Insect pests cause enormous damage to human health (through the transmission of diseases such as dengue fever and malaria) and agriculture (through damage to crops or livestock). Existing control methods include pesticides, biological control, and integrated pest and habitat management. However, while many of these approaches have been highly successful, they also have limitations. For example, the use of pesticides can select strongly for resistance, damage non-pest populations (Hawkins et al., 2018) and the environment. The success of biological control and integrated management programmes may also depend upon whether efficient natural enemies are available and the specific ecological setting. As a result, existing control strategies, particularly chemical control, are likely to become increasingly restricted while simultaneously becoming less effective. Global climate change is also predicted to increase the range and the number of insect pests (Deutsch et al., 2018; Gomez-Zavaglia et al., 2020; Sultana et al., 2020). Therefore, it is clear that there are significant challenges for the future in controlling insect pests, to safeguard against disease and maintain global food security.

In light of these concerns, there has been considerable investment in new and alternative technologies, such as genetics-based approaches to pest control, to protect health and food security (Alphey, 2014; Burt and Crisanti, 2018; Raban et al., 2020). Genetic Pest Management (GPM) aims to harness the natural mating system of the pest species to introduce into the target population traits that will reduce fitness and ultimately lead to a reduction of numbers or elimination. The most widely used GPM systems for suppression to date have been variants of the sterile insect technique (SIT) (Klassen, 2005), including the Wolbachia incompatible insect technique (IIT) (Atyame et al., 2011; De Barro et al., 2011; Zheng et al., 2019) and genetic engineering (Phuc et al., 2007). GPM systems that are transmitted or inherited through one sex and sterilise, kill or change the sex of the other offer the most significant potential for control (Bax and Thresher, 2009). As females are predominantly the agents of damage (via biting or ovipositing), and generally determine the effective population size, most approaches have focused on releasing benign males that produce either male progeny or none at all (Rendón et al., 2004). Elimination of females ensures long-term suppression and immediately reduces the associated damage caused by biting or egg-laying.

GPM technologies for insect population suppression currently under development seek to improve on older systems by spreading female-targeting genetic loads through a population or converting female progeny into functional males. These newer technologies also make wide use of contemporary molecular biology tools—particularly those involved in gene editing such as CRISPR/Cas9. However, what is common to all is that they exploit knowledge of the sex determination pathways of the target species, thus exemplifying the importance of incorporating fundamental biological principles to underpin applied science in GPM (Leftwich et al., 2016; Leftwich et al., 2021).

Here we first introduce the fundamentals of insect sex determination systems, focusing on species of interest to GPM. We detail which components are conserved and which show more rapid evolution, what types of primary signals have evolved and in which “direction” they push downstream cascades (e.g., towards maleness or femaleness). We then provide a framework for understanding how sex determination systems have been used to develop insect population suppression tools. We describe three “generations” of genetic engineering technologies with related components or goals. First-generation systems are genetically engineered analogues of the classical Sterile Insect Technique. Second-generation systems are under development and are made possible by the development of the CRISPR/Cas9 platform. Finally, we discuss the challenges inherent in developing “third-generation” control technologies that seek to achieve the goal of sex-conversion by manipulating master regulators of sex determination.

Sex determination systems can be described as a cascade in the form of a pyramid. There is an initial primary signal (“master regulator”—top of the pyramid) that initiates a limited series of intermediary elements (middle) that then result in diverse downstream sexual differentiation and development (base of the pyramid). In insects, the genes at the base of the pyramid tend to be highly conserved, while the elements at the top show marked diversity, both in identity and function (Adolfi et al., 2021; Hopkins and Kopp, 2021). The hypothesis is that these basal genes—which generally consist of transcription factors—represent ancient mechanisms of sex determination [e.g., doublesex is shared with some non-dipteran arthropods (Price et al., 2015; Wexler et al., 2019)]. At the same time, the primary signal can evolve rapidly, even within a taxonomic order. The differences in conservation between the basal and intermediate elements of the sex determination pathway are shown in Figure 1.

Amongst the downstream basal elements, the most well conserved is doublesex (dsx) (Saccone et al., 2002; Price et al., 2015; Verhulst and van de Zande, 2015; Wexler et al., 2019). Ubiquitous amongst insects, dsx is the “central nexus” between sex determination and sexual differentiation cascades (Verhulst and van de Zande, 2015). It functions as a transcription factor activating or repressing thousands of downstream genes which cause female or male somatic differentiation. Its role in this regulation (male or female biasing) is determined by whether it exists in a male or female “form” as a protein. In most cases, this is determined by sex-specific alternative splicing of the initial dsx pre-mRNA—itself determined by intermediary regulators between the primary signal and dsx. “Male” dsx typically represents the constitutive splicing isoform; while female-specific dsx isoforms require the splice enhancing factor transformer (tra) (Figure 2). However, even within this most conserved member of the insect sex determination cascade, some variation does exist. For example, there are significant differences in the number and “style” of exon skipping between different insect species (Verhulst and van de Zande, 2015). For example, in lepidopterans, the constitutive dsx isoform is female with male determining factors required to shift splicing towards a male form (Lee et al., 2015; Xu et al., 2017; Visser et al., 2021). Further, in at least two species of termites, dsx has evolved towards male-only expression rather than sex-alternate splicing (Wexler et al., 2019; Miyazaki et al., 2021).

An intermediary element that exists directly above dsx is transformer (tra). Although not as highly conserved as dsx (Verhulst et al., 2010b), tra homologues have been identified in a variety of insect orders, e.g., Coleopterans [Tribolium castaneum (Shukla and Palli, 2012)], Hymenopterans [Apis mellifera (Gempe et al., 2009), Nasonia vitripennis (Verhulst et al., 2010a)] and dipterans [Drosophila melanogaster (Sosnowski et al., 1989; Inoue et al., 1990), Musca domestica (Inoue and Hiroyoshi, 1986) and a number of Tephritid fruitflies (Pane et al., 2002; Lagos et al., 2007; Ruiz et al., 2007)]. In these groups, sex-specific splicing of tra leads to a “functional” female-specific or “non-functional” male-specific protein. Interestingly, tra has often been found to self-regulate its splicing (except in D. melanogaster) acting as a positive self-regulatory element to increase its own expression (Pane et al., 2002; Gempe et al., 2009; Salvemini et al., 2009; Verhulst et al., 2010a; Hediger et al., 2010). Functional tra acts as a splicing enhancer, binding dsx pre-mRNA and promoting the inclusion of female-specific exons in the final transcript.

Although tra is an essential gene in the sex determination pathway of many dipterans and hymenopterans; in other species, there may be, as yet, no identified homologue, as is the case for Aedes aegypti (Nene et al., 2007). In these cases, there may be functional quasi-equivalents for transformer. For example, in silkmoth (Bombyx mori) males, P-element somatic inhibitor and IGF-II mRNA binding proteins interact to form a complex which binds dsx pre-mRNA. This complex inhibits internal splice site junctions, excluding female-specific exons to produce “male” form dsx mRNA (Suzuki et al., 2010). This example illustrates that while there may be orthogonal splicing factor/s, analogous to tra, the role the new factor/s plays may be very different (promoting male-form, rather than female-form dsx alternate splicing). The transformer-2 gene (tra2) is also involved in the sex determination pathway of many insects. It is often an additional factor that forms an essential part of the splicing enhancer complex, which helps sustain and regulate the splicing of tra (Salvemini et al., 2009). It is, however, not a homologue of transformer itself. Tra-2, unlike tra, has also been shown to have both expression and function in males (Salvemini et al., 2009).

Above tra (or other intermediary elements) in the sex determination pathway lies the primary signal or master regulator underpinning the sexual determination cascade. The identity and function of these master regulators vary enormously between species even within the same order due to a high turnover rate of the primary signaler at this level (Gempe and Beye, 2011). For example, in four dipteran species, the mosquitos Anopheles gambiae, Ae. aegypti, the Mediterranean fruitfly Ceratitis capitata, and the house fly M. domestica, the master regulators of sex determination are evolutionarily unrelated [Yob, Nix, MoY, and Mdmd respectively (Hall et al., 2015; Krzywinska et al., 2016; Sharma et al., 2017; Meccariello et al., 2019)]. While the exact mechanisms by which these primary signals act remains largely unknown, the mosquito species assessed so far (including those listed above) do not appear to possess a tra homologue (Nene et al., 2007). In contrast, the sex-specific splicing of tra is integral to the sex determination cascade in C. capitata and M. domestica, strongly suggesting divergent functions in regulating intermediary elements between the top and bottom levels of the pyramid in mosquitos and other diptera (Saccone et al., 2011). In D. melanogaster, sex is determined by an X-chromosome counting mechanism. The expression ratios of specific X-linked (sis-a, sis-b, sis-c and run) and autosomal genes determine the correct expression of an autoregulatory-splicing female-determining gene (sex-lethal) (Cline, 1993; Kaiser and Bachtrog, 2010). A 1:1 X: A ratio (implying two X chromosomes) leads to functional sex-lethal expression and the female sex determination cascade (Baker and Ridge, 1980; Sánchez and Nöthiger, 1982; Parkhurst et al., 1990). Although tra plays a crucial intermediary role in D. melanogaster (as in C. capitata), the sex-determining role of sex-lethal appears to be Drosophilid-specific (Meise et al., 1998). In Lepidoptera, primary signals can vary, as both Z-chromosome counting and dominant primary signals exist in different species (Traut et al., 2007). In B. mori, a dominant female-linked (feminizer) piRNA system encoded on the female-specific W chromosome silences Z-linked genes that would otherwise initial male sex-determination (Hasselmann et al., 2008; Katsuma et al., 2018); this then directs dsx splicing. In Hymenoptera, sexual fate is effectively regulated by the presence or absence of a paternal genome. E.g., in the honeybee, A. mellifera, it is determined by heterozygosity at a single locus the complementary sex determiner (csd) gene (Gempe et al., 2009), single sex alleles within an organism result in male development (homozygous/hemizygous) and mismatched sex alleles develop into females (heterozygous). While in the haplodiploid wasp N. vitripennis the sex determination gene wasp overruler of masculinization (wom) is only transcribed from the paternally provided genome (Zou et al., 2020). In both systems, tra and dsx are still employed as intermediate and basal elements (Cho et al., 2007; Zou et al., 2020).

The genomic location of the master regulator and, specifically, whether it exists on a heteromorphic sex chromosome is a further important aspect of insect sex determination systems in this review. In many cases, the evolution of distinct sex-chromosomes necessitates a mechanism for equalising the expression of shared, sex-linked genes between sexes (dosage compensation). In D. melanogaster, the absence of sex-lethal in males initiates hypertranscription from the single X-chromosome to make up for two X’s in females. Inactivation of sex-lethal in females leads to deadly X-chromosome hypertranscription due to carrying two X chromosomes. This “coupled” sex determination and dosage compensation has implications for manipulating these systems for genetic pest control. For example, it is a significant challenge to aim to alter sexual fate without simultaneously “programming” the dosage compensation pathway. As a result, in many cases where the master regulator has been identified and ectopically expressed in females (e.g., Yob in An. gambiae; Guy1 in Anopheles stephensi), the result is the death of the XX individuals rather than their conversion to fertile males (Criscione et al., 2016; Krzywinska and Krzywinski, 2018; Qi et al., 2019). This represents a high hurdle if the most potent application of manipulating sex determination for GPM suppression systems is to convert a population to a single-sex rather than selectively kill off females.

Fortunately, neither fully differentiated sex chromosomes, nor coupled-dosage compensation pathways where heterogametic sex chromosomes exist, appear universal amongst insects that are of concern to human health or agriculture. For example, Aedes albopictus (and other culicine mosquitos) do not possess heterogametic sex chromosomes but rather a small “Male-determining” locus on chromosome 1 (Hall et al., 2014; Gomulski et al., 2018). Transgenic expression of the male-determining gene from within this locus (Nix) in transgenic Ae. albopictus was sufficient to convert females into functioning males (Lutrat et al., 2022). Similarly, in C. capitata, transient ectopic expression of MoY is enough to convert karyotypic females to functional males, suggesting either a lack of dosage compensation or an “un-coupled” version in this species (Meccariello et al., 2019).

Understanding the nature of the sex determination pathway that has evolved in a pest of concern, and its possible interaction with dosage compensation, provides some potential routes for manipulating a species for genetic pest management. If the goal is female to male sex conversion, then the upper levels of the pyramid will likely need to be manipulated to ensure complete sex conversion and the viability/fertility of converted individuals. However, this goal may be difficult or practically impossible for species with lethal dosage compensation. If the goal is simply to kill one of the sexes, then lower, more conserved levels of the pyramid can be targeted, and this may also be beneficial in transferring efficient designs between related pests. Further exploration of the fundamental basis of sex determination mechanisms is, therefore, essential.

First-generation transgenic GPM systems are genetically engineered analogues of the classical sterile insect technique (SIT). In the SIT, the mass release of irradiated (sterilised) males results in a lower proportion of fertilised females in the field due to mating with the sterile males instead of the fertile, wildtype males. SIT is most efficient if only males are released (Rendón et al., 2004). Preventing the introduction into the population of females that can damage fruit crops or transmit disease is an additional advantage. However, male-only releases require an efficient mechanism for sex sorting. For this purpose, genetic sexing strains (GSS- Box 1) were developed that differentiated between males and females with selective markers such as pupal colour or conditional lethality when exposed to high temperatures (Rendon, 1996).

First-generation transgenic systems sought to improve these technologies by creating analogues of GSSs that could also be used as population suppression measures in the field. The most widely adopted of these was the Release of Insects carrying a Dominant Lethal technology (RIDL) (Thomas et al., 2000), but also see (Schetelig and Handler, 2012) (Ogaugwu et al., 2013; Schetelig et al., 2016). The basis behind RIDL is the genetic modification of a pest to carry a deleterious/lethal gene whose expression can be turned off (repressed) during rearing, but which, when inherited by the progeny of released insects, will result in lethality for individuals in the field. As with SIT, mass releases of RIDL insects can thus suppress a target population by continually killing off field-born individuals before they can reproduce. Female-specific RIDL (fsRIDL) and genetic sexing strains have been developed in many species by combining this repressible lethality with sex-alternatively spliced introns from basal genes within the sex determination pathway (Fu et al., 2007; Schetelig and Handler, 2012; Ogaugwu et al., 2013; Tan et al., 2013), specifically dsx and tra. The pre-mRNA of these two genes is spliced differently between males and females—leading to the sex-specific inclusion or exclusion of exonic sequences. Sex-specific transgene expression can be designed by including the sequences responsible for sex-specific splicing (introns) embedded within components integral to the repressible-lethal system. As such, functional fsRIDL proteins are produced in one sex (usually females) and not the other (in the same manner as tra and sex-specific dsx proteins). Released fsRIDL individuals are homozygous fertile males that produce heterozygous male-only offspring when mating to wildtype females following release. These heterozygous fsRIDL males can then produce wildtype males and females as well as heterozygous fsRIDL males, resulting in a steadily diluting suppressive effect without continued releases. Female-specific RIDL lines have been developed in many insect pest species, including tephritid fruitflies (Fu et al., 2007; Ant et al., 2012), blow flies (Yan and Scott, 2020), screwworms (Concha et al., 2016), moths (Morrison et al., 2012; Jin et al., 2013) and mosquitoes (Phuc et al., 2007; Collado, 2013). Proof-of-principle demonstrations have also been made in beetles (Gregory, 2015). Caged and open-release trials have demonstrated that repeated releases of fsRIDL males can cause the rapid suppression of target populations (Harris et al., 2011; Leftwich et al., 2014; Carvalho et al., 2015; Shelton et al., 2020).

For these first-generation technologies, it is not necessary to know the precise means by which the sex-specific processing of dsx or tra components are regulated (i.e., the upstream elements of the pyramid which act upon them). All that is required is a basic understanding of the arrangement of the chosen gene and final mRNA differences between sexes. This conservation of function is a distinct advantage for adapting transgenic constructs, with minimal changes, across multiple species (Tan et al., 2013). A further advantage of using highly conserved basal elements of the pyramid is that the splicing signals which regulate their sex-specific splicing are often shared between closely related species (Ant et al., 2012). For example, an fsRIDL construct built using intronic sequences from pink bollworm (Pectinophora gossypiella) dsx functioned just as effectively in that species as it did in silkworm (B. mori) and diamondback moth (Plutella xylostella) (Jin et al., 2013; Tan et al., 2013). The limitation of using these downstream elements of the pyramid is that these systems are generally limited to killing females rather than their conversion to males. This, coupled with the self-limiting nature of these technologies, makes them far less potent than “second-generation” technologies (next section).

Whereas first- and second-generation technologies seek to manipulate or disrupt basal elements of the sex determination pyramid in order to reduce the fitness of or kill females, future “third-generation” technologies may be designed to manipulate the master regulators of sexual fate, to affect full sex conversion. If attached to an efficient gene drive system, such a technology would spread through a target population causing a growing wave of sex distortion. This is theoretically more efficient than a second-generation system that kills or incapacitiates one sex as homozygotes, because all inheritors, regardless of their genotype, continue to spread the system. This increased efficiency could potentially allow for a dramatic reduction in the number and size of releases required for population control. Depending on the efficiency of sex conversion, this could enable threshold-dependent gene drives to be used, previously discounted for population suppression because of their intolerance to high fitness loads (Leftwich et al., 2018; Champer et al., 2020a).

Many aspects of the theory underpinning such third-generation technologies pre-date second-generation strategies (Lyttle, 1991). However, they have proven challenging to enable in practical terms. Part of the reason is the non-conserved nature of the upstream components that need to be manipulated. This requires deep and specific knowledge of master regulators and their web of interactions with downstream elements for each specific pest species to be targeted. Even then, the unpredictable/inflexible nature of “coupled” sex determination and dosage compensation systems may make such a design unachievable in some species. Hence for third generation systems, the transfer of efficient gene drive designs between pests may not always be possible. In a similar vein, if sex-specific components of fitness are sex-linked it may be the case that efficient sex conversion can be achieved, but the sexual competitiveness of converted individuals is diminished.

The mechanisms for enacting sex conversion will vary greatly depending on whether the targeted gene(s) are the master regulators of sex determination (and whether these initiate a male or female cascade), or those genes directly downstream (tra or tra-2). In an XX/XY system, sex conversion through expression of a master regulator is likely to produce a dominant effect. This is highly likely to affect the dynamics of a gene drive. For example, the effects on population suppression would be seen much earlier than one where conversion is enacted through disruption of a recessive switch e.g., tra or tra-2 (Hoshijima et al., 1991; Pane et al., 2002; Sarno et al., 2010). For systems where the presence of the master regulator determines femaleness [for example, the ZZ/ZW systems common to Lepidoptera (Suzuki et al., 2010)], maleness could be achieved by inactivating the master regulator or making the element below it resistant to its activity [see (Sakai et al., 2016)].

In the next section, we discuss evolutionary and empirical manipulation studies of dosage compensation and sex-linked fitness traits and outline the hurdles these may pose to engineering efficient third-generation technologies.

Manipulating sex determination pathways for genetic pest management has many potential applications. Previous technologies have used the highly conserved “basal” elements of dsx and tra common to almost all insect species to produce reliable mechanisms of biasing sex ratios with the release of modified males carrying factors to generate female sterility or death. Newer technological developments, including homing gene drives, demonstrate these basal elements continue to be predictable and reliable targets for control. Looking forward, development of genetic editing techniques to manipulate “master regulators” of sexual fate have the potential to improve the performance of a wide variety of genetic control methods. However, this approach has potential challenges—different species may exhibit sex-linked genes that are vital for viability or sexual fertility or have strong dosage compensation. However, this is a vibrant field of research and much experimental work is ongoing in a range of different pest species. While it is likely that the application of sex conversion for pest control will inevitably be applied on a case-by-case basis, active investigations on a number of fronts are likely to improve our understanding of the basic biology and evolution of sex determination, as well as expand our genetic toolbox for applied pest management. BOX 1

Two modifications are principally required for Genetic Sexing Strains (GSS) 1) introduction of a recessive conditional lethal gene or viable selectable recessive colour mutations and 2) translocation of a wild-type rescue allele onto the male Y-chromosome. In the final strains, females are homozygous for one or more selectable mutations, while males are heterozygous with a wildtype phenotype (Rendon, 1996). These strains are highly effective at producing substantial bias in the reproductive sex ratios or enabling efficient sex separation, but because of the mutations and chromosome translocations required to generate these strains, high levels of sterility and rearing difficulties were common, many strains were also unstable as a result of these complex chromosomal rearrangements (Robinson, 2002; Nguyen et al., 2021).

Genetic sexing strains produced strong genetic male bias, achieved not through manipulation of the sex determination pathway, but by positioning autosomally lethal alleles onto the sex chromosomes. The advent of powerful genome editing tools and synthetic biology has allowed for the development of other, more refined artificial sex distortions such as X-shredding. These systems exploit the heterogametic nature of these species, where fathers always transmit their X chromosome to their daughters and the Y chromosome to their sons, to cause lethal changes to essential genes, without involving the sex determination pathways directly.

X-shredders were first pioneered in Anopheles gambiae when I-PpoI was discovered to cut the X chromosome in several locations due to its targeting of the repeated ribosomal rDNA (Windbichler et al., 2007). By engineering a destabilised version I-PpoI its activity could be restricted to male meiosis thereby ensuring males were unable to pass on a functional X chromosome (Galizi et al., 2014). As opposed to using an endonuclease which targets a conserved repetitive element; X-shredding can also be driven using CRISPR/Cas9 and targeted sgRNAs, with Cas9 cleavage limited to the male germline (Galizi et al., 2016; Fasulo et al., 2020; Meccariello et al., 2021).

Y-linked editors have also been proposed as a self-limiting strategy significantly more effective than those previously proposed (Burt and Deredec, 2018). If released simultaneously with an autosomal X-shredder this efficiency can be further increased. An alternative approach would be to drive an X-shredder from the Y-chromosome to ensure male offspring inheritance (Gamez et al., 2021). Other proposed Y-linked suppression systems include Medusa; combining a maternally-expressed, X-linked toxin and a zygotically-expressed, Y-linked antidote that causes suppression of the female population and selects for the transgene-bearing Y. At the same time, a zygotically-expressed, Y-linked toxin and a zygotically-expressed, X-linked antidote selects for the transgene-bearing X in the presence of the transgene-bearing Y to create a threshold dependent, highly male-biasing suppression system (Marshall and Hay, 2014) present its own challenges as expression during spermatogenesis can be difficult to achieve from the Y chromosome due to transcriptional repression (Alcalay et al., 2021).