Identification of novel adhesins of M. tuberculosis H37Rv using integrated approach of multiple computational algorithms and experimental analysis

Abstract

Pathogenic bacteria interacting with eukaryotic host express adhesins on their surface. These adhesins aid in bacterial attachment to the host cell receptors during colonization. A few adhesins such as Heparin binding hemagglutinin adhesin (HBHA), Apa, Malate Synthase of M. tuberculosis have been identified using specific experimental interaction models based on the biological knowledge of the pathogen. In the present work, we carried out computational screening for adhesins of M. tuberculosis. We used an integrated computational approach using SPAAN for predicting adhesins, PSORTb, SubLoc and LocTree for extracellular localization, and BLAST for verifying non-similarity to human proteins. These steps are among the first of reverse vaccinology. Multiple claims and attacks from different algorithms were processed through argumentative approach. Additional filtration criteria included selection for proteins with low molecular weights and absence of literature reports. We examined binding potential of the selected proteins using an image based ELISA. The protein Rv2599 (membrane protein) binds to human fibronectin, laminin and collagen. Rv3717 (N-acetylmuramoyl-L-alanine amidase) and Rv0309 (L,D-transpeptidase) bind to fibronectin and laminin. We report Rv2599 (membrane protein), Rv0309 and Rv3717 as novel adhesins of M. tuberculosis H37Rv. Our results expand the number of known adhesins of M. tuberculosis and suggest their regulated expression in different stages.

Figure 1. Selection of target proteins.

A. The selection pipeline used for investigating computationally predicted adhesins. B. Logic table for combining the results from predictions using multiple algorithms. Predictions from various algorithms were considered for the hypothesis of “adhesin and surface localized” in the forms of Claims (supporting) and Attacks (not-supporting) associated with their confidence of predictions in three categories, High, Medium and Low. Several cases could not be resolved and they remained “undecided” in our protocol.

Figure 2. Purification and CD spectral analysis of rRv2599.

A. Protein topology showing the presence of signal peptide and/or transmembrane helices in the protein signal peptide spans from 1–16 amino acids, whereas transmembrane domain spans from 7–29 amino acids. B. Purified protein run on 12% SDS-PAGE stained with coomassie. Lane M is molecular weight marker and Lane 1 is the purified recombinant rRv2599. C. Western blotting with Anti-His antibodies. Lane M is molecular weight marker and Lane 1 is the purified recombinant rRv2599. D. CD spectra of purified rRv2599 at two different temperatures 25°C and 37°C.

Figure 3. Purification and CD spectral analysis of rRv0309.

A. Protein topology showing the presence of signal peptide and/or transmembrane helices in the protein signal peptide spans from 1–34 amino acids, whereas transmembrane domain spans from 7–29 amino acids. B. Purified protein run on 12% SDS-PAGE stained with coomassie. Lane M is molecular weight marker and Lane 1 is the purified recombinant rRv0309. C. Western blotting with Anti-His antibodies. Lane M is molecular weight marker and Lane 1 is the purified recombinant rRv0309. D. CD spectra of purified rRv0309 at two different temperatures 25°C and 37°C.

Figure 4. Purification and CD spectral analysis of rRv3717.

A. Protein topology showing the presence of signal peptide and/or transmembrane helices in the protein. The signal peptide spans from 1–24 amino acids with cleavage site at 24–25 amino acids. B. Purified protein run on 12% SDS-PAGE stained with coomassie. Lane M is molecular weight marker and Lane 1 is the purified recombinant rRv3717. C. Western blotting with Anti-His antibodies. Lane M is molecular weight marker and Lane 1 is the purified recombinant rRv3717. D. CD spectra of purified rRv3717 at two different temperatures 25°C and 37°C.

Figure 5. Phylogenetic analysis of Rv3717 and N-acetylmuramoyl-L-alanine amidase from various bacterial species.

Proteins highlighted have N-acetylmuramoyl-L-alanine amidase activity and are adhesins. Numbers at fork are from bootstrap analysis.

Figure 6. Binding activity of recombinant proteins with extracellular matrix proteins.

Histograms of % color saturation with error bars showing 1 standard deviation (1 SD) of values obtained in triplicate experiments. A. rRv0309, B. rRv2599, C. rRv3717 and D. P-II nitrogen status signaling protein (rRv2919c).

Figure 7. Heatmap of selected adhesin coding genes expressed in various experimental conditions in M. tuberculosis.

Values are Z-scores of log (ratio) of averaged technical replicates of genes expressed in each condition. The color scale is shown against each heatmap.

Figure 8. System of various adhesive molecules having role in adherence of M. tuberculosis to the host.

The pathogen alters its surface adhesins through regulation of gene expression under different conditions of stress faced during its pathogenic life cycle.