Targeting cancer-specific glycans by cyclic peptide lectinomimics.

The transformation from normal to malignant phenotype in human cancers is associated with aberrant cell-surface glycosylation. Thus, targeting glycosylation changes in cancer is likely to provide not only better insight into the roles of carbohydrates in biological systems, but also facilitate the development of new molecular probes for bioanalytical and biomedical applications. In the reported study, we have synthesized lectinomimics based on odorranalectin 1; the smallest lectin-like cyclic peptide isolated from the frog Odorrana grahami skin, and assessed the ability of these peptides to bind specific carbohydrates on molecular and cellular levels. In addition, we have shown that the disulfide bond found in 1 can be replaced with a lactam bridge. However, the orientation of the lactam bridge, peptides 2 and 3, influenced cyclic peptide's conformation and thus these peptides' ability to bind carbohydrates. Naturally occurring 1 and its analog 3 that adopt similar conformation in water bind preferentially L-fucose, and to a lesser degree D-galactose and N-acetyl-D-galactosamine, typically found within the mucin O-glycan core structures. In cell-based assays, peptides 1 and 3 showed a similar binding profile to Aleuria aurantia lectin and these two peptides inhibited the migration of metastatic breast cancer cell lines in a Transwell assay. Altogether, the reported data demonstrate the feasibility of designing lectinomimics based on cyclic peptides.

Unpaired mesh-to-image translation for 3D fluorescent microscopy images of neurons.

While Generative Adversarial Networks (GANs) can now reliably produce realistic images in a multitude of imaging domains, they are ill-equipped to model thin, stochastic textures present in many large 3D fluorescent microscopy (FM) images acquired in biological research. This is especially problematic in neuroscience where the lack of ground truth data impedes the development of automated image analysis algorithms for neurons and neural populations. We therefore propose an unpaired mesh-to-image translation methodology for generating volumetric FM images of neurons from paired ground truths. We start by learning unique FM styles efficiently through a Gramian-based discriminator. Then, we stylize 3D voxelized meshes of previously reconstructed neurons by successively generating slices. As a result, we effectively create a synthetic microscope and can acquire realistic FM images of neurons with control over the image content and imaging configurations. We demonstrate the feasibility of our architecture and its superior performance compared to state-of-the-art image translation architectures through a variety of texture-based metrics, unsupervised segmentation accuracy, and an expert opinion test. In this study, we use 2 synthetic FM datasets and 2 newly acquired FM datasets of retinal neurons.