| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
RIA (CD64A)1
* Department of Autoimmunity and Inflammation,
Department of Protein Biochemistry, and
Department of Pre-Clinical Development, ZymoGenetics, Seattle, WA 98102
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
R promotes cell activation, release of inflammatory mediators, and tissue destruction characteristic of autoimmune disease. To evaluate whether a soluble FcR could block the proinflammatory effects of immune complexes, recombinant human (rh) versions of FcRIA, FcRIIA, and FcRIIIA were prepared. Binding of rh-FcRIA to IgG was of high affinity (KD = 1.7 x 10–10 M), whereas rh-FcRIIA and rh-FcRIIIA bound with low affinity (KD = 0.6–1.9 x 10–6 M). All rh-FcR reduced immune complex precipitation, blocked complement-mediated lysis of Ab-sensitized RBC, and inhibited immune complex-mediated production of IL-6, IL-13, MCP-1, and TNF-
by cultured mast cells. Local or systemic delivery only of rh-FcRIA, however, reduced edema and neutrophil infiltration in the cutaneous Arthus reaction in mice. 125I-labeled rh-FcRIA was cleared from mouse blood with a rapid distribution phase followed by a slow elimination phase with a t1/2 of 
130 h. The highest percentage of injected radioactivity accumulated in blood liver carcass > kidney. s.c. dosing of rh-FcRIA resulted in lower serum levels of inflammatory cytokines and prevented paw swelling and joint damage in a murine model of collagen Ab-induced arthritis. These data demonstrate that rh-FcRIA is an effective inhibitor of type III hypersensitivity. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
R, act as a link between the humoral and cellular arms of the immune system (1, 2). Ligation of these cell surface receptors by the Fc portion of IgG can trigger a variety of immune effector functions, such as Ag presentation, Ab-dependent cellular cytotoxicity, phagocytosis, and the release of inflammatory mediators. Studies of gene targeting in mice and of genetic mutations of the Fcr genes in humans have shown that the activating and inhibitory receptors play opposing roles in immune homeostasis (2). Deletion of the activating receptor -chains, or deletion of the common chain, renders animals resistant to immune complex-mediated autoimmune disease. Deletion of the inhibitory FcRIIb receptor, in contrast, produces animals that are more sensitive to the development of immune complex disease (2). Similarly, mutations in the activating FcRs, that appear to alter their interactions with immune complexes, are risk factors for the development of autoimmune disease within certain human populations (3).
These observations suggested an approach for treating immune complex-mediated disease through the use of agents that block the interaction of immune complexes with cell surface FcRs. As part of a screening effort to identify soluble receptors demonstrating this ability, the ligand-binding domains of the human activating receptors, FcRIA, FcRIIA, and FcRIIIA, were expressed in mammalian cells and purified to homogeneity from their conditioned medium. Each of the recombinant human (rh)3 FcR were tested for their ability to block immune complex-mediated inflammatory events in several in vitro systems and in models of type III hypersensitivity in mice.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
R
Expression constructs encoding soluble monomeric forms of the rh-FcR were generated using DNA sequences encoding their native signal sequence, their extracellular domain, and a C-terminal His6 tag (GSGGHHHHHH). The DNA sequence encoded amino acids 1–282 for rh-FcRIA, 1–212 for rh-FcRIIA, and 1–195 for rh-FcRIIIA. Receptors were purified from supernatants derived from Chinese hamster ovary (CHO) DXB-11 cells (L. Chasin, Columbia University, New York, NY). CHO-conditioned media were sterile filtered, concentrated, and buffer exchanged into 50 mM NaPO4, 500 mM NaCl, 25 mM imidazole (pH 7.5; buffer A). The His-tagged rh-FcR proteins were captured using Ni-NTA His Bind Superflow resin (Novagen) equilibrated in buffer A. Elution of bound protein was accomplished using a gradient of imidazole (0–500 mM) in 50 mM NaPO4, 500 mM NaCl (pH 7.5). Fractions were analyzed for rh-FcR by SDS-PAGE and Western blotting (anti-6x histidine HRP mouse IgG1; R&D Systems).
The Ni-NTA fractions containing rh-FcR were buffer-exchanged into 50 mM NaPO4, 150 mM NaCl (pH 7.5; buffer B) and incubated with Q Sepharose 4FF resin (GE Healthcare) that was pre-equilibrated in buffer B overnight at 4°C. The slurry was transferred to a gravity flow column; the flow-through and wash fractions were combined and assessed for the presence of rh-FcR as described above. The combined fractions were concentrated and injected onto a Superdex 200 Hiload (GE Healthcare) column equilibrated in 50 mM NaPO4, 109 mM NaCl (pH 7.3; buffer C). The column was eluted in buffer C and fractions containing rh-FcR were combined, concentrated, sterile-filtered, and stored at –80°C. rh-FcR were analyzed by SDS-PAGE, Western blotting, N-terminal sequencing, and size exclusion multiangle light scattering. Endotoxin levels were <1.0 endotoxin units/ml for each receptor preparation formulated at 20 mg/ml.
Binding of rh-FcR to immobilized human IgG1
Measurements were performed using a Biacore 3000 instrument. Activation of the sensor chip surface and covalent immobilization of the IgG1 Ab (
from human myeloma plasma; Sigma-Aldrich) was performed using 0.2 M N-ethyl-N'-(3-diethylamino-propyl) carbodiimide and 0.05 M N-hydroxysuccinimide and Biacore control software. The human IgG1 Ab, diluted to 11 µg/ml in 10 mM sodium acetate (pH 5.0), was immobilized to prepare the specific binding flow cell, and a second flow cell was activated, but not exposed to IgG1 to prepare the reference flow cell. The unreacted ester sites on both the specific binding and reference flow cells were blocked with 1 M ethanolamine hydrochloride.
For kinetic analysis of rh-FcRIA binding, the IgG1 Ab was immobilized at a level of 458 resonance units. The rh-FcRIA was injected over both the active and reference flow cells in series. For kinetic analysis of rh-FcRIA binding, a concentration range of 0.16–10.3 x 10–9 M rh-FcRIA in HBS-EP (Biacore) assay buffer (10 mM HEPES (pH 7.4), 0.15 M NaCl, 3.5 mM EDTA, 0.005% polysorbate 20) was used. The rh-FcRIA was injected at a flow rate of 40 µl/min for 3 min. Subsequently, the rh-FcRIA solution was switched to HBS-EP buffer and dissociation was measured for 3 min. Each rh-FcRIA concentration was tested in duplicate using a random sequence. Each measurement was followed by a single 30-s injection of 10 mM glycine-HCl (pH 1.8) at 50 µl/min to regenerate the IgG1 surface.
For equilibrium analyses of rh-FcRIIA and rh-FcRIIIA binding, the IgG1 Ab was immobilized at a level of 1013 resonance units. A concentration range of 0.03–24 x 10–6 M of rh-FcR was used. Each rh-FcR was injected at a flow rate of 10 µl/min for 1 min. The dissociation time for each rh-FcR was 5 min. Each rh-FcRIIA and rh-FcRIIIA concentration was tested in duplicate using a random sequence. Each measurement was followed by a single 30-s injection of HBS-EP at 30 µl/min to regenerate the IgG1 surface.
Binding curves for all three soluble rh-FcRs were processed by subtraction of the reference surface curve from the specific binding surface curve, as well as subtraction of a buffer-injection curve. The processed binding curves were globally fitted to a 1:1 binding model and the resulting kinetic and equilibrium constants were evaluated using Biacore software.
Immune complex precipitation assay
Chicken egg OVA (Calbiochem) was dissolved to a final concentration of 15 µg/ml in PBS and combined with 300 µg of rabbit polyclonal anti-OVA Abs/ml (Rockland Immunochemicals) in a final volume of 200 µl in the presence and absence of the indicated concentration of rh-FcR (4). Immediately thereafter, turbidity of the reaction mixture was monitored at 350 nm every 30 s for 5–10 min at 37°C with the aid of a spectrophotometer. Linear regression was used to calculate the slope of the linear portion of the turbidity curves, and the rh-FcR-mediated inhibition of immune complex precipitation was expressed relative to incubations containing anti-OVA and OVA alone. To assess the effects of IgG isotypes, the immune complex precipitation assay was conducted in the presence and absence of the indicated concentration of unrelated murine IgG1, IgG2a, and IgG2b mAbs (ZymoGenetics) or total mouse or human IgG (Rockland Immunochemicals). Activity of rh-FcR in the absence of IgG was set at 100%, and activity was calculated as described above.
Complement-mediated lysis of SRBCs
Ab-sensitized SRBCs (Sigma-Aldrich) were prepared and were incubated with the indicated concentration of rh-FcR. After 15 min at 4°C, a 25-µl sample of a 1/50 dilution of rat serum (Sigma-Aldrich) was added, and hemolysis was measured by monitoring the absorbance of the mixture at 540 nm as described by the manufacturer.
Immune complex-mediated cytokine release from murine mast cells
Immune complexes were prepared by mixing rabbit polyclonal anti-OVA (300 µl, 1.5 mg) with OVA (75 µl, 75 µg) in a 5-ml final volume of PBS. After 60 min at 37°C, the mixture was placed at 4°C for 18–20 h. The immune complexes were collected by centrifugation at 12,000 rpm for 5 min, the supernatant fraction was removed and discarded, and the immune complex precipitate was resuspended in 1 ml of ice-cold PBS, recentrifuged, and the immune complexes were resuspended in a final volume of 1 ml of ice-cold PBS. Protein concentration was determined using the BCA assay (Pierce).
MC/9 cells (American Type Culture Collection) were subcultured in medium A (DMEM containing 10% FBS, 50 µM 2-ME, 0.1 mM nonessential amino acids, 1 mM sodium pyruvate, 36 µg/ml L-asparagine, 1 ng/ml recombinant murine (rm) IL-3, 5 ng/ml rmIL-4, 25 ng/ml rm-stem cell factor) to a density of 0.5–3 x 106 cells/ml. Cells were collected by centrifugation at 1500 rpm for 5 min and the cell pellet was washed in medium A (without cytokines) and resuspended in medium A at 2.0 x 106 cells/ml. Aliquots of cells (2 x 105 cells) were incubated with 10 µg/well OVA/anti-OVA immune complexes in the presence and absence of rh-FcR in a final volume of 200 µl of buffer A in a 96-well microtiter plate. After 4 h at 37°C, the media were removed and centrifuged at 1500 rpm for 5 min. The cell-free supernatant fractions were collected and aliquots were analyzed for cytokine production using a Luminex cytokine assay kit (Upstate Biotechnology).
Cutaneous reversed passive Arthus reaction in mice
Ten-week-old female C57BL/6 mice (Charles River Laboratories; n = 8 mice/group) were anesthetized with isoflurane, their dorsal skin was shaved, and the back of each mouse was wiped with 70% alcohol. Each mouse received a tail vein injection of a 100-µl final volume of a solution containing 10 mg of OVA/ml and 10 mg of Evan’s blue/ml (Sigma- Aldrich). Immediately thereafter, each mouse received two intradermal (i.d.) injections, of 20 µl each, at distinct sites in the dorsal skin. The i.d. injection solution consisted of PBS containing either 40 µg of rabbit anti-OVA or 40 µg of rabbit nonimmune IgG (Sigma-Aldrich), with and without the indicated concentration of rh-FcR. All Ab solutions were heat-inactivated by incubation at 56°C for 40 min and centrifuged at 14,000 rpm for 10 min before injection. In some instances, the tail vein injection solution also contained 1 mg of dexamethasone/kg (ICN Biomedicals). After 4 h, the mice were euthanized with CO2 gas, the dorsal skin was removed at the level of fascia above the skeletal muscle, and the skin section was reversed. Cutaneous edema was evaluated by measuring the area of vascular leak of Evan’s blue dye (mm2) in the reversed tissue section and by assessing tissue weights (milligrams) of 10-mm punch biopsies encompassing the injection sites. The skin samples were then quickly frozen in liquid N2 and stored at –80°C. Neutrophil infiltration was assessed by measuring myeloperoxidase (MPO) activity in the punch biopsy samples as described (5) using a MPO assay kit (CytoStore). For systemic delivery of rh-FcRIA, mice received tail vein injections of 100 µl containing either vehicle alone or vehicle containing the indicated amount of rh-FcRIA, 1 h before initiating the Arthus reaction.
Pharmacokinetics and tissue distribution of 125I-labeled (125I)-FcRIA in mice
rh-FcRIA was labeled with Na125I (GE Healthcare Biosciences) to a final specific activity of 0.06 µCi/ng using the indirect Iodogen labeling procedure as described by the manufacturer (Pierce). The 125I-rh-FcRIA was homogenous with no evidence of protein aggregation by size exclusion HPLC and >98% of the radioactivity was precipitated with 10% TCA. No significant differences in bioactivity were noted between 125I-labeled and native rh-FcRIA using the immune complex precipitation assay. For the pharmacokinetic studies, the 125I-rh-FcRIA was diluted with unlabeled rh-FcRIA to the desired specific activity.
Female C57BL/6 mice (17–20 gm) received a 100-µl tail vein injection of 125I-rh-FcRIA (100 µg, 5.0 µCi). At each time point, three mice were anesthetized with halothane and blood samples (0.8 ml) were collected by cardiac puncture into serum separator tubes. After 15 min, the samples were centrifuged for 3 min at 16,000 x g and the resulting sera were processed to determine TCA-precipitable radioactivity. These data were converted to microgram equivalents per milliliter based on the specific activity of the dosed material. Mean serum concentration vs time data were analyzed using a three-compartment i.v. bolus model with a 1/y2 (where y is the predicted concentration) weighting function. Analyses were performed using WinNonlin Professional 5.0.1 software (Pharsight). Additional tissues were collected from the animals sacrificed at the 0.25-, 3-, and 24-h time points. Tissues were rinsed, blotted dry, weighed, and analyzed for radioactivity. The percent-injected dose was calculated based on the radioactivity in each tissue and the total radioactive dose administered.
Distribution of 125I-rh-FcRIA in the cutaneous Arthus reaction
rh-FcRIA was labeled with Na125I as described above. Groups of five mice each received a 100-µl s.c. or tail vein injection of 125I-rh-FcRIA (250 µg, 10 µCi). The Arthus reaction was initiated 1 h later as described above. After 4 h, the mice were sacrificed, the extravasation of Evan’s blue dye was measured, and 10-mm punch biopsies encompassing the lesion sites were taken for measurement of total radioactivity.
Collagen Ab-induced arthritis in mice
Male DBA/1J mice (8 wk old; Charles River Laboratories) received a 2-mg dose of Arthrogen-collagen-induced arthritis Ab mixture (Chemicon International) via tail vein injection on day 0 (6). Three days later, each mouse received a 500-µl i.p. injection containing 50 µg of LPS as described by the manufacturer. Mice received s.c. injections of either vehicle alone (PBS) or vehicle containing 0.67 or 2 mg of rh-FcRIA on days 3, 5, 7, and 9. Mice were scored daily for arthritis by measuring paw swelling with the aid of a caliper and by visual assessment. The extent of disease was evaluated in each paw by using a caliper to measure paw thickness, and by assigning a clinical score (0–3) to each paw: 0, normal; 0.5, one or more toes involved, but only the toes are inflamed; 1, mild inflammation involving the paw, and may include a toe or toes; 2, moderate inflammation in the paw and may include some of the toes and/or the wrist/ankle; and 3, severe inflammation in the paw, wrist/ankle, and some or all of the toes. The "average paw score" was calculated for each mouse by adding together the scores of each of the four paws, and then dividing by 4. Before sacrifice of the mice on day 10, blood was collected from each mouse by retro-orbital bleeding under gas isoflurane anesthesia. Blood samples were allowed to clot and serum was isolated by centrifugation and was stored at –80°C before measurement of serum cytokines by Luminex bead technology. Paws were collected into 10% buffered formalin and routinely processed for histology. All animal procedures and care were conducted in accordance with approved ethical guidelines under the auspices of the ZymoGenetics Institutional Animal Care and Use Committee.
Statistical analyses
Unless stated otherwise, all data are expressed as the mean ± SEM. Two group comparisons were assessed by either a two-tailed Student t test or by the Mann-Whitney U test. Multiple group comparisons were by ANOVA. Statistical significance was set at p 
0.05.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
R
Soluble rh-FcR were produced in CHO DXB-11 cells and purified from their conditioned medium. The expression levels for each rh-FcR were 20–100 mg/L and by SDS-PAGE analysis, each rh-FcR migrated as a broad band under reducing and nonreducing conditions that also reacted with an anti-6x histidine HRP tag Ab by Western blotting. The identity and homogeneity of each rh-FcR was confirmed by N-terminal sequencing and amino acid analyses (data not shown). By size exclusion multiangle light scattering analyses, the rh-FcR were of the expected monomeric molecular weights and were highly glycosylated (20–40% by weight), with little or no evidence of protein aggregation. The extensive glycosylation observed is consistent with the large number of potential N-linked sites on each of the rh-FcR (7).
Binding of rh-FcR to monomeric human IgG1
The rh-FcRs bound to immobilized human IgG1 in a manner that was best-fit by a 1:1 binding interaction. The IgG1 exhibited some loss of binding activity upon covalent immobilization and the activity of the surface ranged from 26 to 81% of the theoretical maximum. The association and dissociation phases of rh-FcRIA binding to IgG1 were measurable over a time period of >200 s, allowing kinetic analysis of the binding curves. rh-FcRIA bound to IgG1 with association (ka) and dissociation (kd) rate constants of 2.8 x 106 M–1s–1 and 4.6 x 10–4 s–1, respectively, yielding an equilibrium dissociation constant (KD) of 1.7 x 10–10 M. The association/dissociation rates for rh-FcRIIA and rh-FcRIIIA were too fast to measure accurately, so the equilibrium dissociation constants were determined at steady state. Binding of rh-FcRIIIA and rh-FcRIIA to IgG1 was saturable and of low affinity with estimated KDs of 0.63 x 10–6 M and 1.9 x 10–6 M, respectively. Each rh-FcR bound to immobilized rabbit anti-OVA IgG with rates and affinities similar to that observed with human IgG1 (data not shown).
Inhibition of immune complex precipitation by rh-FcR
Incubation of anti-OVA and OVA at 37°C produced time-dependent increases in immune complex precipitation, evident as increases in the OD of the mixtures. Preliminary studies showed that immune complex precipitation was dependent on the OVA:anti-OVA ratio and the total concentration of OVA and anti-OVA used, and it could be blocked by the addition of complement-containing, but not heat-inactivated, serum (data not shown). Addition of increasing amounts of each rh-FcR produced a concentration-dependent inhibition of immune complex precipitation. Linear regression was used to calculate the slopes of the linear portion of the curves (0–120 s), and the results were expressed as a percent of control (no added rh-FcR). Each rh-FcR decreased the precipitation of OVA/anti-OVA immune complexes with a relative order of potency rh-FcRIIIA > rh-FcRIA > rh-FcRIIA (Fig. 1). Maximal inhibition was seen using 1–1.5 µM for each rh-FcR, a molar ratio of rh-FcR:anti-OVA of 1:1.
|
R
Incubation of the murine mast cell line MC/9 with increasing amounts of anti-OVA:OVA immune complexes increased the amount of the inflammatory cytokines IL-6, IL-13, TNF-, and MCP-1 that accumulated in their extracellular medium. Little or no production of these molecules was observed, however, when MC/9 cells were incubated with an equivalent amount of anti-OVA alone (data not shown). Induction of cytokine production was dependent on length of incubation with immune complexes, with maximal induction seen within 4 h (data not shown). In contrast, incubation of MC/9 cells with immune complexes had no discernable effects on the production of IL-1β, IL-10, IL-12, IFN-, or GM-CSF under any conditions. Incubation of MC/9 cells with immune complexes in the presence of increasing amounts of rh-FcR resulted in a dose-dependent reduction in the levels of IL-6, IL-13, TNF-, and MCP-1 in the mast cell-conditioned medium (Fig. 2) with a relative order of potency rh-FcRIA > rh-FcRIIIA > rh-FcRIIA. For each rh-FcR, the IC50s were similar for each cytokine examined.
|
R
Incubation of Ab-sensitized SRBCs with rat serum at 37°C resulted in complement activation and lysis of the SRBCs. Addition of each rh-FcR to the incubation mixtures blocked SRBC lysis in a dose-dependent manner, with a relative order of potency rh-FcRIA > rh-FcRIIIA > rh-FcRIIA. Little or no inhibition of hemolysis was observed, in contrast, with an unrelated control fusions protein, transmembrane activator and CAML interactor Ig (TACI-Ig) (Fig. 3).
|
RIA blocks inflammation in the cutaneous Arthus reaction in mice
To evaluate whether the anti-inflammatory activities of rh-FcR observed in vitro could be observed in a model of immune complex-mediated hypersensitivity in vivo, the biological activities of rh-FcR were assessed in the cutaneous Arthus reaction in mice. Intravenous administration of a mixture of OVA and Evan’s blue into the tail vein of mice, followed immediately by i.d. injections of rabbit anti-OVA in the dorsal skin, produced a time-dependent increase in edema at the i.d. injection sites. Immune complex-induced edema was apparent as extravasation of the Evan’s blue dye (Fig. 4B, right arrow) and by increases in tissue weight (see below). Little or no edema was apparent following an i.d. injection of vehicle containing an identical concentration of a nonimmune rabbit IgG (Fig. 4A, arrows). The edematous response was specific for immune complexes, as little or no inflammation was observed in the absence of a tail vein injection of OVA (data not shown). Coinjection of anti-OVA Abs with rh-FcRIA abolished extravasation of Evan’s blue dye (Fig. 4B, left arrow). Relative to injection of nonimmune IgG, i.d. injection of anti-OVA Abs produced a >100-fold increase in extravasation of Evan’s blue dye (Fig. 5A), produced a 30-mg increase in tissue weight (Fig. 5B), and increased the tissue content of MPO by 2- to 3-fold (Fig. 5C). Coinjection of anti-OVA with increasing amounts of rh-FcRIA reduced all three measures of inflammation (Fig. 5). The reduction in vascular leak and tissue weight seen with rh-FcRIA was similar to that observed by prior treatment of the animals with dexamethasone (Fig. 5, A and B). Treatment with dexamethasone, however, had no significant effects on the MPO content of the Arthus lesion samples under these conditions (Fig. 5C). In contrast to the reduction in inflammation observed with rh-FcRIA, neither rh-FcRIIIA nor rh-FcRIIA, used over a similar concentration range, reduced anti-OVA induced extravasation of Evan’s blue dye, tissue weight, or tissue MPO activity (Fig. 6). Because local delivery of rh-FcRIA reduced inflammation in the Arthus reaction, this rh-FcR was studied further.
|
|
|
RIA in mice
The data described above demonstrated that, of all the rh-FcR tested by i.d. delivery, only rh-FcRIA reduced inflammation in the murine Arthus reaction. Before assessing the effects of systemically delivered rh-FcRIA, the pharmacokinetics of 125I-rh-FcRIA following a single i.v. dose in mice were evaluated. The serum profile of 125I-rh-FcRIA demonstrated a triexponential decay, with a t1/2 = 0.4 h, t1/2β = 5.9 h, and a t1/2 = 130 h (Fig. 7 and Table I). The serum concentration of 125I-rh-FcRIA fell from 53-µg equivalents/ml at 0.25 h to 3-µg equivalents/ml at 48 h postdose, with the actual and predicted serum concentration of 125I-rh-FcRIA in excellent agreement (Fig. 7, inset). Substantial levels of radioactivity were detectable in serum out to 168 h postdosing, and virtually all of the radioactivity over this time period was protein associated. The steady state volume of distribution was >15 ml, indicating significant distribution of 125I-rh-FcRIA outside the vascular compartment (Table I). At 0.25 h postdosing, radioactivity was highest in the blood > carcass > liver > kidney with little radioactivity found in other tissues (Fig. 8). Recovery of injected radioactivity in these tissues was 100% at the 0.25 h time point.
|
|
|
RIA in the cutaneous Arthus reaction
Intradermal delivery of 13 µg of rh-FcRIA produced maximal inhibition of edema in the cutaneous Arthus reaction (see above). To estimate the amount of rh-FcRIA required to produce a similar effect following systemic dosing, a distribution analysis of 125I-rh-FcRIA was performed in the cutaneous Arthus reaction. To allow for distribution of 125I- rh-FcRIA from the vascular compartment, the Arthus reaction was conducted 1 h following i.v. or s.c. delivery of 125I-rh-FcRIA. In the presence of anti-OVA, vascular leak was increased, and radiolabel accumulation within the lesion sites increased 2- to 3-fold, relative to i.d. injection of nonimmune IgG (data not shown). With either route of administration, the accumulation of 125I-rh-FcRIA within the lesional tissue represented 0.7% of the injected dose. This suggested that i.v. or s.c. delivery of 13 µg/0.007 or 1980 µg of rh-FcRIA would be required to attain an efficacy similar to that observed with i.d. administration.
Systemic delivery of rh-FcRIA blocks inflammation in the cutaneous Arthus reaction
To evaluate whether rh-FcRIA could inhibit cutaneous inflammation following systemic dosing, mice received a single i.v. dose of rh-FcRIA 1 h before initiating the cutaneous Arthus reaction. Systemic dosing of rh-FcRIA produced a concentration-dependent reduction in all three parameters of immune complex-mediated inflammation including extravasation of Evan’s blue dye, tissue weights, and tissue MPO activity (Fig. 9). Under these conditions, the reduction in inflammation seen with the highest dose of rh-FcRIA was similar to that observed with i.d. delivery of 13 µg of rh-FcRIA (Fig. 9). Local or systemic administration of rh-FcRIA can thus block inflammation in the murine Arthus reaction.
|
RIA on joint inflammation in collagen Ab-induced arthritis in mice
The Arthrogen-collagen-induced arthritis Ab mixture is a mixture of four murine mAbs directed against the CB11 cyanogen bromide fragment of type II collagen (6). Three of the four Abs are IgG2a, and the fourth is IgG2b (6). Before evaluating rh-FcRIA in the collagen Ab-induced model of arthritis, the ability of rh-FcRIA to bind to murine total IgG and murine IgG isotypes was assessed using the immune complex precipitation assay. Addition of mouse or human total IgG blocked rh-FcRIA-mediated inhibition of immune complex precipitation in a dose-dependent manner (data not shown). Neither mouse nor human IgG produced significant effects on immune complex precipitation in the absence of rh-FcRIA. Mouse IgG2a and human total IgG blocked the activity of rh-FcRIA with equal potency, whereas neither mouse IgG1 nor IgG2b were active in this regard (data not shown). These data demonstrate that rh-FcRIA binds to total mouse and human IgG and murine IgG2a.
Passive transfer of anti-type II collagen Abs into DBA/1 mice followed 3 days later by LPS treatment induced paw swelling that reached a maximum within 8–10 days (Fig. 10). Subcutaneous injection of rh-FcRIA produced dose-dependent reductions in paw swelling, measured as the clinical paw score (Fig. 10, upper panel) or paw thickness (Fig. 10, lower panel). Reductions in disease severity were evident after one injection of rh-FcRIA. Histologic analysis of the carpal joints of the mice at day 10 revealed extensive inflammation, pannus formation, and joint destruction in animals treated with vehicle alone (Fig. 11, C and D). In contrast, none of the mice that were injected with anti-type II collagen Abs and treated with the highest dose of rh-FcRIA developed significant levels of inflammation, and the cartilage of the carpal joints appeared normal (Fig. 11, A and B). Similar reductions in joint inflammation were observed when dosing with rh-FcRIA was initiated on day 0 (data not shown).
|
|
, GM-CSF, IFN-, IL-10, IL-12, TNF-, IL-15, IL-17, IFN--inducible protein 10, or MCP-1 were noted at this time (data not shown). Treatment with rh-FcRIA produced dose-dependent reductions in the serum concentrations of IL-6 and KC (Table II). The highest dose of rh-FcRIA reduced the serum levels of these inflammatory cytokines to the levels observed in untreated mice (Table II). The serum levels of IL-1β were also reduced 50% by treatment with rh-FcRIA; however, the changes in IL-1β did not reach statistical significance.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
RIA (rh-FcRIA) exhibits anti-inflammatory activities in several models of immune complex-mediated disease. In addition to blocking the precipitation of Ag:Ab complexes in vitro, rh-FcRIA blocked IgG-mediated complement activity, and reduced immune complex-mediated production of inflammatory cytokines by cultured mast cells. These in vitro activities were correlated with efficacy in vivo, as local or systemic delivery of rh-FcRIA reduced vascular leak, neutrophil infiltration, and edema in the cutaneous reverse passive Arthus reaction in mice. Paw inflammation in a murine model of collagen Ab-induced arthritis was also prevented by systemic treatment with rh-FcRIA. Moreover, the latter changes were associated with decreases in the serum levels of inflammatory cytokines.
The rh-FcRIA construct used in the present study is a soluble version of the entire extracellular region of FcRIA, also know as CD64 (8). This region contains three Ig superfamily V-like domains (9), of which the membrane proximal EC2 and EC3 domains appear crucial for high-affinity binding to the Fc domain of monomeric IgG (7, 10). The KD measured for rh-FcRIA binding to human IgG1 was 170 pM, a value similar to that reported previously (11). Analyses of the experimental vs theoretical Rmax levels for binding of rh-FcRIA to human IgG1 and rabbit anti-OVA IgG were in fair agreement, suggesting a 1:1 binding between rh-FcRIA and IgG. These data are consistent with current x-ray crystallographic models depicting a 1:1 interaction between FcR and IgG (7, 10).
With concentrations of IgG in blood in the mg/ml range and with rh-FcRIA displaying a KD for IgG1 in the sub nanomolar range, rh-FcRIA is expected to be saturated with monomeric IgG in vivo. This notion is partially consistent with the pharmacokinetic studies reported herein (Fig. 7), where the elimination half-life of rh-FcRIA in the mouse was estimated at 130 h, a value approximating the half-life of IgG in the mouse (12, 13). However, a significant amount of 125I-rh-FcRIA appeared to distribute out of the vascular compartment at early time points, suggesting that not all of the soluble receptor was sequestered by circulating IgG. The large Vss of 125I-rh-FcRIA, approximately seven times the murine blood volume (Table I), supports this hypothesis. Our current lack of direct information on whether rh-FcRIA circulates within the blood of mice as a complex with IgG, however, limits this interpretation. Mouse IgG appears to bind rh-FcRIA, as total mouse IgG was an effective competitor for rh-FcRIA-mediated inhibition of immune complex precipitation. Moreover, the Ig isotype specificity of these effects, where IgG2a was an effective inhibitor while IgG1 and IgG2b were not, is consistent with the Ig isotype-binding specificity of endogenous mouse FcRI (2).
Although rh-FcRIA may be bound by monomeric IgG within the circulation, this may not limit its effectiveness as an inhibitor of inflammation. Whether the greater efficacy of rh-FcRIA is due to an increased affinity of the soluble receptor for immune complexes or to other pharmacokinetic or cellular parameters (14, 15) is not yet clear. Indeed, the current data shown in Figs. 9–11 demonstrate that systemic administration of rh-FcRIA can block inflammation in two models of immune complex-mediated disease. These data are consistent with the results of gene targeting studies, where FcgrI–/– mice exhibit a phenotype of decreased severity of arthritis and cartilage damage, altered hypersensitivity responses, enhancement of Ab responses, and impaired antibacterial responses (14, 15). These data suggest that in the presence of saturating amounts of IgG, either rh-FcRIA or endogenous murine FcRI can effectively bind Ag/Ab immune complexes and modulate inflammatory responses. Although the data in the present manuscript demonstrate that rh-FcRIA can effectively limit immune complex-mediated inflammation in vivo, it is possible that rh-FcRIA could also negatively impact the response to infection. In this regard, although specific IgG titers against Bordetella pertussis were not different between wild-type and FcrIa–/– mice, the clearance of pathogen from the lungs of FcrIa–/– mice was impaired (15).
There are several nonexclusive mechanisms that may account for the reduction in inflammation observed with rh-FcRIA in the Arthus reaction (Figs. 4, 5, and 9) and collagen Ab-induced arthritis (Figs. 10 and 11). Based on the in vitro activities of rh-FcRIA described above, these mechanisms may include inhibition of immune complex precipitation (Fig. 1), competition with cellular FcRs for binding to immune complexes (Fig. 2), and/or inhibition of complement activation (Fig. 3). The precipitation of Ag:Ab complexes is mediated by noncovalent interactions between the Fc domains of neighboring Ab molecules (16). Inhibition of immune complex precipitation by rh-FcRIA is likely mediated by rh-FcRIA binding to the Fc domain of anti-OVA IgG which prevents Fc-Fc interactions between IgGs. These data are similar to those described previously for a recombinant soluble version of human FcRIIA (4). Mast cells play a key role in mediating inflammation in cutaneous type III hypersensitivity reactions (17, 18, 19). In the cutaneous Arthus reaction in mice, immune complexes elicit inflammation by binding to mast cell FcR, with the subsequent production of inflammatory cytokines, neutrophil infiltration, and tissue damage. Both immune complex-mediated cytokine production from murine mast cells (Fig. 2) and inflammation in the cutaneous Arthus reaction (Figs. 4, 5, and 9) were suppressed by rh-FcRIA, data supportive of a common mechanism for these effects. In human rheumatoid arthritis (20) and in animal models of arthritis (21, 22), immune complexes can induce the production of proinflammatory cytokines through their interaction with cellular FcR (17, 23). The reduction in serum concentrations of IL-6, Il-1β, and KC shown in Table II and prevention of paw inflammation in collagen Ab-induced arthritis (Figs. 10 and 11) support the notion that rh-FcRIA blocks inflammation by interference with binding of immune complexes to cell surface FcR. Soluble versions of human FcRIIA have been shown to inhibit rosette formation and Fc-dependent activation of platelets by blocking the interaction of the Ig Fc domain with cellular FcR (24, 25, 26). Lastly, the potent blockade of complement-mediated lysis of Ab-sensitized erythrocytes shown in Fig. 3 suggests that rh-FcRIA may block inflammation by preventing complement activation. The latter activity may be due to steric hindrance of C1q binding to the Ig Fc domain by rh-FcRIA. The binding sites for FcR and C1q appear to be closely apposed on the lower hinge and upper Cl-type Ig (CH2) domain 2 of Ig Fc (7, 10). Thus, while the anti-inflammatory properties of rh-FcRIA may be due to inhibition of a diverse set of pathways, the common element linking these pathways is the binding of rh-FcRIA to the Fc domain of IgG.
Efficacy of rh-FcRIA in the Arthus reaction and in collagen Ab-induced arthritis in the present report was observed at somewhat high concentrations (33–100 mg/kg) of rh-FcRIA. This could be due to the large amounts of specific Abs used to generate consistent inflammation in these models. For the cutaneous Arthus reaction, for example, each animal received 40 µg of anti-OVA Ab by i.d. injection, and in collagen Ab-induced arthritis, each mouse received an i.v. injection of 2 mg of anti-type II collagen Abs (10% of the total IgG in mouse serum). Maximal inhibition of immune complex precipitation (Fig. 1) and maximal efficacy of local delivery of rh-FcRIA in the Arthus reaction (Fig. 5) was seen at a molar ratio of rh-FcRIA to anti-OVA Ab of 1:1. Thus, neutralization of a large amount of immune complexes may require equimolar amounts of rh-FcRIA. Extrapolation of the stoichiometry established in vitro to that expected in the in vivo models is difficult because the amount of immune complexes present in vivo is not known. In contrast, because rh-FcRIA appears to bind mouse IgG2a, large amounts of rh-FcRIA may be required to saturate the endogenous pool of mouse IgG. The pharmacokinetic profile of 125I-rh-FcRIA in mouse blood shown in Fig. 7 suggests rh-FcRIA binding to IgG and that IgG in the blood may act as a carrier for rh-FcRIA thereby prolonging its biological half-life. These interpretations are only speculative, however, because a single dosing schedule was used in the collagen-Ab-induced arthritis model, and the bioavailability of rh-FcRIA following s.c. dosing is not yet known. The relative efficacy of rh-FcRIA can thus only be established through additional experimentation.
All three of the rh-FcRs used in the present study bound human IgG1 with the expected affinities, inhibited immune complex precipitation, blocked immune complex-mediated cytokine secretion from mast cells, and blocked complement activation in vitro. These data are similar to those reported previously where recombinant soluble versions of FcRIIA exhibited a range of activities in vitro including inhibition of immune complex precipitation, blockade of immune complex binding to B cells, inhibition of rosette formation, Ab-induced platelet activation, and inhibition of rheumatoid factor binding to immune complexes (24, 25, 26, 27, 28, 29). However, as shown in Fig. 6, neither rh-FcRIIA nor rh-FcRIIIA reduced vascular leak, edema, or neutrophil infiltration following their local delivery in the cutaneous Arthus reaction in mice. These data seem to contradict a previous report (24) where a recombinant soluble version of FcRIIA blocked inflammation in the cutaneous Arthus reaction in rats. The reason for the discrepancy between that report and the current study is not clear, but is unlikely to be due to interspecies differences, as the rh-FcRIIA used in the present study was also inactive when tested in a cutaneous Arthus reaction in rats (M. Maurer and J. L. Ellsworth, unpublished information). The lack of efficacy of rh-FcRIIA and rh-FcRIIIA may be due to their low affinity for IgG, which limits their competition with cellular FcRs for binding of immune complexes. The 100-fold difference in relative potencies between rh-FcRIA and rh-FcRIIA for blockade of immune complex-mediated cytokine secretion from mast cells is consistent with this argument.
The abnormal production or clearance of immune complexes from the blood results in a wide array of inflammatory diseases that arise from deposition of immune complexes within tissues (30). These diseases include the connective tissue autoimmune diseases such as systemic lupus erythematosus and rheumatoid arthritis, and a host of other diseases of diverse etiologies. The ability of rh-FcRIA to block inflammatory cytokine production, prevent neutrophil infiltration, and prevent cartilage damage in a model of arthritis suggests that rh-FcRIA, a high-affinity receptor for IgG, may be a useful therapeutic agent for treatment of immune complex-mediated diseases.
|
Acknowledgments |
|---|
|
Disclosures |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
|
Footnotes |
|---|
1 This work was supported in its entirety by ZymoGenetics. 
2 Address correspondence and reprint requests to Dr. Jeff L. Ellsworth, Department of Autoimmunity and Inflammation, ZymoGenetics, 1201 Eastlake Avenue East, Seattle, WA 98102. E-mail address: jefe{at}zgi.com
3 Abbreviations used in this paper: rh, recombinant human; CHO, Chinese hamster ovary; i.d., intradermal; rm, recombinant murine; MPO, myeloperoxidase; TACI, transmembrane activator and CAML interactor.
Received for publication July 11, 2007. Accepted for publication October 25, 2007.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
Rs in autoimmune disorders. Springer Semin. Immunopathol. 28: 305-319. [Medline]RIIA and IIIA polymorphisms in autoimmune diseases. Biomed. Pharmacother. 58: 286-291. [Medline]RII inhibits immune complex precipitation. Clin. Exp. Immunol. 102: 620-625. [Medline]R: current models. Immunol. Lett. 82: 57-65. [Medline] receptor I with picomolar affinity for immunoglobulin G. Biochem. Biophys. Res. Commun. 338: 1811-1817. [Medline]RI-deficient mice show multiple alterations to inflammatory and immune responses. Immunity 16: 379-389. [Medline]RI (CD64) contributes substantially to severity of arthritis, hypersensitivity responses, and protection from bacterial infection. Immunity 16: 391-402. [Medline]RII: production, characterization, and inhibition of the Arthus reaction. J. Exp. Med. 178: 1617-1628. RIIa2 by activated platelets and inhibition of anti-CD9-mediated platelet aggregation by recombinant FcRIIa2. Blood 85: 698-704. RII/CD32) that inhibits the binding of immune complexes to FcR+ cells. J. Immunol. 152: 201-212. [Abstract]RIIa inhibits rheumatoid factor binding to immune complexes. Immunology 109: 246-254. [Medline] receptor II (CD32) isoforms produced in insect cells. Biochemistry 38: 8469-8477. [Medline]This article has been cited by other articles:
|
|
 |
|
  J. L. Ellsworth, N. Hamacher, B. Harder, K. Bannink, T. R. Bukowski, K. Byrnes-Blake, S. Underwood, C. Oliver, K. S. Waggie, C. Noriega, et al. Recombinant Soluble Human Fc{gamma}R1A (CD64A) Reduces Inflammation in Murine Collagen-Induced Arthritis J. Immunol., June 1, 2009; 182(11): 7272 - 7279. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
), or rh-Fc
). Each point represents the mean ± SEM of three separate experiments performed in duplicate. The data are expressed as a percent of the values obtained in the absence of rh-Fc
). Hemolysis of SRBCs was assessed as described in Materials and Methods. Each point represents the average of duplicate determinations from two separate experiments.
) radioactivity was assessed as described in Materials and Methods. Each point represents the mean ± SEM for n = 3 mice. Inset, The actual (•) and predicted (solid line) serum concentration of 125I-rh-Fc
