PREreview of Long-term hybridization in a karst window reveals the genetic basis of eye loss in cavefish

Overall Assessment:

Loss of vision and eyes in organisms adapted to subterranean or cave environments has occurred repeatedly, including at least twice in cave-dwelling populations of Astyanax mexicanus. Kellermeyer et al utilize a rare natural experiment: the co-occurrence of eyeless and eyed Astyanax mexicanus in the Caballo Morro cave to identify candidate genes responsible for the loss of eyes. They identify a candidate mutation in the lens gap-junction protein 43 Connexin-50 (Cx50), and use an impressive variety of comparative and functional analyses to confirm it has an impact on eye phenotype. They demonstrate that Cx50 CRISPants mutants have reduced eye size early in development, and that eye size differs as a function of Cx50 genotype in both lab and natural hybrids between eyed and eyeless fish. Kellermeyer et al identify the first variants directly implicated in eyelessness in cave fish and utilize a variety of techniques to confirm Cx50’s role in cavefish and subterranean mammals more broadly. While the authors claims about Cx50 are well supported, the presence of clear population structure in their data may inflate the false positive rate in their GWAS, impacting interpretation of their results.

Major points:

1. The authors use a genome-wide association study (GWAS) between eyed and eyeless Caballo Moro (CM) fish to identify candidate genes associated with eyelessness (Line 216). However, their results also indicate clear population structure in the cave fish population (Fig. 1F). They also note that CM fish are either fully eyed or completely eyeless with intermediate phenotypes being very rare (Line 90), and that there is spatial segregation of eyed and eyeless fish within the Caballo Moro cave (Line 124). Eyed and eyeless CM fish also appear to differ in phenotypes other than eye presence, most notably pigment (Fig. 1D, E). Together these results seem to suggest some form of reproductive isolation between eyeless and eyed fish, which could inflate false positives in the eye phenotype GWAS. The authors could directly address this issue by including a genome-wide PCA as a covariate in their analysis and subsequent SNP and selective sweep filtering analyses to rationalize the use of these analyses and aid the reader in accurate interpreting of the findings.

2. In Figure 2G, the authors show that CRISPants Cx50 mutants have significantly different eye sizes at 2 days post fertilization, and conclude that disruption of Cx50 resulted in severe eye defects (line 323). However, figure S9 C-D shows that there is no difference in eye size and or presence of lens between Cx50 mutants and wildtype at 6 days post fertilization. The authors could revise this statement to reflect whether Cx50 mutation causes morphological defects or instead only changes in the timing of eye development to more accurately convey this finding. As an alternative, the authors could perform a statistical test to assess whether their finding of cyclops phenotype only in cx50-CRISPants individuals (line 325) is a statistically significant result based on the number of non-mutant individuals they observed.

Minor points:

1. Line 523 - the authors speculate that the Cx50 mutations may be common in subterranean animals because its expression is limited to the lens, so may therefore be less pleiotropic. However, they also state (line 301) that the lens is the master regulator of eye regression in A. mexicanus and impacts the development of the entire eye. Clarifying the effect of Cx50 and the lens on other tissues in the discussion would aid the reader in understanding the impact of this gene.

2. The authors performed a behavior assay to test the aggression of eyed CM fish towards various intruders. Eyed CM fish were found to be equally aggressive to eyeless CM fish and eyed service fish. However, the aggression of eyed CM fish towards other eyed CM fish was not tested. Therefore it is not possible to determine whether eyed CM fish are especially aggressive towards eyeless CM fish, which could potentially explain the spatial segregation they observed. Including this additional test or discussing why it was not included would better connect their behavioral findings to the conditions of the natural environment.

I reviewed this preprint as part of the Transparent and Accountable Peer Review training program. This report was written with the support of my research mentor.

Competing interests

The authors declare that they have no competing interests.

Use of Artificial Intelligence (AI)

The authors declare that they did not use generative AI to come up with new ideas for their review.