How Imaging and Single-Cell Omics Are Reshaping Developmental Biology

Advanced imaging modalities and single-cell omics technologies are fundamentally transforming developmental biology research, transitioning the discipline from static, population-averaged observations into dynamic, spatially resolved systems biology. In a comprehensive review published in Cell Regeneration, researchers Chunling Wang, Xuejing Zhang, Yifan Zhang, and Feng Liu examine how the convergence of super-resolution microscopy, light-sheet fluorescence microscopy, and single-cell omics tools bridges observable cellular behaviors with underlying molecular regulatory programs. Historically, classical developmental investigations relied on descriptive embryology, positional cloning, mutagenesis screens, and bulk biochemical measurements, which successfully uncovered foundational principles of cell fate determination, lineage commitment, tissue patterning, and functional maturation. However, these traditional experimental approaches captured only low-throughput, population-level snapshots that remained blind to individual cell heterogeneity and native spatiotemporal dynamics. Although the genomic revolution of the 2000s enabled systematic reconstruction of gene networks, it still lacked single-cell resolution. The contemporary integration of advanced optical microscopy with single-cell transcriptomics, proteomics, and epigenomics systematically resolves this longstanding dilemma, overcoming the trade-off wherein imaging lacked deep molecular coverage and dissociation-based sequencing erased spatial coordinates.

The authors emphasize that super-resolution microscopy circumvents the fundamental optical diffraction barrier of classical light microscopy, which Ernst Abbe established at approximately 200 nanometers and which historically obscured critical nanoscale developmental events. Modern super-resolution microscopy achieves spatial resolutions ranging from 10 to 100 nanometers through synergistic physical principles and computational innovations. Illumination-field modulation approaches, including stimulated emission depletion microscopy and structured illumination microscopy, employ patterned illumination profiles to achieve spatial resolutions down to 20 nanometers. In parallel, single-molecule localization techniques, such as stochastic optical reconstruction microscopy and photoactivated localization microscopy, attain precision below 20 nanometers by sequentially activating and localizing individual fluorescent emitters. Recent computational developments in deep learning have further accelerated super-resolution reconstructions directly from conventional microscopy datasets. Beyond imaging static structures, super-resolution microscopy enables dynamic real-time tracking of living cellular processes. For example, combining stochastic optical reconstruction microscopy with structured illumination microscopy captured the structural remodeling of vimentin intermediate filaments during B-cell activation in mouse spleen tissue, while deep learning-assisted structured illumination microscopy resolved mitophagy events at 60-nanometer resolution during stem cell differentiation, demonstrating metabolic reprogramming during fate transitions.

While super-resolution techniques illuminate nanoscale structures, embryonic development requires multiscale observation across tissues, organs, and entire living organisms over extended developmental timeframes. Conventional confocal microscopy remains inherently limited for long-term in vivo tracking because of high phototoxicity and slow three-dimensional volumetric acquisition rates. Light-sheet fluorescence microscopy overcomes these obstacles by illuminating exclusively the focal plane of the sample with a thin laser sheet, thereby minimizing photodamage while supporting rapid, volumetric three-dimensional imaging of living specimens. The technical scalability of light-sheet microscopy accommodates continuous imaging from single cells to intact embryos. At the organismal level, light-sheet systems permit non-invasive observation of zebrafish embryos throughout gastrulation until organ primordia emerge, facilitating comprehensive cell lineage tracing and morphogenetic analysis. At the organ level, light-sheet imaging captures three-dimensional vascular networks and immune cell distributions within intact murine brains and hearts. When combined with serial block-face scanning electron microscopy, correlative light and electron microscopy provides ultrastructural perspectives on hematopoietic stem and progenitor cells in zebrafish kidney marrow, whereas tissue clearing methods such as 3DISCO facilitate three-dimensional anatomical atlases of peripheral nerves in human embryos.

An emerging technical direction highlighted in the review involves integrating super-resolution microscopy directly into light-sheet architectures to capture subcellular events across whole living organisms. Adaptive super-resolution light-sheet systems, such as Meta-rLLS-VSIM, deliver near-isotropic 150-nanometer spatial resolution at a one-hertz frame rate in living zebrafish embryos. This technical configuration facilitates minimally invasive tracking of organelle interactions and trafficking dynamics throughout early embryonic development. By reconciling the spatial advantages of nanometer optical resolution with the low phototoxicity and volumetric speed of selective plane illumination, hybrid platforms bridge the historical divide between static subcellular anatomy and living physiological function. Consequently, researchers can monitor dynamic organelle contacts and structural reorganizations in authentic developmental contexts, establishing a dynamic visual record of how subcellular behaviors translate into multicellular morphogenesis without disturbing the living organism under investigation.

Concurrently, single-cell omics technologies serve as an indispensable framework for deciphering cellular heterogeneity, which functions as the primary engine driving developmental fate decisions from initial pluripotency to specialized cell types. Single-cell RNA sequencing profiles gene expression across thousands of individual cells, illuminating rare cellular subtypes and branching differentiation trajectories. In mouse embryogenesis, single-cell RNA sequencing mapped a comprehensive developmental atlas comprising more than 500 distinct cell subtypes and 56 developmental trajectories between embryonic days E9.5 and E13.5, while subsequent studies categorized cellular states across 19 consecutive developmental stages from E3.5 to E13.5. Single-cell transcriptomic investigations in human embryonic development uncovered an early lineage bias toward hematopoiesis among mesodermal progenitors, while placental studies confirmed that embryonic macrophages originate from CD44-positive endothelial cells. Beyond healthy morphogenesis, single-cell sequencing delineates cell-type-specific transcriptional alterations in early pulmonary fibrosis and familial tauopathy pathogenesis. Single-cell proteomics by mass spectrometry measures protein abundance shifts during stem cell differentiation, single-cell DNA sequencing tracks clonal evolution patterns during leukemia progression, single-cell metabolomics assesses dynamic metabolic state alterations, and multimodal single-cell assays demonstrate distinct epigenetic landscapes emerging in ectodermal and mesodermal lineages during gastrulation.

Despite its power in cataloging cellular heterogeneity, dissociation-based single-cell sequencing destroys spatial architecture, severing cells from positional cues, morphological context, and intercellular communication networks essential for morphogenesis. Spatially resolved omics technologies resolve this critical limitation by registering molecular measurements directly to native histological coordinates. Sequencing-based spatial transcriptomic technologies operate across complementary spatial resolutions. Slide-seq utilizes arrays of 10-micrometer polystyrene beads to achieve whole-transcriptome capture at single-cell resolution, whereas spatial enhanced resolution omics-sequencing employs dense spatial barcoding chips to achieve subcellular resolution at or below 500 nanometers within tissue microenvironments. These spatial transcriptomic approaches have systematically charted mammalian embryogenesis, including computational spatial modeling across mouse embryonic stages E6.5 to E7.5, spatial enhanced resolution omics-sequencing from E7 to E8, and Slide-seq profiling between E8.5 and E9.5. In human development, three-dimensional spatial transcriptomic reconstructions of gastrulating embryos resolved positional gene expression profiles and spatial registers of distinct cell types. Spatial sequencing also mapped spatiotemporal transcriptomic dynamics across mouse placental development from E7.5 to E14.5, established spatial developmental trajectories in early zebrafish embryos, and generated dynamic spatiotemporal atlases of zebrafish cardiac regeneration.

Complementing sequencing-array methods, multiplexed image-based spatial technologies achieve subcellular transcript detection directly through in situ fluorescence microscopy. Multiplexed error-robust fluorescence in situ hybridization detects targeted RNA species through sequential hybridization rounds and optical imaging, using fluorescent beads as fiducial registration markers to achieve an alignment error of approximately 20 nanometers. Owing to this nanometer registration precision, multiplexed error-robust fluorescence in situ hybridization delineates cytoarchitecture in complex structures such as synapses and neural circuits. In human fetal development, this technique characterized cellular and molecular development across seven gestational stages and eight cortical regions, uncovering the early emergence of the six-layered neocortical organization and regional specialization, while resolving cellular organization in the fetal dorsal pons. In pathology, multiplexed fluorescence in situ hybridization mapped the spatial distribution, transcriptional signatures, and cell-cell interactions of fibroblasts during murine colitis progression. It also determined the zonal localization of hepatocytes alongside distinct macrophage and mesenchymal subpopulations in fibrotic liver tissues, directly correlating cellular positioning with pathological microenvironments.

The review demonstrates that the relationship between imaging and single-cell omics has evolved beyond parallel experimentation into a continuous, iterative cycle of mutual guidance and cross-validation. In this virtuous feedback loop, dynamic imaging observations frequently guide omics sampling to pinpointed anatomical sites and temporal windows. For example, the Zebrahub project employed light-sheet lineage reconstructions to construct a spatiotemporal map that directly informed tissue sampling strategies and subsequent single-cell sequencing data interpretations in zebrafish embryos. Similarly, coupling in vivo lineage tracing with single-cell RNA sequencing and single-cell assay for transposase-accessible chromatin resolved hematopoietic endothelial cell heterogeneity, with imaging-based lineage tracking providing the necessary interpretative framework. In murine models, genetically encoded, imaging-readable barcodes guided the isolation of specific cell lineages for downstream single-cell sequencing. Reciprocally, unbiased omics analyses identify candidate pathways and regulatory signatures that require spatial and functional validation through imaging. In congenital heart disease research, single-nucleus and single-cell RNA sequencing identified aberrant cell states and immune microenvironments, which researchers subsequently verified at the protein level and mapped spatially using imaging mass cytometry. A comparable omics-first discovery and immunofluorescence validation pipeline confirmed localized immune landscapes in dermatomyositis.

This technological convergence has fueled the development of integrated commercial platforms that combine sequential in situ hybridization, enzymatic decoding chemistries, and automated fluorescence microscopy. Platforms such as Xenium and CODEX achieve subcellular spatial localization of targeted mRNA transcripts at 200-nanometer precision while integrating histological stains, such as hematoxylin and eosin, or fluorescent antibody markers on the same tissue section. By overlaying high-resolution morphological images with spatial transcriptomic profiles, these platforms enhance cell typing and tissue microenvironment characterization. Furthermore, Xenium distinguishes subtle transcriptional differences between nuclear and cytoplasmic compartments across diverse human tissues, including human skeletal specimens, fetal spine structures, kidney sections, and lung epithelium. Similarly, integrating single-cell RNA sequencing with CODEX multiplexed imaging revealed cellular biogeography within human bone marrow niches. In rodent models, Xenium has mapped diverse regions of the mouse brain, identifying unique cellular subpopulations and local intercellular molecular interactions that drive neural organization and developmental patterning.

Looking ahead, the authors outline three interconnected technological frontiers required to transform developmental biology from an observational discipline into an analytical and predictive science. First, future multi-omics methodologies must achieve simultaneous, non-destructive molecular profiling within intact living systems. A critical objective is coupling live-cell transcriptomics, such as Live-seq using fluidic force microscopy and low-input sequencing, with light-sheet fluorescence microscopy to record temporal transcriptomic histories while tracking cell migration and division dynamics in real time. Second, sophisticated artificial intelligence frameworks, including graph neural networks and multimodal variational autoencoders, are essential to fuse multidimensional spatial coordinates, temporal gene expression profiles, and protein interaction networks into predictive computational models capable of forecasting cellular responses to genetic or environmental perturbations. Third, advancing from static, fixed-sample assays toward continuous live-cell multi-omics imaging will provide the temporal resolution necessary to validate mechanistic models. By unifying deep molecular dictionaries with dynamic living observations, this expanding paradigm promises to uncover fundamental mechanisms governing embryogenesis and accelerate translational applications across regenerative medicine, disease modeling, and therapeutic intervention.

Subject of Research: Integration of advanced imaging and single-cell omics in developmental biology

Article Title: When imaging technology meets single-cell omics: new paradigm in developmental biology

Article References: Wang, C., Zhang, X., Zhang, Y., & Liu, F. (2025). When imaging technology meets single-cell omics: new paradigm in developmental biology. Cell Regeneration, 14(1), Article 53. https://doi.org/10.1186/s13619-025-00276-4

Image Credits: AI Generated

DOI: 10.1186/s13619-025-00276-4

Keywords: developmental biology, single-cell omics, super-resolution microscopy, light-sheet fluorescence microscopy, spatial transcriptomics, Stereo-seq, MERFISH, Live-seq, imaging, technology, meets, single-cell