Characterization of Ethanol Extracted Cell Wall Components of Mycobacterium avium Subsp. paratuberculosis

Abstract

Antigens extracted using ethanol (EtOH) and incorporated in the EtOH vortex ELISA (EVELISA) test have previously shown high specificity and sensitivity for detecting Mycobacterium avium subspecies paratuberculosis (Map) and M. bovis infections in cattle. The objective of this study is to define the components present in the EtOH extract. We show that this extract is composed of lipid, carbohydrate, and proteins on the surface of the bacilli, and that EtOH removes the outer layer structure of Map which comprise these elements. To identify proteins, polyclonal antibodies to the EtOH prep were produced and used to screen a Map genomic expression library. Seven overlapping clones were identified with a single open reading frame, MAP_0585, common to all. MAP_0585, which encodes a hypothetical protein, was recombinantly produced and used to demonstrate strong reactivity in sera from hyperimmunized rabbits, but this protein is not strongly immunogenic in cattle with Johne's disease. A panel of monoclonal antibodies was used to determine the presence of additional proteins in the EtOH extract. These antibodies demonstrated that a well-known antigen, termed MPB83, is present in M. bovis EtOH extracts and a fatty acid desaturase (MAP_2698c) is present in Map EtOH extracts, while lipoarabinomannan was common to both. The lipid and carbohydrate components of the extract were analyzed using thin layer chromatography and lectin binding, respectively. Lectin biding and protease treatment of the EtOH extract suggest the antigenic component is carbohydrate and not protein. These results give further insight into this important antigen prep for detecting mycobacterial diseases of cattle.

Keywords: ELISA; Johne’s disease; antigens; lipid; paratuberculosis.

Figure 1

Antibody binding to intact mycobacteria with bovine and rabbit antisera. (a) Shown are the results from 5 Johne’s disease (JD)-positive (gray bars) and 5 JD-negative cows (black bars) against different species and subspecies of mycobacteria. (b) Sera from 2 rabbits immunized with the Map K-10 EtOH extract shows a similar pattern to what was observed with sera from cattle with JD. The pre-immune serum is represented by black histogram bars and post-immune serum is the gray bars. Error bars represent standard deviation of the means. M. avium (706) is TMC706 and M. avium (721) is TMC721, a strain isolated from a child’s lymph node and is serotype 3. Significance is denoted by asterisks, * = p < 0.01; * * = p < 0.001.

Figure 2

Transmission electron microscopy of Map K-10 exposed to rabbit pre-immune and post-immune serum. Detection of antibody binding was with Colloidal gold 10 nm particles. The images are Map strain K-10 labeled with antibody against the EtOH prep from rabbit 3993 (a–c) and 3995 (d,e). Pre-immune serum from 3993 does not react to Map (f). Images are all magnified 49,000×.

Figure 3

Scanning electron microscopy of Map pre and post EtOH extraction. Map were exposed to bovine serum from a JD-positive (a,c,d) and a JD-negative cow (b). The images in (a–c) were taken prior to EtOH extraction while (d) was taken after EtOH extraction. Note the clear zone at the surface indicating the stripped nature of the cell wall in (d). Magnification is indicated by size bars in the lower left corner of each image.

Figure 4

Thin layer chromatography of mycobacterial lipids. Shown are the lipid migration patterns of M. bovis (Mb), Map K-10 (Map), M. kansasii (Mk) and M. avium subspecies avium (Ma). Plates were developed in chloroform-methanol-water (90:10:1) until this solvent phase migrated 14 cm. The R_f = 0.5 indicates the solvent migration halfway point. R_f symbolizes retardation factor.

Figure 5

Lectin binding to the Map EtOH extract. Lectin binding was detected using peroxidase-conjugated secondary antibody and peroxidase substrate (ABTS) followed by measuring absorbance at 415 nm. Histogram bars represent mean lectin binding ± standard deviation of quadruplicate determinations. Binding to both concanavalin A (ConA) (top left) and wheat germ agglutinin (WGA) (top right) was observed predominantly in the chloroform phase with little to no binding in the aqueous phase. Conversely, no Solanum tuberosum lectin (STL) (bottom left) or Limulus Polyphemus lectin (LPL) (bottom right) binding was observed in any fraction of the extract, while a low but detectable binding level was observed in the complete extract with STL. An asterisk denotes p values less than 0.01.

Figure 6

Preparation of EtOH extract for mass spectrometry analysis. (a) Map bacilli were washed 6 times prior to EtOH extraction (labeled × 6) to avoid bovine serum albumin (BSA) contamination. A second extract was prepared after washing Map once (label × 1). Lipids and carbohydrate were removed from the EtOH extract through a second chloroform:methanol:water extraction. This second extraction was run on 12% SDS-PAGE denaturing gels and exposed to Coomassie stain (CBB) and silver stain. The location of BSA migration is identified for both gel images. (b) Proteinase K treatment of Map EtOH extract. Inclusion of proteinase K or the EtOH extract along with staining method is indicated beneath the gel. Kilodalton size markers are shown in the left margins of all gels.

Figure 7

MPB83 is not the most abundant protein in M. bovis EtOH extracts. (a) SDS-PAGE analysis of the M. bovis EtOH extract reveals a high abundant protein migrating between the 50 and 75 kDa size markers (lane 2). (b) Immunoblot analysis with a monoclonal antibody to MPB83 (1F11) identifies the 22-kDa MPB83 protein in the EtOH extract (lane 2). Protein size markers are loaded in lane 1 with sizes identified in kilodaltons (kDa) in the left margins. The unknown and MPB83 proteins are indicated in the right margins.

Figure 8

Sequence alignment of positive clone inserts from a Map phage expression library. Shown are the DNA inserts subcloned from positive plaques obtained from the library screen. Subcloned inserts are aligned and drawn relative to a base pair scale that spans 6.6 kb. There is a 500 and 600 bp break in the scale to amplify the region of interest. The top row shows the coding sequences from the Map genome and their relative positions among the library clones. Highlighted in red is the region of the respective clones that overlap with the coding sequence MAP_0585. Only this coding sequence was universally present in all the clones.

Figure 9

Antigenicity of MAP_0585. Immunoblot analysis shows MAP_0585 reacts strongly to the rabbit antisera developed to the EtOH extract (3993 in lane 2 and 3995 in lane 3) and the protein is also detected by a mAb to the maltose bind protein affinity tag (lane 16). However, the protein is not antigenic in cattle (lanes 9 and 12 are healthy cattle and all other lanes are JD cattle). The position of the MAP_0585 is indicated in the left margin and the host generating antibody is indicated beneath the blot.

Figure 10

Bovine antibody binds to a carbohydrate component of the Map EtOH extract, not to protein. (a) Surface antigens of Map K-10 were extracted with 80% ethanol (EtOH Whole) and fractionated by Folch extraction method into organic (chloroform), interface and aqueous fractions. After evaporating methanol for immobilization of the lipids (and other molecules) onto the wells of a microtitre plate, they were incubated with serum samples (1:100 dilution) collected from JD-positive (solid bar) and negative (open bar) cattle. Histogram bars represent mean antibody binding ± standard deviation of quadruplicate determinations. Most of the antigenicity is in the organic fraction. (b) EtOH extract treated with proteases shows no negative effects on bovine antibody binding. The extract was treated with either 10 µg/mL trypsin or 10 µg/mL proteinase K with each treatment actually enhancing antibody binding. This enhancement was not statistically significant. However, EtOH extract antigens are efficiently removed by ConA-agarose as shown by lack of antibody binding (c). Absorption with agarose only does not affect antibody binding to the EtOH extract. This experiment was repeated in triplicate with qraduplicate measurements for each. p values less than 0.01 are denoted by an asterisk.