Membrane Insertion for the Detection of Lipopolysaccharides: Exploring the Dynamics of Amphiphile-in-Lipid Assays

Abstract

Shiga toxin-producing Escherichia coli is an important cause of foodborne illness, with cases attributable to beef, fresh produce and other sources. Many serotypes of the pathogen cause disease, and differentiating one serotype from another requires specific identification of the O antigen located on the lipopolysaccharide (LPS) molecule. The amphiphilic structure of LPS poses a challenge when using classical detection methods, which do not take into account its lipoglycan biochemistry. Typically, detection of LPS requires heat or chemical treatment of samples and relies on bioactivity assays for the conserved lipid A portion of the molecule. Our goal was to develop assays to facilitate the direct and discriminative detection of the entire LPS molecule and its O antigen in complex matrices using minimal sample processing. To perform serogroup identification of LPS, we used a method called membrane insertion on a waveguide biosensor, and tested three serogroups of LPS. The membrane insertion technique allows for the hydrophobic association of LPS with a lipid bilayer, where the exposed O antigen can be targeted for specific detection. Samples of beef lysate were spiked with LPS to perform O antigen specific detection of LPS from E. coli O157. To validate assay performance, we evaluated the biophysical interactions of LPS with lipid bilayers both in- and outside of a flow cell using fluorescence microscopy and fluorescently doped lipids. Our results indicate that membrane insertion allows for the qualitative and reliable identification of amphiphilic LPS in complex samples like beef homogenates. We also demonstrated that LPS-induced hole formation does not occur under the conditions of the membrane insertion assays. Together, these findings describe for the first time the serogroup-specific detection of amphiphilic LPS in complex samples using a membrane insertion assay, and highlight the importance of LPS molecular conformations in detection architectures.

Fig 1. Representative structure of the molecular components of LPS.

The conserved, hydrophobic lipid A group, core polysaccharide, and hypervariable O polysaccharide antigen. The lipid A group of most E. coli strains has 6 fatty acid tails which anchors LPS into the bacterial cell membrane, and is recognized by host receptor proteins.

Fig 2. Membrane insertion for detection of LPS O157.

(A) Spectral curves demonstrating detection of various concentrations of LPS O157. (B) Integrated values of spectral curves plotted as single points with standard error of the mean. Closed diamonds indicate averaged integrated signal intensity, and open diamonds are integrated NSB.

Fig 3. Concentration dependent detection of LPS O157 in 1 mg/mL beef lysates.

Detection of LPS in beef lysates shows an increase in signal to noise ratios as compared to those seen in the benchmark assay.

Fig 4. O-ag targeted detection of LPS.

Using membrane insertion, two types of LPS were detected using their complement antibodies, polyclonal anti-E. coli O104 (pAb O104) and monoclonal anti-E.coli O111 (mAb O111) labeled with af647.

Fig 5. Assay performance inside flow cells.

Schematics of (A) imaging flow cell and (B) waveguide flow cell. Major differences between these two flow cells include replacement of the waveguide glass substrate with a thin glass coverslip for imaging, and the addition of PDMS and silicone pedestals to create an airtight environment to preserve lipid integrity throughout an assay. (C) 100 μg/mL and (D) 50 μg/mL LPS O111:B4 incubated with BODIPY labeled DOPC lipids. Composite images of (E) 50 μg/mL and (F) 25 μg/mL LPS O157 bound by fluorescently labeled pAb O157-af647.

Fig 6. Imaging LPS O157 with lipid bilayers.

(A) Bilayer prior to incubating with LPS O157. 50 μg/mL LPS (B) O157, (C) O104, and (D) O111:H11.