Genetic manipulation of sex ratio in mammals: the Reaper comes for Mickey

Abstract

In most animals, sexual reproduction results in a 1:1 ratio of females to males. For several sectors of agriculture, for example, milk or egg production, only a single sex is needed. Biasing the sex ratio so that only offspring of the desired sex are produced has the potential to increase breeding efficiency. In this issue of EMBO Reports, Yosef et al [1] demonstrate a genetic approach to bias the sex ratio in mice by specifically disrupting essential genes in male embryos. Their approach is an important first step toward generating sex‐ratio biasing applications for agriculture.

Sexual reproduction is nearly universal among animal species. Sex can be determined genetically, as in mammals and birds, or by environmental mechanisms such as temperature, which is common in reptiles 2. The natural sex ratio of a species can vary depending on its sex determination mechanism and reproductive lifestyle, but usually is 1:1 in animals with XX/XY sex determination (male heterogamety), for example, mammals; or ZW/ZZ sex determination (female heterogamety), for example, birds. Agriculture and aquaculture have long sought to gain efficiency by manipulating the sex ratio in favor of the more profitable sex. Sex‐ratio biasing also has potential for population control of invasive species, agricultural pests, or disease vectors.

In fish and some other aquatic animals, sex hormones can be used to channel sexual development down the male or female path, but in terrestrial vertebrates, this does not work. Indeed, sex ratio is among the few traits that have not been effectively manipulated by conventional selective breeding programs. Currently, the main alternative to culling individuals of the “wrong” sex after birth is to employ artificial insemination using sex‐sorted semen prepared by flow cytometry 3. This process is effective in some mammalian species but is expensive and technically complex and is not yet practical in birds due to their filiform sperm morphology 4. Similarly, technologies are available to determine the sex of unhatched chicks in eggs 5, 6, but these methods are not currently economical at the scale needed for widespread adoption in commercial breeding programs. Approaches involving heritable genetic modifications would potentially be much simpler.

The discovery of sex‐determining genes in a broad variety of species has made it conceptionally straightforward to reduce the abundance of one sex genetically by transforming it into the other. In one application, genetic manipulation of sex determination has been shown to be effective for population control in insects 7. However, in species with genetic sex determination, sex reversal often results in infertility because essential reproductive genes tend to be localized to the sex chromosomes. This potentially limits the use of sex determination modifications in agriculture. Yosef et al 1 have taken a different approach, leaving sex determination intact but using the CRISPR/Cas9 system to selectively kill embryos of one genetic sex. They have tested a strategy that involves crossing animals from two true‐breeding transgenic mouse strains to produce hybrid offspring with a highly biased sex ratio. The system needs further optimization but provides the first proof of concept for genetic sex biasing in mammals.

Yosef et al 1 created a transgene that expresses three guide RNAs (sgRNAs) that target genes known to be essential for embryonic development and integrated it on the Y chromosome in an intronic site that is permissive for gene expression 8 (Fig 1A). The resulting strain is healthy and fertile and has a normal sex ratio. However, when males of this strain are bred to females with a Cas9 nuclease expression vector integrated on an autosome, the sex ratio of the resulting progeny is severely biased toward female. This sex bias presumably results from death of XY embryos (Fig 1B), and consistent with this, a small number of XY pups were born but most died before weaning. Sequencing DNA from males that survived to birth revealed that all had at least two of the three target genes disrupted in a significant proportion of cells in the tail.

Figure 1. A genetic system for biased sex production in mammals.

Design of transgene expression constructs (A) and breeding strategy for sex selection (B) from Yosef et al 1. Female (XX) embryos inherit the Cas9 expression transgene but not the Y‐linked sgRNA transgene and thus survive, while male (XY) embryos inherit both transgenes and are killed by mutation of up to three essential embryonic genes. (C) Mouse developmental cycle, with presumed timing of embryo lethality based on target genes disrupted indicated in cyan. Meiotic mutagenesis of sex‐determining gametes (D) would decrease egg wastage and allow for sex‐ratio biasing without concomitant decrease in litter size.

The approach used by Yosef et al shows the potential of combining programmable nuclease technology with strategic positioning of transgenes on a sex chromosome. It also has advantages over some alternative strategies for inducing sex bias in vertebrates. For example, a previous study showed that expression of ricin from a female‐specific vitellogenin promoter can bias sex toward male in zebrafish 9. However, the late action of this transgene after sexual maturation and the presence of a toxin‐encoding gene in all animals would severely limit the application of similar approaches in aquaculture or agriculture.

The sex‐specific gene targeting approach can potentially be adapted to bias sex in either direction and in species with different sex determination systems. To bias mammalian sex against females, males with sgRNAs encoded on the X chromosome could be mated to females with the autosomal Cas9 transgene. This would cause lethality only in XX progeny, as the XY progeny inherit just a maternal X. To bias avian sex, analogous approaches could be used. Birds have sex chromosomes denoted Z and W, and females (ZW), rather than males (ZZ), are the heterogametic sex. To bias against females, the sgRNAs could be integrated on the female‐specific W chromosome; to bias against males, the maternal Z chromosome could carry the sgRNA construct and the paternal genome would have autosomal Cas9. In addition to versatility and ease of use, an important consideration for genetic sex‐biasing systems is whether they are environmentally safe. In this case, the answer would appear to be a cautious yes, since any escapees would quickly be diluted by the wild population and would have no consequences to the breeding population unless they mated with individuals carrying the other transgene.

While sex biasing by sex‐specific gene targeting has strong potential, several improvements are needed before it is ready for prime time. First, to make the process more humane, targeting needs to be earlier and/or more efficient. This could likely be accomplished by optimizing the genetics of the current approach (e.g., targeting additional essential embryonic genes, using more sgRNAs per gene, and tuning the expression levels of sgRNA and/or Cas9 nuclease components). Second, targeting the embryo is a wasteful process as there is still a maternal investment in eggs that will not produce viable progeny. An approach that addresses both issues would be to move the time of sex selection earlier to the stage of gamete development (Fig 1C). In the “X‐shredder” scheme, for example, sgRNAs that target sequences unique to the X chromosome and are expressed only during gametogenesis are positioned on the Y chromosome 10. The consequence is that in transgenic Y‐bearing males, the X‐bearing sperm are inviable and only Y‐bearing functional sperm are produced, leading to all male progeny (Fig 1D). By targeting the gametes instead of embryos, this approach avoids any late lethality and also eliminates wastage of inviable eggs. In other words, the brood size is not cut in half by the sex‐selection process. Although more complicated to engineer, it may be possible to improve the system even further by incorporating the components of an inducible Y shredder on the Y chromosome itself, allowing production of non‐GMO all‐XX progeny.

In summary, the study by Yosef et al highlights the potential of using modern genome engineering approaches to bias mammalian sex ratio strongly toward females. There are several improvements to their design that could increase the efficiency and ethics of genetic sex‐ratio biasing, and these are conceptually possible with modern genome editing practices. Importantly, for any such strategies to be useful in agriculture, they will have to integrate with modern breeding practices, where true‐breeding strains are often preferred and livestock genetics are continuously improving. Despite these challenges, it is encouraging to see how modern precision genome engineering tools could transform agriculture in the coming decades.

EMBO Reports (2019) 20: e48577