(T1030-03-21) Interactome between Exopolysaccharide, Poly-(1- > 6)-β-N-acetylglucosamine and Proteins in Staphylococcus Epidermidis Biofilms by Live Cell Proximity Labeling

Purpose: Bacterial cells form multicellular communities embedded with self-produced extracellular polymeric substance (EPS) which is preferred as biofilms protecting cells from harsh conditions. However, this poses as a challenge for the medical field due to the increase antibiotic resistance. Understanding how biofilm formation is driven will shed light for treatment development in the future. One of the approaches is understanding the interactions between the macromolecules found in the EPS. This project focuses on studying the interactome between proteins and the exopolysaccharide known as poly-(1- > 6)-β-N-acetylglucosamine (PNAG). Though PNAG is commonly found in bacterial biofilm, there is still a lack of studying about this exopolysaccharide and its role in biofilm formation. S. epidermidis was chosen as the studying species due to its common cause for nosocomial infections. In this experiment, live cell proximity labeling was chosen to study interaction of exopolysaccharide – protein. Utilizing the engineered ascorbate peroxidase protein (APEX2) known for biotinylated any protein in the proximity of 20nm radius through the generation of reactive agents in the presence of peroxide. APEX2 is fused with a catalytically inactive mutant of Dispersin B protein (DspB), a hydrolase which is known to bind to PNAG¹. This allows us to label and identify any protein close to PNAG through combination of proximity labeling and mass spectrometry.

Methods: S. epidermidis RP62A was grown in tryptic soy broth (TSB) at 37 ^oC for 48 hrs for biofilm being fully developed. Biofilm was aliquoted to negative control sample (--), another was used as a positive control sample (++), and the one was(-+). (--) sample throughout the proximity labeling process was incubated with only buffer (100 mM NaCl and 25 mM Na₂HPO₄, pH 6.0). (++) sample was first incubated with the active DspB_wt to degrade all PNAG structure in the biofilm. Then the sample was labeled with our engineered DspB-APEX2 probe. (-+) sample was incubated with only probe. Three samples went through the labeling process beginning with the introduction of biotin phenol, followed up with peroxide, and finally reactions were quenched. All samples were subjected to microscopic imaging using streptavidin-TexasRed conjugate and imaged under microscope. Samples went through cell lysis to collect the lysates. All lysates were run on electrophoresis gel and Western. All samples were subjected to proteomic studying through tandem mass tag (TMT) mass spectrometry.

Results: SDS gel confirm all three samples were loaded with relatively equal amount of lysate (Figure 1). The Western blot images show only proteins that were biotinylated by our proximity labeling process and natively biotinylated proteins in the samples through streptavidin-HRS conjugate.  As showed on Figure 1, the transparent images of (--) and (-+) show larger clustered of cells with multicellular layer, while the (++) sample shows scattered of individual of cells or much smaller cluster of cells, this is the result of introducing of DspB_wt. As for the red channel, we observed a significant labeling only on (-+) sample indicating the labeling process was only done in the presence of PNAG. Three biological replicates of S. epidermidis were subjected to the procedure mentioned above. Each replicate was tagged with a tandem mass tag before combining all together (9 samples in total) to go through LC-MS. A total of 176 S. epidermidis unique proteins were identified across the biological replicates. Among these, 93 proteins were found to be significantly enriched (p < 0.05) by >5-fold in (– +) treated S. epidermidis biofilms compared to control (– –) biofilm. To identify PNAG-dependent biotinylated proteins, we compared (-+) sample with (++) sample in which 47 proteins were significantly enriched by >5-fold in the relative (-+) relative to (++) samples.

Conclusion: Here, we propose proximity labeling approach which was first developed for protein – protein interactome. Utilizing APEX2 probe fused with catalytically inactive DspB, we were able to confirm several unknown protein bands only showed up on our sample lysate and not the two controls (Figure 1). In addition, we also showed the PNAG-dependent labeling capacity and distribution of our probe through microscopic imaging with TxRed (Figure 2). We moved forward with the TMT Mass Spectrometry to identify these PNAG-dependent proteins. Several hits were identified based on Mass Spectrometry data, and we were able to single out the most abundant protein that was found on the sample which was the protein called Elastin Binding Protein (EpbS). As the name suggests, this protein has been known to bind to elastin-abundant tissue such as human skins. Interestingly, this protein on S. epidermidis contains a lysin motif (LysM) domain on the C-terminus of the protein which recognizes and binds to carbohydrates, and the fact this protein was found to be PNAG-dependent based on our proximity labeling result, it requires further study for this protein EbpS and its potential role in biofilm formation through interaction with PNAG.

References: (1). Manuel, S. G. A.; Ragunath, C.; Sait, H. B. R.; Izano, E. A.; Kaplan, J. B.; Ramasubbu, N. Role of Active-Site Residues of Dispersin B, a Biofilm-Releasing β-Hexosaminidase from a Periodontal Pathogen, in Substrate Hydrolysis. FEBS J. 2007, 274, 5987–5999. https://doi.org/10.1111/j.1742-4658.2007.06121.x.
(2) McKenney, D.; Hübner, J.; Muller, E.; Wang, Y.; Goldmann, D. A.; Pier, G. B. The Ica Locus of Staphylococcus epidermidis Encodes Production of the Capsular Polysaccharide/Adhesin. Infect. Immun. 1998, 66, 4711–4720. https://doi.org/10.1128/iai.66.10.4711-4720.1998.
(3) Luan H. Vo, Steven Hong, Kaitlyn E. Stepler, Sureshee M. Liyanaarachchi, Jack Yang, Peter Nemes and Myles B. Poulin*. Mapping protein–exopolysaccharide binding interaction in Staphylococcus epidermidis biofilms by live cell proximity labeling. https://doi.org/10.1101/2023.08.29.555326

Acknowledgements: The project was supported by the National Institute of General Medical Science of the National Institute of Health under award number of R35GM147110. This content is the responsibility of the authors and does not represent the point of view of the National Institutes of Health.


Figure 1. Fluorescence microscopy images of biotin-tagged (red) surface proteins showing APEX2-DspB generate significant labeling over the negative controls where the incubation of DspB-APEX2 is omitted. Incubation of DspB wt causes the loss in labeling signal suggesting the labeling process is PNAG-dependent. Live cell proximity labeling of S. epidermidis biofilms pre-treated with or without DspBwt for 1 h prior to incubation with APEX2-DspB. Total cell lysates (left) were imaged for total protein (Stain-Free image). Total cell lysates (middle) and elution fraction from pulldown assay (right) were blotted for biotin labeling using a streptavidin-HRP conjugate. The results show the presence of proteins that were selectively labeled in presence of PNAG


Figure 2. A total of 176 proteins were enriched by streptavidin pull down from labeled (- +) vs. unlabeled (- -) samples (B) or from labeled vs. DspBwt treated (+ +) samples. A total of 4 proteins were significantly (P>0.01) enriched (> 10-fold) in both sets of experimental conditions.