Analyzing the transcriptome not only of whole tissues but of individual cells, opens completely new doors for understanding gene expression and its regulation. In the last few years, the application of this method exploded, ever increasing our knowledge about gene expression of specific tissues or cell types (Kanter and Kalisky, 2015; Xue et al., 2016; Farrell et al., 2018; Wagner et al., 2018; Adil et al., 2021; Mishra et al., 2022). In SCT analysis, individual cells are loaded onto a microfluidic chip and processed in such a fashion, that resulting cDNA libraries can retrospectively be assigned to a specific cell with high accuracy. Alternatively, a recently developed method named particle-templated instant partition sequencing (PIP-seq) avoids microfluid processing. Instead, it entails barcoding the cDNA after the cells were individually encapsulated in an emulsion (Clark et al., 2023). During the subsequent bioinformatical analysis, cells get sorted into groups, depending on their gene expression profiles (Figure 1). Thereby, we can examine individual and lineage associated cell expression from complex tissues. This leads to a detailed insight into the transcriptomic profiles of the studied tissues by applying current bioinformatics analysis tools like Seurat or Monocle that enable prediction of potential genetic interactions and regulatory networks (Trapnell et al., 2014; Satija et al., 2015). For example, in one of the pioneering studies, Wagner et al. pointed out the importance of chordin for the early development of the zebrafish larvae (Wagner et al., 2018). Ultimately, this can help unravel regulatory interactions and networks of genes crucial for cell development and maintenance. Naturally, this new information can prove to be highly useful when it comes to understanding and treating the occurrence of genetic disorders.

Vision is one of the most fundamental processes in our daily life and vision impairment can have a negative impact on the quality of life. Further, the eye is of particular interest for SCT analysis for multiple reasons. It is a highly complex organ, composed of a variety of tissues that are in turn made of different cell types (Chow and Lang, 2001; Sinn and Wittbrodt, 2013; Miesfeld and Brown, 2019). The probably best-known part of the eye is the retina, the neuronal layer in the posterior segment of the eye. This layer alone already includes the photoreceptor cells (i.e., rods and cones), bipolar cells, horizontal cells, amacrine cells and retinal ganglion cells, which are necessary to detect light and initiate the signal transduction after light detection (Diacou et al., 2022). Non-retinal ocular components such as the often-neglected retinal pigment epithelium (RPE), or the anterior segment are also composed of multiple tissues including cornea, ciliary body, lens and trabecular meshwork which in turn require many specific cell types. All these cell types have their characteristic gene expression profiles and require specific interactions and regulations to differentiate. Moreover, the different components of the eye are derived from different embryonic tissues (Cvekl and Tamm, 2004; Soules and Link, 2005). For instance, while some components of the anterior segment are derived from periocular mesenchyme cells, others originate from cranial mesoderm. Thus, for a healthy eye to develop, these embryonic tissues must be orchestrated and organized by a fine-tuned network of interacting genes (Zuber et al., 2003; Sinn and Wittbrodt, 2013). A single mistake in this developmental process can lead to a severely negative impact on eye function and ultimately even lead to blindness. Furthermore, any attempts to engineer cells in vitro (via approaches such as iPSC) for the purpose of in vivo treatment will require precise understanding of the pathways involved in differentiation and specification. As such, detailed examination of early developmental stages is of particular interest to examine and characterize.

The popularity of SCT analyses has drastically increased within the last few years. This led to the identification of gene expression profiles for several different organ systems and cell types in different species (Klein et al., 2015; Briggs et al., 2018; Pandey et al., 2018; Cao et al., 2019). Novel types of neurons were specificized in the part of the zebrafish forebrain called habenular. The gene nrp1a was found to be specific for the left habenular and there was a general difference in left and right genetic profiles among the neurons (Pandey et al., 2018). Further, SCT analyses of Xenopus tropicalis showed that the onset of several cell types, such as endothelium, is earlier than previously assumed (Briggs et al., 2018). Also, direct comparison between bulk RNA sequencing and SCT analyses showed that the latter should be preferably used when studying individual cell types since the former cannot be as precise (Hegenbarth et al., 2022).

Analyses of retinal tissues alone have been so substantial that metadata analyses have already been performed singularly on retina specific data (Macosko et al., 2015; Welby et al., 2017; Phillips et al., 2018; Clark et al., 2019; Voigt et al., 2019; Lu et al., 2020; Swamy et al., 2021; Yi et al., 2021). The outcomes of these studies have demonstrated for instance, that cones cannot specify properly without the gene atoh7 (Lu et al., 2020), gene expression of Müller glia cells differs among regions within the eye (Yi et al., 2021) and that NFI transcription factors play roles throughout eye development (Clark et al., 2019), pointing to the importance of SCT analyses. However, our understanding of eye development is far from complete. One of the issues with many popular model organisms is gathering sample tissue from multiple subsequent time points during development (Hegenbarth et al., 2022). The main limiting factors for this are the high costs of the SCT procedure and limited access to appropriate tissue, e.g., human donor organs (Welby et al., 2017; Phillips et al., 2018; Lukowski et al., 2019; van Zyl et al., 2022). Especially the limited access and ethical restrictions to embryonic human tissue are the main reasons, why transcriptome studies including early human development are still rare (Welby et al., 2017; Lu et al., 2020). While it can be expected that the general cost for SCT analyses will decrease over the next years, the problematic access to tissue of different developmental stages will remain a limiting factor. However, early developmental time points are of particular importance since gene expression at early stages can characterize the involvement of genes responsible for the early onset of diseases. In this respect zebrafish offer unique advantages.

In the following sections we aim to highlight the unique benefits that zebrafish offer when examining developmental eye transcriptomics including sample accessibility, precise tissue collection and functional analysis of candidate genes.

Zebrafish have been well-established as a model for different human congenital ocular diseases such as coloboma, retinitis pigmentosa, glaucoma, holoprosencephaly and anterior segment dysgenesis (Morris, 2011; Gestri et al., 2012; Link and Collery, 2015; Hong and Luo, 2021; Santhanam et al., 2022). The general similarities in ocular development and genetic conservation make zebrafish a powerful model for the study of ocular disease (see Figure 2A). To date, most zebrafish-based studies have focused on retinopathies, but lately zebrafish are also being employed for the study of anterior segment dysgenesis (Gestri et al., 2012; Richardson et al., 2017; Hong and Luo, 2021; Seese et al., 2021; Santhanam et al., 2022; Vöcking and Famulski, 2023). Despite some differences in anatomy, our own group recently suggested that the annular ligament (AL) of zebrafish can be used as a model for the human trabecular meshwork, highlighting the potential use of zebrafish as a glaucoma model (Vöcking and Famulski, 2023). In this study, we collected anterior segment (AS) mesenchyme cells from several developmental time via FACS and analyzed expression in AS specific GFP + cells of the anterior segment reporter transgenic line Tg(foxc1b:GFP). Not only could we identify cell types of the cornea and AL, but we were also able to follow clusters of these cells throughout the different timepoints and thereby monitor changes in their gene expression profiles. We were able to compare our data to human studies (van Zyl et al., 2020; van Zyl et al., 2022) and found that human marker genes such as krt4, krt5 and myoc are expressed in the zebrafish cornea and AL respectively. Moreover, we found previously unknown markers like hgd and cndp1 expressed in the AL which might also play a role in human TM development (Vöcking and Famulski, 2023). This kind of multiple time point analyses enables the exact determination of the genetic profile of cells during their development and help identify critical moments of their differentiation (Figure 2C) (also see Farnsworth et al., 2020; Farnsworth et al., 2021). Since studies of human samples are mostly limited to mature tissues, insights of early timepoints in model organisms such as zebrafish are crucial to understand early development. Compared to other study animals, this approach is most feasible to perform in zebrafish due to ease of tissue collection and their affordability (Farrell et al., 2018; Wagner et al., 2018; Farnsworth et al., 2020; Farnsworth et al., 2021; Vöcking and Famulski, 2023).