Formation of high-molecular-weight angiotensinogen during ...

Biochem. J. (2013) 449, 209–217 (Printed in Great Britain)

209

doi:10.1042/BJ20120801

Formation of high-molecular-weight angiotensinogen during pregnancy is a result of competing redox reactions with the proform of eosinophil major basic protein Søren KLØVERPRIS*, Louise L. SKOV*, Simon GLERUP*, Kasper PIHL†, Michael CHRISTIANSEN† and Claus OXVIG*1 *Department of Molecular Biology and Genetics, University of Aarhus, DK-8000 Aarhus, Denmark, and †Department of Clinical Biochemistry and Immunology, Statens Serum Institut, Copenhagen, Denmark

The plasma concentration of the placentally derived proMBP (proform of eosinophil major basic protein) increases in pregnancy, and three different complexes containing proMBP have been isolated from pregnancy plasma and serum: a 2:2 complex with the metalloproteinase, PAPP-A (pregnancy-associated plasma protein-A), a 2:2 complex with AGT (angiotensinogen) and a 2:2:2 complex with AGT and complement C3dg. In the present study we show that during human pregnancy, all of the circulating proMBP exists in covalent complexes, bound to either PAPP-A or AGT. We also show that the proMBP–AGT complex constitutes the major fraction of circulating HMW (high-molecular weight) AGT in late pregnancy, and that this complex is able to further associate with complement C3 derivatives post-sampling. Clearance experiments in mice suggest that complement C3-based complexes are removed faster from the circulation compared to

monomeric AGT and the proMBP–AGT complex. Furthermore, we have used recombinant proteins to analyse the formation of the proMBP–PAPP-A and the proMBP–AGT complexes, and we demonstrate that they are competing reactions, depending on the same cysteine residue of proMBP, but differentially on the redox potential, potentially important for the relative amounts of the complexes in vivo. These findings may be important physiologically, since the biochemical properties of the proteins change as a consequence of complex formation.

INTRODUCTION

(proform of eosinophil major basic protein) disulfide linked to AGT have been demonstrated [17,18]. One is a 2:2 proMBP– AGT complex of approximately 200 kDa. The other is a ternary complex with complement C3dg, a 2:2:2 proMBP–AGT– C3dg complex of approximately 300 kDa. However, whether these complexes with proMBP contribute quantitatively to the HMW AGT fraction has been questioned [11,16]. ProMBP is synthesized in large quantity by extravillous trophoblast cells in the placenta [19,20], and its concentration increases throughout pregnancy [17,21]. It is known to circulate in a 2:2 disulfide-linked complex with the metzincin metalloproteinase PAPP-A (pregnancy-associated plasma protein-A) [22]. Uncomplexed PAPP-A regulates the bioavailability of IGFs (insulin-like growth factors) by cleavage of IGFBPs (IGF-binding proteins) [23–25], but PAPP-A in complex with proMBP is proteolytically inactive [26,27]. Thus a network of complexes exists in which AGT, PAPPA and proMBP are key elements. Since these proteins are all present in the placenta, it is reasonable to speculate that formation of the proMBP–AGT and the proMBP–PAPP-A complexes are competing reactions. In this case, a pathological increase in proMBP–AGT may affect the relative amount of uncomplexed, and hence active, PAPP-A, and thereby indirectly the bioavailability of IGF. Further delineation of this network of protein complexes is required to understand its potential role in both normal and complicated pregnancies. We have therefore analysed the complexes present in the circulation of pregnant

AGT (angiotensinogen) is a non-inhibitory serpin of approximately 60 kDa [1,2], known as the precursor of angiotensin I. In addition to systemic effects of the renin– angiotensin system, local and tissue-specific effects, e.g. in the uteroplacental unit, have been described [3]. The plasma concentration of AGT in non-pregnant women is close to the K m value (approximately 1 μM) of its reaction with renin [4,5], but during pregnancy, it is increased up to four-fold [6]. In men and non-pregnant women, the majority of circulating AGT is monomeric. However, 3–5 % is present in a poorly characterized HMW (high-molecular weight) form. Curiously, in normotensive pregnant women, the HMW fraction is increased to approximately 16 %, and under pathological conditions, such as pregnancy-induced hypertension and pre-eclampsia, it may become the predominant form [7,8]. Importantly, renin cleavage of uncharacterized HMW AGT has been reported to progress slower than cleavage of monomeric AGT [9]. The molecular mass of HMW AGT is estimated to be 200– 500 kDa [7,8,10–13], but SDS/PAGE following reduction of disulfide bridges indicates that it is composed of AGT disulfide linked to other proteins, multimeric AGT or a mixture of both [9,14,15]. In vitro, the AGT molecule is known to be able to form disulfide-linked multimers, that are suggested to be the dominant constituent of the HMW fraction [16]. In addition, the presence of two different complexes containing proMBP

Key words: angiotensinogen (AGT), disulfide-linked complexes, pregnancy, pregnancy-associated plasma protein-A (PAPP-A), proform of eosinophil major basic protein (proMBP), renin substrate.

Abbreviations used: AGT, angiotensinogen; DTT, dithiothreitol; HEK, human embryonic kidney; HMW, high-molecular weight; IGF, insulin-like growth factor; IGFBP, IGF-binding protein; mAb, monoclonal antibody; MALDI–TOF-MS, matrix-assisted laser desorption ionization–time-of-flight MS; pAb, polyclonal antibody; PAPP-A, pregnancy-associated plasma protein-A; PBS-T, 0.01 % Tween-20 in PBS; proMBP, proform of eosinophil major basic protein; TST, 50 mM Tris, 500 mM sodium chloride and 0.1 % Tween 20; TST-SM, 2 % skimmed milk powder diluted in TST. 1 To whom correspondence should be addressed (email [email protected]). c The Authors Journal compilation c 2013 Biochemical Society

210

S. Kløverpris and others

women and established a model system based on recombinant proteins, and we have used this system to analyse the process of covalent complex formation. EXPERIMENTAL Plasma, serum and antibodies

Third trimester EDTA-plasma was obtained from seven normotensive pregnant women in the third trimester at Holbæk Hospital, Denmark. In addition, paired EDTA-plasma and serum samples were collected from three normotensive pregnant women in the third trimester. All samples were separated, immediately, placed on ice and frozen at − 20 ◦ C. Term pregnancy blood from pregnant women was collected at Skejby Hospital, Denmark, allowed to form serum at room temperature (22 ◦ C) overnight and stored at − 20 ◦ C. For some of the experiments, samples of third trimester plasma and serum were incubated for 24 h at room temperature in the absence or presence of 50 mM iodoacetamide. Blood samples were collected as part of routine prenatal screening, and used for research in accordance with the Danish Guidelines of the Prenatal Screening Registry. All women gave informed consent and the study was conducted in accordance with the Helsinki Declaration. The antibodies used were F8A2 [28], anti-AGT mouse mAb (monoclonal antibody); SSI 233 [18], anti-AGT chicken pAb (polyclonal antibody); αPP [29], anti-proMBP/PAPP-A rabbit pAb; 234-10 [30], antiproMBP mouse mAb; 234-5 [30], anti-PAPP-A mouse mAb; PM5A [31], anti-proMBP mouse mAb; A063 (Dako), anti-C3d rabbit pAb; 9E10 (ATTC), anti-c-Myc mouse mAb; P0260 (Dako), peroxidase-conjugated rabbit anti-mouse pAb; A9046 (Sigma– Aldrich), peroxidase-conjugated rabbit anti-chicken pAb; P0217 (Dako), peroxidase-conjugated swine anti-rabbit pAb; and A3151 (Sigma–Aldrich), peroxidase-conjugated avidin. Western blotting

AGT Western blotting was performed following separation of proteins by 10–20 % non-reducing or reducing SDS/PAGE. Proteins were blotted on to a PVDF membrane (Millipore) and blocked in TST-SM [2 % skimmed milk powder diluted in TST (50 mM Tris, 500 mM sodium chloride and 0.1 % Tween 20, pH 9.0)]. The membrane was washed in TST, incubated overnight at 4 ◦ C with the F8A2 antibody diluted to 1 μg/ml in TSTSM, washed and incubated for 0.5 h at room temperature with P0260 diluted 1:2000 in TST-SM. The blots were developed using ECL (enhanced chemiluminescence; Amersham Biosciences) and visualized using X-ray film. Detection of proMBP was performed similarly. The antibody 234-10 at 1 μg/ml was used as detecting antibody. DTT (dithiothreitol) was used as reducing agent. ELISA

Protein concentrations were measured by ELISAs performed in Maxisorp polystyrene microtiter plates (Nunc). Coating antibodies were diluted in 0.1 M sodium bicarbonate, pH 9.8, and wells were blocked with 2 % BSA in PBS (20 mM sodium hydrogen phosphate and 150 mM sodium chloride, pH 7.4). Detecting antibodies, samples and calibrators were diluted in PBS-T (0.01 % Tween-20 in PBS) supplemented with 1 % BSA. Washing was carried out with PBS-T. The antibodies used for each ELISA are listed as follows (coating/detecting/secondary antibody): AGT-specific, F8A2 (2 μg/ml)/SSI233 (1:1000 dilution)/A9046 (1:10000 dilution); proMBP specific, αPP c The Authors Journal compilation c 2013 Biochemical Society

(5 μg/ml)/234-10 (1 μg/ml)/P0260 (1:1000 dilution); PAPPA specific, αPP (5 μg/ml)/234-5 (1 μg/ml)/P0260 (1:1000 dilution); AGT–C3d specific, F8A2 (2 μg/ml)/A063 (1:1000 dilution)/P0217 (1:1000 dilution); proMBP/PAPP-A specific, 234-5 (2 μg/ml)/234-10-biotin (1 μg/ml)/A3151 (1:10000 dilution); and proMBP/AGT specific, F8A2 (2 μg/ml)/234-10-biotin (1 μg/ml)/A3151 (1:10000 dilution). In the complex-specific ELISAs, 0.8 M NaCl was added to the buffers to avoid detection of possible non-covalent complexes. Dilution series of third trimester pregnancy plasma was used to establish the standard curves for measurement in blood samples. One arbitrary unit is equivalent to the concentration of the protein in the non-diluted calibrator. For measurement of recombinant proteins, the assays were calibrated with the purified proMBP–PAPP-A complex [29], or immunoaffinity (F8A2) purified recombinant AGT, quantified by amino acid analysis. Size-exclusion chromatography

The plasma from seven normotensive pregnant women was thawed on ice, pooled and diluted 20 times in running buffer (20 mM Tris, 150 mM sodium chloride and 0.02 % Tween20, pH 7.4). To separate monomeric AGT from HMW AGT, size-exclusion chromatography was performed on a column (330 ml) packed with Sephacryl S-200 (Amersham Biosciences) and equilibrated in the running buffer (0.75 ml/min). Individual fractions (3 ml) were analysed by ELISA. Immunodepletion

For depletion of proMBP, chromatographic fractions were incubated overnight at 4 ◦ C with the PM-5A antibody (or 9E10 for the controls) immobilized to CNBr-activated Sepharose 4B (GE Healthcare). Similarly, depletion of PAPP-A and AGT in the fractions was done using a mixture of the mAbs 234-5 and F8A2. For depletion of complement C3 derivatives in serum, samples were incubated overnight at 4 ◦ C with protein–G Sepharose (GE Healthcare) in the absence or presence of A063, and analysed by proMBP Western blotting. Immunoaffinity chromatography

Term pregnancy serum was diluted 10 times in PBS and loaded on to a Sepharose 4B column (4 ml) with immobilized F8A2, or 9E10 as control and equilibrated in PBS. The column was washed in 40 ml of PBS-T followed by 40 ml of PBS and 1 M sodium chloride. The protein was eluted in 0.2 M glycine and 0.1 M sodium chloride (pH 2.5), and immediately neutralized with 1 M Tris (pH 7.4). The eluate was concentrated by ultrafiltration (filter cut-of 10 kDa; Millipore) and dialysed against PBS. Non-reduced and reduced samples were separated by SDS/PAGE and stained with Coomassie Blue for protein identification by MALDI–TOFMS (matrix-assisted laser desorption ionization–time-of-flight MS; see below). Protein identification by MALDI–TOF-MS

Bands were excised from Coomassie-Blue-stained reduced gels, washed in acetonitrile (50 % and 100 %) and dried. The bands from non-reduced gels were incubated with 10 mM DTT (45 min) followed by incubation with 50 mM iodoacetamide (30 min), washed in acetonitrile (50 % and 100 %) and dried. The material was then incubated on ice with 12.5 μg/μl sequence-

Disulfide-linked pregnancy-associated complexes

grade trypsin (Promega) in 10 mM ammonium bicarbonate for 1 h. Excess trypsin was removed, replaced with 50 mM ammonium bicarbonate and incubated overnight at 37 ◦ C. The supernatant was transferred into new tubes and the remaining tryptic peptides were extracted from the gel by the addition of 100 % acetonitrile. All the steps were carried out in siliconized tubes (Sorensen). The peptides were co-crystallized with CHCA (α-cyana-4-hydroxycinnamic acid), and mass spectra were obtained using a Voyager DE-PRO (Applied Biosystems) MALDI–TOF instrument. For protein identification, the peptide masses (m/z) were submitted to the Mascot search engine (http://www.matrixscience.com).

211

of solvent B from 20 % to 90 % over 15 min including a plateau of 5 min at 56 % solvent B. Quantitative analysis was performed by the incubation of different concentrations (0–12 μM) of the AGT monomer or the proMBP–AGT complex with renin (1 nM) in a total volume of 100 μl for 30 min at 37 ◦ C with shaking (800 rev./min). The amount of generated angiotensin I in these experiments did not exceed more than 10 % of the total amount of substrate. The reaction was stopped by the addition of 400 μl of a mixture of 75 % solvent A and 25 % solvent B. The samples were co-injected on the column with pure angiotensin II (Sigma) as an internal standard. A standard curve was prepared by injection of different concentrations of pure angiotensin I (Sigma). The Michaelis–Menten equation was fitted to the data using the Enzyme Kinetics module 1.1 of SigmaPlot 8.0.

Clearance analysis in mice

Female inbred mice (C57BL/6Jbom, Taconic), 7–8 weeks old, were used. The mice were housed in plastic cages under pathogenfree conditions with a 12 h light cycle and fed standard chow (1324; Altromin) and water ad libitum. The AGT monomer, the AGT–proMBP complex and the AGT–proMBP–C3dg complex were purified by AGT immunoaffinity chromatography and further by size-exclusion chromatography (Sephacryl S-300, Amersham). The purity of the complexes was assessed by SDS/PAGE (> 95 %), and the concentrations were determined by amino acid analysis. Samples were diluted in sterile PBS and 1 % BSA. Before tail vein injection (5 μg per injection), the animals were kept at a high ambient temperature to dilate the veins. A volume of 300 μl was slowly injected. Blood samples (20–30 μl) were collected at different time points after the injection (1– 240 min) by retro-orbital plexus puncture into vials (Microvette CB 300 KE). A total of five serum samples were obtained from individual animals. The time course of clearance was analysed in six mice for each protein by specific ELISAs. All proteins displayed a biphasic clearance curve and a two-compartment model, described by the eqn (1) C =C

−k1t 1e

+C

−k2t 2e

(1)

was fitted to the curves, where C is the percentage of injected protein remaining in circulation at a given time (t), k1 and k2 are the kinetic rate constants of the fast (α-phase) and slow phase (β-phase) respectively, and C1 and C2 are the percentages of administered sample removed during the two phases. The values of k1 , k2 , C1 and C2 were derived for each clearance curve by fitting C against t in eqn (1) using SigmaPlot 8.0. The half-lives (t1/2 ) corresponding to the α- and β-phase were calculated as α-phase t1/2 = 0.693/k1 and β-phase t1/2 = 0.693/k2 respectively. The animal experiments were approved by the Danish Animal Experiments Inspectorate.

Cell culture and transfection

HEK (human embryonic kidney)-293T cells (293tsA1609neo) were maintained in high glucose Dulbecco’s modified Eagle’s medium supplemented with 10 % fetal bovine serum, 2 mM glutamine, non-essential amino acids and gentamicin (Invitrogen). Cells were plated onto 6-cm culture dishes and transiently transfected 18 h later by calcium phosphate coprecipitation [32] using 10 μg of plasmid DNA. The cells were transfected with either PAPP-A cDNA [27], cDNA encoding proMBP, mutated variants of proMBP {proMBP(C51S) and proMBP(C169S) [33]} or AGT cDNA, and the culture medium was harvested after 48 h. AGT cDNA [34] was transferred from the original pECE vector to pcDNA3.1( + )/myc-His (Invitrogen) using XbaI and HindIII. The protein concentrations in the culture medium were measured by ELISA, as described above. In vitro complex formation

Formation of the recombinant proMBP–AGT complex was performed by incubating (37 ◦ C) culture medium from HEK-293T cells, transfected separately with cDNA encoding AGT, proMBP, or the mutated variants proMBP(C51S) and proMBP(C169S), to final concentrations of 40 nM of AGT and 400 nM proMBP. The amount of complex formed was measured over time relative to the level reached at the plateau using the proMBP–AGTspecific ELISA. In some experiments, GSH or hydrogen peroxide was included. In similar experiments, recombinant PAPP-A or proMBP were mixed to final subunit concentrations of 20 nM and incubated at 37 ◦ C to allow the covalent complex to form in the presence of 0–240 nM recombinant AGT. The experiment was carried out in the presence of 30–180 μM GSH. RESULTS

Kinetic analysis

Cleavage of the purified AGT monomer and proMBP–AGT complex (see above) was analysed using recombinant human renin (R2779; Sigma). Initial cleavage analysis was performed by the incubation of 1 μM of the AGT monomer or the AGT–proMBP complex with 5 nM renin in PBS containing 0.1 % BSA for 1 h at 37 ◦ C with shaking (800 rev./min). The generated angiotensin I was separated by reversed-phase highpressure liquid chromatography (50 ◦ C, 0.5 ml/min, 218 nm) on a 4mm×250 mm column packed with Nucleosil C18 100-5 (Macherey–Nagel) using a gradient formed from 0.1 % (v/v) trifluoroacetic acid (solvent A) and 0.075 % (v/v) trifluoroacetic acid in 90 % (v/v) acetonitrile (solvent B), increasing the amount

ProMBP is a major constituent of the HMW fraction of AGT in late pregnancy

A pool of seven plasma samples from normotensive women in the third trimester was subjected to size-exclusion chromatography, and the individual fractions were analysed using an AGT-specific ELISA (Figure 1A). AGT antigen eluted as a major peak corresponding to monomeric AGT (fractions 37–46) and a smaller broad peak of HMW AGT (fractions 29–38). The total amount of AGT in the HMW fractions relative to monomeric AGT is in fair agreement with previous studies [7,8]. To determine the proportion of HMW AGT composed of proMBP complexes, fractions 30–35 were pooled and immunodepleted for proMBP. This caused more than a 50 % c The Authors Journal compilation c 2013 Biochemical Society

212


Figure 1 ProMBP is a major constituent of the HMW fraction of AGT in late pregnancy (A) Separation of AGT species in a pool of seven third trimester plasma samples by size-exclusion chromatography (Sephacryl S-200). Elution profile of total protein (broken line, absorption at 280 nm) and AGT antigen (solid line) are shown. The latter was determined by an AGT specific ELISA. (B) Pools of fractions 30–35 (HMW AGT) and fractions 42–45 (monomeric AGT) analysed by an AGT-specific ELISA before (grey bars) and after (white bars) immunodepletion of proMBP. Results are means + − S.D. from three independent experiments. (C) Western blotting using an AGT-specific mAb following reducing SDS/PAGE of the pool of fractions 30–35 before (lane 1) and after (lane 2) immunodepletion of proMBP. Molecular masses are shown on the left-hand side in kDa.

reduction of AGT antigen as shown by ELISA (Figure 1B) and Western blotting following reducing SDS/PAGE (Figure 1C) of the pooled fractions before and after depletion. Thus proMBP is a major constituent of HMW AGT in late pregnancy plasma

ProMBP is disulfide linked to either PAPP-A or AGT in late pregnancy

Measurement of proMBP showed an elution profile of two overlapping peaks (Figure 2A, broken line). No monomeric proMBP, expected to elute after fraction 45, was detected showing that all proMBP is present in HMW complexes. ProMBP Western blotting of the unfractionated material (Figure 2A, insert) showed bands corresponding to the 2:2 proMBP–PAPPA complex (480 kDa) [22], and the 2:2 proMBP–AGT complex (200 kDa) [17]. Upon immunodepletion of both AGT and PAPP-A in individual fractions, proMBP could not be detected (Figure 2A, solid line). Measurement of PAPP-A (Figure 2B, broken line) shows the presence of the proMBP–PAPP-A complex corresponding to the position of the left-hand peak of Figure 2(A), and measurement of the proMBP–AGT complex (Figure 2B, solid line) shows the presence of this complex in the broader right-hand peak. Importantly, these experiments did not reveal the presence of the 300 kDa 2:2:2 proMBP–AGT–C3dg complex previously detected in pregnancy plasma and serum [17,18]. c The Authors Journal compilation c 2013 Biochemical Society

Figure 2

ProMBP is disulfide linked to PAPP-A or AGT in late pregnancy

(A) Individual chromatographic fractions (from Figure 1A) were analysed by proMBP-specific ELISA before (broken line) and after (solid line) combined immunodepletion of PAPP-A and AGT. Unfractionated material was analysed by proMBP Western blotting following non-reducing 10–20 % SDS/PAGE (insert; molecular masses are shown on the left-hand side in kDa). (B) Chromatographic fractions were analysed by ELISA for their content of PAPP-A (broken line) and proMBP/AGT (solid line). The presence of proMBP–AGT complex in individual samples, from which the pool of Figure 1(A) was made, was confirmed (72, 127, 44, 95, 164, 101 and 272 arbitrary units).

Complement C3 containing complexes form as a consequence of post-sampling events

The 2:2:2 proMBP–AGT–C3dg complex might form as a consequence of post-sampling events, explaining its absence in freshly obtained plasma samples. Therefore plasma samples were incubated for 24 h at room temperature and analysed by proMBP Western blotting following non-reducing SDS/PAGE. Dramatic changes in the molecular mass of the proMBP species were observed (Figure 3A, lanes 1 and 2). A similar result was obtained with freshly obtained pregnancy serum (Figure 3A, lanes 3 and 4). However, the observed shift in molecular mass was inhibited by the presence of an alkylating reagent (Figure 3B). Therefore incubation, not processes of blood coagulation, resulted in these changes towards higher molecular mass, possibly involving complement C3dg, which occur as a consequence of disulfide bond formation. To allow analysis of putative complement C3dg containing complexes, serum formed at room temperature from term pregnancy blood was analysed by proMBP Western blotting following non-reducing SDS/PAGE. In addition to the bands corresponding to the proMBP–PAPP-A complex and the proMBP–AGT complex, several additional proMBP-reactive bands were observed (Figure 3C, lane 1). However, following immunodepletion of complement C3d, only bands corresponding to the proMBP–AGT and the proMBP–PAPP-A complex were present, suggesting that the additional complexes all involve derivatives of complement C3 (Figure 3C, lane 2). Non-reducing 7 % SDS/PAGE of affinity purified AGT followed by Coomassie Blue staining revealed five HMW bands (bands 1–5) in addition to monomeric AGT (band 6) (Figure 3D), and MS following in-gel digestion demonstrated the presence of


Figure 3

213

Complement-C3-containing complexes form as a consequence of post-sampling events

(A) ProMBP Western blotting following non-reducing 10–20 % SDS/PAGE of paired third trimester samples of plasma (lanes 1 and 2) and serum (lanes 3 and 4) incubated at room temperature for 0 h (lanes 1 and 3) or for 24 h (lanes 2 and 4). (B) ProMBP Western blotting following non-reducing 10–20 % SDS/PAGE of third trimester serum incubated for 0 h (lane 1) or 24 h at room temperature in the absence (lane 2) or presence (lane 3) of alkylating agent [IAA (iodoaceteamide)]. (C) ProMBP Western blotting following non-reducing 10–20 % SDS/PAGE of term pregnancy serum before (lane 1) and after (lane 2) immunodepletion of C3d. (D) Coomassie-Blue-stained non-reducing 7 % SDS/PAGE of AGT purified from term pregnancy serum by immunoaffinity chromatography. (E) Coomassie-Blue-stained reducing 10–20 % SDS/PAGE of the same material as in (D) and of the eluate from a control column (right-hand gel). The proteins corresponding to numbered bands [1–6 of (D) and 7–16 of (E)] were identified by MALDI–TOF-MS (Supplementary Table S1 at http://www.biochemj.org/bj/449/bj4490209add.htm). In all of the gels molecular masses are shown on the left-hand side in kDa.

proMBP in all of the HMW bands (Figure 3D and Supplementary Table S1 at http://www.biochemj.org/bj/449/bj4490209add.htm). Furthermore, peptides derived from complement C3b were identified in bands 1 and 2, whereas peptides of complement C3dg were identified in bands 3 and 4. Several other proteins were identified as constituents of the HMW complexes (Supplementary Table S1). All of these were also present following reduction (Figure 3E and Supplementary Table S1). It is difficult to assess to what extent complement-C3containing complexes, possibly as a result of complement activation, form in vivo. A faster rate of clearance from the circulation may explain their absence from freshly drawn blood (Figure 3A). To address this question, AGT, proMBP–AGT and proMBP–AGT–C3dg were purified chromatographically (results not shown). Clearance experiments were conducted in mice (Figure 4A), revealing similar kinetics for the clearance of monomeric AGT and the proMBP–AGT complex, but significantly altered kinetics for the proMBP–AGT–C3dg complex (Figure 4A and Table 1). These results strongly indicate that complement C3dg are responsible for the faster clearance rate of complexes in which this molecule is a constituent, and may explain in part why these complexes are not detected in freshly drawn blood samples. Finally, using the purified components, kinetic parameters of renin cleavage were assessed (Figure 4B). The experiment shows an approximately three-fold increase in the K m value and a twofold decrease in the V max value of the cleavage reaction with AGT in complex with proMBP relative to the reaction with monomeric AGT, resulting in a five-fold decrease in catalytic efficiency.

Formation of the proMBP–PAPP-A and proMBP–AGT complexes both depend on proMBP Cys169

Since all of proMBP is contained in covalent complexes with either AGT or PAPP-A, the balance between the two reactions, leading to proMBP–AGT and proMBP–PAPP-A, is important to understand. The mechanism behind the proMBP–PAPP-A complex is known to depend on the presence of proMBP Cys169 [26,33], and micromolar concentrations of GSH accelerates the reaction [33]. To analyse the formation of the proMBP– AGT complex, recombinant AGT and proMBP were mixed and incubated at 37 ◦ C. The rate of complex formation was dramatically increased when the redox potential was shifted by the presence of 100 μM GSH, whereas, in the presence of 100 μM hydrogen peroxide, the reaction was inhibited (Figure 5A). As no ternary proMBP–PAPP-A–AGT complex has been identified, we speculated that proMBP Cys169 is also required for the formation of the proMBP–AGT complex. Recombinant AGT was incubated with recombinant wild-type proMBP or proMBP(C169S) at 37 ◦ C. Interestingly, no complex was formed between AGT and proMBP(C169S) (Figure 5B), suggesting that reactions of proMBP and AGT depend on disulfide bond formation with proMBP Cys169 . Together, these findings indicate a similar mechanism of complex formation between proMBP and PAPP-A and proMBP and AGT. In uncomplexed recombinant proMBP, Cys169 forms a disulfide bond with Cys51 [33]. We therefore carried out an additional experiment using a proMBP mutant proMBP(C51S) in which this disulfide bond is absent, but Cys169 is still present. We found that c The Authors Journal compilation c 2013 Biochemical Society

214


Figure 5 Formation of the proMBP–AGT complex is sensitive to redox potential and depends on proMBP Cys169

Figure 4

Functional analyses of the AGT complexes

(A) Clearance from the circulation of AGT, proMBP–AGT and proMBP–AGT–C3dg was monitored following intravenous injections in mice. Six mice were used for each protein. Serum concentrations of the individual components were measured at defined time points using ELISAs and expressed relative to the concentration at 0 h. In all experiments, a two-compartment model could be fitted to the biphasic curves (AGT, R 2 = 0.9905; proMBP–AGT, R 2 = 0.9949; and proMBP–AGT–C3dg, R 2 = 0.9961). Kinetic parameters for the clearance of the complexes are shown in Table 1. (B) Kinetic analysis of renin cleavage of AGT and proMBP–AGT. Cleavage analysis was carried out by incubating purified AGT or proMBP–AGT with renin (5 nM). Initial cleavage rates are plotted as a function of substrate concentrations. Each data point is the average of two independent experiments. Kinetic parameters (K m and V max ) are based on fitting of the Michaelis–Menten equation to the data. S.D. are shown in parentheses.

the initial rate of AGT complex formation with proMBP(C51S) was increased several fold compared with the reaction with the wild-type proMBP (Figure 5B). Formation of the proMBP–PAPP-A and proMBP–AGT complexes are competing reactions which depend differentially on redox potential

To mimic the conditions of complex formation in vivo, experiments were carried out with all three components. First, recombinant proMBP and PAPP-A were mixed in a 1:1 molar ratio and formation of the covalent proMBP–PAPP-A complex at 37 ◦ C was monitored over time. After approximately 2 h, the concentration of the proMBP–PAPP-A complex had Table 1

(A) Culture supernatants from cells transfected separately with proMBP and AGT cDNA were mixed and incubated at 37 ◦ C in the presence or absence of 100 μM GSH, or 100 μM hydrogen peroxide. The concentrations of AGT and proMBP were 40 nM and 400 nM respectively. Samples were analysed by the proMBP–AGT-specific ELISA at the defined time points. (B) In a similar experiment, culture supernatants from cells transfected with AGT cDNA were mixed with culture supernatants from cells transfected with cDNA encoding wild-type (wt) proMBP, proMBP(C51S) or proMBP(C169S). Previous data show that proMBP(C51S) has increased rate of complex formation with PAPP-A, whereas proMBP(C169S) cannot form a complex with PAPP-A [33]. Results are means + − S.D. for three independent experiments. The concentrations are expressed relative to the amount of complex formed at the plateau of the experiment in (B) (100 %).

reached a maximum (Figure 6A, solid line). Secondly, a similar experiment was carried out in the presence of an eight-fold molar excess of AGT. In this experiment, the maximum reached was approximately 20 % lower (Fig. 6A, broken line), showing that AGT is able to compete for proMBP, but that complex formation with PAPP-A is a faster reaction under these conditions. Thirdly, monitoring the amount of the proMBP–PAPP-A complex formed after 8 h in the presence of increasing concentrations of AGT showed, as expected, a decrease in the amount of the proMBP– PAPP-A complex formed (Figure 6B). Finally, we asked whether the two reactions of proMBP were affected similarly by changes in the redox potential, as the differential sensitivity of the two reactions might determine the balance between the two. An experiment with all three components, similar to the experiment of Figure 6(A), was carried out in the presence of different concentrations of GSH. Interestingly, we observed that with decreasing concentration of GSH, the ability of AGT to compete for proMBP is increased (Figure 6C). Increasing the concentration of GSH to more than 100 μM did not appear

Kinetic parameters characterizing the clearance in mice of the AGT monomer, the AGT–proMBP complex and the AGT–proMBP–C3dg complex

The percentages of the administered sample cleared in the fast phase (C 1 ) and the slow phase (C 2 ), and the values of the kinetic rate constants (k 1 and k 2 ) of each phase were determined by fitting the clearance data to eqn (1) (see the Experimental section). The half lives for the fast phase (α-phase t 1/2 ) and the slow phase (β-phase t 1/2 ) were defined as α-phase t 1/2 = 0.693/k 1 and β-phase t 1/2 = 0.693/k 2 respectively. Standard errors are shown in parentheses. Protein

C 1 (%)

k 1 (min − 1 )

α-Phase t 1/2 (min)

C 2 (%)

k 2 (min − 1 )

β-Phase t 1/2 (min)

AGT AGT–proMBP AGT–proMBP–C3dg

51.7 (17.7) 42.9 (11.8) 46.7 (6.61)

0.0206 (0.00167) 0.0272 (0.00108) 0.0583 (0.0118)

33.6 25.3 11.9

48.3 (18.3) 57.1 (12.3) 53.3 (7.06)

0.00236 (0.00757) 0.00228 (0.0101) 0.00880 (0.00106)

294 303 78.8

c The Authors Journal compilation c 2013 Biochemical Society


Figure 6 Formation of the proMBP–PAPP-A and proMBP–AGT complexes are competing reactions depending on the redox potential (A) Culture supernatants containing recombinant proMBP and PAPP-A were mixed in the absence (molar ratio 1:1:0) or presence (molar ratio 1:1:8) of AGT. The molar concentrations of proMBP and PAPP-A were 20 nM in all experiments. Reactions were carried out at 37 ◦ C in the presence of 100 μM GSH. ProMBP–PAPP-A was assessed by ELISA and 100 % proMBP–PAPP-A complex corresponds to full PAPP-A complex formation. (B) A similar experiment was carried out varying the molar concentration of AGT, as indicated. Samples were analysed following 8 h of incubation. Concentrations are expressed relative to the amount of complex formed in the absence of AGT, corresponding to the plateau of the experiment in (A). (C) A similar experiment, in which the concentration of GSH was varied as indicated. The relative concentration of the proMBP–PAPP-A complex formed in the presence of AGT (PAPP-A/proMBP/AGT molar ratio = 1:1:12) is plotted as a function of GSH concentration. Results are means + − S.D. for three independent experiments.

to have an effect, but the balance was dramatically affected at concentrations below 100 μM.

DISCUSSION

Some proteins, quantitatively absent in non-pregnant women, are abundantly synthesized in the placenta and secreted into the maternal circulation. The concentration of other plasma proteins may change as a result of up-regulation during pregnancy, e.g. increased hepatic synthesis of AGT. In addition to changes in synthesis, the principal proteins of the present study, AGT and placentally derived PAPP-A and proMBP, also engage in the formation of pregnancy-associated protein complexes of increased molecular mass mediated by disulfide bonds. The

215

reactions that lead to these covalent complexes form a network of interactions (Figure 7). As the biochemical properties of individual components, e.g. proteolytic activity of PAPP-A and substrate function of AGT, change as a consequence of complex formation, knowledge about these complexes, their formation and the dynamics of the protein network is important. How the network is regulated in vivo is not known, but differences in the balances between the covalent complexes have been recognized. For example, the fraction of AGT present as HMW complexes (approximately 16 % in normal pregnancy [7,8]) has been found to vary and to be associated with the development of pregnancy-induced hypertension and pre-eclampsia [7,35,36], the latter associated with hypoxia and therefore redox changes [37]. Also, the fraction of PAPP-A present as active uncomplexed PAPP-A is not constant, but varies from about one third in the first trimester to less than 1 % at term [38], causing a many-fold difference in specific PAPP-A activity. We demonstrated in the present study that all of proMBP present in the circulation of pregnant women is complex bound either to PAPP-A or AGT (Figure 2), and we show for the first time that the proMBP–AGT complex is a major constituent of HMW AGT (Figure 1). This conclusion is in agreement with the approximate molar concentrations of AGT (up to 4 μM), PAPP-A (0.1 μM) and proMBP (0.4 μM) [6,17] in the circulation at term. A previously recognized complex, proMBP–AGT–C3dg [17], was not detected in freshly drawn blood. However, we found that storage of serum or incubation of plasma at room temperature for 24 h caused this complement-C3-based complex to form (Figure 3). Curiously, in addition to the quantitatively dominating proMBP–AGT and proMBP–AGT–C3dg complexes, several other proteins, including complement C3b, appeared to be present in HMW complexes (Figure 3, and Supplementary Table S1). On the basis of the present data, it is not possible to further characterize these, but it is reasonable to speculate that the 2:2 proMBP–AGT complex is a part of all complexes, and as a result of complement activation via the alternative pathway [39], this complex reacts with complement C3 derivatives. Further reactions are likely to be mediated by reaction with C3 derivatives of the complexes, but the stoichiometry cannot be predicted based on the current data. The current experiment, in which formed complexes are allowed to accumulate during incubation in vitro, does not reflect the steady state in the circulation where complexes may be cleared differently. We demonstrated that the clearance kinetics of AGT in mice was not affected by formation of the proMBP–AGT complex, but that further complex formation with complement C3dg caused it to clear significantly faster (Figure 4 and Table 1). We therefore speculate that the complementC3-based complexes may form in humans under conditions of increased complement activation, but that they are more likely to be cleared at a faster rate, probably via complement receptors [40]. To address the question of the balance between the two different reactions of proMBP (Figure 7), we used recombinant proteins. First, we found that formation of the recombinant proMBP–AGT complex is accelerated or inhibited by micromolar concentrations of GSH or oxidizing agent respectively (Figure 5A). Secondly, we showed that proMBP Cys169 is required for covalent complex formation and that a mutated variant of proMBP, in which this cysteine residue is in the reduced (-SH) form rather than engaged in the Cys51 –Cys169 disulfide bond of the wild-type proMBP, forms the complex with AGT at a markedly increased rate (Figure 5B). Similar results have been obtained for the formation of the proMBP–PAPP-A complex [33], indicating that the proMBP– AGT and the proMBP–PAPP-A complexes are formed by a similar mechanism. c The Authors Journal compilation c 2013 Biochemical Society

216

Figure 7


Complexes of proMBP present in pregnancy

ProMBP binds covalently to PAPP-A, resulting in the 2:2 proMBP–PAPP-A complex. In addition, proMBP forms a covalent 2:2 complex with AGT (proMBP–AGT), which can further bind derivatives of complement C3. In addition to the two latter complexes, HMW AGT (within the grey circle) is most likely to be composed of polymerized AGT monomer. See the Results and Discussion sections for details.

Of particular interest, these findings suggest that AGT and PAPP-A must compete for covalent interaction with proMBP (Figure 7), explaining the apparent absence from the circulation of a ternary complex, composed of proMBP, AGT and PAPP-A. We demonstrated that such competition occurs by using a model system including PAPP-A, proMBP and a molar excess of AGT. In this system, AGT caused a 20 % reduction in the proMBP–PAPPA complex (Figure 6A), which therefore formed more readily. We further demonstrated that with decreasing concentrations of GSH, less of the proMBP–PAPP-A complex is formed (Figure 6C). However, it is difficult to define in detail how such a mechanism operates in vivo, because both of the reactions of proMBP are likely to be affected by factors other than the redox potential, and they may occur outside the vascular compartment, e.g. in the placental tissue. For example, PAPP-A [41] and proMBP [42] are known to bind to cellular surfaces, and the presence of an AGT receptor on placental cells has been proposed [43]. No data exists for AGT, but surface binding of PAPP-A and proMBP is known to affect complex formation, possibly by means of a proximity effect [31]. Furthermore, the presence on cell surfaces of protein disulfide isomerases [44] may control the local redox environment at the cell surface and thus be a major determinant of the reaction partner of proMBP. Potential biological consequences of a disturbed balance are obvious, in particular the consequences of reduced formation of the proMBP–PAPP-A complex. This would cause increased uncomplexed active PAPP-A and, in turn, increased potential for stimulation of the IGF receptor mediated by the increased cleavage of inhibitory binding proteins by PAPP-A. Several processes of placental development and fetal growth depend on IGF receptor stimulation [45]. The consequence of an increased concentration of proMBP–AGT relates to the kinetics of its cleavage by renin, dramatically reduced compared with uncomplexed AGT (Figure 4B). A reduction in the rate of angiotensin I generation potentially affects systemic blood pressure, but may also have local effects in the placenta [3]. Whether changes in the degree of formation of these complexes have diagnostic potentials or are connected with disease development awaits further studies.

AUTHOR CONTRIBUTION The experimental work was performed by Søren Kløverpris, Louise Skov and Simon Glerup. Blood sampling was done by Kasper Pihl and Michael Christiansen. Data interpretation and paper preparation was done by Søren Kløverpris and Claus Oxvig. c The Authors Journal compilation c 2013 Biochemical Society

ACKNOWLEDGEMENTS We thank Frederik Dagnaes-Hansen for help with clearance studies and Professor Xavier Jeunemaitre (Centre de recherche Cardiovasculaire a` l’HEGP, Paris, France) for AGT cDNA.

FUNDING This work was supported by the Lundbeck Foundation and the Danish Medical Research Council.

REFERENCES 1 Doolittle, R. F. (1983) Angiotensinogen is related to the antitrypsin-antithrombin-ovalbumin family. Science 222, 417–419 2 Zhou, A., Carrell, R. W., Murphy, M. P., Wei, Z., Yan, Y., Stanley, P. L., Stein, P. E., Broughton Pipkin, F. and Read, R. J. (2010) A redox switch in angiotensinogen modulates angiotensin release. Nature 468, 108–11 3 Irani, R. A. and Xia, Y. (2008) The functional role of the renin–angiotensin system in pregnancy and preeclampsia. Placenta 29, 763–771 4 Gould, A. B. and Green, D. (1971) Kinetics of the human renin and human substrate reaction. Cardiovasc Res. 5, 86–89 5 Slater, E. E. and Strout, Jr, H. V. (1981) Pure human renin. Identification and characterization and of two major molecular weight forms. J. Biol. Chem. 256, 8164–8171 6 Weir, R. J., Brown, J. J., Fraser, R., Kraszewski, A., Lever, A. F., McIlwaine, G. M., Morton, J. J., Robertson, J. I. and Tree, M. (1973) Plasma renin, renin substrate, angiotensin II, and aldosterone in hypertensive disease of pregnancy. Lancet 1, 291–294 7 Gordon, D. B. and Sachin, I. N. (1977) Chromatographic separation of multiple renin substrates in women: effect of pregnancy and oral contraceptives. Proc. Soc. Exp. Biol. Med. 156, 461–464 8 Tewksbury, D. A. and Dart, R. A. (1982) High molecular weight angiotensinogen levels in hypertensive pregnant women. Hypertension 4, 729–734 9 Sato, Y., Hiwada, K. and Kokubu, T. (1985) Physicochemical characteristics of human high molecular weight angiotensinogen. Life Sci. 37, 371–377 10 Tetlow, H. J. and Broughton Pipkin, F. (1986) Changing renin substrates in human pregnancy. J. Endocrinol. 109, 257–262 11 Tewksbury, D. A. (1996) Quantitation of five forms of high molecular weight angiotensinogen from human placenta. Am. J. Hypertens. 9, 1029–1034 12 Tewksbury, D. A., Goodman, J. R., Kaiser, S. J., Burrill, R. E. and Brown, H. L. (2000) Quantitation of the five forms of plasma high molecular weight angiotensinogen in women with pregnancy-induced hypertension. Am. J. Hypertens. 13, 221–225 13 Tewksbury, D. A., Tryon, E. S., Burrill, R. E. and Dart, R. A. (1986) High molecular weight angiotensinogen: a pregnancy associated protein. Clin. Chim. Acta 158, 7–12 14 Campbell, D. J., Bouhnik, J., Coezy, E., Menard, J. and Corvol, P. (1985) Characterization of precursor and secreted forms of human angiotensinogen. J. Clin. Invest. 75, 1880–1893 15 Tewksbury, D. A. and Tryon, E. S. (1989) Immunochemical comparison of high molecular weight angiotensinogen from amniotic fluid, plasma of men, and plasma of pregnant women. Am. J. Hypertens. 2, 411–413

Disulfide-linked pregnancy-associated complexes 16 Stanley, P., Serpell, L. C. and Stein, P. E. (2006) Polymerization of human angiotensinogen: insights into its structural mechanism and functional significance. Biochem. J. 400, 169–178 17 Oxvig, C., Haaning, J., Kristensen, L., Wagner, J. M., Rubin, I., Stigbrand, T., Gleich, G. J. and Sottrup-Jensen, L. (1995) Identification of angiotensinogen and complement C3dg as novel proteins binding the proform of eosinophil major basic protein in human pregnancy serum and plasma. J. Biol. Chem. 270, 13645–13651 18 Christiansen, M., Jaliashvili, I., Overgaard, M. T., Ensinger, C., Obrist, P. and Oxvig, C. (2000) Quantification and characterization of pregnancy-associated complexes of angiotensinogen and the proform of eosinophil major basic protein in serum and amniotic fluid. Clin. Chem. 46, 1099–1105 19 Bonno, M., Oxvig, C., Kephart, G. M., Wagner, J. M., Kristensen, T., Sottrup-Jensen, L. and Gleich, G. J. (1994) Localization of pregnancy-associated plasma protein-A and colocalization of pregnancy-associated plasma protein-A messenger ribonucleic acid and eosinophil granule major basic protein messenger ribonucleic acid in placenta. Lab. Invest. 71, 560–566 20 Overgaard, M. T., Oxvig, C., Christiansen, M., Lawrence, J. B., Conover, C. A., Gleich, G. J., Sottrup-Jensen, L. and Haaning, J. (1999) Messenger ribonucleic acid levels of pregnancy-associated plasma protein-A and the proform of eosinophil major basic protein: expression in human reproductive and nonreproductive tissues. Biol. Reprod. 61, 1083–1089 21 Wasmoen, T. L., Coulam, C. B., Leiferman, K. M. and Gleich, G. J. (1987) Increases of plasma eosinophil major basic protein levels late in pregnancy predict onset of labor. Proc. Natl. Acad. Sci. U.S.A. 84, 3029–3032 22 Oxvig, C., Sand, O., Kristensen, T., Gleich, G. J. and Sottrup-Jensen, L. (1993) Circulating human pregnancy-associated plasma protein-A is disulfide-bridged to the proform of eosinophil major basic protein. J. Biol. Chem. 268, 12243–12246 23 Lawrence, J. B., Oxvig, C., Overgaard, M. T., Sottrup-Jensen, L., Gleich, G. J., Hays, L. G., Yates, III, J. R. and Conover, C. A. (1999) The insulin-like growth factor (IGF)-dependent IGF binding protein-4 protease secreted by human fibroblasts is pregnancy-associated plasma protein-A. Proc. Natl. Acad. Sci. U.S.A. 96, 3149–3153 24 Laursen, L. S., Overgaard, M. T., Soe, R., Boldt, H. B., Sottrup-Jensen, L., Giudice, L. C., Conover, C. A. and Oxvig, C. (2001) Pregnancy-associated plasma protein-A (PAPP-A) cleaves insulin-like growth factor binding protein (IGFBP)-5 independent of IGF: implications for the mechanism of IGFBP-4 proteolysis by PAPP-A. FEBS Lett. 504, 36–40 25 Laursen, L. S., Kjaer-Sorensen, K., Andersen, M. H. and Oxvig, C. (2007) Regulation of insulin-like growth factor (IGF) bioactivity by sequential proteolytic cleavage of IGF binding protein-4 and -5 Mol. Endocrinol. 21, 1246–1257 26 Overgaard, M. T., Glerup, S., Boldt, H. B., Rodacker, V., Olsen, I. M., Christiansen, M., Sottrup-Jensen, L., Giudice, L. C. and Oxvig, C. (2004) Inhibition of proteolysis by the proform of eosinophil major basic protein (proMBP) requires covalent binding to its target proteinase. FEBS Lett. 560, 147–152 27 Overgaard, M. T., Haaning, J., Boldt, H. B., Olsen, I. M., Laursen, L. S., Christiansen, M., Gleich, G. J., Sottrup-Jensen, L., Conover, C. A. and Oxvig, C. (2000) Expression of recombinant human pregnancy-associated plasma protein-A and identification of the proform of eosinophil major basic protein as its physiological inhibitor. J. Biol. Chem. 275, 31128–31133 28 Rubin, I., Lykkegaard, S., Olsen, A. A., Selmer, J. and Ballegaard, M. (1988) Monoclonal antibodies against human angiotensinogen, their characterization and use in an angiotensinogen enzyme linked immunosorbent assay. J. Immunoassay. 9, 257–274

217

29 Oxvig, C., Sand, O., Kristensen, T., Kristensen, L. and Sottrup-Jensen, L. (1994) Isolation and characterization of circulating complex between human pregnancy-associated plasma protein-A and proform of eosinophil major basic protein. Biochim. Biophys. Acta 1201, 415–423 30 Qin, Q. P., Christiansen, M., Oxvig, C., Pettersson, K., Sottrup-Jensen, L., Koch, C. and Norgaard-Pedersen, B. (1997) Double-monoclonal immunofluorometric assays for pregnancy-associated plasma protein A/proeosinophil major basic protein (PAPP-A/proMBP) complex in first-trimester maternal serum screening for Down syndrome. Clin. Chem. 43, 2323–2332 31 Glerup, S., Kloverpris, S., Laursen, L. S., Dagnaes-Hansen, F., Thiel, S., Conover, C. A. and Oxvig, C. (2007) Cell surface detachment of pregnancy-associated plasma protein-A requires the formation of intermolecular proteinase-inhibitor disulfide bonds and glycosaminoglycan covalently bound to the inhibitor. J. Biol. Chem. 282, 1769–1778 32 Pear, W. S., Nolan, G. P., Scott, M. L. and Baltimore, D. (1993) Production of high-titer helper-free retroviruses by transient transfection. Proc. Natl. Acad. Sci. U.S.A. 90, 8392–8396 33 Glerup, S., Boldt, H. B., Overgaard, M. T., Sottrup-Jensen, L., Giudice, L. C. and Oxvig, C. (2005) Proteinase inhibition by proform of eosinophil major basic protein (pro-MBP) is a multistep process of intra- and intermolecular disulfide rearrangements. J. Biol. Chem. 280, 9823–9832 34 Gimenez-Roqueplo, A. P., Leconte, I., Cohen, P., Simon, D., Guyene, T. T., Celerier, J., Pau, B., Corvol, P., Clauser, E. and Jeunemaitre, X. (1996) The natural mutation Y248C of human angiotensinogen leads to abnormal glycosylation and altered immunological recognition of the protein. J Biol Chem. 271, 9838–9844 35 Tewksbury, D. A., Kaiser, S. J. and Burrill, R. E. (2001) A study of the temporal relationship between plasma high molecular weight angiotensinogen and the development of pregnancy-induced hypertension. Am. J. Hypertens. 14, 794–797 36 Eggena, P., Hidaka, H., Barrett, J. D. and Sambhi, M. P. (1978) Multiple forms of human plasma renin substrate. J. Clin. Invest. 62, 367–372 37 Young, B. C., Levine, R. J. and Karumanchi, S. A. (2010) Pathogenesis of preeclampsia. Annu. Rev. Pathol. 5, 173–192 38 Gyrup, C., Christiansen, M. and Oxvig, C. (2007) Quantification of proteolytically active pregnancy-associated plasma protein-A with an assay based on quenched fluorescence. Clin. Chem. 53, 947–954 39 Sahu, A. and Lambris, J. D. (2001) Structure and biology of complement protein C3, a connecting link between innate and acquired immunity. Immunol. Rev. 180, 35–48 40 van Lookeren Campagne, M., Wiesmann, C. and Brown, E. J. (2007) Macrophage complement receptors and pathogen clearance. Cell Microbiol. 9, 2095–2102 41 Laursen, L. S., Overgaard, M. T., Weyer, K., Boldt, H. B., Ebbesen, P., Christiansen, M., Sottrup-Jensen, L., Giudice, L. C. and Oxvig, C. (2002) Cell surface targeting of pregnancy-associated plasma protein A proteolytic activity. Reversible adhesion is mediated by two neighboring short consensus repeats. J. Biol. Chem. 277, 47225–47234 42 Glerup, S., Kloverpris, S. and Oxvig, C. (2006) The proform of the eosinophil major basic protein binds the cell surface through a site distinct from its C-type lectin ligand-binding region. J. Biol. Chem. 281, 31509–31516 43 Tewksbury, D. A., Pan, N. and Kaiser, S. J. (2003) Detection of a receptor for angiotensinogen on placental cells. Am. J. Hypertens. 16, 59–62 44 Hogg, P. J. (2003) Disulfide bonds as switches for protein function. Trends Biochem. Sci. 28, 210–214 45 Forbes, K. and Westwood, M. (2008) The IGF axis and placental function. a mini review. Horm. Res. 69, 129–137

Received 14 May 2012/21 August 2012; accepted 4 October 2012 Published as BJ Immediate Publication 4 October 2012, doi:10.1042/BJ20120801

c The Authors Journal compilation c 2013 Biochemical Society

Biochem. J. (2013) 449, 209–217 (Printed in Great Britain)

doi:10.1042/BJ20120801

SUPPLEMENTARY ONLINE DATA

Formation of high-molecular-weight angiotensinogen during pregnancy is a result of competing redox reactions with the proform of eosinophil major basic protein Søren KLØVERPRIS*, Louise L. SKOV*, Simon GLERUP*, Kasper PIHL†, Michael CHRISTIANSEN† and Claus OXVIG*1 *Department of Molecular Biology and Genetics, University of Aarhus, DK-8000 Aarhus, Denmark, and †Department of Clinical Biochemistry and Immunology, Statens Serum Institut, Copenhagen, Denmark

Table S1

Identification of the constituents of HMW AGT present post-sampling in term pregnancy serum

Bands following non-reducing SDS/PAGE (Figures 3D and 3E of the main text) of affinity purified AGT complexes were analysed. Peptide maps were generated by in-gel digest and MALDI–TOF-MS, and the proteins of individual bands were identified by using the Mascot search engine (http://www.matrixscience.com). The band numbering refers to Figure 3(D) (bands 1–6) and 3(E) (bands 7–11) of the main text. Peptide coverage is relative to the full-length proteins. All identifications were significant according to the Mascot search report. Band

Protein name

Number of peptides

Sequence coverage (residues)

MOWSE score

1

AGT ProMBP C3b Diamine oxidase AGT ProMBP C3b Pregnancy-zone protein α-2-Macroglobulin Haptoglobin Haptoglobin-related protein Haemoglobin β-chain AGT ProMBP C3dg Haemoglobin β-chain AGT ProMBP C3dg AGT ProMBP AGT α-2-Macroglobulin Pregnancy-zone protein C3b α-chain Diamine oxidase ProMBP C3b β-chain IgM heavy chain AGT ProMBP Haptoglobin Haptoglobin-related protein C3dg Igλ light chain Igκ light chain Apolipoprotein A1 Haemoglobin β-chain Haemoglobin α-chain

10 11 30 15 9 8 19 17 20 8 7 5 12 9 12 5 10 7 12 10 9 14 19 19 21 12 7 13 10 7 9 9 8 15 5 7 9 10 4

22% (57–262) 40% (98–222) 26% (137–1644) 22% (80–647) 23% (64–262) 37% (98–222) 15% (137–1360) 15% (145–1269) 16% (175–1297) 18% (132–401) 23% (74–343) 55% (1–144) 36% (64–437) 43% (98–222) 9% (979–1260) 41% (18–132) 23% (57–370) 29% (98–222) 8% (955–1260) 22% (64–437) 43% (98–222) 38% (57–437) 14% (215–1297) 20% (16–1136) 16% (749–1582) 17% (80–719) 36% (98–222) 8% (137–657) 23% (65–392) 22% (64–437) 43% (98–222) 21% (132–401) 24% (33–347) 8% (955–1303) 38% (1–208) 39% (46–214) 37% (1–215) 70% (2–147) 28% (1–99)

84 71 163 81 66 69 75 72 91 60 59 69 113 86 70 75 67 53 56 102 100 148 76 91 100 68 82 68 70 82 102 84 60 100 57 68 77 116 52

2

3

4

5 6 7 8 9

10 11 12 13 14 15 16

Received 14 May 2012/21 August 2012; accepted 4 October 2012 Published as BJ Immediate Publication 4 October 2012, doi:10.1042/BJ20120801

1

To whom correspondence should be addressed (email [email protected]). c The Authors Journal compilation c 2013 Biochemical Society