Avian IgY Binds to a Monocyte Receptor with IgG-like Kinetics Despite an IgE-like Structure*

An ancestor of avian IgY was the evolutionary precursor of mammalian IgG and IgE, and present day chicken IgY performs the function of human IgG despite having the domain structure of human IgE. The kinetics of IgY binding to its receptor on a chicken monocyte cell line, MQ-NCSU, were measured, the first time that the binding of a non-mammalian antibody to a non-mammalian cell has been investigated (k+1 = 1.14 ± 0.46 × 105 mol–1sec–1, k–1 = 2.30 ± 0.14 × 10–3 s–1, and Ka = 4.95 × 107 m–1). This is a lower affinity than that recorded for mammalian IgE-high affinity receptor interactions (Ka ∼ 1010 m–1) but is within the range of mammalian IgG-high affinity receptor interactions (human: Ka ∼ 108–109 m–1 mouse: Ka ∼ 107–108 m–1. IgE has an extra pair of immunoglobulin domains when compared with IgG. Their presence reduces the dissociation rate of IgE from its receptor 20-fold, thus contributing to the high affinity of IgE. To assess the effect of the equivalent domains on the kinetics of IgY binding, IgY-Fc fragments with and without this domain were cloned and expressed in mammalian cells. In contrast to IgE, their presence in IgY has little effect on the association rate and no effect on dissociation. Whatever the function of this extra domain pair in avian IgY, it has persisted for at least 310 million years and has been co-opted in mammalian IgE to generate a uniquely slow dissociation rate and high affinity.

IgY is the principal serum antibody of amphibians, reptiles, and birds and shares a common ancestor with both mammalian IgG and IgE (1,2). The cDNA sequence of the chicken upsilon () heavy chain (3) reveals that, like more basal amphibian IgY (4,5), avian IgY contains a domain pair (C2) 4 that has been conserved in mammalian IgE (as C⑀2) but truncated to form the "hinge" region in mammalian IgG (3). Thus, an orthologous domain pair must have existed in the common (IgY-like) ancestor prior to its duplication and subsequent divergence in the mammalian lineage. Avian IgY is the closest extant protein to this ancestor and therefore the most logical choice for study of the evolution of modern IgG and IgE from an ancient IgY-like ancestor.
IgG, the principal mammalian serum antibody, aggregates and facilitates opsonization of antigens, activates complement, and provides protection for the fetus following transport across the placenta. IgE does not activate complement nor cross the placenta but can sensitize effector cells (principally mast cells and basophils) and typically mediates anaphylactic reactions (6). It is well known for its involvement in allergic disease but probably also provides defense against parasitic infections (7).
Avian IgY appears to combine the functions of mammalian IgG and IgE. As the major serum antibody, it provides defense against infection (8) but also mediates anaphylaxis (9). Chicken mast cells activated by IgY are responsible for local anaphylactic reactions in the gut, which play a role in defense against protozoan infections (10,11). Antibody-dependent hypersensitivity and fatal systemic anaphylactic shock have also been demonstrated in chickens (9,12,13) and shown to be mediated by basophils (14), which are much more numerous than mast cells in birds (in contrast to mammals in which the reverse is true) (15). These phenomena constitute indirect evidence for the presence of IgY receptors on both chicken monocytes/macrophages and basophils.
The heavy chain of IgY consists of four constant Ig domains and one variable Ig domain (3), as in mammalian IgE. It is clear from sequence comparisons that after the divergence of the synapsids 310 million years ago (16,17), but before the divergence of the non-eutherian mammals 166 million years ago (18), duplication of the -gene occurred; this was followed by differentiation into the two Ig classes unique to mammals, IgG and IgE, as indicated by their presence in monotremes (19,20). In IgG, the second domain of the heavy chain constant region condensed to form the hinge polypeptide (3), whereas in IgE, the domain was retained as C⑀2 so that, like IgY, the Fc fragment consists of three constant Ig domain pairs. The two interheavy-chain disulfide bridges in the C⑀2 domain are present also in C2 at the same locations and in the hinge of IgG (Fig. 1).
Human mast cells and basophils express the high affinity IgE receptor (Fc⑀RI), which has an affinity for IgE that is uniquely high: K a ϳ 10 10 M Ϫ1 (7). We have shown, through our studies with IgE, IgE-Fc, and a subfragment of IgE-Fc consisting of only the C⑀3 and C⑀4 domains termed Fc⑀3-4 (21), that the high affinity is due to the slow dissociation rate from Fc⑀RI (22,23) and that the C⑀2 domains are in part responsible (24). With a half-life on tissue mast cells of 2-3 weeks (25), IgE can effectively and persistently sensitize these cells for rapid degranulation upon encounter with antigen (allergen). The rapidity of degranulation is responsible for the life-threatening response of anaphylaxis, making an understanding of the mechanism by which the C⑀2 domains alter the dissociation rate of IgE from its receptor a medically important issue.
C2-like domain pairs have persisted since their first appearance in a common ancestor that unites all terrestrial vertebrates. To assess their effect on the antibody-Fc receptor interaction, we have expressed two avian IgY-Fc fragments that differ only by the presence (Fc2-4) or absence (Fc3-4) of the C2 domain pair and measured their binding to the chicken monocyte cell line, MQ-NCSU, as well as that of whole IgY. This is the first report of the cell binding properties of a nonmammalian antibody and also addresses the question of whether the "extra" C2 domain pair in IgY necessarily leads to a high affinity interaction.

EXPERIMENTAL PROCEDURES
Cloning of Fc2-4 and Fc3-4-There is no intron between the CH1 and CH2 exons in chicken (26). Hence, the N terminus of chicken C2, Val-231 by the numbering system of Suzuki and Lee (27), was identified by a Clustal W alignment of the sequences of the heavy chain constant region of human IgE (28) and of duck IgY (29) with that of chicken IgY taken from Parvari et al. (3). The second residue of C⑀2 is a cysteine that in IgE forms an intra-chain disulfide bond with C⑀1; it was therefore mutated to an alanine in Fc⑀2-4 and has similarly been mutated in Fc2-4. The N-terminal 4 amino acids of IgY-Fc were Asp-Ile (from an EcoRV site) Val-Ala.
Fc⑀3-4 included the residues Asp-Ile from an EcoRV restriction site and 2 amino acids from C⑀2, one of which, a cysteine, was thought to be required to form a functionally important inter-chain disulfide bond (21). For this reason, a similar Fc3-4 with an N terminus including Cys-340 was cloned and expressed, but the protein formed intra-chain disulfide bonds with the extra Cys-347 in chicken IgY (Fig. 1). A C347S mutant led to expression of a functional protein, Fc3-4, which has been used for the study reported here.
The polymerase chain reaction was used to engineer EcoRV and PmeI sites onto the 5Ј and 3Ј termini, respectively, of both Fc2-4 and Fc3-4 cDNA from a chicken splenocyte library (kindly donated by Dr. John Young, Institute of Animal Health, Compton UK). Using a holding vector, a PmeI site was introduced after the stop codon of IgE-Fc so that using EcoRV and PmeI, IgY-Fc could be exchanged for IgE-Fc and then inserted into the pRY vector as a HindIII-HindIII cassette (30). Vectors were transfected into NS0 mouse myeloma cells for production of stable clones expressing Fc2-4 or Fc3-4.
Cell Lines-The MQ-NCSU cell line was provided by Dr. M. A. Qureshi, National Program Leader, Animal Genetics, United States Department of Agriculture Cooperative State Research Education and Extension Service, Washington, D.C. (34). Cells were grown at 37°C in modified LM-Hahn medium omitting fetal bovine serum and chick serum, and thereby exogenous IgY, but including 8% ultra-low IgG fetal bovine serum (Invitrogen, catalog number 16250-078).
Flow Cytometry (FACS)-FACS was used to detect immunostained cells following their incubation with specific antibodies and/or antibody conjugates. In all experiments, cells were washed (by centrifugation at 1000 ϫ g for 5 min) and incubated for 10 min in PBS containing 1% bovine serum albumin (BSA) (Sigma, catalog number A7030) before staining for 1 h at 4°C with chicken IgY-(Fab) 2 (Rockland, catalog number 003-0104), (Fab) 2 ϩ whole IgY, IgY, or buffer before incubation with a mouse monoclonal antibody (Sigma, catalog number C-7295), which was found to be an anti-C2, 5 and detection by antimouse fluorophore-conjugated secondary antibodies (Dako, catalog number F0479). Cells were washed twice with PBS-BSA both before and after all antibody incubation steps and fixed in PBS ϩ 4% formaldehyde (Sigma, catalog number F1635) before analysis in a FACSCalibur instrument (BD Biosciences).
Radioiodination of Antibodies and Antibody Fragments-Affinity-purified IgY (Stratech, catalog number 003-000-003) and IgY-Fc were radiolabeled with 125 I (PerkinElmer Life Sciences, catalog number NEZ033A001MC) using the chloramine-T method (35). Desalting spin columns (Pierce, catalog number 89849) were used to remove unincorporated 125 I and exchange the radiolabeled IgY or IgY-Fc into PBS, 0.05% sodium azide following the manufacturer's protocol. The specific activity in cpm/g of each radioligand was determined by estimation of the protein concentration (by UV spectrophotometry) and cpm of radioactivity, using a Wizard 1480 ␥ counter (PerkinElmer Life Sciences).
Cell Binding Assays-MQ-NCSU cells were washed in PBS, 1% BSA, 0.05% sodium azide, pH 7.4, and incubated at 23°C for 10 min before resuspension at 10 7 cells/ml. In experiments to determine the association rate constant, duplicate samples of 10 6 cells were incubated with 10 nM 125 I-labeled IgY, Fc2-4, or Fc3-4 at 23°C. At regular intervals, cells were centrifuged through 200 l of phthalate oil (36), and the radioactivity of the cell pellet was determined using a Wizard 1480 gamma counter (PerkinElmer Life Sciences). The initial rate of association was determined from a linear fit of cpm versus time. Standard errors were generated by the fitting procedure but, due to uncertainty in number of the receptors per cell (see "Discussion"), larger errors have been added to the final values.
Dissociation experiments were performed using the method of Kulczycki and Metzger (37) but with all steps carried out at room temperature; cells were first incubated with 10 nM 125 I-labeled IgY, Fc2-4, or Fc3-4 for 1 h and then resuspended in an excess of unlabeled IgY, and their radioactivity was assayed as above. Nonspecific binding was measured by incubating cells with an excess (350 nM) of unlabeled IgY for 1 h at room temperature before the addition of 10 nM labeled IgY, and the ␥ counts obtained were subtracted from the total bound counts. The mean number of receptor sites per MQ-NCSU cell was determined by titrating to saturation with 125 I-labeled IgY or Fc2-4 of known specific activity. Dissociation rate constants were determined both by non-linear regression analysis from a plot of cpm versus time (according to Equation 20, p88, in Goodrich and Kugel (38)) and from linear regression analysis of a plot of ln(cpm) versus time.

RESULTS
The much higher affinity of IgE for its receptor on mast cells when compared with IgG for any of its receptors and the role of the extra C⑀2 domain pair in IgE prompted this study of chicken IgY, the closest extant immunoglobulin to the ancestor that evolved into all three antibodies.
In a preliminary FACS experiment, it was established that chicken serum IgY binds to MQ-NCSU cells in an Fc-dependent manner; IgY Fabs at 400ϫ molar excess over whole IgY were unable to bind to cells or to inhibit IgY binding (Fig. 2). The binding constants determined in the studies described below are thus for IgY and IgY-Fc binding to an IgY-Fc receptor on a chicken monocyte cell line.
The action of papain on IgY does not give rise to an Fc2-4 fragment since the C2 domain is digested away (27). Hence, to assess the role of C2, Fc2-4 and Fc3-4 were cloned and expressed as recombinant proteins. Purified IgY (Fig. 3, lane 1), 5   recombinant Fc2-4 (lane 2), and Fc3-4 (lane 3), both unreduced (Fig. 3, A and C) and reduced (Fig. 3, B and D), were electrophoresed and stained with Coomassie Brilliant Blue (Fig. 3, A and B) or transferred to nitrocellulose and stained with anti-IgY horseradish peroxidase conjugate (Fig. 3, C and D). Reduced heavy chains gave rise to a single band (Fig. 3C, lanes 1-3), suggesting that the small amount of heterogeneous-sized material in unreduced samples was due to the formation of non-native disulfide bonds on exposure to 100°C in SDS buffer (39) and not to degradation or heterogeneity of glycosylation. Western blots of the same samples (Fig. 3, C and D) showed that the polypeptides are indeed IgY of approximately the expected molecular mass with only a minor band of Fc3-4 monomer (Fig. 3C, lane 3). All were glycosylated as shown by periodic acid Schiff staining, 5 and all had essentially no free sulfydryls per mole of IgY or its fragments, as determined by reduction and carboxymethylation (40). 5 The measured rate constant for the dissociation of 125 I-IgY from MQ-NCSU cells was 2.30 Ϯ 0.14 ϫ 10 Ϫ3 s Ϫ1 (Fig. 4A). A plot of ln(cpm) against time yields a straight line, showing that a single, zero order dissociation event is involved (returning a confirmatory value of 2.05 Ϯ 0.04 ϫ 10 Ϫ3 s Ϫ1 for k Ϫ1 ; Fig. 4B). The same value was obtained when biotinylated IgY, stained with streptavidin-R-phycoerythrin, dissociated from the cells after the addition of unlabeled IgY, 5 suggesting that the derivatization of lysine with biotin, or of tyrosine with iodine, does not perturb the IgY-receptor interaction.
To measure the rate of association of 125 I-IgY with MQ-NCSU cells, an estimate of the number of receptors per cell was made. Measurements on two different days led to an estimate of 20,000, and the likely error in this value is discussed below. The value obtained for k ϩ1 is 1.14 Ϯ 0.04 ϫ 10 5 mol Ϫ1 s Ϫ1 (Fig.  4C). The two rate constants lead to a value for K a ϭ 4.95 ϫ 10 7 M Ϫ1 . When the binding kinetics of the IgY-Fc fragments were determined, Fc2-4 was found to display the same rate of dissociation as Fc3-4 ( Fig. 5A), showing that the C2 domain pair does not have a detectable effect on the rate of dissociation. The association rate constant for Fc2-4 was 3.66 Ϯ 0.16 ϫ 10 5 mol Ϫ1 s Ϫ1 , and for Fc3-4, it was 1.64 Ϯ 0.01 ϫ 10 5 mol Ϫ1 s Ϫ1 , which differ only by a factor of about 2 (Fig. 5B). This is not a significant difference given the errors inherent in this technique (discussed below).

DISCUSSION
These studies determine for the first time the binding kinetics of IgY and its Fc fragments to cells and demonstrate that the affinity is more similar to that of IgG for its high affinity receptor than IgE for its high affinity receptor. Furthermore, the kinetics of the interaction are not affected by the presence or absence of the C2 domain pair, in contrast to the effect that the C⑀2 pair has on the kinetics of IgE binding to Fc⑀RI (24).
IgY dissociates from its receptor on the MQ-NCSU chicken monocyte cell line with a rate constant of k Ϫ1 ϭ 2.30 Ϯ 0.14 ϫ 10 Ϫ3 s Ϫ1 at 23°C (Fig. 4A). Although 10 times faster than the dissociation of human IgG1 from the monocyte cell line U937 (k Ϫ1 ϭ 1.1 Ϯ 0.1 ϫ 10 Ϫ4 s Ϫ1 ) (41), k Ϫ1 for IgY is close to the value 3.5 ϫ 10 Ϫ3 s Ϫ1 (no error reported) measured for the dissociation of mouse IgG subclass 2a from the mouse macrophage cell line P388D (42). Dissociation of human IgE from (cord blood) mast cells is so slow by comparison that even at 37°C, values of 6.05 ϫ 10 Ϫ5 s Ϫ1 (43) (no error given) and 6.93 Ϯ 0.54 ϫ 10 Ϫ5 s Ϫ1 (44) have been recorded.
A value of 1.14 Ϯ 0.04 ϫ 10 5 mol Ϫ1 s Ϫ1 has been determined for the association rate constant of IgY with MQ-NCSU cells, although a much larger error is inherent in this value (and consequently the K a ) than in that of the dissociation rate constant. The number of receptors per cell is a factor in the calculation, and it has been reported that Fc receptor number on human monocytes may vary on average by Ϯ25% and up to 40% (45); the error in k ϩ1 for IgY may therefore be as much as Ϯ40%. The value of k ϩ1 for IgE binding to the Fc⑀RI ␣ chain expressed on Chinese hamster ovary cells at 25°C is 3.1 Ϯ 0.3 ϫ 10 5 mol Ϫ1 s Ϫ1 (30), and for human IgG1 to U937 cells, it is 2.1 Ϯ 0.2 ϫ 10 5 mol Ϫ1 s Ϫ1 (41). Thus, the association rate constant for human IgE differs from the value determined here for IgY binding to MQ-NCSU cells by a factor of only 2.7, in contrast to the difference in affinity, which is about 700-fold (K a ϭ 4.95 ϫ 10 7 M Ϫ1 for IgY; K a ϭ 3.4 ϫ 10 10 M Ϫ1 for IgE) (30). The difference between IgY and IgE in their affinity for cells is therefore due almost entirely to a difference in their rates of dissociation (as it is for IgG when compared with IgE). The affinity of IgY reported here is within the range of mammalian IgG-high affinity receptor interactions: e.g. human K a ϳ 10 8 -10 9 M Ϫ1 ; mouse K a ϳ 10 7 -10 8 M Ϫ1 (46).
The presence of the C⑀2 domain pair slows the rate of dissociation of IgE from Fc⑀RI by a factor of approximately 20, as shown by a comparison of the binding kinetics of two IgE fragments, one with this domain (Fc⑀2-4) and one without (Fc⑀3-4) (22,24). The influence of the C2 domains on the kinetics of IgY binding to cells was therefore investigated using two corresponding IgY fragments, Fc2-4 and Fc3-4, differing only by the C2 domain pair. The dissociation curves of Fc2-4 and Fc3-4 can be superimposed (Fig. 5A), showing that unlike IgE-Fc (24), the presence of the C2 domain pair has no effect on the rate of dissociation. Although the values for the association rate constants differ by a factor of 2.2 (k ϩ1 ϭ 3.66 Ϯ 0.16 ϫ 10 5 mol Ϫ1 s Ϫ1 for Fc2-4; k ϩ1 ϭ 1.64 Ϯ 0.01 ϫ 10 5 mol Ϫ1 s Ϫ1 for Fc3-4), the potential error in these values due to possible variability in cell receptor number of Ϯ40% means that these values are not significantly different.
Although the C2 domains of chicken IgY do not appear to contribute to receptor binding on monocytes, this does not preclude a role in, for instance, binding to mast cells or basophils. However, cutaneous anaphylaxis experiments in chickens (9) have shown that IgY is not retained in tissue for as long as mammalian IgE (i.e. only for days rather than weeks), which suggests that the dissociation rate of IgY from mast cells is also more akin to that of IgG than IgE. A function for C2 is likely, however, as it has persisted in both mammals and birds for the 310 million years since Synapsida and Reptilia diverged, and it has recently been estimated that DNA coding for a non-functional protein domain has a half-life of only 4 million years in the genome (47,48).
The association constant determined in the experiments described in this report (K a ϭ 4.95 ϫ 10 7 M Ϫ1 ) is close to that determined for the interaction between chicken serum IgY and an extracellular domain of a member of the chicken leukocyte receptor complex (LRC), CHIR-AB1, as measured by surface plasmon resonance: K a ϭ 3.70 ϫ 10 7 M Ϫ1 (49). Since CHIR-AB1 mRNA was detectable in chicken M11 primary macrophages (49), it is likely that CHIR-AB1 is the molecule responsible for IgY binding to MQ-NCSU monocytes in the data presented here. One caveat in comparing these two values is that association constants obtained by cell binding and SPR sometimes differ; K a for IgE binding to the Fc⑀RI␣ chain displayed on Chinese hamster ovary cells is 10ϫ higher than the value obtained by surface plasmon resonance (22,24,30). Furthermore, additional polypeptide chains in the native, cell-bound receptor may influence affinity.
Another issue to consider is whether the MQ-NCSU cells express more than one receptor. The human monocyte cell line, U937, expresses a number of both IgG and IgE receptors, and MQ-NCSU cells may express receptors homologous to these, to which IgY may in principle bind. However, in studies of IgG binding to U937 cells, only one receptor, Fc␥RI, is detected, despite the presence of others, Fc␥RII and Fc␥RIII, because the dissociation rates are too fast to be measured (50). Similarly, IgE binding to U937 cells detects neither CD23 (low affinity receptor for IgE) unless induced (because it is expressed at such a low level) (51) nor Fc⑀RI (which requires neuraminidase treatment to unmask it) (52). The kinetic data reported here for IgY binding to MQ-NCSU cells suggest that a single receptor is being detected.
Comparison of the association rate constants measured here in cell binding for IgY with those determined by surface plasmon resonance (49) also shows good agreement: 1.14 ϫ 10 5 mol Ϫ1 s Ϫ1 and 2.4 ϫ 10 5 mol Ϫ1 s Ϫ1 , respectively. These authors also measured the binding kinetics for an IgY-Fc fragment produced by papain digestion, consisting of a dimer of the C3 and C4 domains (49). Their value of k ϩ1 ϭ 3.9 ϫ 10 6 mol Ϫ1 s Ϫ1 for this fragment is 24ϫ faster than that reported here, k ϩ1 ϭ 1.64 Ϯ 0.01 ϫ 10 5 mol Ϫ1 s Ϫ1 , for the recombinant Fc3-4. However, we also generated an Fc fragment by papain digestion and found, by N-terminal sequencing (27), 5 that a substantial proportion of the material consists of an incomplete C3 domain; this may explain the discrepancy between the two k ϩ1 values.
The studies of Viertlboeck et al. (49) and, we believe, the data reported here, concern the binding of an antibody to a receptor coded within the chicken LRC (on microchromosome 31). The mammalian LRC is a 1-megabase gene cluster on human chromosome 19, consisting of 45 genes, of which 30 encode Ig-like receptors that have mostly HLA class I molecules as ligands (53). In mammals, antibody Fc receptors are encoded mainly within the Fc receptor gene family (on human chromosome 1), but there are two examples of Fc receptors encoded within the mammalian LRC: the IgA receptor Fc␣RI (CD89) and bovine Fc␥RIIb (54). There is a significant difference between the way in which human Fc␣RI interacts with IgA and how the "classical" Fc receptors belonging to the FcR gene cluster interact with IgG and IgE. Crystal structures of these Fc-receptor complexes show that in both IgG and IgE, the binding site lies in a similar region at the N terminus of the homologous C␥2 and C⑀3 domains, respectively, whereas IgA interacts with its receptor through a site close to the junction between C␣2 and C␣3 (55). (It is not yet known to which domains of IgG bovine Fc␥RIIb binds, but the interaction does involve domain 1 of the receptor, which is characteristic of LRC receptor interactions, in contrast to FcR, which involve domain 2.) The chicken LRC sequence is incomplete, but already 103 genes highly related to each other have been identified (56); currently, only CHIR-AB1 has a known ligand. As CHIR-AB1 is more closely related to Fc␣RI than the Fc receptors for IgG or IgE, chicken IgY may bind to its receptor in a similar manner to that in which human IgA binds to Fc␣RI. If so, and if the receptor binding site in IgY lies between C3 and C4 (and not at the N-terminal region of C3), this would be consistent with our observation that the presence of C2 plays no role in the binding kinetics.
Intriguingly, the FcR family in chicken appears to be represented by just a single gene (57,58), and our preliminary experiments show that this receptor does not bind IgY. 5 The mam-malian FcR complex contains many receptor homologues whose ligands are not antibodies, and these receptors are thought to play a role in immunoregulation (59). These observations, and the data presented in this report, lead us to speculate that Fc receptor function migrated from the LRC to the FcR locus (which subsequently expanded) during the last 310 million years. The migration of Fc receptor function may have been a consequence of the need to shift the binding site in the antibody Fc from the C3/C4 interface in the IgY-like ancestor of IgG and IgE, due to competition from bacterial Fc-binding proteins. Such proteins are known to bind to the C␣2/C␣3 interface in IgA (60) and the C␥2/C␥3 interface in IgG (61,62), thereby disabling the immune response by blocking FcR interactions. This selection pressure has led to co-evolution of IgA and its receptor (63); a receptor mutating too slowly to keep up with changing bacterial peptide sequences becomes redundant and may persist only as a pseudogene (as in rabbits) (63) or be eliminated altogether (as in mouse) (64). The IgY-like ancestor of IgG and IgE may have overcome this selection pressure by shifting the binding site to the N-terminal region of the domain that evolved into C␥2 and C⑀3. Whatever the reason for the persistence of the C2 domains in IgY and its ancestors, its presence in the ancestor of IgE served as a preadaptation for the evolution of an antibody domain (C⑀2) that contributes to the uniquely slow dissociation rate and high affinity for receptor and the consequent advantages this confers for defense against parasites.