Results

Drug design strategy

We have previously used X-ray crystallography to determine the 2.0-Å resolution structure of the H134 allelic form of FcRIIa.²³ A dimer of receptor monomers was formed around the unique two-fold crystallographic axis of the orthorhombic P2₁2₁2 crystals (Figure 1a). The dimer interface contained a large solvent-filled groove that partially overlapped, and was flanked by, the IgG-binding sites of the two FcRIIa monomers (Figures 1b and c). Thus, molecules designed to bind across this groove (that is, target site) may inhibit the binding of immune complexes and affect the initiation of downstream signaling events that lead to inflammation.

Figure 1.

Target site used for design of small chemical entity (SCE). Solvent-accessible surface views of the FcRIIa dimer showing the target site used for design of SCE. (a) Side view with the IgG-binding sites at the top of the FcRIIa dimer; (b) view looking down onto the IgG-binding site and the target site and (c) close-up view of the groove-shaped target site. Key target residues are F132 (yellow), H134 (green) and K120 (red). Other residues lining the groove are: T122, F124, T155, N157, L162 and S164 (dark grey). A prominent IgG-binding site residue is highlighted (Y160), although F132 and K120 also directly participate in binding IgG. (d) Docking of VIB153, (e) docking of VIB384 and (f) docking of VIB113.

Full figure and legend (276K)

A series of compounds were designed to take advantage of three prominent features of the target site. First, the phenyl rings of F132 lying flat against walls of the groove and at either end of the groove, separated by a distance of approximately 10 Å (1 nm), were used to design SCE containing planar ring systems, capable of participating in – stacking interactions with F132. These hydrophobic planar rings were joined by linkers that could be easily manipulated to increase diversity and vary flexibility while maintaining an optimal distance between hydrophobic moieties.

The second prominent attribute of the target site used was the position of the side chains of K120, located between the two F132 residues and overhanging the central part of the groove. Thus, SCEs featured groups capable of participating in electrostatic interactions with the K120 side chains. These were: (i) acidic moieties such as carboxylic acids for charge neutralization (salt bridge formation) or (ii) polar groups capable of hydrogen bonding with the positively charged amino groups.

Third, the floor of the target site was formed by hydrophobic and uncharged polar groups provided by the side chains of the residues T122, F124, N157, L162 and S164 and these determined placement of the SCE within the groove. Examples of the positions of three of the active the compounds (VIB153, 384 and 113) within the groove are shown (Figures 1d–f, respectively). Using this detailed knowledge of the three-dimensional structure and chemical attributes of the target site, more than 100 compounds were synthesized as sodium salts and screened for inhibition of binding of FcRIIa to immune complexes. As the SCEs were designed to fit within the groove of the FcRIIa dimer,²² and the His/Arg polymorphic side chain is located outside the groove (Figure 1a), this polymorphism is unlikely to influence the binding of SCE. Herein, we present results for the six most active compounds that were tested in both in vitro and in vivo assays for inhibition of FcRIIa-dependent activity (Table 1).

Table 1 - Compounds showing activity in vitro.

Full table (35K)

Strategy for screening SCE in vitro

To select compounds for in vivo testing, SCEs were initially subjected to an in vitro screening to determine the efficiency and specificity of action. Thus, SCEs were selected after meeting the following criteria: (i) inhibition of FcRIIa-dependent platelet activation, (ii) inhibition of platelet aggregation and (iii) failure to inhibit arachidonic acid-induced aggregation of platelets. SCEs selected by platelet assays were also tested for inhibition of heat-aggregated gamma-globulin (HAGG)-induced TNF secretion.

SCE inhibition of IgG-dependent platelet activation and aggregation

Human platelets that are activated by immune complexes express P-selectin, so SCEs were first screened at a high concentration (500 g ml^-1) for inhibition of immune complex-induced expression of P-selectin (Figure 2a). Compounds showing >50% inhibition in three separate platelet tests (from different donors) were then tested for their ability to inhibit immune complex-induced (HAGG) platelet aggregation and for specificity by comparing their effects on other activation pathways, including ADP and arachidonic acid.²⁴ Platelets were collected from different donors each week. There was no difference in activity when SCEs were tested on platelets of Arg/Arg, His/His or Arg/His phenotype (data not shown). Table 1 shows the IC₅₀ results for the fluorescence-activated cell sorting (FACS) tests. All of the selected active SCEs showed 100% inhibition of HAGG-induced aggregation at 500 g ml^-1, with no inhibition of ADP-induced aggregation. Examples of aggregation assays are shown in Figures 2b and c. VIB153 completely inhibited immune complex (HAGG)-dependent platelet aggregation in a dose-dependent manner and specificity was demonstrated by the lack of inhibition of aggregation mediated by ADP. Platelets aggregated normally after subsequent addition of arachidonic acid.

Figure 2.

Inhibition of platelet activation and aggregation by small chemical entities (SCEs). (a) Inhibition of heat-aggregated gamma-globulin (HAGG)-induced P-selectin expression on human platelets as measured by fluorescence-activated cell sorting (FACS): SCE (500 g ml^-1), means.d. ^*Significant inhibition compared with control compound BRI6801, P<0.05 by analysis of variance (ANOVA). (b and c) Representative experiments (of n=6) showing (b) inhibition of HAGG-mediated platelet aggregation by VIB153 at 500, 50 and 5 g ml^-1. Specificity was demonstrated by the lack of inhibition of arachidonic acid-induced aggregation with VIB153 at 500 g ml^-1. (c) VIB153 tested at 500 g ml^-1 with HAGG-induced platelet aggregation, and specificity and lack of toxicity of VIB153 demonstrated by the lack of inhibition of ADP-induced aggregation.

Full figure and legend (101K)

SCE inhibition of TNF- secretion

Although TNF- secretion is not specific for arthritis induction, TNF is a potent, clinically relevant inflammatory mediator in the development of RA, and immune complexes are known to induce TNF secretion.¹⁸ The four most active SCEs, selected by platelet assays, were tested for their capacity to inhibit immune complex-dependent TNF- secretion from human and transgenic mouse macrophages. Using a differentiated human monocytic cell line, U937 that expresses the R134 allelic form of the FcR,²⁵ three compounds, VIB153, 384 and 294 showed dose-dependent inhibition of TNF- secretion (Figure 3a), with activity similar to the anti-FcRIIa antibody IV.3 (71–78% inhibition at the highest dose). VIB197, although inhibiting immune complex-dependent platelet aggregation, was a poor inhibitor of TNF- secretion (Figure 3a). Table 1 also shows the IC₅₀ for the compounds in the TNF- tests with U937 cells. Further analysis of VIB153 and 384 (375 g ml^-1) using HAGG stimulated peritoneal macrophages from R134-expressing huFcRIIa transgenic mice,¹⁵ compared with wild-type C57BL/6 mice, showed that both SCEs inhibited TNF- (Figure 3b; 70% inhibition, P<0.05 compared with untreated cells). Less inhibition (51%) was seen with C57BL/6 macrophages (P<0.5). Thus, SCE had a general anti-inflammatory effect in vitro, possibly acting through other FcRs on both strains of mice. These cross-reactions were not seen in vivo (see below).

Figure 3.

Effect of small chemical entity (SCE) on tumor necrosis factor (TNF)- release in vitro and blood counts in vivo. (a) Heat-aggregated gamma-globulin (HAGG) (250 g ml^-1) stimulated TNF- release from U937 cells following incubation with VIB compounds or IV.3 antibody compared with untreated controls (none) at doubling dilutions from 375 g ml^-1 (1.3–1.4 mM) to 187 g ml^-1 (0.70–0.63 mM), 93 g ml^-1 (0.35–0.31 mM) or 47 g ml^-1 (0.18–0.16 mM) g ml^-1. ^*Values significantly different when compared with untreated controls (analysis of variance (ANOVA), P<0.05). (b) HAGG (250 g ml^-1) stimulated TNF- release from peritoneal macrophage (PEM) following incubation with VIB153 (1.267 mM) or VIB384 (1.377 mM) at a final concentration of 375 g ml^-1. (c) Body weight (wgt) white blood cell count: (WBC), red blood cell count (RBC) and platelets numbers (Plt) of the huFcRIIa mice 24 h after i.p. injection of a single dose of 7.5 mg (diluted in 0.5 ml PBS) of drugs VIB384, 294, 197, 153, 152 or 113, shown as a % of normal parameters measured 12 h prior to treatment. Drug-treated animals were compared with those treated with 0.5 ml of PBS or 30 g per mouse methotrexate in 0.5 ml PBS (2–6 mice per group, mean %s.d.). ^*Values significantly different when compared with untreated controls (ANOVA, P<0.05).

Full figure and legend (105K)

Effects of SCE in vivo

Effects of SCE on blood cells and body weight

In the huFcRIIa transgenic mice treated with SCE or control substances, body weight and full blood counts were monitored at 3, 24 and 48 h, and 4 and 6 days. All mice remained well, with no significant weight loss (Figure 3c). In mice treated with VIB113, platelet and white cell counts decreased significantly, to levels comparable with methotrexate (P<0.05 compared with phosphate-buffered saline (PBS) controls; Figure 3c). No significant changes were seen at other time points, lower doses of VIB113 or with multiple doses of 7.5 mg of VIB197 or 153 (data not shown).

Anti-inflammatory activity of SCE in collagen-induced arthritis

Compounds that specifically inhibited immune complex-mediated activation and aggregation of platelets were tested in vivo for activity in the collagen-induced arthritis (CIA) model. Mice were treated with collagen II in complete Freund's adjuvant (CFA) 21 days prior to treatment with SCE, receiving their first dose of SCE on the same day as a second dose of collagen II/CFA. Although transgenic mice (RR134-expressing huFcRIIa) were highly susceptible to CIA induction¹⁸ and in some cases developed disease after one injection of collagen/CFA (see Figures 4b and d, and 5c and d), all mice received two injections of collagen/CFA and were randomly allocated by cage to treatment groups. No difference was seen in disease incidence between males and females.¹⁸

Figure 4.

Comparison of activity of small chemical entities (SCEs) in collagen-induced arthritis (CIA). (a) The huFcRIIa or C57BL/6 mice immunized with collagen II (days 0 and 21) were treated either with: 7.5 mg of VIB153 on days 21, 24, 27 and 30 (huFcRIIa n=18); 7.5 mg on day 21, then 1 mg day^-1 on days 22–35 (7.5 mg, huFcRIIa n=14); 1 mg day^-1 on days 21–35 (huFcRIIa n=7) or with 0.33 mg day^-1 on days 21–35 (huFcRIIa n=44). Controls were treated with phosphate-buffered saline (PBS), 0.5 ml day^-1 i.p. days 21–35 (huFcRIIa n=24; C57BL/6, n=12). (b) Specificity of action of VIB153: treatment of CIA-susceptible non-transgenic DBA/1 mice with VIB153, 7.5 mg per mouse i.p. on days 21, 24, 27 and 30, following induction of CIA (n=15). The arthritis index at days 35 and 49 (P>0.05) was not significantly different when compared with PBS treatment (n=12). The huFcRIIa mice given VIB153, 7.5 mg per mouse (n=12) on days 21, 24, 27 and 30 showed significantly reduced disease compared with PBS (n=12) controls (P<0.05 at day 49). (c) VIB153 (n=11) compared with VIB152 (n=5), 197 (n=11) and 113 (n=11) at doses of 7.5 mg per mouse i.p. on days 21, 24, 27 and 30 in the huFcRIIa mice, with PBS controls. (d) VIB153 (n=7) compared with VIB294 (n=6) in the huFcRIIa mice at 1 mg day^-1 i.p. on days 21–35. PBS (n=9) controls (0.5 ml i.p. on days 21–35).

Full figure and legend (129K)

Figure 5.

Comparison of activity and specificity of VIB153 and VIB384. (a) The huFcRIIa mice were given 7.5 mg per mouse, i.p. on days 21, 24, 27 and 30 in VIB153 (n=16), VIB384 (n=12) and phosphate-buffered saline (PBS) (n=18) groups (both small chemical entities (SCEs) were P<0.05 compared with PBS at days 35 and 49). Also, mice with index=2 on day 28 were treated with VIB384 (n=5) on days 28, 31, 34, 37 (P<0.05 compared with PBS at days 35 and 42). (b) Dose response of VIB384, from 7.5 mg (n=16), 2.5 mg (n=7) or 1.25 mg (n=5) per mouse in 0.5 ml of PBS, given i.p. on days 21, 24, 27 and 30 after collagen-induced arthritis (CIA) induction in the huFcRIIa mice. PBS (0.5 mls i.p. on days 21, 24, 27 and 30, n=19). (c) Comparison of VIB384 and methotrexate: huFcRIIa mice given VIB384, 7.5 mg per mouse i.p., days 21, 24, 27 and 30 (n=7), or methotrexate, 30 g per mouse i.p., days 21–31 (n=11). These treatments were equally effective at day 35 (P<0.05 compared with PBS controls), but by day 49 the index for methotrexate-treated mice had risen (P>0.05 compared with PBS group), whereas a low mean index was maintained in the VIB153- and 384-treated groups (P<0.05 compared with PBS, n=11 at day 49). (d) Comparison of PBS (n=6) with anti-FcRIIa antibody fragment, 8.7 F(ab')₂ therapy (0.1 mg i.p. days 21, 23, 27 and 30, n=7) with anti-CD3/KT3 (0.5 mg i.p. days 20, 22, 23 and 25, n=6) and VIB153, 7.5 mg per mouse i.p., days 21, 24, 27 and 30 (n=7). These treatments were effective during treatment at day 28, but 8.7 F(ab')₂ and anti-CD3/KT3 showed diminishing inhibition once treatment ceased (>day 35).

Full figure and legend (125K)

Transgenic mice receiving VIB153 (7.5 mg per mouse, 275 mg kg^-1 for a 30 g mouse) on days 21, 24, 27 and 30 showed profound suppression of arthritis development, with a mean arthritis index of 1.2 (maximum possible 12) on day 35, and no further progression of disease at day 49 (Figure 4a). No disease was seen in non-susceptible C57BL/6 mice. To evaluate how changes in the dose of VIB153 may affect arthritis development, huFcRIIa mice were tested with the following treatment regimens: (a) 7.5 mg on day 21 followed by 14 daily doses of 1 mg; (b) 1 mg (33.3 mg kg^-1) daily for 14 days and (c) 0.33 mg (11 mg kg^-1) daily for 14 days. All of the huFcRIIa groups treated with VIB153 showed decreased severity of arthritis (Figure 4a), with 7.5 mg on days 21, 24, 27 and 30 being the most effective treatment.

To test the in vivo specificity of VIB153 for huFcRIIa, VIB153 was tested for inhibition of CIA in non-transgenic DBA/1 mice. This strain is the 'gold standard' CIA-susceptible strain but does not carry the FcRIIA transgene. VIB153 had no significant effect (P>0.05 compared with untreated DBA/1 controls) on the development of CIA in DBA/1 mice (Figure 4b). The lack of effect on CIA in non-transgenic DBA/1 mice argues that the inhibitory activity is predominantly directed at FcRIIa and not other pathways that may be involved in the development of CIA.

Compounds VIB152, 113 and 197 were tested at 7.5 mg per mouse, given intraperitoneally (i.p.) on days 21, 24, 27 and 30 and compared with VIB153 at this dose (Figure 4c). Mean arthritis indices at day 35 were, 6.4 for PBS, 8.4 for VIB152, 4.5 for VIB197 and 1.3 for VIB113, compared with an index of 2.5 for VIB153 at this time point. VIB113 was as effective as VIB153 for inhibition of CIA (P<0.05) at days 35 and 49. The SCE, in order of effectiveness, were: VIB153=113>197>152. VIB152 was not significantly different from PBS controls (P>0.05 at days 35 and 49).

As VIB294 was very effective in vitro (Figures 2 and 3), particularly at inhibiting TNF- secretion from the U937 macrophage cell line (Figure 3a), it was compared in vivo with VIB153 (1 mg 14 doses; Figure 4d). There was no antiarthritic effect with VIB294, whereas VIB153 at this dose again showed significant inhibition of disease (P<0.05 compared with PBS at 35 days). Although reasons for the in vivo compared with in vitro differences are not clear, they are likely to be due to differences in pharmacological properties, as modification of the VIB294 structure, by changing the position of the carboxyl group, created VIB384 with similar in vitro activity to VIB294 (Table 1; Figures 3a and b), but improved in vivo activity. In CIA, VIB384 at all time points was as effective as VIB153 at the 7.5 mg dose. Moreover, VIB384 was effective in delaying disease progression when treatment was commenced at day 28 in mice with mild disease (index=2). In these mice, treatment with 7.5 mg on days 28, 31, 34 and 37 significantly delayed disease, with arthritis indices lower than controls (P<0.05) at days 28 and 35 (Figure 5a). Treatment of mice with more developed disease (scores >4) was not effective with any of the compounds (data not shown).

In a dose–response study, mice were injected on days 21, 24, 27 and 30 with varying doses of VIB384 (7.5, 2.5 or 1.25 mg) or PBS (Figure 5b). The anti-inflammatory effects were seen to be dose dependent. The 1.25-mg dose group had indices similar to the PBS controls (P>0.05), whereas mice injected with the 2.5 mg dose developed moderate arthritis, compared with very mild arthritis in the 7.5 mg dose group (P<0.05 compared with PBS at 49 days). Clearly, VIB384 was as effective as VIB153 at inhibiting CIA in mice, and in vivo testing confirmed the activity of two (VIB153 and 384) of the six compounds selected by in vitro assays.

Comparison of VIB153 and VIB384 with non-SCE treatments of CIA

Methotrexate and immunosuppressive anti-CD3 mAb are known to be effective in treating CIA in DBA/1 mice.^{26, 27} We have shown earlier that fragments of anti-FcRIIa antibodies can inhibit the development of CIA in the huFcRIIa mice.¹⁸ To compare the efficacy of VIB153 and 384 with these treatments of CIA, groups of mice were treated with PBS, VIB153 or VIB384 (both at 7.5 mg day^-1 on days 21, 24, 27 and 30); the published dose and schedule of methotrexate for CIA in mice (30 g day^-1, 1 mg kg^-1, from days 21–31);²⁶ anti-CD3/KT3 (500 g day^-1, 16.5 mg kg^-1)²⁸ or anti-FcRIIa 8.7 F(ab')₂ (100 g day^-1, 3.3 mg kg^-1).¹⁸ VIB153 and 384 were more potent and had a prolonged effect on CIA compared with other treatments (Figures 5c and d). At day 28, all treatments resulted in lower arthritic indices compared with the PBS-treated mice (P<0.05). However, the inhibition of CIA development was sustained only in mice treated with VIB153 or 384 (indices less than 3 at 49 days), compared with the other treatments, where disease progressed rapidly once treatment ceased (P<0.05 compared with VIB153 at 49 days, mean index for non-SCE treatments was >6). Therefore, VIB153 and 384 showed better long-term suppression of CIA than methotrexate, anti-FcR fragments and anti-T cell-antibody. Drug effectiveness was seen in the maintenance of normal movement and shape of limbs, with normal histology, in contrast to untreated controls, where progressive swelling and stiffness were seen, and histology showed cartilage erosion and pannus formation (Figure 6).

Figure 6.

(a) Photograph of the paw of a huFcRIIa transgenic mouse with severe arthritis (score 4), 35 days after collagen-induced arthritis (CIA) induction and (b) paw of a VIB153-treated (7.5 mg per mouse, i.p. on days 21, 24, 27 and 30) huFcRIIa transgenic age-matched mouse at 35 days. (c) Representative section (hematoxylin and eosin-stained) showing the histopathologic features of destructive arthritis in a huFcRIIa transgenic mouse 35 days after CIA induction with mononuclear cell infiltration and advanced pannus. (d) Histopathologic features of a knee joint from a huFcRIIa transgenic mouse 36 days after initial collagen injection and treatment with VIB153 (7.5 mg per mouse, i.p. on days 21, 24, 27 and 30) (hematoxylin and eosin-stained; original magnification 40).

Full figure and legend (220K)

Discussion

FcRIIa plays a pivotal role in immune complex-mediated autoimmune inflammation² and our recent work has demonstrated that antibody fragments specific for FcRIIa can effectively inhibit inflammatory responses in the huFcRIIa transgenic mice.¹⁸ As FcRIIa is expressed on many human leukocytes, modulation of FcRIIa function could have wide-reaching effects on immune activation. The SCEs described herein were designed to inhibit activation through FcRIIa, and can effectively treat CIA in the huFcRIIa transgenic mice, further confirming that this receptor is a valid target for new therapies for diseases such as RA and systemic lupus erythematosus and demonstrating the effectiveness of drug design based on molecular structures defined by X-ray crystallography.

The data presented here clearly demonstrate that compounds designed to bind in the groove formed by dimerization of FcRIIa can selectively interfere with cellular responses following immune complex binding to the FcRIIa receptor, as demonstrated by their specificity in the platelet activation and aggregation assays. Platelets were considered ideal as target 'cells' as they express only FcRIIa, thus avoiding confounding results due to IgG interaction with other FcR. Agonist pathways for platelet responses are well defined, allowing evaluation of specificity and action of SCE on non-FcRIIa pathways (for example, arachidonic acid, ADP or thrombin). Compounds that inhibited IgG-dependent platelet activation and aggregation pathways were selected and tested for TNF- inhibition in vitro and for in vivo testing in the CIA model, using the huFcRIIa transgenic mice.¹⁸

The effectiveness of the in vitro screening techniques was demonstrated, as four of the six compounds selected were active in vivo, causing significant inhibition of CIA. Comparative activity in the assays was as follows:

(a) platelet activation (FACS):

VIB153=384=294>197>152>113>>>BRI6801;

(b) for TNF- inhibition:

VIB294>153=384>197;

(c) CIA inhibition:

VIB153=384=113>197 (152 and 294 inactive).

TNF- is a major component of the inflammatory cascade and FcRIIa has been identified by others as a target for modulation of the expression of TNF-.²⁹ Although inhibition of TNF- production in vitro cannot be extrapolated to predict inhibition of TNF- in vivo, antibodies that block TNF- have been very effective in treating RA,³⁰ so TNF- inhibition was a valuable in vitro screening test to select compounds for in vivo assessment. This approach was validated by showing that SCEs, notably VIB153 and 384, were active in vitro in this assay and also active in vivo. Differential effects of the compounds in vivo could be a consequence of their respective ADME properties. Interaction with other receptors, such as FcRIII, may play a minor role in the inhibition of TNF- and, although this was not seen with CIA. Induction of CIA in DBA/1 mice is FcRIII dependent.³¹

CIA in mice is routinely used as a model of immune complex-dependent autoimmunity, and arthritis development is dependent on binding of autoimmune complexes to FcR,^{32, 33} and deletion or blockade of FcR inhibits the development of CIA.^{18, 34} In vivo, VIB113, 153 and 384 were potent inhibitors of CIA. Both VIB153 and 384 showed better long-term suppression of CIA than methotrexate, anti-T-cell antibody and FcRIIa specific anti-FcR fragments. Given to mice with mild disease (score 2 on day 28) VIB384 significantly delayed disease progression, with arthritis indices lower than controls until day 42 (Figure 5a). However, none of the SCEs was able to control well-established disease (scores >4, data not shown). VIB384 was more effective than methotrexate and anti-CD3, agents that are used primarily in cell depletion therapies in mice,^{26, 27} with disease returning in these animals when therapy ceased and depleted cells were replaced. FcRIIa-specific anti-FcR fragments, although demonstrating that FcRIIa was a valid target for CIA therapy, achieved similar short-term inhibition of disease. The 8.7 F(ab')₂ binds to the FcRIIa immune complex-binding site (see Figure 1), inhibiting receptor interaction with immune complexes.³⁵ The SCEs may act quite differently, as they were designed to fit within the dimeric 'groove', below the immune complex-binding sites. Recent studies of dimer configurations suggest that SCEs may lock the dimer into an inactive form.^{1, 36} We have not been able to block the binding of antibody to receptor with the SCE in in vitro studies using transfected cell lines and human platelets. The antibody-binding site is outside the dimer groove and the drugs were designed to bind in the groove, therefore binding sites should not overlap. We cannot rule out the possibility of some allosteric effects due to the binding of SCEs in the groove that may prevent activation of cells by immune complexes. Affinity constants for binding of SCEs to receptor have not yet been measured, because the target receptor is cell based and labeled ligands are not available. The pharmacokinetic parameters are also yet to be determined. The SCEs are small (molecular weight 250–300) and would be cleared rapidly in vivo, and this is reflected by the higher doses needed for activity.

Rheumatoid arthritis is an autoimmune disorder of unknown etiology, characterized by erosive synovitis, often accompanied with extra articular involvement, such as renal, cardiac, pulmonary and vascular inflammation.³⁷ Treatments for RA include a variety of therapies aimed to relieve pain and swelling of the joints, to slow disease progression and to stop cartilage and bone destruction.³⁸ However, many current drugs, although used long term, do not prevent disease progression.^{39, 40, 41, 42} Novel anti-inflammatory drugs that inhibit FcRIIa function could block disease development early, before activation of the inflammatory cascade. Although FcRIIa antagonists could reduce host resistance to infection, this is unlikely given that humans dispose of immune complexes through C1q binding and subsequent elimination by the liver through capture by CR1 on erythrocytes.⁴³ Blocking FcRIIa should not interfere with these pathways. Increased expression of inhibitory FcRIIb^{44, 45, 46} or blockade of FcRIIa^{29, 47, 48} has been investigated as therapeutic targets for drugs and some have been validated in mouse models^{44, 48} and in human diseases such as systemic lupus erythematosus and RA.^{46, 47} In summary, anti-FcR agents could be less immunosuppressive and act downstream of immune complex formation, blocking FcR-mediated cellular activation. Our earlier studies of immune complex disease in the huFcRIIa mice validate this receptor's role in autoimmune inflammation,^{17, 18} and the data presented herein show that treatments designed to block FcRIIa function can inhibit the development of immune complex-mediated disease.

Methods

Design and synthesis of SCE

The design, synthesis and characterization of the SCE was based on the three-dimensional structure of FcRIIa, identified as a crystallographic dimer and detailed elsewhere.²² The compounds (listed in Table 1) were docked into the FcR dimer using the MINIMIZE_DOCK subroutine in Sybyl 7.2 (Tripos Associates, St Louis, MO, USA) assuming a rigid host site and using Gasteiger–Huckel charges. The ligand (as the dicarboxylate) was then allowed to relax within the confines of the active site (using the Tripos forcefield, default convergent conditions). COMPUTE_ENERGY_DOCK was then used to ascertain that the resulting docked structure was both electrostatically and sterically favorable. Selected compounds were synthesized with >95% purity and tested in vitro and in vivo as aqueous solutions of sodium salts as detailed earlier.⁴⁹

HAGG preparation

HAGG was prepared as described earlier⁵⁰ from Sandoglobulin (Sandoz, Novartis Pharmaceuticals Co., East Hanover, NJ, USA). Briefly, Sandoglobulin (8 mg ml^-1 in PBS) was heated for 30 min at 63 °C and centrifuged at 10 000 g, 4 °C, 10 min, and the supernatant was incubated on ice for 30 min with 1% (w/v) polyethylene glycol 6000 (PEG 6000; Sigma Chemical Co., St Louis, MO, USA) in PBS. The precipitated complexes were centrifuged (10 000 g, 4 °C, 10 min), the supernatant was discarded and complexes were dissolved in PBS at 5 mg ml^-1 and stored at -20 °C.

SCE in HAGG-mediated platelet activation

The platelet-rich plasma from whole blood (centrifuged at 1000 r.p.m. for 10 min) was washed in Tyrodes/Hepes buffer and aliquoted at 5 10⁷ cells per 200 l. Compounds (50 l of 5 mg ml^-1) or PBS was added, plus 1 mM final concentration of EDTA pH 8.0 to prevent aggregation (0.9 l of 0.5 M EDTA) and incubated for 30 min. Activation agents (200 l) were added (HAGG 400 g ml^-1 or arachidonic acid 2.5 g ml^-1) and incubated for 30 min. Samples were fixed (400 l of 4% paraformaldehyde in PBS) and incubated for 30 min before washing with PBS/0.5% bovine serum albumin (2000 r.p.m. for 5 min). Platelets were resuspended in 100 l of conjugated antibody diluted in PBS/bovine serum albumin, for example, fluorescein isothiocyanate-anti-human P-selectin (Seratec, Gottingen, Germany, www.seratec.com.au) and phycoerythrin–anti-gpIIb CD41, isotype controls (IgG1-fluorescein isothiocyanate and IgG1-phycoerythrin, from Silenus, Victoria, Australia, www.chemicon.com/techsupp/Silenus.asp) and platelets were detected by FACScan. The platelet FACS assay was carried out for each compound using three separate platelet donors and average percentage (%) inhibitionstandard deviation (s.d.) was calculated.

Platelet aggregation assay

Platelet suspensions (400 l at 200–300 10⁹ cells l^-1) in a two-channel aggregometer (Chronolog, Havertown, PA, USA) were stimulated with 50 l agonist containing 10 g HAGG or the control agonist at concentrations as follows: 50 mM ADP; 1.6 mM arachidonic acid in 0.2 M Tris pH 7.4 (Sigma Chemical Co.). For stimulation by ADP or arachidonic acid, 10 l of plasma was added. Platelet aggregation was monitored for 15 min.^{51, 52} To evaluate inhibitory activity, SCEs were dissolved in PBS (pH 7.5) and 50 l (500 g ml^-1) of SCE suspension was added to the platelets 5 min before agonist administration.

TNF- release from human cells simulated with HAGG

The human monocyte-like cell line U937 was differentiated in phorbol 12-myristate 13-acetate (Sigma Chemical Co.) at 20 ng ml^-1 in RPMI, for 24 h at 37 °C Non-adherent cells and phorbol 12-myristate 13-acetate were removed by washing with PBS. Adherent cells were harvested, plated into 24-flat well tissue culture plates (4 10⁵ cells per 500 l well) and SCEs were added at 375, 187, 93 or 47 g ml^-1 final concentration. In other wells, anti-FcRIIa mAb (IV.3) was added at 12.5, 6.25, 3.125 or 1.6 g ml^-1 final concentration. Control cultures received an equal volume of PBS. Cultures were incubated for 2 h at 37 °C, stimulated with HAGG (250–500 g ml^-1 in RPMI) and then incubated for 24 h at 37 °C. Supernatants were assayed by ELISA for TNF- on plates coated with mouse–anti-human TNF- mAb (Pharmingen, San Diego, CA, USA, www.pharmingen.com) at 2.5 g ml^-1 (see below).

HAGG stimulation of mouse TNF- and detection by ELISA

To assess TNF- release from peritoneal macrophages, huFcRIIa mice were injected intraperitoneally (i.p.) with 4% thioglycollate and macrophages were lavaged from the peritoneum 4 days later. Adherent peritoneal macrophages were isolated (1 10⁶ cells ml^-1) and incubated with 100 g HAGG, with or without anti-FcRIIa antibody or SCE as detailed above, at 37 °C for 24 h.

Plates (polyvinyl chloride 96 well, Costar-Corning from DKSH, Hallam, VIC, Australia, www.DKSH.com.au) were coated with hamster anti-mouse/rat TNF- (BD Biosciences, Franklin Lakes, NJ, USA) at 2.5 g ml^-1, supernatant was added (50 l for 1 h at 37 °C). Plates were washed 3 with PBS/Tween-20 and detected with secondary biotin anti-mouse TNF- (BD Biosciences) and tertiary streptavidin-HRP (Amersham Lifescience, Buckinghamshire, UK) and absorbance was read at 405 nm by Fluostar optima (BMG LABTECH, Offenburg, Germany).

SCE effects in vivo on mouse body weight and whole blood counts

SCEs were injected i.p. into 8- to 12-week-old huFcRIIa mice. Body weight and full blood counts, measured by Coulter Counter (Coulter Electronics Ltd, Dunstable, Beds, UK), were monitored at 3, 24 and 48 h, and 4 and 6 days. Results were compared with those of mice receiving PBS or methotrexate (30 g per mouse i.p.).

Induction and treatment of CIA

B6.SJL.huFcRIIa (H-2^b) transgenic mice are homozygous for the R134 allelle of FcRIIa.¹⁵ No difference in disease incidence was seen in male and female mice.¹⁸ C57BL/6 (H-2^b) male and female and DBA/1 (H-2^q) male mice at 8–15 weeks of age were used. Collagen type II (Sigma Chemical Co.), 2 mg ml^-1 in 10 mM acetic acid, was emulsified in an equal volume of CFA (Difco Laboratories, Detroit, MI, USA) and 100 l was injected intradermally into the base of the tail. The same dose of collagen II in CFA was administered 21 days later⁵³ and mice were examined daily for arthritis, scored for each limb: score 0=normal, 1=mild inflammation of the paw, 2=severe inflammation of the paw, two or more digits affected, 3=severe inflammation, joint stiffness. Individual mice had a maximum possible score (index) of 12. Mice were randomized to treatment groups by cage, that is, mice in a treatment group were likely to be from the same litter and all of the same age and sex.

For treatment, SCEs were dissolved in PBS (pH 7.5) and injected intraperitoneally, from day 21 or 28 (doses and time points are described in Results). Arthritis progression was monitored up to 49 days (total duration of experiment is 10 weeks). Results were compared with those of mice receiving PBS, methotrexate (30 g per mouse i.p.) FcRIIa antibody fragment, 8.7 F(ab')₂ therapy (0.1 mg i.p.)¹⁸ or anti-CD3/KT3 (0.5 mg i.p.).²⁸

Histology

Joints were preserved in 10% formalin/PBS, decalcified in 5% HCl, 3.5% glacial acetic acid, 95% ethanol and 12.5% (volume/volume) chloroform, and were then embedded in paraffin. Sections (4–6 m) were stained with hematoxylin and eosin.

Statistics

Arthritis index in CIA was expressed as means.d. or standard error of the mean (s.e.m.). Statistical differences were analyzed with the unpaired t-test. For in vitro tests, means.d. for triplicate samples was compared by analysis of variance. All statistical analyses were done using Microsoft Excel analysis tools (www.microsoft.com). A probability of P<0.05 was regarded as significant.

References

1. van de Velde NC, Mottram PL, Hogarth PM. FcgammaRII and multi-system autoimmune disease. Springer Semin Immunopathol 2006; 28: 329–338. | Article | PubMed | ChemPort |
2. Hogarth PM. Fc receptors are major mediators of antibody based inflammation in autoimmunity. Curr Opin Immunol 2002; 14: 798–802. | Article | PubMed | ISI | ChemPort |
3. Bolland S, Ravetch JV. Inhibitory pathways triggered by ITIM-containing receptors. Adv Immunol 1999; 72: 149–177. | PubMed | ISI | ChemPort |
4. Hogarth PM, Ierino FL, Hulett MD. Characterization of FcR Ig-binding sites and epitope mapping. Immunomethods 1994; 4: 17–24. | Article | PubMed | ChemPort |
5. Sylvestre DL, Ravetch JV. Fc receptors initiate the Arthus reaction: redefining the inflammatory cascade. Science 1994; 265: 1095–1098. | Article | PubMed | ISI | ChemPort |
6. Benoist C, Mathis D. Mast cells in autoimmune disease. Nature 2002; 420: 875–878. | Article | PubMed | ISI | ChemPort |
7. Schmitt DA, Bieber T, Cazenave JP, Hanau D. Fc receptors of human Langerhans cells. J Invest Dermatol 1990; 94: 15S–21S. | Article | PubMed | ChemPort |
8. Sallusto F, Lanzavecchia A. Efficient presentation of soluble antigen by cultured human dendritic cells is maintained by granulocyte macrophage colony-stimulating factor plus interleukin-4 and downregulated by tumor necrosis factor-alpha. J Exp Med 1994; 179: 1109–1118. | Article | PubMed | ISI | ChemPort |
9. Unkeless JC, Scigliano E, Freedman VH. Structure and function of human and murine receptors for IgG. Annu Rev Immunol 1988; 6: 251–281. | Article | PubMed | ChemPort |
10. Cambier JC. Antigen and Fc receptor signaling. The awesome power of the immunoreceptor tyrosine-based activation motif (ITAM). J Immunol 1995; 155: 3281–3285. | PubMed | ISI | ChemPort |
11. Rascu A, Repp R, Westerdaal NA, Kalden JR, van de Winkel JG. Clinical relevance of Fc gamma receptor polymorphisms. Ann NY Acad Sci 1997; 815: 282–295. | Article | PubMed | ChemPort |
12. van Lent PL, Licht R, Dijkman H, Holthuysen AE, Berden JH, van den Berg WB. Uptake of apoptotic leukocytes by synovial lining macrophages inhibits immune complex-mediated arthritis. J Leukoc Biol 2001; 70: 708–714. | PubMed | ChemPort |
13. Bendele A, McComb J, Gould T, McAbee T, Sennello G, Chlipala E et al. Animal models of arthritis: relevance to human disease. Toxicol Pathol 1999; 27: 134–142. | Article | PubMed | ISI | ChemPort |
14. McKenzie SE. Humanized mouse models of FcR clearance in immune platelet disorders. Blood Rev 2002; 16: 3–5. | Article | PubMed | ChemPort |
15. McKenzie SE, Taylor SM, Malladi P, Yuhan H, Cassel DL, Chien P et al. The role of the human Fc receptor Fc gamma RIIA in the immune clearance of platelets: a transgenic mouse model. J Immunol 1999; 162: 4311–4318. | PubMed | ChemPort |
16. Taylor SM, Reilly MP, Schreiber AD, Chien P, Tuckosh JR, McKenzie SE. Thrombosis and shock induced by activating antiplatelet antibodies in human FcgammaRIIA transgenic mice: the interplay among antibody, spleen, and Fc receptor. Blood 2000; 96: 4254–4260. | PubMed | ChemPort |
17. Tan Sardjono C, Mottram PL, Hogarth PM. The role of FcgammaRIIa as an inflammatory mediator in rheumatoid arthritis and systemic lupus erythematosus. Immunol Cell Biol 2003; 81: 374–381. | Article | PubMed |
18. Tan Sardjono C, Mottram PL, van de Velde NC, Powell MS, Power D, Slocombe RF et al. Development of spontaneous multisystem autoimmune disease and hypersensitivity to antibody-induced inflammation in Fcgamma receptor IIa-transgenic mice. Arthritis Rheum 2005; 52: 3220–3229. | Article | PubMed | ChemPort |
19. Hulett MD, Hogarth PM. Molecular basis of Fc receptor function. Adv Immunol 1994; 57: 1–127. | PubMed | ISI | ChemPort |
20. Hulett MD, Witort E, Brinkworth RI, McKenzie IF, Hogarth PM. Identification of the IgG binding site of the human low affinity receptor for IgG Fc gamma RII. Enhancement and ablation of binding by site-directed mutagenesis. J Biol Chem 1994; 269: 15287–15293. | PubMed | ISI | ChemPort |
21. Powell MS, Barton PA, Emmanouilidis D, Wines BD, Neumann GM, Peitersz GA et al. Biochemical analysis and crystallisation of Fc gamma RIIa, the low affinity receptor for IgG. Immunol Lett 1999; 68: 17–23. | Article | PubMed | ChemPort |
22. Maxwell KF, Powell MS, Hulett MD, Barton PA, McKenzie IF, Garrett TP et al. Crystal structure of the human leukocyte Fc receptor, Fc gammaRIIa. Nat Struct Biol 1999; 6: 437–442. | Article | PubMed | ISI | ChemPort |
23. Pietersz GA, Powell MS, Ramsland PA, Hogarth PM. Receptor structure and design of anti-inflammatories: new therapeutics for autoimmune disease. Annu Rep Med Chem 2002; 37: 217–224. | ChemPort |
24. Jin J, Quinton TM, Zhang J, Rittenhouse SE, Kunapuli SP. Adenosine diphosphate (ADP)-induced thromboxane A(2) generation in human platelets requires coordinated signaling through integrin alpha(IIb)beta(3) and ADP receptors. Blood 2002; 99: 193–198. | Article | PubMed | ChemPort |
25. Gosselin EJ, Brown MF, Anderson CL, Zipf TF, Guyre PM. The monoclonal antibody 41H16 detects the Leu 4 responder form of human Fc gamma RII. J Immunol 1990; 144: 1817–1822. | PubMed | ChemPort |
26. Neurath MF, Hildner K, Becker C, Schlaak JF, Barbulescu K, Germann T et al. Methotrexate specifically modulates cytokine production by T cells and macrophages in murine collagen-induced arthritis (CIA): a mechanism for methotrexate-mediated immunosuppression. Clin Exp Immunol 1999; 115: 42–55. | Article | PubMed | ISI | ChemPort |
27. Utset TO, Auger JA, Peace D, Zivin RA, Xu D, Jolliffe L et al. Modified anti-CD3 therapy in psoriatic arthritis: a phase I/II clinical trial. J Rheumatol 2002; 29: 1907–1913. | PubMed | ISI | ChemPort |
28. Mottram PL, Murray-Segal LJ, Han W, Maguire J, Stein-Oakley AN. Remission and pancreas isograft survival in recent onset diabetic NOD mice after treatment with low-dose anti-CD3 monoclonal antibodies. Transpl Immunol 2002; 10: 63–72. | Article | PubMed | ChemPort |
29. Belostocki K, Park MS, Redecha PB, Masuda E, Salmon JE, Pricop L. FcgammaRIIa is a target for modulation by TNFalpha in human neutrophils. Clin Immunol 2005; 117: 78–86. | Article | PubMed | ChemPort |
30. Feldmann M. Development of anti-TNF therapy for rheumatoid arthritis. Nat Rev Immunol 2002; 2: 364–371. | Article | PubMed | ISI | ChemPort |
31. Stahl TD, Andren M, Martinsson P, Verbeek JS, Kleinau S. Expression of FcgammaRIII is required for development of collagen-induced arthritis. Eur J Immunol 2002; 32: 2915–2922. | Article | PubMed |
32. Myers LK, Rosloniec EF, Cremer MA, Kang AH. Collagen-induced arthritis, an animal model of autoimmunity. Life Sci 1997; 61: 1861–1878. | Article | PubMed | ISI | ChemPort |
33. Nakamura A, Takai T. A role of FcgammaRIIB in the development of collagen-induced arthritis. Biomed Pharmacother 2004; 58: 292–298. | Article | PubMed | ChemPort |
34. van Lent PL, Nabbe K, Blom AB, Holthuysen AE, Sloetjes A, van de Putte LB et al. Role of activatory Fc gamma RI and Fc gamma RIII and inhibitory Fc gamma RII in inflammation and cartilage destruction during experimental antigen-induced arthritis. Am J Pathol 2001; 159: 2309–2320. | PubMed | ISI | ChemPort |
35. Ierino FL, Hulett MD, McKenzie IF, Hogarth PM. Mapping epitopes of human Fc gamma RII (CDw32) with monoclonal antibodies and recombinant receptors. J Immunol 1993; 150: 1794–1803. | PubMed | ISI | ChemPort |
36. Powell MS, Barnes NC, Bradford TM, Musgrave IF, Wines BD, Cambier JC et al. Alteration of the Fc gamma RIIa dimer interface affects receptor signaling but not ligand binding. J Immunol 2006; 176: 7489–7494. | PubMed | ChemPort |
37. Harris EJ. Clinical features of rheumatiod arthritis. In: Ruddy S, Harris E, Sledge B (eds). Kelley's Textbook of Rheumatology. WB Saunders Company: Philadelphia, Pennsylvania, 2001, pp 967–1000.
38. Mottram PL. Past, present and future drug treatment for rheumatoid arthritis and systemic lupus erythematosus. Immunol Cell Biol 2003; 81: 350–353. | Article | PubMed | ChemPort |
39. Aletaha D, Kapral T, Smolen JS. Toxicity profiles of traditional disease modifying antirheumatic drugs for rheumatoid arthritis. Ann Rheum Dis 2003; 62: 482–486. | Article | PubMed | ChemPort |
40. Neeck G. Fifty years of experience with cortisone therapy in the study and treatment of rheumatoid arthritis. Ann NY Acad Sci 2002; 966: 28–38. | PubMed | ChemPort |
41. Simon LS, Smolen JS, Abramson SB, Appel G, Bombardier C, Brater DC et al. Controversies in COX-2 selective inhibition. J Rheumatol 2002; 29: 1501–1510. | PubMed | ChemPort |
42. Solomon DH, Avorn J, Sturmer T, Glynn RJ, Mogun H, Schneeweiss S. Cardiovascular outcomes in new users of coxibs and nonsteroidal antiinflammatory drugs: high-risk subgroups and time course of risk. Arthritis Rheum 2006; 54: 1378–1389. | Article | PubMed | ChemPort |
43. Clynes R, Maizes JS, Guinamard R, Ono M, Takai T, Ravetch JV. Modulation of immune complex-induced inflammation in vivo by the coordinate expression of activation and inhibitory Fc receptors. J Exp Med 1999; 189: 179–185. | Article | PubMed | ISI | ChemPort |
44. van Lent P, Nabbe KC, Boross P, Blom AB, Roth J, Holthuysen A et al. The inhibitory receptor FcgammaRII reduces joint inflammation and destruction in experimental immune complex-mediated arthritides not only by inhibition of FcgammaRI/III but also by efficient clearance and endocytosis of immune complexes. Am J Pathol 2003; 163: 1839–1848. | PubMed | ChemPort |
45. Moll T, Nitschke L, Carroll M, Ravetch JV, Izui S. A critical role for Fc gamma RIIB in the induction of rheumatoid factors. J Immunol 2004; 173: 4724–4728. | PubMed | ChemPort |
46. Stefanescu RN, Olferiev M, Liu Y, Pricop L. Inhibitory Fc gamma receptors: from gene to disease. J Clin Immunol 2004; 24: 315–326. | Article | PubMed | ChemPort |
47. Pricop L, Li L, Salmon JE, Jacob CO. Characterization of the FcgammaRIIA promoter and 5'UTR sequences in patients with systemic lupus erythematosus. Genes Immun 2002; 3 (Suppl. 1): S47–S50. | Article | PubMed | ChemPort |
48. van Royen-Kerkhof A, Sanders EA, Walraven V, Walraven V, Voorhorst-Ogink M, Saeland E et al. A novel human CD32 mAb blocks experimental immune haemolytic anaemia in FcgammaRIIA transgenic mice. Br J Haematol 2005; 130: 130–137. | Article | PubMed | ChemPort |
49. Hogarth PM, Pietersz G, Baell JB, Garrett TPT, Matthews BR, McCarthy TD. Fc receptor modulators and uses thereof. International publication number WO2000/015214.
50. Wines BD, Gavin A, Powell MS, Steinitz M, Buchanan RR, Mark Hogarth P. Soluble FcgammaRIIa inhibits rheumatoid factor binding to immune complexes. Immunology 2003; 109: 246–254. | Article | PubMed | ChemPort |
51. Gratacap MP, Herault JP, Viala C, Ragab A, Savi P, Herbert JM et al. FcgammaRIIA requires a Gi-dependent pathway for an efficient stimulation of phosphoinositide 3-kinase, calcium mobilization, and platelet aggregation. Blood 2000; 96: 3439–3446. | PubMed | ChemPort |
52. Gross BS, Wilde JI, Quek L, Chapel H, Nelson DL, Watson SP. Regulation and function of WASp in platelets by the collagen receptor, glycoprotein VI. Blood 1999; 94: 4166–4176. | PubMed | ISI | ChemPort |
53. Campbell IK, Bendele A, Smith DA, Hamilton JA. Granulocyte-macrophage colony stimulating factor exacerbates collagen induced arthritis in mice. Ann Rheum Dis 1997; 56: 364–368. | Article | PubMed | ChemPort |

Acknowledgements

Funding to synthesize SCE was provided by Arthron Ltd. We thank Tessa Bradford and Nick Tsipouras for excellent technical assistance. This study was funded by the National Health and Medical Research Council (Australia), PrimaBiomed Ltd (Melbourne, Australia) and Trillium Therapeutics Inc., (Toronto, Ontario, Canada). MSP and PLM received support from the Arthritis Foundation, Australia. CTS was supported by a philanthropic grant from PaperlinX Pty Ltd. PAR is the recipient of an NH&MRC R Douglas Wright Career Development Award (365209).