A Surface-Enhanced Raman Spectroscopy-Based Biosensor for the Detection of Biological Macromolecules: The Case of the Lipopolysaccharide Endotoxin Molecules

Abstract

The development of sensitive methods for the detection of endotoxin molecules, such as lipopolysaccharides (LPS), is essential for food safety and health control. Conventional analytical methods used for LPS detection are based on the pyrogen test, plating and culture-based methods, and the limulus amoebocyte lysate method (LAL). Alternatively, the development of reliable biosensors for LPS detection would be highly desirable to solve some critical issues, such as high cost and a long turnaround time. In this work, we present a label-free Surface-Enhanced Raman Spectroscopy (SERS)-based method for LPS detection in its free form. The proposed method combines the benefits of plasmonic enhancement with the selectivity provided by a specific anti-lipid A antibody (Ab). A high-enhancing nanostructured silver substrate was coated with Ab. The presence of LPS was quantitatively monitored by analyzing the changes in the Ab spectra obtained in the absence and presence of LPS. A limit of detection (LOD) and quantification (LOQ) of 12 ng/mL and 41 ng/mL were estimated, respectively. Importantly, the proposed technology could be easily expanded for the determination of other biological macromolecules.

Keywords: LPS; Nano immunoassay; Raman spectroscopy; SERS; antibody; biosensor.

Figure 1

(a) Comparison of average SERS spectra acquired at different concentrations of LPS. Spectra were background corrected by a second-order polynomial curve. (b) Bar graph showing the average intensities and standard deviation of the peak at 2900 cm⁻¹ obtained in 5 SERS maps of the well corresponding to [LPS] = 1.25 µg/mL. (c) Plot of SNR of the peak centered at 2900 cm⁻¹ versus LPS concentration. Standard errors are included in the size of the experimental points. (d) Linear fit performed in the low concentration range used to estimate the limit of detection in correspondence with a SNR = 1.

Figure 2

Average SERS spectrum of Ab (10 µg/mL) + BS deposited on the silver substrate, according to the procedure described in the text. Spectrum was obtained using laser power impinging on the substrate of 100 µW and an integration time of 1 s.

Figure 3

Principal component analysis of Ab@c-SERS substrate. Score plots obtained for the samples (a) “Ab + BS”, (b) “Ab”, and (c) “BS” as reported in the text. For “Ab+BS” sample, the Ab concentrations were 0.1, 1, and 10 µg/mL. In the lower panel, we report three cartoons, which illustrate the coating degree of Ab (BS molecules are represented by red circles).

Figure 4

(a) Averaged SERS spectra of Ab@c-SERS upon exposure to LPS at different concentration values (for the sake of simplification, we have selected only some of the concentration values analyzed). Each spectrum, normalized to the prominent peak at ~2900 cm⁻¹, corresponds to the average of 400 spectra acquired by raster scanning. The blue area highlights the most relevant spectral changes induced upon LPS binding to Ab@c-SERS that were observed. (b) Zoomed image of the dashed area reported in part (a) that highlights the spectral shifts of the band envelope maxima in the high frequency region.

Figure 5

Principal component analysis of spectra at different LPS in the 1100–1800 cm⁻¹ spectral region. (a) Projection of the score plot in the PC1-PC3 plane. Spectra due to samples at different LPS concentration values are represented by dots of different colors, as specified in the legend. The yellow arrow indicates the direction of the apparent shift of the score clouds observed as the LPS concentration values increase. (b) Trend of the projections along PC3 axes of the centroids of score cloud as a function of LPS concentration value. (c) PC3 loadings resulting from the PC analysis previously described. (d) Loading plot in the PC1-PC3 plane, highlighting the direction of variation of significant features in the PC3 vectors.

Figure 6

(a) Outcome of the fitting procedure of the spectral envelope in the region around 3000 cm⁻¹ with four Gaussian peaks (blue dashed lines). The fitted envelope is reported as a continuous red line. (b) Trends of Δ₂₉₃₃ as a function of the LPS concentration value. The inset also reports the line resulting from a linear regression of data in the low concentration region in linear scale.

Figure 7

Schematic illustration of the procedure followed for the SERS detection of LPS at different concentrations. Step 1: Preparation of c-SERS substrate and wells setting up, using a parafilm mask. The inset shows a SEM image of coral-like SERS substrate (scale bar: 500 nm); Step 2: Immobilization of Ab by physical adsorption on each surface well; Step 3: Adding of BS to each surface well; Step 4: Adding of LPS to the functionalized SERS surface wells.