(76e) Quasi-3D Plasmonic Nanostructures for Strain Specific Identification of Marine Bacteria Vibrio Parahaemolyticus Using SERS

Surface-enhanced Raman spectroscopy (SERS) enabled by the advance of nanotechnology has immerged as a powerful analytical and sensing tool for detecting a verity of analytes from toxins to entire microorganisms. The extreme strong local electric fields, raising from the local surface plasmon resonance (LSPR), in the vicinity of noble metallic nanostructures can significantly increase the detection sensitivity by 10⁶ -10¹⁴-fold, which enables the detection of single molecules.^1,2 Since the strength of the electromagnetic field decays exponentially from the metal surface, SERS is limited to the near-surface of diffusely scattering objects. As a consequence, the SERS technique is well suited for probing cell membrane architectures and constituents that directly contact the surface of metal nanostructures. Several studies have demonstrated the capability of SERS for distinguishing various bacteria at the species-level by either mixing bacteria with  colloidal nanoparticle suspensions or  dropping bacteria on SERS-active substrates.^3,4 However, for many important applications such as food safety, environment monitoring, and homeland security, the strain-level identification is desired due to the large variation of pathogenicity among strains.

Here, we present the in situ identification of different strains of marine bacterium Vibrio Parahaemolyticus using SERS based on the unique quasi-3D plasmonic nanostructure arrays developed in our group.  The quasi-3D nanostructure arrays are composite of physically separated gold thin film with nanoholes on top and gold nanodisks at the bottom of the wells. Unlike previous SERS substrates of random gold nanoclusters, these SERS substrates are fabricated via electron beam lithography (EBL) which allows us to optimize nanostructures for SERS sensitivity of large organisms, and to precisely control the dimension that enables quantitative and reproducible analysis. We conducted finite-different time-domain (FDTD) electromagnetic calculations to determine the optimal nanostructures for detecting microorganisms. The optimal structure, which is 400 nm in diameter, 100 nm in spacing and 400 nm in depth, was fabricated via EBL followed by evaporation of 50 nm gold.  Seven Vibrio Parahaemolyticus strains obtained from environment including oyster, plankton and water as well as from clinical patients stool were selected to set up an in situ bacteria “fingerprint” database. 

Vibrio Paraheamolyticus anabiosis cultures were grown 12-16 hours in tryptic soy broth (TSB) at 37 ^oC with shaking to achieve exponential-phase growth. Bacteria solution was washed for 3 times with 1 g/L of 0.5% NaCl and then diluted to 10⁸ CFU/ml. The SERS-active substrate with four 50 μm x 50 μm quasi-3D nanostructure arrays was immersed in the bacteria solution in a custom-built Teflon cell covered with a piece of cover glass, allowing the collection of in situ SERS spectra of bacteria. Total 32 spectra were collected on each substrate and two substrates were used for each strain.  Results showed highly reproducibility due to the well controlled dimension of the nanostructure arrays.

All the raw spectra were baseline-corrected using cubic spline interpolation. The intensity and wavenumber of 25 main peaks of baseline-corrected spectra were used for principle component analysis (PCA) ^5,6   to generate the principle components (PCs) translation matrix. The first three PCs (PC1, PC2, and PC3), which account for greater than 90% of the variance in the analysis, were used to characterize these spectra and to differentiate the Vibrio Parahaemolyticus strains. All seven strains can be clearly differentiated in the 3D-PCA plot. Furthermore, we conducted the blind test by randomly selecting three strains from the seven tested strains. Following the same method, the PCA values of three blind samples overlap with the clusters of three known strains, and thereby, the strains of the blind samples were determined. The successful determination of the strains of blind samples indicate that quasi-3D plasmonic nanostructure arrays can be used as SERS-active substrates for strain specific identification of marine pathogens. 

       (1)  Kneipp, K.; Wang, Y.; Kneipp, H.; Perelman, L. T.; Itzkan, I.; Dasari, R. R.; Feld, M. S. Phys. Rev. Lett. 1997, 78, 1667.

       (2)  Nie, S.; Emory, S. R. Science 1997, 275, 1102.

       (3)  Patel, I.; Premasiri, W.; Moir, D.; Ziegler, L. J. Raman Spectrosc. 2008, 39, 1660.

       (4)  Jarvis, R.; Brooker, A.; Goodacre, R. Faraday Discuss. 2006, 132, 281.

       (5)  Xu, J. J.; Zhang, L.; Gong, H.; Homola, J.; Yu, Q. M. Small 2011, 7, 371.

       (6)  Jolliffe, I. Principal component analysis; Springer verlag, 2002.