Machine learning-based estimation of spatial gene expression pattern during ESC-derived retinal organoid development.

Organoids, which can reproduce the complex tissue structures found in embryos, are revolutionizing basic research and regenerative medicine. In order to use organoids for research and medicine, it is necessary to assess the composition and arrangement of cell types within the organoid, i.e., spatial gene expression. However, current methods are invasive and require gene editing and immunostaining. In this study, we developed a non-invasive estimation method of spatial gene expression patterns using machine learning. A deep learning model with an encoder-decoder architecture was trained on paired datasets of phase-contrast and fluorescence images, and was applied to a retinal organoid derived from mouse embryonic stem cells, focusing on the master gene Rax (also called Rx), crucial for eye field development. This method successfully estimated spatially plausible fluorescent patterns with appropriate intensities, enabling the non-invasive, quantitative estimation of spatial gene expression patterns within each tissue. Thus, this method could lead to new avenues for evaluating spatial gene expression patterns across a wide range of biology and medicine fields.

Designing receptor agonists with enhanced pharmacokinetics by grafting macrocyclic peptides into fragment crystallizable regions.

Short half-lives in circulation and poor transport across the blood-brain barrier limit the utility of cytokines and growth factors acting as receptor agonists. Here we show that surrogate receptor agonists with longer half-lives in circulation and enhanced transport rates across the blood-brain barrier can be generated by genetically inserting macrocyclic peptide pharmacophores into the structural loops of the fragment crystallizable (Fc) region of a human immunoglobulin. We used such 'lasso-grafting' approach, which preserves the expression levels of the Fc region and its affinity for the neonatal Fc receptor, to generate Fc-based protein scaffolds with macrocyclic peptides binding to the receptor tyrosine protein kinase Met. The Met agonists dimerized Met, inducing biological responses that were similar to those induced by its natural ligand. Moreover, lasso-grafting of the Fc region of the mouse anti-transferrin-receptor antibody with Met-binding macrocyclic peptides enhanced the accumulation of the resulting Met agonists in brain parenchyma in mice. Lasso-grafting may allow for designer protein therapeutics with enhanced stability and pharmacokinetics.