Patino Alejandra V, Chapman Henry, Ginzel Joshua D, Sills Joelle E, Lyerly Herbert Kim, Rogers Bruce W, Drut Joaquin E, Snyder Joshua C
Department of Pharmacology and Cancer Biology, Durham, NC, USA.
Department of Surgery, Duke University, Durham, NC, USA.
bioRxiv. 2025 May 14:2025.05.09.652935. doi: 10.1101/2025.05.09.652935.
Fluorescent proteins (FPs) have revolutionized spatiotemporal observations in biology. Yet, the design of multiplexed assays remains constrained by limited spectral characterization and palette validation. Although over 1,000 FPs have been catalogued, systematic resources for characterizing their use in multiplexed approaches are lacking. Here we present a resource and methodology for selecting and decoding FPs in multiplexed imaging experiments. A library of forty-four FPs was built for rapid assembly into mammalian expression vectors and transposase-mediated integration. Hyperspectral imaging was performed for each FP and spectral space was characterized mathematically. To support experimental design and data interpretation, we developed the Cosine Similarity Decoder of XFP () toolbox, to predict spectrally resolvable FP palettes and decode and re-color hyperspectral images. Using this approach, we demonstrate live-cell imaging of 12 uniquely labeled clones. Our work offers a scalable platform for selecting optimal FP palettes for multiplex experiments, with broad utility across diverse biological systems and hyperspectral imaging techniques.
荧光蛋白（FPs）彻底改变了生物学中的时空观测。然而，多重分析的设计仍然受到有限的光谱表征和调色板验证的限制。尽管已编目了1000多种荧光蛋白，但缺乏用于表征其在多重方法中使用情况的系统资源。在这里，我们展示了一种在多重成像实验中选择和解码荧光蛋白的资源和方法。构建了一个包含44种荧光蛋白的文库，用于快速组装到哺乳动物表达载体中，并通过转座酶介导进行整合。对每种荧光蛋白进行了高光谱成像，并对光谱空间进行了数学表征。为了支持实验设计和数据解释，我们开发了XFP余弦相似性解码器（Cosine Similarity Decoder of XFP, CSDeXFP）工具箱，以预测光谱可分辨的荧光蛋白调色板，并对高光谱图像进行解码和重新上色。使用这种方法，我们展示了12个独特标记克隆的活细胞成像。我们的工作为选择用于多重实验的最佳荧光蛋白调色板提供了一个可扩展的平台，在各种生物系统和高光谱成像技术中具有广泛的实用性。
