Answers are grounded in this paper and cite their sources.
Identifies genetic factors influencing eye loss in cavefish, suggesting complex trait evolution.
Cave dwelling animals experiencing constant darkness show the common adaptation of vision loss (Krishnan and Rohner, 2017). Eye loss is an active developmental process in Mexican tetra (Astyanax mexicanus; cavefish), where the eye forms during embryogenesis but then degenerates due to lens apoptosis (McGaugh et al., 2014). As such, the cavefish is a valuable organism to study evolutionary rewiring of developmental programmes.In evolutionary genetics, it is a major challenge to identify casual genes and variants for complex traits within broad trait-associated genomic regions. The Mexican tetra is a powerful system to overcome this challenge since surface and cave forms are interfertile, genetically tractable and amenable to genome editing (Gross, 2012; Protas et al., 2008; Rohner, 2018; Stahl et al., 2019; Warren et al., 2021). Despite all these advantages, the genetic basis for cavefish eye loss remains only partly resolved.To address this problem, two recent preprints provide complementary approaches. Shennard et al. (2025a preprint) combined quantitative trait locus (QTL) mapping, gene expression, population genetics, single-nucleus RNA-sequencing and CRISPR screening to identify genes involved in cavefish eye degeneration. Kellermeyer et al. (2024 preprint) used a natural hybrid population to uncover a functional mutation in connexin-50 linked to eye loss. Together, these studies show how experimental and natural genetic approaches can reveal genotype–phenotype relationships.Cavefish eye loss is a polygenic and developmentally complex trait. Previous work has implicated pathways such as sonic hedgehog signalling, rx3 and cbsa (Ma et al., 2020; Shennard et al., 2025b; Yamamoto et al., 2004), but many causative genes remain unknown. Shennard et al. (2025a preprint) have addressed this problem by designing a multi-step strategy to prioritize candidate genes for functional testing in A. mexicanus.The research team began with previously mapped eye QTL from surface fish×Pachón cavefish crosses, covering traits such as eye size, lens size and retinal thickness. To narrow down broad intervals, two filters were applied: differential expression in developing eyes at 54 h post-fertilization (hpf) and signatures of positive selection in wild Pachón populations. This resulted in 29 candidate genes, not previously linked to eye development. Single-nucleus RNA-sequencing at 6 days post-fertilization (dpf) showed that these genes are enriched in relevant retinal cell types, including photoreceptors and retinal pigment epithelium-like cells. The key strength of this study is its functional screen using F0 CRISPR-Cas9 mutagenesis in surface fish. The authors targeted 28 of 29 candidate genes and measured effects on eye size, pupil size and visual behaviour. Validation with rx3 produced severe eye and behavioural defects. The screen identified some genes affecting eye size: otpb, fbln7 and vtg3 reduced eye size, whereas lin7a, spdef, cyp24a1, lhx6, and pfkmb increased it. Seven effects were confirmed with independent guide RNAs.These results demonstrate that cavefish eye degeneration may involve several developmental processes rather than regulation by one gene. It appears that most candidate genes do not have cavefish-specific coding variations, making it possible to state that eye degeneration might be a result of a regulatory shift rather than protein-coding variations. The gene cyp24a1 appears to have substantial coding differences and thus proves that both protein-coding and regulatory variations can be involved in trait development. From among the validated genes, fbln7 had a temporary impact on eye and pupil development, affecting the early larva phase only.Whereas Shennard et al. used an experimental screening strategy, Kellermeyer et al. took advantage of an unusual natural experiment. In the Caballo Moro cave, a collapsed ceiling created a karst window, allowing light into parts of an otherwise subterranean pool. As a result, eyed and eyeless Astyanax fish coexist within the same cave system, with illuminated and dark habitats located side by side and no obvious geographic barrier between them.This ecological setting provides an exceptional opportunity. Despite potential gene flow, most fish in the cave are either fully eyed or completely eyeless, with very few intermediate phenotypes. Genetic analysis revealed that the eyed Caballo Moro fish are long-standing hybrids derived from surface and cave lineages. This natural hybridization effectively created a recombination experiment in the wild, allowing the authors to narrow the candidate genetic architecture of eye degeneration far more sharply than is usually possible with laboratory QTL mapping.Using whole-genome sequencing of eyed, eyeless and surface fish, the researchers identified genomic regions associated with the eyeless phenotype and then filtered for variants that distinguished eyeless fish from both eyed cavefish and surface fish. This reduced the search space to 203 candidate single-nucleotide polymorphisms (SNPs) across 41 genes. Among these, one variant stood out: a nonsynonymous mutation in cx50, which encodes connexin-50, a lens gap-junction protein also known as GJA8 in mammals.The biological plausibility of cx50 is strong. In vertebrates, CX50 is highly expressed in the lens and is required for lens development, transparency and homeostasis. Human mutations in this gene are associated with congenital cataracts and microphthalmia. In Caballo Moro cavefish, the authors identified a serine-to-lysine substitution (S89K) in a conserved transmembrane region of cx50. CRISPR disruption of cx50 in surface fish caused severe early eye defects, confirming its role in eye development. The study further shows independent targeting of cx50 in different cavefish lineages: Guatemala cavefish carry S89K, while El Abra cavefish carry a distinct T39M mutation associated with reduced eye size in both lab and natural populations.Kellermeyer et al. extended this pattern across taxa, identifying additional nonsynonymous cx50 variants in cave-adapted fish and subterranean mammals. Structural modelling suggested these mutations may disrupt the connexin channel pore and lens physiology. Finally, introducing the S89K mutation into mice caused cataracts and reduced eye and lens size, functionally validating its effect in a mammalian system. Together, these findings identify cx50 as a convergent, lens-specific target, mutations of which reduce eye development with limited pleiotropic effects.These two preprints offer complementary evidence regarding trait evolution. In the first study, Shennard et al. demonstrated by employing QTL mapping, expression studies, single-cell profiling, and CRISPR validation that eye degeneration in cavefish entails many genes and regulatory pathways. In contrast, Kellermeyer et al. discovered a large effect coding variant through natural hybridization.Several key questions remain. How do regulatory changes in multiple eye-development genes interact with coding variants such as cx50? Are candidate gene effects additive, epistatic or stage-specific? Do repeated cases of eye degeneration use the same molecular targets or distinct developmental routes, and is selection acting directly on eye loss or indirectly via linked traits such as behaviour or metabolism?Overall, these studies shift the field from mapping eye-loss loci toward identifying causal genes, variants and developmental mechanisms. More broadly, they show how natural evolutionary systems, single-cell genomics and genome editing together can link genotype to phenotype in evolution.
No takes yet. Share an insight, caveat, or question.
Jain et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: