INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Humans and closely related mammals possess natural anti-alpha galactosyl (Gal1-3Gal) antibodies (22). It is suggested that these human antibodies are developed in response to normal bacterial flora in the intestine (8) in a way similar to the blood group antigen-specific antibodies (34). The anti-alpha galactosyl (Gal1-3Gal) antibodies are found in humans, apes, and Old World monkeys but not in New World monkeys and other mammals (7, 9). This is consistent with the fact that humans, apes, and Old World monkeys lack a functional alpha (1-3) galactosyltransferase (14, 19) and therefore do not possess the Gal1-3Gal structure on their tissues. The opposite is true in other mammalian species, including cattle (13, 18), and the Gal1-3Gal structure is therefore a self antigen in these species.

The human anti-Gal1-3Gal antibodies have an unusually high titer, accounting for as much as 1% of total serum immunoglobulin G (IgG) (20 to 100 µg/ml) (5). They have been suggested to play an important role in innate immunity. They are responsible for inactivation of retrovirus and other enveloped viruses of animal origin by human serum (21, 28, 32) and rejection of pig-to-human xenografts (23). They may also be involved in preventing microbial and parasite infections (1, 30).

Although the human anti-alpha galactosyl antibodies react most avidly with the Gal1-3Gal structure, they also react strongly with melibiose (Gal1-6Glc) (6, 35). As a result, melibiose has been used as a ligand for the affinity purification of these human antibodies. Recently, antimelibiose antibodies have been detected in cattle and chickens (26, 27), suggesting that there are similar anti-alpha galactosyl antibodies in animals. However, their specificity and distribution have not been extensively studied and no comparison has been made with the Gal1-3Gal epitope. Here we report the characterization of the specificity and distribution of the bovine antimelibiose antibodies and their potential involvement in host defense against microorganisms.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Bacteria. Streptococcus agalactiae types II (ATCC 27541), IV (ATCC 49446), and V (ATCC 700048) were obtained from the American Type Culture Collection. Type II is a bovine strain isolated from a cow with mastitis, whereas types IV and V are strains from humans. A clinical isolate of S. agalactiae obtained from the milk of a cow suspected of having mastitis and designated 5247 was provided by the Texas Veterinary Medical Diagnostic Laboratory (College Station). Escherichia coli (ATCC 25922 and ATCC 35215), Staphylococcus aureus (ATCC 29213 and ATCC 25923), and Pseudomonas aeruginosa (ATCC 27853) were also used. All bacteria were maintained on blood agar. They were cultured in tryptose broth for 24 h at 37°C before used in dot blot analysis.

Serum samples. One fetal, three newborn (1 to 10 days old), and five calf (~6 months old) bovine pooled serum samples were obtained from Gibco-BRL (Grand Island, N.Y.). Nineteen adult sera were obtained from cows at ages of 1 to >10 years in the Veterinary Medicine Park, Texas A&M University. All adult samples were collected in October 1996. An older sample collected in May of 1992 from one cow was also used and designated 16-1. The new sample from this same cow was designated 16-2. All serum samples were stored at 20°C. A pooled adult sample was made by mixing equal volumes of individual adult samples, excluding samples 1, 6, 12, and 16-1 because of the limited amounts available. Two pooled adult human serum samples were used, one (I) from Sigma Chemical Co. (St. Louis, Mo.) and another (II) from Scantibodies Lab, Inc. (Santee, Calif.).

ELISA. An indirect enzyme-linked immunosorbent assay (ELISA) was used to measure serum antimelibiose antibodies. The assay was performed in a reagent excess manner; both the antigen and secondary antibody were used in excess amounts as determined by titration assays. Plates (NUNC, Inc., Naperville, Ill.) were coated overnight at 4°C with melibiose (Gal1-6Glc)-bovine serum albumin (BSA; Sigma Chemical Co.) or Gal1-3Gal-BSA (V-LAB, Covington, La.) at a concentration of 5 µg/ml in 50 mM carbonate buffer (pH 9.6) (100 µl/well). Plates were washed three times with wash buffer (phosphate-buffered saline [PBS] with 0.02% NaN₃) and then blocked for 1 h with dilution buffer containing 3% BSA (gamma globulin free; Sigma Chemical Co.), 0.05% Tween 20, and 0.02% NaN₃ in PBS. Serum samples diluted in the same buffer were added to the plates (100 µl/well) and incubated at room temperature for 2 h. Following washing as described above, affinity-purified alkaline phosphatase-conjugated anti-human IgG whole molecule (Sigma Chemical Co.), anti-bovine IgG whole molecule (Sigma Chemical Co.), or alkaline phosphatase-conjugated anti-bovine IgM heavy (µ)-chain antibodies (KPL, Gaithersburg, Md.) at a 1:1,000 dilution were added (100 µl/well) and incubated at room temperature for 1 h. After being washed, each well on the plate received 100 µl of the substrate p-nitrophenylphosphate prepared in accordance with the instructions from the manufacturer (Pierce, Rockford, Ill.). The reaction was stopped after a 10-min incubation by addition of 50 µl of 2 M NaOH. Optical density (OD) at 405 nm was determined using a Dynatech microplate reader (MR600).

Statistical analysis of ELISA data. A pooled calf serum sample (no. 2) was used as a reference for determining titers of individual samples. The titration endpoint was arbitrarily chosen as the highest serum dilution that gave an OD twofold higher than that of the background (fetal bovine serum). The endpoint was assigned as 1 titer unit/100 µl. This gave calf sample 2 a unit range of 1 (1:640) to 64 (1:10). The OD values were plotted against the logarithm of the titer units. A linear fit was made, and the log unit value of each sample was obtained by extrapolating against the regression line. The titer units per 0.1 ml of serum were calculated by an anti-log transformation, followed by multiplication with the sample dilution. All samples were tested at a 1:80 or 1:160 dilution so that OD values were within the range of the standard curve. The geometric means of different age groups were compared by the unpaired Student t test.

Measurement of anti-alpha galactosyl IgA, IgG1, and IgG2. Anti-alpha galactosyl IgG1, IgG2, and IgA were measured by an indirect ELISA using affinity-purified anti-bovine IgA, IgG1, and IgG2 antibodies conjugated to horseradish peroxidase (Bethyl Laboratories, Montgomery, Tex.). The ELISA was performed as described above in a reagent excess manner. The conjugates or secondary antibodies were used at a 1:1,000 dilution. All serum samples were diluted 1:40. Tetramethylbenzidine (Pierce) was used as the substrate. The reaction was stopped after a 30-min incubation by addition of 100 µl of 2 M H₂SO₄. OD at 450 nm was determined using a Dynatech microplate reader (MR600).

Affinity chromatography. A 10-ml column of melibiose-agarose (Sigma Chemical Co.) was used for isolation of the bovine antimelibiose antibodies. The column was first washed with 100 ml of TN buffer (20 mM Tris, 150 mM NaCl, pH 7.4) containing 0.02% NaN₃. Bovine serum (20 ml) mixed with an equal volume of the same buffer was loaded onto the column at a rate of 15 ml/h. The column was then washed with 200 ml of the buffer, and bound proteins were eluted with 200 mM melibiose in the same buffer. One-milliliter fractions were collected. Those fractions containing detectable proteins were pooled, dialyzed against TN buffer, and concentrated with polyvinylpyrrolidone (molecular weight, 360,000; Sigma Chemical Co.). Protein concentrations were determined by the bicinchoninic acid protein assay (Pierce).

Gel electrophoresis and immunoblotting. Electrophoretic analysis of isolated proteins was carried out under denaturing conditions as described previously (17). Proteins were detected by Coomassie blue staining.

Dot blot assay. A dot blot assay was used to determine if the purified bovine antimelibiose antibodies reacted with bacteria. Bacteria were grown in tryptose broth at 37°C for 24 h. They were washed three times with PBS and suspended in PBS at 10⁸/ml. Bacteria were killed by incubation at 60°C for 1 h. One and one-half microliters of the killed-bacterial preparations were spotted onto dry cellulose membranes and air dried. Coomassie blue staining was performed to ascertain that the amount of bacteria attached to the membrane was consistent. The membranes were blocked with 3% BSA and 0.05% Tween 20 in PBS at room temperature for 2 h. They were then probed at room temperature for 1 h with purified bovine antimelibiose antibodies (1 µg/ml) in PBS containing 1% BSA and 0.05% Tween 20. Melibiose was used as an inhibitor at a concentration of 10 mM. The remaining steps with affinity-purified, alkaline phosphatase-conjugated anti-bovine IgG whole molecule and BCIP-nitroblue tetrazolium substrate were the same as those described above.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

The bovine anti-alpha galactosyl antibodies in serum were measured with an ELISA using melibiose (Gal1-6Glc)-BSA as an antigen (see Materials and Methods). The serum titration curves of seven serum samples, two pooled newborn, two pooled calf, two adult, and one pooled adult, are presented in Fig. 1. They were parallel to each other, suggesting that the measured antibodies reacted against the same epitope. Calf sample 2, a pooled serum sample, was used as the reference sample and included in all assays. The specificity of this assay was confirmed by inhibition assays with melibiose (see below).

View larger version (30K): [in this window] [in a new window]   FIG. 1.   Titration curves of seven bovine serum samples (two newborn, two calf, and three adult).

Specificity of bovine anti-alpha galactosyl antibodies. Inhibition assays with various monosaccharides, disaccharides, and oligosaccharides were used to determine the specificity of the bovine anti-alpha galactosyl antibodies. The results are shown in Table 1 and Fig. 2. Based on the lowest concentrations showing a 50% reduction in OD, melibiose was the strongest inhibitor, followed equally by stachyose (Gal1-6Gal1-6Glc1-2Fru) and raffinose (Gal1-6Glc1-2Fru) and then almost equally by Gal1-6Gal, Gal, and Gal1-2Gal (Table 1). Others, including Gal1-3Gal, Gal1-4Gal, and lactose (Gal1-4Glc), exhibited only minor inhibition (Table 1). Since the Gal1-6Glc or Gal1-6Gal structure was present at the nonreducing end of stachyose or raffinose, these results suggest that these bovine anti-alpha galactosyl antibodies preferentially react with galactose in the 1-6 linkage (Gal1-6Glc or Gal1-6Gal).

View larger version (26K): [in this window] [in a new window]   FIG. 2.   Sugar inhibition assays. Calf serum sample 2 was used at a 1:80 dilution.

View larger version (19K): [in this window] [in a new window]   FIG. 3.   Specificity comparison of human and bovine anti-alpha galactosyl antibodies. Panels: A, bovine sera (calf sample 2 and pooled adult sample); B, human sera (samples I and II). Each serum sample was tested against both antigens on the same plate, and each plate was used for two serum samples (one bovine and one human).

Presence of bovine anti-alpha galactosyl antibodies in different age groups. Twenty-seven bovine serum samples of three age groups were tested (Fig. 4). They included 3 pooled newborn (1 to 10 days old) samples, 5 pooled calf (~6 months old) samples, and 19 adult (>1 year old) samples. All of them, including those from the newborn, contained the antimelibiose antibodies. The antibody titers gradually increased from newborn to adult samples (Fig. 4), although the differences were not statistically significant. Titers in individual adult samples varied widely, with an average of 754 and a standard deviation of 546 (Fig. 4), although they exhibited a normal distribution (95% of the values were within the mean ± 2 standard deviations). Those at an age of 5 to 9 years (no. 12 to 16-1) exhibited the highest titers. It was noted that the titer of the freshly obtained sample (16-2) was fourfold lower than that of the one (16-1) obtained 4 years ago. By 1996, this cow was more than 10 years old.

View larger version (20K): [in this window] [in a new window]   FIG. 4.   Presence of anti-alpha galactosyl antibodies in bovine serum samples. Samples were tested at a 1:80 or 1:160 dilution. The titer unit was obtained by extrapolation against the regression curve of the reference sample (calf sample 2; see Materials and Methods). The horizontal bars indicate the average titers for each age group, and the error bars show the standard deviations.

Purification of bovine anti-alpha galactosyl antibodies. The anti-alpha galactosyl antibodies were purified by affinity chromatography using melibiose-agarose. Gel electrophoretic analysis of proteins eluted from the column with melibiose showed that the IgG heavy chain was the dominant band (Fig. 5, lane 1). The antibody heavy and light chains were identified by immunoblotting (Fig. 5, lanes 2 and 3). The IgM heavy chain was barely detectable, compared to the IgG heavy chain (Fig. 5). This observation is consistent with the results described above, suggesting that the bovine antimelibiose antibodies are mainly of the IgG isotype. The protein close to 200 kDa (Fig. 5, lanes 1 and 2) was most likely the undenatured IgG whole molecule, since it reacted with anti-IgG whole molecule antibody (Fig. 5, lane 2) but not with anti-IgM heavy (µ)-chain antibody (Fig. 5, lane 3).

View larger version (38K): [in this window] [in a new window]   FIG. 5.   Gel electrophoresis and immunoblotting analysis of bovine serum proteins eluted from the melibiose-agarose column. Lanes: 1, proteins stained with Coomassie blue; 2, proteins probed with alkaline phosphatase-conjugated anti-bovine IgG whole molecule; 3, proteins probed with alkaline phosphatase-conjugated anti-bovine IgM heavy (µ) chain. The open arrowheads indicate the IgG whole molecule, the closed arrowheads indicate the IgM heavy chain, the empty arrows indicate the IgG heavy chain, and the filled arrows indicate the  light chain. Molecular size standards in kilodaltons are indicated on the left.

Anti-alpha galactosyl IgA, IgG1, and IgG2. The anti-alpha galactosyl IgA, IgG1, and IgG2 antibodies were examined using 2 pooled newborn samples, 2 pooled calf samples, and 1 pooled adult sample along with 10 individual adult samples. The anti-alpha galactosyl antibody was primarily of the IgG1 isotype in all samples, except for two adult samples, 7 and 10, which had more IgG2 than IgG1 (Fig. 6A and B). However, it was noted that the IgG2 level was significantly increased in calves and adults compared to newborns (P < 0.05). In fact, the increase in IgG2 accounted for 81% of the total IgG increase in calves and adults. IgA was present in all of the samples, and its level was much higher in newborns than in calves or adults (Fig. 6A and B).

View larger version (55K): [in this window] [in a new window]   FIG. 6.   Measurement and comparison of anti-alpha galactosyl IgA, IgG1, and IgG2 responses. The anti-alpha galactosyl IgA, IgG1, and IgG2 antibodies were measured using class- or isotype-specific antibodies conjugated to horseradish peroxidase as described in Materials and Methods. All sera were diluted 1:40. The serum sample numbers correspond to those in Fig. 4. Panels: A, Anti-alpha galactosyl IgA, IgG1, and IgG2 levels expressed as direct OD values; B, Anti-alpha galactosyl IgA, IgG1, and IgG2 levels expressed as percentages.

Reaction of purified bovine anti-alpha galactosyl antibodies with S. agalactiae. Many bacteria, including Streptococcus spp., are known to have immunoreactive terminal galactose residues in capsular polysaccharides (2, 20). S. agalactiae and group B streptococci are major causes of human neonatal infections (24) and also bovine mastitis (16). To determine if the bovine anti-alpha galactosyl antibodies react with these bacteria and others, three American Type Culture Collection S. agalactiae strains (serotypes II, IV, and V) and a bovine S. agalactiae clinical isolate (5247) were tested in a dot blot test along with strains of E. coli, S. aureus, and P. aeruginosa. The type II S. agalactiae strain is an isolate from a bovine with mastitis, whereas the type IV and V strains are from humans. The results of the dot blot tests showed that these bovine antibodies reacted with the type II bovine strain and the clinical bovine isolate (5247) (Fig. 7A). No reaction was detected with the type IV or V human strain (Fig. 7A) or with the E. coli, S. aureus, and P. aeruginosa strains tested (data not shown). The reaction with bovine strains was inhibited by melibiose (Fig. 7B).

View larger version (57K): [in this window] [in a new window]   FIG. 7.   Reaction of bovine anti-alpha galactosyl antibodies with S. agalactiae. Bacteria spotted onto the membrane were probed with bovine anti-alpha galactosyl antibodies in the absence (A) or in the presence (B) of melibiose (10 mM).

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

The present experiments demonstrated that cattle possess an anti-alpha galactosyl antibody with a serum concentration close to that of human anti-Gal1-3Gal antibodies (20 to 100 µg/ml). However, the specificity of these bovine antibodies is distinctly different from that of human antibodies. The bovine anti-alpha galactosyl antibodies preferentially reacted with the galactose in 1-6 linkage, i.e., Gal1-6Glc or Gal1-6Gal.

Stachyose and raffinose are the second strongest inhibitors (next to melibiose) (Table 1). It is not known why stachyose (Gal1-6Gal1-6Glc1-2Fru) and raffinose (Gal1-6Glc1-2Fru) are less efficient inhibitors than melibiose, especially the latter, which has a structure identical to that of melibiose (Gal1-6Glc) at the nonreducing end. It may be that the third sugar affects the binding; i.e., the third sugar found in stachyose and raffinose may not be the right one for the optimal binding of these antibodies. Alternatively, it could be due to the fact that larger molecules are simply less efficient competitors than the smaller ones bearing the same functional domain. Similar observations have been made with human anti-alpha galactosyl antibodies (35).

Gal1-6Gal and Gal1-2Gal were moderate inhibitors (Table 1), suggesting that a significant reaction may also be found with these two structures. On the other hand, Gal1-3Gal, Gal1-4Gal, and lactose (Gal1-4Glc) were weak inhibitors. Together, it seems that these bovine antibodies favor the linkage positions more than the anomeric configuration of the linkage since Gal1-6Gal is a more efficient inhibitor than Gal1-3Gal and Gal1-4Gal.

The lack of a strong reaction of bovine anti-alpha galactosyl antibodies to the Gal1-3Gal structure is consistent with the fact that the Gal1-3Gal epitope is a self antigen in animals, including cattle, due to the presence of alpha (1-3) galactosyltransferase (13). By the same token, cattle might lack the Gal1-6Glc or Gal1-6Gal structure in their tissues; i.e., they might lack an alpha (1-6) galactosyltransferase. Whether humans have such an enzyme also remains to be determined, although the Gal1-6Gal structure has been reported in human tumor cells (10, 15).

The origin of these bovine antibodies is not known. They could develop in response to any environmental agents bearing this epitope. The positive reaction of these antibodies with bovine S. agalactiae suggests that responses to bacteria, either pathogenic or nonpathogenic, can be one of the sources, although a positive reaction itself does not necessarily prove the possession of an identical epitope (Gal1-6Glc or Gal1-6Gal) by the bacteria. S. agalactiae is a primary pathogen of bovine mastitis and is of major economic concern in the dairy industry (16). It is also the leading cause of human neonatal mortality and morbidity (24). Although all serotypes can be found in both human and animals, serotype II is one of the most frequently isolated serotypes (II and III) in bovine mastitis cases (12, 31).

Structures of capsular polysaccharides of bovine S. agalactiae have not been elucidated. Studies on the human strains of all six serotypes showed that there is no terminal (1-6)-linked galactose residue present in the capsular polysaccharides (11, 33). However, terminal galactose residues in 1-6 linkage are present in the type II strain (11). Gal1-6Gal was found to be a moderate inhibitor of these bovine anti-alpha galactosyl antibodies. Thus, it seems consistent with the finding that these bovine antibodies reacted negatively with the human type IV and V S. agalactiae strains but positively with the bovine type II strain. Studies with S. cricetu and S. sobrinus have also shown that only certain serotypes possess the terminal galactose residues in a particular linkage (2, 20).

Bovine antimelibiose antibodies are clearly present in newborns and are most likely from the mother's colostrum. The titers of these antibodies varied widely in adults, which may reflect various degrees of exposure to microorganisms bearing this epitope or the related epitopes. It was noted that among the adults (>1 year old), those 5 to 9 years old exhibited the highest titers. Furthermore, the titer may vary in an individual at different times, as shown with two samples (16-1 and 16-2) collected 4 years apart from the same cow. The titer of the second sample (16-2), collected when the cow was >10 years old, was severalfold lower than that of the first one (16-1). This was not the case when antimannan (yeast) antibodies were measured in the same samples; the titer in sample 16-2 was actually higher than that in sample 16-1 (25).

There are two IgG isotypes in cattle, IgG1 and IgG2 (3). IgG2 may be further divided into IgG2a and IgG2b. IgG1 was found to be the primary anti-alpha galactosyl IgG in newborns, calves, and adults. IgG1 is the major antibody in bovine serum, with an IgG1/IgG2 ratio of 1.2:1 (3). In cattle, there is no antibody transfer through the placenta and newborns obtain maternal antibodies through colostrum. IgG1 is also the major antibody in the colostrum (IgG1/IgG2 ratio, 16:1) and in milk (IgG1/IgG2 ratio, 116:1) (3). IgA is a very minor immunoglobulin compared to IgG in both serum and colostrum, although the IgA concentration in colostrum is significantly higher than that in serum (colostrum IgA/serum IgA ratio, 15:1) (3). This seems consistent with the finding that IgG1 was the major anti-alpha galactosyl antibody in newborns, calves, and adults. However, a significant increase in anti-alpha galactosyl IgG2 was observed in calves and adults, which accounted for 81% of the total anti-alpha galactosyl IgG increase compared to newborns. Whereas IgG1 has been, in general, found to be the primary IgG isotype induced against infections or immunizations, the IgG2 response varies. Bovine IgG2 appears to be more related to the innate immunity functions. Under T-cell-independent conditions, gamma interferon increases the B-cell production of IgG2 but not that of IgG1 (4) and only IgG2 has receptors on resting neutrophils and monocytes (29, 36). Thus, an increase in the IgG2 level may enhance the innate immunity response and the immune surveillance of the host against microorganisms bearing this epitope (Gal1-6Glc or Gal1-6Gal) or a related epitope.