Primary Epitopes of Chicken Egg Yolk Antibodies to Peptidophosphogalactomannan

Abstract

Egg yolks from hens immunized with peptidophosphogalactomannan (pPGalManⁱⁱ), which contains 10 phosphocholine diester residues and is secreted by Penicillium fellutanum, contain antibodies against 5-O-β-d-galactofuranosyl epitopes. These epitopes were the only significant determinants in pPGalManⁱⁱ. Approximately 60-fold less pPGalManⁱⁱ (1.6 μM galactofuran chains) was required for 50% inhibition than galactofurano-oligosaccharides or pPGalMan containing two galactofuranosyl residues per chain.

Filamentous fungi produce soluble extracellular polysaccharides and glycopeptides (1, 9, 10, 14, 19, 21, 23). Many of these polymers have active antigenic determinants (3, 14, 17, 20, 27). Penicillium fellutanum (formerly Penicillium charlesii) peptidophosphogalactomannans (pPGalMan; M_w, 25,000 to 70,000) (9, 19, 21, 23, 25, 26, 32) contain a mannan with about 80 α-1,2- and α-1,6-mannopyranosyl residues and 12 small manno-oligosaccharidyl units, each attached to a 3-kDa peptide (Fig. 1). Eight to ten 5-O-β-d-galactofuranosyl-containing chains with 2 to 20 residues branch from the mannan. pPGalManⁱⁱ and pPGalManⁱⁱⁱ (26, 31) contain approximately 10 and 2 phosphocholine diester residues, respectively, and a variable number of galactofuranosyl-6-O-phosphodiester residues (5).

FIG. 1.

Diagram of pPGalMan. The mannopyranosyl residues in each tetrasaccharide are attached by α-1,2 linkages and the tetrasaccharides are attached by α-1,6 linkages. The mannan is attached to the peptide by an O-glycosidic linkage to a seryl residue. Manno-oligosaccharides are attached to seryl and threonyl residues. An average of one galactan chain branches from each manno-octasaccharide and one phosphocholine phosphodiester is attached to C-6 of a mannopyranosyl residue.

Sera from rabbits immunized with whole-cell preparations from P. fellutanum reacted with galactofuranosyl-containing heteropolysaccharide (20). Sera from guinea pigs injected with purified pPGalMan conjugated to bovine gamma globulin reacted weakly to manno-oligosaccharides of pPGalManⁱⁱ (11) and were unreactive to galactofuranosyl residues. Soluble pPGalMan did not elicit antibody in any of several species. This preparation, pP(Gal₂)Man, was later shown to contain an average of two galactofuranosyl residues per galactan chain (unpublished data).

Antisera from rabbits immunized with extracellular polysaccharides of Penicillium sp. cell walls react with synthetic β-d-galactofuranosyl-containing oligosaccharides (17). The 5-O-β-d-galactofuranosaccharides resulted in the most inhibition.

Antibodies that react specifically with furanosyl residues of parasites are of increasing clinical importance (4, 6–8, 14–16, 22, 28–30).

The purpose of this investigation was to determine if stable antibody could be elicited from purified glycopeptides, such as pPGalManⁱⁱ in phosphate-buffered saline (PBS) without adjuvant, and to determine the polymers’ epitope(s). Laying hens challenged with immunogenic substances during the laying season produce eggs that contain immunoglobulin Y (IgY), which is similar but not identical to IgG, in their yolks. Antibody is selectively deposited in egg yolk and is obtained by noninvasive means (2, 18).

Preliminary experiments.

No immunological response was obtained in laying hens injected subcutaneously and in the footpad at weeks 1 and 3 with solutions of pPGalManⁱⁱ (200 μg/ml in PBS) and with whole P. fellutanum cells at weeks 6 and 9. A response to subcutaneous injections of rabbit IgG in PBS was obtained in these hens. In contrast, other chickens responded to a course of two subcutaneous and two intravenous injections of either pPGalManⁱⁱ or pPGalManⁱⁱⁱ in PBS. The immune responses to pPGalManⁱⁱ and pPGalManⁱⁱⁱ were similar. Yolks from eggs stored at 4°C for a year retained antibody with little loss of activity. In these experiments, anti-pPGalMan activity was tested routinely by an enzyme-linked immunosorbent assay (ELISA) procedure (24, 27) in microtiter plates (Dynatech Laboratories, Inc.) coated with 0.4 μg (0.057 nmol) of either pPGalManⁱⁱ or pPGalManⁱⁱⁱ (26) in 0.14 M NaCl–0.02% NaN₃. After incubation for 24 h at 4°C, the wells were washed with PBS containing 0.05% Tween 20. Unoccupied wells were blocked with 1 mg of bovine serum albumin in 0.1 ml of a solution of PBS, 0.01% NaN₃, and 0.05% Tween 20. Incubation at 24°C for 45 min followed. Plates were washed with PBS-NaN₃-Tween 20. Primary antibodies, prediluted with PBS, were added to all wells except those in the row that served as the secondary-antibody control. Plates were incubated for 60 min at 24°C. After the wells were washed, the quantity of chicken anti-pPGalMan antibody adsorbed to pPGalManⁱⁱ in each well was determined with rabbit anti-chicken IgG (whole molecule) alkaline phosphatase conjugate with p-nitrophenylphosphate as the substrate in 10% diethanolamine buffer (pH 9.8)–0.2% NaN₃. p-Nitrophenol released in each well was quantified with a Bio-Rad ELISA model 2550 enzyme immunoassay reader set at 405 nm.

Purification of chicken egg yolk anti-P. fellutanum antibody.

Antibodies were fractionated by polyethylene glycol precipitation, hydrophobic-interaction chromatography, and gel permeation chromatography (12). Anti-pPGalManⁱⁱ activity from permeation chromatography resulted in a 31-fold increase in ELISA units per microgram of protein. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (13) showed anti-pPGalManⁱⁱ activity at 28 and 62 kDa.

Immunochemical studies.

The reaction between pPGalManⁱⁱ or pPGalManⁱⁱⁱ and anti-P. fellutanum pPGalMan antibodies was quantified with 5 μg of protein/well. pPGalManⁱⁱ or pPGalManⁱⁱⁱ (0 to 1 μg/well) was used in an indirect ELISA system. Both pPGalMan species bound to Immulon wells in a hyperbolic concentration-dependent manner. Half saturation of the wells occurred with 26 nmol of either pPGalMan species (data not shown). Approximately 57 nmol (0.4 μg/well) of pPGalManⁱⁱ or pPGalManⁱⁱⁱ was used to coat the wells.

Competitive inhibition experiments with a range of concentrations of soluble phosphogalactomannan (PGalManⁱⁱ) or pPGalManⁱⁱ as the inhibitor of antibody interaction with bound pPGalManⁱⁱ or pPGalManⁱⁱⁱ, respectively, showed 50% inhibition at 0.14 and 0.16 μM (1.4 and 1.6 μM galactofuran chains), respectively (Table 1). This suggests that phosphocholine phosphodiester is not a major epitope because pPGalManⁱⁱ, which contains at least fivefold more phosphocholine phosphodiester than pPGalManⁱⁱⁱ (26, 31), is not a better inhibitor than pPGalManⁱⁱⁱ. The epitope(s) on pPGalManⁱⁱ was determined with fragments derived by chemical or enzymatic degradation of pPGalManⁱⁱ. A range of concentrations of each fragment was tested as a hapten inhibitor of binding of anti-pPGalManⁱⁱ antibodies to pPGalManⁱⁱ in a competitive ELISA inhibition system. The concentration of inhibitor or galactofuran chains required to inhibit 50% of antibody binding to Immulon-bound pPGalManⁱⁱ (Table 1) was determined from plots of the percentages of inhibition versus log micromolar values of inhibition or chain.

TABLE 1.

Inhibition of antibody binding to pPGalManⁱⁱ by modified pPGalManⁱⁱ and by oligosaccharide fragments^a

Inhibitorb	50% Inhibitory concn (μM)		Residues/ chain (n)
	Saccharide	Galactofuran chains
pPGalManii	0.16	1.6	20
pPGalManii in PBS	0.13	1.3	20
PGalMan	0.14	1.4	12
pPMan	NI	NI	N/A
Peptide	NI	NI	N/A
pP(Gal2)Man	9.8	98	2
Galactofurano-oligo- saccharides
Tetrasaccharide(s)	55	55	4
Trisaccharide(s)	100	100	3
Disaccharide	180	180	2
1-O-β-CH3-d-Man	3,600	N/A	N/A
Anionic saccharide	125	125	2

^aNI, no inhibition; N/A, not applicable. 

^bMolecular masses are as follows: pPGalManⁱⁱ, 65 kDa; PGalMan, 62 kDa; pPMan, 18.6 kDa; and pP(Gal)₂Manⁱⁱ, 21.9 kDa. 

Peptide or peptidophosphomannan (pPMan), obtained by treatment of pPGalManⁱⁱ with dilute acid, did not inhibit the immune response. In contrast, 9.8 μM pP(Gal₂)Manⁱⁱ (98 μM galactofuran chain) resulted in 50% inhibition of pPGalManⁱⁱ binding to anti-pPGalManⁱⁱ. Ten of 20 galactofuranosyl residues in pP(Gal₂)Manⁱⁱ are phosphodiesters (5). Considering that each chain contains two galactofuranosyl residues, the neutral galactofuranotrisaccharide binds with about the same avidity as the average of each chain in pP(Gal₂)Manⁱⁱ. This is further evidence that mannopyranosyl-6-O-phosphocholine phosphodiester in pPManⁱⁱ is not a significant epitope. Furthermore, pPMan, which also contains phosphodiester residues, was not inhibitory.

Anionic galactofurano-oligosaccharide, obtained from an anion-exchange resin following dilute-acid treatment of pPGalManⁱⁱ, was comparable to neutral galactofuranotrisaccharide as an inhibitor. Considered collectively, the data suggest that 5-O-β-d-galactofuranosyl residues are the primary epitopes in egg yolks from chickens challenged with pPGalManⁱⁱ. The peptide region has no influence on antibody binding. The concentration of galactofurano-oligosaccharide or galactan chains in pP(Gal₂)Manⁱⁱ required for 50% inhibition of anti-pPGalManⁱⁱ binding is more than 60-fold greater than that in pPGalManⁱⁱ or PGalManⁱⁱ. This suggests that galactofuran chains with a large number of residues have greater avidity for anti-pPGalManⁱⁱ.

Although the chicken egg yolk anti-pPGalManⁱⁱ antibody-antigen interaction in this study was not as sensitive as that from the rat on fungal galactomannan (27), the use of chickens may have utility in some situations in which antibody can be stored in the egg for long periods. The noninvasive means of obtaining antibody, the ease of isolation and purification of antibody, and the fact that adjuvant is not required to elicit significant antibody activity all may be of value in some situations.