(715c) Multiplex Analysis of Raman Nanoparticle Signatures for Protein Imaging In Tissues

Lutz, B. R., University of Washington
Dentinger, C., Intel Corporation, Digital Health Group
Nguyen, L., Intel Corporation, Digital Health Group
Sun, L., Intel Corporation, Digital Health Group
Zhang, J., Intel Corporation, Digital Health Group
Chan, S., Intel Corporation, Digital Health Group
Allen, A., Fred Hutchinson Cancer Research Center
Knudsen, B., Fred Hutchinson Cancer Research Center

The ability to simultaneously detect multiple biomolecule targets in a single sample reduces demands on precious samples and allows new insight into expression colocalization in applications including cell and animal imaging, tissue histology, and flow cytometry. Nanoparticle probes based on surface-enhanced Raman scattering (SERS) are emerging as a new class of optical labels with unique spectral fingerprints that are well-suited for multiplex detection. Since the discovery of nanoparticle SERS in 1997, a variety of fabrication approaches and biomolecule conjugation methods have been developed to enable early applications in immunoassays, nucleic acid hybridization, cell and tissue staining, and in vivo imaging. Despite the significant potential of Raman probes for improved multiplex quantification, they have primarily been used for qualitative detection of single targets, due in part to the lack of analysis methods that take full advantage of the complex spectral fingerprint.

We describe application of antibody-conjugated Composite Organic-Inorganic Nanoparticles (COINs) for multiplex detection in fixed human tissue samples, and we present a simple but robust spectral fitting method that quantitatively separates overlapping probe signatures and background autofluorescence. Simplified solution and protein plate binding assays show that multiplex spectral analysis allows sensitive detection of small signals (<2% of total signal) and provides quantitative multiplex detection of at least four overlapping probe signatures. In a triplex assay on human prostate tissue, two antibody-conjugated COINs and a fluorophore are used to image expression of prostate-specific antigen, cytokeratin-18, and DNA. Spectral analysis allows point-by-point removal of unknown background autofluorescence and direct measurement of quantification error at each point in the sample (~2% of total signal). The multiplex assays are carried out by single-laser excitation using a small fraction of the visible window (~10%). This suggests that multiple-laser excitation of Raman probes and quantification by spectral analysis will allow high-order multiplexing beyond that currently possible with fluorophores.