Embryonic Lethality, Juvenile Growth Variation, and Adult Sterility Correlate With Phylogenetic Distance of Danionin Hybrids

Abstract

Hybrid incompatibility, which plays a pivotal role in speciation, is expected to correlate with greater phylogenetic distance. Here, we investigate the fitness of interspecies hybrids within the Danionin subfamily, which includes the model species, Danio rerio, and its relatives - Danio kyathit, Danio albolineatus, Danio margaritatus, and Devario aequipinnatus. We generated hybrids through in vitro fertilization, using Danio rerio as the maternal species, with normal fertilization rates showing no incompatibilities in sperm-egg interactions within these two genera. Generally, all hybrids exhibit normal patterns and timelines in early developmental transitions, from cleavage stages to the initiation of epiboly, although inter-genera Danio-Devario hybrids subsequently exhibit fully penetrant embryonic lethality. Intra-genus Danio hybrids, on the other hand, can survive through embryogenesis and into adulthood. However, rates of survival during these stages diminish according to phylogenetic distance, with increasing early lethality in hybrids from more distantly related species. Additionally, Danio hybrids exhibit increased growth rate variability during juvenile stages. All Danio hybrids have reduced testes sizes, sperm counts, and sperm viabilities, with sperm displaying defects in flagellum formation and integrity. Adult male intra-genus hybrids are invariably sterile, except in the case of Danio rerio hybrids with the closely related Danio kyathit, which produced a backcrossed F2 generation that did not survive juvenile stages. Our studies highlight a loss of hybrid compatibility at various life stages in the Danio and Devario genera, based on deleterious effects and reduced developmental robustness, emphasizing a correlation between the severity of incompatibility outcomes and the degree of phylogenetic relatedness.

Keywords: danionin; phylogeny; postzygotic incompatibility.

Figure 1

CytB Danionin phylogenetic tree. The evolutionary history of Danionin species used for this study was inferred using the neighbor‐joining method. The optimal tree is shown based off cytochrome b sequences. The tree is drawn to scale, with branch lengths in the same units as those of the evolutionary distances used to infer the phylogenetic tree. The evolutionary distances were computed using the p‐distance method and are in the units of the number of base differences per site. All ambiguous positions were removed for each sequence pair (pairwise deletion option). There were a total of 1141 positions in the final data set. Evolutionary analyses conducted in MEGA11 (Tamura, Stecher, and Kumar 2021). [Color figure can be viewed at wileyonlinelibrary.com]

Figure 2

Conserved early developmental features in Danionin species. Intraspecies embryo face‐view micrographs of D. rerio, D. kyathit, D. albolineatus, D. margaritatus, and Dev. aequipinnatus through the midblastula transition. (A) Two‐ to four‐cell stages (0.75–1 hpf). (B) Eight‐ to sixteen‐cell stages (1.25–1.5 hpf). (C) Thirty‐two cell stage (1.75 hpf). (D) High stage (3.3 hpf). (E) Sphere to dome stages (3.8–4.3 hpf) when development has transitioned from maternal to zygotic control. (F) Developmental staging through the midblastula transition (MBT; gradient gray box) traced by timelapse imaging. The average timepoints of the first 10 embryonic stages (1‐cell through 1K‐cell stages) were traced computationally through object tracking of cellular divisions (designated by the line plots). The remaining four stages, presented as dot plots, were classified manually by determination of diagnostic morphological features (per Kimmel et al. 1995) and average timepoints in timelapse images. 250 µm scale bar. [Color figure can be viewed at wileyonlinelibrary.com]

Figure 3

Initiation of early developmental incompatibilities in interspecies Danionin hybrids. Embryo face view micrographs of D. rerio/D. rerio (Dr x Dr) controls, D. rerio/D. kyathit (Dr x Dk) hybrids, D. rerio/D. albolineatus (Dr x Da) hybrids, D. rerio/D. margaritatus (Dr x Dm), and D. rerio/Dev. aequipinnatus (Dr x Dev) hybrids through the midblastula transition. (A) Two‐cell to four‐cell stages (0.75–1 hpf). (B) Eight‐cell to sixteen‐cell stages (1.25–1.5 hpf). (C) Thirty‐two cell stage (1.75 hpf). (D) High stage (3.3 hpf) (E) Dome stage (3.8–4.3 hpf) when development has transitioned from maternal to zygotic control. (F) Developmental staging through the midblastula transition (MBT; gradient gray box) traced by timelapse imaging. Average timepoints as in Figure 2. 250 µm scale bar. [Color figure can be viewed at wileyonlinelibrary.com]

Figure 4

Early hybrid incompatibilities result in embryonic defects. (A) At 1 dpf, survival of hybrid embryos from Dr x Dk and Dr x Da crosses was similar to that of time‐matched D. rerio (Dr x Dr), while hybrid embryos from Dr x Dm and Dr x Dev crosses exhibited a higher fraction of developmental defects. (B) D. rerio control embryo imaged on 1 dpf. Clutches of Dr x Dr control embryos were consistently viable at 1 dpf embryos (B'). (C–F) Hybrid embryos (Dr x Dk, Dr x Da, Dr x Dm, and Dr x Dev) at 1 dpf. 1 dpf, embryos from Dr x Dk (C, C') and Dr x Da (D, D') hybrid crosses are largely normal and indistinguishable from control D. rerio embryos, clutches from Dr x Dm hybrid crosses consist of a mixture of viable and inviable embryos (E, E'), and clutches from Dr x Dev hybrid crosses show predominantly severe gastrulation defects and are inviable (F, F'). Embryos were scored for phenotypic severity (inset diagrams in B'–F') from Grade 0 (G0; wild type) to Grade 3 (G3; severe gastrulation defects) according to classification at the top right (see Results for details). Statistically significant differences to intraspecific Dr x Dr controls [392–G0: 392], by two‐way ANOVA's; Dr x Dm [243 – G0: 51, G1: 68, G2: 90, G3: 34; p < 0.001]; Dr x Dev [n = 280–G0: 0, G1: 26, G2: 47, G3: 207; p < 0.001]. [Color figure can be viewed at wileyonlinelibrary.com]

Figure 5

Intra‐genus Danio hybrid juveniles exhibit reduced and variable growth rates. (A) Threshold images of D. rerio (Dr x Dr) controls and hybrids of Dr x Dk, Dr x Da, and Dr x Dm crosses at the larval‐to‐juvenile transition (6 wpf, A, A'), juvenile stages (8 wpf, B, B'; 12 wpf, C, C') and young adults (16 wpf, D, D'). Fry of interspecies hybrid crosses show smaller body length (TL) measurements compared to D. rerio controls. TL variances (S²) among hybrids appear to be increased compared to control (Statistically significant differences among hybrid and control variances; Levene's F‐tests; p < 0.01). A subset of hybrid fry appears to exhibit slower growth rates compared to control, an effect that becomes more prominent as juveniles grow and which leads to the segregation of discrete groups of slow‐growing individuals that is most prominent in older fish and in hybrids from species with greater phylogenetic divergence. Asterisks (*) in (A'–D') represent statistically significant difference between TL variances of hybrid and D. rerio controls. [Color figure can be viewed at wileyonlinelibrary.com]

Figure 6

Intra‐species Danio hybrid adults exhibit reduced fertility. (A) Average fertilization rates in IVF backcrosses using hybrid sperm and D. rerio eggs, showing reduced fertilization potential in hybrid sperm. (B) Threshold images of testes in intraspecies (Dr x Dr, Dk x Dk, Da x Da, Dm x Dm) controls and interspecies (Dr x Dk, Dr x Da, Dr x Dm) hybrids. (C, D) Quantification using pixel area measurements show reduced proportional testes sizes of all hybrids relative to D. rerio controls (C) and their respective intraspecies controls (D). [Color figure can be viewed at wileyonlinelibrary.com]

Figure 7

Reduced sperm density and viability in intra‐species Danio hybrids. (A, B) Representative images of Eosin‐Nigrosin (EN) live‐dead staining of D. rerio control and hybrid (Dr x Da) sperm. Using this stain, viable/live sperm retain a white sperm head (white arrow) and nonviable/dead sperm have a purple influx in the sperm head (black arrowhead). (C) D. rerio control sperm are largely viable and sperm from hybrids exhibit drastic decreases in viability. [Color figure can be viewed at wileyonlinelibrary.com]

Figure 8

Flagellar defects in sperm from intra‐species Danio hybrids. (A) D. rerio (Dr x Dr) control sperm. (B–D) Intraspecies control sperm of D. kyathit, D. albolineatus, and D. margaritatus males. (E, F) Sperm from interspecies hybrid males (Dr x Dk, Dr x Da, and Dr x Dm) exhibiting truncated tails. (E', F') Representative sperm from Dr x Dk and Dr x Da hybrids exhibiting flagellar microtubule breaks (white arrowheads) along the tail region. (G') Dr x Dm sperm exhibiting accumulated tubulin around the sperm head and no flagellar elongation. (H) All interspecies hybrids show reduced sperm flagellum length, with greater prevalence in sperm pools from Dr x Da and Dr x Dm hybrids. (I) Dr x Da hybrid sperm have a range of widths with some measuring two times that of D. rerio control sperm (*). (J) Quantification of the number of breaks observed among randomly selected sperm tails from hybrid crosses. Breaks were most prevalent among Dr x Da and Dr x Dm hybrid sperm. (K–M) Counts sorted by three regions along the sperm tail, proximal, mid, and distal, show no obvious distribution of flagellar break events. [Color figure can be viewed at wileyonlinelibrary.com]

Figure 9

F2 Hybrid breakdown in D. rerio/D. kyathit juvenile progeny. (A) Schematic of hybrid crosses to generate F2 hybrid backcrossed progeny. (B) D. rerio (Dr x Dr) control at 1 dpf. (C) Danio rerio (Dr) x Danio rerio/Danio kyathit (Dr*Dk) F2 hybrid embryo at 1 dpf showing indistinguishable morphology as D. rerio control. (D) Juvenile counts of time‐matched D. rerio controls (solid line) and Dr x Dr*Dk F2 hybrids (dotted line). Loss of F2 hybrids occurs throughout the observed period, with no fry surviving beyond 16 wpf.