PROJECT SUMMARY The tumor microenvironment (TME) in essentially all epithelial cancers is associated with significant biochemical and structural changes in the extracellular matrix (ECM). Many tumors including those of the breast, pancreas and ovary are characterized by profound changes in the collagen architecture. ECM changes (~micron scale) are below the resolution of conventional imaging modalities but analysis of this structure is critical for understanding carcinogenesis and metastasis. We have used the collagen-specific modality of Second Harmonic Generation (SHG) optical microscopy to discriminate cancer specimens from normal tissues based on changes in supramolecular structure, fibril structure, and fiber morphology, where we have focused on high grade serous ovarian cancer (HGSOC). However, SHG cannot identify the specific molecular alterations, which could provide critical information on disease etiology, prognosis, and response to therapy. Now we will develop a novel method that combines spatially registered SHG and surface enhanced mid-infrared spectral imaging (SE-MIRSI) correlating morphometric and chemometric information to elucidate tumor-promoting ECM alterations. The latter spatially probes specific molecular signatures from vibrational spectroscopy and provides increased sensitivity using nanophotonic substrates, allowing rapid and large-area chemical imaging of whole tissue sections. Specifically, SE-MIRSI can quantitatively identify specific changes in isoform distribution, posttranslational modifications and altered crosslinking of the collagen fibers. Spatial registration of SHG and SE-MIRSI then will provide a comprehensive, ultrasensitive, label free, non-destructive, high-resolution structural and biochemical imaging platform to investigate the role of ECM alterations in promoting tumor carcinogenesis and metastasis. Here, we will develop a multivariate data processing workflow that identifies the specific signatures of collagen and other ECM components from the two modalities establishing the basis of an accurate classifier. We will validate the multimodal characterizations on HSGOC tissue samples. At the end of this project, we will have developed a multimodal imaging platform that will uniquely identify collagen and other ECM biochemical alterations in the TME. We will establish performance measures based on imaging speed and throughput, sensitivity and classification accuracy. These structural and biochemical analyses will provide new insight into carcinogenesis and disease progression in several carcinomas. We propose these Aims: Aim 1. Identify specific structural and biochemical signatures of in vitro ECM models through the combined use of SHG and SE-MIRSI. Aim 2. Validate spatially registered SHG/SE-MIRSI method on high grade serous ovarian cancer and identify specific associated structural morphology and biochemical signatures.