An antagonistically pleiotropic gene regulates vertebrate growth, maturity, and lifespan

Introduction

Vertebrates exhibit a remarkable degree of variation in their life history traits, including size, age at maturity, and lifespan. This extreme diversity is exemplified by the ~1000-fold difference in lifespan between the short-lived turquoise killifish¹ and the centuries-old Greenland shark². The antagonistic pleiotropy theory of aging³ (AP) aims to explain part of this variation by proposing that some genes may enhance fitness early in life while being detrimental late in life, leading to positive selection of genes that ultimately limit lifespan^{4,5,6,7,8,9,10}.

Insight into the regulatory networks that may include AP alleles can be gained, for example, through genome-wide association studies (GWAS) that link early-life performance and age-related diseases^{11,12,13,14,15,16,17}. Other approaches, such as experimental evolution in model organisms, have demonstrated that selection for late-life reproduction can increase lifespan and reduce early-life fecundity^18,19,20. However, such studies do not address the underlying genetic mechanisms at a single gene resolution, which could be determined through genetic perturbations (e.g., the C. elegans trl-1 mutant²¹).

Several targeted gene mutations with increased longevity have been identified in laboratory animals²². Some of these mutations display robust trade-offs with growth and reproduction, as seen in the Ames and Snell dwarf mice^23,24,25,26, and following genetic manipulation of IGF1 signalling^27,28. However, there are a few examples in vertebrates where putative molecular mechanisms of antagonistic pleiotropy are directly linked to the regulation of specific life-history traits within a normal, biologically relevant range of expression (but see refs. ^24,29). For instance, Ames and Snell dwarf mice are long-lived but sterile, indicating that such deleterious mutations will be strongly selected against in natural populations.

Recent GWAS in humans^30,31 and wild male salmon populations^32,33 link polymorphism in the vestigial-like family member 3 (vgll3) locus with a variation in age at maturity. These findings suggest VGLL3 might have evolutionarily conserved functions regulating pubertal timing in vertebrates. Fish species exhibit a vast range of evolutionary adaptations, providing an experimental playground for modeling complex traits in health and disease^{22,34,35,36,37}.

Here, we leverage the turquoise killifish (Nothobranchius furzeri) as a model to investigate the role of VGLL3 as a regulator of antagonistically pleiotropic effects in vertebrates. With rapid sexual maturation (~2–4 weeks³⁸) and a naturally short lifespan (4–6 months¹), the genetically tractable killifish model^39,40,41 offers a powerful platform to investigate the molecular underpinnings of vertebrate aging^{39,42,43,44,45,46,47,48} and life-history differences⁴⁹.

Targeting two evolutionarily conserved vgll3 isoforms accelerates somatic growth, specifically in males, and advances puberty onset in a dose-dependent manner. Transcriptomic profiling, together with supporting cellular and physiological analyses, suggested involvement in regulating cell cycle, somatic stem cells and germline proliferation, and DNA repair. Aged male mutants exhibit melanoma-like tumors with high penetrance, validated through cancer engraftment studies into a newly developed immunodeficient rag2 mutant killifish. Consistent with the concept of antagonistic pleiotropy, we show that vgll3 is under positive selection in killifish and that male vgll3 mutants exhibit a significantly reduced lifespan along with an increased incidence of cancer.

These experimental findings identify vgll3 as a major-effect gene with antagonistically pleiotropic, age-opposed effects. These findings are consistent with a trade-off between early-life benefits, such as accelerated growth and reproductive maturity, and the heightened risk of late-life disease and mortality, warranting further investigation in wild populations.

Results

Cell-type and sex-specific gonadal expression of vgll3

In humans, VGLL3 protein expression is primarily found in the male gonads⁵⁰. By analyzing our recently generated single-cell data from killifish testes and ovaries⁴⁴, we identified the specific gonadal cell types that express vgll3 (Fig. 1a, left, and Supplementary Fig. 1a). Our analysis revealed that vgll3 is primarily expressed in the somatic cells of the gonad, particularly in male-specific Sertoli and Leydig cells (Fig. 1a, right, and Supplementary Fig. 1b), consistent with a sexually dimorphic role.

Fig. 1: Identification and disruption of vgll3 isoforms in killifish. — **Fig. 1: Identification and disruption of *vgll3* isoforms in killifish.**

We then investigated the spatial distribution of the vgll3 transcript in male gonads using single-molecule fluorescence in situ hybridization chain reaction (smFISH HCR, Supplementary Fig. 1c). To identify specific cell types, we compared vgll3 expression with known markers, including amh (anti-Müllerian hormone) for Sertoli cells, and the classical germ cell marker ddx4/vasa (dead-box helicase 4) as a reference. Notably, unlike antibody staining, smFISH detects individual mRNA molecules; therefore, true overlap of signals is not expected⁵¹, especially for transcripts expressed at lower levels. Together, our single-cell analyses confirm that vgll3 is expressed in male gonadal support cells.

Genetic manipulation of distinct vgll3 exons in killifish

Recent findings in salmon suggest that the vgll3 genotype, associated with the timing of male puberty^32,33, may influence the expression levels of vgll3 transcript isoforms^52,53. These include a shorter isoform identified in salmon that arises from an alternative 5′ UTR within the first intron and is likely translated from an alternative start codon in exon 2, providing an additional strategy for controlling expression levels⁵².

Interestingly, we observed that a shorter conserved vgll3 isoform is present across many vertebrate species (Fig. 1b), including fish, such as the predatory carp (Chanodichthys erythropterus) and zebrafish (Danio rerio). Other vertebrates include birds (e.g., the chicken Gallus gallus and Japanese quail Coturnix japonica), smaller mammals, such as the naked mole-rat (Heterocephalus glaber) and the golden spiny mouse (Acomys russatus), as well as bats (the Velvety free-tailed bat Molossus molossus). It also appears in larger mammals, such as the lion (Panthera leo), and marsupials like the common wombat (Vombatus ursinus). Transcripts mostly originate from an alternative 5’ UTR within the first intron (e.g. carp, Fig. 1b, left), and are translated from a conserved alternative start site (Fig. 1b, right).

To further investigate whether killifish also express a short vgll3 isoform, we re-analyzed publicly available RNA-seq data from adult male gonads⁵⁴ using Cufflinks⁵⁵ (Fig. 1c). This analysis predicted both the long and short vgll3 isoforms, supporting the presence of an alternative transcription start site within the first intron (Fig. 1c). To quantify expression levels, we performed qPCR on RNA pooled from the testis of prepubertal and pubertal killifish, revealing a significant predominance of the longer transcript (Fig. 1d). It is important to note that our data cannot determine whether these isoforms are co-expressed in the same cell type (or in different cells).

To explore whether VGLL3 has a functional role in vertebrate life-history, we edited the killifish vgll3 using established CRISPR protocols^56,57 (Fig. 1e). Specifically, we either edited the first or the third exons of vgll3 (vgll3^Ex1 and vgll3^Ex3), introducing frameshift mutations predicted to result in loss-of-function alleles (Fig. 1e, and see Methods). Importantly, targeting the third exon is expected to disable both isoforms. We then outcrossed heterozygous fish for several generations to reduce the burden of possible off-target mutations.

Vgll3 mutants display accelerated growth and maturity

Nuptial coloration, a body-color change associated with permanent or seasonal sexual maturation, is common across various vertebrates and invertebrates and plays a crucial role in reproductive success. In killifish, nuptial coloration is mostly visible in the male caudal fin and is correlated with the onset of sperm production and mating^45,58. Therefore, as a proxy for sexual maturity, the variability in the onset of tail coloration was visually scored at the age of one month (Fig. 2a).

Fig. 2: Molecular and phenotypic profiling of vgll3 isoform mutants. — **Fig. 2: Molecular and phenotypic profiling of *vgll3* isoform mutants.**

Focusing on male physiology (as predicted by the salmon GWAS^32,33), our data indicated that a significant acceleration in male maturation occurs in homozygous exon 1 mutants (vgll3^Ex1/Ex1, p = 0.008, Fig. 2a), and heterozygous exon 3 mutants (vgll3^Ex3/+, p = 0.005, Supplementary Fig. 2a). Interestingly, homozygous exon 3 mutants (vgll3^Ex3/Ex3) displayed a reverse trend (p = 0.049, Fig. 2a). Heterozygotes for the Exon 3 mutation also show accelerated puberty, suggesting a similarity between Exon 3 heterozygotes and Exon 1 homozygotes. This supports the interpretation that the isoform effect could result from changes in the vgll3 gene dosage.

Transcriptional analysis of vgll3 isoforms suggests distinct and overlapping functions

To gain further molecular insight, we performed RNA sequencing on primary tail fibroblasts isolated from WT fish and both homozygous vgll3 mutants. While fibroblasts may not directly contribute to the organism-level phenotypes, they provide a controlled cell population, allowing us to focus on cell-autonomous phenotypes following the different vgll3 manipulations. Additionally, this strategy minimizes confounding effects from tissue-specific differences in cellular composition that can arise during sexual maturation.

Similar to the onset of nuptial colors, principal component analysis (PCA) suggested linear relationships between the genotypes, with wild-type (WT) samples positioned centrally along the first principal component (PC1, Figs. 2b, left and S2b). The transcriptomic analysis indicated partially reduced vgll3 expression levels in mutant-derived primary cells, likely due to nonsense-mediated decay (NMD, Supplementary Fig. 2b, as previously reported for other killifish mutations⁴⁵).

The selective effect of NMD is expected according to established NMD prediction rules^59,60 (e.g., the position of the exon within the gene, indel position within an exon, and distance to exon–exon junctions). Specifically, although both indels introduce premature stop codons, the exon-1 lesion sits below the ~200 bp from the start-site threshold^59,60, and is therefore unlikely to trigger NMD. By contrast, the exon-3 mutation lies 491 bp upstream of the final exon–exon junction, exceeding the >55 bp rule, and is predicted to elicit NMD in both isoforms.

A Venn diagram of differentially expressed genes between each mutant and the WT revealed only partial overlap (Fig. 2b, right). These findings are consistent with the reversed trends of the observed puberty phenotypes, where homozygous exon 3 mutants (Ex3) delay puberty, while exon 1 mutants (Ex1) accelerate it (Fig. 2a). Based on this, we applied a linear model, followed by gene set enrichment analysis (GSEA). These enrichments highlighted multiple biological processes associated with linear expression trends across genotypes (Fig. 2c), particularly when expression is highest in exon 1 mutants (Ex1 > WT > Ex3).

Genes upregulated in Exon 1 mutants were enriched for pathways supporting cell-cycle progression, including mitotic cell cycle, cellular respiration, and ribosome/translation. Additional pathways were linked to DNA repair, reproduction, and steroid hormone biosynthesis (Fig. 2c, Supplementary Data 1). In contrast, genes upregulated in Exon 3 mutants were enriched for pathways related to growth regulation and cell death (Fig. 2c, left, and Supplementary Data 1).

Several representative genes upregulated in exon 1 (Fig. 2b, right, and Supplementary Data 1) include the androgen-responsive spermatogenesis genes strbp⁶¹ (Spermatid Perinuclear RNA Binding Protein), possibly linking molecular changes to male reproduction and spermatogenesis (Fig. 2c, right). Meanwhile, the androgen receptor (ar), which is critical for the male pubertal growth spurt and maturation of secondary sexual characteristics⁶², was downregulated in both mutants, suggesting modified androgen signaling (Fig. 2c, right).

Several genes with significantly downregulated expression in exon 1 mutants are also linked to the DNA damage response (DDR), including hdac4 (Histone Deacetylase 4, Fig. 2c, right), a chromatin-regulator that modulates homologous recombination⁶³. Loss of HDAC4 impairs proper repair and leads to persistent DNA damage. Recent studies have linked vgll3 to the Hippo pathway in Atlantic salmon^64,65. Similarly, several downregulated genes, such as ar and glut3 (slc2a3) are known Hippo targets^66,67 (Fig. 2c, right).

In summary, VGLL3 manipulation drives distinct transcriptional programs with predicted effects on cell proliferation, hormone signaling, spermatogenesis, and DNA repair. To initially examine whether VGLL3 is directly recruited to sites of DNA damage, we used laser-induced damage⁶⁸ (LID) combined with live-cell imaging in primary cells expressing a GFP-tagged short or long VGLL3 isoform (Addgene # 241148, 241149). While the macro domain of macroH2A⁶⁹ (used as a positive control), localized to the damage site, neither of the VGLL3 isoforms showed recruitment (Supplementary Fig. 2c). We next explored whether VGLL3 plays a functional role in cell cycle or DNA damage repair.

Cell proliferation was quantified using incorporation of 5-ethynyl-2′-deoxyuridine (EdU; Fig. 2d, left), and the DNA damage response was characterized via γH2AX nuclear staining following genotoxic stress (Etoposide treatment, Fig. 2d, right). Our data revealed that, consistent with the trends observed in the transcriptomic data, vgll3^Ex1/Ex1-derived cells proliferate faster than WT and vgll3^Ex3/Ex3-derived cells, and exhibit increased γH2AX foci following damage. Next, we investigated whether these molecular and cellular signatures are associated with physiological outcomes in vivo.

vgll3 ^Ex1/Ex1 males display accelerated growth

Following the pubertal and cellular phenotypes, we decided to focus on vgll3^Ex1/Ex1 mutants. Measuring body length of adult males and females suggested a male-specific trend (p = 0.51 in females, and p = 0.07 in males, Supplementary Fig. 3a). Therefore, as suggested by the male-specific role of vgll3 in Atlantic salmon^32,33 and its sexually dimorphic expression in killifish (Fig. 1a, right, and Supplementary Fig. 1b), we decided to concentrate on male physiology. We measured the cumulative effect on male growth, which demonstrated a significant increase in body length (p = 0.015, Fig. 3a). Similarly, we observed an increase in other related attributes, including weight (at 3 months of age, p = 0.0047, Fig. 3b), and depth (in old fish, p = 0.037, Supplementary Fig. 3b).

Fig. 3: Pleiotropic effects of vgll3 exon 1 mutants. — **Fig. 3: Pleiotropic effects of *vgll3* exon 1 mutants.**

The gonadosomatic index (GSI) measures gonad mass as a percentage of total body mass and is calculated with the GSI formula: (gonad weight / total body weight) × 100. This index is used to directly assess the sexual maturity of animals, as it correlates with gonadal development. While both body and gonadal weight increased in vgll3^Ex1/Ex1 mutant males (at 2.5-3 months old, Figs. 3b, S3c, left), GSI was significantly higher, confirming accelerated sexual maturation (p = 0.0024, Fig. 3c).

Quantifying the relative proportion of proliferative testicular cysts revealed an increase in germline proliferation in exon 1 mutants (p = 1.55 × 10^-6, Fig. 3d, top), which was correlated with an increase in mature sperm (p = 0.000048, Supplementary Fig. 3c, right). Similar to the results observed in cells, an increase in γH2AX nuclear staining was also detected in the testis (p = 0.04, Supplementary Fig. 3d). Beyond the germline, an increase in proliferating somatic stem cells was observed in intestinal crypts (p = 3.4×10^-7, Fig. 3d, bottom), suggesting that increased growth might be mediated through stem cell proliferation. These findings are notable given that killifish hatchlings already exhibit one of the fastest growth rates among vertebrates³⁸, reaching sexual maturity in as little as ~3 weeks^1,70. Together, our data suggest that vgll3 expression levels may act to fine-tune vertebrate maturity.

vgll3 ^Ex1/Ex1 males exhibit an increase in age-related melanoma-like expansion

Killifish naturally exhibit rapid growth and early maturation, an adaptation that enhances fitness by enabling breeding before habitat desiccation. However, does accelerating growth and maturation through vgll3 disruption come at a cost? The antagonistic pleiotropy theory of aging proposes that certain genes enhance an organism’s performance early in life but later have harmful effects. These genes can be favored by natural selection because the force of selection on traits is stronger during early life⁷, and such genes can be selected for, even if they contribute to a mild decline in late-life performance.

To explore this paradigm, we aged WT and vgll3^Ex1/Ex1 mutant populations. Interestingly, we observed that older mutant males exhibited a significant increase in melanoma-like expansion incidences in the caudal fin (Fig. 3e), suggesting a potential evolutionary trade-off. Importantly, tracking hundreds of aged fish revealed that naturally occurring expansions are relatively rare in WT individuals (Supplementary Fig. 3e). However, confirming whether these melanocytic expansions are truly cancerous requires developing a set of tools for functional interrogation.

Generation and characterization of a rag2 immunodeficient model

Tumor transplantation studies into an immunodeficient host are a gold-standard approach for investigating cancer biology in model organisms^71,72,73. Therefore, we generated an immunocompromised model to enable cancer engraftment in killifish by mutating the rag2 gene^73,74,75 (Fig. 4a). Specifically, we generated a ~ 250 base-pair deletion allele (Δ250) in the single exon of the killifish rag2 coding sequence (Fig. 4a, left). We outcrossed this line for several generations and generated a fertile homozygous fish.

Fig. 4: Generation of a rag2 immunodeficient killifish model. — **Fig. 4: Generation of a *rag2* immunodeficient killifish model.**

Murine rag2 mutants display hematopoietic and immune system defects, including arrested B-cell and T-cell development⁷⁶. Therefore, we compared the transcriptome of the killifish head kidney, a dual-function organ responsible for blood filtering and hematopoiesis in teleosts⁷⁷, from three male WT and three male rag2 mutants. Principal component analysis (PCA) demonstrated that both groups segregate according to PC3 (Figs. 4a, right, and S4a). The differential expression analysis revealed 14 downregulated and 3 upregulated genes (fold change > 2, FDR < 0.05, Fig. 4b, Supplementary Data 1). Interestingly, downregulated genes were directly related to rag2 functions, including members of the immunoglobulin gene family (Supplementary Fig. 4b). These data suggest that, as expected, failure to rearrange the V(D)J gene segments also perturb expression.

Upregulated genes were sparse and linked to immune and signaling-related pathways, including the novel immune-type receptors (NITRs)^{78,79,80,81,82,83} (i.e., nitr13.2, Supplementary Fig. 4b). NITRs have been identified in all major lineages of teleost fish⁸⁴, and are structurally similar to the killer immunoglobulin-like receptors (KIR) on mammalian natural killer (NK) cells⁸². These receptors do not depend on V(D)J recombination⁸⁴ and are expressed in teleost NK-like cells^82,85. Interestingly, a similar expansion of NK-like cells was detected in zebrafish rag2 mutants⁸¹, suggesting an evolutionarily conserved mechanism.

Lack of V(D)J recombination in rag2 mutants

V(D)J recombination is a genetic mechanism that rearranges gene segments within immunoglobulin and T cell receptor loci⁸⁶. This process occurs in developing B and T lymphocytes and is essential for generating the vast diversity of antibodies (in B cells) and T cell receptors (in T cells) needed to recognize a wide array of pathogens. Specifically, the V (Variable), D (Diversity), and J (Joining) gene segments are randomly selected and joined together through the action of the recombination-activating genes 1 and 2 (rag1 and rag2).

The killifish immunoglobulin heavy chain (igh) gene locus⁸⁷ consists of a constant region (C), and several variable V(D)J regions (Fig. 4c, top). To assess V(D)J recombination in rag2 mutants, we adapted a recently developed 5’ Rapid Amplification of cDNA Ends (5’ RACE) protocol⁸⁸. This technique is used to identify the 5′ end of an mRNA transcript, particularly when the transcription start site or upstream exon structure is unknown, as in cases involving specific combinations of V(D)J gene segments. Briefly, we extracted RNA from the head kidneys of WT and rag2 mutant fish and performed cDNA amplification using a specific primer for the constant region. Template-switching further allowed us to amplify all igh recombination variants (Fig. 4c, and see Methods).

This protocol produced a ~ 600 base pair amplicon in WT fish, consistent with mature immunoglobulins⁸⁸. Conversely, we did not detect these amplicons in rag2 mutants (Fig. 4c, bottom, the shorter, faint bands are likely a result of non-specific amplification). This suggests the absence, or significant reduction, of V(D)J recombination in rag2 mutants. This suggests that the killifish rag2 mutant is an immunodeficient genetic model.

Functional characterization of vgll3-mediated melanoma

To explore the tumorigenicity of the melanocyte expansion seen in old vgll3 mutants, we first developed intramuscular transplantations⁷³. We initially used RFP⁺-derived⁸⁹ primary killifish fibroblasts to facilitate visual confirmation of successful injections (Fig. S5a, top). Injected fish were monitored for 3 months. As expected for non-malignant cells, the injected cells were initially detectable in both rag2 mutants and WT controls but did not persist (n = 0/10 and n = 0/8, respectively, see example in Fig. S5a, bottom).

In contrast, cells isolated from a melanocyte expansion in the tail of vgll3^Ex1/Ex1 mutants were successfully engrafted. Small engraftments were first observed one-month post-injection and developed into an extensive melanoma-like tumor in rag2 mutants after two months (Fig. 5a). Thus, these “expansions” can be classified as tumors. While successful engraftments were consistently observed when expansion-derived melanocytes were injected into rag2 mutants (WT or vgll3^Ex1/Ex1, n = 3/3, see Methods), these cells failed to engraft when transplanted into WT fish (0/3). Similarly, phenotypically normal melanocytes (see examples in Supplementary Fig. 3e) did not engraft into rag2 mutants (0/3).

Fig. 5: Age-related effects of vgll3Ex1/Ex1 mutants. — **Fig. 5: Age-related effects of *vgll3*^Ex1/Ex1 mutants.**

Following successful engraftment, melanocytes were detected in multiple organs, invading skeletal muscle and blood vessels (Fig. 5b, c). To visualize cell proliferation, tumor cells were engrafted into a new rag2 recipient, and six weeks following engraftment, the fish were injected with EdU, which is incorporated into the DNA during cell division (see Methods). Many EdU-positive nuclei were detected within the melanocyte-positive area (Fig. 5c, and see quantification in Supplementary Fig. 5b).

Vgll3-mediated effect on age-specific mortality rate

A survival assay revealed reduced lifespan in vgll3^Ex1/Ex1 mutants, highly significant in males and with a more modest effect in females (log-rank: males, p = 0.0004, −15% in median lifespan; females, p = 0.044, −7% in median lifespan, Fig. 5d). Weibull regression analyses (see Methods) confirmed these findings (females: z = 2.41, p = 0.016; males: z = 4.02, p < 0.001), with no significant interaction between the effects of sex and genetic background (z = 0.52, p = 0.605). While female mutants had a 35% higher risk of death (hazard) than WTs, male mutants had a 55% higher risk of death (Supplementary Fig. 5c).

Both vgll3^Ex1/Ex1 females and males displayed a higher increase in age-related mortality relative to WTs, indicative of actuarial senescence (Figs. 5e and S5d, Wald test for Weibull shape parameter: females z = 2.36, p = 0.018; estimate (standard error) – WT 2.364 (0.095); mutant 2.654 (0.078); males z = 2.23, p = 0.026; estimate (standard error) – WT 3.442 (0.087); mutant 3.769 (0.118); and see Methods). In the Weibull formula (µ₀(x | b) = b₀b1^b0x^b0-1), beta parameters describe either baseline mortality (b₀) or mortality rate change with age (b₁, shape parameter). Finally, Bayesian survival trajectory analysis (BaSTA)⁹⁰ further supported the increase in age-related mortality, and found no difference in baseline mortality between vgll3^Ex1/Ex1 mutants relative to WTs for either sex (Supplementary Fig. 5e, f, and see Methods).

From an ecological perspective, an elegant study investigating the selective forces shaping life-history trait evolution in killifish examined the genomic basis of adaptation to seasonal habitats in 45 African killifish species⁹¹. Analyzing their data, we identified vgll3 as being positively selected in annual Nothobranchius species when compared to non-annual killifish species (Fig. 5f, in the top 5% of all significantly selected genes, p = 4.9×10^-6). It is worth noting that these signals of positive selection do not, on their own, demonstrate that alleles are directly selected to favor accelerated aging in wild populations, rather than persisting due to mutation–selection balance. Distinguishing antagonistic pleiotropy (AP) from mutation accumulation (MA) ultimately requires evidence that, in nature, the relevant allele confers a net fitness advantage when early-life benefits and late-life costs are considered together (see Discussion).

Using a codon-based multiple-sequence alignment of vgll3 orthologs across killifish species (see Supplementary Data 1 for trimmed protein sequences, extracted from Cui et al.⁹¹), we identified eleven amino-acid positions in VGLL3 that are specifically conserved among 7 short-lived Nothobranchius, relative to 22 long-lived killifish species (Supplementary Fig. 5g; see Methods). These sites occur predominantly in low-complexity regions (IUPred2A⁹², Supplementary Fig. 5g), including the N-terminal E-rich region (aa 42–70, Supplementary Fig. 5g). Similarly, in Atlantic salmon, two vgll3 missense variants, one in the N terminus (aa 54) and one in the C terminus (aa 323), are strongly associated with sea age at maturity³².

Human studies⁹³ demonstrate that intrinsically disordered regions (IDRs) frequently undergo adaptive amino-acid substitutions, often at higher rates than structured cores, which can drive functional changes such as altered phase behavior or cofactor interactions (e.g., in VGLL3⁹⁴). Along these same lines, pleiotropic architectures were recently predicted to have evolved more easily in transcriptional cofactors, such as vgll3⁵³, which regulate genes through protein-protein interactions.

Together, our findings raise the intriguing possibility that vgll3 contributed to the adaptation of killifish to an ephemeral habitat by promoting rapid growth and maturation, coupled with a compressed lifespan. These results identify vgll3 as an antagonistically pleiotropic gene that scales somatic growth, maturation, lifespan, and age-related disease (Fig. 5g), which may be evolutionarily conserved across vertebrates⁴⁹.

Discussion

The antagonistic pleiotropy theory of aging proposes that aging evolves through the selection of genes that enhance traits that increase fitness by their effect early in life, at the expense of late-life performance. While experimental evolution and GWAS studies provide broad support for this theory, we often lack functional studies that explicitly link variation in individual genes to complex age-specific life histories.

In this study, we mutated two evolutionarily conserved isoforms of vgll3 in the turquoise killifish. Our data suggest that mutating the long isoform of vgll3 by editing exon 1 results in a pleiotropic effect. On one hand, the mutants exhibited accelerated growth, stem cell proliferation, and maturity; on the other, they showed an increased incidence of age-related cancer and reduced lifespan. Reducing the expression of both the short and long isoforms of vgll3, through a heterozygous mutation in exon 3, produced a similar effect on maturation, indicating a probable dose-dependent effect of vgll3.

It is worth mentioning that isoform- and dosage effects are not mutually exclusive, as evidenced by salmon vgll3 studies⁵², which indicate that isoform variation, polymorphisms, and quantitative expression differences act in concert across tissues in determining the timing of puberty^32,52,95. Likewise, many other complex traits can be modulated by either isoform variation or gene mutations. For example, isoforms of FOXO/DAF-16 regulate longevity in C. elegans, while human FOXO3 variants are associated with exceptional lifespan^96,97.

Resolving the evolutionary mechanism

Our experimental data establish causal, age-opposed effects of vgll3, but do not provide definitive population-level evidence for how vgll3 variants are maintained in nature. Accordingly, vgll3 variants could reflect classical antagonistic pleiotropy (AP; early-life benefit that is favored by selection despite late-life costs), or late-acting deleterious effects that accumulate under mutation accumulation (MA) when selection is weak at older ages.

The Nothobranchius genome shows evidence of relaxed purifying selection, plausibly driven by recurrent bottlenecks and reduced effective population sizes in ephemeral habitats⁹¹. When considering the micro-evolutionary context of MA, genomic landscapes of recent divergence among annual populations^38,98, suggest that even within annual Nothobranchius, shorter-lived populations exhibit a significant accumulation of deleterious variants compared to longer-lived populations (consistent with MA/relaxed selection). Against this ‘noisy’ background, population-level allele frequency patterns can be dominated by recent drift and demography, and vgll3 was not found as an outlier in recent population divergence.

When considering the macro-evolutionary context by comparing short-lived annuals to long-lived non-annual species, signals of positive selection are identified, and vgll3 is among the upper tail (top 5%) of genes across the annual Nothobranchius clade (Fig. 5f). Importantly, such comparative tests support episodic selection on coding sequence but do not, on their own, establish that the selected changes specifically reflect aging-related trade-offs.

Physiologically, vgll3 mutants exhibit earlier sexual maturation and produce more mature sperm at the age that corresponds to peak fertility⁹⁹ (Figs. 3a–d, S3C). Moreover, male killifish are territorial, and traits such as size and coloration play important roles in dominance and mating success^100,101,102. These phenotypes can confer a net fitness gain in common killifish habitats, which favor early maturity and reproduction as they frequently restrict effective breeding time to just a few weeks (most habitats dry within 1–4 months^98,103). Shorter generation time itself can increase fitness, even when the number of offspring is the same (per generation)²⁹. Given the evolutionary conservation of vgll3’s function, this pleiotropic gene might be a broad regulator of maturity and aging across vertebrates.

Classic and modern syntheses present AP and MA as complementary forces along the same declining-selection gradient with age, emphasizing that many empirical patterns can reflect contributions from both processes⁴. Accordingly, genome-wide drift/relaxed selection coexists with signatures consistent with positive selection of vgll3 during annual-life-history evolution. Thus, phylogenetic and comparative signals alone cannot fully exclude the passive accumulation of deleterious variants (consistent with MA), even when individual loci show signatures compatible with selection.

The fundamental challenge is to determine whether natural selection has favored variants with age-accelerating late-life costs because they confer early-life benefits (AP). While our experimental findings are consistent with AP, they do not definitively exclude mutation accumulation (MA). Future work should therefore seek population-level evidence showing that putative aging-accelerating alleles are maintained by ongoing selection (rather than persisting at low frequency under mutation–selection balance), and directly quantify fitness trade-offs in ecologically relevant settings, including genotype-by-environment effects, so that net fitness effects can be evaluated across the life course. These efforts should be paired with functional characterization of naturally occurring vgll3 alleles (e.g., effects on protein function and stability, and splice-isoform usage).

The role of VGLL3 in cancer

The rag2 immunodeficient model facilitated the identification of melanoma-like cancer, which sporadically occurs during the aging of vgll3^Ex1/Ex1 mutants. Yet, how vgll3 can modulate cancer in killifish remains to be elucidated. Recent studies have implicated VGLL3 in tumor development^104,105, either by promoting tumor cell proliferation and motility^105,106,107 or, conversely, as a tumor suppressor through interactions between the Hippo pathway and estrogen receptor alpha (ERα)¹⁰⁴.

Vgll3 is a member of the VGLL family, whose members serve as cofactors for TEA domain–containing transcription factors (TEADs)¹⁰⁸. It also features a unique glutamate-rich motif at the N-terminus, believed to mediate liquid-liquid phase separation¹⁰⁹, and a low complexity histidine-rich motif at the C-terminus, associated with localization to nuclear speckles¹¹⁰.

Consistent with the altered DNA damage response we observed in vgll3 mutants (Fig. 2d), a proposed mechanism could be via the role of VGLL3 in DNA double-strand break repair⁹⁴. We also note that, although we focus on testing the antagonistic pleiotropy theory of aging and not aging per se (early-life benefit vs. late-life cost), integrating the survival data with the observed alteration in DNA damage response (a hallmark of aging) and the increased incidence of an age-related disease (cancer), suggests a possible link to aging.

However, the role of VGLL3 in tumor development within the context of a physiologically healthy individual requires further investigation. Future studies employing RNA-seq and/or whole-genome sequencing of melanocytes derived from vgll3 mutants may uncover gene expression changes and additional mutations that contribute to tumorigenicity.

The role of VGLL3 in maturity and age-related mortality rate

The antagonistically pleiotropic phenotypes of vgll3 may arise from its effects on growth, reproductive maturation, and age-related disease. In natural salmon populations, vgll3 accounts for approximately 40% of the variation in male maturation age^32,33. In humans, VGLL3 is associated with sexually dimorphic effects^111,112,113, and GWAS studies have linked vgll3 to various physiological traits, including puberty^{12,30,114,115,116}, testosterone and sex hormone levels^117,118, cancer^119,120, metabolic functions (triglyceride levels^121,122 and HDL cholesterol¹²³), waist-hip ratio¹¹⁶ and hip circumference¹²⁴, and body height¹²⁵. However, the precise mechanisms by which vgll3 functionally influences downstream gene regulation remain under intense investigation⁵³.

Our findings in turquoise killifish offer a functional complement to previous correlative studies in Atlantic salmon and humans, highlighting the conserved yet context-dependent role of VGLL3 in regulating maturation timing. While VGLL3 associations in salmon are largely sex- and population-specific, our manipulations in killifish suggest that isoform expression and dosage are key modulators of this trait. Notably, as heterozygotes for the Exon 3 mutation also accelerate puberty (i.e., Exon 3 het ≈ Exon 1 homozygote), a dosage effect is more plausible. Accordingly, any selection pressure on the locus that alters VGLL3 function (e.g. amino acid changes) or expression levels could yield a similar outcome.

These results suggest that a partial reduction in total VGLL3 function, which exhibits sexually dimorphic expression, can promote earlier sexual maturation. Mechanistically, our transcriptomic, cellular, and physiological data further implicate vgll3 in cell cycle regulation, DNA damage response, and the reproductive axis, among other related pathways. This insight bridges population-level associations in salmon with causal genetic evidence in killifish. It suggests that VGLL3’s regulation of reproductive timing is finely tuned by isoform composition and expression levels, potentially explaining the variability observed across species, sexes, and populations.

Downstream of vgll3, several studies have begun to elucidate possible molecular mechanisms, such as involving the Hippo and TGF-β signaling pathways, and Sertoli cell function^52,53. It was recently predicted that pleiotropic architectures might evolve more easily in transcriptional cofactors, such as vgll3⁵³, which regulate genes through protein-protein interactions rather than direct DNA binding (further supported by the position of the amino acids under positive selection, Supplementary Fig. 5g). The rationale behind this prediction is that cofactors can interact with multiple transcription factors, and simple changes, such as in expression levels, can orchestrate a strong effect on regulatory networks without compromising DNA binding.

Complex age-specific life histories are generally predicted to be highly polygenic. However, the AP theory postulates the existence of single genes with broad pleiotropic effects across the life cycle. The developmental theory of aging¹²⁶, a physiological extension of AP, proposes that selection acting on early-life traits, from zygote to the onset of reproduction, can drive the evolution of aging. This occurs because genes beneficial during development may later become detrimental in adulthood.

Because the late-life cost (i.e., the increased cancer incidence) is driven by continued vgll3 function, this supports the hyperfunction scenario of the developmental theory of aging¹⁰. However, we cannot exclude the possibility that reduced resource allocation to maintenance and repair contributes to the elevated cancer risk, which would be consistent with the ‘disposable soma’ theory (another physiological account of antagonistic pleiotropy).

Williams (1957)³ illustrated this concept with a hypothetical example of a gene involved in calcium metabolism in vertebrates: an allele that enhances bone calcification during development may later contribute to artery calcification, causing harm in old age. While it remains unclear whether calcium metabolism operates through such a mechanism, we identify vgll3 as an antagonistically pleiotropic gene. This gene, involved in cell growth, promotes beneficial developmental traits but imposes late-life costs by increasing tumor incidence and reducing lifespan, ultimately shaping vertebrate life history.

Methods

Experimental model and subject details

African turquoise killifish strain, husbandry, and maintenance

The African turquoise killifish (GRZ strain) were housed as previously described^44,45,56. Fish were grown at 28 °C in a central filtration recirculating system with a 12 h light/dark cycle at the Hebrew University of Jerusalem (Aquazone Ltd, Israel). Fish were fed once a day with live Artemia until the age of 2 weeks (#109448, Primo), and starting at week 3, fish were fed three times a day on weekdays (and once a day on weekends), with GEMMA Micro 500 Fish Diet (Skretting Zebrafish, USA), supplemented with Artemia twice a day. All turquoise killifish care and use was approved by the Subcommittee on Research Animal Care at the Hebrew University of Jerusalem (IACUC protocols #NS-18-15397-2, #NS-22-16915-3, and #HU-24-17607-4).

CRISPR/Cas9 target prediction and gRNA synthesis

CRISPR/Cas9 genome-editing protocols were performed according to published protocols^45,56. In brief, for gene targeting, a gRNA target site was identified using CHOPCHOP¹²⁷ as listed below, and if needed, the first two nucleotides were altered to have the gRNA start with a GG. Design of variable oligonucleotides and hybridization with a universal reverse oligonucleotide was performed according to⁵⁶, and the resulting products were used as a template for in vitro transcription. Each gRNA was in-vitro transcribed and purified using TranscriptAid T7 High Yield Transcription Kit (Thermo Scientific #K0441), according to the manufacturer’s protocol.

Gene

▓

name	gRNA sequence
rag2 (Exon 1, XM_015964119.1)	GGGAGTTACATCATTCACGGCGG
rag2 (Exon 1, XM_015964119.1)	GGGTCTCACAAGAAAGCAAGGGG
Vgll3 (Exon 1, XM_015965528.1)	[G/T]GCGCTCCATAGTCATTGTATGG
Vgll3 (Exon 3, XM_015965528.1)	G[G/T]GAGGTGCCTTGACCTTCAGGG

Production of Cas9 mRNA

Experiments were performed according to refs. ^{45,47,56,57,128,129}. The pCS2-nCas9n expression vector was used to produce Cas9 mRNA (Addgene, #47929)¹³⁰. Capped and polyadenylated mRNA was in-vitro transcribed and purified using the mMESSAGE mMACHINE SP6 ULTRA (ThermoFisher # AM1340).

Microinjection of turquoise killifish embryos

For the generation of the mutant fish using CRISPR/Cas9, microinjection of turquoise killifish embryos was performed according to ref. ⁵⁶. Briefly, nCas9n-encoding mRNA (300 ng/μL) and gRNA (30 ng/μL) were mixed with phenol-red (P0290, Sigma-Aldrich) and co-injected into two-cell stage fish embryos. We used several compound heterozygous mutations for Exon 3 indels, primarily vgll3^ins1 and vgll3^ins5, surrounding the PAM site. Sanger DNA sequencing was used for detecting successful germline transmission on F1 embryos using the following primer sequences:

rag2F: TGATGTTTTCTGTTTTGACCCA

rag2R: TATCTGTGGGCAGGACCTGTA

Vgll3_Ex1F: GGACACAGCCCGTCTGAAGTCT

Vgll3_Ex1R: GAAGTTCGCCTCACCTGTTG

Vgll3_Ex3F: CTGTTCACCTACTTTAAGGGCG

Vgll3_Ex3R: TGTAGGTCCAGGTATCAGGGAG.

Maturity, growth, and survival assays

Maturity assessment

Since fish needed independent evaluation before genotyping, each fish was individually housed in a 1-liter tank starting at week 2. For assessing sexual maturity in males, the onset of tail coloration was visually scored at 30 days post-hatch. Examples of juvenile and mature coloration are shown in Fig. 2a. Confirmation of sex for slower maturing fish was done by scoring them again at the age of 60 days.

Growth and GSI measurements

Fish were euthanized with 500 mg/l tricaine (MS222, #A5040, Sigma), water was blotted using Kimwipes, and then weighed. The testes were dissected, washed in PBS, briefly blotted using Kimwipes, and then weighed. GSI was calculated as follows: GSI = (gonad weight / total body weight) × 100. For standard length and depth measurement, fish were imaged with a Canon EOS 250D digital camera. A ruler was included in each image to provide an accurate scale. Standard length was measured from the tip of the snout to the posterior end of the last vertebra. Depth was measured from the base of the dorsal fin to the base of the anal fin. The actual size was then calculated using ImageJ (1.52a), by converting pixel number to millimeters using the reference ruler.

Melanoma incidence assessment

Male fish from both vgll3^Ex1/Ex1 and WT backgrounds were routinely monitored starting at six weeks of age. Fish in which the black stripe in the tail exceeded its normal boundaries and invaded the yellow stripe in the tail were considered to exhibit melanoma-like expansion. Examples can be found in Supplementary Fig. 3e.

Lifespan measurements

Constant housing parameters are important for reproducible lifespan experiments^56,131. After hatching, fish were raised with the following density control: up to 30 fish in a 1-liter tank for week 1, 5 fish in a 1-liter tank for week 2. Since these fish were also used for independent maturity assessment, each fish was individually housed in a 1-liter tank starting at week 2. At the age of 4 weeks, adult fish were genotyped and individually housed in a 1-liter tank for the rest of their life. Both male and female fish were used for lifespan experiments and were treated identically. Fish mortality was documented daily starting at week 4. Lifespan analyses were performed using GraphPad Prism for all survival curves with a Kaplan-Meier estimator. A log-rank test was used to examine any significant differences between the overall survival curves.

Age-specific mortality analysis

For robustness, parametric survival analyses were also run, fitting a Weibull regression model with the survreg function, and diagnostic graphical plots verified the adequacy of model fit, with generally linear and parallel lines, using the WeibullDiag function in the event history analysis package^132,133,134. Hazard ratios and shape parameters (rate of mortality rate increase with age) from the Weibull regression were determined using the weibreg function. We first fitted a model with sex, genotype and their interaction as covariates and then fitted separate Weibull regression models for each sex, with genotype as a covariate and the shape parameter allowed to vary by genotype. Formal Wald tests were used to determine significant shape parameter differences between genotypes, for each sex-specific model. To estimate the baseline mortality rate, we used Bayesian survival trajectory analysis in the BaSTA package⁹⁰ (ver. 1.9.5). We used an unbiased approach to compare ten models of age-specific mortality rates using Gompertz, Logistic, and Weibull mortality functions, with either a simple, Makeham¹³⁵ or bathtub shape¹³⁶, or an exponential model with a simple shape, representing no senescence. We ran four parallel simulations of each model, for 1,500,000 iterations, with a burn-in of 150,001 iterations and a thinning of 1,500, allowing for robust convergence of all models in the ‘multibasta’ function of the BaSTA package, except Weibull models with simple and bathtub shapes. Models were then compared using deviance information criteria (DIC values¹³⁷). The best-fitting model (with the lowest DIC) was identified as the Weibull model with simple shape parameters (see Supplementary Data 1). This simple Weibull model had low serial autocorrelation (<5%) and the posterior distributions converged for both baseline mortality and increase in mortality with age, when run for 1,000,000 iterations, with a burn-in of 100,001 iterations and thinning of 1000 (see Supplementary Data 1). The posterior distributions were compared between treatments using Kullback–Leibler divergence calibration (KLDC; Supplementary Fig. 5f). We considered a KLDC value > 0.85 to indicate a substantial difference between treatments, following established practice^138,139.

Generation of primary cell culture

Cells were isolated as described previously⁴⁵. Briefly, one-month-old fish were sedated with MS-222 (200 mg/L Tricaine), and a 2–3 mm piece of their tail fin was trimmed using a sterile razor blade. Following disinfection with 70% ethanol, tissue samples were incubated for 2 h with 1 mL of an antibiotic solution containing Gentamicin (50 µg/mL Gibco) and Primocin^TM (50 µg/mL, InvivoGen) in PBS at room temperature. Tissues were then transferred into an enzymatic digestion buffer (200 µL, in a 24-well plate) containing Dispase II (2 mg/mL, Sigma Aldrich) and Collagenase Type P (0.4 mg/mL, Merck Millipore) in Leibovitz’s L-15 Medium (Gibco), minced with a sterile pair of scissors, and incubated for 15 min. The digested tissue was then mixed with 400 µL of complete Leibovitz’s L-15 growth medium (Gibco), supplemented with Fetal Bovine Serum (15% FBS, Gibco), penicillin/streptomycin (50 U/ml, Gibco), Gentamicin (Gibco, 50 µg/ml), and Primocin^TM (50 µg/ml, InvivoGen). On the following day, the media was carefully replaced, and during the first 7 days, cells were washed daily with fresh media before adding new media. When cells reached 85–90% confluency, they were passaged with Trypsin-EDTA 0.05% (0.25% Trypsin-EDTA, diluted in PBS). Cells were incubated in a 28 °C humidified incubator (Binder, Thermo Scientific) with normal air composition. Primary cells were derived from ten individuals visually identified as males based on early secondary sexual characteristics at the onset of reproduction. However, given their young age and the absence of histological confirmation, we refer to this cohort as presumptive males.

Cell transplantation

For the calibration of fibroblast transplantation, RFP⁺ cells⁸⁹ were isolated from the tail, cultured as a primary cell line⁴⁵, and injected into rag2 homozygous fish (40,000 cells/µl, at 4 µl per fish). For the injection, adult male recipient fish (~1.5-month-old) were anesthetized, and intramuscular injection was performed using a Nanofil syringe (WPI, #NANOFIL).

For transplantation of melanoma-like cells, 8-month-old males (vgll3^Ex1/Ex1 or WTs), with a melanocyte expansion in the tail fin, were euthanized. Part of the fin was dissected, finely minced with scissors, and digested with 0.2% Collagenase Type P (Merck Millipore) and 0.12% Dispase II (Sigma Aldrich) in Leibovitz’s L-15 Medium (Gibco) for 30 min at RT. Cells were filtered using a Falcon 40 µm filter, centrifuged at 0.4 g for 10 min, resuspended in L15 with 10% FBS, and injected dorsally in rag2^Δ250/Δ250 and WT recipients. Following in vivo expansion of the melanoma cells in the rag2^Δ250/Δ250 mutant recipients, dorsal muscle tissue containing melanoma was dissected and prepared for serial transplantation into consecutive recipients as described above. Fish were monitored following engraftment for ethical purposes, and melanoma development was visually confirmed. Tumor burden was estimated by monitoring feeding, swimming patterns, and overall behavior, in accordance with the ethics protocol. Tumors were determined to have no detectable effect on well-being. As with any invasive procedure, the injection itself may be lethal. Therefore, the melanocyte mix from each donor was injected into at least three recipients. However, results were uniform across all surviving recipients, either they all developed a melanoma around the injection site when the donor was from a melanocyte expansion, or they did not, if the donor displayed a normal black stripe.

Hematoxylin and eosin

Tissue samples were processed as described previously^{39,40,43,44,45,56,99,128,140,141,142,143,144,145,146,147,148}. Briefly, fish were euthanized with 500 mg/l tricaine (MS222, #A5040, Sigma). The body cavity of the fish was opened and fixed for 72 h in 4% PFA solution at 4 °C with mild agitation. Samples were dehydrated and embedded in paraffin using standard procedures. Sections of 5-10 μm were stained with Hematoxylin and Eosin and examined by microscopy. A fully motorized Olympus IX83 microscope with an Olympus DP28 camera was used to collect images. Only images of the central part of the gonad were used since the gonads of fish in the genus Nothobranchius have a different composition in the distal part of the gonad¹⁴⁹. The total area and the area containing Spermatozoa in images of three-month-old WT and vgll3^Ex1/Ex1 mutant testes were calculated using ImageJ.

EdU staining

For melanoma proliferation, a transplanted rag2^Δ250/Δ250 mutant fish received an IP injection of EdU (50 mg/kg in DMSO). After 6 hours, the fish was sacrificed, its trunk was dissected, fixed for 48 h in 4% formaldehyde at 4 °C, dehydrated, embedded in paraffin, and sectioned. Consecutive 7 μm sections were collected individually on slides, deparaffinized, and rehydrated. One section was immediately mounted with Fluoroshield containing DAPI (Sigma-Aldrich # F6182). An adjacent section was treated overnight with 30% H₂O₂ to bleach the melanoma pigment (according to ref. ¹⁵⁰), and then stained with ClickTech EdU Cell Proliferation Kit (baseclick, BCK-EdU488IM100) and mounted with Fluoroshield containing DAPI.

For cell proliferation assay, cells were seeded on an 8-chamber slide (ibidi GmbH #80841) at comparable densities, allowed to recover for 24 hours, and treated with EdU for 8 hours. The cells were then fixed in 4% PFA (bioWorld #30450002) for 15 min, stained with the EdU-Click Kit, and mounted with Fluoroshield containing DAPI. Imaging was performed as described above and analyzed in ImageJ.

Single-molecule FISH

smFISH was performed as described previously⁴⁴. In brief, split-initiator hybridization probes for FISH chain reaction (HCR version 3.0) were designed and manufactured by Molecular Instruments for the following genes: amh (XM_015977279.1), ddx4 (XM_015957842.1), and vgll3 (XM_015965528.2). Then, 10-µm paraffin slices (as described above) were baked for 1 h at 60 °C, and smFISH was performed according to the manufacturer’s guidelines. In brief, after rehydration, slides were boiled for 15 min in 0.01 M citrate buffer (Sigma-Aldrich, C8532) and permeabilized with 20 µg/ml proteinase K (A&A Biotechnology, 1019-20-5) in PBS for 15 min at 37 °C, before being washed with PBS. Slides were pre-hybridized in the hybridization buffer provided by the manufacturer (Molecular Instruments) and then hybridized with the indicated probe (20 nM, in hybridization buffer) at 37 °C overnight. After washing, the signal was amplified with either custom-made green (488) or red (546) fluorophores at room temperature overnight. Slides were washed with 5×SCCT, and autofluorescence was quenched using a TrueVIEW Autofluorescence Quenching Kit (Vector Labs, SP8500) according to the manufacturer’s protocol. Slides were mounted with VECTASHIELD containing DAPI (Vector Labs, 30326), and images were collected with an Olympus FV-1200 confocal microscope and processed in ImageJ software.

Immunohistochemistry

For staining of primary cells, cells were seeded on an 8-chamber slide, allowed to recover for 24 hours, and treated with Etoposide (Sigma, E1383) at 50 uM for 1 hour. Samples were then washed with PBS and fixed in 4% PFA for 15 min, permeabilized for 15 min in 0.1% Triton (Avanator Performance Materials, X198-07) in PBS, followed by blocking (Dako, #X0909¹⁵¹) for 10 min, and incubation for 1 hour with primary antibody (rabbit anti-Histone H2A.XS139ph, GeneTex 127342, at 1:100). Slides were washed and incubated with secondary antibody (Goat anti-rabbit AlexaFluor 488, Invitrogen A11088, at 1:2000). After several washes, slides were mounted with Fluoroshield containing DAPI. For each culture, random regions were imaged at 400x magnification on a fully motorized IX-83 Olympus microscope with a Lumencor Spectra X fluorescent light system (Lumencor, Beaverton, OR, USA). The quantification of the γ-H2AX coverage of the nucleus was carried out using the FIJI package (NIH, ImageJ) “Foci Analyzer” (https://github.com/BioImaging-NKI/Foci-analyzer). Normalized intensity was calculated as mean foci intensity (total foci intensity/total foci area) normalized to nucleus area. At least 100 cells per genotype were analyzed. Statistics were calculated using one-way ANOVA with a Tukey post-hoc.

For tissue section staining, tissue sections were processed as described previously^141,145. Following rehydration, slides were washed in PBS and permeabilized for 15 min in 0.25% Triton (Avanator Performance Materials, X198-07) and 1% BSA (Sigma-Aldrich, A7906 in PBS), followed by blocking (Dako, X0909) for 10 min and incubation with primary antibodies overnight. The following primary antibodies were used: mouse anti-PCNA antibody (Abcam 29, 1:200); rabbit anti-Histone H2A.XS139ph (GeneTex 127342, 1:100) antibody. Secondary antibodies: Goat anti-rabbit AlexaFluor 488 (Invitrogen A11088); donkey anti-rabbit Alexa Fluor 549 secondary antibody (Abcam, 150064, 1:500). Secondary antibodies were incubated for 1 h at room temperature. After several washes, autofluorescence was quenched using the TrueVIEW Autofluorescence Quenching Kit (Vector Labs, SP8500) and mounted with Fluoroshield containing DAPI. Samples were imaged with the Olympus FV-1200 confocal microscope and processed in ImageJ.

Laser micro-irradiation

Cells were seeded in a 12-well plate and transfected with either the long or short isoform of vgll3 tagged with GFP (Addgene #241148, #241149), or a control MacroH2A-mKate plasmid⁶⁹ (a generous gift from the Toiber lab). Cells were subjected to laser micro-irradiation as previously described⁶⁸. Briefly, cells were plated on a flourodish (Ibidi; #81158) and pre-sensitized with 1 μg/μl Hoechst 3334 dye for 10 min at 37 °C. Laser microirradiation was executed using an LSM-700 inverted confocal microscope. DNA damage was induced by micro-irradiating a single region in the nucleus with 15 iterations of 405 nm laser beam. Time-lapse images were acquired, and signal intensity at damaged sites was measured using Zen 2009 software.

Profiling of V(D)J recombination

Amplification of the V(D)J locus was performed according to⁸⁸ with several modifications. RNA was collected from the head kidneys of 3 WT and 3 rag2 mutant males and purified using the Direct-zol RNA Miniprep kit (Zymo Research, #R5052). 45 ng of RNA was combined with 2 μl of 10 μM gene-specific primer (GSP, Supplementary Data 1), homologous to the second constant-region exon of n. furzeri ighm. The reaction volume was brought to a total of 8 μl with nuclease-free water, and the resulting mixture was incubated for 2 min at 70 °C to denature the RNA, then cooled to 42 °C to anneal the GSP¹⁵².

Following annealing, the RNA-primer mixture was combined with 12 μl of reverse transcription master-mix (SMARTScribe Reverse Transcriptase, Takara-Clontech #639538) including the reverse-transcriptase enzyme and template-switch adapter SmartNNNa primer (Supplementary Data 1). This primer consists of a constant sequence used for further PCR steps, a variable sequence containing multiple random nucleotides (N) that function as a unique molecular identifier (UMI), and several deoxyuridine bases (U), which are used to degrade the primer following amplification, and a 3’-terminal sequence of riboguanosine residues, which anneal to cytidine residues added by the reverse transcriptase and enable template switching (see Fig. 4c). The complete reaction mixture was incubated for 1 h at 42 °C for the reverse transcription reaction, then mixed with 2 μl of uracil DNA glycosylase (NEB, #M0280S, 1000 units) and incubated for a further 40 min at 37 °C to digest the template-switch adapter.

Following reverse transcription, the reaction product was purified using a standard PCR purification kit (QIAGEN #28104). The reaction mixture then underwent PCR amplification with the PCRBio HS Taq Mix Red (#PB10.23) using the IGH-B and M1SS primers (Supplementary Data 1), according to the manufacturer’s instructions, and resolved on a 1% agarose gel.

RNA-seq library preparation

For the rag2 ^Δ250/Δ250 transcriptomics, the head kidney was isolated as described previously⁵⁶. Samples were disrupted by bead beating in 300 μl of TRI Reagent (Sigma, T9424) and a single 3 mm metal bead (Eldan, BL6693003000) using TissueLyzer LT (QIAGEN, 85600) with a dedicated adaptor (QIAGEN, #69980). Beating was performed at 50 Hz for 5 min.

For the vgll3 transcriptomics, fibroblast cultures in a 10 cm plate at 90% confluency were harvested by washing them in ice-cold PBS, and then incubating them in 1 mL of TRI Reagent for 5 min. RNA purification was performed using Direct-zol RNA Miniprep Kit (Zymo Research # R2052) according to the manufacturer’s instructions. RNA concentration and quality were determined by using an Agilent 2100 bioanalyzer (Agilent Technologies).

Library preparation was performed using KAPA mRNA HyperPrep Kit (ROCHE-08105936001) according to the recommended protocols. Library quantity and pooling were measured by Qubit (dsDNA HS, Q32854), and size selection at 4% agarose gel. Library quality was measured by Tape Station (HS, 5067-5584). Libraries were sequenced by NextSeq 2000 P2, 100 cycle,75 bp single end (Illumina, 20024906) with ~40 million reads per sample for rag2 and ~30 million reads per sample for vgll3.

RNA sequencing analysis

Quality control and adapter trimming of the fastq sequence files were performed with FastQC (v0.11.8)¹⁵³, fastx-toolkits (v0.0.13), Trim Galore! (v0.6.4)¹⁵⁴, and Cutadapt (3.4)¹⁵⁵. Options were set to remove Illumina TruSeq adapters and end sequences, to retain high-quality bases with phred score > 20, and a remaining length > 20 bp. Successful processing was verified by re-running FastQC. Reads were mapped and quantified to the killifish genome Nfu_20140520^40,41 using STAR 2.7.6a¹⁵⁶.

For the rag2 analysis, differential gene expression between rag2 mutant and WT fish was performed using the edgeR package, (v3.32.1)^157,158 by the classic model (exactTest function). For the vgll3 analysis, the average of the two technical replicates of each sample was calculated for further analysis, except for a single technical replicate (Ex1 sample 3 A), which was removed as it displayed a large difference between the replicates.

Differential gene expression analysis between the genotypes (WT vs. Ex1, WT vs. Ex3, and Ex1 vs. Ex3) was performed using the classic model as described before. In addition, a linear model between Ex1, WT, and Ex3 was performed using glmQLFTest function.

Gene Ontology enrichment analysis

For rag2 analysis, enriched Gene Ontology (GO) terms associated with transcript level were identified using GO implemented in R package clusterProfiler (v3.18.1)¹⁵⁹ with the threshold of FDR < 0.05 and log2(FC) > 1. For vgll3 analysis, GO terms associated with transcript levels were identified using Gene Set Enrichment Analysis (GSEA) implemented in the R package clusterProfiler (v3.18.1)¹⁵⁹. All the transcripts were ranked and sorted in descending order based on the multiplication of log2 transformed fold change and -log10(FDR). Note that due to the random seeding effect in GSEA, the exact p-value and rank of the enriched terms may differ for each run. These random seeds did not qualitatively affect the enrichment analyses. GO terms were based on human GO annotations from org.Hs.eg.db (v3.13.0)¹⁶⁰ and AnnotationDbi (v1.54.1)¹⁶¹.

Isoform prediction

To confirm the expression of vgll3 isoforms, we re-analyzed the available RNA-seq data (gonads) from¹⁶². We aligned the sequences to the killifish genome Nfu_20140520^40,41 using STAR 2.7.11b¹⁵⁶ with the option --outSAMstrandField intronMotif, and using Cufflinks (v2.2.1), we identified predicted isoforms¹⁶³, and confirmed their presence in several samples.

Quantitative PCR

To measure expression of different vgll3 isoforms, we performed real-time qPCR using RNA extracted from the testis of WT male fish, from several ages around puberty (1 month ±1 week). RNA was extracted using TRI reagent (Sigma-Aldrich, T9424). Samples were disrupted by bead beating in 300 μl of TRI Reagent (Sigma, T9424) and a single 3 mm metal bead (Eldan, BL6693003000) using TissueLyzer LT (QIAGEN, #85600) with a dedicated adaptor (QIAGEN, #69980). Beating was performed at 50 Hz for 5 min. RNA was purified using a Direct-zol RNA Miniprep Kit (Zymo Research # R2052) according to the manufacturer’s instructions. cDNA was prepared using a Verso cDNA Synthesis Kit (Thermo Scientific # AB1453A). qPCR was performed using Fast SYBR Green Master Mix (ThermoFisher Scientific #4385612) on a QuantStudio5 Real-Time PCR System (ABI #A34322) using the following primers:

vgll3(Ex1)-F:	TCAAGGAGAGCTGACTTTCTCTC,
vgll3(Ex1)-R:	TAGCTCTGGTGGTACATCACTTC,
vgll3(Ex2)-F:	CAATGTCGCCCTTAAAGTAGGTG,
vgll3(Ex2)-R:	TAGCTCTGGTGGTACATCACTTC,
vgll3(5’UTR)-F:	AAACAAACCGGAGTTAGTCTGC,
vgll3(5’UTR)-R:	TCCTACACGACCTGTTAGGTTTC,
vgll3(Ex2-3)-F:	GCTAACTATTATATGGAAAGGTGGTGG,
vgll3(Ex2-3)-R:	GAGTTGGCTTGAGAGGGAAAG,
actin-F:	ATGTTTGAGACCTTCAACACACC,
actin-R:	TCCATCACGATACCTGTGGTTC.
tbp-F:	CGGTTGGAGGGTTTAGTCCT
tbp-R:	GCAAGACGATTCTGGGTTTG
insr-F:	TGCCTCTTCAAACCCTGAGT
insr-R:	AGGATGGCGATCTTATCACG.

Identification of positive selection in VGLL3

We analyzed a codon-based multiple sequence alignment of VGLL3 orthologs from 7 foreground fish species and 22 background species (strictly non-annuals, see Supplementary Data 1), identifying amino acid positions showing foreground-specific conservation. All species names and protein sequences from Cui et al.⁹¹ are provided in Supplementary Data 1. We searched for strict foreground-specific sites, where all foreground residues matched, differed from the background consensus, and were present in ≤10% of the background, identifying 11 strict sites (E5D, S40C, P47L, V108I, P136A, S140A, I154V, T194S, T285S, P294S, P310S). The classification into annual and strictly non-annual, and all protein sequences were according to Cui et al.⁹¹. Low complexity IDRs and the TONDU motif were predicted according to https://iupred2a.elte.hu/.

Statistics and reproducibility

The number of biological replicates and statistical tests for each experiment is presented in the corresponding figures and figure legends. All experiments were repeated independently with similar findings at least twice. All attempts at replication were successful. This excludes the RNA-seq datasets and fish lifespan that had independent biological replicates. All fish were randomly assigned to each experimental group while controlling for age and genotype. Visual scoring of melanoma presence and size, maturity, fish size, sperm content measurements, and scoring of images, including proliferation and DNA damage, were performed independently and blindly by two researchers. Sample sizes for all experiments are similar to those reported in previous publications for similar experiments performed in killifish^44,45.

Fish were excluded from the lifespan cohort if they did not develop normally (e.g., did not inflate the swim bladder or were deformed) or did not die a natural death (e.g., jumped out of a tank during routine maintenance). These criteria were pre-established. Otherwise, no data was excluded. Data distribution was assumed to be normal in bench experiments, as measured parameters, such as size and age at maturity, are considered normal and are treated as such in the field³⁸. In RNA-seq analysis, the data were normalized (see Methods).

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Data availability

All raw RNA sequencing data, as well as processed datasets, are found in the GEO database, accession numbers GSE226279 and GSE300596. Source data are provided with this paper.

Materials availability

All fish lines are available upon request, and plasmids are available from Addgene.

Code availability

References

Harel, I. The turquoise killifish. Nat. Methods 1150–1151 https://doi.org/10.1038/s41592-022-01631-y (2022).
Nielsen, J. et al. Eye lens radiocarbon reveals centuries of longevity in the Greenland shark (Somniosus microcephalus). Science 353, 702–704 (2016).
Article ADS CAS PubMed Google Scholar
Williams, G. C. Pleiotropy, natural selection, and the evolution of senescence. Evolution 11, 398–411 (1957).
Article Google Scholar
Austad, S. N. & Hoffman, J. M. Is antagonistic pleiotropy ubiquitous in aging biology? Evol. Med. Public Health 2018, 287–294 (2018).
Article PubMed PubMed Central Google Scholar
Maklakov, A. A. & Chapman, T. Evolution of ageing as a tangle of trade-offs: energy versus function. Proc. R. Soc. B 286, 20191604 (2019).
Article CAS PubMed PubMed Central Google Scholar
Chen, H. Y., Jolly, C., Bublys, K., Marcu, D. & Immler, S. Trade-off between somatic and germline repair in a vertebrate supports the expensive germ line hypothesis. Proc. Natl. Acad. Sci. USA. https://doi.org/10.1073/pnas.1918205117 (2020).
Hamilton, W. D. The moulding of senescence by natural selection. J. Theor. Biol. 12, 12–45 (1966).
Article ADS CAS PubMed Google Scholar
Rose, M. R. Life history evolution with antagonistic pleiotropy and overlapping generations. Theor. Popul. Biol. 28, 342–358 (1985).
Article MathSciNet Google Scholar
Gaillard, J.-M. & Lemaître, J.-F. The Williams’ legacy: a critical reappraisal of his nine predictions about the evolution of senescence. Evolution 71, 2768–2785 (2017).
Article PubMed Google Scholar
Lemaître, J.-F., Moorad, J., Gaillard, J.-M., Maklakov, A. A. & Nussey, D. H. A unified framework for evolutionary genetic and physiological theories of aging. PLoS Biol. 22, e3002513 (2024).
Article PubMed PubMed Central Google Scholar
Day, F. R., Elks, C. E., Murray, A., Ong, K. K. & Perry, J. R. Puberty timing associated with diabetes, cardiovascular disease and also diverse health outcomes in men and women: the UK Biobank study. Sci. Rep. https://doi.org/10.1038/srep11208 (2015).
Elks, C. E. et al. Thirty new loci for age at menarche identified by a meta-analysis of genome-wide association studies. Nat. Genet. 42, 1077–1085 (2010).
Article CAS PubMed PubMed Central Google Scholar
He, C. & Murabito, J. M. Genome-wide association studies of age at menarche and age at natural menopause. Mol. Cell. Endocrinol. 382, 767–779 (2014).
Article CAS PubMed Google Scholar
Hartge, P. Genetics of reproductive lifespan. Nat. Genet. 41, 637–638 (2009).
Article CAS PubMed Google Scholar
Visscher, P. M. & Yang, J. A plethora of pleiotropy across complex traits. Nat. Genet. 48, 707–708 (2016).
Article CAS PubMed Google Scholar
Dönertaş, H. M., Fabian, D. K., Fuentealba, M., Partridge, L. & Thornton, J. M. Common genetic associations between age-related diseases. Nat. Aging 1, 400–412 (2021).
Article PubMed PubMed Central Google Scholar
Rodríguez, J. A. et al. Antagonistic pleiotropy and mutation accumulation influence human senescence and disease. Nat. Ecol. Evol. 1, 55 (2017).
Article PubMed Google Scholar
Rose, M. R. Laboratory evolution of postponed senescence in Drosophila melanogaster. Evolution 38, 1004–1010 (1984).
Partridge, L., Prowse, N. & Pignatelli, P. Another set of responses and correlated responses to selection on age at reproduction in Drosophila melanogaster. Proc. R. Soc. Lond. Ser. B Biol. Sci. 266, 255–261 (1999).
Article CAS Google Scholar
Rose, M. R. Evolutionary Biology of Aging (Oxford University Press, 1994).
Wu, D. et al. An antagonistic pleiotropic gene regulates the reproduction and longevity tradeoff. Proc. Natl. Acad. Sci. USA. 119, e2120311119 (2022).
Article CAS PubMed PubMed Central Google Scholar
Tacutu, R. et al. Human ageing genomic resources: new and updated databases. Nucleic Acids Res. 46, D1083–D1090 (2018).
Article CAS PubMed PubMed Central Google Scholar
Flurkey, K., Papaconstantinou, J. & Harrison, D. E. The Snell dwarf mutation Pit1dw can increase life span in mice. Mech. Ageing Dev. https://doi.org/10.1016/S0047-6374(01)00339-6 (2002).
Flurkey, K., Papaconstantinou, J., Miller, R. A. & Harrison, D. E. Lifespan extension and delayed immune and collagen aging in mutant mice with defects in growth hormone production. Proc. Natl. Acad. Sci. USA. https://doi.org/10.1073/pnas.111158898 (2001).
Article PubMed PubMed Central Google Scholar
Bartke, A. & Brown-Borg, H. Life extension in the dwarf mouse. Curr. Top. Dev. Biol. (04)63006-7 https://doi.org/10.1016/S0070-2153(04)63006-7 (2004).
Shi, C. & Murphy, C. T. Sex and death. Curr. Top. Dev. Biol. 144, 353–375 (2021).
Stearns, S. C., Ackermann, M., Doebeli, M. & Kaiser, M. Experimental evolution of aging, growth, and reproduction in fruitflies. Proc. Natl. Acad. Sci. USA. 97, 3309–3313 (2000).
Article ADS CAS PubMed PubMed Central Google Scholar
Yuan, R. et al. Genetic coregulation of age of female sexual maturation and lifespan through circulating IGF1 among inbred mouse strains. Proc. Natl. Acad. Sci. USA. 109, 8224–8229 (2012).
Article ADS CAS PubMed PubMed Central Google Scholar
Long, E. & Zhang, J. Evidence for the role of selection for reproductively advantageous alleles in human aging. Sci. Adv. 9, eadh4990 (2023).
Article PubMed PubMed Central Google Scholar
Cousminer, D. L. et al. Genome-wide association and longitudinal analyses reveal genetic loci linking pubertal height growth, pubertal timing and childhood adiposity. Hum. Mol. Genet. 22, 2735–2747 (2013).
Article CAS PubMed PubMed Central Google Scholar
Tu, W. et al. Associations between menarche-related genetic variants and pubertal growth in male and female adolescents. J. Adolesc. Health 56, 66–72 (2015).
Article PubMed Google Scholar
Ayllon, F. et al. The vgll3 locus controls age at maturity in wild and domesticated Atlantic salmon (Salmo salar L.) males. PLoS Genet. 11, e1005628 (2015).
Article ADS PubMed PubMed Central Google Scholar
Barson, N. J. et al. Sex-dependent dominance at a single locus maintains variation in age at maturity in salmon. Nature 528, 405–408 (2015).
Article ADS CAS PubMed Google Scholar
Schartl, M. Beyond the zebrafish: diverse fish species for modeling human disease. Dis. Model Mech. 7, 181–192 (2014).
ADS PubMed Google Scholar
Podrabsky, J. E., Riggs, C. L. & Wagner, J. T. Tolerance of environmental stress. in Annual Fishes. Life History Strategy, Diversity, and Evolution (eds Berois, N., García, G. & De Sá, R. O.) 159–184 (CRC Press, Taylor & Francis, 2016).
Jaggard, J. B. et al. Cavefish brain atlases reveal functional and anatomical convergence across independently evolved populations. Sci. Adv. 6, eaba3126 (2020).
Article ADS CAS PubMed PubMed Central Google Scholar
Wang, W. et al. Changes in regeneration-responsive enhancers shape regenerative capacities in vertebrates. Science 369, eaaz3090 (2020).
Vrtílek, M., Žák, J., Pšenička, M. & Reichard, M. Extremely rapid maturation of a wild African annual fish. Curr. Biol. 28, R822–R824 (2018).
Article PubMed Google Scholar
Harel, I. et al. A platform for rapid exploration of aging and diseases in a naturally short-lived vertebrate. Cell 1013–1026 https://doi.org/10.1016/j.cell.2015.01.038 (2015).
Valenzano, D. R. et al. The African turquoise killifish genome provides insights into evolution and genetic architecture of lifespan. Cell 163, 1539–1554 (2015).
Article ADS CAS PubMed PubMed Central Google Scholar
Reichwald, K. et al. Insights into sex chromosome evolution and aging from the genome of a short-lived fish. Cell 163, 1527–1538 (2015).
Article ADS CAS PubMed Google Scholar
Chen, Y. R. et al. Tissue-specific landscape of protein aggregation and quality control in an aging vertebrate. Dev. Cell 1892–1911.e13 https://doi.org/10.1016/j.devcel.2024.04.014 (2024).
Harel, I. et al. Identification of protein aggregates in the aging vertebrate brain with prion-like and phase-separation properties. Cell Rep. 112787 https://doi.org/10.1016/j.celrep.2023.112787 (2024).
Moses, E. et al. The killifish germline regulates longevity and somatic repair in a sex-specific manner. Nat. Aging 791–813 https://doi.org/10.1038/s43587-024-00632-0 (2024).
Astre, G. et al. Genetic perturbation of AMP biosynthesis extends lifespan and restores metabolic health in a naturally short-lived vertebrate. Dev. Cell 58, 1350–1364.e10 (2023).
Ripa, R. et al. Refeeding-associated AMPKγ1 complex activity is a hallmark of health and longevity. Nat. Aging 3, 1544–1560 (2023).
Article CAS PubMed PubMed Central Google Scholar
Bergman, M. et al. The cGAS-STING pathway is an in vivo modifier of genomic instability syndromes. bioRxiv https://doi.org/10.1101/2024.10.16.618655 (2024).
Di Fraia, D. et al. Altered translation elongation contributes to key hallmarks of aging in the killifish brain. Science 389, eadk3079 (2025).
Article PubMed Google Scholar
Harel, I. Scaling life as an interspecies hallmark of aging. Genes Dev. 39, 930–932 (2025).
Uhlén, M. et al. Tissue-based map of the human proteome. Science 347, 1260419 (2015).
Article PubMed Google Scholar
Raj, A., Van Den Bogaard, P., Rifkin, S. A., Van Oudenaarden, A. & Tyagi, S. Imaging individual mRNA molecules using multiple singly labeled probes. Nat. Methods 5, 877–879 (2008).
Article CAS PubMed PubMed Central Google Scholar
Verta, J.-P. et al. Cis-regulatory differences in isoform expression associate with life history strategy variation in Atlantic salmon. PLoS Genet. 16, e1009055 (2020).
Article CAS PubMed PubMed Central Google Scholar
Verta, J.-P. et al. A complex mechanism translating variation of a simple genetic architecture into alternative life histories. Proc. Natl. Acad. Sci. USA. 121, e2402386121 (2024).
Article CAS PubMed PubMed Central Google Scholar
Xu, A. et al. Transcriptomes of aging brain, heart, muscle, and spleen from female and male African turquoise killifish. Sci. Data 10, 695 (2023).
Article CAS PubMed PubMed Central Google Scholar
Trapnell, C. et al. Differential gene and transcript expression analysis of RNA-seq experiments with TopHat and Cufflinks. Nat. Protoc. 7, 562–578 (2012).
Article CAS PubMed PubMed Central Google Scholar
Astre, G., Moses, E. & Harel, I. The African turquoise killifish (Nothobranchius furzeri): biology and research applications. in Laboratory Fish in Biomedical Research (eds. D'Angelo, L. & de Girolamo, P.) 245–287 (Elsevier, 2022).
Rozenberg, I., Moses, E. & Harel, I. CRISPR-Cas9 genome editing in Nothobranchius furzeri for gene knockout and knock-in. Cold Spring Harb. Protoc. 2023, 90–99 (2023).
Blazek, R., Polacik, M. & Reichard, M. Rapid growth, early maturation and short generation time in African annual fishes. Evodevo 4, 24 (2013).
Article PubMed PubMed Central Google Scholar
Lindeboom, R. G. H., Supek, F. & Lehner, B. The rules and impact of nonsense-mediated mRNA decay in human cancers. Nat. Genet. 48, 1112–1118 (2016).
Article CAS PubMed PubMed Central Google Scholar
Lindeboom, R. G. H., Vermeulen, M., Lehner, B. & Supek, F. The impact of nonsense-mediated mRNA decay on genetic disease, gene editing and cancer immunotherapy. Nat. Genet. 51, 1645–1651 (2019).
Article CAS PubMed PubMed Central Google Scholar
Zhou, Y., Zhang, X., Xiong, S., Zeng, X. & Zhang, X. Predicted gene 31453 (Gm31453) and the gene encoding carboxypeptidase A5 (Cpa5) are not essential for spermatogenesis and male fertility in the mouse. Reprod. Fertil. Dev. 33, 401–409 (2021).
Article CAS PubMed Google Scholar
Hughes, I. A. et al. Androgen insensitivity syndrome. Lancet 380, 1419–1428 (2012).
Article CAS PubMed Google Scholar
Di Giorgio, E. et al. HDAC4 influences the DNA damage response and counteracts senescence by assembling with HDAC1/HDAC2 to control H2BK120 acetylation and homology-directed repair. Nucleic Acids Res. 52, 8218–8240 (2024).
Article CAS PubMed PubMed Central Google Scholar
Kjærner-Semb, E. et al. Vgll3 and the Hippo pathway are regulated in Sertoli cells upon entry and during puberty in Atlantic salmon testis. Sci. Rep. 8, 1–11 (2018).
Article Google Scholar
Ahi, E. P. et al. Hippo-vgll3 signaling may contribute to sex differences in Atlantic salmon maturation age via contrasting adipose dynamics. bioRxiv https://doi.org/10.1101/2024.10.21.619349 (2024).
Ibar, C. & Irvine, K. D. Integration of Hippo-YAP signaling with metabolism. Dev. Cell 54, 256–267 (2020).
Article CAS PubMed PubMed Central Google Scholar
Kuser-Abali, G., Alptekin, A., Lewis, M., Garraway, I. P. & Cinar, B. YAP1 and AR interactions contribute to the switch from androgen-dependent to castration-resistant growth in prostate cancer. Nat. Commun. 6, 8126 (2015).
Article ADS PubMed PubMed Central Google Scholar
Abu-Zhayia, E. R. et al. CDYL1-dependent decrease in lysine crotonylation at DNA double-strand break sites functionally uncouples transcriptional silencing and repair. Mol. Cell 82, 1940–1955 (2022).
Article CAS PubMed Google Scholar
Onn, L. et al. SIRT6 is a DNA double-strand break sensor. Elife 9, e51636 (2020).
Article CAS PubMed PubMed Central Google Scholar
Reichard, M. & Polačik, M. Nothobranchius furzeri, an ’instant’ fish from an ephemeral habitat. Elife 8, e41548 (2019).
Article PubMed PubMed Central Google Scholar
Sato, Y. et al. Tumorigenicity assessment of cell therapy products: the need for global consensus and points to consider. Cytotherapy 21, 1095–1111 (2019).
Article CAS PubMed Google Scholar
Amatruda, J. F. & Patton, E. E. Genetic models of cancer in zebrafish. Int. Rev. Cell Mol. Biol. 271, 1–34 (2008).
CAS PubMed Google Scholar
Tang, Q. et al. Optimized cell transplantation using adult rag2 mutant zebrafish. Nat. Methods 11, 821–824 (2014).
Article CAS PubMed PubMed Central Google Scholar
Couëdel, C. et al. Analysis of mutations from SCID and Omenn syndrome patients reveals the central role of the Rag2 PHD domain in regulating V (D) J recombination. J. Clin. Invest. 120, 1337–1344 (2010).
Article PubMed PubMed Central Google Scholar
Villa, A. et al. V (D) J recombination defects in lymphocytes due to RAG mutations: severe immunodeficiency with a spectrum of clinical presentations. Blood J. Am. Soc. Hematol. 97, 81–88 (2001).
CAS Google Scholar
Hao, Z. & Rajewsky, K. Homeostasis of peripheral B cells in the absence of B cell influx from the bone marrow. J. Exp. Med. 194, 1151–1164 (2001).
Article CAS PubMed PubMed Central Google Scholar
Press, C. M. & Evensen, Ø The morphology of the immune system in teleost fishes. Fish. Shellfish Immunol. 9, 309–318 (1999).
Article Google Scholar
Cannon, J. P. et al. A bony fish immunological receptor of the NITR multigene family mediates allogeneic recognition. Immunity 29, 228–237 (2008).
Article CAS PubMed PubMed Central Google Scholar
Yoder, J. A. Form, function and phylogenetics of NITRs in bony fish. Dev. Comp. Immunol. 33, 135–144 (2009).
Article CAS PubMed Google Scholar
Renshaw, S. A. & Trede, N. S. A model 450 million years in the making: zebrafish and vertebrate immunity. Dis. Model Mech. 5, 38–47 (2012).
Article CAS PubMed PubMed Central Google Scholar
Moore, F. E. et al. Single-cell transcriptional analysis of normal, aberrant, and malignant hematopoiesis in zebrafish. J. Exp. Med. 213, 979–992 (2016).
Article PubMed PubMed Central Google Scholar
Carmona, S. J. et al. Single-cell transcriptome analysis of fish immune cells provides insight into the evolution of vertebrate immune cell types. Genome Res. 27, 451–461 (2017).
Article CAS PubMed PubMed Central Google Scholar
Pereiro, P. et al. Zebrafish Nk-lysins: first insights about their cellular and functional diversification. Dev. Comp. Immunol. 51, 148–159 (2015).
Article CAS PubMed Google Scholar
Dornburg, A. & Yoder, J. A. On the relationship between extant innate immune receptors and the evolutionary origins of jawed vertebrate adaptive immunity. Immunogenetics 1, 18 (2022).
Google Scholar
Yoder, J. A. et al. Resolution of the novel immune-type receptor gene cluster in zebrafish. Proc. Natl. Acad. Sci. USA. 101, 15706–15711 (2004).
Article ADS CAS PubMed PubMed Central Google Scholar
Bassing, C. H., Swat, W. & Alt, F. W. The mechanism and regulation of chromosomal V (D) J recombination. Cell 109, S45–S55 (2002).
Article CAS PubMed Google Scholar
Bradshaw, W. J. & Valenzano, D. R. Extreme genomic volatility characterizes the evolution of the immunoglobulin heavy chain locus in cyprinodontiform fishes. Proc. Biol. Sci. 287, 20200489 (2020).
CAS PubMed PubMed Central Google Scholar
Bradshaw, W. J., Poeschla, M., Placzek, A., Kean, S. & Valenzano, D. R. Extensive age-dependent loss of antibody diversity in naturally short-lived turquoise killifish. Elife 11, e65117 (2022).
Rozenberg, I. et al. Exploring life-long tissue homeostasis through lineage tracing and cell transplantation. bioRxiv https://doi.org/10.1101/2023.05.01.538839 (2023).
Colchero, F., Jones, O. R. & Rebke, M. BaSTA: an R package for Bayesian estimation of age-specific survival from incomplete mark–recapture/recovery data with covariates. Methods Ecol. Evol. 3, 466–470 (2012).
Article Google Scholar
Cui, R. et al. Relaxed selection limits lifespan by increasing mutation load. Cell 178, 385–399.e20 (2019).
Article CAS PubMed Google Scholar
Mészáros, B., Erdős, G. & Dosztányi, Z. IUPred2A: context-dependent prediction of protein disorder as a function of redox state and protein binding. Nucleic Acids Res. 46, W329–W337 (2018).
Article PubMed PubMed Central Google Scholar
Afanasyeva, A., Bockwoldt, M., Cooney, C. R., Heiland, I. & Gossmann, T. I. Human long intrinsically disordered protein regions are frequent targets of positive selection. Genome Res. 28, 975–982 (2018).
Article CAS PubMed PubMed Central Google Scholar
Wu, W. et al. VGLL3 modulates chemosensitivity through promoting DNA double-strand break repair. Sci. Adv. 10, eadr2643 (2024).
Article CAS PubMed PubMed Central Google Scholar
Mobley, K. B. et al. Maturation in Atlantic salmon (Salmo salar, Salmonidae): a synthesis of ecological, genetic, and molecular processes. Rev. Fish. Biol. Fish. 31, 523–571 (2021).
Article Google Scholar
Kwon, E.-S., Narasimhan, S. D., Yen, K. & Tissenbaum, H. A. A new DAF-16 isoform regulates longevity. Nature 466, 498–502 (2010).
Article ADS CAS PubMed PubMed Central Google Scholar
Willcox, B. J. et al. FOXO3A genotype is strongly associated with human longevity. Proc. Natl. Acad. Sci. USA. 105, 13987–13992 (2008).
Article ADS CAS PubMed PubMed Central Google Scholar
Vrtilek, M. et al. Longitudinal demographic study of wild populations of African annual killifish. Sci. Rep. 8, 4774 (2018).
Article ADS PubMed PubMed Central Google Scholar
Moses, E., Franek, R. & Harel, I. A scalable and tunable platform for functional interrogation of peptide hormones in fish. Elife 12, e85960 (2023).
Article CAS PubMed PubMed Central Google Scholar
Searle, C. Social Behaviour of the African Turquoise Killifish (Nothobranchius Furzeri) (University of Exeter, 2019).
Fassiau, G., Silvestre, F. & Mouchet, S. Coloration impacts sexual behavior through dominance without any color preferences in the short-living fish Nothobranchius furzeri. (2023).
Krug, J. et al. Generation of a transparent killifish line through multiplex CRISPR/Cas9-mediated gene inactivation. Elife 12, e81549 (2023).
Vrtilek, M. et al. Extremely rapid maturation of a wild African annual fish. Curr. Biol. 28, R822–R824 (2018).
Article CAS PubMed Google Scholar
Ma, S. et al. Transcriptional repression of estrogen receptor alpha by YAP reveals the Hippo pathway as therapeutic target for ER+ breast cancer. Nat. Commun. 13, 1061 (2022).
Article ADS CAS PubMed PubMed Central Google Scholar
Hori, N. et al. Vestigial-like family member 3 stimulates cell motility by inducing high-mobility group AT-hook 2 expression in cancer cells. J. Cell. Mol. Med. 26, 2686–2697 (2022).
Article CAS PubMed PubMed Central Google Scholar
Hori, N. et al. Vestigial-like family member 3 (VGLL3), a cofactor for TEAD transcription factors, promotes cancer cell proliferation by activating the Hippo pathway. J. Biol. Chem. 295, 8798–8807 (2020).
Article CAS PubMed PubMed Central Google Scholar
Haque, R. et al. VGLL3 expression is associated with macrophage infiltration and predicts poor prognosis in epithelial ovarian cancer. Front. Oncol. 13, 1152991 (2023).
Article CAS PubMed PubMed Central Google Scholar
Yamaguchi, N. Multiple roles of vestigial-like family members in tumor development. Front. Oncol. 10, 1266 (2020).
Article PubMed PubMed Central Google Scholar
Horii, Y. et al. VGLL3 is a mechanosensitive protein that promotes cardiac fibrosis through liquid–liquid phase separation. Nat. Commun. 14, 550 (2023).
Article ADS CAS PubMed PubMed Central Google Scholar
Salichs, E. et al. Genome-wide analysis of histidine repeats reveals their role in the localization of human proteins to the nuclear speckles compartment. PLoS Genet. 5, e1000397 (2009).
Article PubMed PubMed Central Google Scholar
Liang, Y. et al. A gene network regulated by the transcription factor VGLL3 as a promoter of sex-biased autoimmune diseases. Nat. Immunol. 18, 152–160 (2017).
Article CAS PubMed Google Scholar
Billi, A. C. et al. The female-biased factor VGLL3 drives cutaneous and systemic autoimmunity. JCI Insight 4, e127291 (2019).
Takakura, Y. et al. VGLL3 activates inflammatory responses by inducing interleukin-1α secretion. FASEB J. 35, e21996 (2021).
Article CAS PubMed Google Scholar
Pickrell, J. K. et al. Detection and interpretation of shared genetic influences on 42 human traits. Nat. Genet. 48, 709–717 (2016).
Article CAS PubMed PubMed Central Google Scholar
Perry, J. R. B. et al. Parent-of-origin-specific allelic associations among 106 genomic loci for age at menarche. Nature 514, 92–97 (2014).
Article CAS PubMed PubMed Central Google Scholar
Kichaev, G. et al. Leveraging polygenic functional enrichment to improve GWAS power. Am. J. Hum. Genet. 104, 65–75 (2019).
Article CAS PubMed Google Scholar
Ruth, K. S. et al. Using human genetics to understand the disease impacts of testosterone in men and women. Nat. Med. 26, 252–258 (2020).
Article CAS PubMed PubMed Central Google Scholar
Haas, C. B., Hsu, L., Lampe, J. W., Wernli, K. J. & Lindström, S. Cross-ancestry genome-wide association studies of sex hormone concentrations in pre-and postmenopausal women. Endocrinology 163, bqac020 (2022).
Article PubMed PubMed Central Google Scholar
Michailidou, K. et al. Association analysis identifies 65 new breast cancer risk loci. Nature 551, 92–94 (2017).
Article ADS CAS PubMed PubMed Central Google Scholar
Wang, A. et al. Characterizing prostate cancer risk through multi-ancestry genome-wide discovery of 187 novel risk variants. Nat. Genet. 55, 2065–2074 (2023).
Article CAS PubMed PubMed Central Google Scholar
Sinnott-Armstrong, N. et al. Genetics of 35 blood and urine biomarkers in the UK Biobank. Nat. Genet. 53, 185–194 (2021).
Article CAS PubMed PubMed Central Google Scholar
Richardson, T. G. et al. Evaluating the relationship between circulating lipoprotein lipids and apolipoproteins with risk of coronary heart disease: a multivariable Mendelian randomisation analysis. PLoS Med. 17, e1003062 (2020).
Article PubMed PubMed Central Google Scholar
Graham, S. E. et al. The power of genetic diversity in genome-wide association studies of lipids. Nature 600, 675–679 (2021).
Article ADS CAS PubMed PubMed Central Google Scholar
Christakoudi, S., Evangelou, E., Riboli, E. & Tsilidis, K. K. GWAS of allometric body-shape indices in UK Biobank identifies loci suggesting associations with morphogenesis, organogenesis, adrenal cell renewal and cancer. Sci. Rep. 11, 10688 (2021).
Article ADS CAS PubMed PubMed Central Google Scholar
Yengo, L. et al. A saturated map of common genetic variants associated with human height. Nature 610, 704–712 (2022).
Article ADS CAS PubMed PubMed Central Google Scholar
Zwaan, B. J., Bijlsma, R. & Hoekstra, R. F. On the developmental theory of ageing. I. Starvation resistance and longevity in Drosophila melanogaster in relation to pre-adult breeding conditions. Heredity 66, 29–39 (1991).
Article PubMed Google Scholar
Labun, K. et al. CHOPCHOP v3: expanding the CRISPR web toolbox beyond genome editing. Nucleic Acids Res. 47, W171–W174 (2019).
Article CAS PubMed PubMed Central Google Scholar
Harel, I., Valenzano, D. R. & Brunet, A. Efficient genome engineering approaches for the short-lived African turquoise killifish. Nat. Protoc. 2010–2028 https://doi.org/10.1038/nprot.2016.103 (2016).
Rozenberg, I. et al. A genetic toolbox for the turquoise killifish identifies sporadic age-related cancer. bioRxiv https://doi.org/10.1101/2023.05.01.538839 (2024).
Jao, L. E., Wente, S. R. & Chen, W. Efficient multiplex biallelic zebrafish genome editing using a CRISPR nuclease system. Proc. Natl. Acad. Sci. USA. 110, 13904–13909 (2013).
Article ADS CAS PubMed PubMed Central Google Scholar
Dodzian, J., Kean, S., Seidel, J. & Valenzano, D. R. A protocol for laboratory housing of turquoise killifish (Nothobranchius furzeri). J. Vis. Exp. https://doi.org/10.3791/57073 (2018).
Article PubMed PubMed Central Google Scholar
Carroll, K. J. On the use and utility of the Weibull model in the analysis of survival data. Control. Clin. Trials 24, 682–701 (2003).
Article PubMed Google Scholar
Klein, J. P. & Moeschberger, M. L. Multivariate survival analysis. in Survival Analysis: Techniques for Censored and Truncated Data 425–441 (Springer, 2006).
Zhang, Z. Parametric regression model for survival data: Weibull regression model as an example. Ann. Transl. Med. 4, 484 (2016).
Article PubMed PubMed Central Google Scholar
Pletcher, S. D. Model fitting and hypothesis testing for age-specific mortality data. J. Evol. Biol. 12, 430–439 (1999).
Article Google Scholar
Siler, W. A competing-risk model for animal mortality. Ecology 60, 750–757 (1979).
Article Google Scholar
Millar, R. B. Comparison of hierarchical Bayesian models for overdispersed count data using DIC and Bayes’ factors. Biometrics 65, 962–969 (2009).
Article MathSciNet PubMed Google Scholar
Rodríguez-Muñoz, R., Boonekamp, J. J., Fisher, D., Hopwood, P. & Tregenza, T. Slower senescence in a wild insect population in years with a more female-biased sex ratio. Proc. R. Soc. B 286, 20190286 (2019).
Article PubMed PubMed Central Google Scholar
Spagopoulou, F. et al. Silver-spoon upbringing improves early-life fitness but promotes reproductive ageing in a wild bird. Ecol. Lett. 23, 994–1002 (2020).
Article PubMed Google Scholar
Van Keymeulen, A. et al. Epidermal progenitors give rise to Merkel cells during embryonic development and adult homeostasis. J. Cell Biol. 187, 91–100 (2009).
Article PubMed PubMed Central Google Scholar
Harel, I. et al. Distinct origins and genetic programs of head muscle satellite cells. Dev. Cell 16, 822–832 (2009).
Article CAS PubMed PubMed Central Google Scholar
Gruenbaum-Cohen, Y. et al. The actin regulator N-WASp is required for muscle-cell fusion in mice. Proc. Natl. Acad. Sci. USA. 109, 11211–11216 (2012).
Article ADS CAS PubMed PubMed Central Google Scholar
Nathan, E. et al. The contribution of Islet1-expressing splanchnic mesoderm cells to distinct branchiomeric muscles reveals significant heterogeneity in head muscle development. Development 135, 647–657 (2008).
Article CAS PubMed PubMed Central Google Scholar
Theis, S. et al. The occipital lateral plate mesoderm is a novel source for vertebrate neck musculature. Development 137, 2961–2971 (2010).
Article CAS PubMed Google Scholar
Harel, I. & Tzahor, E. Head muscle development. in Craniofacial Muscles (eds. McLoon, L. K. & Andrade, F.) 11–28 (Springer, 2012).
Harel, I. et al. Pharyngeal mesoderm regulatory network controls cardiac and head muscle morphogenesis. Proc. Natl. Acad. Sci. USA. 109, 18839–18844 (2012).
Article ADS CAS PubMed PubMed Central Google Scholar
Benayoun, B. A. et al. Remodeling of epigenome and transcriptome landscapes with aging in mice reveals widespread induction of inflammatory responses. Genome Res. 29, 697–709 (2019).
Article CAS PubMed PubMed Central Google Scholar
Harel, I. & Brunet, A. The African turquoise killifish: a model for exploring vertebrate aging and diseases in the fast lane. Cold Spring Harb. Symp. Quant. Biol. https://doi.org/10.1101/sqb.2015.80.027524 (2015).
Uribe, M. C., Grier, H. J. & Mejía-Roa, V. Comparative testicular structure and spermatogenesis in bony fishes. Spermatogenesis 4, e983400 (2014).
Article PubMed Google Scholar
Hu, L. et al. Retaining antigenicity and DNA in the melanin bleaching of melanin-containing tissues. Int. J. Clin. Exp. Pathol. 13, 2027–2034 (2020).
Lenti, E. et al. Fate mapping and scRNA sequencing reveal origin and diversity of lymph node stromal precursors. Immunity 55, 606–622 (2022).
Article CAS PubMed Google Scholar
Turchaninova, M. A. et al. High-quality full-length immunoglobulin profiling with unique molecular barcoding. Nat. Protoc. 11, 1599–1616 (2016).
Andrews, S. FastQC: a quality control tool for high throughput sequence data. https://www.bioinformatics.babraham.ac.uk/projects/fastqc/
Krueger, F. Trim Galore. https://www.bioinformatics.babraham.ac.uk/projects/trim_galore/
Martin, M. Cutadapt removes adapter sequences from high-throughput sequencing reads. EMBnet J. 17, 10–12 (2011).
Article Google Scholar
Dobin, A. et al. STAR: ultrafast universal RNA-seq aligner. Bioinformatics 29, 15–21 (2013).
Article CAS PubMed Google Scholar
McCarthy, D. J., Chen, Y. & Smyth, G. K. Differential expression analysis of multifactor RNA-Seq experiments with respect to biological variation. Nucleic Acids Res. 40, 4288–4297 (2012).
Article CAS PubMed PubMed Central Google Scholar
Robinson, M. D., McCarthy, D. J. & Smyth, G. K. edgeR: a Bioconductor package for differential expression analysis of digital gene expression data. Bioinformatics 26, 139–140 (2010).
Article CAS PubMed Google Scholar
Yu, G., Wang, L.-G., Han, Y. & He, Q.-Y. clusterProfiler: an R package for comparing biological themes among gene clusters. OMICS 16, 284–287 (2012).
Article CAS PubMed PubMed Central Google Scholar
Carlson, M., Falcon, S., Pages, H. & Li, N. org. Hs. eg. db: Genome wide annotation for Human. R Package version 3 (2019).
Pagès, H., Carlson, M., Falcon, S. & Li, N. AnnotationDbi: manipulation of SQLite-based annotations in Bioconductor. R package version 1 (2021).
Teefy, B. B. et al. Dynamic regulation of gonadal transposon control across the lifespan of the naturally short-lived African turquoise killifish. Genome Res. 33, 141–153 (2023).
Article CAS PubMed Google Scholar
Tilgner, H. et al. Comprehensive transcriptome analysis using synthetic long-read sequencing reveals molecular co-association of distant splicing events. Nat. Biotechnol. 33, 736–742 (2015).
Article CAS PubMed PubMed Central Google Scholar

Download references

Acknowledgements

We thank the Harel lab for stimulating discussion and feedback on the manuscript. We thank Fatma Idrees and Reem Barakat for killifish maintenance, N. Melamed-Book from the imaging facility (HUJI), and A. Turjeman from the genomic center (HUJI). I.H. discloses support for the research of this work from ERC StG #101078188, Zuckerman STEM Program, ISF 2178/19, Israeli Ministry of Science 3-17631, 3-16872, BSF-NSF 2020611, Ministry of Innovation, Science and Technology #0008852. E.M. discloses support for the research of this work from Levi Eshkol scholarship of the Israeli Ministry of Science. R.F. discloses support for the research of this work from the Czech Science Foundation (#22-01781O) and the Ministry of Education, Youth and Sports of the Czech Republic (#CZ.02.1.01/0.0/0.0/16_025/0007370). T.A. discloses support for the research of this work from the Pamela and Paul Austin Research Center on Aging fellowship.

Author information

Authors and Affiliations

Department of Genetics, The Silberman Institute, The Hebrew University of Jerusalem, Givat Ram, Jerusalem, Israel
Eitan Moses, Marva Bergman, Tehila Atlan, Roman Franěk, Omer Ben Dor, Henrik von Chrzanowski, Shay Kinreich & Itamar Harel
Department of Obstetrics & Gynecology, Shaare Zedek Medical Center and Faculty of Medicine, The Hebrew University of Jerusalem, Jerusalem, Israel
Marva Bergman & Ido Ben-Ami
School of Biological Sciences, University of East Anglia, Norwich, UK
Elizabeth M. L. Duxbury & Alexei A. Maklakov
Department of Zoology, University of Cambridge, Cambridge, UK
Elizabeth M. L. Duxbury
University of South Bohemia in Ceske Budejovice, South Bohemian Research Center of Aquaculture and Biodiversity of Hydrocenoses, Vodnany, Czech Republic
Roman Franěk
Department of Biology, Technion-Israel Institute of Technology, Haifa, Israel
Enas R. Abu-Zhayia & Nabieh Ayoub

Authors

Eitan Moses
Marva Bergman
Tehila Atlan
Elizabeth M. L. Duxbury
Roman Franěk
Omer Ben Dor
Henrik von Chrzanowski
Enas R. Abu-Zhayia
Nabieh Ayoub
Shay Kinreich
Ido Ben-Ami
Alexei A. Maklakov
Itamar Harel

Contributions

E.M. and I.H. designed the study. E.M. performed experiments with help from R.F. and M.B. M.B. was co-supervised by I.B.A. and I.H. T.A. designed and performed the analysis of all RNA-seq of vgll3 and rag2 experiments. O.B.D. performed the positive selection analysis. H.C. generated the rag2 mutant. S.K. and H.C. assisted with maturity and lifespan experiments. E.R.A.Z. performed the nuclear recruitment assay under the supervision of N.A. All experiments were conducted under the supervision of I.H. E.M.L.D. performed BaSTA and parametric Weibull analysis under the supervision of A.A.M. E.M., R.F., T.A., E.M.L.D. and I.H. wrote the manuscript. All authors commented on the manuscript.

Corresponding author

Correspondence to Itamar Harel.

Ethics declarations

Competing interests

The authors declare no competing interests.

Peer review

Peer review information

Nature Communications thanks the anonymous reviewers for their contribution to the peer review of this work. A peer review file is available.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Source data

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Cite this article

Moses, E., Bergman, M., Atlan, T. et al. An antagonistically pleiotropic gene regulates vertebrate growth, maturity, and lifespan. Nat Commun 17, 4410 (2026). https://doi.org/10.1038/s41467-026-72381-0

Download citation

Received: 06 January 2026
Accepted: 15 April 2026
Published: 02 June 2026
Version of record: 02 June 2026
DOI: https://doi.org/10.1038/s41467-026-72381-0