The JI
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     
 

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by van Mirre, E.
Right arrow Articles by Hack, C. E.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by van Mirre, E.
Right arrow Articles by Hack, C. E.
The Journal of Immunology, 2004, 173: 332-339.
Copyright © 2004 by The American Association of Immunologists

Monomeric IgG in Intravenous Ig Preparations Is a Functional Antagonist of Fc{gamma}RII and Fc{gamma}RIIIb

Edwin van Mirre1,*, Jessica L. Teeling*, Jos W. M. van der Meer{dagger}, Wim K. Bleeker* and C. Erik Hack*,{ddagger}

* Department of Immunopathology, Sanquin Research, Amsterdam, {dagger} Department of Internal Medicine, University Medical Center St. Radboud, Nijmegen, and {ddagger} Department of Clinical Chemistry, Vrije Universiteit Medical Center, Amsterdam, The Netherlands


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
Intravenous Ig preparations (IVIg), originally developed as a substitution therapy for patients with low plasma IgG, are nowadays frequently used in the treatment of various immune diseases. However, the mechanism of action of IVIg in these diseases remains elusive and is often referred to as "immunomodulatory." We hypothesized that monomeric IgG may act as a low-affinity Fc{gamma}R antagonist and sought experimental evidence for this hypothesis. Human neutrophils as well Fc{gamma}RIIa-transfected IIA1.6 cells were used as Fc{gamma}R-positive cells and aggregated IgG (aIgG) or stable dimeric IgG as Fc{gamma}R-specific agonists for these cells. We found that monomeric IgG purified from IVIg at concentrations similar to that of IgG in plasma, diminished the binding of stable dimeric IgG to Fc{gamma}RIIa transfectants, reduced aIgG-induced influx of Ca2+ ions into the cytosol of neutrophils, and attenuated the aIgG-induced release of elastase. Notably, monomeric IgG by itself did not elicit these responses, nor did it affect these processes in response to fMLP. Absorption of IgG from normal plasma revealed that plasma IgG exerted similar effects as monomeric IgG in IVIg. In addition, adding monomeric IgG to blood of healthy volunteers showed a dose-dependent decrease of aIgG-induced elastase release. Finally, we observed decreased aIgG-induced polymorphonuclear neutrophil responses in two hypogammaglobulinemic patients upon treatment with IVIg. We conclude that monomeric IgG at physiological levels acts as a low-affinity Fc{gamma}R antagonist. Moreover, Fc{gamma}R antagonism constitutes an immunomodulatory effect of IVIg.


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
Polyspecific IgG for i.v. use (i.v. Ig preparation (IVIg))2 was originally developed as a substitution therapy for hypo- and agammaglobulinemic patients. However, Imbach et al. (1) observed that IVIg supplementation was in fact effective as a treatment in idiopathic thrombocytopenic purpura (ITP). This has led to the widespread use of IVIg in various immune diseases. However, the biological effects of IVIg explaining its efficacy in these diseases, are still poorly understood, although they are frequently referred to as "immunomodulatory" effects. One could postulate various mechanisms for these immunomodulatory effects of IVIg, some of which are dependent on a productive interaction of the Fc{gamma} portion of infused Ig with the Fc{gamma}R on effector cells (1, 2, 3, 4) or with proteins of the complement system (5, 6).

Blockade of the Fc{gamma}R on phagocytic cells, preventing the removal of sensitized platelets by the macrophages in the spleen and liver, is believed to be the mechanism of action by which IVIg is effective as a treatment of ITP (1). Indeed, blockade of Fc{gamma}R by IVIg on macrophages has been shown to inhibit macrophage-mediated phagocytosis of Ag-bearing target cells (2, 7). Depending on their expression on effector cells, Fc{gamma}R exert different effects. For example, on phagocytes, they mediate phagocytosis, endocytosis, Ab-dependent cellular cytotoxicity, and induction of respiratory burst. Three types of Fc{gamma}R, types I, II and III, are differentiated based on their affinity for monomeric IgG. Type I (Ka = 108–109 M–1) is considered a high-affinity receptor, whereas types II (Ka = 106 M–1) and III (Ka = 5.5 x 105 M–1) (8, 9) are considered to be low-affinity receptors. Therefore, it is postulated that Fc{gamma}RI binds monomeric IgG in vivo, whereas Fc{gamma}RII and Fc{gamma}RIII preferentially interact with immune complexes (ICs). Thus, regarding IVIg, Fc{gamma}RII and Fc{gamma}RIII will interact predominantly with di- or polymeric IgG, whereas Fc{gamma}RI likely reacts with monomeric IgG as well.

IVIg contain monomeric IgG as well as a variable amount of dimeric and polymeric IgG. In a mouse model for ITP, we have shown that dimeric IgG in IVIg potently inhibits removal of Ab-sensitized platelets, whereas preparations with low dimeric content were hardly active (2). Notably, the removal of sensitized platelets in this model is dependent on the low-affinity Fc{gamma}RIII. Hence, one could postulate that dimeric IgG is the active principle in IVIg, explaining the efficacy of this drug in ITP. To what extent monomeric IgG in IVIg may contribute to the blockade of Fc{gamma}RII and Fc{gamma}RIII is not known. Although dimeric IgG is more potent in reducing IC-mediated anaphylaxis in rats, we also observed that dimeric IgG induces anaphylaxis itself, whereas high-dose monomeric IgG did not, but still had a protective effect.3 In the present study, we investigated whether monomeric IgG at high concentrations could act as a low-affinity Fc{gamma}R antagonist. We sought for experimental evidence for this hypothesis, using Fc{gamma}RIIa-transfected cells as well as neutrophils as a model. Neutrophils express both Fc{gamma}RIIa and Fc{gamma}RIIIb (1–4 x 104 and 1–3 x 105 molecules per cell, respectively), and under resting conditions, no Fc{gamma}RI (10). Possible antagonistic effects by IVIg as well as by plasma IgG on neutrophils were investigated. Our results indicate that monomeric IgG, in plasma as well as in IVIg, has sufficient affinity to displace the binding of ICs to Fc{gamma}RII and Fc{gamma}RIII, resulting in an attenuation of signal transduction via these receptors.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
Patients and blood sampling

Two patients with low plasma IgG due to late-onset common variable immune deficiency, who had not been treated with IVIg before, received IVIg at 0.4 g per kg of body weight. Before, halfway through (~1 h after start of the infusion), and at the end of IVIg infusion (~1.5 h after start), blood was collected in heparin-coated tubes. Responsiveness of neutrophils for aggregated IgG (aIgG; as a model for ICs) was tested ex vivo by incubating whole blood with aIgG as described below. The patients as well as healthy volunteers contributed to this study after informed consent.

Ig preparations

IVIg (lot no. 01H03H443A; Sanquin CLB, Amsterdam, The Netherlands) contains 60 g/L protein, of which at least 95% is IgG.

To obtain monomeric IgG, IVIg was reconstituted and set at low pH by dialysis against 10 mM acetate buffer containing 0.24 M glucose and 0.037 M NaCl (pH 4.15). Monomeric content was verified by HPLC gel filtration (Superdex 200 HR 16/30; Amersham Biosciences, Uppsala, Sweden); no dimeric or polymeric content was detected. Throughout the experiments, all preparations used were analyzed for actual content of mono-, di-, and polymeric IgG using this gel filtration procedure. All experiments with monomeric IgG preparations reported in this paper were done with preparations that did not contain measurable amounts of di- or polymeric IgG. A part of monomeric IgG was biotinylated according to standard procedures (Pierce, Rockford, IL) and used for direct binding studies (see below). Gel filtration of the biotinylated preparation revealed it did not contain measurable amounts of di- or polymeric IgG.

IgG dimers were prepared as described by Huizinga et al. (10). In short, mAb K37, directed against the {lambda}-L chain of human IgG, was digested by pepsin (Cooper Biomedical, Malvern, PA) for 20 h at 37°C to generate F(ab')2. The F(ab')2 were incubated with IVIg for 72 h at 4°C in a 1:2 ratio. Subsequently, complexes were purified on a Ultropac TSK-G4000SWG column (LKB-Producter, Bromma, Sweden) connected to a fast protein liquid chromatography system using PBS-0.02% sodium azide as running buffer. Fractions were collected and concentrated by dialysis against polyethylene glycol. Purity of the stable dimeric fraction was checked by gel filtration using a Superose 12 HR 10.30 column (Pharmacia, Peapack, NJ). The chromatogram was analyzed by a computer program (EZChrom, version 6.5; Pharmacia). The stable dimeric IgG fraction was pooled and biotinylated according to standard procedures (Pierce).

aIgG was obtained by incubating IVIg at 10 mg/ml in PBS (pH 7.4) for 30 min at 63°C (11). Gel filtration chromatography on a Superdex 200 HR 16/30 column revealed that the preparation contained 43% aIgG, no dimeric and 57% monomeric IgG, as analyzed by a computer program (Unicorn, version 4.5; Amersham Biosciences).

mAbs and reagents

The following mAb against human Fc{gamma}R were used: CD16 (anti-Fc{gamma}RIII, clone 3G8, prepared as F(ab')2; a generous gift from Dr. M. de Haas, Sanquin Research); CD16-PE (anti-Fc{gamma}RIII, IgG2a isotype, clone CLB-FcR-gran/1, 5D2; Sanquin Research), CD32-biotin (anti-Fc{gamma}RII, Fab of clone IV.3; also a generous gift from Dr. M. de Haas), and anti-CD64-FITC (anti-Fc{gamma}RI, IgG1 isotype, clone 10.1; InstruChemie, Delftzijl, The Netherlands).

FITC-conjugated rabbit-anti-human IgG polyclonal F(ab')2 were obtained from Sanquin Research and were reduced with 1 mM DTT, and subsequently, 2.5 mM iodoacetamide was added to obtain Fab. The Fab fragments were then dialyzed against PBS overnight to remove DTT and iodoacetamide. Relevant isotypes controls were obtained from Sanquin Research: isotype control IgG2a-PE (clone 713) and IgG1-FITC (clone 203). Streptavidin-allophycocyanin (strep-APC; BD Pharmingen, San Diego, CA) was used to visualize CD32-biotin on the cell surface.

Fab of clone anti-C1q-85, a generous gift from Drs. F. McGrath and D. Wouters (Sanquin Research), were used to block the classical pathway of complement activation.

Cell lines

IIA1.6 cells transfected with human Fc{gamma}RIIa R131 were a generous gift from Dr. J. Van de Winkel (University Medical Center Utrecht, Utrecht, The Netherlands) (12). This IIA1.6 transfectant was cultured in IMDM (BioWhittaker, Verviers, Belgium) supplemented with 10% (v/v) FCS (Bodinco, Alkmaar, The Netherlands), 100 U/ml penicillin (Invitrogen Life Technologies, Paisley, U.K.), and 100 µg/ml streptomycin (Invitrogen Life Technologies) under selection of 0.8 mg/ml geneticin (G-418-sulfate; Invitrogen Life Technologies).

Isolation of neutrophils and preparation of plasma and IgG-depleted plasma

Blood was obtained from healthy volunteers by venous puncture in heparin- or EDTA-containing tubes (Vacuette; Greiner Bio-One, Alphen a/d Rijn, The Netherlands).

Heparinized blood was diluted 1/1 in PBS containing 10% sodium citrate, layered on Percoll ({delta} = 1.078), and centrifuged at 2500 rpm for 15 min without brake. The pellet containing erythrocytes and neutrophils, was collected, and erythrocytes were lysed in ice-cold NH4Cl buffer (155 mM NH4Cl, 10 mM KHCO3, and 0.1 mM EDTA at pH 7.4). Subsequently, neutrophils were washed in HEPES buffer (123 mM NaCl, 5 mM KCl, 1 mM CaCl2, 1 mM MgCl2, 25 mM HEPES, and 10 mM glucose (pH 7.4)) and counted. Samples were taken and assessed by flow cytometry for the expression of Fc{gamma}R. Purity and viability was >95%, as determined by flow cytometry and tryptan blue exclusion, respectively. All cells used in the experiments were negative for Fc{gamma}RI and cell surface-bound IgG. EDTA-plasma was obtained by centrifugation at 3000 rpm for 10 min (no brake), and supplemented with 100 µg/ml anti-C1q Fab and 15 U/ml hirudin (a generous gift from Dr. H. Te Velthuis, Sanquin Research) to block classical pathway activation and clotting, respectively. The anti-C1q mAb will be described in more detail elsewhere (F. McGrath, G. Arland, M. Daha, and E. Hack, manuscript in preparation). The plasma was then recalcified with 20 mM CaCl2. IgG was depleted from plasma by batchwise incubation with protein G coupled to Sepharose (Amersham Biosciences). As a control, plasma was absorbed onto human serum albumin (Sanquin Research) coupled to CNBr-Sepharose (Amersham Biosciences).

Detection of IgG on the cell surface of neutrophils

Blood was obtained as described above and diluted 1/10 in FACS lysis buffer (BD Biosciences, San Jose, CA). Cells were then washed minimally, i.e., one time, or more extensively, and stained with FITC-conjugated rabbit anti-human IgG Fab. IgG binding to neutrophils was then analyzed by flow cytometry. Neutrophils were discriminated from other cell populations in the blood by their typical forward-scatter/side-scatter pattern, and checked for their absence of Fc{gamma}RI expression.

Binding of stable IgG dimers to Fc{gamma}R

Wells of a 96-well round-bottom plate were preincubated with PBS containing 2% (w/v) human serum albumin (Sanquin Research). Then, 50 µl of dilutions of biotinylated stable dimeric IgG were added into the wells, together with 50 µl of FACS buffer (PBS, 0.5% (w/v) BSA, 0.1% (w/v) NaN3) with or without monomeric IgG (0.1 mg/ml). Next to that, 75 µg/ml biotinylated monomeric IgG with or without 4.9 µg/ml Fc{gamma}RII-specific F(ab')2 (AT10; a generous gift from Dr. J. Van de Winkel), FACS buffer only, or solely AT10 F(ab')2 were put in precoated wells. Subsequently, IIA1.6 cells transfected with Fc{gamma}RIIa R131 were added and incubated for 15 min on ice. After washing, the cells were incubated with streptavidin-allophycocyanin (strep-APC; BD Pharmingen) for 15 min at 4°C in the dark, washed, and analyzed by flow cytometry on FACSCalibur (BD Biosciences).

Measurement of intracellular calcium in neutrophils

Neutrophils were suspended at a concentration of 107 cells/ml in HEPES buffer and loaded with 1 µM fura 2-AM (Molecular Probes, Leiden, The Netherlands) by incubation at 37°C for 45 min. The cells were then washed twice with HEPES buffer and suspended at 2 x 106/ml. Loaded neutrophils were transferred to a cuvette and stimulated either with monomeric IgG (1 and 10 mg/ml) or aIgG (1 mg/ml). Also, cells were pretreated with either 1 or 10 mg/ml monomeric IgG for 5 min at 37°C, and then stimulated with 1 mg/ml aIgG. The influx of Ca2+ ions was measured with a fluorometer (Luminescence Spectrometer LS55; PerkinElmer, Fremont, CA), and data were analyzed with FLWinlab software (PerkinElmer).

Ex vivo model to assess sensitivity of neutrophils for aIgG

A titration of aIgG with either 1, 10, or 20 mg/ml monomeric IgG or an equivalent amount of water to correct for dilution with monomeric IgG (in 90 µl of IMDM) was prepared in wells of a 96-well round-bottom plate. The wells had been precoated with 2% (v/v) human serum albumin in PBS for 1 h at 37°C. When whole blood was used, 15 U/ml hirudin and 100 µg/ml anti-C1q Fab were added. Whole blood or an equivalent amount of blood depleted for plasma, was then diluted 1/10 in the wells. The mixtures were then incubated for 2 h at 37°C in humidified air containing 5% CO2. Elastase release was then measured with sandwich ELISA as described (3). As a positive control for degranulation, 1 µM fMLP (Sigma-Aldrich, St. Louis, MO) and 5 µg/ml cytochalasin B (Sigma-Aldrich) were added. In the experiments where plasma and IgG-depleted plasma were used, 10% plasma was added to the titrated aIgG. Thereafter, the cells were diluted 1/10 as described above.

Analysis of data

Results are depicted as mean ± SD; where applicable, Student’s t test was used. A two-sided p value of <0.05 was considered to indicate a significant difference.


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
IgG is present on neutrophils in the circulation

We postulated that monomeric IgG at plasma concentration has sufficient affinity to bind to low-affinity Fc{gamma}R. Hence, it is expected that circulating cells having low-affinity Fc{gamma}R carry surface-bound IgG. To assess this, we collected fresh blood samples from three healthy donors and tested the neutrophils in these for surface-bound IgG, using limited and more extensive washing procedures. Neutrophils had expression levels of 40.5 ± 21.8 mean fluorescence intensity (MFI) and 255.8 ± 66.7 MFI for Fc{gamma}RII and Fc{gamma}RIII, respectively. Results obtained with one donor are shown in Fig. 1. As can be seen, IgG was present on the cell surface on neutrophils that were washed one time, to become rapidly dissociated from the cells after more extensive washing procedures. For example, after 10 times washing in FACS lysing buffer, IgG was undetectable on the cell surface of neutrophils by flow cytometry. Notably, the neutrophils used in this experiment were negative for the high-affinity Fc{gamma}RI (Fig. 1B). Results obtained with three other healthy donors yielded similar results.



View larger version (23K):
[in this window]
[in a new window]
  		FIGURE 1. IgG is detectable on the surface of circulating neutrophils. Blood was obtained from a healthy volunteer in heparin-coated tubes by venous puncture and diluted 1/10 in FACS lysis buffer. Cells were then washed as indicated, and stained with FITC-conjugated rabbit anti-human IgG Fab. A, IgG binding on neutrophils was analyzed by flow cytometry. Results obtained with three other volunteers were comparable. B, Neutrophils, in gate R1 in whole blood were Fc{gamma}RI negative, whereas monocytes, in gate R2, are Fc{gamma}RI positive. The dotted line indicates the isotype control, whereas the thick line indicates CD64-FITC mAb.

 
Fc{gamma}R binding studies

To evaluate whether the binding of ICs to surface Fc{gamma}R was altered by the presence of monomeric IgG, we performed binding studies in vitro using an Fc{gamma}RIIa-transfected cell line. The cells had a high expression of Fc{gamma}RIIa (1441.8 ± 170.8 MFI) as assessed by flow cytometry. Stable dimers were prepared as described in Materials and Methods, and used as a model to determine binding of IgG complexes to the transfected cells. Dose-dependent binding of the dimers to the transfected cells was observed (Fig. 2). Binding to the transfected cells was less when IgG dimers were incubated in the presence of IgG monomers. To verify that the decreased binding of dimeric IgG in the presence of monomeric IgG was caused by blockade of the Fc{gamma}RIIa, direct binding studies using biotinylated monomeric IgG were performed. Monomeric IgG, incubated at a concentration of 75 µg/ml, bound to the Fc{gamma}RIIa, which could be blocked by addition of F(ab')2 of a mAb against Fc{gamma}RII (AT10; Fig. 2B), suggesting that monomeric IgG, at the concentrations tested, is indeed able to directly bind to the low-affinity Fc{gamma}RIIa, interfering with subsequent binding of ICs.



View larger version (15K):
[in this window]
[in a new window]
  		FIGURE 2. Monomeric IgG inhibits binding of IgG dimers by occupation of Fc{gamma}RIIa. A, Several concentrations of biotinylated stable dimeric IgG were incubated with Fc{gamma}RIIa-R131-transfected cells. Binding of the dimeric IgG was visualized by flow cytometry, and expressed as MFI. {blacksquare}, Binding of dimeric IgG without monomeric IgG. {blacktriangleup}, Binding in the presence of 0.1 mg/ml monomeric IgG. *, p < 0.05. B, Monomeric IgG binds specifically to Fc{gamma}RIIa. First, the background fluorescence was determined (panel 1). Next, binding of monomeric IgG was measured (panel 2). After that, binding was blocked with AT10 F(ab')2 to determine specificity of binding (panel 3). Last, background fluorescence of AT10 F(ab')2 was determined (panel 4). The MFI in the second panel was significantly higher (p < 0.0005) compared with the other panels. There was no significant difference between panels 1, 3, and 4.

 
aIgG-mediated rise of intracellular Ca2+ is reduced by monomeric IgG

Upon cross-linking by specific mAbs or ICs, both Fc{gamma}RII and Fc{gamma}RIIIb on neutrophils will induce a rise in cytosolic free Ca2+ (13). To assess its possible antagonist effects on Fc{gamma}R, we investigated the effect of monomeric IgG on the increase of cytosolic Ca2+ upon triggering of Fc{gamma}R by aIgG, as a model for ICs. aIgG indeed induced an increase of intracellular Ca2+ in neutrophils, whereas treatment with monomeric IgG did not (Fig. 3A). Pretreatment of the cells with monomeric IgG reduced the influx of Ca2+ induced by aIgG (Fig. 3B). To assess whether this effect was due to desensitization by monomeric IgG, an Fc{gamma}R-independent stimulus, fMLP, was used to stimulate the cells. Monomeric IgG did not have a significant effect on the fMLP-mediated signal (Fig. 3C).



View larger version (11K):
[in this window]
[in a new window]
  		FIGURE 3. Monomeric IgG attenuates on aIgG-induced Ca2+ influx in neutrophils. A, Neutrophils (107 per milliliter) were loaded with 1 µM fura-2/AM. Under continuous stirring, background fluorescence was measured in a fluorometer for 100 s. The cells were then stimulated with either 1 mg/ml aIgG (dotted line) or 10 mg/ml monomeric IgG (solid line). B, Loaded neutrophils were either pretreated with 0 (solid line), 1 (dashed line), or 10 (dotted line) mg/ml monomeric IgG for 5 min. After measurement of background fluorescence for 100 s, cells were stimulated with 1 mg/ml aIgG. C, Loaded neutrophils were pretreated with 0 (dotted line) or 10 (solid line) mg/ml monomeric IgG for 5 min. Cells were then stimulated with 1 µM fMLP. The experiments shown were repeated six times with identical results.

 
aIgG-triggered neutrophil degranulation is attenuated by monomeric IgG

Downstream of the signaling cascade of Fc{gamma}R is the release of azurophilic granule content (13, 14). To study the effect of monomeric IgG on Fc{gamma}R-induced degranulation, blood was depleted from plasma and stimulated for 2 h with aIgG with or without monomeric IgG. Degranulation was assessed by measuring the release of elastase. aIgG-mediated elastase release was significantly reduced by the addition of monomeric IgG (Fig. 4A). To study whether this effect was due to desensitization by monomeric IgG, the effect on fMLP/cytochalasin B-induced degranulation was also studied. As can be seen in Fig. 4B, monomeric IgG had no effect on fMLP/cytochalasin B-mediated elastase release. Also, monomeric IgG itself did not induce degranulation of neutrophils in the blood (Fig. 4B). Furthermore, addition of blocking Fab or F(ab')2 of mAbs against Fc{gamma}RII and Fc{gamma}RIII (IV.3 Fab and 3G8 F(ab')2, respectively) attenuated the release of elastase (data not shown and Ref. 3).



View larger version (18K):
[in this window]
[in a new window]
  		FIGURE 4. aIgG-mediated elastase release by neutrophils is diminished by monomeric IgG. A, Blood cells were washed three times with an excess of IMDM to remove plasma, and added 1:10 to wells precoated with human serum albumin and IMDM containing aIgG at the indicated concentrations with ({blacksquare}) or without ({diamondsuit}) 1 mg/ml monomeric IgG. *, p < 0.05. B, The light gray bar indicates 1 mg/ml monomeric IgG was added. The dark gray bars represent the wells where no monomeric IgG was added. Either medium or 1 µM fMLP with 5 µg/ml cytochalasin B was added to the wells. The wells were then incubated for 2 h at 37°C. Finally, elastase concentrations in the well were measured with ELISA. The experiment was performed three times. Data represent mean ± SD. C, Blood was obtained from two patients with low IgG during a first treatment with IVIg, before, halfway through (~1 h after start of the infusion), and at the end of infusion (1.5 h after the start). Whole blood was diluted 1/10 in IMDM containing 1 mg/ml aIgG, final concentration, and incubated for 2 h at 37°C. Thereafter, elastase content of the mixtures was measured. Data represent mean ± SD.

 
Two patients diagnosed with late-onset common variable immune deficiency resulting in hypo- or agammaglobulinemia were treated for the first time with IVIg. Blood was sampled prior, during, and after infusion, and stimulated ex vivo with aIgG. Neutrophils of both patients responded to stimulation with aIgG before infusion with IVIg (Fig. 4C). However, at the end of infusion, responsiveness to stimulation with aIgG had decreased in both patients.

Plasma IgG down-modulates aIgG-induced neutrophil degranulation

One may postulate that the tertiary structure of monomeric IgG in IVIg differs from that of plasma IgG due to manufacturing artifacts, and that the effects observed with monomeric IgG purified from IVIg are not representative for plasma IgG. In addition, one may postulate the presence of anti-neutrophil Abs in the IVIg preparation, which via a Kurlander phenomenon (15) can cause a blockade of Fc{gamma}R. Hence, we studied whether autologous plasma IgG could exert an effect similar to that of monomeric IgG in IVIg. Therefore, we depleted plasma for IgG, as described in Materials and Methods. IgG concentration, as determined with nephelometry, in this plasma was <0.04 vs 9.58 ± 3.57 g/L IgG before depletion. As a control, plasma was absorbed over inactivated CNBr-Sepharose and used as IgG containing plasma. Plasma, either depleted for IgG or not, was mixed with IMDM and with peripheral blood cells to yield 10% plasma, final concentration. Before incubation with neutrophils, a mAb that inhibits C1q, was added to block the classical pathway.

Blood reconstituted with 10% control plasma or 10% IgG-depleted plasma, was stimulated for 2 h with aIgG or fMLP/cytochalasin B. Neutrophils stimulated with aIgG in the absence of plasma IgG released more elastase than neutrophils stimulated in the presence of plasma IgG (Fig. 5A), whereas cells stimulated with fMLP/cytochalasin B responded equally irrespective of the presence or absence of plasma IgG (B). Furthermore, plasma that had been adsorbed over HSA-coated CNBr-Sepharose did not affect aIgG-induced degranulation of neutrophils (Fig. 5B).



View larger version (13K):
[in this window]
[in a new window]
  		FIGURE 5. Plasma IgG reduces aIgG-induced elastase release by neutrophils. A, EDTA blood was centrifuged, and plasma was obtained and depleted for IgG as described in Materials and Methods. Blood cells were washed with IMDM, and added 1:10 to wells precoated with albumin and containing a titration of aIgG in IMDM, supplemented with either 10% control plasma ({blacktriangleup}) or 10% IgG-depleted plasma ({blacksquare}). *, p < 0.05. B, Buffer (light gray bars) or 1 µM fMLP with 5 µg/ml cytochalasin B (dark gray bars) were used as blank and positive control, respectively. The wells were incubated for 2 h at 37°C. C, Blood was added 1:10 to wells precoated with human serum albumin and IMDM containing 15 U/ml hirudin, anti-C1q Fab, 1 mg/ml aIgG, and various concentrations of monomeric IgG, as represented by the symbols in the curves. The wells were then incubated for 2 h at 37°C. Single dots represent blanks. Each curve represents an independent individual experiment. Elastase content of the wells was then measured with ELISA. Data represent mean ± SD.

 
Although physiological levels of IgG in plasma already occupy most Fc{gamma}R on the blood cells, ICs can still bind to leukocytes and exert, albeit attenuated, a biological response. However, ITP is nowadays treated with high-dose IVIg while these patients more or less have normal IgG levels.

Therefore, we used blood from healthy volunteers and stimulated with aIgG in the presence or absence of various doses monomeric IgG and evaluated responses by elastase release.

As shown in Fig. 5C, addition of monomeric IgG resulted in a dose-dependent decrease in aIgG-induced elastase release. Furthermore, when a dose of 20 mg/ml, a dose easily reached during infusion therapy, was added to the blood, elastase release was decreased to background levels (as shown by single dots).


   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
The low-affinity Fc{gamma}RII and Fc{gamma}RIII are supposed to be stimulated by polyvalent interactions with multimerized IgG, for example, IgG in ICs (16, 17). Conversely, monomeric IgG should be unable to induce signal transduction via these receptors. Indeed, also in the present study, we observed that monomeric IgG did not stimulate Ca2+ influx or release of elastase by neutrophils. However, we also show that, despite its inability to induce signal transduction via these receptors, monomeric IgG has sufficient affinity for Fc{gamma}RII and Fc{gamma}RIIIb to bind to these receptors, thereby attenuating signal transduction by aIgG. This antagonism of Fc{gamma}RII and Fc{gamma}RIIIb by monomeric IgG was observed with IgG in IVIg as well as with plasma IgG, and at concentrations that occur in vivo. Furthermore, infusion of IVIg in patients with low plasma IgG who had never received this drug before, resulted in diminished responsiveness of neutrophils to aIgG, supporting that the observed effects are relevant in vivo. In addition, at therapeutic concentrations, monomeric IgG, the major constituent of IVIg, attenuates aIgG-induced effects in blood. Therefore, functional blockade of low-affinity Fc{gamma}R might be an important early mechanism of action of IVIg.

ICs rather than aIgG constitute major agonists for low-affinity Fc{gamma}R in vivo. However, it is well established that aIgG and IC are very much alike; both activate the classical pathway of the complement system (18), and both fully interact with Fc{gamma}R on phagocytes (19, 20). As it is more stable and easier to prepare with human Abs than IC, we decided to use aIgG in our studies. aIgG or IC can activate complement in plasma to become opsonized with complement factors such as C3b (18). Hence, in the experiments with plasma, aIgG potentially could interact with complement receptors on cells as well, which may blur the effects mediated by Fc{gamma}R. To prevent this, we added a mAb that blocked the interaction of C1q with aIgG, and thereby prevented activation of complement by aIgG.

Monomeric IgG derived from IVIg contains a variable amount of contaminating dimeric IgG. To remove this contaminating dimeric IgG, IVIg was dialysed to low pH, resulting in the disruption of any dimers formed (21). Furthermore, control experiments, in which we tested the effect of small amounts of dimeric IgG, revealed that small amounts of contaminating dimers (<0.5%) in the monomeric IgG fraction cannot explain the inhibiting effect of monomeric IgG on the binding of dimeric IgG to Fc{gamma}RIIa-transfected cells (data not shown). Such an amount of dimeric IgG in the monomeric preparation would have been detected upon gel filtration, but was not found in the monomeric IgG preparation we used for the experiments. Hence, contaminating dimers do not explain the effects of monomeric IgG on the binding to Fc{gamma}RIIa. Moreover, because these dimers have the capability to cross-link low-affinity Fc{gamma}RII or Fc{gamma}RIII and induce signal transduction (3, 22), the observed inhibiting effect of monomeric IgG on aIgG-induced activating effects cannot be explained by the presence of dimeric or polymeric IgG.

Although obtained from plasma pools of >1000 donors, IgG in IVIg may be slightly distinct from plasma IgG. For example, the manufacturing processes may alter the tertiary structure of IgG somewhat as compared with native IgG in plasma. However, in the experiments in which plasma was either depleted for IgG or not, we also found an inhibiting effect of autologous nonpurified plasma IgG on aIgG-induced neutrophil stimulation. This suggests that, although there may be structural differences between monomeric IgG in plasma, and that in IVIg, the effect of monomeric IgG on low-affinity Fc{gamma}R is not unique for IVIg-derived IgG, but apparently is a general property of monomeric IgG.

The lack of fucose groups on human IgG improves binding of this IgG to human Fc{gamma}RIII and enhances Ab-dependent cellular cytotoxicity (23). However, as far as we know, this effect is unique for the interaction with Fc{gamma}RIIIa and not shared by other low-affinity Fc{gamma}R. Fc{gamma}RIIIa is not expressed by neutrophils (24). Hence, it is unlikely that the effects of monomeric IgG as described in the present article, are related to the presence or absence of this fucose group in the IgG molecules. In addition, it is commonly accepted that Fc{gamma}R cross-linking is required for signal transduction and activation of the cells. There is some debate whether monomeric IgG can interact with just one or possibly two Fc{gamma}R (25, 26). Our results show that monomeric IgG does not activate neutrophils, although it binds to at least Fc{gamma}RIIa (Fig. 2). Thus, monomeric IgG apparently is incapable of cross-linking the low-affinity Fc{gamma}R. This strongly suggests that monomeric IgG binds low-affinity Fc{gamma}R in a 1:1 ratio without inducing activation.

Activation of human granulocytes by IVIg is reported to occur via stimulation of Fc{gamma}RII and Fc{gamma}RIIIb (3, 22). This activation is mainly mediated by di- and polymeric IgG present in these preparations. Monomeric IgG cannot activate neutrophils unless these express Fc{gamma}RI, for example, in inflammatory disease (27, 28, 29). We found that some donors constitutively express this high-affinity receptor (data not shown). The presence of the high-affinity Fc{gamma}RI would be an obvious explanation for the observed binding of monomeric IgG. However, we screened potential donors for Fc{gamma}RI expression on neutrophils, and only used neutrophils that had undetectable expression of Fc{gamma}RI. Hence, interaction of monomeric IgG with Fc{gamma}RI is excluded to explain our results. In contrast, even if all neutrophils would express Fc{gamma}RI, and all these receptors would bind monomeric IgG, this would scavenge ~0.1% of the monomeric IgG present under physiological conditions, thus leaving sufficient monomeric IgG to bind Fc{gamma}RII and Fc{gamma}RIIIb. In the case of monomeric IgG purified from IVIg, one could postulate that the inhibiting effect was mediated via anti-neutrophil Abs in IVIg, which after binding via their Ag-binding sites to Ags on the neutrophils, interacted with low-affinity Fc{gamma}R on the cells via their Fc fragments, the so-called Kurlander phenomenon (15). However, the inhibiting effect was also observed with autologous IgG in plasma. Because Abs against autologous neutrophils in general do not occur in normal plasma, a Kurlander phenomenon cannot explain the inhibiting effect by monomeric IgG.

We showed that physiological levels of monomeric IgG (10 mg/ml) are unable to trigger an Fc{gamma}R-mediated intracellular Ca2+ rise. It could be argued that monomeric IgG cannot bind to low-affinity Fc{gamma}R at all. However, we observed binding of monomeric IgG to cells transfected with Fc{gamma}RIIa, and this binding was completely blocked by the addition of a Fab specific for Fc{gamma}RII. In addition, the studies of Galon et al. (8) and Maenaka et al. (9) showed that monomeric IgG as well as Fc fragments bind both Fc{gamma}RIIa and Fc{gamma}RIII with fast kinetics. Importantly, at physiological concentrations, given the affinities mentioned above, IgG would occupy 98.5 and 97.4% of Fc{gamma}RIIa and Fc{gamma}RIIIb, respectively. Furthermore, our current data show that pretreatment with high-dose monomeric IVIg, down-modulates aIgG-induced changes of intracellular Ca2+. This suggests that monomeric IgG indeed binds in a dose-dependent manner to low-affinity Fc{gamma}R. It could also be stated that monomeric IVIg induces a functional refractory state of the neutrophils. Nevertheless, our experiments show that, although pretreated with high-dose monomeric IgG, neutrophils still respond to fMLP stimulation as well as untreated neutrophils. Thus, monomeric IgG does not render neutrophils refractory to further stimulation. Another explanation for the effects of monomeric IgG could be down-modulation of the Fc{gamma}R. However, binding experiments were performed in the cold in the presence of NaN3 and mAb against Fc{gamma}RII revealed expression levels compared with untreated cells (data not shown) rendering down-modulation highly unlikely. Furthermore, it is unlikely that monomeric IgG exerts its effect through induction of Fc{gamma}RIIb expression on the neutrophils, because the preincubation time (5 min) presumably was too short to establish this (30).

In conclusion, we show that monomeric IgG is able to bind and saturate low-affinity Fc{gamma}R to attenuate signal transduction by aIgG via these receptors. Hence, monomeric IgG at physiological as well as therapeutical levels acts as a functional receptor antagonist for low-affinity Fc{gamma}R. This also holds in vivo as shown by the results of two patients who received a first dose of IVIg. Importantly, these results together suggest that, under normal conditions, low-affinity Fc{gamma}R on neutrophils in the circulation are partially saturated, making it difficult for ICs to bind. Patients with hypo- or agammaglobulinemia lack this inhibiting effect of monomeric IgG, and consequently may be more prone than persons with normal IgG levels to develop side effects after administration of IVIg containing some dimeric or polymeric IgG. Observations in patients indeed support this notion (16).


   Acknowledgments
 
We thank Mr. Anton J. Tool and Dr. Arthur Verhoeven for their excellent assistance with the measurement of calcium transients and for providing us with fMLP and cytochalasin B.


   Footnotes
 
1 Address correspondence and reprint requests to Dr. Edwin van Mirre, Department of Immunopathology, Sanquin Research, P.O. Box 9190, Plesmanlaan 125, 1066 CX Amsterdam, The Netherlands. E-mail address: e.vanmirre{at}sanquin.nl Back

2 Abbreviations used in this paper: IVIg, i.v. Ig preparation; ITP, idiopathic thrombocytopenic purpura; aIgG, aggregated IgG; MFI, mean fluorescence intensity. Back

3 J. Teeling, E. van Mirre, W. Bleeker, G. Rigter, T. Küypers, and E. Hack. Amelioration of immune complex-mediated anaphylaxis by intravenous gammaglobulin (IVIg) in rat. Submitted for publication. Back

Received for publication November 24, 2003. Accepted for publication April 30, 2004.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Imbach, P., S. Barandun, V. d’Apuzzo, C. Baumgartner, A. Hirt, A. Morell, E. Rossi, M. Schoni, M. Vest, H. P. Wagner. 1981. High-dose intravenous gammaglobulin for idiopathic thrombocytopenic purpura in childhood. Lancet 1:1228.[Medline]
  2. Teeling, J. L., T. Jansen-Hendriks, T. W. Kuijpers, M. de Haas, J. G. Van de Winkel, C. E. Hack, W. K. Bleeker. 2001. Therapeutic efficacy of intravenous immunoglobulin preparations depends on the immunoglobulin G dimers: studies in experimental immune thrombocytopenia. Blood 98:1095.[Abstract/Free Full Text]
  3. Teeling, J. L., E. R. De Groot, A. J. Eerenberg, W. K. Bleeker, G. Van Mierlo, L. A. Aarden, C. E. Hack. 1998. Human intravenous immunoglobulin (IVIG) preparations degranulate human neutrophils in vitro. Clin. Exp. Immunol. 114:264.[Medline]
  4. Lazarus, A. H., J. Freedman, J. W. Semple. 1998. Intravenous immunoglobulin and anti-D in idiopathic thrombocytopenic purpura (ITP): mechanisms of action. Transfus. Sci. 19:289.[Medline]
  5. Mollnes, T. E., I. H. Andreassen, K. Hogasen, C. E. Hack, M. Harboe. 1997. Effect of whole and fractionated intravenous immunoglobulin on complement in vitro. Mol. Immunol. 34:719.[Medline]
  6. Basta, M., P. Kirshbom, M. M. Frank, L. F. Fries. 1989. Mechanism of therapeutic effect of high-dose intravenous immunoglobulin: attenuation of acute, complement-dependent immune damage in a guinea pig model. J. Clin. Invest. 84:1974.[Medline]
  7. Dalakas, M. C.. 1998. Mechanism of action of intravenous immunoglobulin and therapeutic considerations in the treatment of autoimmune neurologic diseases. Neurology 51:S2.[Abstract/Free Full Text]
  8. Galon, J., M. W. Robertson, A. Galinha, N. Mazieres, R. Spagnoli, W. H. Fridman, C. Sautes. 1997. Affinity of the interaction between Fc{gamma} receptor type III (Fc{gamma}RIII) and monomeric human IgG subclasses: role of Fc{gamma}RIII glycosylation. Eur. J. Immunol. 27:1928.[Medline]
  9. Maenaka, K., P. A. van der Merwe, D. I. Stuart, E. Y. Jones, P. Sondermann. 2001. The human low affinity Fc{gamma} receptors IIa, IIb, and III bind IgG with fast kinetics and distinct thermodynamic properties. J. Biol. Chem. 276:44898.[Abstract/Free Full Text]
  10. Huizinga, T. W., M. Kerst, J. H. Nuyens, A. Vlug, B. von dem, D. Roos, P. A. Tetteroo. 1989. Binding characteristics of dimeric IgG subclass complexes to human neutrophils. J. Immunol. 142:2359.[Abstract]
  11. Hack, C. E., A. J. Belmer. 1986. The IgG detected in the C1q solid-phase immune-complex assay is not always of immune-complex nature. Clin. Immunol. Immunopathol. 38:120.[Medline]
  12. Van Den Herik-Oudijk, I. E., N. A. Westerdaal, N. V. Henriquez, P. J. Capel, J. G. Van de Winkel. 1994. Functional analysis of human Fc{gamma}RII (CD32) isoforms expressed in B lymphocytes. J. Immunol. 152:574.[Abstract]
  13. Vossebeld, P. J., C. H. Homburg, R. C. Schweizer, I. Ibarrola, J. Kessler, L. Koenderman, D. Roos, A. J. Verhoeven. 1997. Tyrosine phosphorylation-dependent activation of phosphatidylinositide 3-kinase occurs upstream of Ca2+-signalling induced by Fc{gamma} receptor cross-linking in human neutrophils. Biochem. J. 323:87.
  14. Naucler, C., S. Grinstein, R. Sundler, H. Tapper. 2002. Signaling to localized degranulation in neutrophils adherent to immune complexes. J. Leukocyte Biol. 71:701.[Abstract/Free Full Text]
  15. Kurlander, R. J.. 1983. Blockade of Fc receptor-mediated binding to U-937 cells by murine monoclonal antibodies directed against a variety of surface antigens. J. Immunol. 131:140.[Abstract]
  16. Barandun, S., P. Kistler, F. Jeunet, H. Isliker. 1962. Intravenous administration of human gammaglobulin. Vox Sang. 7:157.[Medline]
  17. De Grandmont, M. J., C. Racine, A. Roy, R. Lemieux, S. Neron. 2003. Intravenous immunoglobulins induce the in vitro differentiation of human B lymphocytes and the secretion of IgG. Blood 101:3065.[Abstract/Free Full Text]
  18. Cowdery, J. S., Jr, P. E. Treadwell, R. B. Fritz. 1975. A radioimmunoassay for human antigen-antibody complexes in clinical material. J. Immunol. 114:5.[Abstract/Free Full Text]
  19. Daha, M. R., L. A. van Es. 1984. Fc- and complement receptor-dependent degradation of soluble immune complexes and stable immunoglobulin aggregates by guinea pig monocytes, peritoneal macrophages, and Kupffer cells. J. Leukocyte Biol. 36:569.[Abstract]
  20. Gormus, B. J., R. L. Vessella, L. N. Martin, M. E. Kaplan. 1982. Heterogeneity of human lymphocyte Fc receptors: studies using heat-aggregated and antigen-complexed IgG from human, rabbit, guinea pig, horse and goat. Comp. Immunol. Microbiol. Infect. Dis. 5:483.[Medline]
  21. Tankersley, D. L., M. S. Preston, J. S. Finlayson. 1988. Immunoglobulin G dimer: an idiotype-anti-idiotype complex. Mol. Immunol. 25:41.[Medline]
  22. Nemes, E., F. Teichman, D. Roos, L. Marodi. 2000. Activation of human granulocytes by intravenous immunoglobulin preparations is mediated by Fc{gamma}RII and Fc{gamma}RIII receptors. Pediatr. Res. 47:357.[Medline]
  23. Shields, R. L., J. Lai, R. Keck, L. Y. O’Connell, K. Hong, Y. G. Meng, S. H. Weikert, L. G. Presta. 2002. Lack of fucose on human IgG1 N-linked oligosaccharide improves binding to human Fc{gamma}RIII and antibody-dependent cellular toxicity. J. Biol. Chem. 277:26733.[Abstract/Free Full Text]
  24. Huizinga, T. W., C. E. van der Schoot, C. Jost, R. Klaassen, M. Kleijer, A. E. dem Borne, D. Roos, P. A. Tetteroo. 1988. The PI-linked receptor FcRIII is released on stimulation of neutrophils. Nature 333:667.[Medline]
  25. Kato, K., W. H. Fridman, Y. Arata, C. Sautes-Fridman. 2000. A conformational change in the Fc precludes the binding of two Fc{gamma} receptor molecules to one IgG. Immunol. Today 21:310.[Medline]
  26. Kato, K., C. Sautès-Fridman, W. Yamada, K. Kobayashi, S. Uchiyama, H. Kim, J. Enokizono, A. Galinha, Y. Kobayashi, W. H. Fridman, et al 2000. Structural basis of the interaction between IgG and Fc{gamma} receptors. J. Mol. Biol. 295:213.[Medline]
  27. Reilly, M. P., S. E. McKenzie. 2002. Mechanisms of action of IVIg: physiology of Fc receptors. Vox Sang. 83:(Suppl. 1):57.
  28. Basta, M., F. Van Goor, S. Luccioli, E. M. Billings, A. O. Vortmeyer, L. Baranyi, J. Szebeni, C. R. Alving, M. C. Carroll, I. Berkower, et al 2003. F(ab)2-mediated neutralization of C3a and C5a anaphylatoxins: a novel effector function of immunoglobulins. Nat. Med. 9:431.[Medline]
  29. Lamoureux, J., E. Aubin, R. Lemieux. 2003. Autoimmune complexes in human serum in presence of therapeutic amounts of intravenous immunoglobulins. Blood 101:1660.[Free Full Text]
  30. Samuelsson, A., T. L. Towers, J. V. Ravetch. 2001. Anti-inflammatory activity of IVIG mediated through the inhibitory Fc receptor. Science 19:291.



This article has been cited by other articles:


Home page
 BloodHome page
W. B. Breunis, E. van Mirre, M. Bruin, J. Geissler, M. de Boer, M. Peters, D. Roos, M. de Haas, H. R. Koene, and T. W. Kuijpers
Copy number variation of the activating FCGR2C gene predisposes to idiopathic thrombocytopenic purpura
Blood, February 1, 2008; 111(3): 1029 - 1038.
[Abstract] [Full Text] [PDF]

Home page
 BloodHome page
T. Tha-In, H. J. Metselaar, H. W. Tilanus, Z. M. A. Groothuismink, E. J. Kuipers, R. A. de Man, and J. Kwekkeboom
Intravenous immunoglobulins suppress T-cell priming by modulating the bidirectional interaction between dendritic cells and natural killer cells
Blood, November 1, 2007; 110(9): 3253 - 3262.
[Abstract] [Full Text] [PDF]

Home page
 J Am Coll CardiolHome page
A. Staudt, P. Eichler, C. Trimpert, S. B. Felix, and A. Greinacher
Fc{gamma} Receptors IIa on Cardiomyocytes and Their Potential Functional Relevance in Dilated Cardiomyopathy
J. Am. Coll. Cardiol., April 24, 2007; 49(16): 1684 - 1692.
[Abstract] [Full Text] [PDF]

Home page
 JDRHome page
Y.-T.A. Teng
Protective and Destructive Immunity in the Periodontium: Part 1--Innate and Humoral Immunity and the Periodontium
Journal of Dental Research, March 1, 2006; 85(3): 198 - 208.
[Abstract] [Full Text] [PDF]

Home page
 Infect. Immun.Home page
A. Casadevall and L.-a. Pirofski
New Concepts in Antibody-Mediated Immunity
Infect. Immun., November 1, 2004; 72(11): 6191 - 6196.
[Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by van Mirre, E.
Right arrow Articles by Hack, C. E.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by van Mirre, E.
Right arrow Articles by Hack, C. E.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS