HinzovS, E., and Voràcova, В. (1973):

The role of H-2 linked Ss-Slp region in the control of mouse

complement.


Dodds, W.J., Raymond, S.L., Moynihan, A.C., Pickering, R.J., and

Amiraian, K. (1977): Coagulation studies in C4-deficient guinea

pigs.

Proc. Soc. Exp. Biol. Med. 154, 382.

Donaldson, V.H., and Evans, R.R. (1963): A biochemical abnorma

lity in heriditary angioneurotic edema: absence of serum in

hibition of CI esterase.

Am. J. Med. 35, 37.

49

Donaldson, V.H., and Rosen, Г.S. (1966): Heriditary angioneurotic

edema. Л clinical survey.

Pediatrics 27, 1017.

Donaldson, V.H., Ratnoff, 0., Dias da Silva, W., and Rosen, F.S.

(1969): Permeability increasing activity in HANE-plasna.

J. Clin. Invest. 48, 642.

Dukor, P., Schumann, G., Gisler, R.H., Dierich, M., König, W.,

Hadding, U., and Bitter-Suermann, D. (1974): Complement dependent

В cell activation by cobra venom factor and other mitogens.

J. Exp. Med. 1Z9, 337.

Ellerson, J.R., Yasmean, D., Painter, R.H., and Domington, K.J.

(1972): Л fragment corresponding to the CH2 region of IgC with

complement fixing activity.

FEBS Lett. 24, 319.

Ellman, L., Green, I., and Frank, M. (1970): Genetically control

led total deficiency of the fourth component of complement in the

guinea pig.

Science (Washington) 170, 74.

Ellman, L., Green, I., Judge, F., and Frank, Ч.М. (1971): In vivo

studies in C4 deficient guinea pigs.

J. Exp. Med. 134, 162.

Erickson, R.P., Tachibana, D.K., Herzenberg, L.A., and Rosenberg,

L.T. (1964): Л single gene controlling hemolytic complement and

a serum antigen in the mouse.

J. Immuno1. 9 2, 611.

Fearon, D.T., and Austen, K.F. (1975): Properdin: binding to C3b

and stabilization of the C3b-dependent convertase.

J. Exp. Med. 142, 856.

Fearon, D.T., and Austen, К.Г. (1976): The human complement

system: Biochemistry, biology and pathobiology.

Essays Med. Blochen. 2, 1.

50

Fearon, D.T., Daha, M.R., Weiler, J.M., busten, K.F. (1976):

The natural modulation of the amplification Dhase of complement

activation.

Transplant. Rev. 32, 12.

Ferluga, J., Schorlemmer, H.U., Baptista, L.C., and Allison, A.C.

(1976): Cytolytic effects of the complement cleavage product,

C3a.

Br. J. Cancer 34, 626.

Ferluga, J., Schorlemmer, H.U., Baptista, L.C., and Allison, A.C.

(1978): Production of the complement cleavage product C3a, by

activated macrophages and its tumorolytic effects.

Clin. Exp. Immunol. 31, 512.

Ferreira, Α., and Nussenzweig, V. (1975): Genetic linkage between

serum levels of the third component and the H-2 complex.

J. Exp. tied. 141, 513.

Fong, J.S.C., and Good, Ч.А. (1975): Congenital and experinental-

ly induced complement deficiencies m rabbits.

Birth Defects 9, 571.

Frank, M.M., Hay, J., Gaither, B.S., and Ellman, L. (1971):

In vitro studies of complement function in sera of C4 deficient

guinea pigs.

J. Exp. lied. 134, 171.

Gelfand, M.C., Sachs, D.H., Lieberman, R., and Paul, W.E. (1974):

Ontogeny of В lymphocytes. III. H-2 linkage of a gene controlling

the rate of appearance of complement receptor lymphocytes.

J. Exp. lied. 139, 1142.

Gewürz, H., Pickering, R.J., Mergenhagen, S.E., and Good, R.A.

(1968): The complement profile in acute glomerulonephritis,

systemic lupus erythematosus, and hypocomplementemic glomerulo

nephritis .

Int. Arch. Allergy Appi. Immunol. 34, 556.

51

Giertz, H., Seutter, W., and Schmutzler, Vi. (1964): Release of

histanine from rabbit blood cells by an anaphylatoxin-like sub

stance formed in nomai and C3 defect rabbit plasma.

Nature ?,02, 1346.

Gigli, I., 4uddy, S., and Austen, Κ.Γ. (1968): The stoichioiretric

measurement of the serum inhibitor of the first component of

complement by the inhibition of immune hemolysis.

J. Immunol. 100, 1154.

Gigli, I., and Nelson, R.A. (1968): Complement dependent immune

phagocytosis.

Exp. Cell Res. 51, 45.

Gigli, I., and Austen, K.F. (1969): Fluid phase destruction of

C2 by CI . 1 . Its enhancement and inhibition by homologous

and heterologous C4.

J. Exp. Med. 128, 679.

Goldman, M.B., and Goldman, J. (1976): Pelationship of functional

levels of early components of complement to the H-2 complex of

mice.

J. Immunol. 117, 1584.

Götze, 0., and Müller-Eberhard, H.J. (1974): The role of proper

din in the alternate pathv.'ay of complement activation.

J. Exp. Med. 139, 44.

Götze, 0., Medicus, R.G., and Müller-Eberhard, H.J. (1976):

Comparative analysis of the properties of active and precursor

properdin.

J. Immunol. 116, 1735.

Götze, 0., and Müller-Eberhard, H.J. (1977): The alternative

pathway of complement activation.

Adv. Inm. 24, 1.

Hammer, C.A., Nicholson, Α., and Mayer, M.M. (1975): On the

mechanism of cytolysis by complement.

52


Hammer, C.H., Shin, M.L., Abramovitz, A.S., and Mayer, 4.4.

(1977): On the mechanism of cell membrane damage by complement.

J. Immunol. 119, 1.

Hannema, A.J., Pondman, K.W., Döhmann, U., Gadner, H., and

Dooren, L.J. (1975): C7 deficiency in man.

Prot. Biol. Fluids 22, 581.

Henson, P.M. (1970): Mechanisms of release of constituents from

rabbit platelets by antigen-antibody complexes and complement.

J. Immunol. 105, 476.

Henson, P.M., and Cochrane, CG. (1971): Acute inmune complex

diseases in rabbits.

J. Exp. ried. 133, 554.

Henson, P.M. (1972): Complement dependent adherence of cells to

antigen and antibody. Mechanism and consequences.

In: Biological activities of complement (Ingram ed.)

S. Karger, Nev/ York, p. 173.

Herzenberg, L.A., Tachibana, D.K., Herzenberg, L.A., and Rosen

berg, L.T. (1963): A gene locus concerned with hemolytic comple

ment in Mus musculus.

Genetics 48, 711.

Heusinkveld, R.S., Leddy, J.P., Klemperer, M.R., and Brecken-

ridge, R.T. (1974): Heriditary deficiency of the sixth component

of complement in man. II. Studies of hemostasis.

J. Clin. Invest. S3, 554.

Howie, J.В., and Helyer, B.J. (1965): Autoimmune disease in mice.

Ann. N.Y. Acad. Sci. 124, 167.

Hurst, M.M., Volanakis, J.E., Stroud, R.M., and Bennet, J.C.

(19 75): Cl fixation and classical complement pathway activation

by a fragment of the Cp4 domain of I gil.

J. Exp. Med. 142, 1322.

53

Janoff, Λ. (1970): Mediators of tissue damage in human poly

morphonuclear neutrophils.

Series Haem. 3, 96.

Jersild, C , Rubinstein, P., and Day, N.K. (1976): The HLA system

and inherited deficiencies of the complement system.


Kaplan, Λ.Ρ., Kay, A.B., and Austen, K.F. (1972): A prealbumine

activator of prekallikrein. 3. Appearance of chemotactic activity

for human neutrophils by the conversion of human prekallikrein

to kallikrein.

J. Exp. Med. 135, 81.

Kaplan, A.P., Goetzl, E.J., and Austen, K.F. (1973): The fibrino

lytic pathway of human plasma. II. The generation of chemotactic

activity by activation of plasminogen proactivator.


Kaplan, A.P., and Austen, K.F. (1975): Activation and control

mechanisms of Hageman factor-dependent pathways of coagulation,

fibrinolysis, and kinin generation and their contribution to the

inflammatory response.

J. Allergy Clin. Immunol. 56, 491.

Kitamura, H., Itakura, N., and Inai, S. (1976): A new theoretical

model of immune hemolysis: Application of the reaction between

EAC 1-8 and C9.

Immunochemistry 13, 771.

Klemperer, M.R., Donaldson, V.H., and Rosen, F.S. (1968): Effect

of CI esterase on vascular permeability in man; studies in normal

and complement deficient individuals and in patients with heri-

ditary angioneurotic edema.


Kociba, G.J., Loeb, W.F., and Wall, R.L. (1972): Development of

procoagulant (tissue thromboplastin) activity in cultured leuco

cytes.

J. Lab. Clin. Med. 79, 778. 54

Kolb, W.P., and Müller-Eberhard, H.J. (1973): The membrane attack

mechanism of complement. Verification of a stable C5-9 complex

in free solution.

J. Exp. Med. 138, 4 38.

Lachmann, P.J., Grennan, D., Martin, Α., and Demant, P. (1975):

Identification of Ss protein as murine C4.

Nature 258, 242.

Lai A Fat, R.F.M., and van Fürth, R. (1975): In vitro synthesis

of some complement components (Clq, C3, and C4) by lymphoid

tissues and circulating leucocytes in man.

Immunology 28, 359.

Lepow, I.H., Naff, G.B., Todd, E.W., Pensky, J., and Hinz, C.F.

(1963): Chromatographic resolution of the first component of

human complement into three activities.

J. Exp. Med. 117, 983.

Lepow, I.H. (1971): Permeability producing peptide byproduct of

the interaction of the first, fourth and second components of

complement.

In: Biochemistry of the acute allergic reactions

Eds.: Austen and Becker

Blackwell Scientific Pubi. Oxford p. 205.

Lim, D., Gewürz, Α., Lint, T.F., Ghaze, M., Sepheri, В., and

Gewürz, H. (1976): Absence of the sixth comnonent of complement

in a patient with repeated episodes of meningococcal meningitis.

J. Pediat. 89, 42.

Lindberg, L.H., and Rosenberg, L.T. (1968): Nephrotoxic serum

nephritis in mice with a genetic deficiency in complement.

J. Immunol. 100, 34.

Linscott, W.D., and Cochrane, CG. (1964): Guinea pig 61c-globulin:

its relation to the third component of complement and its altera

tion following interaction with immune complexes.


55

Marder, V.J. (1971): Fibrinogen and fibrin degradation products.

Physicochemical and physiological considerations.

Thromb. Diath. Haemorrh. (suppl.) 47, 85.

Mayer, Π.Μ. (1972): Mechanism of cytolysis of complenent.


Mayer, Γ1.Μ. (1977): Mechanism of cytolysis of lymphocytes:

A comparison with complement.

J. Immunol. 119, 119 5.

McLean, Я.Н., and Michael, A.F. (1973): Prooerdin and C3 pro-

activator: alternate pathway components in human glonerulo-

nephritis.


Medicus, R.G., Götze, 0., and Müller-Eberhard, H.J. (1976a):

Activation of properdin (P) and assembly and regulation of the

alternative pathway C5 convertase.

J. Immunol. 116, 1741.

Medicus, R.G., Schreiber, R.D., Götze, 0., and Müller-Eberhard,

H.J. (1976b): A molecular concept of the properdin pathway.


Medicus, R.G., Götze, 0., and Müller-Eberhard, H.J. (1976c):

Alternative pathway of complement: recruitment of precursor

properdin by the labile C3/C5 convertase and the potentiation of

the pathway.

J. Exp. Med. 144, 1076.

Miller, G.W., and Nussenzweig, V. (1975): A new complement

function: solubilization of antigen-antibody aggregates.


Müller-Eberhard, H.J., and Biro, C.E. (1963): Isolation and des

cription of the fourth component of human complement.

J. Exp. Med. 118, 447.

56

Müller-Eberhard, H.J., Polley, M.J., and Calcott, Ν.Λ. (1967):

Formation and functional significance of a molecular complex

derived from the second and fourth components of human complement.

J. Exp. Med. 125, 359.

Müller-Eberhard, H.J., and Götze, 0. (1972): C3 proactivator

convertase and its mode of action.

J. Exp. Med. 135, 1003.

Naff, G.B., and Ratnoff, O.D. (1968): The enzymatic nature of

C'lr.

J. Exp. Med. 128, 571.

Nilsson, Α., and Müller-Eberhard, H.J. (1967a): Studies on the

mode of action of the fifth, sixth and seventh components of

human complement in immune hemolysis.

Immunology 13, 101.

Nilsson, U.R., and Müller-Eberhard, H.J. (1967b): Deficiency of

the fifth component of complement in mice v;ith an inherited

complement defect.

J. Exp. Med. 125, 1.

Nomenclature of complement (1968).

Bull. W.H.O. 39, 935.

Pensky, J., Levy, L.R., and Lepow, I.H. (1961): Partial purifi

cation of a serum inhibitor of C'l esterase.

J. Biol. Chem. 236, 1674.

Pepys, M.B. (1974): Role of complement in induction of antibody

production in vivo.

J. Exp. Med. 140, 126.

Polley, H.J., and Müller-Eberhard, H.J. (1968): The second com

ponent of human complement: its isolation, fragmentation by C'l

esterase and incorporation into С'З convertase.

J. Exp. Med. 128, 533.

57

Prydz, H., Allison, A.C., and Schorlemmer, H.U. (1977): Further

link between complement activation and blood coagulation.

Nature 270, 173.

Reid, K.B.M., and Porter, R.R. (1976): Subunit composition and

structure of subcomponent Clq of the first component of human

complement.

Biochem. J. 155, 19.

Rittner, Ch. (1976): Genetic loci of components of the classical

and alternate pathway of complement activation: A new dimension

of the immunogenetic linkage group (HLA) on chromosome 6 in man.

Hum. Genet. 35, 1.

Rosen, F.S. (1975): Immunodeficiency.

In: Immunogenetics and immunodeficiency.

Ed.: В. Benacerraf

M.T.P. (London) 1975, p. 229.

Rother, U., and Rother, К. (1961): Über einen eingeborenen

Komplementdefekt bei Kaninchen.

Z. Immunitätsforsch. Exp. Ther. 121, 224.

Rother, K., Rother, U., Petersen, K.F., Gemsa, П., and Mitze, F.

(1964a): Immune bactericidal activity of complement: separation

and description of intermediate steps.


Rother, К., Rother, U., and Schindera, F. (1964b): Passive

Arthus-reaktion bei Komplement-defekten Kaninchen.

Ζ. Immunitäts-Allergieforsch. 126, 473.

Rother, К., and Rother, U. (1965): Studies on complement defec

tive rabbits. IV. Blood clearance of i.v. injected S. typhi by

the reticuloendothelial system.


Rother, К., Rother, U., Müller-Eberhard, H.J., and Nilsson, U.R.

(1966a): Deficiency of Сб in rabbits v/ith an inherited comple-

58

ment defect.

J. Exp. Med. 124, 773.

Rother, К., Vassalli, P., Pother, U., and McCluskey, R.T. (1966b):

Masugi neohritis in С'б defective rabbits.

Fed. Proc. 2S, 309.

Rother, К., Rother, U., Vassali, P., and McCluskey, R.T. (1967):

Nephrotoxic serum nephritis in C6 deficient rabbits. I. Study of

the second phase of the disease.


Rother, К. (1972): Leukocyte mobilizing factor: a new biological

activity derived from the third component of comnlement.

Eur. J. Immunol. 2, 550.

Ruddy, S., and Austen, K.F. (1971): C3b inactivator of man.

II. Fragments raroduced by C3b inactivator cleavage of cell-bound

or fluid phase C3b.

J. Immunol. 107, 742.

Rubin, D., Rother, U., and Rother, К. (1967): Reactivity of comple

ment in cytotoxicity by isoantibodies.

Fed. Proc. 26, 602.

Sandberg, A.L., and Osler, A.G. (1971): Dual pathways of comple

ment interaction with guinea pig immunoglobulins.

J. Immunol. 107, 1268.

Schorlemmer, H.U., and Allison, A.C. (1976): Effects of activated

complement components on enzyme secretion by macrophages.

Immunology 31, 781.

Schorlemmer, H.U., Bitter-Suerman, D., and Allison, A.C. (1977):

Complement activation by the alternative pathway and macrophage

enzyme secretion in the pathogenesis of chronic inflammation.

Immunology 32, 929.

59

Schreiber, P.D., and Müller-Eberhard, H.J. (1974): Fourth com

ponent of human complement: description of a three polypeptide

chain structure.

J. Exp. Med. 140, 1324.

Schreiber, P..D., Medicus, P..G., Götze, 0., and Müller-Eberhard,

H.J. (1975): Properdin- and nephritic factor dependent C3 conver-

tases: requirement of native C3 for enzyme formation and the

function of bound C3b as properdin receptor.

J. Exp. Med. 14 2, 760.

Schreiber, R.D., Götze, О., and Müller-Eberhard, H.J. (1976a):

Alternative pathway of complement: Demonstration and characteri

zation of initiating factor and its properdin-independent function.

J. Exp. Med. 144, 1062.

Schreiber, R.D., Götze, 0., Müller-Eberhard, H.J. (1976b):

Nephritic factor: its structure and function and its relationship

to initiating factor of the alternative pathway.

Scand. J. Immunol. 5, 70 5.

Shin, H.5., Pickering, R.J., and Mayer, M.M. (1971): The fifth

component of the guinea pig complement system.

J. Immunol. 106, 473.

Sissons, J.G.P., West, R.J., Fallows, J., V7illiams, D.G., Baucher,

B.J., Amos, N., and Peters, D.K. (1976): The complement abnormali

ties of lipodystrophy.

N. Engl. J. Med. 29 4, 461.

Sitomer, G.R., Stroud, R.M., and Mayer, M.M. (1966): Reversible

adsorption of С'З by EAC'4 : Role of Mg

Immunochemistry 3, 57.

Snijderman, R., Philips, J., and Mergenhagen, S.E. (1969): Poly

morphonuclear leucocyte chemotactic activity in rabbit serum and

guinea pig serum treated with immune complexes: Evidence for C5a

as the major chemotactic factor.

Infect. Immun. 1, 521.

60

Snijderman, R. , Shin, H.S., and Hausiran, M.H. (1971a): Λ chemo-

tactic factor for mononuclear leukocytes.


Snijderman, M.D., Phillips, J.K., and Ilergenhagen, S.E. (1971b):

Biological activity of complement in vivo. Role of C5 in the

accumulation of polymorphonuclear leukocytes in inflammatory

exsudâtes.

J. Exp. Med. 134, 1131.

Spitzer, R.E., Vallota, E.H., Forristal, J., Sudora, E., Stitzel,

Α., Davis, N.C., and West, C D . (1969): Serum C3 lytic system in

patients with glomerulonephritis.

Science 164, 436.

Stecher, V.J., and Sorkin, E. (1969): Studies on Chemotaxis.

XII. Generation of chemotactic activity for polynorphonuc]ear

leucocytes in sera with complement deficiencies.

Immunology 16, 231.

Stewart, G.J., Ritchie, W.G.M., and Lynch, P.R. (1974): Venous

endothelial damage produced by massive sticking and emigration

of leukocytes.

Aro. J. Path. 74, 507.

Stiffel, C , Biozzi, G., Mouton, D., Bouthillier, Y., and

Decreusefond, C. (1964): Studies on phagocytosis of bacteria by

the reticulo-endothelial system in a strain of mice lacking hemo

lytic complement.


Takahashi, M., Czop, J., Ferreira, Α., and Nussenzweig, V. (1976):

Mechanism of solubilization of immune aggregates by complenent.

Implications for immunopathology.


Takahashi, M., Tack, B.F., and Nussenzweig, V. (1977): Require

ments for the solubilization of immune aggregates by complement:

assembly of a factor В dependent C3 convertase on the immune

61

complex.

J. Exp. Med. 146, 86.

Tamura, N., and Nelson, R.?. (1967): Three naturally occurring

inhibitors of components of coT.pJement in guinea pig and rabbit

serum.


Thompson, R.A., and Lachmann, P.J. (1970): Reactive lysis: the

complement mediated lysis of unsensitized cells,

J. Exp. .'led. 131, 629.

Till, G., and Ward, Ρ.Λ. (1975): Two distinct chemotactic factor

inactivators in human serum.

J. Exp. tied. 114, 843.

Unanue, E.R., Mardiney, M.P., and Dixon, F.J. (1967): Nephrotoxic

serum nephritis in complement intact and deficient mice.


Vallota, E.H., and Müller-Eberhard, H.J. (1973): Formation of C3a

and C5a anaphylatoxins in whole human serum after inhibition of

the anaphylatoxin inactivator.

J. Exp. Med. 137, 1109.

Vogt, W., Dames, V/., Schmidt, G., and Dieminger, L. (1976): Com

plex formation of properdin factor В with C3b, an essential step

in the generation of C3b dependent C3 cleaving enzyme of the

properdin system.

J. Immunol. 116, 1753.

Volk, H., Mauersberger, D., Rother, К., and Rother, U. (1964):

Prolonged survival of skin homografts in rabbits defective in the

third component of complement.


Ward, P.A., Cochrane, CG., and Müller-Eberhard, H.J. (1965):

The role of complement in Chemotaxis of leukocytes in vitro.

J. Exp. Med. 122, 327.

62

Ward, P.A. (1972): Complement-derived chemotactic factors and

their interactions with neutrophilic granulocytes.

In: Biological activities of complement.

Ed.: Ingram, E.D.

S. Karger, Basel, p. 108.

Weiler, J.M., Daha, M.R., Austen, K.F., Fearon, П.Т. (1976):

Control of the amplification convertase of complement by the

plasma protein S1H.

Proc. Natl. Acad. Sci. U.S.A. 7Z, 3268.

Zimmerman, T.S., Arroyave, C.'l., and Hüller-Eberhard, H.J.

(1971): A blood coagulation abnormality in rabbits deficient in

the sixth component of complement, and its correction by purified

C6.

J. Exp. ned. ¿4, 1591.

63

Chapter III

PARTICIPATION OF COMPLEMENT IN GRAFT REJECTION

A REVIEW OF THE LITERATURE

65

3.1. INTRODUCTION

Organs and tissues transplanted across a histocompatibility

barrier elicit as a rule an immune response directed against the

graft, which in most instances results in its destruction. This

process may be mediated by cells that are cytotoxic to the graft

(cellular immunity) or it can be initiated by antibodies speci

fically directed against the graft antigens (humoral immunity).

Which of these reactions will occur or will predominate depends

on a host of variables such as the animal model, the type of allo

graft, the presence of specific presensitization, or the concomi

tant use of immunosuppressive treatment.

The rejection of first set allografts in unmodified recipients

is predominantly cellular, although signs of humorally mediated

destruction (vascular lesions, granulocyte invasion, platelet

aggregation, and fibrin deposits) can be found (Carpenter et al,

1976). However, if an individual is sensitized and has antibodies

against the graft, then the destruction after transplantation is

clearly a humoral process. Depending on the concentration of

circulating antibody the graft shows signs of hyperacute rejection.

This phenomenon occurs in primarily vascularized grafts in

several experimental systems and also in human transplantation.

It is most dramatically illustrated by the grafting of a human

kidney to a recipient who has circulating antibodies against the

HLA-antigens of the donor. Within minutes after restoration of

the vascular connections a violent and rapid destruction occurs.

(Kissmeyer-Nielsen et al, 1966; Williams et al, 1968; Patel and

Terasaki, 1969) . Comparable events take place in first set,

primarily vascularized xenografts, where rejection is caused by

naturally existing antibodies against the graft (Perper and

Najarían, 1966; Giles et al, 1970; Rosenberg et al, 1971).

The rejection of skin-grafts is somewhat different from that

of other organ grafts, since in both first- and second set grafts

cellular rejection has already developed before vascular con

tinuity and consequent accessibility to humoral antibodies are

established.

66

In certain experimental models humorally mediated graft des

truction can be induced by the passive administration of anti

bodies. Klassen and Milgrom (1971) and Holter et al (1972)

induced hyperacute rejection of kidney allografts in rabbits by

injection of antidonor serum. In this kind of experiments skin

grafts again react somewhat differently, since they become only

sensitive to the destructive action of antibodies after the graft

has become vascularized, which occurs 4-5 days after transplanta

tion (Gerlag et al, 1975). To study antibody mediated rejection

of skin grafts cellular rejection has to be prevented by immuno

suppressive treatment (neonatal thymectomy, anti lymphocyte serum).

In such systems it has been shown that administration of anti

bodies can indeed induce acute rejection of both xenografts and

allografts of skin (Winn et al, 19 73; Baldamus et al, 19 73; Koene

et al, 1973).

The tissue injury induced by antibodies is a consequence of

the activation of secondary mediator systems of which the comple

ment system is the most important. Activation of the complement

sequence not only generates a number of immunobiological acti

vities important for the inflammatory reaction, but can also

trigger other effector pathways such as coagulation and kinin

generation (see chapter 2). Since this study focuses on the role

of complement in antibody mediated graft destruction we will here

review the relevant literature concerning its participation in

the rejection process. In the subsequent section we will discuss

the literature concerning the participation of antibodies and

complement in the rejection of various grafts as studied by

immunofluorescence techniques. The following section deals with

the reports of the absence of destruction in experimental animal

systems, if non-complement fixing antibodies are administered.

The literature in which complement activity is measured during

graft rejection is also reviewed, and finally attention will be

given to graft experiments in recipients with congenital comple

ment deficiency, or experimental complement depletion.

67

3 . 2 . IIIMUNOFLUORESCENCE STUDIES OF GRAFTS

By immunofluorescence techniques complement components,

particularly C3 and Clq, can frequently be demonstrated in asso

ciation with IgG and IgM m rejected kidney allografts both m

man(Porter et al, 1968; McKenzie and Whittingham, 1968; HcPhaul

et al, 1970; Andres et al, 1970) and in experimental animals

(Lmdquist et al, 1968; Pbbas et al, 1974). The presence of these

deposits does not by itself mean that these immunoglobulins and

complenent have participated in the allograft rejection. However,

their pathogenetic role is supported by a number of observations.

These deposits in the glomerular capillaries are frequently found

at sites where active destruction of the graft occurs and where

the density of HLA antigens is maximal (Sijbesma et al, 1974) .

Immunoglobulin eluates of these grafts react m vitro with donor

cells (Hager et al, 1964; Hampers et al, 1967; Goldman et al,

1971; Pedersen and Morris, 1974) or, alternatively, produce a

lesion indistinguishable from rejection when injected into ani

mals of the donor strain (Spong et al, 1968). Furthermore it was

found that deposits of IgM and C3 in vessel walls in one hour

biopsies correlated in 89% of the cases studied with 1 month

graft failure, while, if these deposits were absent, only 11% of

the grafts failed after one month (Tourville et al, 1977).

Although it is likely that similar mechanisms play a role in

the humoral destruction of skin grafts, this cannot be substanti

ated, since reports on immunofluorescence studies of such grafts

are lacking.

3.3. PREVENTION OF COMPLEMENT FIXATION

A number of experiments is reported which suggest that des

truction cannot occur if the administered antibodies are unable

to activate complement. This was found with F(abl)2 fragments of

antibodies and with antibodies or antibody subclasses that do not

or very poorly fix complement. These preparations can even block

the destructive action of complement fixing antibodies by their

68

prior binding to the antigenic target. Thus, allografts perfused

with donor specific F(abl)2 antibodies prior to transplantation,

did not show acute rejection after transplantation in specifically

sensitized monkeys (Kobayshi et al, 1972; Habal et al, 1973) or

rabbits (Holter et al, 1973). In experiments with rat skin grafts

in mice Winn and coworkers (1973) found that injection of F(ab,)2

fragments of mouse anti rat serum did not induce acute rejection

in contrast to intact antibodies, which gave rise to a violent

humoral destruction. Moreovor, if F(ab1)2 was given prior to the

intact antibodies the rejection process could be blocked. Similar

results were found in our laboratory in an allogeneic skin graft

model (Capei et al, 1979). Results analogous to those with F(ab,)2

fragments were found with fowl-antibodies which do not fix

mammalian complement e.g. chicken anti rat serum in rat skin grafts

in mice (Winn et al, 1973), and with non-complement-fixing 7S Igl

mouse alloantibodies in skin allografts in mice (Jansen et al,

1975) .

These studies convincingly demonstrate that complement acti

vation plays an important role in graft destruction by passively

administered antibodies.

3.4. MEASUREMENT OF COMPLEMENT ACTIVITY DURING GRAFT REJECTION

Another approach to determine whether complement is involved

in graft rejection, is the measurement of total complement acti

vity (CH50 titer) and the determination of isolated complement

factors with functional assays or with immunochemical methods

(radial immunodiffusion assay) during graft rejection. However,

interpretation of the results is difficult for a number of reasons.

Certain complement factors can already be decreased while the

CH50 titer is still unaltered. Furthermore, the activity found

is the resultant of synthesis, catabolism, and tissue distribution.

Thus, a decrease in complement activity does not by itself mean

that there exists an accelerated catabolism, while alternatively

increased catabolism may go undetected when there is also an in

crease in synthesis. Some complement factors react as acute phase

69

-J о

TABLE I. Complement activity during graft réfection

Author Transplantation model

Grafted organ Recipient species

Complement alterations

A. ALLOGRAFT EXPERIMENTS

A.l. Animals

Simonsen, 1953

Guinea et al, 1964

Gewürz et al, 1966

Coppola and Villegas, 1967

Rother et al, 1967b

Coppola et al, 196Θ

Glovsky et al, 1973

Kux et al, 1973

kidney

skin

kidney

skin

skin

skin

skin

kidney

dog

rat

dog

guinea pig

rat

guinea pig

guinea pig

rabbit

no changes

no changes (neither in first- nor in

second set grafts)

no changes

depression of CH50 titer until

rejection was complete

no changes

depression of CH50 and C3 up to C9,

C2 levels unaltered

no changes

intrarenal complement consumption

(arterio-venous differences in CH50)

A.2. Human

Guinea et al, 1964 kidney slight reduction of CH50 titer, more

pronounced reduction of C2

Austen and Russell, 1966

Gewürz et al, 1966

Carpenter et al, 1967

Carpenter et al, 1969

Shehadeh et al, 1970

Kux et al, 1974

Fearon et al, 1977

Ooi et al, 1977

kidney

kidney

kidney

kidney

kidney

kidney

kidney

kidney

nan

man

man

man

man

man

C2 depression

C2 and CH50 depressed in cases with

"accelerated" rejection

initial rise, followed by a depression

of C3

hypercatabolism of C4 and C3

initial rise of C3 and C4, later on

depression

intrarenal complement consumption

indicative for subsequent rejection

modest depressions of factor В and 01H

depression of factor C3 and C4

B. XENOGRAFT EXPERIMENTS

Gewürz et al, 1966 rabbit kidney

sheep kidney

dog

dog

depression of CI, C4, C2, C3 and CH50

depression of CH50

Table II. Influenae of congenital complément def-iciency or complement devletion on allograft survival

Author Transplantation model

Grafted

organ

Recipxent

species

Presensiti-

zation

Method of decomple-

nentation or defi

cient factor

Effect on graft sur

vival

Ä. COMPLEMENT DEPLETION

A.l. Fvrst-set grafts

Fuji et al, 1966

Gewürz et al, 1967

Pepys, 1972

Glovsky et al, 1973

A.2. Second-set grafts

Maillard and Zarco, 1968

Dempster and Brown, 1970

Kux et al, 1971

skin

kidney

skin

skin

skin

skin

dog

chicken

mouse

guinea pig

guinea pig

skin

kidney

kidney

rat

dog

dog

sodiumcopper-

chlorophyllin

CoVFb)

С о VF

CoVF

CoVF

fumaropimaric acid

slight prolongation

no effect

no effect

prolongation during

C3 depression

c) prolongation of MST

from 8.1 to 10.6 days

no effect

skin grafts CoVF

skin graft CoVF

kidney graft citrate intra-

artenally

Kobayashi et al, 1970 kidney monkey skin grafts CoVF

no effect

no effect

prolongation in 50%,

no rejection in 50%

during citrate in

fusion

no effect

Thomas et al, 1977 kidney rabbit

Whittum and Lindquist, 1977 heart rat

B. COMPLEMENT DEFICIENCY

Volk et al, 1964 skin rabbit

Crisler and Frank, 1965

Caren and Rosenberg, 1965

Biro, 1966

skin

skin

skin

mouse

mouse

rabbit

Rother et al, 1967a skin rabbit

Salerno, 1969 spleen cells mouse

Weitzel and Rother, 1970 skin mouse

a) induction of donor specific antibodies

b) CoVF = cobra venom factor

c) MST = median survival time

skin grafts CoVF

skin grafts CoVF

C6

none

none

skin grafts

C5

C5

C6

C6

none

none

C5

C5

no rejection during

complement depletion

prolongation of MST

from 1.04 to 66 hours

delayed rejection in

50%, graft acceptance

in 50%

no effect

no effect

no effect on first,

second and third

set grafts

no effect in 50%,

prolongation in 35%,

acceptance of graft

in 15%

no effect

slight prolongation

reactants, and can be raised by a number of non specific stimuli.

On the other hand complement factors can be lower as a consequence

of immunosuppressive treatment with prednison (Atkinson and Frank,

1973; Sneiderman and Wilson, 1975), during infection, and in

uremic patients (Kult et al, 1974).

It is therefore not surprising that on considering the litera

ture concerning complement activity during graft rejection one

finds contradictory results. The observations are summarized in

table 1.

In experimental allografts in animals in most instances no

changes in complement activity are found, which is in agreement

with the observation that first-set allografts are predominantly

rejected by cellular mechanisms. In all reports on clinical trans

plantation, however, a decreased complement activity has been

found, especially of factor C2. There is no proper explanation

for this difference. Possibly this observed difference is due to

the fact that in clinical situations attempts are made to treat

the rejection, which is not done in the reported experimental

animal systems. Treatment of rejection in man consists mainly of

increasing the dosis of corticosteroids and it has been shown

that these drugs, especially in high doses, depress the levels of

all complement components (Atkinson and Frank, 1973). A further

argument for this might be that in most instances there is a

delay between the onset of clinically diagnosed rejection and the

decrease of complement levels. However, not all complement altera

tions during graft rejection can be contributed exclusively to a

decreased synthesis since Carpenter et al (1969) found an in

creased catabolism of radiolabelled C4 and C3 during rejection.

It is also possible that humorally mediated rejection is more

important in renal allografts in man because the immunosuppressive

therapy tends to suppress cellular immunity more strongly than the

humoral response (Russell and Winn, 1970). As stated above, immuno

suppression was not used in the reported animal systems.

During the rejection of primarily vascularized xenografts in

animals a clearcut and impressive decrease in complement activity

has been observed. We have found analogous results in a xenogeneic

skin graft system and these will be presented in chapter 7.

74

3.5. INFLUENCE OF COMPLEMENT DEPLETION OR CONGENITAL COMPLEMENT

DEFICIENCY OF THE HOST ON ALLOGRAFT SURVIVAL

The best method to prove participation of complement in graft

rejection, is trying to influence graft survival by manipulation

of the host's complement system. For this purpose cobra venom

factor (CoVF) has been mostly used. CoVF is a factor from certain

snake venoms that, via activation of the alternative pathway,

induces a stable C3 convertase, which is not susceptible to the

control mechanisms. It therefore leads to continuous C3 cleavage

and induces C3 depletion (Cochrane et al, 1970).

The results reported in the literature concerning complement

depletion and graft rejection are rather confusing (see table 2).

For first set allografts three authors (Fuji et al, 1966; Pepys,

1972; Glovsky et al, 1973) report prolongation of graft survival,

while Gewürz et al, 1966, and Glovsky et al, 1973 did not find

any delay in rejection. At first sight it is surprising that in

second set grafts, where humoral rejection is generally more domi

nant, three of six authors do not find prolongation after comple

ment depletion of the recipients.

Also, when recipients with congenital complement deficiencies

are used, almost all authors find an unchanged rejection pattern.

The reason for this becomes however clear, if one realizes that

all authors except one made use of skin grafts in their experi

ments. As we already mentioned, humoral rejection is only seen in

skin grafting experiments if the cellular immune response is

suppressed until vascular connections between graft and recipient

are established. If no immunosuppression is given the rejection

process will be predominantly cellular, and therefore not influenced

by complement deficiency. Two authors (Volk et al, 1964; Rother

et al, 1967a) reported prolonged survival of skin grafts, carried

by C6 deficient rabbits, but they also had in their groups a

number of animals that accepted their grafts indefinitely. This

raises serious doubts whether there existed equal differences in

histoincompatibility between all donors and recipients, and for

this reason it must be questioned whether the observed prolongation

75

σι TABLE III. Influenae of complement depletion, or congenital complement deficiency on xenograft survival

Author Transplantation model Method of deconple- Effect on graft survival

Graft Recipient species mentation or defi

cient factor

A. COMPLEMENT DEPLETION

Snijder et al, 1966

Gewürz et al, 1966,1967

L m n et al, 1970

Moberg et al, 1971

Kux et al, 1971

Mejia-Laguna et al,1972

Winn et al, 1973

pig kidney

rabbit kidney

pig kidney

pig kidney

pig kidney

dog kidney

rat skin

dog

dog

dog

dog

dog

rabbit

mouse

CoVF

CoVF

aggregated IgG

carrageenin

citrate

CoVF

citrate

CoVF

CoVF

prolongation from 9 to 123 m m .

prolongation from 5 to 72-150 min.

prolongation from 5 to 22-90 min.

prolongation from 5 to 51 m m .

20 fold prolongation

prolongation from 5 to 30-60 m m .

no rejection during citrate infusion

prolongation from 17 to 74 min.

delayed or no rejection depending

on the dose of CoVF and passively

administered antibodies

B. COMPLEMENT DEFICIENCY

Mejia-Laguna et al,1970 dog kidney rabbit

Winn et al, 1973 rat skin mouse

C6

C5

no effect

delayed rejection depending on the

amount of passively administered

antibodies

was rightly attributed to the C6 deficiency.

3.6. INFLUENCE OF COMPLEMENT DEPLETION OR CONGENITAL COMPLEMENT

DEFICIENCY OF THE HOST ON XENOGRAFT SURVIVAL

In the studies reported (see table 3) acute rejection could be

prevented by depletion of the complement system (in most instances

C3 depletion). However the protection was not complete, since

rejection eventually occurred m all cases.

The few studies that were performed in animals with congenital

complement deficiencies suggest that rejection of xenografts

occurred in spite of a deficiency in the terminal complement com

ponents. Thus, in a study with dog kidneys transplanted to C6

deficient rabbits the rejection was the same as in normal rabbits

(Me^ia-Laguna et al, 1970). Comparable results were obtained by

Winn et al (1973) with established rat skin grafts m C5 deficient

mice. With high dosis of mouse anti rat antiserum humoral destruc

tion of the grafts could be induced.

3.7. CONCLUSION

On reviewing the literature we have come to the conclusion

that complement clearly is involved in the rejection of xenografts.

This is supported by the histologic findings in rejected xeno

grafts. These show the characteristic picture of an Arthus reac

tion, that is known to be complement dependent. From the results

of xenograft studies in complement deficient recipients it seems

likely that not all complement factors are necessary, and that

generation of factors C3a and C5a, with their biologic important

functions, suffices to cause graft damage. However, complement

inhibition did never prevent a rejection, which suggests that

complement independent processes are also involved.

The literature concerning complement participation in allo

graft rejection is contradictory. With regard to skin allografting

in non-immunosuppressed animals most of the available data indi

cates that humoral rejection, and thus complement involvement, is

77

of minor importance.

The reports on primarily vascularized allografts in animals

do not allow a definite conclusion, although the results in sen

sitized recipients make it likely that humoral mechanisms contri

bute to graft damage.

In human kidney transplantation complement participation is

probably of greater importance than in analogous experimental

animal systems. Most rejected human kidneys show characteristics

of antibody mediated graft destruction. This is possibly due to

the fact that the recipients are treated with immunosuppressive

agents which better suppress cell-mediated than humoral immunity.

Manipulation of the complement system might, therefore, be a means

to interfere with this destructive process.

78

REFERENCES

Abbas, Л.К., Carson, J.M., Carpenter, C.B., Galvanek, E.G.,

Merrill, J.P., and Damnin, G.J. (1974): Immunologic enhancement

of rat renal allografts. II. Immunohistologv of acutely rejecting

and passively enhanced grafts.

Amer. J. Path. 75, 271.

Andres, G.A., Accinni, L., Hsu, K.C., Penn, I., Porter, K.A.,

Rendali, J.M., Seegal, B.C., and Starzl, Т.Е. (1970): Human renal

transplants. 3. Immunopathologic studies.

Lab. Invest. 22, 558.

Atkinson, J.P., and Frank, H.M. (1973): Effect of cortisone

therapy on serum complement components.


Austen, K.F., and Russell, P.S. (1966): Detection of renal allo

graft rejection in man by demonstration of a reduction in the

serum concentration of the second component of complement.


Baldamus, CA., McKenzie, I.F.С, Wmn, H.J., and Russell, P.S.

(1973): Acute destruction by humoral antibody of rat skin grafted

to mice.

J. Immunol. 110, 1532.

Biro, C E . (1966): The role of the sixth component of complement

in some types of hypersensitivity.

Immunology 10, 563.

Capel, P.J.Α., Tamboer, W.P.M., de Waal, R.M.W., Jansen, J.L.J. ,

and Koene, R.A.P. (1979) : Passive enhancement of mouse skin allo

grafts by alloantibodies is Fc dépendent.

J. Immunol. 122, 421.

79

Caren, L.D., and Rosenberg, L.T. (1965): Complément in skin

grafting in nice.

Immunology 9, 359.

Carpenter, C.B., Gill, T.J., Merrill, J.P., and Dammin, G.J.

(1967): Alterations in human serum ßlc-globulin (C3) in renal

transplantation.

Am. J. Med. 43, 854.

Carpenter, C.B., Ruddy, S., Shehadeh, I., Müller-Eberhard, H.J.,

Merrill, J.P., and Austen, K.F. (1969): Complement metabolism of

the fourth (C4) and third (C3) components in patients with renal

allograft rejection and heriditary angioedema (HAE).


Carpenter, C.B., d'Apice, A.J.F., and Abbas, ̂ .K. (1976): The role

of antibodies in the rejection and enhancement of organ allografts.

Adv. Immunol. 22, 1.

Cochrane, CG., Müller-Eberhard, H.J., and Aiken, B.S. (1970):

Depletion of plasma complément in vivo by a protein of cobra

venom: its effect on various immunologic reactions.


Coppola, E.D., and Villegas, G.R. (1967): Serum complement:

changes after skin grafting m guinea pigs.

Proc. Soc. Exp. Biol. Med. 12S, 1071.

Coppola, E.D., Villegas, G.R., and Rosato, F.E. (1968): Changes

in serum complement and fucose after skin grafting m guinea pigs.

Surg. Res. 8, 308.

Cnsler, C , and Frank, Ч.М. (1965): Skin graft rejection and

Arthus reaction in mice deficient in the third component of com

plement.

Fed. Proc. 24, 447.

Dempster, to.J., and Brown, D.L. (1971): The nature of experimen

tal second set kidney transplant rejection. 3. The role of comple-

80

ment.

Br. J. Exp. Path. 52, 280.

Fearon, D.T., Daha, M.R., Strom, T.B., Weiler, J.M., Carpenter,

C.B., and Austen, K.F. (1977): Pathways of complement activation

in membranoproliferative glomerulonephritis and allograft rejection.


Fuji, G., Suzuki, M., Hirose, Y., Goto, Sh., Ishibashi, J., Haga,

К., and Sindo, T. (1966): Effect of sodium-copper-chlorophyllin

as a complement inhibitor on the allograft rejection.

Jap. J. Exp. Med. 36, 499.

Gerlag, P.G.G., Koene, Я.A.P., Hagemann, J.H.F.M., and Wijdeveld,

P.G.А.В. (1975): Hyperacute rejection of skin allografts in the

mouse. Sensitivity of ingrowing skin grafts to the action of

alloantibody and complement.

Transplantation 20, 308.

Gewürz, H., Scott-Clarck, D., Finstad, J., Kelly, W.D., Varco,

R.L., Good R.A., Gabrielsen, A.E. (1966): Role of the complement

system in graft rejections in experimental animals and man.


Gewürz, H., Scott-Clarck, D., Cooper, M.D., Varco, R.L., Good,

R.L. (1967): Effect of cobra venom-induced inhibition of comple

ment activity on allograft and xenograft rejection reactions.


Giles, G.K., Boehmig, H.J., Lilly, J., Amemiya, H., Takagi, H.,

Colrug, A.J., Hathaway, W.E., Wilson, C.B., Dixon, F.J., and

Starzl, Т.Е. (1970): Mechanism and modification of rejection of

heterografts between divergent species.


Glovsky, M.M., Ward, P.A., and Fudenberg, H.H. (1973): Role of

complement in guinea pig skin allograft rejection.

Clin. Imm. Immunopath. 1, 165.

81

Goldman, В.S., Shumak, Κ.Η., Silver, M.D., Cartier, J., Lixfield,

W., Bigelow, W.G., and Crookston, J.H. (1971): Antibodies in the

serum and on the heart of dogs with cardiac allografts.

Transplant. Proc. J, 515.

Guinea, E.J., Austen, K.F., and Russell, P.S. (1964): Measurement

of serun complement during homograft rejection in man and rat.


Habal, M.B., Carpenter, C.B., Kobayashi, K., Bush, G.J., Migra, M.,

and Birtch, A. (1973): The protective effect of r(ab)2 and succi-

nylated IgG in the early treatment of hyperacute rejection in the

primate.


Hager, E.B., DuPuy, M.P., and Wallach, D.F. (1964): Immunologic

suicide and studies on the role of antibody and complement in

canine kidney homograft rejection.


Hampers, C.L., Kolker, P., and Hagar, E.B. (1967): Isolation and

characterization of antibodies and other immunologically reactive

substances from rejecting renal allografts.


Holter, A.R., McKearn, T.J., Neu, M.R., Fitch, F.W., and Stuart,

F.P. (1972): Renal transplantation in the rabbit.


Holter, A.R., Neu, M.R., McKearn, T.J., Lynch, A.F., and Stuart,

F.P. (1973): Abrogation of hyperacute rejection of renal allo

grafts by pepsin digest fragments of antidonor antibody.


Jansen, J.L.J., Koene, R.A.P., van Kamp, G.J., Tamboer, W.P.M.,

and Wijdeveld, P.G.А.В. (1975): Isolation of pure IgG subclasses

from mouse alloantiserum and their activity in enhancement and

hyperacute rejection of skin allografts.

J. Immunol. 115, 387.

82

Kissmeyer-Nielsen, F., Olsen, S., Petersen, V.P., and Fjeldborg,

0. (1966): Hyperacute rejection of kidney allografts, associated

with pre-existing humoral antibodies against donor cells.

Lancet 2, 662.

Klassen, J., and Milgrom, F. (1971): Studies on cortical necrosis

in renal grafts.

Transplant. Proc. Z, 598.

Kobayashi, K., Hricko, G.M., Habal, M.B., Lukl, P., Busch, G.J.,

Hunsicker, L., Reisner, G.S., and Birth, A.G. (1972): Hyperacute

rejection of renal allografts in the primate.


Koene, R.A.P., Gerlag, P.G.G., Hagemann, J.H.F.M., van Haelst,

U.J.G., and Wijdeveld, P.G.А.В. (1973): Hyperacute rejection of

skin allografts in the mouse by the administration of alloanti-

body and complement.


Kult, J., Richter, U., Scheitza, E., Hennemann, H., and Heidland,

Α. (1974) : Störungen im Komplementsystem bei Niereninsuffizienz

und ihre Beeinflussung durch Aminosäurensubstitution.

Dtsch. Med. Wochenschr. 99, 3 39.

Kux, M., Boehmig, H.J., Amemiya, H., et al (1971): Modification

of hyperacute canine renal homograft and pig to dog heterograft

rejection by the intra-arterial infusion of citrate.

Surgery 70, 103.

Kux, M., Piza, F., Ranks, R., Simma, VJ., Sternberger, H., Stockinger,

L., and Wiedermann, G. (1973): Komplementaktivierung und Endothel-

schädigung in den ersten zwei Stunden nach experimenteller Nieren

homotransplantation .

Z. Immunitätsforsch. 14S, 432.

Kux, M., Piza, F., Sternberger, H., Wagner, 0., and Wiedermann, G.

(1974): Immediate complement consumption in human renal allografts.

A diagnostic and prognostic test.

83

Eur. Surg. Res. 6, 5.

Lindquist, R.R., Guttmann, R.D., Merrill, J.P. (1968): Renal

transplantation in the inbred rat. II. An immunohistochemical

study of acute allograft rejection.

Amer. J. Path. 52, 531.

Linn, B.S., Jensen, J., Pardo, V., Levi, F., and Hutson, D.G.

(1970): Prolongation of renal xenografts by citrate.

JAMA 212, 864.

Maillard, J.L., and Zarco, R.M. (1968): Dêcomplenentation par un

facteur extrait du venin de cobra. Effet sur plusieurs reactions

immunes du cobaye et du rat.

Ann. Inst. Pasteur. 114, 756.

McKenzie, I.F.С, and Whittinghan, S. (1968): Deposits of immuno

globulin and fibrin in human allografted kidneys.

Lancet 2, 1313.

McPhaul, J.J., Dixon, F.J., Brettschneider, L., and Starzl, Т.Е.

(1970): Immunofluorescent examination of biopsies from long-term

renal allografts.

N. Engl. J. Med. 282, 412.

Mejia-Laguna, J.E., Garcia-Cornejo, M., Lopez-Soriano, F., and

Biro, C.E. (1970): The role of the sixth component of complement

in the rejection of kidney xenografts.

Immunology 19, 767.

Mejia-Laguna, J.E., Martinez-Palomo, Α., Biro, C.E., Chavez, В.,

Lopez-Soriano, F., and Garcia-Cornejo, M. (1972): Morphologic

study of the participation of the complement system in hyperacute

rejection of renal xenotransplants.

Amer. J. Path. 69, 71.

Moberg, A.W., Shons, A.R., Gewürz, H., and Mozes, M., and Najarían,

J.S. (1971): Prolongation of renal xenografts by the simultaneous

sequestration of preformed antibody, inhibition of complement,

84

coagulation and antibody synthesis.

Transplant. Proc. Ъ, 538.

Ooi, У.M.; Ooi, В.S., Vallota, E.H., First, M.R., and Pollak, V.E.

(1977): Circulating immune comDlexes after renal transplantation. 125

Correlation of increased I-Clq binding activity with acute

rejection characterized by fibrin deposition in the kidney.


Patel, P., and Terasaki, P.I. (1969): Significance of the positive

crossmatch test in kidney transplantation.

N. Engl. J. Med. 280, 735.

Pedersen, N.C., and Morris, B. (1974): The role of humoral anti

body in the rejection of primary renal allografts in sheep.

J. Exp. Med. 1Û0 , 619.

Pepys, M.B. (1972): Role of complement in induction of the aller

gic response.

Nature 237, 157.

Perper, R.J., and Najarían, J.S. (1966): Experimental renal hetero

transplantation.


Porter, K.A., Andres, G.A., Calder, M.V7., Dossetor, J.В., Hsu,

K.C., Rendali, J.M., Seegal, B.C., and Starzl, Т.Е. (1968):

Human renal transplants. II. Immunofluorescent and immunoferritin

studies.

Lab. Invest. IS, 159.

Rosenberg, J.C., Hawkins, E., and Rector, F. (1971): Mechanism of

immunological injury during antibody-mediated hyperacute rejection

of renal heterografts.


Rother, U., Ballantyne, П., Cohen, С , and Rother, К. (1967a):

Allograft rejection in C6 deficient rabbits.

J. Exp. Med. 12S, 565.

85

Rother, К., Rother, U., and Ballantyne, D.I.. (1967b): Serum com

plement activity in rat recipients of small and massive skin

allografts.

Proc. Soc. Exp. Biol. 124, 439.

Russell, P.S., and Winn, H.J. (1970): Transplantation.

N. Engl. J. Med. 282, 786, 848, 896.

Salerno, A. (1969): Survival of transplanted spleen cells in allo

genic mice. The effect of strength of histoincompatibility and of

complement deficiency.


Shehadeh, I.H., Carpenter, C.B., Monterio, C.H., and Merrill, J.P.

(1970): Renal allograft rejection. An analysis of lysozymuria,

serum complement, lymphocyturia, and heterophil antibodies.

Arch. Int. Med. 2 25, 850.

Simonsen, M. (1953): Biological incompatibility in kidney trans

plantation in dogs. II. Serological investigations.

Acta Path. Microb. Scand. 32, 36.

Sneiderman, CA., and Wilson, J.A. (1975): Effects of cortico

steroids on complement and the neutrophilic polymorphonuclear

leukocyte.


Snijder, G.B., Ballesteros, E., Zarco, R.M., and Linn, В.S. (1966):

Prolongation of renal xenografts by complement suppression.

Surg. Forum 17, 478.

Spong, F.W., Feldman, J.D., and Lee, S. (1968): Transplantation

antibody associated with first-set renal homografts.

J. Immunol. 101, 418.

Sybesma, J.P., Kater, I,., Borst-Eilers, E., de Planque, Β.Α.,

van Soelen, T., and Tuit, T. (1974): HLA antigens in kidney tissue.

Localisation by means of immunofluorescence technique.


86

Thomas, F., Naff, G., Thomas, J., and Dvorak, K. (1977): Prevention

of hyperacute kidney rejection by decomplementation using purified

cobra venom factor.

J. Surg. Res. 22, 189.

Tourville, D.R., Kim, D.U., Viscuso, M., Jacobs, 4., Schoen, Ξ.,

and Fillepone, D. (1977): anticipation of renal transplant failure

by postanastomosis biopsy and immunofluorescence.


Volk, Η., Mauersberger, D., Rother, К., Rother, IJ. (196 4):

Prolonged survival of skin homografts in rabbits defective in the

third component of complement.


Weitzel, H.K., and Rother, К. (1970): Studies on the role of

serum complement in allograft rejection and in immunosuppression

by antithymocyte serum.

Eur. Surg. Res. 2, 310.

Whittum, J.Α., and Lindquist, R.R. (1977): Mechanisms of cardiac

allograft rejection in the inbred rat. The effect of complement

depletion by cobra venom factor on hyperacute cardiac allograft

rejection.


Williams, G.M., Hume, D.M., Hudson, R.P., Morris, P.J. and

Milgrom, F. (1968): "Hyperacute" renal homograft rejection in man.

N. Engl. J. Med. 2 79, 611.

VJinn, H.J., Baldamus, C.A., Jooste, S.V., Russell, P.S. (1973):

Acute destruction by humoral antibody of rat skin grafted to mice.

J. Exp. Med. 137, 89 3.

87

Chapter IV

ROLE OF ANTISERUM AND COMPLEMENT IN THE ACUTE ANTIBODY-MEDIATED

REJECTION OF MOUSE SKIN ALLOGRAFTS IN STRAIN COMBINATIONS WITH

INCREASING HISTOINCOMPATIBILITY

J.H.M. Berden, P.G.G. Gerlag, J.F.H.M. Hagemann, and R.A.P. Koene

Department of Medicine, Division of Nephrology, Sint Radboud

Ziekenhuis, University of Nijmegen, Nijmegen

published in: Transplantation (1977) 24, 175

89

ROLE OF ANTISERUM AND COMPLEMENT IN THE ACUTE ANTIBODY-MEDIATED REJECTION OF

MOUSE SKIN ALLOGRAFTS IN STRAIN COMBINATIONS WITH INCREASING

HISTOINCOMPATIBILITY1

J. H. M B E R D E N , P. G G GERLAG, J F H M H A G E M A N N , A N D R A Ρ

KOENE

Department of Medicine, Division of Nephrology, Sint Radboud Ziekenhuis University of Nijmegen, Nijmegen, The Netherlands

SUMMARY

Acute antibody-mediated rejection (AAR) of mouse skin allografts was studied in nine donor-recipient combinations with increasing histoincompatibihty ranging from an Η-Y to a complete H-2 plus non-H-2 disparity AAR was induced by the injection of specific alloantiserum along with a heterologous complement on day 7 after grafting Sera from rabbits, guinea pigs, and from a human volunteer were used as complement sources The recipients were treated with antilymphocyte serum on days 0, 2, and 4 to postpone cell-mediated rejection With increasing histoincompatibihty the mean survival time of untreated grafts decreased, the in vitro cytotoxic activity of the alloantiserum rose, and AAR could be induced with lower amounts of antiserum The higher efficiency of rabbit complement compared with guinea pig complement and human complement, that is known to exist in in vitro cytotoxicity, was also found in vivo Rabbit complement could induce AAR m combination with relatively weak histoincompatibihty (H-2K, H-2D, or non H-2 differences), where guinea pig complement and human complement were ineffective All three complement species elicited AAR if there was a disparity for H-2D plus non-H-2, H-2K plus non-H-2, H-2, or H-2 plus non-H-2 The rules for immunogemcity of the different histocompatibility loci as they have been described for cell-mediated graft destruction also apply to this humorally mediated rejection process

Since the first description of hyperacute rejection of human renal allografts (14), there has been a growing interest in the role of humoral antibodies in graft destruction Primarily, vascularized grafts are clearly susceptible to antibody-mediated damage, not only in man, but also in many experimental animals (2) The role of antibody in the destruction of skin grafts is more controversial Until recently it was assumed that skin grafts were resistant to antibody and that their destruction was brought about by cellular mechanisms However, several models have now been described In which antibodies are responsible for acute skin graft destruction Jooste and Winn ill) and Jooste et al (12) were able to induce acute

1 This work was supported by grants from the Netherlands Foundation of Medical Research (FUNGO) and from the Netherlands Kidney Foundation

destruction of rat skin grafted onto immuno-suppressed mice by the injection of mouse anti-rat serum or of rat antiserum specifically directed against the alloantigens of the donor In a subsequent study they showed that allogeneic rat skin grafted onto immunosuppressed rats could also be destroyed by alloantibody (11) Furthermore, they found evidence that the final damage is caused by the activation of mouse complement (27) In a previous report we have shown that mouse skin allografts can be successfully destroyed by alloantibody provided a heterologous complement, ι e , rabbit complement, is administered along with the antiserum (16) Apparently the mouse endogenous complement is inefficient in this situation and this correlates well with the low efficiency of this complement species in in vitro cytolysis systems (9) In the mouse the study of acute rejection of allografts by antibody is of particu-

91

TRANSPLANTATION

lar interest because its many inbred strains permit detailed genetic analysis The genetics of cell-mediated skin grail destruction has been extensively studied in this species, and it has become evident that there exists a close relationship between the time required for graft, destruction and the extent of histocompatibility differences (7,20,21) Our model of acute antibody-mediated, complement-dependent rejection (AAR) enabled us to study whether an analogous correlation also exists for humorally mediated allograft damage To this end we have studied the occurrence of AAR in nine donor-recipient combinations with increasing histoincompatibihty

Complement activation is necessary for AAR and whether this occurs seems to depend upon the degree of histoincompatibihty between donor and recipient The studies of Jooste and Winn ill) and Jooste et al (12) have shown that mouse complement can be come activated in a xenogeneic system and also in certain allogeneic combinations In our system with limited antigenic differences, rabbit complement was required for the de struction of allografts by antibody Analogous results have been obtained in renal grafts in rats August ra t kidneys in AS recipients were not destroyed by antibody, unless guinea pig complement was administered with the antiserum (6) On the other hand, DA grafts in Lewis recipients, which represent a stronger histoincompatibihty than August to AS, could be destroyed by administration of antibody alone (5) Differences in efficiency of various complement species have also been observed in many in vitro cytolytic systems (4, 9, 10, 17, 26) With mouse alloantibody rabbit complement is superior to guinea pig complement, while rat and mouse complement hardly ever induce complement-dependent cytotoxicity (9, 17) The second aim of this study was, therefore, to test several complement species rabbit complement (RC), guinea pig complement (GPC), and human complement ( H O , for their ability to evoke AAR of skin grafts m combinations with increasing histoincompatibihty and to compare these results with the complement activity in in vitro cytotoxicity

MATERIALS AND METHODS

Mice Inbred lines of BIO D2/new Sn, A/HeJ, BIO Br, and BIO LP mice were originally ob

tained from the Jackson Laboratory, Bar Harbor, Maine C57BL6/Rij mice were obtained from the Radiobiological Institute TNO, Rijswijk, The Netherlands BALB/c and DBA/2 mice were originally obtained from the National Institutes of Health, Bethesda, Maryland CBA/J mice were bought from a commercial animal farm (G L Bomholtgard Ltd, Rij, Denmark)

In our animal laboratory these mouse strains were kept by continuous brother-sister mat-ings (C57BL6/R1J x A/HeJJF, (B6AF1) hybrids were raised in the laboratory All of these strains were analysed with various monospecific alloantisera (obtained from National Insti tutes of Health, Bethesda, Maryland) against private and public specificities of the H-2 complex to rule out the possibility of differences between the presumed and real haplotypes of these strains

Selection of donor-recipient combination Information concerning H 2 haplotypes, H-2 recombinants, and non H 2 differences used in selecting donor-recipient combinations came from Demant (3), Graff and Bailey (7), Graff and Snell (8), and Klein (15) The nine combinations chosen and their antigenic differences are listed in Table 1 The first three combinations comprise only non-H-2 differences in an increasing number Combination 3 represents disparity of almost all known non H 2 loci Combinations 4, 6, and 8 differ in the H-2D region, the H 2K region, and the complete H-2 complex, respectively Since C57BL10 congenie strains served as donors on C57BL6 recipients there is also a difference at the H 9 locus (8) This difference, however, is weak so that there will be no cumulation of immunogemcity between the non-H-2 and H-2 differences ( 7) Furthermore, absorption of the C57BL6 anti-B10 D2 serum with C57BL10 cells did not change the titer of the antiserum in vitro For practical reasons we therefore consider these combinations as differing only at the H 2 complex In combinations 5, 7, and 9 there exists not only a difference at the H 2 complex (H-2D end, H 2K end, and complete H-2, respectively), but there are also a number of non-H-2 differences

Production of alloantisera Alloantisera were prepared by injecting female recipient mice at weekly intervals, ι ρ , with a suspension of approximately 5 χ IO7 donor lymphoid cells in complete Freund's adjuvant For the

92

BERDEN ET AL.

TABLE 1. Antigenic differences of strain combinations used

Combination

1 2

3

4 5

6 7 θ

9

Donor (M)

B6AF, (M) C57BL6

B 1 0 D 2

BIO Br A/HeJ

B10.D2 DBA/2 B 1 0 D 2

DBA/2

• Only "classic" H-2 4 According toU).

Recipient (F)

B6AF, (F) BIO LP

BALBc

B6AF, CBA/J

B6AF, B6AF, C57BL6

C57BL6

specificities

Haplotype

a χ b -> a χ 6

ь — b

d->d

k — a χ 6 0 - *

d - » α χ 6 d -» a χ 6

< i - > 6

H-2 coi

Private

-

H-2D 32 II-2D 4

H.2K 31 H-2K 31 H-2K31 H-2D4

H-2K 31 H-2D4

nplex

Specifici ties"

Public

-

H-2 6, 10, 13, 14, 27, 29, 35, 36, 40, 41, 43.44

H-2 34 H-2 34 H-2 3 ,8 , 10, 13,34, 4 1 , 4 2 . 4 3 , 4 4 , 4 7 , 4 9

H-2 3, 8, 10, 13, 34, 41, 42, 43, 44, 47, 49

28, 42,

40,

40,

are listed Specificties 34 and 47 are probably la specific

Non-H-2 loci

H-Y H-3, H-9, H-13 and probably

two other*

H-1.H-3,H-4,H-7,H-8,H-' 13, H-15, tlm 30, Il-X

H-9 H-I, H-4, H-7, H-9 and prob

ably also H-3 and H-8

H-9 HI, H-3, H-4, H-8, H-13 H-9

HI, H-3, H-4, H-8, H-13

ι ties

anti-H-Y antiserum B6AF, sperms cells were used instead of lymphoid cells. After five injections most animals developed ascites and could be tapped every 1 to 2 weeks. The production of several weeks was pooled and, before use, all antisera were heated at 56 С for 45 min and sterilized by passing through a sterile 0.20-Γημ filter. The cytotoxic titer of the ascites fluid was the same as the titer in the serum of these mice. Antisera were stored at -20 С.

Anhlymphocyte serum (ALS). ALS was prepared by a s.c. injection of 5 to 8 χ IO7 C57BL6 lymphoid cells (thymus and mesenterial lymph nodes) in complete Freund's adjuvant into a goat. This priming dose was followed by weekly i.v. boosters of the same number of C57BL6 lymphoid cells. The animals were bled 1 week after the fourth immunization, and the serum was heated at 56 С for 45 min and absorbed once with C57BL6 RBC (5 volumes of serum/1 volume of packed cells). This ALS was not toxic to the recipients if injected i.p.

Complement. Fresh or fresh frozen sera from New Zealand White rabbits, random-bred SPF Albino guinea pigs, and from a human volunteer were used as sources of complement. Before use all sera were tested for toxicity in vitro (a 1/4 dilution should give less than 10% cytoly-sis of the various donor lymphocytes) and in vivo by injecting 0.5 ml of serum i.v. Blood was obtained in rabbits and guinea pigs by cardiac puncture, in the human by venous puncture. After clotting for 45 min at room temperature

and centnfugation, the serum was divided in aliquots of 0 5 ml and, if not immediately used, stored at -30 С Complement was used only once after thawing.

Serology. Cytotoxic tests were performed using the trypan blue exclusion test. Except in the B6AF! female anti-male antiserum, where B6AF, sperm cells were used, spleen cells were used as the targets. They were prepared free of RBC with Tris-NH,C1. One hundred microliters of cells (5 x 105) in Hank's balanced salt solution (BSS) were incubated with equal volumes of antibody dilutions and RC, GPC, or HC (diluted 1/4) for 30 min at 37 С The titer was defined as the antiserum dilution giving 50% lysis. All tests were performed in triplicate and the following controls were performed: (1) cell control, 100 μ\ of cells in 200 μΐ of Hank's BSS; (2) antibody control, 100 μΐ of cells, 100 μΐ of undiluted antiserum, and 100 μΐ of Hank's BSS; and (3) complement control, 100 μΐ of cells, 100 μΐ of complement, and 100 μΐ of Hank's BSS. In these controls cytolysis should not exceed 10%.

Skin grafting technique. Donor tail skin was grafted onto the flank of the recipients by a modification of the "fitted graft" technique. Instead of plaster bandage we used a band-aid to cover the graft. In all experiments except in combination 1, donor and recipient were females and between 6 to 10 weeks old. The band-aid was removed on the 6th day after transplantation. The fate of the grafts was followed by

93

TRANSPLANTATION

daily macroscopic inspection The grafts were considered to be rejected when no viable epidermis remained The median survival time (MST) and the standard deviations were calculated according to the method of Litchfield (ÍS)

Hyperacute rejection studies All experiments were performed on well established grafts that showed no signs of rejection In all combinations, except combinations 1 and 2, the survival of these grafts was prolonged by nonspecific immunosuppression with ALS (0 25 ml ι ρ ) on days 0, 2, and 4 after grafting Seven days after grafting, the mice received 0 01 to 2 0 ml of alloantiserum directed against the donor antigens and 0 25 ml of RC, 0 5 ml of GPC, or 0 5 ml of HC by ι ν injection through the tail vein The choice of the volumes of complement was based on dose response studies in combination 8, where we found that these amounts provided excess of complement if injected with excess antibody Volumes of alloantiserum lower than 0 5 ml were given in an appropriate dilution in saline, with an end volume of 0 5 ml In each donor-recipient combination control experiments were performed by injecting antiserum alone, complement alone, and 0 25 ml of RC with a particular for that combination nonspecific alloantiserum AAR was defined as the complete necrosis of the skin graft 24 to 72 hr after ι ν injection In most cases this process was completed within 48 hr

Histological examination In the two combinations where we did not see macroscopic signs of acute rejection, histological examinations were performed Vz, 2, and 8 hr after ι ν injec

tion of alloantiserum and complement The skin graft and surrounding recipient skin were removed and fixed in 4% buifered formalin The paraffin sections were stained with hematoxylin and eosin

RESULTS

Cytotoxic titers of alloantisera The titers of the various alloantisera are given in Table 2 In all cases we could detect cytotoxic antibodies, even in the three combinations with only non-H-2 differences Antiserum 1 induced cytolysis of B6AF| sperm In antiserum 2, prepared across an H-3, H-9, and H-13 difference, we could demonstrate cytotoxic activity by increasing the sensitivity of the test with the use of less target cells (W instead of 5 χ IO5) With many non-H-2 antigenic differences present, as is the case in combination 3, the antiserum became highly cytotoxic (i/128 with RC)

Antisera produced across H-2 differences showed a gradual rise in cytotoxicity as the antigenic difference increased The titer was lowest with only an H-2D difference, higher with an H-2K difference, and highest when there was incompatibility for the whole H-2 complex (antisera 4, 6, and 8) Addition of non-H-2 differenceb to each of these H-2 differences (antiserum 5, 7, and 9) gave a further nse in cytotoxic activity with GPC and HC, whereas this increase was absent in the tests with RC The situation is somewhat special with antiserum 5, since this serum does not simply represent an extension of antiserum 4 with non-H-2 antibodies The H-2D 32 difference of combina-

TABLE 2 Cytotoxic titers of alloantisera with different complement species

Combination Alloantiserum

B6AF, (F) anti-BSAF, BIO LP anti-C57BL6 BALB/c anti-BIO D2 B6AF, anti-B10 Br CBA/J anti-A/HeJ B6AF, anti BIO D2 B6AF, anti-DBA/2 C57BL6 anti-BIO D2 C57BL6 anti-DBA/2

(M)

Cytotoxic titer

RC GPC

+° + ••

1/128 1/16 1/512 1/512 1/512 1/3072 1/4096

ND» ND + « 0 1/16 1/12 1/32 1/64 1/128

HC

ND ND + « 0 1/16 1/8 1/32 1/48 1/96

α Cytolysis of sperm cells up to 90% in dilution 1/2 * ND, not done r Cytolysis up to 60% above controls in dilution 1/2, with lower number of target cells (10 /̂100 μΐ), controls

less than 10% lysis '* Cytolysis up to 30% in dilution 1/2

9 4

BERDEN ET AL

tion 4 has been replaced here by an H-2D 4 difference, which is more immunogenic (20) and, furthermore, many public specificities are introduced This could explain the relatively high rise in cytotoxic activity of antiserum 5 as compared with antiserum 4

In accordance with previous studies (4,9,10, 17), we found that in all instances the cytotoxicity was highest if measured with RC as compared with GPC or НС To test the activity of mouse complement, each test was also performed with fresh recipient mouse serum, but cytolysis clearly above control levels was never observed

Acute antibody-mediated rejection The results are summarized in Table 3 Control grafts placed on recipients that received ALS survived with MST's ranging from 17 8 ± 1 2 to 33 9 ± 1 4 days Thus, in no case were there signs of rejection on the day of ι ν injection (day 7) and the following 3 days Not shown are the control experiments performed in each donor-recipient combination with administration of alloantise-rum alone, RC, GPC, or HC alone, or nonspecific alloantiserum along with RC In no case was AAR observed and the MSTs of these ex

periments did not differ significantly from the MST's of the ALS-treated animals

Non-H-2 differences (combinations 1, 2, and 3) We were not able to induce AAR across an Η-Y barrier nor across an H-3, H-9, and H-13 barrier In the B6AF, male onto female combination it was, however, possible to reduce the survival t ime of the skin grafts from 126 to 99 8 days by administration of alloantiserum and RC Microscopic examination 11г, 2, and 8 hr after ι ν injection of antiserum and RC did not show any signs of rejection In the second combination, C57BL6 onto BIO LP, where macro-scopically no signs of rejection occurred, we found microscopically 2 hr elfter administration of antibodies and RC a slight increase of granulocytes in the blood vessels of the graft and in the surrounding tissues of the graft with focal extravasation of erythrocytes Eight hours after the injection, these rejection signs were already less extensive This suggests the induction of a reaction that was not severe enough to cause visible rejection of the graft When skin was grafted across a great number of non-H-2 differences (combination 3), the MST of the untreated animals came into the range of

TABLE 3 AAR of mouse skin allografts in strain combinations with increasing histoincompatibihty

Combi nation

1 2

3

4

5

6

7

8

9

Donor -» Recipient (M-.F)

B6AF, (M) - B6AF, (F) C57BL6 — BIO LP

BIO D2 -» BALB/c

BIO Br - B6AF,

A/HeJ - CBA/J

BIO D2 ^ B6AF,

DBA/2 -. B6AF,

BIO D2 — C57BL6

DBA/2 -> C57BL6

MST in untreated (I) and immunosuppreseed

(II)

No of animals

HO

из 112

II 20

no II 20 110

1110

115 II 18

no ino

no и i3

115 II 14

animals"

Days ±

>120 15 4 ±

10 4 ± 24 0 ±

12 8 ± 31 9 ±

9 8 ± 24 0:!:

10 0 ± 21 2 ±

9 5 ± 22 0 ±

9 3 ± 19 0 ±

85 ± 17 8 ±

SD

5

1 1 2

RC

Antiserum (ml)

2 5" 2 5"

0 5 10 15 10 15 0 01 0 05

0 01 0 05

0 01 0 05

0 005 0 01

0 005 0 01

b

AAR'

0/15 0/8

0/6 0/11 6/6 0/12 8/8 0/6 6/6

0/6 6/6

0/6 11/11

0/8 8/8

0/8 6/6

JFC

Anti serum

(ml)

15

15

05 10 15 05 10 15 2 0 05 10 15 0 01 0 05

0 01 0 05

ND' ND

ъ

AAR'

0/7

0/β

0/7 0/6 7/7 0/7 0/8 0/9 0/6 0/6 0/6 6/6 0/5 8/8

0/5 6/6

HC'

Anti serum (ml)

15

15

15

15

15

05

05

ND

ND

AAR'

0/5

0/5

8/8

0/8

8/8

10/10

8/8

" Immunosuppression by ι ρ injection of 0 25 ml of GAMLS on day 0, 2, and 4 after grafting 6 AAR was induced by ι ν injection of listed volumes of antiserum along with 0 25 ml of RC, 0 5 ml of GPC, or 0 5 ml of

HC on day 7 after grafting r Number of mice that show complete graft necrosis within 24 to 72 hr after injection/number of mice tested d 1 5 ml was given ι ν and 1 0 ml ι ρ ' ND, not done

9 5

TRANSPLANTATION

MST's of grafts acrctes an H-2 b a r n e r In this combination AAR could be induced with 1 5 ml of antiserum and RC Lower doses of antiserum did not cause AAR, although there was a decrease of several days in the MST of the skin grafts With GPC and HC no AAR could be induced

H-2 differences (combinations 4, 6, and 8) Induction of AAR across an H-2D incompatibility was only possible with 1 5 ml of alloantise-rum and RC, while across an H-2K incompatibility it was obtained with 0 05 ml of alloantise-rum If there existed a complete H-2 complex disparity, the amount of alloantiserum necessary for induction of AAR with RC was further reduced to 0 01 ml In this latter combination it was also possible to induce AAR with GPC and HC but the dose of alloantiserum had to be higher than with RC

H-2 and non-H-2 differences (combinations 5, 7, and 9) AAR with RC of skin grafted across H-2D plus non-H-2 differences or across H-2K plus non-H-2 differences was obtained with 0 05 ml of alloantiserum whereas only 0 01 ml was necessary for rejection of a graft differing at the whole H-2 complex plus non-W-2 loci Thus, the introduction of non-H-2 differences did not lead to a lower amount of alloantibody necessary for AAR with RC The A/HcJ -» CBA/J combination formed an exception, but as already mentioned in this combination not only H-2 differences are introduced but there is also a change in the H-2 difference compared with BIO Br -> B6AF, For GPC and HC we found that the amount of alloantiserum required for AAR decreased when non-H-2 differences were introduced In BIO D2 onto BBAF, we could not induce AAR with 2 0 ml of alloantiserum, but in the DBA/2 onto B6AF, combination 1 5 ml of antiserum given with GPC or HC was effective This must be the result of the cumulative effect of non-H-2 differences In the in vitro lymphocytolysis we observed a similar pattern The addition of non-H-2 differences to H-2 differences increased the cytotoxicity if the antiserum was titrated with GPC and HC, but the titer with RC did not change (Table 2) Figure 1 gives a graphical summary of the results of this study

DISCUSSION

Our results show that the sensitivity of skin grafts to the destructive effects of alloantibody and heterologous complement increases with

the extent of antigenic differences between donor and recipient In a combination with a few non-H-2 differences, we found only microscopic signs of tissue damage, even if a high amount of specific antiserum was administered In the combinations with H-2 differences, К end-in-compatible grafts were more easily destroyed than D end-incompatible grafts, whereas with complete H-2 and non-H-2 differences a violent rejection could be induced with minimal amounts of antiserm This increase in sensitivity was accompanied by a gradual decrease in MST of the skin grafts in untreated animals in which rejection is presumed to be mediated mamly by cellular mechanisms The decrease in MST that occurs when the antigenic differences become greater has been described earlier (7, 20, 21), and it has been shown in those studies that the MST can even be used as a measure for the degree of immunogemcity of the graft The rules of immunogemcity that have been found for these cellular immune responses (7) also apply to our model where the destruction is mediated by humoral mechanisms

A second aspect that deserves to be mentioned is the finding that there were clear-cut differences in the activity of a given antiserum if tested with different complement species The lower in vitro activity of GPC compared with RC has been described by several authors (4,9, 10,17) and is confirmed in this study Furthermore, we found that HC behaves more or less like GPC in this respect In all of the combinations studied, we found similar differences in complement activity in vivo RC was always more active than GPC or HC These two latter complement species only did work in combinations with strong antigenic differences The argument about the relationship between in vitro and in vivo complement activity extends to mouse complement This complement species is very inefficient in vitro, and was apparently not sufficiently activated in our in vivo system to induce antibody-mediated destruction

The correlation between in vitro and in vivo activity raises the question whether the mechanism of AAR is exactly the same as the in vitro cytolysis, or in other words, whether the complement pathway has to become completely activated to induce AAR Winn et al (27) found in their model, where mouse complement was the active complement, that in C5-deficient mice the process was delayed, while severe de-

9 6

BERDEN ET AL

Minimal volume of a n t i s e r u m ( m l )

lo

i s 1

O l 0 05

О 01

• RC

D GPC

Ш не

и (cytotoxic t i t e r ) ' of a n t i s e r u m

4 0 9 6 -

2 0 4 8

1024 -

512 -

256

12Θ -

6 4

32

16

8 -

4 -

2 -<2 -

Histocompatibil ity barr ier

Η-Y H-3 H-2D multiple H 2K H 2D H-2K H 9 non H-2 plus plus H-13 ηοπΗ2ηοπΗ-2

H-2 H-2 plus

non H-2

MST untreated animals (days) >126 15 4 12 8 10 4 10 0 9 8 9 5 9 3 8 5

FIGURE 1 Acute antibody-mediated rejection in strain combinations with increasing histoincompatibil-ity Relationship between MST of grafts in untreated recipients, in vitro cytotoxic titer of antiserum, and minimal amount of antiserum required to induce AAR with different complement species (RC, GPC, and HC)

pletion of granulocytes could abolish rejection.

They concluded from these observations that

the role of complement is merely one of gener

ating the chemotactic substances C3a and C5a

Histological studies show that in our model

similar mechanisms play a role, since we found

earlier a dense accumulation of granulocytes in

the graft after injection of antibody and comple

ment (76) It is, however, not excluded that, in

our experiments with heterologous comple

ment, direct lysis of the graft attributable to

complement activation beyond C5 played an

important role. Further support for this as

sumption stems from our unpublished findings

that with complement from C6-deficient rab

bits, antibody-mediated graft destruction could

not be induced and that the destructive capac

ity was reestablished after reconstitution of

this serum with C6. It might be argued that C6

also contributes to the leucotactic activity in

the C567 complex (25), but there is evidence

(23, 24) that the leucotactic property of C6-

deficient serum is unimpaired For the time be

ing, this problem remains unresolved and

awaits studies with sera that are deficient only

in the terminal components of the complement

9 7

TRANSPLANTATION

system, such as C9.

In clinical transplantation little is known about the role of non-HLA differences in the acceptance of allografts There are a few reports of severe rejection episodes of renal grafts from HLA mixed lymphocyte culture-identical siblings U9, 22). Kano et al. (13) reported that in recipients with a rejection of an HLA-identical kidney, antibodies could be detected against non-HLA antigens. From models in mice it is known that, if there exist multiple non-H-2 differences, skin graft survival can be equivalent to that of grafts differing at the complete H-2 complex (7). In our study it was possible to induce AAR across only non-H-2 differences. These differences can, therefore, not only cause cellular graft rejection but can also be responsible for antibody-mediated graft damage. This phenomenon should be bome in mind when the antibody-mediated rejection in clinical transplantation is studied.

Acknowledgments The technical assistance of the staff of the animal laboratory (Head; Dr. W. J. I v. d Gulden) is gratefully acknowledged We thank Mrs. J. Koene-Bogman for performing the histological examinations.

LITERATURE CITED

1. Bevan M J· 1976 Immunogenetics 3: 177 2 Carpenter CB, d'Apice AJF, Abbas AK: 1976

Adv Immunol 21 1 3. Demant Ρ- 1973 Transplant Rev 15: 162 4 Detnck-Hooks B, Borsos T, Rapp HF: 1975 J

Immunol 114: 287

5. Fabre JW, Morris PJ 1974 Transplantation 18: 429

6 French ME. 1972 Transplantation 13: 447 7. GrafTRJ, Bailey DW: 1973 Transplant Rev 15- 26 8 Graff RJ, Snell GD: 1969 Transplantation 8: 861 9 Grant CK: 1976 Transplantation 21: 323

10 Haughton G, McGhee MP. 1969 Immunology 16: 447

11. Jooste SV, Winn HJ: 1975 J Immunol 114: 933 12. Jooste SV, Winn HJ, Russell PS: 1973 Trans

plant Proc 5· 713 13. Капо К, Kussmaul CA, Milgrom F: 1975 Trans

plant Proc 7· 185 14. Kissmeyer-Nielsen F, Olsen S, Petersen VP, et

al· 1966 Lancet 2: 662 15. Klein J: 1973 Transplantation 15 137 16. Koene RAP, Gerlag PGG, Hagemann JFMH, et

al 1973 J Immunol 111: 520 17. Koene RAP, McKenzie IFC: 1973 J Immunol

111: 1894 18. Litchfield JT: 1949 J Pharmacol Exp Ther 97: 399 19. Lundgren G, Magnussen G, Moller E, et al: 1972

Tissue Antigens 2: 32 20. McKenzie IFC, Snell GD: 1973 J Exp Med 138:

259 21. McKenzie IFC, Snell GD: 1973 Transplantation

17. 328 22. Salaman JR, Godfrey AM, Russell RB, et al:

1976 Tissue Antigens 8: 233 23. Snijderman R, Phillips J, Mergenhagen SE:

1970 Infect Immun 1: 521 24. Stecher VJ, Sorkin E: 1969 Immunology 16: 231 25. Ward PA, Cochrane CG, Muller-Eberhard HJ:

1966 Immunology 11: 141 26. Winn HJ: 1965, ρ 133 In Wolstenholm GEW,

Knight J (eds). Ciba foundation symposium on complement. Churchill Ltd, London

Received 18 January 1977. Accepted 5 May 1977.

Chapter V

THE ROLE OF COMPLEMENT FACTORS IN ACUTE ANTIBODY-MEDIATED

REJECTION OF MOUSE SKIN ALLOGRAFTS

J.H.M. Berden, P.J.A. Capel, and R.A.P. Koene



published in: European Journal of Immunology (1978) S, 158

99

J.H.M Berden, Ρ J.A. Capel and

R.A.P. Koene

Department of Medicine, Division of

Nephrology, Sint Radboud Ziekenhuis,

University of Nijmegen, Nijmegen

The role of complement factors in acute antibody-mediated rejection of mouse skin allografts*

Established, Η 2 incompatible, mouse skin allografts are desiroyed in 2 4 - 4 8 h by a single intravenous injection of specific alloantibody and heterologous complement (C) from rabbits (RC), guinea pigs (GpC) or humans (HC) Mouse complement (MC) itself is inefficient in this respect since grafts are not destroyed by the injection of antibody alone This is in keeping with the low efficiency of MC in in vitro cytolysis

We have studied the role of different С factors in this model Two patterns were observed (a) Three C-deficient sera (GpC-R4, GpC-R3, RC-R6) showed neither activity m vitro (C-mediated lymphocytoloxicily in a trypan blue assay) nor did they induce graft destruction Activity in vitro and in vivo could be completely reconstituted with the appropriate С component (b) Two C-deficient sera (GpC-R 1 and HC-R9) were inactive ш in vitro cytotoxicity but, nevertheless, induced destruction of the skin grafts if injected with antibody Thus, in vivo, the mouse seemed to provide the deficient factor, and this was confirmed by the observation that m vitro cytotoxicity could be reconstituted not only with the appropriate С component, but alto with normal mouse serum The results show (hat the inefficiency of MC m vitro and in vivo does not reside in the factors CI or C9 Furthermore, in our model С is activated via the classical pathway, and at least exogenous C4, C3 and C6 are required to induce acute antibody-mediated graft destruction

From the experiments with C6-deficicnt serum which has normal chemo-tactic properties, it is concluded that activation of chemotactic factors with attraction of granulocytes is not sufficient to induce rejection, but that the activation of the membrane attack unit is required

1 Introduction

The rejection of skin grafts in the mouse is considered a primarily cellular immune event The antibody-mediated destruction of mouse skin allografts is of less importance, as is shown by the failure of passively administered allo-antibodies to induce graft rejection [ I ] This is at least partly due to the inefficiency of the mouse complement (MC) system We have previously shown that well-established H-2-incompatible allografts earned by immunosuppressed mice can be quickly (within 24 h) destroyed by an intravenous (ι ν ) injection of alloantiscmm together with rabbit complement (RC) [ 1 ] There are several arguments which support the assumption that the exogenous С is necessary for this antibody-mediated graft rejection First, the histology of the grafts shows the charactenstics of an Arthus reaction [ 1 ] which is known to be C-dependent Second, after inactivation of the С system, by heating, treatment with zymosan, carra-

[I 1873]

* Tins study was supported in part by grants from the Netherlands Poundation of Medical Research FUNGO, and from the Netherlands Kidney Foundation

Correspondence: J Η M Berden, Department of Medicine, Division of Nephrology, Sint Radboud Ziekenhuis, Geert Grooteplein Zuid 16, Nijmegen, The Netherlands

Abbrevia lions: AAR Acute antibody-mediated rejection ALS: Anti-lymphocyie serum GpC· Guinea pig complement GVB++: Veronal-buffered saline with gelatin, Ca++ and Mg++ Hanks' BSS. Hanks" buffered salt solution HC- Human С MC: Mouse С MST: Median survival time NMS: Fresh normal mouse serum RC* Rabbit С Ri С deficient in Cl R4· С deficient in C4 R3: С deficient in СЭ R6: С deficient in C6 R9: С deficient in C9 RSO Relative salt concentration SD. Standard deviation S RBC: Sheep red blood cells

geenan or aggregated human IgG, the rabbit serum no longer induced acute graft destruction if injected with antibody [2] Third, acule rejection was only seen with C-fixing antibodies (7 S IgGj) and not with 7 S IgG l antibodies that do not fix

с[з] Heterologous С is apparently not required in all skin grafting systems in mice Winn et al [4, S] could destroy rat skin grafted onto mice by injection of mouse anti-rat serum or by rat alloantiserum without exogenous С However, m none of our allogeneic skin graft experiments was MC sufficiently activated, since injections of alloantibodies alone never induced graft destruction [6] This is in keeping with the very low activity of MC in m vitro lymphocytotoxicity tests with mouse alloantibodies [7]

The donor recipient combination (BIO D2 ->C57BL/6) used in the current study, consists of a complete H 2 complex disparity in which we can induce acute antibody-mediated rejection (AAR) by the injection of RC, guinea pig complement (GpC), or human complement (HC) together with alloantibody [6]

We have in this model tested various sera which were made deficient in isolated С factors, and sera from animals that were congenitally deficient in certain С factors We have used these С factor-deficient sera (R sera) to study the role of individual С components in our model of AAR Furthermore, this approach could delineate which components m the mouse endogenous С system are responsible for its low efficiency in cytotoxic tests and in AAR For this purpose we compared the m vivo results with the lymphocytotoxic activity of the various R sera in the presence or absence of normal mouse serum (NMS)

100

Complement factors in humoral rejection of skin grafts

2 Materials and methods

2 1 Animals

Inbred BIO D2/new Sn (H-2d) mice were originally obtained from the Jackson Laboratory (Bar Harbor, ME) C57BL6/Rij (II 2 b ) mice were obtained from the Radiobiological Institute TNO (Rijswijk, NL) C4 deficient guinea pigs came from the National Institutes of Health, (Bethesda, MD) The C6 deficient rabbits were provided by Dr Lachmann (Cambridge, G B )

2 2 Production of alloantiserum

Alloanliserum was prepared by injecting female C57BL/6 mice at weekly intervals, intrapenloneally (ι ρ ), with a suspension of approximately 5 χ IO7 BIO D2 lymphoid cells in complete Freund's adjuvant After five injections, most animals developed ascites and could be tapped every 1 to 2 weeks The ascitic fluid of several weeks was pooled and, before use all antisera were heated at 56 0 C during 45 mm and sterilized by passing them through a sterile 0 20 mji filter The cytotoxic titer of the ascitic fluid was the same as the titer in the serum of these mice Antisera were stored at - 20 0 C

2 3 Anti-lymphocyte serum (ALS)

This was prepared by a subcutaneous injection of 5 χ IO 7 -8 χ IO7 C57BL/6 lymphoid cells (thymus and mesenterial lymph nodes) in complete Freund's adjuvant into a goat This priming dose was followed by weekly ι ν boosters of the same number of C57BL/6 lymphoid cells The animals were bled one week after the fourth immunization, the serum heated at 56 0 C for 45 mm, and absorbed once with C57BL/6 red blood cells (5 vol of serum/1 vol of packed red blood cells) After absorption, this ALS was not toxic to the recipients if injected ι ρ

2 4 Complement

Fresh or fresh frozen sera from New Zealand white rabbits, random bred SPF Albino guinea pigs, and from a human volunteer were used as sources of С Blood was obtained in rabbits and guinea pigs by cardiac puncture, in the human by venous puncture After clotting for 45 mm at room temper ature, and centnfugation the serum was divided into aliquots of 0 5 ml and, if not immediately used stored at - 3 0 0 C for maximally 3 weeks С was used only once after thawing Be fore use all sera were tested for toxicity m vitro (a 1 4 dilution should give less than 10 % cytolysis of В10 D2 lympho cytes), and in vivo by ι ν injection of 0 5 ml serum

2 5 Reagents

Veronal-buffered salme (GVB + + ), pH 7 35, containing 0 1 %

gelatin 0 15 mM Ca',"r and 0 5 m M Mg+ +, relative salt concentration (RSC) 0 147, was prepared accordmg to Mayer [8] The buffer was freshly prepared each day The rabbit anli-sheep red blood cell antiserum and rabbit anti-human C9 were commercially obtained from Behrmgwerke, Marburg, FRG Sheep red blood cells (SRBC) were drawn monthly from the same sheep, and collected in an equal volume of

Alsever's solution Before use, SRBC were washed four times in ten volumes of G V B + +

2 6 Production of R sera

The Rl serum was prepared from pooled guinea pig serum by euglobulin precipitation (pH 7 5, RSC 0 04) accordmg to the method of Nelson [9] The precipitate was redissolved m a volume of GVB + + equal to the original serum volume The С1 free supernatant was brought back to a RSC of 0 15 with NaCl after concentration to the original volume R4 serum was obtained from C4-deficient guinea pigs R3 serum was prepared by incubation of guinea pig serum with boiled zymosan (Koch Light Laboratories, Colnbrook, Bucks , GB), 15 mg/ml pooled guinea pig serum, for 60 min at 37 0 C After incubation, the zymosan was spun down in a refnger-ated centrifuge The R6 serum was obtained from C6-de fluent rabbits R9 serum was prepared by absorption of C9 from human serum by Sepharose coupled rabbit anti-human C9 antiserum The absorption procedure was performed at 0 0C in the presence of 0 01 M fcDTA and was repeated until, after reconstitution with Ca + + and Mg't"t\ hemolytic activity was no longer detectable R9 serum was concentrated to the original volume at 0 0 C The R sera were sterilized by passage through a 0 20 π\μ filter AU R sera, when tested in a microhemolytic assay, were unable to lyse sensitized SRBC, while they were functionally active after reconstitution with the appropnate factor(s)

2 7 Reconstitution of the R sera

The reconstitution experiments were performed with functionally pure GpC and HC factor preparations from Cordis Ltd (Miami, FL) In reconstitution experiments, Rl was reconstituted with the in G V B + + dissolved euglobulin precipitate, R4 with GpC4, R3 with guinea pig R4, R6 with GpCó and R9 with HC9 The R sera were serially diluted in GVB + + or Hanks' buffered salt solution (BSS) to which a constant amount of the deficient С factor had been added R l , R3 and R9 serum and the reconstituted sera were always compared in the same test with the batch of С from which they were prepared Th R4 serum was compared with pooled normal guinea pig serum, and the R6 serum with pooled normal rabbit serum In the lymphocytotoxicity tests an attempt was also made to reconstitute the R serum with NMS This was done in a two stage procedure to eliminate the anticomplementary activity that is often present in NMS No lymphocytotoxicity was found if NMS alone was used as a source of С

2 S Serology

A microhemolytic assay was performed on microliter plates Twenty five μ\ of the test sample were serially diluted in 2 5 y l G V B + + Thereafter, 25 ді of SRBC (2 χ 10B/ml) sensitized with anti SRBC antiserum (diluted 1 600) were added After mixing, the plates were incubated for 30 mm at 37 0 C After incubation, the plates were placed on ice to stop the reaction ¿nd centnfuged for 1 min in the cold The 5 0 % lysis point was taken as the titer R sera were also tested in a lymphocytotoxicity assay using the trypan blue exclusion test BIO D2 spleen cells were used as targets prepared free of red cells with Tm-NH^CI One hundred μΐ of cells (5 χ 105) in Hanks' BSS were incubated during 30 mm at 37 0 C with

101

J H M Berden, Ρ J A Capel and R A Ρ Koene

equal volumes of CS7BL/6 anti В10 D2 antiserum (diluted 1 16) and senal dilutions of whole C, R serum, reconstituted R serum, or С factors The titer was defined as the dilution giving 50 % cell lysis Tests were done in triplicate, and the following controls were performed (a) cell control 100 μ) cells in 200 μΐ Hanks' BSS, (b) antibody control 100 μ\ anti serum (diluted 1 16), 100 μΐ Hanks' BSS and 100 μΐ cells, (с) С control 100 μΐ undiluted C, R serum, reconstituted R serum, or pure С factor with 100 μΐ Hanks' BSS and 100 μΐ cells In these controls cytolysis did not exceed 10 %

2 9 Skin grafting technique

Female BIO D2 tail skin was grafted onto the nght dorsal flank of female C57BL/6 recipients by a modification of the fitted graft technique The graft was fixed by Nobecutane spray Instead of plaster badage, we used band-aid to cover the graft The donor and recipient mice were between 6 and 10 weeks old The band aid was removed on day 6 after transplantation and the fate of the grafts followed by daily macroscopic inspection Grafts were considered to be rejected when no viable epithelium remained Median survival time (MST) and the standard deviation (SD) were calculated by the method of Litchfield [ 10] The level of significance was calculated with Student's t-test

2 10 Acute rejection studies

The experiments were performed on well established grafts that showed no signs of rejection Survival of grafts was prolonged by nonspecific immunosuppression with ALS (0 25 ml ι ρ on days 0, 2 and 4 after grafting) from 9 3 + 1 1 days to 19 0 ± 1 1 days (Table 1) Thus, an ongoing cellular rejection could not interfere with the antibody-mediated destruction induced at day 7 after grafting and completed within 2 4 - 7 2 h On day 7 the recipients received 0 5 ml C57BL/6 anti-B10 D2 antiserum and 0 5 ml C, R serum, С factor, or reconstituted R serum by an ι ν injection through the tail vein Control experiments were performed by injecting antiserum, C, R serum and reconstituted R serum, alone, or the deficient С factor and alloantiserum AAR was defined as the complete necrosis of the skin graft 2 4 - 7 2 h after the iv injection

3 Results

3 1 Lymphocytotoxic activity of the R sera in vitro

The titration curves of the va nous R sera are given in Fig I

Rl serum The R l serum showed no cytotoxicity towards BIO D2 cells sensitized with alloantiserum After reconstitution with CI (100 μΐ of the redissolved euglobulin precipitate/ 100 μΐ R l ) , the cytotoxic titer became 1 24 This is close to 1 32, the original titer that was found with the batch of pooled guinea pig serum from which the R1 and С1 fractions were prepared The GpCl fraction itself had no cytotoxic activity If the cells were incubated with alloantiserum and NMS and washed, subsequent addition of GpC R l also produced a cytotoxicity of 1 24 Reversal of the procedure, ι e

incubation of cells with alloantiserum and GpC-RI, followed by washing and addition of NMS, did not result in cytotoxicity This confirms earlier reports that the reconstituting

factor in NMS is С1 [16], and is further supported by our finding that an euglobulin precipitate of NMS was as active in the reconstitution of GpC-RI as whole NMS

>90-Θ 0 -

6 0 -

4 0 -

20-<10-

• — ^ ^ч

^ 5 β ι , α ^ ν ,

В

ч

•ν *л х Л л\ Ч

RI

»90-80·

6 0 -

4 0 -

20-<10-

S "Ч итз ч и

\ \

R3

ч,

- ! 1 7—I 1 1 Г-

НИ* ti

Figure 1 Lymphocytotoxic activity of the various R sera against BIO 02 spleen cells sensitized with C57BL/6 anü BIO D2 antiserum Each diagram represents the activity of normal C, R serum, reconsti tuted R serum, R serum in the presence of NMS, or the deficient factor

102

Complement factors in humoral rejection of skin grafts

R4 serum Neither the R4 serum nor punfted GpC4 (12S0 CH50 U/ml), or GpC-R3 showed cytotoxic activity against BIO D2 spleen cells Reconstitution of 100μ] GpC-R4 with 100 μΐ GpC4 ( 1250 CHso U/ml) or with 100 μΐ GpC-R3 restored cytotoxicity up to 1 24 It was not possible lo reconstitute GpC-R4 with NMS

R3 serum With the R3 serum no cytotoxicity was found, while addition of excessive amounts of GpC-R4 restored the cytotoxicity to the same titer as that of pooled GpC Reconstitution of the R3 serum with NMS was not possible R6 serum Neither the R6 serum norGpC6 (2500 CHSo U/ml) were able to lyse sensitized B\0 D2 spleen cells Retonbtitu-tion of 100 μΐ Rt-R6 with 100 jul GpC6 ( 1250 CHso U/ml) restored the cytolytic activity up to 1 12, while normal RC in this (est system had a cytotoxic titer of I 32 Thus, we found a functional synergism between RC-R6 and ОрСб, and this is analogous to the findings of Rother et al who could reconstitute RC R6 with purified HC6 [11] Reconstitution with NMS was not possible

R9 serum In the experiments with R9 serum the same pattern was found as with GpC-Rl The R serum itself had no cytotoxic activity, neither had pure HC9 (5000 CHso U/ml) When 100 μ1€9 (5000 tH S o U% I) were added to 100 μΐ R9, the cytotoxicity was restored up to a titer of I 12, while nor mal HC had a titer of 1 16 NMS was also able to restore the cytolytic activity of human R9 up to a titer of 1 12, if NMS was added after incubation of sensitized target cells with the R serum

3.2 Ability of R sera to induce AAR

By the administration of 0 S ml RC, GpC, or HC together with 0 5 ml alloanlisemm, AAR could be induced within 48 h,

while alloantiserum alone, or the various С species alone gave negative results The MST of the control groups did not differ significantly from the ALS-treated group (Table 1) The in vivo results with the different R sera are summarized in Table 2 It was possible to induce AAR with GpC-R 1, while injection of R serum without alloantiserum was ineffective A similar pattern was observed with the HC-R9 serum

Table 1. Occuircnce of AAR and MST of BIO D2 skin giafu on C57BL/6 recipients m va no us control groups

No of ¡murali

150 13-1)

10

16 10

10 β

10 11

C57BL/6 anti-BIO D2 antlierum1)

-

+ +

+

+

Complément·)

-

-RC RC

GpC GpC

HC HC

AARb)

-+

+

+

MSTiSD

9 3 i 1 1 1 9 0 i 1 1

23.0 1 1 1

18 6 1 1.1

2 1 5 1 1 1

19 4 t 1.1

a) 0 5 ml ι ν on day 7 after grafting b) Defined as complete necrosis of the skin graft 24-72 h after 1 ν

injection c) Untreated animals d) Recipients received 0 25 ml ALS ι ρ on days 0, 2 and 4 after graft

ing

Table 2 Induction of AAR by 1 ν injection of the various R sera and alloantiserum on day 7 after grafting

R serum AARa) R serum and R senim Reconstituted Reconstituted alloandbody only R serum and R serum only

alloantibody

GpCRl 7/ 7b) 0/8 HC-R9 6/ 6 0/6

GpC-R4 0/12 11/11 0/8 GpC-R3 Q/ll Π/11 0/8 RC-R6 0/11 10/10 0/8

a) Defined as complete necrosis of the graft 24 72 h aften ν injection

b) Number ot animals that showed complete necrosis of Mie graft/ number animal* tested

Administration of GpC R4 with alloantiserum did not induce AAR Addition of pure GpC4 to the R4 serum before the injection did not reconstitute the destructive capacity of the R serum Even doses as high as 5000 CH50 U were ineffective This was somewhat surprising since m vitro, the lymphocyto-toxicity of the R4 scrum was completely restored by C4 A possible explanation might be that after the injection of the reconstituted R serum, GpC4 is rapidly inactivated in the mouse circulation [12] However, AAR was obtained with R4 serum that had been reconstituted with guinea pig R3 serum Injection of this mixture without alloantiserum did not lead to AAR

The R3 serum did noi induce AAR if injected together with alloantiserum Active C3 is not available since its isolation leads to inactivation Thus, in this instance wc were unable to reconstitute the defect with the appropriate factor How ever, in the experiments with R4 serum it had already been shown that combination of R3 and R4 resulted η compie mentation and led to induction of AAR

Induction of AAR with R6 serum and alloantiserum was also impossible Reconstitution of 0 5 ml R6 serum with an equal amount of guinea pig C6 ( 1250 CH50 U/ml) lead to destruction of the graft if the mixture was injected with alloantiserum In the control experiments, injections of only C6 with alloantibody or reconstituted R serum alone did not result in rejection

4 Discussion

Our results show that for the induction of AAR in our model not all heterologous С factors have to be administered GpC Rl and HC R9 were as effective as normal GpC and HC. showing that the mouse can reconstitute the defect On the other hand, the limiting effect of MC does not he at a single point in the С sequence since three sera deficient in different С components were only able to induce rejection after rcconsti tution with the appropriate factor

In all but one experiment (R4 reconstituted with C4) we found in R sera a close correlation between С activity m intro and tn vivo AAR occurred only if the R serum had cytotoxic activity in vitro after reconstitution with the deficient factor,

103

J H M Berden, Ρ J A Capel and R A Ρ Koene

or with NMS This suggests that the С system is activated similarly in vivo and in vitro From the expenments with GpC*R4 it can be concluded that m vivo С activation in our model is initiated via the classical pathway, as it is ш іл vitro cytotoxicity of cells sensitized with antibody If alternative pathway activation alone were sufficient to induce AAR, then administration of the R4 serum should have induced rejection, since activation of the alternative pathway is normal in this serum [13] At least three murine С factors, C4, C3 and C6, are relatively inefficient in this system since NMS could not reconstitute these R sera neither m virro nor m vivo With regard to C4 and C3 these findings are in accord with those of Rice, who in a hemolytic system found that CpC-R4 and R3 were not reconstituted by NMS [14] Mouse CI, on the other hand, seems not to be at fault NMS and also its euglobuhn precipitate, reconstituted the GpC-RI serum in vitro This complémentation confirms previous findings of other investigators who obtained similar results in a hemolytic system with sensitized SRBC [ 1 4 - 1 6 ] Since the Rl serum tould induce graft rejection without reconstitution with exogenous С1, we must conclude that also in vivo the mouse effectively provided the deficient CI factor

It has been suggested by Winn [4] that for antibody-mediated graft destruction, activation of the complete С pathway is unnecessary He assumed that the generation of chemotactic С components is sufficient for the destruction of the graft This was based on expenments with CS-deficient mice that rejected rat xenografts after the administration of mouse anti-rat serum These mice are completely deficient in CS, and thus activation beyond this component is impossible Care was taken to prevent administration of exogenous С components with the anti-rat serum On the basis of his findings, we expected that our R6 serum would also induce rejection, but this was clearly not the case This cannot be the result of deficient chemotactic properties of the R6 serum, Since it has been shown to generate normal Chemotaxis [ Π Ι 9] and lacks only lytic activity We have preliminary unpublished data from histological studies that after administration of R6 serum, the granulocyte infiUration in the graft is as abundant as after the injection of whole RC Despite this, destruction did not occur with R6 We therefore assume that in our model complete С activation with generation of not only chemotactic but also membranolytic activity is necessary to induce AAR

We think that the only possible explanation for the induction of AAR by R9 serum is that the mouse endogenous C9 is efficient enough to complement exogenous R9-serum especially after this is what happened in the in virro test It is highly unlikely that graft destruction occurred without activation of C9, but this possibility could not be formally excluded through lack of a negative control in the m vivo experiments This would require a mouse completely deficient in C9, and in such an animal the R9 serum would fail to induce AAR

We may thus conclude that for the induction.of AAR, the complete С pathway has to become activated via the classical pathway, and that generation of chemotactic С components alone is not sufficient Our study gives some indications for a useful approach to influence the occurrence of antibody-mediated graft destrULtion by manipulation of the С system If it becomes possible to inhibit С activity in vivo it seems wise to concentrate on the early factors such as С1 or C4 because deficiency of these factors might lead to inhibition of humor ally mediated rejection Such an approach would leave the alternative pathway fully intact so that the C-dependent defense mechanisms against bacterial infections remain unaltered This would be of great ad\antage in clinical transplantation

The skillful technical assistance of Ms Jacqueline Hage mann and the staff of the animal laboratory (Head Dr WJ/ van de Cuiden) is gratefully acknowledged

Received August 30 1977

5 Reference*

1 Koene, R A P , Gerlag Ρ G G , Hagemann, J h H M , van Haelst, UJG andWydcveld.PG A B , / Immunol 1973 JIJ 520

2 Gerlag, Ρ G G , Thesis University of Nijmegen 1975 Nijmegen

3 Jansen, J L J Koene, R Α Ρ , van Kamp, G J , Tamboer W Ρ M andWijdeveld, PC AB,У Immunol 1975 /75 3β7

4 Winn, Η J , Β al da mu s, С А , Jóos te, S V and Russell P S , / Exp Med 1973 137 893

5 Jooste.SV and Winn, H J, ƒ Immunol 1975 J14 933

6 Berden J H M Gerlag, Ρ G G , Hagemann, J F H M and Koene, R AP, Transplantation 1977 24 175

7 Grant CK , Transplantation 1976 21 323

β Kabat, E A and Mayer, M M , Experimental Immunochermstry, 2nd Edit, Thomas, С Γ , Springfield 1961

9 Nelson R A , Jensen, J Gigli, I and Tamura, N , Immunochem istry 1966 S 111

10 Litchfield, J Τ, У Pharmacol Exp Ther 1949 97 399

11 Rother, К , Rother, Ь Muller Eberhard, Η J and Nilsson, U R , J Exp Med 1966 ¡24 773

12 Mandle, R J , McConell Mapes, J A and Nilsson, L R J Immunol 1977 ¡19 180

13 brank, M M , May J , Gaither, Τ and Ellman. L J Exp Med 1971 J 34 176

14 Rice.CE andCrowson C N , y Immunol 1950 65 201

15 Brown, G С J Immunol 1943 46 319

16 Winn H J in Wolstenholme, G E W andKmght.J (fcdsi.Oto Foundation Symposium on Complement Churchill Ltd , London 1965.ρ 133

17 Stecher, VJ and Sorkin, E .Immunology 1969 ¡6 231

18 Snyderman, R , Philips J and Mergenhagen, S Ь , Infect Immun 1969 У 521

19 Lim, D . Gewürz A Lint, Τ F , Ghaze, M , Scphen В and Gewürz, H, J Pediat 1976 89 42

104

Chapter VI

A SENSITIVE HAEMOLYTIC ASSAY OF MOUSE COMPLEMENT

J.H.M. Berden, J.F.H.M. Hagemann, and R.A.P. Koene



published in: Journal of Immunological Methods (1978) 23, 149

105

A SENSITIVE HAEMOLYTIC ASSAY OF MOUSE COMPLEMENT *

J.H.M. BERDEN, J.F.H.M. HAGEMANN and R.A.P. KOENE Department of Medicine, Division of Nephrology, St. Radboud Ziekenhuis, University of Nijmegen, Nijmegen, The Netherlands (Received 1 March 1978, accepted 20 March 1978)

Under appropriate conditions the haemolytic activity of mouse complement can be measured in a 5ICr release assay. Modifications that increase the sensitivity are the following: (a) mouse 7 S anti-SRBC antibodies as amboceptor; (b) low amount of target cells; (c) ionic strength of 0.13 M NaCl in the test medium; (d) incubation temperature of 30oC; (e) incubation time of 90 min; (f) total reaction volume of 1 ml.

INTRODUCTION

In tests that are generally used to measure the haemolytic activity of human and guinea pig complement, mouse serum seems to lack complement activity (Brown, 1943; Rice and Crowson, 1950). In 1952 McGhee reported that in a system with a very low number of sensitized sheep red blood cells (SRBC) one CH50 U/ml could be detected in some mouse sera, while other sera from mice of the same strain were still unable to induce haemolysis.

Several authors claimed that some of the then known complement factors (CI, C4, C2, C3) were absent from the mouse complement system, and that, therefore, no haemolytic activity could be demonstrated (Ritz, 1911; Brown, 1943; Rice and Crowson, 1950; Muschel and Muto, 1956). In 1961, however, Borsos and Cooper found, by using cellular intermediates, that all these factors were present in at least the mouse strain they tested. Later on it became evident that some inbred strains lack haemolytic activity due to lack of C5 (Nilsson and Müller-Eberhard, 1967). Although this C5 deficiency could explain the inability to demonstrate haemolytic activity of mouse complement in these strains it did not explain why in other strains that were not C5 deficient, С activity was so difficult to measure.

In 1962 Rosenberg and Tachibana described an assay for the measurement of haemolytic activity of mouse complement, but its sensitivity was rather low. In 1975 Stolfi et al. mentioned a method that used spectrophotometric determination of the haemoglobin release but details of the method have not been published. Mouse sera are always spontaneously, and sometimes

* This study was partly supported by grants from the Netherlands Foundation for Medi cal Research FUNGO and from the Netherlands Kidney Foundation.

107

strongly, haemolysed. This interferes with the measurement of haemoglobin release due to complement activity, and makes the above mentioned approach less attractive. To overcome this problem we have developed an assay that makes use of slCr-labelled sheep red blood cells. This technique has a greater sensitivity than the determination of haemoglobin release (Weinrach et al., 1958). We describe here the optimum conditions for this 51Cr-release test. We have also studied how the haemolytic activity of mouse complement is influenced by the use of different amboceptors and by variation of the reaction conditions.


Mice Inbred lines of B10.D2/old Sn and A/HeJ mice were originally obtained

from the Jackson Laboratory (Bar Harbor, Maine, U.S.A.). C57BL6/Rij mice were obtained from the Radiobiological Institute TNO (Rijswijk, The Netherlands). DBA/2 and A/J mice were brought from Zentral Institut für Versuchstierzucht, Hanover, G.F.R. In our animal laboratory these mouse strains were maintained by continuous brother sister matings. (C57BL6/Rij X A/HeJ)Fi (=B6AF1) hybrids were raised in the laboratory.

Production of mouse amboceptor Mouse amboceptor was prepared in C57BL6, A/J and B6AF1 mice in two

different ways. The first method consisted of weekly intraperitoneal injections of 4 X 107 SRBC in 0.1 ml 0.9% saline. One week after the 5th immunization the mice were bled from the retro-orbital venous plexus. After clotting for 45 min at room temperature, the separated serum was heated for 45 min at 56°C to destroy complement activity. The sera were stored at —30oC. For the production of larger amounts of amboceptor in the same strains we gave weekly intraperitoneal injections of 4 X 107 SRBC in 0.1 ml 0.9% saline mixed with 0.1 ml complete Freund's adjuvant. After 5 injections most animals developed ascites, which was tapped every week. The product of several weeks was pooled. Before use, the ascites fluid was heated at 56°C for 45 min, and sterilized by passing through a sterile 0.20 nm filter. The haemolytic titer of the ascites fluid was the same as that of the serum of these mice. The sera were divided into 1 ml aliquots and stored at —30oC.

Complement All experiments were performed with pooled mouse sera. The mice were

bled from the retro-orbital venous plexus. After clotting for exactly 45 min at room temperature, the serum was collected by centrifugation at 40C, and stored in liquid nitrogen. Fresh serum from special pathogen free random bred Albino guinea pigs served as a source of guinea pig complement.

Reagents If not otherwise stated all titrations were performed in dextrose gelatin-

108

veronal buffer (DGVB2+) containing 0.00015 M Ca2+ and 0.0005 M Mg2+

(pH 7.35), ionic strength of 0.13 M NaCl, and prepared according to Rapp and Borsos (1970). The ionic strength was determined by measuring specific conductance at 20oC. The buffer was freshly prepared each day. Rabbit amboceptor was obtained commercially from Behring Werke.

SRBC were obtained aseptically from the same sheep each month. Blood was collected in an equal amount of Alsever's solution, and stored at 40C. Before labelling with s l Cr (Sodium Chromate, 50—400 mCi/mg/Cr, Radiochemical Center, Amersham, U.K.) the SRBC were washed four times with DVGB2+. One millilitre SRBC (containing 8 X 107 cells) was labelled with 6 μΟ 51Cr, by incubation during 60 min at 37°C. This resulted in 90% binding of the radioactivity (Ebaugh et al., 1953). With this dose of sodium Chromate no injury to erythrocytes occurs (Necheles et al., 1953). After incubation the labelled SRBC were washed three times with 10 volumes of DGVB2+, and the pellet resuspended in the original volume. To prepare sensitized cells (EA), an equal volume of an appropriate dilution of amboceptor in DGVB2+ was slowly pipetted into the labelled SRBC suspension with continuous mixing. The EA cell suspension was then incubated for 10 min at 370C with frequent mixing. The spontaneous release of 5 1Cr from this cell suspension during the first 48 h was 2—4% of the total activity bound.

Separation of mouse amboceptor in IgG-IgM antibodies This was performed with isokinetic sucrose gradients, prepared by addi

tion of 33% (w/w) sucrose in phosphate buffered saline (PBS) to 12 ml 5% (w/w) sucrose in PBS in a mixing chamber at constant volume (Noll, 1973). Test serum (0.5 ml) was layered on top of a 12 ml gradient and centrifuged in a SW 41 rotor (IEC) at 35,000 rev./min for 16 hours at 40C. The gradient was sampled from the bottom; 0.8 ml fractions were collected and each fraction was tested for agglutinating and haemolytic activity.

Serologic tests Basic outline of the haemolytic assay of mouse complement. The tests

were performed in sterile plastic tubes. The mouse serum to be tested was diluted 1 : 10 in DGVB2+ and 10—800 μΐ was put in tubes containing varying amounts of DGVB2+ to produce dilutions with a final volume of 800 μΐ. The dilutions were chosen such that at least 4 dilutions gave partial lysis. After adding 200 μΐ of 51Cr-labelled EA suspension (4 Χ 107 cells/ml) the tubes were incubated in a waterbath at different temperatures, and for different incubation periods. After incubation the tubes were placed on ice, and 2 ml of ice-cold DGVB2+ were added to quench the reaction. The tubes were then centrifuged for 5 min at 3,000 rev./min at 4°C. From each tube 2 ml supernatant was collected to measure the released s l C r activity. Each test included the following controls (in triplicate): (a) 200 μΐ of EA suspension in 800 μΐ DGVB2+ to determine spontaneous s l C r release; (b) 200 μΐ of EA suspension

109

lysed in 800 μΐ water to measure 100% lysis; (c) 200 μΐ of SRBC incubated with the highest amount of mouse serum used in that test, to exclude the presence of naturally existing mouse antibodies against SRBC. The released s l Cr activity in each tube was measured for 10 min in a gamma-ray well type scintillation counter (MS 588, Beard Atomic). The fraction of SRBC lysed in each dilution was calculated according to the formula:

cpm test serum—cpm control cpm complete lysis—cpm control

The data were plotted according to the logarithmic transformation of the Von Krogh equation to determine the CH50 titer per ml of undiluted mouse serum.

Antibody titrations of different amboceptor species. We studied the influence of the amboceptor species by comparing rabbit amboceptor in a limited complement system with different mouse amboceptors (A/J, C57BL6 and B6AF1). One hundred microliters of amboceptor were diluted in 2-fold steps in 100 μΐ DGVB2 +. To these serial dilutions 100 μΐ of 51Cr-labelled SRBC (8 X 107/ml) were added. After incubation for 10 min at 370C, 6μ1 of MC (C57BL6 male mouse serum) and 794 μΐ DGVB2+ were added, and incubated for 90 min at 30oC. Percentage lysis (fraction X100) was calculated for each amboceptor dilution.

Antibody titrations of different anti-SRBC antibody classes. After ultra-centrifugation of the different mouse amboceptors on a sucrose gradient 25 μΐ of each collected fraction was diluted in 2-fold steps in 25 μΐ DGVB2* in microtiter plates. To test the agglutinating activity 25 μΐ of SRBC (108/ ml) were added, and after mixing, the plates were incubated for 30 min at 370C. To test the haemolytic activity 25 μΐ SRBC (108/ml) were added to the amboceptor dilutions, incubated for 10 min at 370C, and thereafter 25 μΐ guinea pig serum or 25 μΐ B10.D2 male serum were added. After mixing, the plates with GPC were incubated for 30 min at 37CC, and those with MC for 90 min at 30CC. The plates were centrifuged for 1 min at 1000 rev./min at 40C, and read macroscopically. The titer was defined as the last dilution that gave complete haemagglutination or lysis. All tests were performed in duplicate.

RESULTS

Amboceptor species and class of anti-SRBC antibody As can be seen in Fig. 1, mouse and rabbit amboceptor give comparable

haemolysis. Both antisera exert strong anticomplementary effects, and maximal lysis requires relatively high concentrations of antibodies (dilution of 1/8 to 164). Mouse amboceptors had the advantage of giving less haemagglutination during sensitization of the SRBC than rabbit amboceptor. A difference in activity was observed between the three mouse amboceptors. With C57BL6 amboceptor maximal lysis occurred at a lower dilution than

110

·/. cell lysis

ЮОл

- ι 1 1 1 1 г 16 64 256 1024 256 1024

(amboceptor dilution)"

Fig. 1. Titration of three mouse and one rabbit amboceptor in a limited complement system (measured with 6 μΐ C57BL6 male MC). Each point represents the mean value of 4 different tests, the vertical bars give the S.E. M.

hemagglutinin — ( t i t e r ) - ' ,9s

4 0 9 6

1024

256

64

16

4

hemolytic ( t i t e r ) - 1

19S

I

I ' ' I ' I I ' ' ' I

5 10 15 fraction number

Fig. 2. Haemagglutinin and haemolytic titers of A/J, B6AF1 and C57BL6 amboceptor fractions, collected after separation on an isokinetic sucrose gradient. Haemolysis was measured with GPC and MC.

I l l

with B6AF1 and A/J, and the maximal lysis reached was also lower. Silver et al. (1972a; 1972b) reported that the antibody response to SRBC in C57BL6 mice was different from that in mice of the A strain'. In C57BL6 they found a predominantly 2-ME sensitive (IgM) response, while in A mice the antibodies formed were 2-ME resistant (IgG). Furthermore, Winn (1965) observed that mouse complement (MC) was better activated by IgG anti-SRBC antibodies, while by contrast, GPC was better activated by IgM anti-SRBC antibodies. To study whether the higher haemolytic activity of A/J amboceptor as compared to C57BL6 amboceptor in our system was due to the class of antibody present in these amboceptors, we separated IgG and IgM antibodies of A/J, C57BL6 and B6AF1 amboceptor with isokinetic sucrose gradients. The results are shown in Fig. 2. In the haemagglutination test the A/J amboceptor contains more 7 S anti-SRBC antibodies, while C57BL6 amboceptor contains more 19 S antibodies. The values for B6AF1 amboceptor are intermediate between those of the other two strans. A comparable pattern was found in the haemolytic test with MC, where in A/J and B6AF1 serum only 7 S antibodies were haemolytic, while in the C57BL6 serum lysis by 19 S was predominant. It is also clear from this figure that MC is better activated by IgG than by IgM antibodies, since the haemolytic titer of 7 S A/J antibodies with MC was 1 : 32, while with 19 S C57BL6 antibodies, of comparable haemagglutinating activity, the titer was only 1 : 2. For GPC a better activation by 19 S anti-SRBC was found, which is in agreement with findings of other investigators (Winn, 1965; Borsos and Rapp, 1965; Ishizaka et al., 1968). The results thus suggest that for the optimal determination of MC amboceptor containing 7 S antibodies (e.g., A/J anti-SRBC serum) should be preferentially used to sensitize SRBC.

C H 5

120-

100-

8 0 -

6 0 -

4 0 -

2 0 -

0 titer ( U/ml)

/ r

15 22 30 37 incubation temperature (°C)

Fig. 3. Haemolytic activity of B6AF1 male MC from a single batch measured at different incubation temperatures. Each point represents the mean of 4 different tests; the vertical bars are the S.E.M.

112

с·, ьсн

ι •30 J 120н

Т Ю -

•oc-9 0 -

ao-7 0 -

r

•e^

* ^

С Э 5

U г- ι

/ ^Y

0 0 /

X

0 0 9

У /

C-i

/ \ \

013

IO"iC s

\ \

0 1 5

' e r g i · ·

"

(•vi NaCi

Fig. 4. Haemolytic activity of B6AF1 male MC from a single batch at different ionic strengths. Each point represents the mean value of 4 different tests; the vertical bars are the S.E.M.

Incubation temperature It has been shown for several complement species that a lower incubation

temperature than 370C gives better haemolysis (Leon, 1956; Borsos et al., 1961; Colten et al., 1967; Cooper and Müller-Eberhard, 1968; Cunniff and Stollar, 1968; Sakamoto, 1975). To find out which incubation temperature was optimal for mouse complement we performed a haemolytic titration of the same batch of B6AF1 MC m quadruplicate at 15°C, 2 2 ° ^ 30oC and 370C (Fig. 3). The figure shows that for MC an incubation temperature of 30oC is optimal.

Ionic strength Lepow et al., (1956) stated that a lower ionic strength of the test medium

resulted in a higher haemolytic activity of human complement (HC). Later many reports confirmed these findings for GPC and HC (Becker and Wirtz, 1959; Rapp and Borsos, 1963; Inoue and Nelson, 1965; 1966; Gölten et al., 1968; Shin et al., 1971; Segerling and Muller, 1976). To find out whether this also holds true for MC we measured the CH50 titer of the same batch of B6AF1 MC (in quadruplicate) at different ionic strengths. The buffers were prepared by adding appropriate amounts of veronal buffered dextrose 5% (with 0.1% gelatine, 0.0005 M Mg2+, 0.00015 M Ca2+) to GVB2+. We found an optimum at 0.13 M NaCl (Fig. 4).

Incubation time The time necessary to obtain maximal lysis was measured in the presence

of one CH50 unit of MC (9 μΐ B6AF1 male serum) at 300C with 8 X 106

sensitized SRBC, total reaction volume 1 ml. After 90 min of incubation no further increase in haemolysis wis found (Fig. 5). In the presence of a lower amount of MC (7 μΐ B6AF1 serum) the time necessary to obtain maximal lysis increased to 120 min.

113

yss

6 0 1

2 0 -

1Θ0 210 ?40 mcubalion time (min)

Fig. 5. Percentage lysis of SRBC measured in a limited complement system at increasing incubation times. (I) Measured with 9 μΐ B6AF1 male mice MC, (II) with 7 jul B6AF1 MC. Each point represents the mean percentage lysis of four different tests. The S.E.M. for each point was < 1%.

Reaction volume By decreasing the total reaction volume haemolytic tests become more

sensitive (Mayer et al., 1946), and less serum is necessary to determine the CH50 titer. This is of special advantage in measuring haemolytic complement activity of mouse serum, because in most instances there is only a small volume of serum available. The same batch of mouse complement measured in total reaction volumes of 1 and 3 ml respectively gave a CH50 titer of 148.6 ± 2.0 and 69.9 ± 1.5 U/ml (means ± S.E.M. of 4 determinations).

Number of sensitized SRBC To determine the influence of the number of sensitized SRBC on the

CH50 titer ( U/ml)

300

250

200

150

100 H

50

10° 510 10' 510 Ю sensitized SRBC

Fig. 6. Relation between CH50 titer (U/ml) of male C57BL6 MC and number of labelled SRBC added. Each point gives the mean titer of four measurements, the vertical bars are the S.E.M.

1 1 4

TABLE 1

COMPLEMENT LEVELS IN DIFFERENT INBRED MOUSE STRAINS *

B10.Br C57BL6 B6AF1 A/HeJ DBA/2

Ss-low Ss-high Ss-high X C5 deficient C5-deficient CS-deficient

48.1 ± 4.1 167.5 ±10 .7

77.3 ± 4.1 0 0

Strain Complement status CH50 titer »* ± SEM U/ml Male Female

34.9 ±4 .9 63.8 ± 3.9 45.9 ± 1.7

0 0

* All mice were 6 weeks old. ** Each value is the mean of determinations in 6 mice.

CH50 titer we measured the CH50 titer of the same batch of C57BL6 MC to which different numbers of sensitized SRBC were added, under otherwise similar test conditions. We found a linear relationship between the CH50 titer and the number of SRBC if the results were plotted semi-logarithmi-cally (Fig. 6). It should be noted that although sensitivity is highest with 106

SRBC the reproducibility is less at this concentration.

Reproducibility of the test The intra-assay precision, determined by the coefficient of variation of

25 measurements of sera with complement levels ranging from 4—300 CH50 U/ml varied from 0.3—5.1%. Inter-assay variation with the same sera was 3.1—7.1%.

Complement levels in different inbred mouse strains Since the investigations of Demant et al. (1973) it has become evident

that a correlation exists between complement activity and H-2 haplotype, especially with the Ss-region allotype. To demonstrate the applicability of the method we measured complement levels in Ss-high and Ss-low mice. Furthermore we measured complement activity in some C5 deficient mice, and in a hybrid of a C5 positive and C5 negative parental strain (Table 1). In agreement with Nilsson and Müller-Eberhard (1967) the results also show a difference in complement level between male and female mice.

DISCUSSION

Mouse complement is very difficult to measure in standard haemolytic assays and test conditions have therefore to be modified. An important improvement in the system is the replacement of measurement of haemoglobin release by that of s lCr release. The first advantage of this is that the number of target cells may be decreased which leads to higher sensitivity. For example, the CH50 titer of C57BL6 mice measured with our method

1 1 5

http://B10.Br

was 10 times higher than in the method of Rosenberg and Tachibana (1962). Second, the error introduced by the fact that many mouse sera are haemo-lysed is eliminated. A third advantage is that in spite of the increased sensitivity of the assay, its coefficient of variation is very small.

The results of antibody titrations with amboceptors from different mouse strains stress the importance of the antibody class used. Mouse complement is much more readily activated by IgG than by IgM. We found that C57BL6 mice make predominantly IgM antibodies, while A/J mice produce mostly IgG after multiple immunization with SRBC.

Our findings confirm those of Silver et al. (1972a; 1972b), although there are quantitative differences. These might be accounted for by differences in technique since we used sucrose-gradient centrifugation instead of 2-mercapto-ethanol sensitivity to distinguish the two classes. The differential activity of 7 S and 19 S amboceptors with regard to the activation of mouse complement has practical implications because a sample of the same batch of MC tested with SRBC sensitized with A/J amboceptor gave a CH50 titer approximately 50% higher than with C57BL6 amboceptor.

On the basis of our findings we may summarize practical guidelines for the use of the assay as follows: (a) number of sensitized SRBC: 8 X 106 cells; (b) amboceptor class: IgG; (c) incubation temperature: 30oC; (d) ionic strength: 0.13 M NaCl; (e) incubation time: 90 min; (f) reaction volume: 1 ml. With this test substantial differences in complement activity between different mouse strains were found. The method offers an opportunity to measure sensitively complement levels in an animal experimental system widely used in immunological research.

REFERENCES

Becker, E.L. and G.H. Wirtz. 1959, Biochim. Biophys. Acta 35, 291. Borsos, T. and M. Cooper, 1961, Proc. Soc. Exp. Biol. Med. 107, 227. Borsos, T. and H.J. Rapp, 1965, Science 150, 505. Borsos, T., H.J. Rapp and M.M. Mayer, 1961, J. Immunol. 87, 326. Brown, G.C., 1943, J. Immunol. 46, 319. Colten, H.R.,T. Borsos and H.J. Rapp, 1967, Protides Biol. Fluids 15, 471. Gölten, H.R., T. Borsos and H.J. Rapp, 1968, J. Immunol. 100, 799. Cooper, N.R. and H.J. Müller-Eberhard, 1968, J. Immunol. 101, 813. Cuniff, R.V. and B.D. Stollar, 1968, J. Immunol. 100, 7. Demant, P., J. Capkova, E. Hinzova and B. Voracova, 1973, Proc. Natl. Acad. Sci. U.S.A.

7 0 , 8 6 3 . Ebaugh, F.G., C.P. Emerson and J.F. Ross, 1953, J. Clin. Invest. 32, 1260. Inoue, K. and R.A. Nelson, 1965, J. Immunol. 95, 355. Inoue, K. and R.A. Nelson, 1966, J. Immunol. 96, 386. Ishizaka, T., T. Tada and K. Ishizaka, 1968, J. Immunol. 100, 1145. Leon, M.A., 1956, Proc. Soc. Exp. Biol. Med. 91, 150. Lepow, I.H., O.D. Ratnoff, F.S. Rosen and L. Pillemer, 1956, Proc. Soc. Exp. Biol. Med.

92, 32. Mayer, M.M., A.G. Osier, O.G. Bier and M. Heidelberger, 1946, J. Exp. Med. 84, 535. McGhee, R.B., 1952, Proc. Soc. Exp. Biol. Med. 80, 419.

1 1 6

Muschel, L.H. and T. Muto, 1956, Science 123, 62. Necheles, T.F., I.M. Weinstein and G.V. Le Roy, 1953, J. Lab. Clin. Med. 42, 358. Nilsson, U.R. and H.J. Müller-Eberhard, 1967, J. Exp. Med. 125, 1. Noll, H. 1967, Nature 251, 360. Rapp, H.J. and T. Borsos, 1963, J. Immunol. 91, 826. Rapp, H.J. and T. Borsos, 1970, The Molecular Basis of Complement Action (Appleton,

Century and Crofts, New York). Rice, C E . and C.N. Crowson, 1950, J. Immunol. 65, 201. Ritz, H., 1911, Ztschr. f. Immunitaetsf. 9, 321. Rosenberg, L.T. and D.K. Tachibana, 1962, J. Immunol. 89, 861. Sakamoto, M., 1975, Japan. J. Exp. Med. 45, 183. Segerling, M. and F. Müller, 1976, Immunochemistry 13, 117. Shin, H.S., R.J. Pickering and M.M. Mayer, 1971, J. Immunol. 106, 480. Silver, D.M., I.F.C. McKenzie and H.J. Winn, 1972a, J. Exp. Med. 136, 1063. Silver, D.M. and H.J. Winn, 1972b, Cell. Immunol. 7 237. Stolfi, R.L., R.A. Fugmann, L.M. Stolfi and D.S. Martin, 1975, J. Immunol. 114, 1824. Weinrach, R.S., M. Lai and D.W. Talmage, 1958, J. Inf. Dis. 102, 60. Winn, H.J., 1965, in: Ciba Foundation Symposium on Complement, eds. G.E.W. Wolsten·

holm and J. Knight (Churchill, London) p. 133.

1 1 7

Chapter VII

COMPLEMENT DEPENDENT AND -INDEPENDENT MECHANISMS IN ACUTE

ANTIBODY-MEDIATED REJECTION OF SKIN XENOGRAFTS IN THE MOUSE

Jo H.M. Berden*, Jacqueline F.H.M. Hagemann*, Wim P.M. Tamboer*,

M.José J.Th. Bogman , Robert A.P. Koene*

Departments of Medicine, Division of Nephrology and Pathology,

Sint Radboud Ziekenhuis, University of Nijmegen, Nijmegen

119

INTRODUCTION

Two experimental systems have been described in which skin

grafts carried by mice can be acutely destroyed by the administra

tion of antisera directed against the graft antigens (1,2). One

of these consisted of allogeneic skin grafted onto mice, that

were treated with anti mouse lymphocyte serum to suppress cellular

immunity (2). On day 7 after transplantation, when the grafts

showed no signs of rejection the recipients received an injection

of alloantiserum along with fresh rabbit serum as a source of

complement. Acute graft destruction occurred within 24-48 h.

Administration of exogenous complement was essential, since in

jection of antiserum alone did not destroy the grafts. This system

enabled us to study the mechanism of humorally mediated graft

destruction. Histologically, the grafts showed the characteristics

of an Arthus reaction with dense infiltration of polymorphonuclear

leukocytes (PMN) as early as 30 minutes after the antiserum in

jection, followed by extravasation of erythrocytes and necrosis.

It was subsequently shown that activation of the complete comple

ment pathway was necessary for destruction to occur (3).

Our experimental system was based on an earlier study, in

which rat skin grafted onto immunosuppressed mice was destroyed

by injection of anti rat serum (1). Histological changes in these

grafts were also of the Arthus type. Apparently the mouse endoge

nous complement was activated in this xenogeneic system of wide

antigenic disparity, since administration of exogenous complement

was not necessary to cause destruction. Further work of Winn in

this system showed that acute destruction could be induced in C5

deficient mice, while on the other hand, C3 depletion by cobra

venom factor, or PMN depletion with anti granulocyte serum, in

hibited the graft destruction (4). This suggested that activation

of C3 with consequent PMN attraction sufficed to induce tissue

damage in this xenogeneic graft model. Complement mediated cyto-

lysis of the grafts by the late acting complement components did

probably contribute to the damage because the rejection in mice

with a normal complement status had a more violent course, but

120

it played no essential role. Although these experiments show that

complement is a very important factor, it was not completely ex

cluded that graft destruction by antibody in this system also

occurred via complement independent mechanisms. The finding that,

in C3 depleted mice, grafts that were not rejected all showed

signs of inflammation might serve as an argument for such a com

plement independent destruction.

We have used a comparable xenogeneic system, to study antibody

mediated rejection and now present evidence that another, comple

ment independent, pathway of graft damage exists. This complement

dependency of graft rejection in this system was studied by using

complement deficient or complement depleted recipients, or alter

natively, by injecting non-complement fixing antibodies. Histo

logic studies were done in addition to macroscopic inspection of

the grafts, to obtain information on the mediators that were res

ponsible for the destruction.


Animals. Inbred C57BL6/RÌJ mice were originally obtained from the

Radiobiological Institute TNO, Rijswijk (The Netherlands), and

DBA/2 (C5 deficient) mice from the N.I.H. Bethesda (Maryland,

U.S.A.). Inbred PVG/c rats came originally from the Institute of

Psychiatry, Bethlem Royal Hospital, Beckingham (Kent, U.K.).

In our laboratory these strains were kept by continuous brother-

sister matings.

Transplantation procedure. Male PVG/c tail skin was grafted onto

the flank of male C5 7BL6 or DBA/2 recipients by a modified fitted

graft technique (3). The fate of the grafts was followed by daily

macroscopic examination. Grafts were considered to be rejected

when no viable epithelium remained. Median survival times (MST)

and standard deviations (SD) were calculated according to the

method of Litchfield (5). To postpone cellular rejection of the

grafts all recipients were treated with rabbit anti mouse lympho

cyte serum (RAMLS) or the Ig fraction thereof RAMLG) on days -2,

121

-1, О, 2 and 4 after grafting in a dose of 0.25 ml i.p. This

immunosuppressive regimen prolonged the MST of PVG/c grafts to

23.0 + 1.2 days in C57BL6 recipients and to 24.5 + 1.3 days in

DBA/2 mice.

Acute reiection studies. On day 7 after grafting, when all grafts

were well healed m and showed no signs of rejection, the recip

ients received mouse anti rat serum (MARS) or its Ig fraction

(MARG) via an intravenous (i.V.) injection in the tail vein. In

mice that are not complement deficient this results in a rapid

and violent graft destruction. In this study we define this acute

antibody mediated rejection (AAR) as the complete necrosis of the

graft within 24-72 h after the i.v. injection of anti rat serum.

Complement depletion. Attempts to induce complement depletion in

vivo were first made with four agents that are known to have anti

complementary effects in vitro. These agents were: suramin

(Germanin ) provided by Bayer, Mijdrecht, the Netherlands;

flufenamic acid (Arlef ) provided by Parke-Davis, Mijdrecht, the

Netherlands; pentamidine (Lomidine ) provided by Specia, Amstelveen,

the Netherlands, and carrageenin commercially obtained from Sigma

St. Louis, U.S.A. In vivo dosages were as high as possible, but

such that there were no toxic effects to the recipients. In a

second series of experiments cobra venom factor (CoVF), 100 U/ml

(Cordis Ltd, Miami, U.S.A.) was used as an anticomplementary agent.

Antisera. RAMLS was prepared according to the method of Gray et

al (6). MARS was prepared in C57BL6 mice by weekly i.p. injections

of 5 χ 10 PVG/c lymphoid cells suspended m complete Freund's

adjuvant. After five injections the animals developed ascites,

that was tapped weekly. The lymphocytotoxic titer of the ascites

fluid was the same as the titer in the serum of these mice. The

antisera, if not used for preparation of Ig fractions, were pooled

and heated at 56 С for 45 min to destroy complement activity, and

sterilized by passage through a sterile 20 mp filter (Scheicher

and Schüll, Dassel, West Germany).

122

P r e p a r a t i o n of Ig f r a c t i o n s from a n t i s e r a . From RAMLS and MARS

Ig f r a c t i o n s (RAMLG and MARG) were p r e p a r e d i n o r d e r t o o b t a i n

p r e p a r a t i o n s f ree of complement f a c t o r s . The Ig f r a c t i o n s were

p r e c i p i t a t e d in 50% s a t u r a t e d ammoniurnsulphate, and s u b s e q u e n t l y

d i a l y s e d a g a i n s t phospha te b u f f e r e d s a l i n e (PBS). The d i a l y s a t e

was p u t on a Sepharose 4B-CL coupled P r o t e i n A column (Pharmacia ,

Uppsa la , Sweden). A f t e r s u f f i c i e n t washing wi th PBS, immuno

g l o b u l i n s were e l u t e d from t h i s column w i t h 0.05 M s o d i u m c i t r a t e

b u f f e r pH 2 . 8 , c o n t a i n i n g 1 M NaCl. The e l u a t e was immedia te ly

n e u t r a l i z e d and d i a l y s e d a g a i n s t PBS. The MARG-dialysate was

c o n c e n t r a t e d t o a volume t h a t had the same c y t o t o x i c i t y i n v i t r o

towards PVG/c lymphoid c e l l s as the MARS p r e p a r a t i o n from which

i t o r i g i n a t e d . RAMLG was c o n c e n t r a t e d and used in dosages t h a t

gave s i m i l a r p r o l o n g a t i o n of g r a f t s u r v i v a l as the o r i g i n a l

RAMLS p o o l .

P r e p a r a t i o n of IgGl and IgG2 s u b c l a s s an t ibody from MARG. Non

complement f i x i n g (IgGl) and complement f i x i n g (IgG2) a n t i b o d i e s

were p r e p a r e d from MARG by a f f i n i t y chromatography acco rd ing t o

an e a r l i e r d e s c r i b e d method ( 7 ) . Contamina t ing immunoglobulins

were removed from the IgGl and IgG2 p r e p a r a t i o n s by r e p e a t e d runs

on Sepharose columns coa ted wi th a p p r o p r i a t e immunoglobulin a n t i

b o d i e s . The p u r i t y of t he f i n a l p r e p a r a t i o n s was t e s t e d by r a d i a l

immunodiffusion. Contamina t ing IgA o r IgM could no t be demon

s t r a t e d (lower l i m i t of d e t e c t i o n < 0.05%). The f i n a l IgGl p r e

p a r a t i o n t h a t c o n t a i n e d 13,000 yg I g G l / m l , was con tamina ted wi th

0.0 5% IgG2, and the IgG2 p r e p a r a t i o n (8,700 pg IgG2/ml) w i th 0.1%

IgGl .

Se ro logy . The measurement of h e m o l y t i c complement a c t i v i t y of

t he mouse was performed wi th a C r - r e l e a s e assay t h a t has been

d e s c r i b e d e l sewhere i n d e t a i l ( 8 ) . Lymphocyto tox ic i ty of the a n t i -

s e r a was a s s e s s e d in a t r y p a n b lue e x c l u s i o n a s s a y , w i th PVG/c

s p l e e n c e l l s and r a b b i t complement ( 9 ) .

H i s t o l o g i c examina t ion of t he g r a f t s . For t h i s purpose s k i n g r a f t s

123

and surrounding recipient skin were removed at different times

after the i.V. injection of MARG and fixed in 4% buffered forma

lin. From paraffin embedded tissue blocks 4 ym sections were cut,

and stained with hematoxylin and eosin, with elastic van Gieson

stain to visualize the vessel walls, and with Mallory's phospho-

tungistic acid hematoxylin stain to detect fibrin. The histologi

cal sections were studied by the pathologist (M.J. B.) who had

no knowledge of the experimental protocol used. At each tirae-

interval three grafts were studied, and all experiments were per

formed twice.

RESULTS

Complement consumption du r ing g r a f t r e j e c t i o n . Mice were b l e d

from the r e t r o - o r b i t a l ve in a t v a r i o u s i n t e r v a l s a f t e r the i n -

time after ¡.v. injection

0 1 2 8 24 48 72 (hr) I 1 1 1 — I — ¡ — г - " — ι — " — ι — " — ι 1 1 1

'U c h a n g e CH50 titer T

.20

О

,20-

-40

.60-

-80-

-100-

Fig. 1. Changes of CHSO titer in CS7BL6 -n.ale mice during the resection of a

P'/G/a skin graft induced by the i.V. administration of 0.25 ml mouse

anti rat serum (ш). Control experiments: CHSO titers in C57BL6 mice

carrying a FVG/a skin graft (o), and in CS7B16 nice without a skin

graft that received 0.2S ml mouse anti rat serum (·). Each point is

the mean + S.E.M. of determinations in 8 mice. All determinations were

done in triplicate.

124

jection of MARS, and hemolytic complement activity was measured

in these sera. The results are given in Fig. 1. Two control ex

periments were performed. One group of mice that received MARS

but did not carry a graft, showed no changes in complement levels,

The second group was grafted and treated with RAMLS but did not

receive MARS. This treatment induced a 40% decrease of the CH50

titer on day 7, with a gradual rise to normal values on day 9.

The injection of MARS in mice carrying a PVG/c graft resulted in

an immediate decrease of the CH50 titer. After 24 h when the

graft was completely rejected, the CH50 titer returned towards

normal values. These results show that there was consumption of

complement during AAR.

Influence of treatment with anticomplementary agents on hemolytic

complement activity and AAR. C57BL6 male mice received three i.p.

injections of 0.5 mg pentamidine, 2 mg carrageenin, 2.5 mg flu-

-4Θ -Z4 О

t ime after i.p. injection anticomplementary agent

72 9 6 (hr) - i

% change CH50 t i t e r

* 2 0 1

- 2 0

- 4 0

-60

- 8 0

-ίσο

Fig, 2. Changes of the CHSO titer in CS7BL6 mice after three i, p. administra

tions (-48 h, -24 h, and 0 h) of various antiaomplementary agents.

• 0, S mg pentamidine

m 2 mg aarrageenin

a 2. 6 mg flufenamia aaid

о 5 mg suramin

Each point represents the mean + S.E.M. of determinations in four

different miae. Each determination was performed in dupliaate.

125

ТЛВЬЕ I. Influenae of treatment with anti eomplenentary agents

on AAR of PVG/a skin grafted onto CS7BL6 recipients

Group number Anticomplementary agent Occurrence of AAR

1 none 19/19

2 pentamidine (0.5 mg) Β/θ

3 carrageenin (2 mg) 9/9

4 flufenamic acid (2.5 mg) 8/8

5 suramin (5 ng) 5/7

a) injected i.p. 48, 24, and 2 h before the i.v. injection of HARG, on day 7

(daily dose listed in ng)

b) AAR was induced by i.v. injection of 0.1 ml MARG on day 7 after grafting

c) number of mice with complete graft necrosis within 24-72 h after IIARG

injection/number of mice tested

fenamic acid, or 5 mg suramin at 24 h intervals. The CH50 titers

measured 24, 48, 72 and 96 h after the third gift are presented

in Fig. 2. Only with suramin a substantial decrease in CH50 titer

was found. C57BL6 mice carrying a PVG/c graft received a similar

treatment with these anticomplementary agents at 48, 24, and 2 h

before the injection of 0.1 ml MARG on day 7.

Table I shows that AAR was not prevented by pretreatment with pen

tamidine, flufenamic acid, or carrageenin, which were the drugs

that scarcely influenced the CH5Û titer. Suramin that induced a

substantial depletion of hemolytic activity of complement pre

vented AAR in 2 out of 7 mice.

Influence of CoVF treatment on hemolytic complement activity.

Prior to the experiments in which it was tried to inhibit AAR with

CoVF we studied the effects of increasing CoVF dosages on comple

ment activity in otherwise untreated C57BL6 mice. Fig. 3 shows

that, 24 h after a dose of 20 U, hemolytic activity had almost

time after i.p. injection of CoVF 72 96 120 (hr)

-100

Fig. 3. Influence of CoVF treatment on the CH50 titer of С57БЬв mice.

• single dose of 2.S U i.p.

a single dose of 5 U i.p.

m single dose of 20 U i.p.

о daily dose of 20 U i.p.

Each point represents the mean * S.E.M. of determinations in four

different mice. Each determination uas performed in triplioate.

127

00

TABLE I I . Influenae of CoVF treatnent on AAR of P/G/o

skin grafted onto СБ7вІв геаг^гепЬв

Group number Dose MARC (nl) CoVF treatment a)

Occurrence of AAR

0.1 ml

0.1 ml

0.25 ml

0.5 ml

0/15 (MST: 23.0)

19/19

0/8

6/11

10/10

d)

a) 20 U CoVF daily i.p., started on day 5 after grafting and continued for 7 days or

until rejection of the graft was complete

b) AAR was induced by i.v. injection of listed volumes of MARG on day 7 after grafting

c) number of mice with complete graft necrosis within 24-72 h after MARG injection/

number of mice tested

d) MST = median survival time

dropped to zero. Daily injections of this dose ensured a complete

depletion of hemolytic complement activity during at least 5 days.

Influence of CoVF treatment on AAR in normal and C5 deficient mice.

C5 7BL6 mice carrying PVG/c skin grafts were treated with daily

injections of 20 U CoVF from day 5 after transplantation. Injection

of 0.1 ml MARG at day 7 did not induce AAR, while such a dose

induced rapid destruction in recipients with a normal complement

status (Table II). However, the grafts in CoVF treated recipients

were not completely protected against the destructive effects of

the anti rat serum since at higher doses AAR occurred. A dose of

0.25 ml destroyed the grafts in about half of the recipients,

while 0.5 ml MARG induced AAR in 100% of the cases. Since complete

and sustained loss of hemolytic complement activity had been in

duced by CoVF treatment, these results suggested that graft des

truction could be mediated through complement independent mecha

nisms. More arguments for this assumption were sought by performing

similar experiments in C5 deficient mice. The results are shown

in Table III. Similarly to the findings in CoVF treated mice, 0.1

ml MARG did not induce AAR in these animals. With increasing

doses of MARG the grafts again became sensitive to the antibody.

That AAR in these animals only occurred at higher antibody doses

was apparently a consequence of the C5 deficiency since 0.1 ml

MARG injected along with 0.5 ml of non complement deficient nor

mal mouse serum induced AAR in all recipients (group 5). The

results in these C5 deficient mice confirm the findings of Winn

et al (4). He presumed that in these animals activation of C3

was sufficient to induce destruction and that, although activa

tion of lytic components of complement contributed to the damage,

these were not essential. If this were the case, one would expect

that C3 depletion in C5 deficient mice would further decrease the

sensitivity of the grafts to AAR, or would even protect them

completely against destruction. Therefore we treated DBA/2 mice

with a protocol identical to that used in the mice that had no

genetic complement deficiency. The results given in Table III

(groups 6 and 7) show that this CoVF treatment caused no sub

stantial change in the sensitivity to MARG.

129

TABLE III. Influence of Ci deficiency and CoVF treatment on AAR

of PVG/c skin grafted onto DBA/I reaipients

Group number Dose MARG (nl) CoVF treatment Occurrence of AAR

1 - - 0/15C) (HST: 24.5)

2 0.1 ni - 0/16

3 0.25 ml - 10/17

4 0.5 ml - 11/11

5 0.1 ml + 0.5 nl NMSd) - 10/10

6 0.25 ml + 2/7

7 0.5 nl + 9/9

a) 20 U CoVF daily i.p., started on day 5 after grafting, and continued for 7 days or

until rejection was complete

b) AAR was induced by i.v. infection of the listed volumes of MARG on day 7 after grafting

c) number of mice with complete graft necrosis within 24-72 h after MARG injection/

number of mice tested

d) NHS = normal mouse serum not deficient in C5# injected along with MARG

Destructive activity of non-complement fixing IgGl subclass anti

bodies . We surmised that non-complament fixing antibodies could

also cause AAR if complement independent mechanisms indeed played

a role in our system. For this purpose we tested the activity of

purified IgGl antibodies prepared from MARG. As stated in the

methods the preparation was slightly contaminated with IgG2. If

tested in a cytotoxic assay against PVG/c lymphocytes residual

complement dependent cytotoxicity was found resulting in 50% cell

lysis up to a dilution of 1:8. This was attributed to the conta

minating IgG2, present in a concentration of 6.5 ug/ml. The ori

ginal MARG pool contained 6 ,900 pg IgG2/ml and had a cytotoxic

titer of 1:20,480. The destructive effects of increasing doses

of IgGl were studied in C57BL6 recipients with a normal complement

status (Table IV). The limiting dose for the induction of AAR was

approximately 800 pg IgGl. This dose was contaminated with 0.4

yg of IgG2. From the results of the dose response study for IgG2,

also presented in Table IV, it can be seen that the amount of

IgG2 contamination of the IgGl preparation cannot be held res

ponsible for the destruction, since the limiting dose for IgG2 is

between 65 and 131 yg and thus at least 160 times higher.

Histology of AAR in mice with a normal complement status and in

complement depleted animals. Induction of AAR in complement de

ficient animals did not only require higher amounts of antibodies

but the grafts also showed a different rejection pattern. In mice

with a normal complement status there was rapid necrosis of the

graft that was complete in 8 to 24 h. In C3 or C5 deficient

animals the process occurred more slowly. It started with hyper-

aemia of the graft followed by cyanosis and necrosis that was

complete after 48-72 h. The histological findings reflected these

macroscopic differences. Grafts, that were carried by mice with

a normal complement status, showed intravascular coagulation with

damage of vessel walls and extravasation of erythrocytes at 30

min after the injection of 0.25 ml of MARG. There was no intra

vascular increase of PMN. Necrosis was found after 2 h and was

complete after 8 h. In grafts of mice in which C3-depletion was

induced by CoVF, the first event was a dense intravascular accu-

131

ы

TABLE IV. AAR of FVG/o skin grafted onto CS7BL6 recipients induced by the administration of

IgGl and IgG?, subclass antibodies specifically directed against the donor antigens

Preparation IgGl (pg) IgG2 (pg) Occurrence of AAR a)

IgGl

1625

812

406

0.Э

0.4

0.2

5/5"

2/6

0/5

IgG2

0.3

0.2

0.1

261

131

65

5/5

5/5

0/5

a) AAR was induced by i.v. injection of the listed amounts of IgGl or IgG2 on day 7 after

grafting

b) nunber of nice with complete graft necrosis within 24-72 h after the IgGl or IgG2

injection/nunber of mice tested

mulation of PMN at one half and two hours after the injection of

0.25 ml of MARG. There were no signs of intravascular coagulation

at that time. Slight vessel damage with incipient extravasation

of erythrocytes and intravascular coagulation started at 8 h.

Partial or complete necrosis was present at 36 and 48 h after the

injection. The accumulation of PMN was not caused by the CoVF

treatment since sections of normal skin of the same recipient,

and grafts in mice that were treated with CoVF but not with MARG

showed no increase of PMN. These results show that dependent on

the complement status of the recipients two distinct types of

antibody mediated graft destruction occur.

DISCUSSION

It is clear from this and previous studies (2-4,9) of anti

body mediated rejection, that in recipients with a normal comple

ment status, activation of complement is an important effector

system that leads to graft damage. The decrease of the CH50 titers

during AAR of PVG/c skin grafts indicates that complement con

sumption took place. The CH50 titer remained depressed until re

jection was complete. Similar results have been reported during

rejection of xenogeneic kidneys in dogs (10).

An in vitro anticomplementary effect has been described for

pentamidine (11-15), flufenamic acid (16) and carrageenin (17,18),

but we were not able to induce in vivo complement depletion with

these drugs. With respect to carrageenin it has been reported

earlier that its anticomplementary activity in vivo is not very

impressive (19) . Only treatment with suramin induced a substantial

decrease of the CH50 titer in vivo that was in line with its

reported anticomplementary activity in vitro (20,21). The residual

complement activity was obviously sufficient to induce graft des

truction, since only 2 out of 7 grafts were protected against the

destructive effect of a relatively low dose (0.1 ml) of MARG.

Only with long lasting depletion of complement activity by CoVF

protection against 0.1 ml of MARG was obtained. However, when the

dose was raised to 0.5 ml, destruction of all grafts again occurred,

although the reaction was not as abrupt as seen in animals with

133

a normal compleirent status. It is possible that traces of comple

ment activity not measurable in the hemolytic complement assay

in vitro, mediated the destruction at higher doses of antibody.

Two findings argue against this explanation. First, in C5 defi

cient animals the decrease in sensitivity was similar to that in

C3 depleted animals and there was no further drop when these ani

mals were also treated with CoVF. One would anticipate a decrease

to occur if C3 activation, without activation of terminal comple

ment components, played an important role in the destruction.

A second and stronger argument for the involvement of complement

independent mechanisms, was our finding that non-complement

fixing IgGl antibodies also caused destruction. This destructive

effect of non complement fixing IgGl is not present in all systems.

We have previously shown that mouse skin allografts that can be

destroyed by administration of specific alloantiserum together

with rabbit complement, are completely unaffected if IgGl anti

bodies are used (7). These studies were performed in a congenie

model with limited antigenic differences and this suggests that

the occurrence of destruction by IgGl depends on the antigenic

disparity of the donor-recipient combination. Λ further argument

for the fact that complement-independent mechanisms can only

become operative in graft models with a wide antigenic disparity,

is our previous finding that AAR of mouse allografts only occurred

in the presence of a completely intact complement system (3).

Histologic studies can give insight into the spectrum of

reactions that may lead to antibody mediated vascular damage. The

results in animals with a normal complement status show that acute

intravascular coagulation can occur, without involvement of PMN.

It is likely that under these circumstances there is direct cyto-

lysis of vascular endothelium by antibody and complement with

consequent rapid platelet aggregation and coagulation. This has

happened before PMN, attracted by activated chemotactic factors,

can reach the graft. It, thus, seems that PMN are not absolutely

necessary to cause graft destruction initiated by antibody. The

destruction can, however, also take place via a classic Arthus

type reaction with activation of chemotactic complement compo

nents and secondary attraction of PMN. We have shown previously

134

that this type of reaction occurs in an allogeneic system, where

destruction is seen only in the presence of exogenous complement

(2). After accumulation of PMN in the graft vessels, there is

vascular damage, extravasation of erythrocytes and necrosis. Sim

ilar histologic changes have been found in another model of anti

body mediated destruction of rat skin grafts (1,4). On the basis

of our findings we propose that a third pathway of vascular damage

should be added to this spectrum. Although this pathway is comple

ment independent it is mediated by PMN as shown by the histologic

studies in animals depleted of complement by treatment with CoVF.

It is not clear which factors are responsible for the attraction

of PMN under these circumstances. It might be the result of immune

adherence of PMN to the Fc part of the immunoglobulins bound to

the vascular endothelium (22). Since there is no amplification of

this process by the generation of chemotactic factors of the com

plement system, the reaction develops more slowly and requires a

relatively high amount of antibodies. This does not exclude that

the process is amplified by other mechanisms. It has been shown

that mouse IgG antibodies can induce a complement independent

release of histamin and serotonin from leukocytes (23), and this

could obviously contribute to the rejection process. Furthermore,

immune adherence of PMN can cause release of lysosomal enzymes

(22), that can enhance further accumulation of PMN, causing damage

of the vascular endothelium (24). Elucidation of the mechanism in

this type of rejection clearly requires further study. Our findings

are in apparent contrast to earlier observations in reversed pas

sive Arthus reactions and in experimental nephrotoxic nephritis,

where depletion of the complement system could prevent PMN accu

mulation (25-29). This stresses the fact that in graft rejection

other mechanisms can become operative, whose activation is proba

bly influenced by the extent of antigenic disparity that exists

between donor and recipient.

The existence of this mechanism might have practical implica

tions for clinical transplantation. Attempts to influence humoral

graft rejection by complement depletion might fail because this

approach does not provide absolute protection against antibody

mediated graft destruction.

135

SUMMARY

Complement dependency of acute antibody mediated rejection

(AAR) was studied in a xenogeneic skin graft model in the mouse.

PVG/c rat skin grafted to immunosuppressed mice is quickly des

troyed by intravenous administered mouse anti rat serum on day 7

after grafting. We demonstrated depression of the CH50 titer in

the recipients, indicative for complement consumption during the

rejection process. Of four anticomplementary agents: pentamidine,

carrageemn, flufenamic acid and suramin, only treatment with

suramin led to a substantial complement depletion in vivo.

Administration of suramin prior to anti rat serum protected 30%

of the grafts against AAR, while the three other drugs were in

effective in this regard. Complete and long lasting complement

depletion induced by treatment with cobra venom factor (CoVF)

decreased the sensitivity of the grafts for AAR. However, complete

protection was not achieved, since high doses of anti donor serum

again induced destruction. Similar results were obtained in C5

deficient recipients, and m CoVF treated C5 deficient recipients.

These results indicated that complement independent rejection

mechanisms were operative. This was further substantiated by the

finding that pure non-complement fixing IgGl subclass antibodies

were able to elicite AAR. Histologic studies showed that AAR of

grafts in mice with a normal complement status was caused by

intravascular coagulation without primary involvement of poly

morphonuclear leukocytes. In complement depleted animals an

Arthus-like reaction was seen, with dense intravascular accumula

tion of PMN, that, apparently, were attracted to the graft through

complement independent mechanisms.

ACKNOWLEDGMENTS

The skillfull technical assistance of Miss Cathy N. Maass and the

staff of the animal laboratory (head: Dr. W.J.I, van der Gulden)

is gratefully acknowledged.

136

REFERENCES

1. Baldamus, CA., I.F.С. McKenzie, H.J. Winn, and P.S. Russell.

1973. Acute destruction by humoral antibody of rat skin grafted

to mice. J. Immunol. 120:1532.

2. Koene, R.A.P., P.G.G. Gerlag, J.H.F.M. Hagemann, U.J.G. van

Haelst, and P.G.A.В. Wijdeveld. 1973. Hyperacute rejection of

skin allografts in the mouse by the administration of allo-

antibody and complement. J. Immunol. 111:520.

3. Berden, J.H.M., P.J.A. Capel, and R.A.P. Koene. 1978. The role

of complement factors in acute antibody-mediated rejection of

mouse skin allografts. Eur. J. Immunol. 5:158.

4. Winn, H.J., C.A. Baldamus, S.V. Jooste, and P.S. Russell. 1973.

Acute destruction by humoral antibody of rat skin grafted to

mice. The role of complement and polymorphonuclear leukocytes.

J. Exp. Med. 237:893.

5. Litchfield, J.T. 1949. A method for rapid graphic solution of

time-per cent effect curves. J. Pharmacol. Exp. Ther. 97:399.

6. Gray, J.G., A.P. Monaco, and M.L. Wood. 1966. Studies on

heterologous antilymphocyte serum in mice. I. In vitro and

in vivo properties. J. Immunol. 96:217.

7. Jansen, J.L.J., R.A.P. Koene, G.J. van Kamp, W.P.M. Tamboer,

and P.G.A.B. Wijdeveld. 1975. Isolation of pure IgG subclasses

from mouse alloantiserum and their activity in enhancement and

hyperacute rejection of skin allografts. J. Immunol. 22£:387.

8. Berden, J.H.M., J.F.Η.M. Hagemann, and R.A.P. Koene. 1978.

A sensitive hemolytic assay of mouse complement. J. Immunol.

Meth. 2 3:149.

9. Berden, J.H.M., P.G.G. Gerlag, J.F.H.M. Hagemann, and R.A.P.

Koene. 19 77. Role of antiserum and complement in the acute

antibody-mediated rejection of mouse skin allografts in strain

combinations with increasing histoincompatibility. Transplan

tation. 24:175.

10. Gewürz, H., D. Scott Clarck, J. Finstad, W.D. Kelly, R.L.

Varco, R.A. Good, and A.E. Gabrielsen. 1966. Role of the com

plement system in graft rejections in experimental animals and

man. Ann. N.Y. Acad. Sci. 223:673.

137

11. Coats, Ε.Α. 1973. Comparative inhibition of thrombin, plasmin,

trypsin, and complement by benzamidines using substituent

constants and regression analysis. J. Med. Chem. 25:1103.

12. Asghar, S.S., K.W. Pondiran, and R.H. Cormane. 1973. Inhibition

of Clr, Cls and generation of Cls by amidino compounds.

Biochem. Biophys. Acta. 317:539.

13. Hausch, С , and M. Yoshimoto. 1974. Structure-activity rela

tionships in immunochemistry. 2. Inhibition of complement by

benzamidines. J. Med. Chem. 17:1160.

14. Glovsky, M.M., M. Cory, and A. Alenty. 1974. Inhibition of

guinea pig complement by dérivâtes of benzamidine. Immunology.

20:819.

15. Hauptman, J., and F. Markwardt. 1977. Inhibition of the hemo

lytic complement activity by derivatives of benzamidine.

Biochem. Pharmacol. 2 5:325.

16. Harnty, Th.W., and M.B. Goldlust. 1974. Anti-complement

effects of two anti-inflammatory agents: Niflumie and flu-

fenamic acids. Biochem. Pharmacol. 2 3: 3107.

17. Davies, G.E. 1963. Inhibition of guinea pig complement in

vitro and in vivo by carrageenm. Immunology. 6:561.

18. Davies, G.E. 1965. Inhibition of complement by carrageenm:

Mode of action, effect on allergic reactions, and on comple

ment of various species. Immunology. 5:291.

19. Borsos, T., H.J. Rapp, and C. Crisler. 1965. The interaction

between carrageenm and the first component of complement.

J. Immunol. 94:662.

20. Fong, J.S.C., and R.A. Good. 1972. Suramin, a potent reversible

and competative inhibitor of complement system. Clin. Exp.

Immunol. 10:121.

21. Eisen, V., and C. Loveday. 1973. Effects of suramin on comple

ment, blood clotting, fibrinolysis, and kinin formation.

Br. J. Pharmac. 43:678.

21. Henson, P.M. 1971. Interaction of cells with immune complexes:

adherence, release of constituents, and tissue injury. J. Exp.

Med. 134:114.

23. Casey, F.B., and S. Tokuda. 1973. A comparative study of the

mechanisms of passive cutaneous anaphylaxis induced by mouse

138

IgG, rabbit IgG, and rabbit F(ab,)2 antibodies. Int. Arch.

Allergy. 44:644.

24. Janoff, A. 1970. Mediators of tissue damage in human poly-

norphonuclear neutrophils. Ser. Haemat. 3:96.

25. Ward, P.A., and C.G. Cochrane. 1965. Bound complement and

immunologic injury of blood vessels. J. Exp. Med. 121:215.

26. Cochrane, C.G., E.R. Unanue, and F.J. Dixon. 1965. A role of

polymorphonuclear leukocytes and complement in nephrotoxic

nephritis. J. Exp. Med. .222:99.

27. Cochrane, C.G. 1967. Mediators of the Arthus and related

reactions. Progr. Allergy. ll:i.

28. Cochrane, C.G. 196 8. The role of immune complexes and comple

ment in tissue injury. J. Allergy. 42:113.

29. Cochrane, C.G., H.J. Müller-Eberhard, and B.S. Aiken.

Depletion of plasma complement in vivo by a protein of cobra

venom: its effect on various immunologic reactions. J. Immunol.

105:55.

139

Chapter Vili

SUMMARY

141

Rejection mediated by antibodies accounts for a considerable

amount of graft losses in renal transplantation. The detrimental

effects of antibodies that are already circulating at the time of

grafting or that are formed after transplantation might be di

minished by interfering with the effector systems triggered by

these antibodies. Of these effector systems the complement system

is one of the most important, and an understanding of its precise

role in antibody mediated rejection (AAR) might provide additional

tools for the management of the rejection process. The aim of

this study was to investigate the role of the complement system

in AAR in an experimental animal model.

A protocol was used in which graft destruction could reproduci-

bly be elicited by the passive administration of anti donor anti

serum to skin graft recipients, whose cellular immune response

was suppressed by treatment with anti mouse lymphocyte serum.

The first set of experiments was performed in an allogeneic skin

graft model in the mouse. In this model administration of exogenous

complement together with antiserum is required for AAR, since the

recipient's endogenous complement system is inefficiently activa

ted by mouse alloantisera.

The influence of antigenic disparity on AAR was studied in nine

donor-host combinations whose histoincompatibility ranged from a

single non-H-2 difference to a disparity for the complete H-2

complex plus multiple non-H-2 differences. With the exception of

combinations with a single or few non-H-2 differences AAR could

be induced in all combinations. It was found that with increasing

histoincompatibility, the amount of antibodies required for the

induction of AAR decreased, parallel to the decrease of the median

survival time of skin grafts in untreated animals, and also paral

lel to the increase of the cytotoxic titer of the anti donor serum

in vitro. It is concluded that the results for immunogenicity of

the different histocompatibility loci as described for cell-mediated

rejection, also apply to our model of humoral graft destruction.

Complements from various species differ in their effectiveness

in in vitro lymphocytotoxicity tests with mouse alloantisera. All

antisera, when tested with rabbit complement (RC), gave rise to a

higher cytotoxic titer than with guinea pig complement (GPC) or

142

human complement (HC). These differences in effectivity were

mirrored in the in vivo studies. In some donor-host combinations

with relatively small antigenic disparity (multiple non-H-2 dif

ferences, H-2D or H-2K) , where AAR could be elicited by injection

of RC along with anti donor serum, GPC and HC were ineffective.

In the combinations with a greater histo-incompatibility (H-2D

plus non-H-2, H-2K plus non-H-2, H-2 and H-2 plus non-H-2) GPC

and HC were also effective. However, dose response studies showed

that GPC was less efficient than RC, for a greater amount of anti

serum was necessary to induce AAR with it. In none of the allo

geneic donor-host combinations were we able to induce AAR by in

jection of antiserum alone. This is in agreement with the fact

that mouse complement never caused lymphocytotoxicity of target

cells in vitro, not even with the strongest antiserum. In general

terms one can conclude that the cytotoxic titer of an antiseruir.

as tested in vitro gives a good indication about its destructive

power in vivo.

The question whether all components of the complement system

are necessary for the induction of AAR was studied in an allo

graft model in which a complete H-2 disparity existed between

donor and recipient (B10.D2 -* C57BL6). As stated above AAR can be

evoked in such a combination by injection of anti donor serum

along with RC, GPC or HC. Experiments were done in which allo-

antiserum was injected together with sera from animals deficient

in single complement factors or, alternatively, with sera that

were artificially depleted of one of the complement components

(R-sera). The in vivo effects of the various R sera were again

compared with their in vitro activity in the lymphocytotoxicity

assay. None of the R-sera (Rl, R4, R3, R6 or R9) was able to

mediate lymphocytotoxicity when tested with B10.D2 target cells

and C57BL6 anti B10.D2 serum. Reconstitution with the deficient

complement factor restored the lymphocytotoxic capacity of all

the R sera. If normal mouse serum was added to Rl and R9 serum,

these regained their cytotoxicity, demonstrating that mouse CI

and C9 could replace the deficient factor. For CI this was further

substantiated by the finding that a crude mouse CI preparation

had the same restoring effect on Rl serum as whole mouse serum.

143

The R4, R3 and R6 sera could not be reconstituted with normal

mouse serum. If the different R sera were injected with anti donor

serum in C57BL6 recipients of a B10.D2 skin graft results analogous

to the in vitro findings were obtained. Rl and R9 serum induced

AAR without prior reconstitution. On the other hand C4, C3 and C6

depleted complement sources did not evoke AAR. This was only

obtained if, prior to administration, these R sera were reconsti

tuted with the appropriate complement factor. These experiments

permit the following conclusions. The close correlation between

the in vivo activity and the cytotoxicity in vitro strongly sug

gests that activation of the classical complement pathway is

necessary to induce AAR in this allogeneic model. This is corro

borated by the finding that R4 serum, which can be activated nor

mally via the alternative pathway, did not elicit AAR. Furthermore,

the complete complement sequence has to become activated in this

model. Generation of chemotactic complement factors (C3a and C5a)

alone, is not sufficient to cause AAR, since R6 serum, that can

generate normal chemotactic activity, did not induce AAR. Lastly,

the low cytotoxic properties of mouse complement do not reside

in factors CI and C9, because these factors could effectively

restore Rl and R9 serum.

The aim of the second set of experiments was to elucidate the

role of endogenous complement in AAR. For this purpose a xeno

geneic skin graft model was used in which a PVG/c rat skin graft

was placed on immunosuppressed C57BL6 or DBA/2 mice. In this model

of greater antigenic disparity administration of exogenous comple

ment is not required and the injection of mouse anti rat serum

alone causes a rapid and violent destruction. Complement consump

tion was measured during AAR, and complement dependency of the

destruction was studied by using complement depleted recipients

and by the administration of non-complement fixing antibodies.

Before this study was undertaken a sensitive assay of mouse

complement had to be developed in order to measure complement con

sumption and complement inhibition. Because under test conditions

normally used for the measurement of hemolytic activity of human

and guinea pig complement no activity is found with mouse comple

ment, several conditions of the test were investigated to increase

144

its sensitivity. It was found that 7S IgGl amboceptor activated

mouse complement better than 19S IgM amboceptor. The use of a

Cr release assay allowed the reduction of the number of target

cells. Furthermore it was found that optimal incubation conditions

were: an ionic strength of the incubation medium of 0.13 M NaCl,

an incubation time of 90 minutes, and an incubation temperature

of 30 C. With this test it was possible to measure hemolytic ac

tivity of mouse complement sensitively and reproducibly.

With the above mentioned method we found complement consumption

during AAR of a PVG/c skin graft carried by C57BL6 recipients.

Ten minutes after the i.v. injection of mouse anti rat serum there

was a sharp decrease in CH50 titer, which remained depressed during

the rejection process. After the rejection had been completed the

CH50 titer gradually rose to normal levels. Subsequently, attempts

were made to inhibit complement activity and to prevent AAR with

four agents which are anticomplementary in vitro. With pentamidine,

carrageenin and flufenamic acid no significant decrease of the

complement activity in C57BL6 mice was obtained. Treatment of

these agents did not prevent the occurrence of AAR. Although ad

ministration of suramin caused a substantial decrease in CH50 titer

(20% of normal values), it protected only 2 of 7 grafts against

AAR. Prevention of AAR was only found if a complete and prolonged

depletion of the complement system of the C57BL6 recipient was

induced by treatment of CoVF prior to the administration of anti

donor serum. Similar results were obtained in C5 deficient DBA/2

recipients. The protective effect of CoVF pretreatment or of C5

deficiency, was, however, overcome if a greater dose of mouse anti

rat globulin (MARG) was given to the recipients. This was even the

case if C5 deficient recipients were also treated with CoVF.

With 0.5 ml MARG, AAR was obtained in all recipients despite C3

depletion, C5 deficiency, or C3 depletion plus C5 deficiency.

These results indicated that complement independent mechanisms

became operative in this xenogeneic skin graft model, provided

sufficient amounts of antibody were presented to the graft.

A further argument for this was obtained by the finding that non-

complement fixing IgGl antibodies, prepared from MARG, could in

duce AAR. Although it is as yet not clear which mechanisms are

145

operative in the complement independent destruction, we have

acquired some clues from histologic examinations of the grafts.

In animals with a normal complement status the destruction of the

xenografts was brought about by acute intravascular coagulation

without primary involvement of polymorphonuclear leukocytes (PMN).

Xenografts rejected by complement depleted recipients showed

dense intravascular accumulation of PMN, with only secondary signs

of intravascular coagulation. These PMN must have been brought

into the graft by complement-independent mechanisms. This is

possibly the result of immune adherence of PMN to the Fc part of

the antibodies bound to the vascular endothelium. Elucidation of

the mechanisms in this type of rejection clearly requires further

s t udy.

When extrapolating these results to the clinical transplanta

tion situation, one may conclude that it is worthwile to continue

the search for effective anticomplementary agents, since these

might lead to attenuation of the humoral rejection process.

Agents that inhibit the complement components of the classical

activation unit are preferable, because they have the advantage

of leaving the alternative pathway intact. This is important for

the complement dependent defense mechanisms against bacteria.

Whether complement independent mechanisms can become operative in

the vascular rejection of human kidney grafts remains speculative,

until we know more about the role of immunoglobulin subclasses

and about the extent of antigenic disparity necessary to initiate

this process.

146

WOORDEN VAN DANK

Allen die hebben bijgedragen aan het tot stand komen van dit

proefschrift wil ik gaarne bedanken.

De transplantatie experimenten en de ontelbare serologische be

palingen konden worden verricht dankzij de enthousiaste hulp van

Jacqueline Hagemann. Zij werd in de loop der jaren bijgestaan door

Trees Borggreve, Ton Lensen en Peter Daamen.

Voor het verkrijgen van diverse sera en de bereiding van meerdere

antisera werd ik op deskundige wijze geholpen door de medewerkers

van afdeling IV (hoofd: dhr. J. Koedam) van het Centraal Dieren

laboratorium van de Universiteit (hoofd: Dr. W.J.I. v.d. Gulden).

Bij het selecteren en verkrijgen van de diverse dierstammen deed

ik nooit vergeefs een beroep op de kennis en hulp van dhr. J. van

Gaaien en dhr. F. Janssen van het Centraal Dierenlaboratorium.

De discussies over de resultaten van het onderzoek met de mede

werkers van het laboratorium voor nefrologie (Rob de Waal,

Simon Lems en Wim Tamboer) waren steeds erg plezierig en vormden

een essentiële bijdrage aan de voortgang en de tot stand koming

van dit proefschrift. Ondanks vele tegenslagen isoleerde Wim Tam

boer met onnavolgbare vasthoudendheid de verschillende immuun-

globulinen en subklas antilichamen.

Van de adviezen van Mej. Drs. W.C.A. Buys (Ir. Reichertlaborato-

rium) bij het ontwikkelen van de Cr release assay werd dankbaar

gebruik gemaakt. Evenmin mag onvermeld blijven de vriendschappe

lijke hulp die ik kreeg van Mevr. A.C. Feiten (Ir. Reichert-

laboratorium) en dhr. G. Grutters (Isotopenlaboratorium van het

Centraal Dierenlaboratorium) bij de praktische uitvoering van de 5 ΙΟ r assay.

De tekeningen voor dit proefschrift werden met zorg gemaakt door

dhr. H. Berris van de afdeling Medische Illustratie (hoofd: dhr.

W. Maas). De medewerkers van de afdeling Medische Fotografie

(hoofd: dhr. A. Reynen) vervaardigden de foto's.

De Heer E. de Graaff (hoofd Medische Bibliotheek) en zijn staf

waren behulpzaam bij het verzamelen van de literatuur.

Het typen van de talrijke versies van de tekst werd met grote

toewijding en nauwkeurigheid verricht door Angèle Wentholt.

148

CURRICULUM VITAE

De auteur van dit proefschrift werd geboren op 26 mei 1948 te

Blenck. Hij bezocht het St. Thomascollege te Venlo en behaalde

het einddiploma HBS-B in 1965. Daarna studeerde hij geneeskunde

aan de Katholieke Universiteit te Nijmegen, waar hij in 1971 het

doctoraal- en in 19 73 het artsexamen aflegde. Vanaf 1 augustus

1973 was hij m opleiding tot internist, aanvankelijk m het

St. Joseph Ziekenhuis te Eindhoven (hoofd van de opleiding:

Dr. P.F.L. Deckers) en vanaf 1 februari 1975 aan de Universiteits

kliniek voor Inwendige Ziekten (hoofd: Prof.Dr. C.L.H. Majoor).

Op 1 augustus 19 78 werd hij ingeschreven als internist in het

specialistenregister. Vanaf 1 januari 1979 is hij werkzaam op de

afdeling nierziekten (hoofd: Prof.Dr. P.G.A.B. Wijdeveld) van de

Universiteitskliniek voor Inwendige Ziekten.

Hij is getrouwd met Tilla Crommentuyn, en heeft 2 kinderen:

Joost en Bas.

149

STELLINGEN

1

De regels voor de immunogeniciteit van de diverse loci binnen het

H-2 complex gelden niet alleen voor de cellulaire maar ook voor de

humorale afstoting van huidtransplantaten.

2

Acute antilichaam afhankelijke afstoting van allogene huidtrans

plantaten in de muis vereist een aktivatie van de volledige

klassieke route van het complement systeem.

3

Acute antilichaam afhankelijke afstoting van xenogene huidtrans

plantaten in de muis kan via complement onafhankelijke mechanismen

verlopen.

4

Transplantaatbeschadiging door antilichamen kan zowel het gevolg

zijn van een lokale Shwartzman reactie als van een klassieke

Arthus reactie. Welke van de twee tot ontwikkeling zal komen hangt

af van het antigene verschil tussen donor en ontvanger, en van de

titer en subklas specificiteit van de circulerende antilichamen.

5

In tegenstelling tot allogene niertransplantaties bij het konijn,

induceert Cyclosporine A geen blijvende tolerantie bij muizen die

een allogeen huidtransplantaat dragen.

- S.P.M. Lems en R.A.P. Koene (1979)

ter publicatie aangeboden

- C.J. Green en A.C. Allison

Lancet (1978) , 1, 1182

6

Bij een patient met een niet-purulente mono- of oligoarthritis

dient men ook een Campylobacter jejuni infectie als mogelijke

oorzaak uit te sluiten.

- J.H.H. Berden, H.L. Muytjens en L.B.A. v.d. Putte

British Medical Journal (1979), !_, 380

7

Immuuncomplexen en complement aktivatie spelen een belangrijke

rol bij de ontwikkeling van de complicaties van een Plasmodium

falciparum infectie.

- WHO Technical Report (1975) 579

- eigen waarneming

8

Bij een onbegrepen acute nierinsufficientie moet als mogelijke

oorzaak de ziekte van McArdle (niet-traumatische rhabdomyolyse)

overwogen worden.

- eigen waarneming

9

Er zijn voldoende aanwijzingen voor een immunologische Pathogenese

van multiple sclerose om goed opgezette pogingen ter behandeling

van deze ziekte met immunosuppressiva te rechtvaardigen.

- P. Delmotte, O.R. Hommes en R. Gonsette (redacteuren)

Immunosuppressive treatment in multiple sclerosis.

European Press, Gent, 1977

10

In een toekomst ige t r a n s p l a n t a t i e wetgeving v e r d i e n t t en a a n z i e n

van d o n a t i e van organen h e t "geen-bezwaar-sys teem" de voorkeur

boven h e t " toesternmingssysteem" .

- H.J.J. Leenen

Nederlands Juristen Blad (1978) , 53_, 871

11

Het onderwijs in de medische geschiedenis verdient een ruimere

plaats in de medische opleiding.

Nijmegen, 27 april 19 79 J.H.M. Berden

the role of complement in acute antibody ...skin allografts could not be destroyed by injection of...

Documents