Two Murine Monoclonal Antibodies against Serogroup E Salmonellae

Prior to the reclassification ofSalmonella enterica subsp. enterica serogroup E, serogroup E₁, E₂, E₃, and E₄ strains were identified separately by serogroup-specific antisera to O antigens 10, 15, 15 plus 34 (15,34), and 19, respectively (2). Serogroup E₂ refers to a group of serogroup E₁ strains that are lysogenized by phage ɛ15, while strains from serogroup E₃ are serogroup E₂strains that have been further lysogenized by phage ɛ34. These two serogroups are now grouped into a single serogroup, E₁, while serogroup E₄ remains unchanged. O factors 15 and 15,34 in the phage-lysogenized strains replace the original O factor 10 in the parental group E₁ strain. Lysogenic conversion by ɛ15 results in a change in the diester linkage between the galactose and mannose residues (Gal-Man) of the repeating trisaccharide unit on the backbone of the polysaccharide (PS) moiety from α-1,6 to β-1,6. It also results in a loss of an O-acetyl group from the Gal residue (Fig. 1). This change is believed to be responsible for the loss of O-10, and in its place, these organisms express a new O antigen, O-15. Some of these organisms may additionally harbor another phage, ɛ34, and this imparts to these organisms an additional O antigen 34 specificity which is related to the glycosylation of the Gal residue on the ɛ15-lysogenized variants' lipopolysaccharide (LPS) backbone (8).

• Open in new tab
• Download powerpoint

Fig. 1.

Structure of an O-antigenic polysaccharide unit ofS. enterica subsp. enterica serogroup E LPS. Gal, galactose; Man, mannose; Rha, rhamnose. Footnotes: a, X = O-acetyl→6 (nonlysogenic serogroup E₁), X − (ɛ15-lysogenized serogroup E₁), d-glucose α1→4 (ɛ15- and ɛ34-lysogenized serogroup E₁), ord-glucose α1→6 (serogroup E₄); or anomeric linkage is α for nonlysogenic serogroup E₁ and serogroup E₄, β for lysogenic serogroup E₁.

Serotyping of O antigens to identify the phage-modifiedSalmonella variants within serogroup E₁ is still of epidemiological significance, especially for public health reasons. This work is complicated by the fact that strains may express the O factors 1, 3, 10, 15, 19, and 34 in different combinations (i.e., 3,10; 3,15; 3,15,34; 1,3,19; 1,3,10,19; and 1,3,15,19) (2, 7). With some single factor O antisera being withdrawn from the market by commercial producers, more-detailed serotyping can be achieved only at reference laboratories at which the tedious task of production and purification of O-factor-specific antisera is still carried out with animals. Monoclonal antibodies (MAbs), with their exquisite specificity, can be an efficient alternative to conventional antisera in Salmonella serogroup differentiation (10, 13). Their monospecificities allow detailed mapping of epitopes within the LPS structures, and their simple, high-yielding production makes generation of large volumes of these highly specific reagents of consistent quality a simple task. In addition, MAbs can be produced in vitro by using artificial capillary systems (1) and thus will not require the continued use of laboratory animals in antibody production once the hybridoma cell line is established.

The use of MAbs in Salmonella serology, however, has not gained broad popularity. This is in part due to the fact that their reaction patterns are not identical to those of conventional antisera. The intrinsic high specificity of MAbs is the main factor contributing to this discrepancy. Within a Salmonella colony there may exist subpopulations exhibiting different somatic antigens (2, 5). This phenomenon, called from variation, is usually associated with postpolymerization modifications such as acetylation and glycosylation (2, 4). Once such variation is introduced, the monospecific MAbs may no longer be able to react with the subpopulation that has undergone this change, whereas a conventional antiserum will intrinsically be able to compensate because of its polyclonal nature.

(This work was conducted by S. P. Ng in partial fulfillment of the requirements for a Ph.D. from the University of Hong Kong, Hong Kong, People's Republic of China.)

An immunoglobulin G1 MAb, MO15, was produced by the method of Kohler and Milstein (6), using S. enterica subsp.enterica serovar London var.15⁺, a phage ɛ15-lysogenized serogroup E₁Salmonellastrain, as the immunizing strain. The serogroup specificity of MO15 was confirmed by enzyme-linked immunosorbent assay (ELISA) (11) and immunoblotting against polyacrylamide gel electrophoresis-resolved purified LPS (3, 16, 18). The epitope structure of this MAb is best described by O antigen 15, with its specific anomeric linkages of Gal to Man to rhamnose on the backbone of the ɛ15-lysogenized serogroup E₁ LPS (Fig. 1). This MAb epitope structure is reflected in the exclusive reaction of MO15 with serogroup E₂ strains that exhibit O antigen 15 (Table 1) and inhibition of the reaction by O antigen 15 antiserum (Table 2). The reduced reactivity of MO15 against the ɛ34-converted strains may have arisen from structural modification due to glycosylation of the Gal residues on their LPS backbone PS chains. Nevertheless, this glycosylation did not seem to completely destroy the epitope for MO15, indicating that the glycosylation is not completely stoichiometric (12), with enough MO15 epitopes being left intact for reaction with the MAb. The β linkage between Gal and Man in the LPS of ɛ15- and ɛ34-lysogenized serogroup E₁ salmonellae is probably essential for reaction with MO15, since neither the nonlysogenic E₁ nor the E₄ strains, which have the same backbone except for the α linkage between the Gal and Man residues, reacted with the MAb.

There was one nonlysogenic E₁ strain that reacted with MO15 in the slide agglutination assay but not in the capture ELISA. This particular strain may contain cells with different levels of expression of the O-acetyl group on the Gal residues in the LPS backbone, and it is possible that some of their Man residues are in the β anomeric configuration. The O-acetyl residues in the nonlysogenic serogroup E₁ strains' LPS are believed to be added postpolymerization in a nonstoichiometric manner (9, 12). The coexistence of both α- and β-mannosyl residues on the LPS of an individual Salmonella strain had also been reported (14, 15). In either case, the result would be a nonlysogenic serogroup E₁ strain containing a subpopulation of cells with LPS structures more closely resembling that of the ɛ15-lysogenized variants. This probably explains why this strain was positive in the slide agglutination assay but not in the capture ELISA, since detection by the former assay would be favored because it uses excess antigen to detect this microheterogeneity.

MO10 was produced as described by Tsang et al. (17). Its epitope structure was demonstrated to be independent of the conventional O antigen 10, i.e., O-acetylated Gal (17). The lack of reaction of MO10 with ɛ15- or ɛ34-lysogenized variants indicates that the linkage between Gal and Man residues in the O-PS backbone is also an essential component in the MO10 epitope. The weaker reactivity of MO10 with group E₄Salmonellastrains is probably due to the fact that part of the Gal residues in the E₄Salmonella LPS were replaced by glucose to form O antigen 1, thereby altering the normal epitope for MO10. Reactions of MO10 with serogroup E₄ and of MO15 with serogroup E₃Salmonella strains were both favored by the slide agglutination test, which uses an excess of antigen. This indicates that there exists heterogeneity in the population of cells being tested. Like the O-acetyl group in serogroup E₁ LPS, nonstoichiometric addition of glucose to the Gal in the serogroup E₄Salmonella LPS would allow some MO10 epitopes to be present, thus permitting their detection by an assay which uses an excess of antigen, such as the slide agglutination test.

It is slightly misleading to use the name MO10, since its antigenic structure does not involve the conventionally defined O-10, in which the O-acetyl-substituted Gal plays a dominant role (Fig. 1). However, since there is no single exclusive common antigen for the nonlysogenic E₁ and E₄ strains that could be used to differentiate them from the phage-lysogenized E₁ strains as described in the Kauffmann-White scheme (7), and the MAb is primarily reactive against the nonlysogenic E₁ strains, it was decided that MO10 would be the most appropriate name until a new antigenic factor is defined. This undefined new antigenic factor for MO10 reinforces the existence of epitopes besides those described in the Kauffmann-White scheme (2), and it provides evidence that conventional antisera, being polyclonal in nature, may not be the best tool for defining antigenic determinants.

If each epitope on an antigen could instigate an individual immune reaction, and hence a clone of antibodies, then each of these epitopes would be an antigen factor under the Kauffmann-White scheme, provided they are of significance in identification of the different serotypes (2). In the not-so-distant past, technology did not allow separate labeling of these epitopes, but as MAbs have become commonly available, we are now ready to reevaluate the Kauffmann-White scheme based on MAb-defined epitopes. We will probably be confronted with too many MAb-defined antigenic factors for rational use initially, but as these are evaluated for their significance in serotype identification, we should be able to eliminate nonspecific antigenic factors and assemble a collection of those that will together facilitate efficient serogroup and subgroup differentiation. Furthermore, due to the exquisite specificities of MAbs, it will also be possible to pinpoint and characterize the detailed molecular structure of each MAb-defined antigenic factor. This new collection of MAb-defined antigenic factors will undoubtedly contribute to the improvement of the current serological classification scheme of Kauffmann and White.

In the case of serogroup E salmonellae, in slide agglutination assays, MO10 will identify the nonlysogenic E₁ and E₄strains, possibly via a new MAb-based antigenic factor involving the α-1,6-linked Gal-Man backbone as the dominant epitope structure. Similarly, MO15 will identify the phage ɛ15- and ɛ34-lysogenic strains via an antigenic factor that is similar to that of O-15, with the β-1,6-linked Gal-Man backbone acting as the dominant epitope structure. To allow for further differentiation between the nonlysogenic E₁ and E₄ strains and between ɛ15- and ɛ34-lysogenic strains, additional MAbs are needed to recognize the dominant epitopes of O antigens 19 and 34, respectively. Together, these MAbs would facilitate complete differentiation of serogroup E Salmonella strains. OtherSalmonella serogroups might be similarly differentiated by using MAbs.

• Copyright © 2000 American Society for Microbiology