Label-free Optical Imaging and the Tumor Microenvironment

Cancer is a devastating disease with approximately 19.3 million new cancer diagnoses and 10 million deaths worldwide in 2020.¹ Screening and early detection improve the 5-year survival rates; however, 70-90% of all cancer deaths result from secondary metastasis. The tumor microenvironment drives the growth and metastatic spread of tumor cells through a series of interactions between stromal and cancer cells. Collagen, the most abundant protein in the tumor microenvironment, provides cues for several of these processes. Second Harmonic Generation (SHG) microscopy imaging, a label-free, collagen specific imaging modality is a powerful tool in probing collagen organization and has been used to describe collagen organization in several cancers including, breast, ovarian, pancreatic, and melanoma. The Tilbury Lab at the University of Maine seeks to quantitatively explore the dynamics of collagen remodeling in both 3D in-vitro models and patient biopsy tissues using label-free SHG imaging and advanced image analysis algorithms built in collaboration with the CompuMAINE lab at the University of Maine. Recently, we have adapted the 2D Wavelet Transform Maximum Modulus (2D WTMM) to quantify collagen fiber morphologies across multiple size-scales demonstrating increased sensitivity over commonly used 2D FFT image analysis approaches. In addition to the advanced image analysis algorithm development; we are focused on leveraging the physical underpinnings of SHG generation to interrogate the microstructural features of collagen via the SHG directionality and polarization responses. Of particular interest is the relationship of collagen microstructural features and a class of cryptic integrin binding sites which provide additional acellular signaling cues to promote tumorigenesis. In conjunction with collagen remodeling, we also aim to incorporate optical metabolic imaging with the hierarchical collagen remodeling with cellular metabolism in 3D cancer models to increase knowledge of the bi-directional integrin-mediated cancer dynamics.  Overall, our goal is to further develop label-free optical imaging and quantitative image analyses for use in cancer biology.