Functional effects of domain deletions in a multidomain serine protease, Clr

Molecular Immunolog.v, Vol. 33. No. 415, pp. 351-359, 1996 Copyright c 1996 Elsevier Science Ltd. All rights reserved Printed in Great Britain 0161-5890/96 $15.00 +O.OO

Pergamon 0161~5890(95)00160-3

FUNCTIONAL EFFECTS MULTIDOMAIN SANDOR CSEH,* ZSOLT LbRINCZ,*

PETER VERNE

OF DOMAIN DELETIONS SERINE PROTEASE, Clr

IN A

GAL,* MIKL& SARVARI,* J6ZSEF DOB&* N. SCHUMAKERf and PETER ZAVODSZKY*$

*Institute of Enzymology, Biological Research Center, Hungarian Academy of Sciences, Budapest Pf. 7., H- 1518 Hungary; TThe Molecular Biology Institute and Department of Chemistry and Biochemistry, University of California, Los Angeles, CA 90024, U.S.A. (First received 6 September

1995; accepted in revisedform

24 November

1995)

C 1r subcomponent of the first component of complement is a complex, multidomain glycoprotein containing five regulatory or binding modules in addition to the serine protease domain. To reveal the functional role of the N-terminal regulatory domains, two deletion mutants of C 1r were constructed. One mutant comprises the N-terminal half of domain I joined to the second half of the highly homologous domain III, resulting in one chimeric domain in the N-terminal region, instead of domains I-III. In the second mutant most of the N-terminal portion of domain I was deleted. Both deletion mutants were expressed in the baculovirus-insect cell expression system with yields typical of wild type Clr. Both mutants maintained the ability of the wild type Clr to dimerize. The folding and secretion of the recombinant proteins was not affected by these deletions, and Clinhibitor binding was not impaired. The stability of the zymogen was significantly decreased however, indicating that the N-terminal region of the Clr molecule contains essential elements involved in the control of activation of the serine protease module. Tetramer formation with Cl s in the presence of Ca*+ was abolished by both deletions. We suggest that the first domain of Clr is essential for tetramer formation, since the deletion of domain I from Clr impairs this interaction. Copyright cc) 1996 Elsevier Science Ltd. Abstract-The

Key words: Complement

activation, glycoprotein, zymogen, baculovirus.

INTRODUCTION The first component of complement Cl, plays an important role in the initiation of the classical pathway activation. Cl is a complex heteropentameric protease consisting of one Clq, two Clr and two Cls subunits. (It was reviewed by Arlaud et al., 1987; Schumaker et al., 1987.) The non-enzymatic Clq, which can bind to immune complexes, is the recognition part of the complex and provides a framework for the C 1s-C 1r-C 1r-C 1s tetramer (Reid et al., 1977). Clr and Cls are serine proteases which form the Cl r,Cl s2 tetramer in the presence of Ca2+ (Villiers et al., 1985). This tetramer binds to the

IAuthor to whom correspondence should be addressed. AcMNPV: Autographa californica nuclear polyhedrosis virus; Cl: first component of complement; Clq, Clr, Cls: subcomponents of Cl; Clrcr: a fragment of Clr containing the N-terminal part of the A chain; ClryB: a fragment of Clr containing the C-terminal part of the A chain and the B chain; CCP: complement control protein; EGF: epidermal growth factor; ELISA: enzyme linked immunosorbent assay; FPLC: fast protein liquid chromatography; Sf9: Spodoptera frugiperda 9 cell line; SDS-PAGE: sodium dodecyl sulfate-polycrylamide gel electrophoresis; TBS: Tris-buffered saline; Tris: Tris (hydroxymetyl)aminomethane.

Abbreviations:

collageneous arms of Clq to form the functionally active Cl. The first step in the classical pathway is the autoactivation of the Clr zymogen by the cleavage of an ArgIle bond in the catalytic part of the molecule. Activated Clr then cleaves the corresponding Arg-Ile bond in the Cls zymogen converting it into an active serine protease which will activate subsequent complement components (C4 and C2) of the classical pathway. Like other serum proteases, Clr and Cls are mosaic proteins as a result of exon shuffling (Patthy, 1985; Leytus et al., 1986; Tosi et al., 1987). The trypsin-like catalytic domain is preceded by five non-catalytic or interacting domains, including two loo-residue repeats which are rather specific for Cl r and Cl s (domain I and III), one EGF-like domain II, and two CCP motifs (domain IV and V) which are closely associated with the serine protease module (Leytus et al., 1986; Tosi et al., 1987). Previous studies showed (Arlaud et al., 1986; Lacroix et al., 1989; Busby and Ingham, 1990) that Clr and Cls can be fragmented into two functionally distinct regions:

351

(1) The catalytic module yB (including the two CCP motifs and the serine protease domain). (2) The a-fragment (domain I, domain II (EGF-like) and a part of domain III). The a-regions

are responsible

for the Ca’+-dependent

352

S. CSEH et al.

Clr-Cls interaction and probably for the interaction between the Clr,Cls, tetramer and Clq. In the case of Clr, the yB region is responsible for the dimerization and has autoactivation capacity (Lacroix et al., 1989). The exact role of the individual domains in these specific interactions is unknown, and the available models describing the Cl structure and activation contain speculative elements. Recently we have expressed cDNA clones for human Clr and Cl s in insect cells using baculovirus vectors. The post-translational modifications and the biological activity of the recombinant proteins were checked (Gal et al., 1989; Sarvari et al., 1990; Luo et al., 1992; Monkovic et al., 1992). The objective of the present work was to clarify further the role of the N-terminal domains of Clr in the stuctural organization and function of Cl. For this purpose we have constructed two deletion mutant Clr cDNA clones. The mutants were expressed in the baculovirus-insect cell system and their properties were characterized.

MATERIALS

AND METHODS

Cells and viruses

The insect cells Spodoptera frugiperda (X9) and wild type Autographa californica nuclear polyhedrosis virus (AcMNPV) were provided by Max Summers (A&M University, TX) and maintained as summarized by (Summers and Smith, 1987). The baculovirus transfer vector pAcYM1 was a generous gift from David H. L. Bishop (Institute of Virology, Oxford, U.K.) and is described in (Matsuura et al., 1987). Cells were grown at 27°C in TNM-FH [supplemented Grace’s insect medium with 10% fetal calf serum (Serva, Heidelberg, Germany), 3.3g/l lactalbumin hydrolysate (Sigma, St Louis, MO) and 3.3 g/l yeastolate (Oxoid, Basingstoke, U.K.)]. Recombinant viruses were produced as described (Summers and Smith, 1987). Clr cDNA The full length cDNA coding sequence HClr2200 for human Clr in pUC8 was generously provided by Prof. Earl Davie (University of Washington, Seattle, WA), and is described in (Leytus et al., 1986).

concentrated to 50-fold by a PM10 ultrafiltration brane (AMICON, Beverly, MA).

mem-

ELBA

The Clr, Cls and Cl-inhibitor contents of the samples were determined by a solid-phase ELISA “sandwich” system. The assay was carried out essentially as described previously (Nakamura et al., 1986). Goat anti-Clr, antiCls and anti-Cl-inh antibodies were purchased from Atlantic Antibodies (Scarborough, ME). Rabbit anti-Clr and anti-Cl s antibodies were obtained from Calbiochem (San Diego, CA), anti-Cl-inh from Sigma. Goat antirabbit IgG antibody conjugated to horseradish peroxidase was a Sigma product. In the case of Clr-ELISA, purified Clr (Calbiochem) was used as a standard to determine the Clr concentration. Western blot

Gels containing 11% acrylamide were prepared as described by Laemmli (1970). For immunoblotting we followed the method of Towbin (Towbin et al., 1979). The nitrocellulose strips were first incubated with goat anti-Clr antibody, then with horseradish peroxidase labeled rabbit anti-goat IgG conjugate. 4-Chloro- 1-naphthol (Sigma) was used as substrate for the peroxidase. GelJiltration

Size exclusion chromatography was performed on a Superose 12 column using a Pharmacia FPLC system. Unless specifically noted, experiments were performed in 0.10 M Tris-HCl, pH 6.5 and 0.15 M NaCl (TBS). The proteins were eluted at a flow rate of 0.4 ml/min with TBS containing 10 mM CaCl, or 50 mM EDTA. To monitor the protein of interest, fractions were collected every minute and measured by ELISA. The column was calibrated by the following proteins: (1) IgG: MW = 160 kDa, t = 30 min; (2) human serum albumin: MW = 67 kDa, t = 33.2 min; (3) /3-lactoglobulin: MW = 35 kDa, t = 35.7 min; (4) cytochrome c: MW = 12.4 kDa, t=40.9 min. RESULTS

DNA constructions

Construction

Restriction enzymes, T4 DNA ligase, Klenow polymerase and mung bean nuclease were purchased from New England Biolabs (Beverly, MA). DNA manipulations were carried out essentially as summarized by Maniatis (Sambrook et al., 1989).

Mutant I. Our purpose in preparing this mutant was to delete the EGF-like domain II from the Clr molecule by taking advantage of the high degree of homology between domains I and III of Clr. A convenient BglII site occurs in both modules (I and III) at equivalent positions allowing deletions of 179 residues (Ile78Gln256) without affecting the reading frame. The mutant cDNA lacks the C-terminal half of domain I, the entire EGF-like module II and the N-terminal half of module III. In other words, it contains a chimeric domain composed of the N-terminal half of domain I and the C-terminal half of domain III, and domains IV, V, VI (Fig. 1A).

Expression

of recombinant

protein

Sf9 cells (2 x 107) were infected in a 175 qcm flask at a multiplicity of infection l-3. After 1 hr at 27°C the medium was changed to Grace’s medium. The cell culture supernatant was harvested 72 hr later. The sample was

of the mutant cDNAs

353

Domain deletions and regulation of serine protease Cl r

Expression of the recombinant proteins

Mutant II. In this case we deleted the DNA fragment corresponding to most of module I from the Clr Nterminal region. In this experiment we used a modified Clr cDNA in which the PstI site at module IV was eliminated by oligonucleotide-directed previously mutagenesis. We deleted the sequence Cys13-Pro99 by PstI-Xmal digestion. In order to restore the reading frame, the Xmal 5’ protruding end was end-filled by Klenow polymerase and the PstI 3’ single stranded end was removed by digestion with mung bean nuclease before circularization of the plasmid. The DNA construct was checked by sequencing. The resulting cDNA contains the leader sequence and encodes 22 amino acid residues from domain I preceding domains II, III, IV, V and VI (Fig. IB).

DELETED

(A)

The deletion mutant C 1r cDNAs were subcloned into the pAcYM1 transfer vector under the control of the strong polyhedrin promoter. After cotransfection of St9 cells with the recombinant plasmids and wild type AcMNPV DNA, recombinant virus clones were isolated. Sf9 cells were infected with the recombinant virus stocks at a MO1 of 1-3. Three days post-infection, cell culture medium was collected and analysed for the presence of the recombinant proteins. Both mutant proteins were secreted into the cell culture medium. The concentrations of the mutant proteins were estimated by ELISA using polyclonal anti-Clr antibody. The yields of both the mutant proteins were about 3 pg/ml, equivalent to that

SEQUENCE

+

i

K

P

K

ct

s

Y

P

P Nb

i P t S G G A LFCR P

F” r’ 1 T

p” L c LY I. W

‘F C G

p” G’ y R

FR

E’ L 0 F 0

LOE E Y

0 F

y

P

N I

0

N” “c

0

0 5 M

k s P I_ G N

SRG WK

F” G” E : G

R

;

:

I

; T

tFtIT

t’ L’

T

0 T F

EE N* : ci

E K K

Y

ii T

K

PPG

: L 0 V

:: 0

t ht F

S sNA

P VKLV

r Nk?

~GOP AT I

crrV

Co

R% P ‘a

NG

E S

K” : 0 K G E K

E R S G A

0 F R Y T T

T”

I:

IJo

V

,”

I P EG@l

R

ALLELENsvTLGPNLLP~~~G~ v

CATALYTIC bOYAW4

P

L

C-C

pYyK tt E YYOI

N AKyT

R

PL

A~KEEMVGFGSVYGMLGLOYFTO~~ . on

Fig. 1. Primary structure of the deletion mutant Clr molecules. Deletion mutant I (A), deletion mutant II (B). (Partly adopted from Leytus et ai., 1986) The framed sequences were deleted from the cDNA of C 1r and the remaining parts were fused.

354 W

S. CSEH et al.

DELETED SEQUENCE

” T n F S

EDP

:

E

E N* G T 1 M F

“s

L”E D E R Y ki G PR s 0 c-C P S OAE 0 c-c O H L

KS ‘?s

FY GG

ACE_CH NYv

D Y

L D V

KG FL AYyQA

V*N

’

T * L

p’

E s

0’ L

s c-c

R

K N

L 0 D I

N. r

d

P

:

r”, V

K

00 Ii D D

: R V E

: P P 0 L 0 T

:,

c” L T L H

G

0

PF

LK

F

Ss

L’

N

Go

G

s

D E R s G A TR

K NF

557 EG~~IALLELENSVTLGPNLLPJ CATALYTIC

L

DH

K G c-c ,L

OOYAU(

R v

P

L

C-C

11 E

V

M’ G

MP NV

Y

P

F R Y T T

T

PY KY

YYOIRA

AIKEEMVGFGSVYGMLGLOYFTDYD

WI

FR V R L P ”

KK E ! E

705 YE CCXM

Fig. l-Continued.

of the wild type Clr. SDS-PAGE and Western blot were also used to identify the Clr mutants (Fig. 2). The molecular weight of mutant I and II was about 65 kDa and 75 kDa as it was expected (Fig. 2a). In the case of mutant I we could detect a smeared band (double band) on the blot at 30-35 kDa under reducing conditions (Fig. 2b), representing the Clr B chain (32 kDa) and the shortened A chain (34 kDa). In the case of mutant II the pattern was similar, although reflecting the larger size of the A chain (45-50 kDa) (Fig. 2~). On reducing gels both mutant proteins appeared to be totally cleaved, reflecting the fast activation, while the wild type Clr was only partially cleaved under similar conditions. Interaction

with Cl subcomponents

A major purpose of our study was to characterize the interactions between these mutant proteins and the other Cl subcomponents (e.g. Cls, Cl-inhibitor) in order to

map within the N-terminal portion of the A chain of Clr those regions participating in these interactions. Dimerization. Gel filtration experiments were used to characterize the self-associating properties of the deletion mutants. At pH 6.5, where the wild type Clr was known to exist as a dimer, it eluted from the Superose 12 column at 30 min (Fig. 3A). At pH 5.0, the peak shifted to 34 min due to the dissociation of the Clr dimers. Both mutants showed the same behaviour: at neutral pH the proteins eluted at about 30 min, whereas at pH 5.0 the elution time increased accordingly (Fig. 3 B,C). These results demonstrated that both mutant proteins formed dimers at neutral pH and physiological ionic strength. Tetramerformation. To test whether the mutant dimers could form the Cls-(Clrmut)z-Cls tetramer, we incubated the mutant proteins with human serum Cls at 30°C for 20 min in the presence of 10 mM Ca*+. The Cls and the mutant dimers eluted separately from the Superose 12 column, we could not detect any shift toward the void

Domain

deletions

and regulation

of serine protease

Clr

355

mutant I

1 -

1

2

43 kD

(a)

F

94 kD

A and B chains of mutant I -

1

-

A chain

-

B chain

30 kD

2 (b)

- - 94 kD -

Clr

- - 67 kD -

A chain

A chain of mutant II - 43 kD

-

B chain of mutant II -

- - 30 kD

B chain

: 1

2

Fig. 2. Western blot of the deletion mutant Clr proteins. (a) Deletion mutant I and II under nonreducing conditions. Lane 1: deletion mutant II. Lane 2: deletion mutant I. (b) Deletion mutant I and Clr under reducing conditions. Lane 1: deletion mutant I. Lane 2: wild type recombinant Clr. (c) Deletion mutant II and Clr under reducing conditions. Lane 1: deletion mutant II. Lane 2: wild type recombinant Clr.

S. CSEH et al.

356 Wild type Clr

2 1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2

(A)

0 20

22

24

26

28

30

Time

32

34

36

38

40

(min)

Figure 3/A

Mutant I

1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0

20

22 24 26 28 30 32 34 Time

36 38 40

(min)

Figure 3/B

Mutant II

1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0

20

22

24 26 28 30 32 34 36 Time

38 40

(min)

Fig. 3. Identification of wild type and deletion mutant Clr dimers (at pH6.5) and monomers (at pH 5.0) by size exclusion chromatography. Recombinant Clr (A), deletion mutant I (B), and deletion mutant II (C) were examined. Dark circles indicate the elution profiles at pH6.5. Two hundred microliter samples containing approximately 20 pg recombinant protein were loaded onto a Superose 12 column equlibrated in TBS-50 mM EDTA and eluted at a flow rate of 0.4 ml/min. Empty circles show the results at pH 5.0. In these experiments, each sample was first dialysed against acetate buffer (0.05 M Na-acetate/acetic acid, pH 5.0 and 0.15 M NaCl), then a 200~1 solution containing about 2Opg recombinant protein was loaded onto the Superose 12 column equlibrated in acetate buffer. Fractions (0.4 ml) were collected and the elution of Cl r was determined by ELISA.

volume, which indicated that there was no detectable complex formation between these subcomponents. In the control experiments with wild type Clr, the elution times for the Clr, and the Clr,Cls, tetramer were 30 min and

26 min, respectively (Fig. 4). These results indicated that the regions deleted from the Clr were essential for tetramer formation. Cl-inhibitor binding. In order to check whether these

351

Domain deletions and regulation of serine protease Cl r 0.6

DISCUSSION Studies with the yB proteolytic fragment of Clr have revealed that the lack of modules I,II,III results in a

0.4

complete

a

0.2

0 20

22

24

26

28

30

32

34

36

38

40

Time/min

Fig. 4. Demonstration of the lack of interaction between Cls and deletion mutants. Samples of 10 pg deletion mutant I (empty circles); and II (empty triangles) and recombinant wild type Clr (dark circles) were incubated (20 min, 30°C) with 10 pg C 1s in TBS-10 mM CaCl, ~. The samples were applied to Superose 12 column equilibrated with the same buffer and eluted at a flow rate of 0.4ml/min. Fractions of 0.4 ml were collected and the elution of Clr was determined by ELISA.

deletions affected the C-terminal serine protease module, we studied the C 1-inhibitor binding ability of the deletion mutants. The highly glycosylated Cl-inhibitor, which is a member of the serpin family, can form a heat- and denaturantstable complex with activated Clr and Cls (Salvesen et al., 1985). On the Superose 12 column, the Cl-inhibitor eluted at 28 min, which is rather high for a molecule of 105 kDa, reflecting the high level of glycosylation of this molecule. After incubation (lOmin, 37°C) with any of the deletion mutants, the elution time decreased to 24 min (Fig. 5) indicating complex formation between C 1-inhibitor and the deletion mutant proteins.

0.6

: cv

loss of interactions

with Cls

(Arlaud

et al.,

1980) and Clq (Busby and Ingham, 1990). Our two deletion mutants represent a gradual approach to the problem using well defined recombinant subunits. Mutant I comprises a single Clr-Cls type module followed by the whole yB region, while in mutant II only

04

%

a 0.2

0 16

18

20

22

24

26

28

30

32

34

36

Timelmin

Fig. 5. Demonstration of the ability of deletion mutants to bind Cl-inhibitor by gel-filtration. Ten micrograms Cl-inhibitor in 200 ~1 TBS was loaded onto a Superose 12 column equlibrated in TBS and eluted at a flow rate of 0.4ml/min (dark circles). Samples of 10 pg deletion mutant I or deletion mutant II mixed and incubated (10 min, 37°C) with 10 pg Cl-inhibitor in 200 ~1 TBS were applied to the Superose 12 column (empty circles; empty triangles). The elution of Cl-inhibitor was determined by ELISA.

the first C 1r-C 1s module has been deleted. Both mutants maintained their ability to dimerize. Like the intact C 1r, these dimers dissociated at acidic pH (pH 5.0). Therefore we conclude that these deletions in the CIregion did not affect the folding and the self association potential of the yB domain. Arlaud and co-workers have published related results using (yB), fragments of Clr produced by limited proteolysis of serum Clr (Arlaud et al., 1986). The deletions within the a-region affected the stability of the zymogen form of the mutant proteins. In the case of recombinant Clr, a significant amount of the product was in zymogen form, as judged from Western blots (Fig. 2b). In the case of the two deletion mutants, however, the recombinant components were extensively activated. We could only detect a slight amount of the zymogen form even though the concentration of Cal+ in the medium (9 mM) was high (Fig. 2a). It is well known that the autoactivating ability of isolated serum Clr is markedly decreased in the presence of Ca’+ (Cooper, 1985) in contrast to the isolated (yB),, which is insensitive to Ca’+ ions. Arlaud and co-workers have suggested, that all the structural elements that are necessary for autoactivation are localized in the (yB), domains and the N-terminal cc-region is responsible for the Ca”+ dependent downregulation of Clr autoactivation (Lacroix et al., 1989).

Our results support this view and provide evidence for the regulatory role of the cc-region in the activation of the zymogen. Both deletions led to accelerated cleavage of the ArggIle bond in the serine protease domain. Therefore we conclude that only the intact a-region stabilizes the zymogen form, i.e. protects the Arg-Ile bond from cleavage. Both deletions studied, have decreased this stabilizing effect. These results suggest that the x-region and the yB part in Clr are not independent and a conformational signal from the a-region can be transmitted to the yB catalytic region. The suspected candidate for Ca” binding in Clr was the EGF-like domain II, based on the present study it becomes likely that both domains I and II participate in the Ca’+ binding and Ca2+-dependent interactions, since even the deletion of domain I could decrease the Ca’+ dependent down-regulation. Cl-inhibitor binds to both deletion mutants, indicating that they retained their serine protease activity. CI-inhibitor is a highly glycosylated protein consisting of a compact “head” and an elongated N-terminal “tail” (see Schumaker and Phillips, 1993 for a recent electron micrograph). The C-terminal “head” of the Cl-inhibitor contains the reactive site, which presents a substrate-like structure to the target proteases while the role of the N-

358

S. CSEH et al.

terminal “tail” is unknown. It has been suggested that the N-terminal “tail” interacts with the non-catalytic domains of Clr and Cls reducing their binding affinity to Clq (Perkins et al., 1990). Our results demonstrate that deletions in the a-region of Clr do not abolish the binding of Cl-inhibitor to the activated Clr, i.e. the presence of the N-terminal region of Clr is not essential for the association of the C-terminal parts of the regions. None of the two deletion mutant Clr proteins forms stable tetramer with human serum Cls in the presence of Ca2+ ions. Proteolytic fragmentation studies have shown that the cc-regions of both C 1r and C 1s mediate the Cl rCls interaction and the interaction of the tetramer with Clq (Busby and Ingham, 1990). The cc-region comprises motifs I, II and the N-terminal part of motif III. Domains I and III are both about 100 residues long, and represent a motif which was found first in Clr and later in a few other proteins (Bork, 1991). The functions of these domains are unknown. but they are supposed to participate in interactions specific to C lr/Cls (e.g. tetramer formation, Clq binding). Domain II is an EGF-like sequence found in a variety of secreted or membranebound glycoproteins. It harbours an unusual amino acid, fi-hydroxy-asparagine in position 150 in Clr and 134 in Cls (Thielens et al., 1990). It was suggested that these domains contained the Ca2+ binding site of the a-regions mediating this way the Clr-Cls interaction in the tetramer (Arlaud et al., 1987). Recently, it was shown that the fl-hydroxylation of the Asn residue was not essential for the Ca2+ binding (Luo et al., 1992; Monkovic et al., 1992). In the case of C 1s, the EGF-like motif alone (a2-fragment) did not contain all the structural elements required for Ca2+ binding and Ca2’-dependent protein-protein interaction since fragments of the A-chain of Cls, crl (domain I) and ~12(domain II), were clearly unable to form stable Ca2+-dependent complexes with Clr or Cls (Thielens et al., 1990). In the case of Clr, the contribution of individual domains of the N-terminus can not be resolved from the fragmentation studies and sequence comparisons. With our combined domain deletion experiments the effect of the first and the EGF modules of Clr on the C lr-Cls interaction could be assessed individually. Deletion of module I abolished the ability of Clr to bind Cls, demonstrating that the EGF-like domain in combination with domain III cannot maintain the Clr-Cls interaction. Our results also confirm the significance of the EGF-like module of Clr in the tetramer formation since the deletion of this domain also impaires Clr-Cls binding. Two models have been suggested for the Cl r-C 1s interaction so far. Illy et al. (199 1) have stressed the possible significance of the charge complementarity between Clr and Cls which exists on motif I (Fig. 6A), while Tosi et al. (1987) have emphasized the role of domain II in this interaction (Fig. 6B). Our experiments clearly indicate the essential role of both, domain I and II in the stabilization of Clr-Cls interaction, supporting this way the head to tail Clr-Cls arrangement, as it is shown in Fig. 6A.

A

Clr

&ii$-p Cls

B

I

I

Cls

Clr II

II

88 Fig. 6. Schematic representation of two different proposals for the Clr-Cls interaction (only domain I and II are indicated). (A) Interaction through the first and second domains (adopted from Illy et al., 1991). (B) Interaction through the second domains (adopted from Tosi et al., 1987).

Acknowledgements-We are indebted to Prof Earl Davie, who provided us with the plasmid containing the full-length coding sequence for Clr, to Prof Max Summers for sending us the baculovirus expression system, and to Prof David Bishop for the pAcYM1 baculovirus transfer vector. We thank MS Julia Balczer for her skillful technical assistance.This research has been supported by grants F0173 16, F006303 and TO17478 from the Hungarian Scientific Research Fund (OTKA), a grant N 92-97-48-340 from the Hungarian National Committee for Technological Development (OMFB) and a grant from The National Science Foundation NSF DMB 9120491.

REFERENCES Arlaud G. J., Villiers C. L., Chesne S. and Colomb M. G. (1980) Purified proenzyme Clr some characteristics of activation and subsequent proteolytic cleavage. Biochim. Biophvs. Acta 616, 116129. Arlaud G. J., Gagnon J., Villiers C. H. and Colomb M. G. (1986) Molecular characterization of the catalytic domains of human complement serine-protease Clr. Biochemistry 25, 5177-5182.

Arlaud G. J., Colomb M. G. and Gagnon J. (1987) A functional model of the human Cl complex. Zmmun. Today 8,106-l 11. Bork P. (1991) Complement components Clr/Cls, bone morphogenic protein 1 and Xenopus faeuis developmentally regulated protein UVS.2 share common repeats. FEBS Left. 282, 9-12.

Busby T. F. and Ingham K. C. (1990) NH,-terminal calciumbinding domain of human complement Cls mediates the interaction of Clr with Clq. Biochemistry 29,46134618. Cooper N. R. (1985) The classical complement pathway: Activation and regulation of the first complement component. Adv. Zmmun. 37, 151-216.

Gal P., Sarvari M., Szilagyi K., Zavodszky P. and Schumaker V. N. (1989) Expression of hemolytically active human comp-

Domain

deletions

and regulation

lement component Clr proenzyme in insect cells using a baculovirus vector. Complement InJihmmation 6,433441. Illy C. H., Thielens N. M.. Gagnon J. and Arlaud G. J. (1991) Effect of lactoperoxidase-catalysed iodination on the calcium-dependent interaction of human Cls. Biochemistry 30, 7135-7141. Lacroix M. B., Aude C. A.. Arlaud G. J. and Colomb M. G. (1989) Isolation and functional characterization of the proenzyme form of the catalytic domains of human Clr. Biochem. J. 257,885-89 1. Laemmli U. K. ( 1970) Cleavage of structural proteins during the assembly of the head of bacteriophage Tl. Nature 227, 680-685. Leytus S. P., Kurachi K., Sakariassen S. and Davie E. W. (1986) Nucleotide sequence of the cDNA coding for human complement Clr. Biochemistry 25,48554863. Luo C., Thielens N. M.. Gagnon J., Gal P., Sarvari M.: Tseng Y., Tosi M., Zavodszky P., Arlaud G. J. and Schumaker V. N. (1992) Recombinant human complement subcomponent Cls lacking /3-hydroxyasparagine, sialic acid, and one of its two carbohydrate chains still reassembles with C 1q and C 1r to form a functional Cl complex. Biochemistry 31, 4254 4262. Matsuura Y., Possee R. D., Overton H. A. and Bishop D. H. L. (1987) Baculovirus expression vectors: The requirements for high level expression of proteins. including glycoproteins. J. gen. Virol. 68, 1233-1250. Monkovic D. D.. VanDusen W. J., Petroski C. J., Garsky V. M.. Sardana M. K.. Zavodszky P.. Stern A.M. and Friedman P. A. (1992) Invertebrate aspartyl/asparaginyl P-hydroxylase: Potential modification of endogeneous epidermal growth factor-like modules. Biochem. Biophys. Res. Commun. 189,2333 241 Nakamura R. M., Voller A. and Bidwell D. E. (1986) Handbook qf Experimentul Immunology (Edited by Weir, D. M.) pp 27.1-27.20, Blackwell, Oxford. Patthy L. (1985) Evolution of the proteases of blood coagulation and fibrinolysis by assembly from modules. Cell 41, 657-663. Perkins S. J., Smith K. F., Amatayakul S., Ashford D., Rademacher T. W., Dwek R. A., Lachmann P. J. and Harrison R. A. (1990) The two domain structure of the

of serine protease

C 1r

359

native and reaction centre cleaved forms of Cl-inhibitor of human complement by neutron scattering. J. molec. Biol. 214, 751-763. Reid K. B. M., Sim R. B. and Faiers A. P. (1977) Inhibition of the reconstitution of the hemolytic activity of the first component of human complement by a pepsine-derived fragment of subcomponent Clq. Biochem. J. 161,239-245. Salvesen G. S., Catanese J. J., Kress L. F. and Travis J. (1985) Primary structure of the reactive site of human Cl-inhibitor. J. biol. Chem. 260,2432-2436 Sambrook J., Fritsch E. F. and Maniatis T. (1989) Molecular Cloning: A Laboratory Manual, 2nd ed., Cold Spring Harbor Laboratory Press, Cold Spring Harbor. New York. Sarvari M., Csikos Gy., Sass M., Gal P., Schumaker V. N. and Zavodszky P. (1990) Ecdysteroids increase the yield of recombinant protein produced in baculovirus insect cell expression system. Biochem. Biophvs. Rex Cornmun. 167, 11541161. Schumaker V. N. and Phillips M. L. (1993) Electron microscope studies of Cls, Cls,, Clr?, Clr,Cls, and Cl-inhibitor. Behring Inst. Mitt. 93, 17-24. Schumaker V. N., Zavodszky P. and Poon P. H. (1987) Activation of the first component of complement. ‘4. Reu. Immunol. 5, 2142 Summers M. D. and Smith G. E. (1987) A Manual qf Methods jbr Baculooirus Vectors and Insect Cell Culture Procedures, Tex. Agric. Exp. Stn.. [Bull.]. 1555, College Station, Texas. Thielens N. M.. Van Dorsselaer A., Gagnon J. and Arlaud G. J. (1990) Chemical and functional characterization of fragment of Cls containing the epidermal growth factor homology region. Biochemistry 29, 3570-3578. Tosi M., Duponchel C. and Meo T., Julier C. (1987) Complete cDNA sequence of human complement Cls and close physical linkage of the homologous genes Cls and Cl r. Biochemistry 26, 85168524. Towbin H., Staehelin T. and Gordon J. (1979) Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc. natn. Acad. Sci. U.S.A. 76,435&4354. Villiers C. L., Arlaud G. J. and Colomb M. G. (1985) Domain structure and associated functions of subcomponents Clr and C 1s of the first component of human complement. Proc. natn. Acad. Sci. U.S.A. 82, 4477448 1.