A Codominant Role of FcγRI/III and C5aR in the Reverse Arthus Reaction

Mice

FcγRIII-deficient mice were generated as decribed (17). They were bred for eight generations onto C57BL/6 mice under pathogen-free conditions in the animal facility of Hannover Medical School. The homozygous FcγRIII^−/− mice were selected and WT FcγRIII^+/+ C57BL/6 littermates were used for all comparisons. C57BL/6 mice homozygous for FcRγ^−/− were obtained from Taconic (Germantown, NY). All these mice were male and were used at 8–12 wk of age. Experiments were conducted in accordance with the regulations of the local authorities.

Reverse passive Arthus reaction in skin and lung

Mice were anesthetized with ketamine and xylazine and shaved at their basolateral sides. Rabbit IgG anti-OVA Ab (30 μg; Sigma, Munich, Germany) was injected intradermally at multiple sites. In addition, the trachea was cannulated and 150 μg Ab was applied. Immediately thereafter, 200 μl of 0.25% Evans blue together with 20 mg/kg of OVA Ag were given i.v. Where indicated, mice were injected twice with 4.25 μg purified cobra venom factor (CVF) (Naja naja, Calbiochem-Novabiochem, Bad Soden, Germany) i.p. at 24 and 16 h before the Arthus reaction to deplete complement. Serum complement levels were determined as described (21). In additional experiments, mice received C5aRA ΔpIIIA8 to inhibit C5aR-triggered activation (10). Hereby, 200 μl antagonist was given at a concentration of 7.3 × 10⁻⁶ M before application of anti-OVA IgG, and then 100 μl antagonist was given at 60 and 120 min after IC challenge. Mice were killed 4 h after initiation of the Arthus reaction and were assayed for infiltration of PMN, hemorrhage, and plasma exudation in skin, lavaged lung tissue, and bronchoalveolar lavage (BAL) fluid.

BAL and quantitation of hemorrhage and PMN accumulation in bronchoalveolar space

Pulmonary vasculature was gently flushed with PBS with a catheter positioned in the root pulmonary artery. Lungs were lavaged with PBS (1 ml, five times at 4°C) after cannulation of the trachea. The volume of collected BAL fluid was measured in each sample and total cell count was assessed with a hemocytometer (Neubauer Zählkammer, Gehrden, Germany). The amount of erythrocytes represented the degree of hemorrhage. For quantitation of PMN accumulation, differential cell counts were performed on cytospins (10 min at 55 × g) stained with May-Grünwald/Giemsa using 300 μl of BAL fluids.

Myeloperoxidase (MPO) assay in skin and lung

Skin punches of the injection sites 1 cm in diameter, lavaged lung tissues, and BAL fluids were assessed for PMN accumulation by MPO activity as described (22). Briefly, homogenized tissue was suspended in 50 mM potassium phosphate buffer (pH 6.0, 0.5% hexadecyltrimethylammoniumbromide). Cells were broken by three cycles of freezing and thawing and subsequent sonication. The supernatant was mixed with 0.167 mg/ml o-dianisidine dihydrochloride (Sigma) and 0.0005% hydrogen peroxide. The amount of MPO was calculated by assessing the change in absorbance at 450 nm. A serial dilution of MPO from human PMN (Calbiochem-Novabiochem) served as standard. Samples were run in duplicate.

Measurement of plasma exudation in skin and lung

Evans blue dye binds avidly to albumin and serves as a marker of extravasation of plasma proteins into skin and lung. This technology compares favorably with the methodology involving radiolabeled albumin (23). One-centimeter skin punches were harvested, placed in 1 ml of formamide, sonicated for 10 min, and incubated at 37°C for 16 h. The concentration of dye in appropriate dilutions of skin eluate, BAL fluid, and serum was measured spectrophotometrically at 620 nm. Serum samples were also measured at 405 nm to account for hemolysis. Extinction at 405 nm was multiplied with factor 0.014 and deducted from the extinction at 620 nm. This factor corrects for extinction caused by murine hemoglobin at 620 nm. Plasma exudation was quantitated by multiplying the ratio of extinction in BAL fluid or skin punches to extinction in plasma by the dilutions of the specimens.

Histologic and immunohistochemistry studies

To process lung tissues for histological examination after lavage, they were fixated in 4% buffered paraformaldehyde, embedded in paraffin, and stained with hematoxylin and eosin according to conventional procedures. In additional experiments, lavaged lung tissues were prepared for immunohistochemistry. The alkaline phosphatase antialkaline phosphatase (APAAP) technique was used to phenotype neutrophils (Gr1; Dianova, Hamburg, Germany). Cryostat sections were incubated with the primary Abs for 30 min at 21°C. After washing with TBS-Tween (0.05% Tween 20; Serva, Heidelberg, Germany), incubations were done with the bridging Ab Z0494 and the rat-APAAP Ab complex D0488 for 30 min. To increase the staining intensity, the last two steps were repeated once. Fast blue (Sigma) served as substrate for alkaline phosphatase. Positive and negative controls produced the expected results.

Statistical analysis

Statistical analysis was performed using the SPSS v. 8.0 statistical package (SPSS, Chicago, IL). To analyze differences of mean values between groups, a two-sided unpaired Student‘s t test was used; p < 0.05 was considered significant and p < 0.001 was considered highly significant. Independent contribution of FcγR and C5aR was assessed by two-way, two-factorial ANOVA (univariate general linear model procedure). Function or dysfunction of either receptor was coded in a binary variable (factor). Between-subject effects indicated independent contribution of a factor.

Arthus reaction in the skin

Reverse passive Arthus reactions in the skin were performed to dissect the relative contribution of FcγRI, FcγRIII, and complement during cutaneous, IC-triggered inflammation. C57BL/6 mice deficient in the ligand-binding α-chain of FcγRIII (FcγRIII^−/−) or deficient in the FcRγ-chain critical for signaling of both FcγRIII and FcγRI (FcRγ^−/−) were compared with WT C57BL/6 mice. Inflammatory responses were quantitated by measurement of MPO in homogenized tissue as a marker of PMN infiltration and by assessment of Evans blue in tissue extracted by formamide, indicating the degree of microvascular permeability. In WT mice, a strong accumulation of PMN with a concomitant increase of plasma exudation was observed within 4 h after IC challenge (Fig. 1⇓, A and B). In FcγRIII^−/− mice, recruitment of PMN was markedly decreased and comparable to that seen in FcRγ^−/− mice (Fig. 1⇓A). Cutaneous plasma exudation differed in that the attenuation observed in FcγRIII^−/− was further reduced to background levels in FcRγ^−/− mice (Fig. 1⇓B). These data identify FcγRIII as the dominant FcR in PMN influx and suggest that both FcγRI and FcγRIII can trigger enhanced vascular permeability.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1.

Passive reverse cutaneous Arthus reaction in WT, FcγRIII^−/−, and FcRγ^−/− mice receiving CVF and C5aRA. Mice were injected intracutaneously with 30 μg of IgG anti-OVA Ab and then with systemic 20 mg/kg OVA Ag and 0.25% Evans blue. Recruitment of PMN (A) and plasma exudation (B) were assessed 4 h after initiation of the cutaneous Arthus reaction by MPO assay or formamide extraction of extravasated blue dye, respectively (IC, ▪). Treatments with CVF and C5aRA are indicated (IC + CVF, ▦; and IC + C5aRA, ▨). Mice receiving only Ag or Ab served as controls (Ag, □; and Ab, dotted bars). Ag control group and Ab control group each comprised five WT mice per group, and IC treatment groups comprised eight to twelve mice per group (except for IC + C5aRA, which had five to six mice per group). Data are expressed as mean ± SEM. Differences for both parameters were significant or highly significant for the IC treatment groups of WT mice compared with FcγRIII^−/− and FcRγ^−/− mice (∗, p < 0.05 to p < 0.001), and were significant for IC compared with IC + CVF and IC + C5aRA treatment groups (†, p < 0.05). In addition, FcγRIII^−/− and FcRγ^−/− mice differed significantly for plasma exudation but not PMN infiltration (‡, p < 0.05).

Recently, we verified the proposed role of complement (24) in addition to FcγR in cutaneous Arthus reaction. This was based on the observation that complement depletion with CVF completely abrogated the inflammatory response, as defined by Evans blue extravasation, only in FcγRIII^−/− mice and not in control mice (17). The FcγRIII^−/− mice and their normal littermates used in that study had a mixed genetic background of C57BL/6 and 129 strains. Interestingly, although rather ineffective in mixed C57BL/6 and 129 strains, the treatment with CVF resulted in a significant attenuation of inflammation in WT mice of C57BL/6 strain. By using two i.p. injections of 4.25 μg CVF 24 and 16 h before IC challenge, a strong reduction of PMN influx and a complete suppression of plasma exudation were observed (Fig. 1⇑, A and B). This finding illustrates that variations in the genetic background of C57BL/6 and 129 mice are of strong influence of the inflammatory response to the Arthus reaction. In support, the hemolytic activity is significantly lower in C57BL/6 mice compared with that in 129 mice (17), which may indicate that strain-specific differences in complement activity can contribute to the observed heterogeneity. The importance of complement was further confirmed with the specific C5aRA ΔpIIIA8 (10). In WT mice, MPO activity and plasma exudation decreased by more than 60 and 80%, respectively, after C5aRA treatment (Fig. 1⇑, A and B). This indicates a critical role for C5a, as recently suggested in C5aR^−/− mice (9). A complete suppression for both parameters was similarly observed in FcγRIII^−/− mice and in FcRγ^−/− mice after inhibition of C5aR-triggered activation by C5aRA (Fig. 1⇑, A and B). Taken together, these results demonstrate that FcγR and C5aR are codominant receptor pathways in the initiation of the inflammatory cascade in skin of C57BL/6 mice.

Arthus reaction in the lung

It is still unclear whether the pathogenesis of the inflammatory response in IC diseases follows a general pattern (13) or if tissue-specific differences exist (9). Thus, the immunopathological cascade was induced not only in the skin, but also simultaneously in the lung in the same animals, allowing for an assessment of organ-specific differences under exactly the same experimental conditions. In WT mice, the profound pulmonary IC inflammatory response with enhanced PMN infiltration, plasma exudation, and hemorrhage into the bronchoalveolar space was markedly suppressed, although not completely abolished, by application of either CVF or C5aRA (Fig. 2⇓A, C, and D). Interestingly, interstitial PMN influx, as defined by MPO activity in lavaged lung tissue, was substantially attenuated after C5aRA treatment, whereas systemic complememt depletion by CVF resulted in increased MPO activity (Fig. 2⇓B). This finding is in accordance with the recognized ability of CVF to induce accumulation of PMN, especially in the lung (25).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 2.

Pulmonary IC inflammation in WT mice after CVF and C5aRA treatment. The induction of the inflammatory response in the lung was performed by challenge with 150 μg anti-OVA Ab intratracheally and then with OVA Ag in WT mice (IC, ▪), WT mice treated with CVF (IC + CVF, ▦), and C5aRA (IC + C5aRA, ▨). Mice receiving only Ag or Ab served as controls (Ag, □; and Ab, dotted bars). After 4 h, mice were killed and PMN infiltration in the alveolar space (A), assessed by differential cell counts in Giemsa stains of BAL fluid, PMN infiltration in lung tissue (B), measured by MPO activity in lavaged lung, hemorrhage (C), measured by total cell count of erythrocytes present in BAL fluid, and plasma exudation (D), quantitated by the amount of Evans blue dye extravasation in BAL fluid, were evaluated. Ag and Ab control groups comprised 5 animals per group, and IC treatment groups comprised 8–12 animals per group (except for C5aRA, which had five to six animals per group). Data are presented as mean ± SEM. Differences in IC compared with IC + CVF and IC + C5aRA treatment groups were significant or highly significant for all parameters (∗, p < 0.05 to p < 0.001).

Given the primary role of C5a among complement mediators in both skin and lung, the contribution of FcγRI and FcγRIII to the pathophysiology in the lung was further evaluated. In both FcγRIII^−/− and FcRγ^−/− mice, alveolar PMN infiltration as well as interstitial PMN accumulation in lavaged lung tissue was substantially decreased (Fig. 3⇓, A and B). The reduction in hemorrhage seen in FcγRIII^−/− mice was further attenuated in FcRγ^−/− mice (Fig. 3⇓C). This result shows that both FcγRI and FcγRIII contribute significantly to this parameter. Unexpectedly, plasma exudation in the lung was largely independent of FcγRI and FcγRIII (Fig. 3⇓D), which is different than what has been observed in the skin (Fig. 1⇑B). This finding was confirmed by histological studies demonstrating marked perivascular edemas in WT mice (Fig. 4⇓A) which were also evident in FcγRIII^−/− mice (Fig. 4⇓B) and FcRγ^−/− mice (Fig. 4⇓C). The residual proportion of hemorrhage and plasma exudation seen in FcγRIII-deficient mice was completely abrogated after C5aRA treatment (Fig. 3⇓, C and D). In contrast, interstitial MPO activity of FcγRIII^−/− mice and FcRγ^−/− mice was not affected, and alveolar PMN influx was totally absent after inhibition of C5aR (Fig. 3⇓, A and B). As shown by immunohistochemistry, Gr1-positive PMN in C5aRA-treated FcRγ^−/− mice retarded in the pulmonary vasculature, apparently unable to migrate along a chemotactic gradient into the alveolar space (Fig. 4⇓D).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 3.

Pulmonary IC inflammation in WT, FcγRIII^−/−, and FcRγ^−/− mice receiving C5aRA. Inflammation in the indicated mice (IC, ▪) was induced and analyzed for PMN infiltration in alveoli (A), MPO activity in lavaged lung tissue (B), alveolar hemorrhage (C), and alveolar plasma exudation (D) as described in Fig. 2⇑. Where indicated, mice were additionally treated with 3× C5aRA in a total volume of 400 μl at a concentration of 7.3 × 10⁻⁶ M (IC + C5aRA, ▨). WT mice receiving PBS instead of Ag (not shown) or Ab served as controls (Ab, dotted bars). Ab control groups comprised 5 animals per group, IC treatment groups comprised 8–12 animals per group, and C5aRA-treated mice comprised five to six animals per group. Data are presented as mean ± SEM. Differences for hemorrhage, and alveolar and interstitial PMN infiltration were significant for the IC treatment groups of WT mice compared with FcγRIII^−/−, and FcRγ^−/− mice (∗, p < 0.05), whereas the decrease in plasma exudation was not significant with p = 0.274 and p = 0.097, respectively. FcγRIII^−/− and FcRγ^−/− mice only differed significantly for hemorrhage (‡, p < 0.05). Animals treated with C5aRA differed significantly compared with untreated mice of the same genotype in alveolar PMN infiltration, alveolar plasma exudation, and pulmonary hemorrhage (†, p < 0.05) with the exception of hemorrhage in FcRγ^−/− mice (p = 0.202). No differences were observed for MPO activity in lavaged lung tissue. As in the skin, Ag and Ab control values did not differ between WT and FcR^−/− mice (data not shown).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 4.

Histopathology of lung from FcγR-deficient mice. A–C, Representative hematoxylin and eosin-stained sections of lungs from WT (A), FcγRIII^−/− (B), and FcRγ^−/− (C) mice after induction of IC inflammation as described in Fig. 2⇑ (v, vessel; b, bronchiolus; and e, edema). Marked perivascular edemas around pulmonary vessels are present in all mouse strains analyzed. D, Representative immunostained section of lungs from FcRγ^−/− mice treated with C5aRA. GR1-positive PMN (blue) are detectable sticking at the vascular wall (v). As assessed by two blinded reviewers in three to five mice per C5aRA treatment group, this finding could be obtained in both FcRγ^−/− (D) and FcγRIII^−/− (not shown) mice, but not in WT mice (not shown).

Together, the results give strong evidence that in C57BL/6 mice FcγR and C5aR are codominant receptors in the initiation of the Arthus response in both skin and lung and further suggest that the sequelae of IC-triggered inflammation are not necessarily the endpoints of the same reaction. Tissue site-specific differences exist with respect to plasma exudation, which is almost entirely dependent on C5aR in the lung and on both C5aR and FcγR in the skin. FcγRIII is the dominant FcγR for cutaneous and alveolar PMN influx. Interestingly, FcγRI is of additional importance for plasma exudation in the skin and hemorrhage in the lung. These observations, combined with the recent report that FcγRI contributes significantly to enhanced PMN infiltration in IC-induced peritonitis (20), indicate that effector cells expressing FcγRI are involved in IC-triggered inflammation. In the lung, the alveolar macrophage expressing FcγRI, FcγRII, and FcγRIII is the most prominent cell type in the alveoli and is known to promote IC injury through the production of cytokines (TNF-α, platelet-activating factor, etc.), chemokines (macrophage-inflammatory protein-2, cytokine-induced neutrophil chemoattractant-1, etc.), and vasoactive substances (reviewed in Ref. 26). In the skin, resident mast cells, epidermal Langerhans cells (both of which express FcγRIII), and macrophages are the most likely involved cell types (15, 27). These mast cells have already been demonstrated to contribute significantly to IC-induced injury in skin (15, 28). However, because FcγRIII, but not FcγRI, is expressed on mast cells, it appears that skin macrophages are of additional importance.

One might assume that FcγR and complement represent a redundant system, leading to less attenuation by dysfunction of one pathway with the remaining pathway partly substituting the response. However, as shown in this study, the sum of attenuations caused by FcγR deletion and C5aR blockade is far above 100% in all parameters analyzed. Therefore, both pathways may in part act independently, but also have to act in a cascade of events with one pathway dependent on triggering by the other as first suggested by Sylvestre and Ravetch (14). Action independent from the other pathway can be derived from the inflammatory response present with one pathway functional and the other dysfunctional. This has been formally assessed by ANOVA (p values for between-subject effects of the factors FcγR and C5aR, respectively: alveolar PMN infiltration, 0.007, 0.002; pulmonary interstitial PMN infiltration, 0.026, 0.307; alveolar hemorrhage, 0.003, 0.058; pulmonary plasma exudation, 0.196, 0.002; cutaneaous PMN infiltration, 0.007, 0.002; and cutaneous plasma exudation, 0.017, 0.15). This demonstrates that the FcγR and C5aR pathways contribute significantly to the inflammatory response, with the exceptions of C5aR in interstitial PMN influx and FcγR in pulmonary plasma exudation. On the other hand, the sum of the independent contributions obtained by adding the mean values of single groups does not reach the full strength of the inflammatory response as seen in WT mice. This fact suggests a dependent interaction between FcγR and C5aR which may explain the controversy about their relative roles (9, 13).

In summary, we have used FcγRIII- and FcRγ-deficient mice in combination with a specific antagonist against C5aR to distinguish between FcγRI-, FcγRIII-, and C5aR-mediated effects. This approach enabled us to demonstrate that the initiation of the inflammatory cascade in IC disease is mainly determined through the action of FcγRIII and C5aR pathways. Furthermore, the comparison between FcγRIII^−/− and FcRγ^−/− mice shows that FcγRI plays an additional role. The earlier observations of C5aR^−/− mice (9) suggested a strict tissue-site dependency of C5aR. We observed minor differences for both FcγR and C5aR, with the exception that vascular permeability was more prominently mediated by C5aR in lung and by FcγR in skin. In addition, our findings not only support the current idea that FcγRIII is a dominant receptor in acute inflammation (17) but also extend this concept by integrating the interplay between FcγRI, FcγRIII, and C5aR with dependent and independent actions. Finally, the data strengthen the view that both FcγR and C5aR have to be considered as potential targets in immunotherapy of IC disease in humans.