CRISPR-based genetic control strategies for insect pests

Ying YAN,Roswitha A.AUMANN,Irina H?CKER,Marc F.SCHETELIG

Department of Insect Biotechnology in Plant Protection,Institute for Insect Biotechnology,Justus-Liebig-University Giessen,Giessen 35394,Germany

Abstract Genetic control strategies such as the sterile insect technique have successfully fought insect pests worldwide. The CRISPR (clustered regularly interspaced short palindromic repeats) technology,together with high-quality genomic resources obtained in more and more species,greatly facilitates the development of novel genetic control insect strains that can be used in area-wide and species-specific pest control programs. Here,we review the research progress towards state-of-art CRISPR-based genetic control strategies,including gene drive,sex ratio distortion,CRISPRengineered genetic sexing strains,and precision-guided sterile insect technique. These strategies’ working mechanisms,potential resistance development mechanisms,and regulations are illustrated and discussed. In addition,recent developments such as stacked and conditional systems are introduced. We envision that the advances in genetic technology will continue to be one of the driving forces for developing the next generation of pest control strategies.

Keywords: insect pest,genetic control,sterile insect technique,gene drive,genetic sexing strains,CRISPR-Cas9

1.Introduction

Climate change,global food security,environmental sustainability,and vector-borne diseases have become major challenges in the 21st century. For the last two decades,insect pests are playing significant roles in the agricultural losses,and infectious diseases transmitted by vector insects have been spreading globally (Oerke 2006;WHO 2014;Deutschet al.2018). Those problems are expected to be aggravated with rising temperatures and other climate change factors (Jactelet al.2019;Tonnanget al.2022). Currently,one of the most important control methods for insect pest is the use of insecticides,which could pose risks to the environment and human health (Bonner and Alavanja 2017;Raniet al.2021).Moreover,heavy use of insecticides has led to resistance development in many pest species (Liu 2015;Jurat-Fuenteset al.2021). Therefore,alternative pest control approaches that are cost-effective,environmentally friendly,and more sustainable,are urgently needed and have been one of the focuses of modern pest management research. Genetic control,also known as genetic pest management (GPM),is a biological control method for pest species that introduces desirable and heritable genetic modifications into a wild populationviaintraspecific mating (Kniplinget al.1968;Vanseventer 1975;Cockburnet al.1984;Curtis 1985;Harvey-Samuelet al.2017;Devoset al.2022). The sterile insect technique (SIT) was the first implemented and one of the most successful genetic control strategies for insect pests.The SIT was used to eradicate the New World screwwormCochliomyiahominivorax(Coquerel),a devastating fly endangering livestock,wildlife,and human health,from North and Central America,which was considered an unprecedented achievement in insect pest management(Klassen and Curtis 2005;Scottet al.2017). In an SIT program,insects of the target species are mass-reared,sterilized by radiation,and released into the field to mate with their wild-type (WT) counterparts,which consequently will produce no offspring (Knipling 1955,1959) (Fig.1-A).Such an approach is species-specific,highly effective,and has been successfully incorporated into area-wide integrated pest management programs (AW-IPM) to battle insect pests worldwide (Klassen and Curtis 2005;Vreysenet al.2021).

Fig.1 Sterile insect technique (SIT) and CRISPR-based genetic control strategies. A,for SIT control programs,the insects are sterilized by radiation and mating between sterilized and wild-type (WT) insects produce no offspring. However,the sterilized females that are co-released would mate with the sterilized males and reduce the control efficacy. B,the homing CRISPR-based gene drive (GD) promotes the super-Mendelian inheritance and can be used to target genes that control the female fertility or viability. C,CRISPR-Cas9 sex-ratio distortion (CRISPRSRD) selectively targets sequences that are exclusively located on the X chromosome thus causes “X-shredding” or “X-poisoning” (modified from Galizi et al. 2016). Such method results in mostly Y chromosome-bearing sperm and produces male-biased population. D,CRISPR-engineered genetic sexing strains (CRISPR-GSS)can be based on phenotypic markers for mechanical sorting,and/or temperature-sensitive (ts) mutations. Mutations in autosomal phenotypic marker genes as well as ts mutations,generated using CRISPR-mediated homology-directed repair (HDR),in essential autosomal genes,need to be rescued via a translocation of a functional WT allele onto the Y chromosome (A/Y) to restore the WT phenotype in males. When the ts mutation is introduced to a sex determination gene,the restrictive temperature can theoretically transform females into phenotypic males thus produce a male-only population for release. E,the precision guided SIT (pgSIT)simultaneously targets genes that confer female specific (fs) lethality and male specific (ms) sterility by crossing Cas9-expressing and gRNA-expressing lines. Such an approach could eliminate all females and only produce sterile males that can be released.

For SIT programs,females are considered inefficient control agents since they would mate with co-released males and thus prevent the sterilized males from mating with the WT females in the field (Rendonet al.2004;Franzet al.2021). Moreover,released females could cause undesired damage to crops or contribute to disease transmission in the case of vector insects. In addition,massive numbers of insects are needed for SIT release to ensure that most of the matings involve one sterile partner (Knipling 1955,1959;Schliekelmanet al.2005).Therefore,an efficient sex separation of insects in massrearing facilities and the release of only males have been important factors that improve SIT efficiency. Genetic sexing strains (GSS) based on sex separation enabled by sex-specific phenotypes,or transgenic sexing strains(TSS) based on conditional female-specific lethality conferred by transgene(s),have been developed to remove up to 100% females from release populations(Harvey-Samuelet al.2017;Alphey and Bonsall 2018;Franzet al.2021;H?ckeret al.2021). It is also known that radiation often reduces the insects’ fitness,resulting in less competitive insects for release compared to their WT counterparts (Lanceet al.2000;Klassen and Curtis 2005;Diaset al.2021). This is especially true for Lepidopteran pests since they are highly resistant to ionizing radiation and thus require high doses of gammaor X-rays to induce complete sterility (Simmonset al.2010;Marec and Vreysen 2019). Therefore,genetic control strategies that generate more effective and robust release agents would further improve the overall efficacy of the SIT program and extend the technology's range of targeted species and control goals (e.g.,suppressionvs.eradication).

The last decade has seen rapid development and broad application of CRISPR (clustered regularly interspaced short palindromic repeats)-based geneediting technologies in the life sciences (Doudna and Charpentier 2014;Hsuet al.2014;Knott and Doudna 2018;Cannon and Kiem 2021). Originally,the CRISPR system is derived from the adaptive immune system of bacteria and archaea. The applicable derivative is composed of a CRISPR-associated (Cas) nuclease that cleaves DNA by forming site-specific doublestranded breaks (DSB) and a single chimeric guide RNA(sgRNA) to recognize its target sequence (Horvath and Barrangou 2010;Wiedenheftet al.2012). The sgRNA is a simplification of the natural CRISPR system created by molecular biologists by combining the two separately expressed RNAs,a CRISPR RNA (crRNA) and a transactivating crRNA (tracrRNA),which are processed by the endogenous bacterial machinery to yield the mature gRNA(Deltchevaet al.2011). The DSB induced by the CRISPR system can be repaired by one of the two cellular DNA repair pathways: non-homologous end-joining (NHEJ)or homology-directed repair (HDR) (Sander and Joung 2014). The NHEJ pathway can cause random insertions or deletions (InDels) at the DSB site. NHEJ-targeting of the coding sequence (CDS) of a gene can therefore lead to a so-called gene knock-out,as InDels can interrupt the reading frame and thus prevent correct protein production.The HDR pathway can generate specific mutations or introduce genes of interest into the genome by providing the corresponding DNA template for the repair,which is called sequence knock-in. Both repair strategies have been rapidly adopted in different insect species and used for functional genomics (Sunet al.2017;Taninget al.2017). Furthermore,both approaches were explored in different genetic control strategies against insect pests,e.g.,CRISPR-based homing gene drives,which work through HDR (Gantz and Bier 2015),or CRISPR-based sex-distortions (CRISPRSD),which rely on NHEJ-based gene disruption (Galiziet al.2016).

Substantial progress also has been achieved on the genomics end. Through completed and ongoing insect genome projects as well as databases such as NCBI BioProjects,i5K,InsectBase,and IAS1000,high-quality genomes and transcriptomes of many insect species were assembled and made available to researchers (Poelchauet al.2015;Yinet al.2016;Qian and Wan 2018;Liet al.2019). These resources greatly facilitate the research of genetic networks underlying the reproduction,sex determination,and development in different insect species,which provides crucial knowledge for developing a particular genetic control strategy for pest insects. The availability of genomic data and new findings in genetic research boosts the enthusiasm to use these novel molecular tools and new target sites/genes for genetic control,leading to many promising new approaches.Here,we review state-of-art strategies for genetic control of pest insects using CRISPR-based systems. Recent findings,resistance mechanisms,regulatory aspects,and possible future developments of those strategies are discussed.

2.CRISPR-based genetic control strategies

To overcome potential disadvantages of the SIT,such as reduced efficiency caused by bi-sexual releases (in cases the sex separation at mass-rearing scale is not available)or the necessity of repeatedly releasing large amounts of insects (due to the non-optimal control agents),CRISPRbased systems have been developed to introduce genetic modifications and generate genetic strains for insect control (Table 1). Those strategies have mainly been evaluated under lab conditions,but could be promising for field applications in the future.

2.1.CRISPR-based gene drive

Gene drive for pest control works by introgressing desirable gene(s) into natural populations based on super-Mendelian inheritance of the gene(s) using selfish genetic element(s) (Curtis 1968;Sinkins and Gould 2006;Champeret al.2016). The gene drive strategy can be used for population suppression,by introducing an effector gene that causes lethality or sex bias leading to a reduction of offspring in each generation,or for population modification,by spreading a genetic variant that eliminates a harmful trait but sustains the population of the target species (Champeret al.2016;Rabanet al.2020;Devoset al.2022). The selfish genetic elements that have been adopted or proposed for engineering gene drive systems include naturally-occurring elements such as transposons (Ribeiro and Kidwell 1994;Sinkins and Gould 2006;Bier 2022),a region of nuclear DNA found only in some flour beetles calledMedea(maternal effect dominant embryonic arrest) (Beemanet al.1992;Wade and Beeman 1994;Beeman and Friesen 1999),or certain intracellular bacteria likeWolbachia(Turelli and Hoffmann 1991;Sinkins and Godfray 2004;Wanget al.2022). In addition,synthetic systems such as artificialMedeasystems,engineered underdominance or homing endonuclease genes (HEGs) have been developed to drive genes into a population (Sinkins and Gould 2006;Champeret al.2016;Rabanet al.2020;Hayet al.2021).

The artificialMedeasystems demonstrated inDrosophilaspecies contain a maternal microRNA (miRNA)(the “toxin”) inMedea-bearing females targeting an essential embryonic gene (e.g.,myd88) in all embryos,and a recoded target gene (the “antidote”) that is not silenced by the miRNA. The two elements are tightly linked and always inherited together (Chenet al.2007;Akbariet al.2014;Buchmanet al.2018). The recoded target gene is expressed at the zygotic stage thus rescues the embryos that inherit theMedeasystem. Due to the killing effect on non-Medea-bearing progeny,the frequency of Medeabearing progeny within the population would increase.Therefore,a cargo gene linked to theMedeasystem can spread in a population (Wardet al.2011;Akbariet al.2014). In addition to the RNAi-mediated silencing of essential genes,CRISPR-based gene targeting was used to engineer the “Toxin-Antidote” (Champeret al.2020a,2021),or the “Cleave and Rescue” systems (Oberhoferet al.2019,2020,2021). These strategies can produce a lethal effect by targeting an essential gene using the CRISPR system,and the rescue is achieved by providing a recoded version of the essential gene,resistant to cleavage.

The engineered underdominance drive is based on an underdominant trait that leads to lower fitness of heterozygotes (or their progeny) compared to that of the parental homozygotes. Therefore,if a transgene is linked to such an underdominant trait,it will spread in the population (Magori and Gould 2006;Altrocket al.2010;Akbariet al.2013). Underdominance systems need high initial release levels for the drive to spread and therefore can be used to locally restrict the population manipulation(Champeret al.2016;Hayet al.2021;Bier 2022). The underdominance systems include reciprocal chromosomal translocations,single-and two-locus toxin-antidote,as well as doubleMedeasystems (Champeret al.2016,2020b;Hayet al.2021). A reciprocal chromosomal translocation eliminates translocation-bearing heterozygotes which do not inherit a complementary gene set. This approach was proposed to fix desirable genes in insect pest populations (Curtis 1968;Robinson 1976). The doubleMedeauses two RNAi-basedMedeatoxin-antidote sets and each antidote is tightly linked to the toxin for the otherMedea(Akbariet al.2013).Therefore,progeny must inherit both constructs to survive,whereas the ones inheriting no or only single antidote construct die,mimicking an underdominance situation. For underdominance drives relying on toxin and antidote,CRISPR-based gene targeting can be used to engineer the toxin cassette as described before. Many of the above CRISPR-based gene drive (CRISPR-GD)strategies were shown to efficiently introgress a gene or a genetic system into aDrosophilapopulation. For more details on these examples and to obtain a more complete picture on existing systems,we refer the interested reader to in-depth reviews on this topic (Champeret al.2016;Rabanet al.2020;Hayet al.2021;Bier 2022;Devoset al.2022;Verkuijlet al.2022).

The HEGs gene drive system incorporates an endonuclease gene into the genome,whose product then cleaves at a target site on the homologous chromosome opposite to the position of the HEG,allowing the HEG to be copied to the homologous chromosomeviaHDR(Gimble 2000;Sinkins and Gould 2006). This process was termed “homing” and results in more HEG-containing gametes than those without HEG and,therefore,a super-Mendelian inheritance (Sinkins and Gould 2006;Champeret al.2016;Hayet al.2021). The development of such homing-based gene drives is greatly facilitated by the CRISPR system,which allows the construction of HEG-like Cas9-gRNA gene cassettes. Such a cassette can be inserted into the targeted locus and is then further copied to the second allele by HDR,converting a heterozygous to a homozygous genotype (Gantzet al.2015;Gantz and Bier 2015). A proof-of-principle study inD.melanogastershowed efficient HDR-directed genomic insertion of a Cas9-gRNA cassette from the chromosome of origin to the homologous chromosome (Gantz and Bier 2015). Different efforts have been made to translate this technology into a pest management tool. For example,female fertility genes can be targeted (Fig.1-B).This was used for the control of the malaria mosquitoAnophelesgambiaeby targeting three ovary-enriched genes,includingAgAP005958(ortholog ofDrosophila yellow-g),AgAP007280(ortholog ofDrosophilanudel),andAgAP011377,as well as the sex determination genedoublesex(Agdsx) (Hammondet al.2016;Kyrouet al.2018). In these studies,the cage populations collapsed,because theAgAP005958,AgAP011377,andAgdsxhomozygous mutant females did not produce eggs,whereasAgAP007280mutants laid eggs that did not hatch. In another mosquito species,An.stephensi,an autosomal genekynurenine3-hydroxylase(Ask3h)involved in the tryptophan metabolism pathway,was targeted using the homing CRISPR-GD (Gantzet al.2015). The homozygous mutant females showed reduced survival and reproduction. Thus,population elimination was achieved when the CRISPR-GD and WT mosquitos were combined at a 1:1 ratio in cage experiments (Phamet al.2019).

Similarly,inD.melanogaster,theyellow-ganddeformedgenes were targeted for female sterility and embryonic lethality,respectively (Oberhoferet al.2018).However,the homing was not efficient due to the high fitness load in heterozygous mutants conferred by the activity of Cas9 in the female germline. Specifically,most of the flies that inherited the Cas9/gRNA complex from the mother were either sterile or died,suggesting that a male germline-specific promoter might be preferred to drive the Cas9 expression,so the carryover of Cas9 from sperm into the zygote is minimal (Oberhoferet al.2018).Indeed,limiting the Cas9 expression to the male germline effectively suppressed the drive conversion in the maternal germline when targeting an autosomal and eye-marker genecinnabar(Champeret al.2018). Besides genes that directly confer lethality or sterility,sex determination genes such astransformer(tra) were targeted,whereupon the females developed as intersex and became sterile(Carramiet al.2018). Further simulations showed that the fertility of heterozygous females carrying atramutation might play a significant role in the suppression of insect pests such as the Mediterranean fruit flyCeratitis capitata,in which the sex-converted individuals (XX karyotype) are fertile and contribute to the drive efficiency(Paneet al.2002;Carramiet al.2018;Meccarielloet al.2019). Also,simulations in the red flour beetleTribolium castaneumshowed that targetingangiotensin-converting enzyme2(TcAce2)andTc010993,which are involved in female and male fertility,respectively,could suppress the population effectively (Druryet al.2017). A split homing CRISPR-GD,in which the Cas9 and gRNAs components are introduced separately,was demonstrated in a Lepidopteran species,the diamondback mothPlutellaxylostella(Xuet al.2022). By targeting the pigmentation genesk3handyellow(y),heritable Cas9-mediated germline cleavage was observed. However,the driving efficiency was low,possibly due to the insufficient expression of Cas9 in early meiosis (Xuet al.2022). Nevertheless,this study pointed to the feasibility of homing CRISPR-GD in Lepidopteran species,which could be improved by using germline-specific regulatory elements to control the Cas9 expression.

2.2.CRISPR-Cas9 sex-ratio distortion (CRISPRSRD)

To manipulate the insect sex ratio for pest control,a synthetic sex-ratio distortion system was previously engineered inAn.gambiaeby ectopic expression of a homing endonuclease,I-PpoI,identified initially from the slime mouldPhysarumpolycephalum(Flicket al.1998;Galiziet al.2014).I-PpoIis encoded by a mobile and self-splicing intron located in ribosomal DNA (rDNA),and it functions by recognizing and cleaving DNA of the corresponding intron-lacking allele by inducing a sitespecific double-stranded break (Belfort and Perlman 1995;Flicket al.1998). Notably,I-PpoIwas also found to precisely cut a conserved rDNA sequence located on the X chromosome inAn.gambiae(Windbichleret al.2007). A variant ofI-PpoIwith optimized activity was expressed inAn.gambiaeduring male meiosis to specifically cleave such X-linked rDNA (Galiziet al.2014).This process,termed X-shredding,resulted in mostly Y chromosome-bearing sperm and produced up to 95%male offspring when engineered males mated with WT females (Galiziet al.2014). To mimic theI-PpoI-mediated X-shredding,a CRISPRSRDsystem was developed also in theAn.gambiaeby using aβ2-tubulin(β2t) promoter for the spermatogenesis-specific expression of a Cas9 endonuclease and gRNA cassette targeting rDNA clusters on the X chromosome,resulting in a strong sexratio distortion similar to theI-PpoIstrategy (Galiziet al.2016) (Fig.1-C). For Lepidopteran pests that have a ZW(females) and ZZ (males) sex chromosome system (Trautet al.2007),‘W-shredding’ was proposed,which would suppress populations by cleaving the W chromosome and thus cause sex distortion towards males (Holman 2019).

Since the rDNA genes are not limited to a single cluster on the X chromosome in most insects,developing a CRISPRSRDstrategy in new target species would rely on searching sequence repeats exclusively located on the X chromosome (Papathanos and Windbichler 2018). A bioinformatic pipeline has been designed to identify short and highly abundant sequence elements that meet the requirements of the CRISPRSRDSystem (Papathanos and Windbichler 2018). In addition,a detailed workflow starting from genomic data to the generation of CRISPRSRDstrains was illustrated and can be considered for any XX/XY heterogametic species (Tsoumaniet al.2020). Based on these methods,CRISPRSRDwas adopted inD.melanogasterandC.capitata,and for the latter,up to 80% male bias was achieved (Fasuloet al.2020;Meccarielloet al.2021). InD.melanogaster,two distinct models for sex distortion were created. In addition to X-shredding,which led to 68% male bias,an‘X-poisoning’ approach was designed by targeting an X-linked and putatively haplolethal gene,Rps6(Fasuloet al.2020).RpS6is a ribosomal protein gene identified from a sex-linked haplolethal locus since females die due to a heterozygous deficiency (Lefevre and Johnson 1973;Stewart and Denell 1993). Fasuloet al.(2020)achieved a knock-out ofRpS6by CRISPR that killed most female offspring at early developmental stages and resulted in a strong male bias (＞92%). Interestingly,most female survivors carried mutations that are not expected to restore theRpS6function (Fasuloet al.2020).Therefore,more insights into the lethal mechanism ofRpS6or a better haplolethal gene target may be needed to achieve a male bias efficiency desired for practical applications. Notably,X-poisoning acts at the postzygotic stage since female lethality is caused by an insufficient dose of theRpS6product. In contrast,the X-shredding acts prezygotically (already during spermatogenesis)(Fasuloet al.2020). Consequently,half of the zygotes are expected to die in the X-poisoning system,while most zygotes should survive and develop into males in the X-shredding system. Such differences in the population size of the next generation,together with the sex-bias efficiencies,would have an impact on the overall male output and should be evaluated for the male production of any CRISPRSRDstrategy. Modeling suggests that the X-poisoning approach can be more efficient than several other self-limiting strategies by placing the gene-editing cassette on the Y chromosome,generating so-called‘Y-linked genome editors’ (YLEs) (Burt and Deredec 2018). Thus,the YLEs would solely be found in males but confer the harmful effect(s) only to female descendants.The selection against YLEs is expected to be absent or weak if the transgene itself confers no or only low fitness costs. Therefore,the system could persist longer than some other self-limiting strategies after the release of engineered insects is stopped (Burt and Deredec 2018).

2.3.CRISPR-engineered genetic sexing strains(CRISPR-engineered GSS)

The use of GSS (such as VIENNA-7 and VIENNA-8 for medfly) in SIT programs is a classic example of successful area-wide integrated pest management(Rendonet al.2004;Franzet al.2021). These medfly‘classic GSS’ were developed using a naturally occurring mutation,mutagenesis,and classical genetic approaches.They are based on two recessive phenotypic markers,whitepupae(wp) andtemperature-sensitivelethal(tsl),located on chromosome 5 of the medfly. To achieve sex-specific phenotypes,the corresponding WT alleles were translocated onto the Y chromosome by induction of chromosome breaks using irradiation (T(Y;A)). GSS females are homozygous for both traits,have white puparia and are sensitive to elevated temperatures. In contrast,heterozygous males have brown puparia and can survive elevated temperatures due to functional WT alleles on the Y chromosome (Franzet al.2021).Importantly,meiotic recombination could transfer thewpand/ortslWT allele back to the autosome,thus leading to the breakdown of the sexing mechanism. The frequency of such recombination depends on the positions of the T(Y;A) breakpoints,the sexing genes,and their relative distance (Franzet al.1994).tslmutations are often generated by EMS mutagenesis,a random process that is time and labor-consuming. The same is true for placing the WT allele on the Y chromosome using radiation to induce random chromosomal translocations. However,if the genes and the mutations responsible for the desired phenotypic traits are known,the CRISPR system could provide a more convenient,flexible,and transferable method to engineer ‘neo-classical’ GSS (CRISPRengineered GSS) (Fig.1-D). CRISPR-based genome editing approaches can be used to: i) precisely create the desired mutations (likewportsl) in the respective genes of different pest species (H?ckeret al.2021;Wardet al.2021);and ii) link the rescue allele to the male sex. The sex linkage could be achieved either by an HDR-mediated knock-in of the respective WT allele at a suitable position on the Y chromosome (thus creating a third copy of the gene) or by inducing CRISPR-mediated DSBs in targeted positions on the Y chromosome and the autosome carrying the marker gene. The concurrent induction of two DSBs can lead to a reciprocal translocation,and the WT allele of the marker gene could be translocated to the Y chromosome (Chenet al.2015;Brunet and Jasin 2018).

Recent progress towards the development of marker-based CRISPR-engineered GSS includes the identification of the genes responsible for two valuable sexing markers,whitepupae(Wardet al.2021),and red eye (cardinal) (Chenet al.2022). Previously,bothwpandcardinalphenotypes (based on naturally found or chemically induced mutations) have been used to create GSS (Franzet al.2021;Koskiniotiet al.2021).Furthermore,severalD.melanogastergenes with verified recessive or dominanttemperature-sensitive(ts) phenotypes,includingshibire,Notch,RpII215,ple,andDsor1,have been proposed as potential targets to engineertsl-based CRISPR-engineered GSS in other species (Nguyenet al.2021). For insects with an XX/XY sex chromosomal system,such as some Tephritid,Drosophila,and mosquito pests,recessive mutations would be needed since males are heterogametic. In contrast,for Lepidopteran pests in which females are heterogametic,a dominanttslmutation (trans)located to the W chromosome would be needed to engineer a GSS(Marecet al.2005;Trautet al.2007;Marec and Vreysen 2019). Thus,it has been proposed to place one of theNotchtsl-mutant alleles,N60g11,together with a marker gene (EGFP),onto the W chromosome of the codling mothCydiapomonellausing thepiggyBactransposon system. In such a scheme,under permissive temperature conditions (forN60g11),the progeny would consist of WT males (ZZ) and transgenic females (WZ) showing green fluorescence,while switching to restrictive conditions(e.g.,15°C forN60g11) would then kill all female progeny at the embryonic stage and produce only WT males for release (Marecet al.2005). However,sincepiggyBactransposon mediates random transgene insertion and shows low transformation efficiency inC.pomonella,this approach faces the challenge of unpredictable transgene integration into the genome (Marecet al.2005;Marec and Vreysen 2019). More recently,the gene-editing tool transcription activator-like effector nucleases (TALENs)was successfully used to insert theDsRed2gene into the W chromosome of the silkworm,Bombyxmori,for the generation of a GSS carrying a female-specific marker(Zhanget al.2018). Since CRISPR-mediated knock-in has been achieved in different Lepidopteran species (Li J Jet al.2021),it should be possible to placetslmutation(s)or marker gene(s) into the W chromosome using this technique. In addition,precise and efficient knock-in by CRISPR requires the availability of genomic resources(especially high-quality sequences of sex chromosomes)in different pests and pest strains. The recently reported genomes,such as the high-quality one fromC.pomonella(Wanet al.2019) and those from 114 geographical populations of the diamondback moth,Plutellaxylostella(Youet al.2020),would greatly facilitate such geneediting systems in Lepidopteran pests.

Besides introducingtsmutations in essential genes to engineertsl-based neo-classical GSS,tsmutations can also be introduced in genes involved in the sex determination pathway to conditionally transform female embryos into phenotypic male adults (engineered sex conversion). Such modifications were,for example,createdviaCRISPR-HDR in thetransformer2(tra2) gene,which co-operates withtraand determines the sexual development in the spotted-wingDrosophila,Drosophila suzukii,andC.capitata(Li and Handler 2017;Aumannet al.2020). The CRISPR-introducedtra2ts2mutation is a proline to serine substitution,a knowntsallele fromD.melanogaster(Amreinet al.1990),and exposure ofD.suzukiiandC.capitatatra2ts2flies to the restricted temperatures transformed females into males or intersex.InD.suzukii,tra2ts2flies maintained at 20°C and switched to 26°C produced sterile XX pseudo-males and XY males with compromised viability (Li and Handler 2017). ForC.capitata,it was not possible to identify a permissive temperature and establish a homozygous mutant line due to the severely limited fertility of the WTEgypt-IIand the mutant flies at reduced temperatures (below 18.5°C) (Aumannet al.2020). ForD.melanogaster,the permissive temperature fortra2ts2flies was 16°C(Amreinet al.1990). Therefore,the general fitness of the WT and the performance of thetsmutant insects under certain temperature conditions would be critical for the development of anyts-based CRISPR-engineered GSS or sex conversion system.

2.4.Precision guided SIT (pgSIT)

To develop genetic sexing systems for SIT application in Lepidopteran pest species,a female-specific and embryonic lethal system was demonstrated in the silkwormB.mori(Zhanget al.2018). It consists of an‘a(chǎn)ctivator’ strain containing ananos(nos)-Cas9cassette and aDsRedmarker on the W chromosome produced by TALENs,and an ‘effector’ strain carrying a gRNAexpressing cassette that targets an sex determination geneBmtra2(Zhanget al.2018). Although the molecular mechanism is not yet clear,Bmtra2was found to be essential for the embryonic development of females.Crossing the activator (nos-Cas9) to the effector line(U6-gRNABmtra2) interruptedBmtra2expression and killed all females during embryogenesis (Xuet al.2017;Zhanget al.2018). Furthermore,to generate sterileP.xylostellathat can be released without radiation,a sexdetermination geneP-elementsomaticinhibitor(PSI)was identified,which regulates the splicing of a malespecific isoform ofdsx(Wanget al.2021). Knock-out of this gene using the CRISPR/Cas9 system resulted in complete male sterility since mating WT females with mutant males produced eggs that did not hatch. More recently,aPMFBP1(polyaminemodulatedfactor1 bindingprotein1) gene was identified inB.moriwhich is essential for spermatogenesis. Crossing theB.morinos-Cas9line with a U6-sgRNA line that targetingBmPMFBP1generated sterile males whereases the fertility of mutant females was comparable to WT females (Yanget al.2022). Further analysis indicated thatBmPMFBP1deficiency showed no impact on the anucleated unfertile apyrene sperm but affected the migration of nucleated fertile eupyrene sperm,leading to male infertility (Yanget al.2022). Therefore,bothPSIandPMFBP1were proposed as potential targets for inducing male sterility for the genetic control of Lepidopteran pests (Wanget al.2021;Yanget al.2022).

Such Cas9 ‘a(chǎn)ctivator’ and gRNA ‘effector’ scheme for genetic control was also demonstrated inD.melanogasterandAedesaegyptiand named ‘precision-guided SIT’ (pgSIT) (Kandulet al.2019;Li Met al.2021).This approach is designed to kill 100% of females and sterilize 100% of males by knocking out genes essential for female viability and male fertility using the CRISPR system (Fig.1-E). ForD.melanogaster,the sex determination geneSex-lethal(Sxl) andβ2-tubulin(β2t) were chosen as target genes. SinceDmSxlalso controls the X chromosome dosage compensation,loss-of-function mutations kill all females due to X chromosome hyperactivation (Cline 1978,1993). On the other hand,theβ2tgene is involved in the function of sperm microtubules,and loss-of-function mutations produce immotile sperm and consequently cause male sterility (Kemphueset al.1982;Fackenthalet al.1995).In the pgSIT strategy forD.melanogaster,the transgenic line expressing gRNAs that target bothSxlandβ2tand the transgenic line expressing Cas9 in the germline are maintained separately,and the crossing between the gRNA and Cas9 lines eliminated all females and produced only sterile males that were mating-competent(Kandulet al.2019). ForAe.aegypti,myosinheavychain(myo-fem),which controls flight muscle development in females,was targeted to generate flightless females described as unlikely to mate or transmit diseases (Li Met al.2021). Moreover,knocking-outAaβ2tproduced sterile males with comparable fitness to WT males. Cage studies with different releasing ratios (e.g.,pgSIT:WT mosquitoes at 20:1 or 40:1) showed that the populations can be eliminated within three generations (Li Met al.2021). Therefore,pgSIT could serve as a straightforward and robust control strategy that can be transferred to other species.

3.What’s new and what’s yet to come?

3.1.Resistance development

For control strategies based on insect strains carrying transgene constructs including gene drive,CRISPRSRD,and pgSIT,spontaneous mutations in the inserted regulatory or coding sequence such as promoters,Cas9,or gRNA can occur and abolish the transgenic system’s function,thus leading to resistance development against the control system in the population. Spontaneous mutations sufficient to shut off a promoter are estimated to occur at a rate of～10-7or less,while higher rates(1-5×10-6) of missense or nonsense mutations silencing the transgene are expected in a～1 kb coding sequence(Handler 2016). InD.melanogaster,F1survivors from parents carrying a Tetracycline-off (Tet-off) Embryonic Lethal System occur at a 5.8×10-6frequency under restrictive conditions due to deletions and InDels in the coding sequence of the two functional transgenestTA(1 kb) andhidAla5(1.2 kb) (Zhaoet al.2020). The commonly usedCas9variant fromStreptococcus pyogenes(SpCas9) is 4.2 kb (Hsuet al.2014). Therefore,the mutation rate inSpCas9can be estimated to be 2.4×10-5. Several new variants ofCas9were reported,such as the codon-optimizedCas9fromSaCas9andStreptococcusthermophilus(St1Cas9),which are 3.2 and 3.4 kb,respectively (Murovecet al.2017). Those variants could contribute to a smaller vector size and reduce the size-dependent mutation rate. Most ongoing SIT programs release 107-108insects per week for efficient population suppression. The number of individuals that escape control due to certain resistance mechanism(s)would depend on the operational scale of the insect population,which is further subject to the efficacy of the control program. Therefore,the negative effect of the potential ‘escapers’ due to the spontaneous mutations should be considered for the modeling,application,and monitoring of CRISPR-based control strategies (Gould and Schliekelman 2004;Robinsonet al.2009;Alpheyet al.2011;WHO 2021).

In addition to transgene-inactivating mutations(primary site mutations),Zhaoet al.(2020) identified similarly frequent resistance to the engineered lethality at unknown,so-called secondary sites inD.melanogaster.This effect was inherited maternally (Zhaoet al.2020).Besides spontaneous primary and secondary site mutations,pre-existing resistance alleles were reported in wildD.melanogasterpopulations that could invalidate the Tet-off controlled lethality inD.melanogaster(Knudsenet al.2020). Survivors from primary or secondary site or pre-existing resistances could build a population that may be resistant to further control by the same or a similar lethality system. The existence or development of such resistances may be subject to species-specific mutation rates,allelic diversity of populations,and the sensitivity of the control system components to mutational inactivation.Therefore,those elements should be considered and tested for each novel control system in a target species,taking into account also the genetic background in the release area.

Resistance mechanisms in different gene drive systems and their impact on the driving efficacy depend on the molecular architecture of the drive but can also be subject to the behavior and life history of a target species.For an in-depth summary of the current knowledge we refer the interested reader to some recent articles(Champeret al.2016;Priceet al.2020;Hayet al.2021;Bier 2022;Devoset al.2022;Verkuijlet al.2022). For homing CRISPR-GD,for example,the control efficacy is mainly subject to the NHEJ-induced resistance alleles,e.g.,in-frame mutations that remove the gRNA target site but preserve the gene function or genetic variability within and among targeted populations (Druryet al.2017;Priceet al.2020). The resistance due to alleles that the gRNA can no longer target could evolve rapidly when only a single site is targeted (Hammondet al.2017;Carramiet al.2018). It was demonstrated experimentally that multiplexing of gRNAs could reduce the formation of resistance alleles and increase drive efficiency (Champeret al.2018).

For CRISPRSRDX-shredder approaches,the target sites (e.g.,rDNA sequences) are X-chromosome-specific and highly repetitive (Galiziet al.2016;Meccarielloet al.2021). Therefore,resistance due to target site mutation is considered unlikely,although the possibility cannot be completely ruled out (Priceet al.2020). In addition,the functionality and efficacy of the potential CRISPRSRDtarget sites identified using bioinformatic approaches(Papathanos and Windbichler 2018;Tsoumaniet al.2020)may depend on the unknown genetic variation within the wild populations. Different copy numbers of the target site in wild populations could lead to a reduced cutting of the X chromosome and thereby not lead to an efficient shredding effect,thus serving as a pre-existing resistance that can be identified by sequencing the target population.The recently demonstrated ‘X-poisoning’ can serve as a redundant system in addition to X-shredding to further reduce the chance of resistance (Fasuloet al.2020).

As for traditional GSS approaches,the major concern for CRISPR-engineered GSS is meiotic recombination that could revert the translocation of the WT marker allele by relocating it from the Y chromosome back to the autosome,thus rescuing the WT phenotype in females and leading to female survivors. Since the recombination frequency is lower if the autosomal breakpoint is closer to the location of the gene(s) that is used for sexing (Franzet al.1994;Franz 2005),generating translocations with breakpoints close to such gene(s) might help to improve the stability. In addition,irradiation-induced inversions in the genomic region can be used as recombination suppressors (Augustinoset al.2020;Koskiniotiet al.2021). If the genomic information is available,such recombination-reducing strategies,including targeted translocations or inversion,could also be realized by CRISPR-mediated approaches (Iwataet al.2016).Furthermore,the accumulation of recombinants can be prevented by a management strategy known as a filter rearing system (FRS). Here,a small population(mother colony) is maintained,in which recombinants are physically removed. This colony can be used to refresh the main insect production,thus significantly improving the strain stability and production efficiency (Aketarawonget al.2020;Franzet al.2021). If the trait used for sexing does not create a visible phenotype,like for atsl-based CRISPR-engineered GSS,the FRS could be greatly facilitated by including a visible marker,such aswp,which should be closely linked to thetsl(Franzet al.2021).

For pgSIT,the Cas9-and gRNA-expressing strains are maintained separately,i.e.,the two components of the CRISPR system remain inactive. Crosses of the two strains then produce completely sterile males,which should not pass the transgene on to the WT population (Kandulet al.2019;Li Met al.2021). Such a protocol theoretically neither allows the accumulation of suppressors during mass-rearing nor their selection after release. However,like the NHEJ-induced resistance alleles in gene drive,the scalability of pgSIT may be subject to the rate of in-frame mutations that retain the gene function for viability or fertility but prevent gRNA recognition. Therefore,targeting multiple sites or genes for each designated phenotype would safeguard pgSIT operations.

3.2.Stacked systems

For the optimal suppression effect,stability of the suppression system,and ecological safety,genetic strains with stacked and mechanistically independent systems were suggested for a control program (Handler 2016). For example,two different lethal mechanisms were introduced into a transgenic sexing strain of Australian sheep blowflyLuciliacuprina,including the Tet-off system regulatortTA,which kills insects due to “transcriptional squelching”,and a pro-apoptotic genehid,which kills insects by induction of apoptosis (Yan and Scott 2020). It was proposed that such a Tet-Off-based lethal system,and a multifactorial reproductive sterility system,based on the CRISPR/Cas9,can be combined to reduce resistance development in a genetic control program (Eckermannet al.2014).A novel genetic control technology was developed inD.melanogasterthat combines a female-killing system based on thetTA-induced lethality and an incompatible male system based on the lethal overexpression of endogenous genes induced by a dCas9-based programmable transcription activator (Upadhyayet al.2022). InAn.gambiae,the CRISPR-GD and CRISPRSRDwere stacked by using a Cas9-gRNA construct that targetsAgdsxand expresses the site-specific nucleaseI-PpoIfor X-shredding. Consequently,female sterility and male-biased inheritance were achieved simultaneously,leading to a faster drive and more efficient population suppression than a bi-sexual drive (Simoniet al.2020).

3.3.Conditional systems

For many CRISPR-based strategies,maintaining the modified insect population is challenging due to difficulties controlling the strains’ lethality,sterility,or sex sorting effect. For example,pgSIT requires Cas9 and gRNA lines to be kept separately and crossed to produce sterile males for release. Such a scheme is difficult to be implemented in an operational control program if a sexing procedure for large populations is not available.Therefore,conditional expression systems that allow efficient induction and suppression of the functional genes can facilitate the development and implementation of control strategies that are designed to suppress a population. Currently,the xylose-,light-,rapamycin-,doxycycline-,4-hydroxytamoxifen-,and small moleculeinducible CRISPR systems have been explored mostly in cell cultures (Richteret al.2017;Millset al.2020).Conditional Cas9-expression has been achievedinvivoinD.melanogasterusing the UAS/Gal4 system (Meltzeret al.2019;Portet al.2020). Since a third transgene(Gal80) is needed to inhibit Cas9 expression in the UAS/Gal4 system,it is not a practical method to develop genetic control strategies.

Temperature-inducible gene targeting based on theinvivoinduction of the CRISPR system inD.suzukiiwas demonstrated by the expression of the Cas9 gene under the regulation of theD.melanogasterheatshock protein70(Dmhsp70) promoter and its 3′UTR (Yanet al.2021). While basal Cas9 expression was detected in the embryos of transgenicD.suzukiiat 25°C,applying a heatshock treatment by a 1-h exposure to 37°C considerably increased the Cas9 expression and boosted the somatic and germline mutagenesis rates of an exogenousEGFPgene (Yanet al.2021). TheDmhsp70-Cas9 approach was also utilized for pgSIT,whereby the necessary sexseparation step in the original version could be omitted(Kandulet al.2021a). Specifically,the Dmhsp70-Cas9 line and gRNA line targeting bothtraandβ2twere crossed and bred to double-homozygosity at 18°C.Rearing this strain at 26°C with heat-shock treatment(1 or 2 h exposure to 37°C) during early larval stages eliminated all females/intersexes and sterilized all males.In contrast,the Dmhsp70-Cas9 and gRNA lines targeting bothSxlandβ2tcan only be maintained as doubleheterozygous strain at 18°C,and applying similar heatshock treatments to this strain also produced only sterile males. A leaky baseline expression of Cas9 conferred by theDmhsp70promoter in somatic cells was observed and thought to be responsible for the sterile phenotype at 18°C of double homozygous females that express gRNAs targetingSxlandβ2t(Kandulet al.2021a).

Other molecular conditional methods that can be considered for insect species include the Tet-off and Teton,Erythromycin-off,Biotin-on,Vanillic-acid regulated,Phloretin-off,Bile acid-off,and the Quinic-acid systems(Venkenet al.2011;Eckermannet al.2014;Jaffriet al.2021;Leeset al.2022). Some binary transactivational systems derived from bacteria,such as cymTA,pipTA,ttgTA,and vanTA systems,which are regulated by the small molecules cumate (C10H11O2),virginiamycin M1,phloretin,and vanillic acid,respectively,were evaluated inD.melanogaster(Gamezet al.2021). Tissue-specific transactivation of reporter gene expression was observed when employing those systems. However,the gene expression could not be effectively suppressed by any of these regulatory molecules (Gamezet al.2021). Up to date,the Tet-off system is the most used conditional method for the genetic pest management in insects and showed tight control for the expression of lethal effector genes in an optimal design (Harvey-Samuelet al.2017;Alphey and Bonsall 2018;Yanet al.2020;Jaffriet al.2021). It can be foreseen that development of conditional systems that effectively work with CRISPR-based control strategies will be in demand.

3.4.Regulation

Despite the potential resistance development to CRISPRGD strategies,insilicomodeling suggests that existing GD approaches are highly invasive in wild populations,even if only a small number of individuals are released or escaped due to its self-propagating mechanism (Nobleet al.2018). Therefore,caution is emphasized for the field application of this technology,and contamination of nature concerns are raised for the GD research in the laboratory (Oyeet al.2014;Akbariet al.2015;Webberet al.2015;Taninget al.2017). Thus,CRISPRGDs currently face technical,ethical,regulatory,and governance challenges (Gutzmannet al.2017;Lunshof and Birnbaum 2017;Devoset al.2022). Both CRISPRSRDand pgSIT would release transgenic insects (e.g.,carrying fluorescent markers or a Cas9 transgene) but are selflimiting because the transgenes would not persist in the wild. Strategies applying SRD can be used as sexseparation methods to produce a male-only population for a sterile release (by radiation) or for fertile releases.The latter is considered to be more effective for the control than the former because the SRD effect can be passed through a few generations in the field before it breaks down due to the dilution of the SRD transgene.Because of such difference,fertile releases of males carrying an SRD system would receive a stricter risk assessment since the transgene(s) would also be carried through limited generations in the wild. pgSIT,on the other hand,produces sterile males and thus does not pass the transgene to wild insects. It was suggested that pgSIT could be regulated similarly to previous transgenic technologies such as the Release of Insects carrying a Dominant Lethal (RIDL),which has been implemented in several countries (Li Met al.2021). Unlike others,CRISPR-engineered ‘neo-classical’ GSS of insects that contain point mutations in endogenous genes aimed to confer the sexing effect but do not carry transgenes should be handled and regulated similar to classical GSS,which have been successfully released for decades (Franz 2005,2021).

4.Conclusion

Using selfish genetic elements and manipulating sex ratios for insect pest management are not new ideas(Craiget al.1960;Hickey and Craig 1966;Curtis 1968;Kniplinget al.1968). However,before modern molecular biology,exploration of these ideas primarily relied on naturally occurring genetic mutations or elements. The employment of molecular genetic engineering and insect transformation in multiple species has overcome many previously intractable problems using traditional methods.In addition,CRISPR technology further revolutionized the possibilities to perform precise,efficient,flexible,and transferrable gene or genome edits that allow researchers to translate ideas into practice faster than ever before.At the same time,the recent advances in sequencing technologies and bioinformatics promote the assembling of high-quality genomes and transcriptomes,which greatly facilitate the design and development of genetic control strategies. The possible application of a certain approach,such as the herein introduced CRISPR-GD,CRISPRSRD,CRISPR-engineered GSS,and pgSIT,would depend on the efficacy and the regulations applying for each specific case,which is further subject to its working mechanism,potential risks,and the different regulations that currently apply to identical genetic systems worldwide. With the rising concerns about global food security and public health,it can be foreseen that genetic control will remain a fast-evolving research area that battles agricultural pests and human disease vectors,and evaluations of every single system should be considered for transnational,science-based regulations.

Acknowledgements

This work was funded by the Deutsche Forschungsgemeinschaft (DFG,German Research Foundation)within project numbers 470105316/YA 502/3-1 (to Ying Yan) and SCHE 1833/7-1 (to Marc F.Schetelig).

Declaration of competing interest

The authors declare that they have no conflict of interest.