Production of Haploid Zebrafish Embryos by In Vitro Fertilization

The production of haploid embryos can be implemented to identify recessive mutations in the F2 generation when the haploids are obtained from mature zebrafish females that are the offspring of mutagenized parents (Figure 1). This saves time and space compared to the traditional diploid screen, in which researchers need to raise F2 families whose diploid offspring are then evaluated for phenotypes of interest (Figure 1).

The major steps in the protocol described in Steps 1-5 above for obtaining zebrafish haploid embryos through in vitro fertilization are schematized as a flowchart (Figure 2). These steps involve preparation of solutions and priming of females by exposure to male hormones in mating tanks (Day 1), followed by sperm and egg procurement and in vitro fertilization (Day 2), and finally rearing of the haploids (Days 3-5) (Figure 2). When combined with a genetic screen, a common mutagenesis procedure in the zebrafish includes previous exposure of the male zebrafish wild type adults to a mutagen (e.g., to a chemical mutagen like ENU, although other procedures, such as retroviral based mutagenesis can be implemented which also facilitate cloning of the chromosomal defect), followed by breeding with wild type females to create a mutagenized F1 generation. Such preparations require several months of work (approximately 6 months) based on the generation times needed to rear wild types, perform mutagenesis, and rear the subsequent mutagenized F1^15,17,22.

Another major point for consideration in employing haploids for a screen is the choice of zebrafish strain. The AB* (AB star) strain has become well known for producing haploid embryos with better overall characteristics and health than other strains¹⁵. However, in the derived Tübingen strain bred in our laboratory, we have recently observed embryological features that have enabled successful screens for developmental anomalies of the renal system (G. Gerlach and R. Wingert, unpublished). Adult zebrafish Tübingen strain males were mutagenized by a series of three ENU treatments, and then crossed to wild type Tübingen females to generate an F1 generation for screening. The F1 females were squeezed to obtain eggs and these F2 eggs were fertilized with UV inactivated sperm. Compared to diploid clutches obtained by natural matings of wild type Tübingen parents (Figure 3A), embryos in these haploid Tübingen strain clutches displayed a shorter, stocky trunk characteristic of the haploid state (Figures 3B,C). Further, haploid clutches typically consist of embryo siblings that display a range of defects, which we observed as well (Figures 3B,C), which is known to vary within strains, as discussed further below^15,17.

Previous studies have established a canonical grading series, corresponding to grades A through D, to categorize the range of haploid phenotype defects^15,17, and additionally haploids have also been referred to in terms of the adjectives good, intermediate and poor quality²². Grade A haploids, corresponding to high quality haploids, are the most normal in appearance, with a short and stocky body but overall show a morphological appearance similar to diploids^15,17,22 (compare diploid siblings (d) to haploid grade A (labeled A) siblings, Figure 3B). Haploid grade A embryos also develop modest pericardial edema, which is a typical feature of haploid zebrafish development^15,17,22 (Figure 3B). In comparison to grade A haploids, intermediate or so-called grade B haploid embryos (labeled B, Figures 3B,C) are distinguished by development of a further shortened or kinked/twisted trunk, though features of a head and tail are clearly distinguishable^15,17,22. Finally, grade C or poor quality haploids (also listed in some references as C/D)^15,17,22 are extremely defective and display a disorganized mass of cells in association of with a yolk ball (labeled C, Figure 3C). Even amongst the same strain, haploid clutches vary in the distribution of A, B, and C phenotypes that one will observe. For example, the haploid clutch from one Tübingen female displayed better development overall, with mostly A and B grade haploid offspring (Figure 3B) compared to haploids obtained from a second Tübingen female, whose clutch consisted of numerous grade C haploid offspring as well as A and B haploid phenotypes (Figure 3C). Similar strain variability of haploid embryo grades have been previously observed in AB and AB* strain fish as well¹⁵.

Despite these morphological differences, organogenesis of many structures proceeds relatively normally in the haploid embryo. Although they have a shorter body trunk, development of organs like the eyes and heart is relatively similar to that of wild type diploid embryos, and cell types such as pigment cells (melanophores and xanthophores) can also be evaluated¹⁵. In addition, we have documented recently that the pronephros, or embryonic kidney, is developed by 1 day post fertilization in the wild type haploid, similarly to wild type diploids. As evidence of this, pronephros cell type patterning displays the prototypical pattern of segments (Figure 4). Further, the phenotype of recessive mutations that alter pronephros development, such as the lib mutation that reduces retinoic acid production, show a similar pattern in the haploid state compared to the diploid state (Figure 4)^23,24. This observation is in keeping with that of other recessive mutations that lead to similar phenotypes in both the haploid and diploid condition, though this is not always the case with all mutations and should be kept in mind if evaluating this type of screen strategy¹⁵.

A number of tissues and organs are typically abnormal in haploids. For example, haploids display defects in circulation, commonly exhibiting blood pooling, which may make their ability to study some aspects of vasculogenesis limited¹⁵. In addition, it has been noted that haploids can have irregularities in ear development, such that they can be evaluated for mutations that prevent otic vesicle formation entirely, but not for mutations that affect subtle processes involved with morphogenesis of this structure¹⁵. As another example of this, brain morphogenesis is abnormal in haploids, and thus haploid screens to identify requisite brain development pathways are limited¹⁵. Despite these limitations, our analysis of kidney development in the haploid state (Figure 4) adds this organ to the list of ‘screenable’ embryonic structures. This finding highlights that a haploid screen methodology can be used to dissect the genetic components of renal progenitor patterning and adds emphasis to the sentiment that ingenious screen approaches can circumvent the limitations of haploid embryo ontogeny to uncover valuable new mutant models to study vertebrate development¹⁵.

Figure 1. Schematic of diploid and haploid screen strategies in the zebrafish model. Following mutagenesis of the parental generation, the F1 generation can be raised and used either to generate F2 mutant families that are used to screen the F3 generation in a diploid state (left) or directly screened in a haploid strategy (right). In a haploid screen, the sexually mature F1 generation females are used to collect haploid F2 clutches for genetic analysis. Female heterozygous carriers are identified if approximately half of the F2 haploids display a mutant phenotype (red embryos) compared to the wild type haploid siblings (yellow embryos).

Figure 2. Haploid embryo production flow chart. The major steps of in vitro fertilization to produce haploid embryos are schematized as tasks performed on Day 1 (pink boxes) to prepare sperm solutions and prime adult zebrafish females, tasks performed on Day 2 (orange boxes) to collect and UV inactivate sperm, to collect eggs, to fertilize the eggs with UV inactivated sperm to generate haploids, and then to revive the female mother, and finally tasks performed on subsequent Days 3-5 (yellow box) when the clutches are incubated to the desired time point(s) for observation and analysis.

Figure 3. Comparison of diploid and haploid Tübingen strain zebrafish embryos. A) Diploid wild type embryos were produced by natural spawning and then incubated until approximately 30 hpf. B, C) Haploid wild type embryos were produced by in vitro fertilization of clutches obtained from F1 generation mutagenized females with UV-inactivated sperm, and incubated until approximately 30 hpf. Haploids exhibited a range of abnormalities in development compared with diploid wild type embryos (black arrowheads, d to indicate diploid). Relatively normal haploids had a shortened, more stock body analogous to the grading scheme of an A haploid¹⁵ (red arrowheads, labeled A), while haploids with gross abnormalities exhibited a severely truncated body axis (purple arrowheads, labeled B) corresponding to a grade of B, and finally grade C (blue arrowheads, labeled C) based on published descriptions¹⁵. All embryos were photographed live at a 2.5X magnification.

Figure 4. Developmental patterning of cell types that comprise the pronephros embryonic kidney organ in haploid Tübingen strain zebrafish. Haploid wild types displayed a normal pattern of cell types in the embryonic kidney structure (top row). The embryonic kidney consists of a pair of segmented functional units known as nephrons. The segmented pattern of discrete cell types present in nephrons can be visualized based on the whole mount in situ hybridization expression pattern of gene transcripts that are unique to the differentiated nephron cell types, which include the podocytes (P) marked by wt1b and nephrin, the proximal convoluted tubule (PCT) marked by slc20a1a, the proximal straight tubule (PST) marked by trpm7, the distal early (DE) marked by slc12a1, and the distal late tubule (DL) marked by slc12a3. lib haploid mutant embryos (bottom row) display reduced podocytes, PCT and an absent PST, along with an expanded DE and DL segment, similar to diploid lib embryos¹⁸. Embryos were photographed in a dorsal view, with anterior to the left, at a 10X magnification.