Wolbachia pipientis Associated With Tephritid Fruit Fly Pests: From Basic Research to Applications

Members of the true fruit flies (family Tephritidae) are among the most serious agricultural pests worldwide, whose control and management demands large and costly international efforts. The need for cost-effective and environmentally friendly integrated pest management (IPM) has led to the development and implementation of autocidal control strategies. These approaches include the widely used sterile insect technique and the incompatible insect technique (IIT). IIT relies on maternally transmitted bacteria (namely Wolbachia) to cause a conditional sterility in crosses between released mass-reared Wolbachia-infected males and wild females, which are either uninfected or infected with a different Wolbachia strain (i.e., cytoplasmic incompatibility; CI). Herein, we review the current state of knowledge on Wolbachia-tephritid interactions including infection prevalence in wild populations, phenotypic consequences, and their impact on life history traits. Numerous pest tephritid species are reported to harbor Wolbachia infections, with a subset exhibiting high prevalence. The phenotypic effects of Wolbachia have been assessed in very few tephritid species, due in part to the difficulty of manipulating Wolbachia infection (removal or transinfection). Based on recent methodological advances (high-throughput DNA sequencing) and breakthroughs concerning the mechanistic basis of CI, we suggest research avenues that could accelerate generation of necessary knowledge for the potential use of Wolbachia-based IIT in area-wide integrated pest management (AW-IPM) strategies for the population control of tephritid pests.


INTRODUCTION The Economic Importance and Management of Tephritid Pest Species
Flies in the family Tephritidae (Diptera) include some of the world's most important agricultural pests. The family is comprised of ∼4,900 described species within 481 genera, of which six (Anastrepha, Bactrocera, Ceratitis, Dacus, Rhagoletis, and Zeugodacus) contain ∼70 major pest species (White and Elson-Harris, 1992;Norrbom, 2004aNorrbom, ,b, 2010Mengual et al., 2017). Pest tephritids represent an enormous economic cost because they cause direct losses to a diversity of crops (fruits, vegetables, and flowers) (White and Elson-Harris, 1992). Furthermore, they hamper the development of agriculture in numerous countries, due to the strict quarantines imposed by countries importing affected crops, and to the huge costs associated with efforts aimed at prevention, containment, suppression, and eradication.
To prevent or minimize the harmful effects of tephritid pests, growers must comply with health and safety standards required by the market, applying an area-wide management approach involving chemical, biological, cultural, and autocidal control practices (Reyes et al., 2000;Enkerlin, 2005). Autocidal refers to methods that use the insect to control itself, by releasing insects that are sterile or induce sterility upon mating with wild insects in the next or subsequent generations (Black et al., 2011;Leftwich et al., 2014;Handler, 2016). Autocidal strategies include the sterile insect technique (SIT) (Knipling, 1955;Hendrichs and Robinson, 2009); one of the most widespread control methods used against fruit flies (reviewed in Dias et al., 2018). SIT relies on the mass-rearing production, sterilization and recurrent release of insects (preferentially males) of the targeted species. Sterilization is typically attained by radiation (Bakri et al., 2005), in a way that does not impair male mating and insemination capabilities. Wild females that mate with sterilized males lay unfertilized eggs. At the appropriate sterile:wild (S:W) ratio, the reproductive potential of the target population can be reduced (Knipling, 1955;Klassen and Curtis, 2005;Cáceres et al., 2007). Historically, at least 28 countries have used the SIT at a large-scale for the suppression or eradication of pests (Hendrichs et al., 1995Suckling et al., 2016). SIT has been applied successfully for several non-tephritid insect pests including the New World screw worm Cochliomyia hominivorax (Coquerel), several species of tsetse fly (Glossina spp.), the codling moth Cydia pomonella (L.) (reviewed in Robinson, 2002b;Dyck et al., 2005;Bourtzis and Robinson, 2006), and mosquitoes (Benedict and Robinson, 2003;Lees et al., 2015). Successful SIT programs as part of Area-wide Integrated Pest Management (AW-IPM) strategies have also been implemented for several tephritids: Ceratitis capitata (Wiedemann); Anastrepha ludens (Loew); Anastrepha obliqua (Macquart); Zeugodacus cucurbitae (Coquillett); Bactrocera dorsalis Hendel; and Bactrocera tryoni (Froggatt) (Enkerlin, 2005;Hendrichs et al., 2005;Klassen and Curtis, 2005;Cáceres et al., 2007). SIT is currently being developed for three additional tephritid species: Anastrepha fraterculus (Wiedemann) (Cladera et al., 2014); Dacus ciliatus Loew (Rempoulakis et al., 2015) and Bactrocera tau (Walker) (Du et al., 2016). The advantages of the SIT over other pest control approaches (e.g., use of pesticides) are that it is species-specific and environmentally friendly (Lees et al., 2015;Bourtzis et al., 2016), and resistance is less likely to evolve (but see Hibino and Iwahashi, 1991;McInnis et al., 1996).
Another autocidal strategy where mating between massreared and wild insects can be used to suppress pest populations is the incompatible insect technique (IIT); coined by Boller et al. (1976). The earliest successful pilot application of IIT was in Culex mosquitoes (Laven, 1967), and interest in applying it to mosquitoes has resurged in recent years (reviewed in Ross et al., 2019b). IIT also relies on the principle of reducing female fertility, but utilizes endosymbiotic bacteria instead of radiation, to induce a context-dependent sterility in wild females. It is based on the ability of certain maternally inherited bacteria (mainly from the genus Wolbachia) to induce a form of reproductive incompatibility known as cytoplasmic incompatibility (CI; explained in the section below). Herein we review the current knowledge on taxonomic diversity of Wolbachia-tephritid associations and their phenotypic consequences, and identify gaps in knowledge and approaches in the context of potential application of IIT, alone or in combination with SIT, in AW-IPM programs to control tephritid pests.

The Influence of Wolbachia on Host Ecology
Insects and other arthropods are common hosts of maternally inherited bacteria (reviewed in Duron and Hurst, 2013). These heritable endosymbionts can have a strong influence on host ecology. Typically, such vertically transmitted bacteria are vastly (or fully) dependent on the host for survival and transmission. Certain associations are obligate for both partners, and generally involve a nutritional benefit to the host. Other heritable bacteria are facultative, with such associations ranging from mutualistic to parasitic from the host's perspective. Among these, Wolbachia is the most common and widespread facultative symbiont of insects and arthropods (Hilgenboecker et al., 2008;Zug and Hammerstein, 2012;de Oliveira et al., 2015;Weinert et al., 2015).
Wolbachia is a diverse and old genus (possibly older than 200 million years; Gerth and Bleidorn, 2016) of intracellular Gramnegative Alphaproteobacteria (within the order Rickettsiales) associated with arthropods and filarial nematodes. Wolbachia cells resemble small spheres 0.2-1.5 µm, occur in all tissue types, but tend to be more prevalent in ovaries and testes of infected hosts, and are closely associated with the female germline (reviewed by Harris et al., 2010; see also Sacchi et al., 2010). Wolbachia is estimated to infect 40-66% of insect species (Hilgenboecker et al., 2008;Zug and Hammerstein, 2012;de Oliveira et al., 2015;Weinert et al., 2015). Within a species or population, the infection prevalence of Wolbachia can be quite variable over space (e.g., Kriesner et al., 2016) and time (e.g., Hoffmann, 1991, 1995).
The most commonly documented effects of Wolbachia on arthropod hosts fall under the category of reproductive parasitism, which involves manipulation of host reproduction to enhance symbiont transmission and persistence, in general by increasing the relative frequency of Wolbachia-infected vs. uninfected females. Females are typically the sex that can transmit Wolbachia and other heritable bacteria, although rare exceptions exist (Hoffmann and Turelli, 1988;Moran and Dunbar, 2006;Chafee et al., 2010). Wolbachia employs all four types of reproductive manipulation (reviewed by Werren et al., 2008;Saridaki and Bourtzis, 2010;Schneider et al., 2011). Feminization converts genetic males into functional females, and occurs in the orders Hemiptera, Lepidoptera, and Isopoda. Wolbachiainduced parthenogenesis occurs in haplo-diploid hosts (e.g., Acari, Hymenoptera, and Thysanoptera), where unfertilized eggs, which would otherwise develop into males, develop into females. Male killing causes death of infected males to the presumed advantage of surviving infected female siblings, and occurs in Coleoptera, Diptera, Lepidoptera, and Pseudoscorpionida. Cytoplasmic incompatibility (CI) (Yen and Barr, 1971) prevents infected males from producing viable offspring upon mating with females lacking Wolbachia (or a compatible strain of Wolbachia; see below; Figure 1). CI is the most commonly reported Wolbachia-induced reproductive phenotype, and is found in Acari, Coleoptera, Diptera, Hemiptera, Hymenoptera, Isopoda, Lepidoptera, and Orthoptera.
Cytoplasmic incompatibility was discovered almost half a century ago (Yen and Barr, 1971), but its mechanism has not been fully elucidated. A useful conceptual model to understand the observed patterns of CI is "mod/resc" (Hurst, 1991;Werren, 1997). It postulates that Wolbachia has two functions: mod (for modification), which acts as a toxin or imprint of the male germline; and resc (for rescue), which acts as an antidote. The mod function acts on the nucleus in the male germline, before Wolbachia are shed from maturing sperm (Presgraves, 2000). When a sperm nucleus affected by mod enters the egg of an uninfected female, this nucleus encounters problems such as delays in DNA replication and cell-cycle progression, leading to embryo death. In contrast, if the appropriate resc ("the antidote") function is active in the egg, the defect caused by mod in the sperm is rescued, and the embryo proceeds through normal development. In the case of unidirectional CI (uni-CI), all or some of the eggs from uninfected females that are fertilized by sperm from Wolbachia-infected males (the "CI cross") fail to develop ( Figure 1A). Wolbachia-infected females are compatible with uninfected males, and thus have a reproductive advantage over their Wolbachia-uninfected counterparts. Consequently, above a certain threshold of Wolbachia infection frequency in a host population, Wolbachia frequencies can rapidly increase to a stable equilibrium frequency. When CI is strong (e.g., all embryos from the CI cross fail), fitness costs of Wolbachia are low, and maternal (vertical) transmission is high, the threshold Wolbachia frequency to achieve invasion can be close to zero, and the stable equilibrium frequency can be close to 100% (Caspari and Watson, 1959;Turelli and Hoffmann, 1999;Rasgon, 2008). Bi-directional CI (bi-CI) results from crosses involving two different (incompatible) Wolbachia strains ( Figure 1B). Crosses between females and males infected with the same or compatible Wolbachia strains are viable. Under bi-CI between two Wolbachia strains with equivalent fitness effects on a host, the infection FIGURE 1 | (A,B) Qualitative illustration of uni-and bidirectional cytoplasmic incompatibility (CI) on the basis of the Wolbachia infection status of the parent generation. Empty male and female symbols signify absence of Wolbachia. Blue and yellow ovals represent distinct Wolbachia strains. Green tick marks = Successful offspring production. Red crosses = no offspring production. (C) A special case of unidirectional incompatibility in which one Wolbachia strain (see text) can rescue another strain (i.e., the yellow one), but not vice versa.
frequency of an introduced strain must exceed 50% to achieve invasion (Rousset et al., 1991;Dobson et al., 2002). Special cases of uni-CI and bi-CI patterns can occur. For example, a strain may not induce CI, but is able to rescue the defect caused by a different strain ( Figure 1C) (Zabalou et al., 2004a).
Several recent breakthrough studies have collectively identified Wolbachia-encoded genes (of viral origin) that contribute to the induction and rescue of CI (Beckmann et al., 2017;LePage et al., 2017;Bonneau et al., 2018Bonneau et al., , 2019Lindsey et al., 2018;Shropshire et al., 2018;Beckmann et al., 2019c;Chen et al., 2019;. Wolbachia-encoded genes that rescue CI are labeled as cifA, cidA, or cindA, depending on whether they rescue a defect caused by deubiquitylase (d), nuclease (n), both (nd); "f " is used by certain authors and/or when the nature of the defect is unknown (see Beckmann et al., 2019a,b;. In CI-inducing Wolbachia strains, each of the above genes occurs upstream of a gene (its "cognate") similarly labeled, but with a "B" replacing the "A" (i.e., cifB, cidB, or cindB, respectively) that seems to function as a toxin. Certain Wolbachia strains have more than one "A-B" pair, and the combination of these is consistent with patterns of incompatibility in Drosophila (LePage et al., 2017) and Culex (Bonneau et al., 2018). Knowledge accrued to date indicates that more than one Wolbachia-encoded mechanism of CI exists, and thus, information on the genes encoded by Wolbachia genomes can help predict expected patterns of incompatibility among strains that have not been experimentally characterized.
In addition to its reproductive phenotypes on arthropods, Wolbachia engages in obligate mutualistic interactions with filarial nematodes  and with members of five insect orders (reviewed in Zug and Hammerstein, 2015). As a facultative symbiont, Wolbachia can provide direct fitness benefits to its insect hosts by influencing development, nutrition, iron metabolism, lifespan, and fecundity (Dean, 2006;Aleksandrov et al., 2007;Weeks et al., 2007;Brownlie et al., 2009;Ikeya et al., 2009;Kremer et al., 2009;Newton and Rice, 2020), and most notably, by conferring resistance or tolerance to pathogens, particularly single-stranded RNA viruses (Hedges et al., 2008;Teixeira et al., 2008;Moreira et al., 2009). The interference of Wolbachia with the replication and transmission of certain viruses, along with its ability to spread in populations via CI, form the basis of several population replacement programs (reviewed in Ross et al., 2019b;Chrostek et al., 2020). Wolbachia in Drosophila appears to confer an additional fitness benefit in the form of increased recombination (Bryant and Newton, 2019;Singh, 2019).

Methods to Assess Wolbachia Infection Status
For purposes of this review, we consider a host species or population as "infected" with Wolbachia, even if the infection is transient or found at low titer. Wolbachia, and most cytoplasmically transmitted endosymbionts, are fastidious to culture outside host cells, such that their study typically relies on culture-independent methods. A recommended flowchart of steps is depicted in Figure 2. The most utilized approach to date for identifying hosts infected with Wolbachia is through PCR screening of Wolbachia genes in DNA extracts of hosts. Different PCR primers have been used to perform such surveys, traditionally targeting a portion of the 16S ribosomal (r)RNA gene or of a ubiquitous protein-coding gene (e.g., wsp or ftsZ). Simoes et al. (2011) evaluated the relative sensitivity and specificity of different primer pairs aimed at Wolbachia detection and identification, revealing that no single PCR protocol is capable of specific detection of all known Wolbachia strains. A related method known as "loop mediated isothermal amplification" (LAMP; not shown in Figure 2), which requires less infrastructure than PCR, has been successfully employed for Wolbachia detection in several insects (da Silva Gonçalves et al., 2014).
The two major shortcomings of utilizing solely PCR (or LAMP) to detect Wolbachia presence are the occurrence of false negatives and false positives. A false negative occurs when a specimen is infected by Wolbachia, yet the screening approach fails to detect its presence. The efficiency of the PCR can be affected by the presence of inhibitors (Marcon et al., 2011;Beckmann and Fallon, 2012), by low concentration/poor quality of the target DNA molecule, as well as type and concentration of the polymerase and other PCR reagents. At the very least, negative Wolbachia detection PCRs should be validated by evaluating the quality of the DNA extract, through positive amplification of a host-encoded gene (e.g., the mitochondrial Cytochrome Oxidase subunit I or single-copy nuclear genes). Several higher sensitivity approaches have been devised, particularly for low-titer infections, such as: long PCR (Jeyaprakash and Hoy, 2000); nested PCR (Arthofer et al., 2009a); quantitative PCR (Mee et al., 2015); or the design of alternative and/or more specific primers, including the use of Wolbachia multi-copy genes as PCR targets (Schneider et al., 2014). These methods, however, have not been widely implemented, likely due to the higher effort or cost involved.
False positives occur when a specimen not harboring Wolbachia is identified as Wolbachia-infected. Several instances have been reported where insect chromosomes carry Wolbachiaderived fragments, presumably from a horizontal gene transfer event that occurred at some point in the host lineage as the result of an active infection that was subsequently lost. The size of the horizontally transmitted fragment can range from ca. 500 bp to the equivalent of an entire Wolbachia chromosome (Dunning Hotopp et al., 2007). In some cases, entire Wolbachia chromosomes have been transferred more than once onto the same host genome (Brelsfoard et al., 2014;International Glossina Genome Initiative, 2014). The range of hosts carrying Wolbachia-derived genome fragments is broad and includes several dipterans (tephritids, Glossina morsitans Westwood; Drosophila spp., mosquitoes), other insects, as well as nematodes (Fenn et al., 2006;Dunning Hotopp et al., 2007;Nikoh et al., 2008;Brelsfoard et al., 2014;Morrow et al., 2015;Attardo et al., 2019). It is therefore desirable to corroborate PCRbased inferences with approaches that detect Wolbachia cells in host tissues. Such microscopy approaches can be based on nucleic acid hybridization (e.g., Chen et al., 2005) or antibodybased detection of Wolbachia proteins (e.g., wsp; Veneti et al., 2003;and ftsZ;Newton et al., 2015). A major drawback of these methods is that they require substantial investment in time and equipment compared to PCR-based approaches. False positives can also occur if the primers targeted at Wolbachia turn out to amplify a fragment of the genome of the host (not derived from Wolbachia) or of another symbiont of the host. Such false FIGURE 2 | Recommended steps to screen for Wolbachia infections in tephritids and other arthropods. A PCR is performed with Wolbachia-specific primers on DNA isolated from whole, or parts of (e.g., abdomens), insect. Agarose gel electrophoresis of the PCR products is used to determine whether the amplicon is of the expected size. Amplicons of expected size are directly sequenced (e.g., Sanger method). High sequence identity to other Wolbachia suggests Wolbachia infection. Clean chromatograms are consistent with a single Wolbachia strain. Otherwise, a cloning step to identify different Wolbachia alleles is required. Other genes are then amplified and sequenced for further genetic characterization of the strain. As an optional step, localization of Wolbachia cells within host tissues can be achieved by Fluorescent In Situ Hybridization (FISH) with Wolbachia-specific rRNA probe or immunolabeling with antibody specific to Wolbachia protein. An amplicon of an unexpected size might indicate the occurrence of a horizontally transmitted Wolbachia genome fragment to the insect chromosome, rather than a current infection. Similarly, multiple nucleotide polymorphisms (NP) or insertions/deletions, compared to known strains, are suggestive of Wolbachia pseudogenes (e.g., horizontally transferred to host genome). This can be further tested by in situ hybridization of Wolbachia-specific probe to host chromosomes, and/or by Whole Genome Sequencing of host. Photo of fly (Anastrepha obliqua) by Fabiola Roque (ECOSUR-UT). Image from Fast et al. (2011) freely available at https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4030408/bin/NIHMS391830-supplement-Supporting_Online_Material.pdf (accessed April 01, 2019). Original sources of other photographs are available in open access journals (Doudoumis et al., 2012;Brelsfoard et al., 2014;Joubert et al., 2016).
Frontiers in Microbiology | www.frontiersin.org positives are relatively easy to rule out upon sequencing and analysis of the amplified product. Finally, as with any PCR work, false positives can result from contamination of the specimen, the DNA template, or the PCR reagents. Thus, it is important to implement adequate sterile practices and negative controls.
The above approaches require destruction of specimens for DNA isolation or for tissue fixation. As a rapid and nondestructive alternative, Near-Infrared Spectroscopy (NIR) has been developed for identification of specimens infected with Wolbachia, including the distinction of two different Wolbachia strains (Sikulu-Lord et al., 2016). This method, however, requires standardization according to species, sex, age, or any other condition that may affect absorbance, and is not 100% efficient. To our knowledge, this method has not been employed to assess Wolbachia infection in tephritids.

Methods to Taxonomically Characterize Wolbachia Strains
The main evolutionary lineages of Wolbachia are assigned to "supergroups" . Sixteen supergroups have been recognized to date (Glowska et al., 2015;Bleidorn and Gerth, 2017). Supergroups A and B are widespread in arthropods and are common reproductive manipulators (Baldo et al., 2006;Werren et al., 2008). Supergroups C and D are obligate mutualists of filarial nematodes, whereas supergroup F is found in both arthropods and nematodes (Ros et al., 2009). Other supergroups have more restricted host distributions . Wolbachia are generally compared and classified on the basis of Multilocus Sequence Typing (MLST) systems (Baldo et al., 2006;Paraskevopoulos et al., 2006). The most commonly used MLST is based on the PCR amplification of fragments of five ubiquitous genes: coxA, fbpA, ftsZ, hcpA, and gatB. However, this MLST system has limitations, in that not all genes are readily amplified in all Wolbachia strains, and it fails to distinguish among very closely related strains Bleidorn and Gerth, 2017). Several additional genes commonly amplified and reported are the 16S rRNA, groEL, gltA, and the wolbachia surface protein (wsp) (O'Neill et al., 1992;Braig et al., 1998;Zhou et al., 1998;Augustinos et al., 2011). The wsp gene is highly variable and shows evidence of intragenic recombination (Werren and Bartos, 2001;Ros et al., 2012). An MLST database 1 is available to compare sequences of alleles for the five MLST loci and the wsp gene. Upon submission to the MLST database, new alleles for the wsp and for each of the MLST loci are assigned a unique number. A Wolbachia sequence type (ST) is defined on the basis of MLST allele combinations, with each allele combination assigned a unique ST number. Further characterization of each MLST-defined strain can be achieved by examination of four hypervariable regions (HVRs) of the wsp gene (Baldo et al., 2005).
Hosts can be infected by one or more distinct strains of Wolbachia. Traditionally, direct Sanger sequencing of PCR products that resulted in sequences with ambiguous base pairs would be subjected to cloning followed by sequencing. The allele intersection analysis method (AIA; Arthofer et al., 2011) can then be used to assign MLST alleles to Wolbachia strains, but it requires a priori knowledge on the number of strains present. AIA identifies pairs of multiply infected individuals that share Wolbachia and differ by only one strain. Alternative approaches to circumvent cloning include the use of strain-specific primers (e.g., for the wsp gene; Zhou et al., 1998;Arthofer et al., 2009b), or of high throughput sequencing approaches (e.g., Illumina HiSeq, MiSeq, and NovaSeq) to sequence MLST or other marker PCR amplicons (e.g., Gibson et al., 2014;Brandon-Mong et al., 2015). Primer bias, however, where the fragment of one Wolbachia strain is preferentially amplified over the other, has been reported , such that presence of certain Wolbachia strains might be missed.
Use of the MLST system alone has two major drawbacks. First, strains of Wolbachia sharing identical MLST or wsp alleles can differ from each other at other loci (Paraskevopoulos et al., 2006;Riegler et al., 2012). Secondly, the MLST, 16S rRNA, and wsp loci contain limited phylogenetic signal for inferring relationships within Wolbachia supergroups (Bleidorn and Gerth, 2017). Therefore, to assess such intra-ST variation and to infer evolutionary relationships among closely related Wolbachia strains, additional (more variable) loci must be evaluated. The multiple locus variable number tandem repeat analysis developed by Riegler et al. (2012) allows distinction of closely related Wolbachia strains based on PCR and gel electrophoresis.
Whole genome sequencing represents a powerful approach to distinguish closely related Wolbachia strains, infer their evolutionary relationships, test for recombination, and identify genes of interest (e.g., Klasson et al., 2009;LePage et al., 2017). Due to its fastidious nature (but see Uribe-Alvarez et al., 2018 for a recent breakthrough) and occurrence of repetitive elements, genome sequencing and assembly of Wolbachia (and other host-associated uncultivable bacteria) has proven difficult. Recent advances, particularly those based on targeted hybrid enrichment (Lemmon and Lemmon, 2013) prior to highthroughput sequencing (Bleidorn, 2016;Goodwin et al., 2016) have been successfully applied to Wolbachia for short-read technologies (i.e., Illumina; Kent et al., 2011;Geniez et al., 2012). The combination of targeted hybrid capture and longread technologies, such as Pacific Biosciences' Single Molecule Real-Time (e.g., Wang et al., 2015) or Oxford Nanopore Technologies' platforms is expected to greatly advance Wolbachia genomics research.

Methods to Functionally Characterize Wolbachia Strains
A major challenge to investigating the effects of Wolbachia on a host is to generate Wolbachia-present and Wolbachiafree treatments while controlling for host genetic background. The challenge stems from the difficulty of adding or removing Wolbachia to/from particular hosts. Addition of Wolbachia to a particular host background can be achieved by transinfection (reviewed in Hughes and Rasgon, 2014). Because the vertical transmission of Wolbachia appears to be dependent on its close association to the host germline, successful artificial transfer of Wolbachia typically relies on injection of cytoplasm from a donor egg (but see Frydman et al., 2006) or early embryo into FIGURE 3 | Backcrossing procedure. Wolbachia infection is indicated by blue oval. Host nuclear backgrounds are indicated by colors: white represents the initial nuclear background of Wolbachia-infected host; red (darkest) indicates the host background of the Wolbachia-free line contributing males every generation. Different shades of red represent the increasing replacement of "white" nuclear background over backcrossing generations (F 1 to F 8 ) by "red" nuclear background.
a recipient embryo via microinjection (reviewed in Hughes and Rasgon, 2014). The success rate of the transinfection procedure is generally very low; in tephritids it is 0-0.09% (calculated as the proportion of injected embryos that emerged as Wolbachiainfected adult females that transmitted Wolbachia to offspring) (Zabalou et al., 2004b(Zabalou et al., , 2009Apostolaki et al., 2011;Martinez et al., 2016). The low success rate is generally a result of the low survival of injected embryos, the low proportion of Wolbachiapositive survivors, and the low/incomplete transmission of Wolbachia to their offspring. Intra-species (or between sibling species) transfer of Wolbachia to a particular host nuclear background can also be achieved through introgression, whereby males of the desired background are repeatedly backcrossed with Wolbachia-infected females (e.g., Dobson et al., 1999;Jaenike, 2007). Under this scheme, after eight generations of consistent backcrossing, ∼99.6% of the host nuclear background is expected to have been replaced (Figure 3). The drawback of this approach is that the mitochondrial genome will not be replaced. Therefore, the effects of mitochondrial type and Wolbachia infection cannot be separated.
Due to less than perfect transmission, passive loss of Wolbachia in certain host individuals may be used to obtain Wolbachia-free and Wolbachia-infected hosts of equivalent genetic background. Wolbachia removal has also been achieved by "extreme" temperature treatment (e.g., 30 • C; Ribeiro, 2009). The most common way of removing Wolbachia, however, is achieved via antibiotic treatment, but several potential biases must be addressed (reviewed by Li et al., 2014). Antibiotic treatment is likely to alter the microbiota, other than Wolbachia, associated with the host. In addition, antibiotics may affect the host in a microbe-independent manner. For instance, antibiotic treatment can affect host mitochondria (Ballard and Melvin, 2007), which in turn can reduce host fitness. A common practice to circumvent these problems is to wait several generations after antibiotic treatment, and to promote "restoration" of the host's pre-antibiotic microbiota, excluding Wolbachia (e.g., exposing the insects to the feces of non-treated individuals). Wolbachia does not appear to be efficiently transmitted via ingestion (e.g., Faria et al., 2016), but see discussion on horizontal transmission routes below. It is essential to monitor Wolbachia infection status of antibiotic-treated host strains, because antibiotics may not always fully remove infection. Instead, they may reduce Wolbachia densities to non-detectable levels in one or few generations (Li et al., 2014); this has been our experience in both Anastrepha (S.B. Lanzavecchia, C. Conte, and D.F. Segura, pers. obs.) and Drosophila (M. Mateos, pers. obs.).
Unidirectional CI is tested by comparing the embryo hatching rates of the CI cross (uninfected female X infected male) to that of one or more control crosses. For testing bidirectional CI, the reciprocal crosses of hosts infected by the different Wolbachia strains are assessed. A significantly lower embryo hatching rate of the CI cross(es) compared to that of the control cross(es) constitutes evidence of CI. CI can be partial or complete (100% embryo failure). As with any fitness assay, care must be taken to prevent potential biases, including crowding and age effects; which have been shown to influence CI (Hoffmann et al., 1990;Turelli and Hoffmann, 1995;Reynolds and Hoffmann, 2002). Adequate assessment of fertilization must be performed to ensure that failed embryos are not confused with unfertilized eggs. This may require testing for insemination of females that produce no larval progeny (e.g., Zabalou et al., 2009;Conte et al., 2019), or exclusion of females that predominantly lay unfertilized eggs, such as old Drosophila melanogaster virgin females (Menon et al., 2014).
For the genus Ceratitis, two species have been screened for Wolbachia. No evidence of Wolbachia was found in Ceratitis fasciventris Bezzi. Also, no evidence of infection was found in several populations of C. capitata, the Mediterranean fruit fly (medfly), in the early '1990s' (Bourtzis et al., 1994). PCR amplification and sequencing of the 16S rRNA gene in several field and lab specimens of C. capitata from Brazil suggested infection with Wolbachia supergroup A (Rocha et al., 2005). However, recent thorough surveys of wild populations and lab colonies indicate that Wolbachia is absent in C. capitata from numerous localities in different continents (Supplementary File S1).

Wolbachia Prevalence in Tephritids (in Time/Space)
Numerous studies report Wolbachia infection frequencies (or data from which this measure can be estimated) in natural populations of tephritids. Few of these studies, however, have adequate sample sizes for such inferences (e.g., many such studies are based on 10 or fewer individuals). Notwithstanding, inferred Wolbachia prevalence in tephritid populations is highly variable. In Anastrepha, ∼10 species harbor at least one population with prevalence ∼100%, whereas populations of three species reported lower frequencies (e.g., 88%, 51-60%, and 8.7%) (Supplementary File S1). In Bactrocera, one population of B. caudata had 100% prevalence, whereas all other species with positive Wolbachia results exhibited low prevalence.
The best studied tephritid system in terms of spatial and temporal variation in Wolbachia prevalence is that of R. cerasi in Europe, which was surveyed over a ∼15-year-period in 59 localities (Schuler et al., 2016b). Collectively, at least six strains wCer1-6 have been identified from Europe and the Middle East. In an early (1998) survey, Riegler and Stauffer (2002) found all European R. cerasi individuals infected by one strain (wCer1), most central and southern European populations harbored an additional strain wCer2 (i.e., wCer1 + wCer2), and at least one Italian population harbored wCer1 + wCer4 (Zabalou et al., 2004b). A rapid spread of wCer2 (a strain associated with cytoplasmic incompatibility) has been detected. Multiple infections, in various combinations of all five known Wolbachia strains from Europe, have been revealed recently. Samples from Poland, Italy, and Austria, are infected with strains wCer1-5 those from Czech Republic (prior to 2009) and Portugal lacked wCer2 only, while the Swiss samples lacked wCer3 (Arthofer et al., 2009b). A more recent study of the Czech Republic (2015) and Hungary (2016) revealed that wCer2 is spreading at a speed of 1.9 and 1 km/yr, respectively (Bakovic et al., 2018). Analysis of 15 Greek, two German and one Russian population confirm fixation for wCer1 in all R. cerasi populations, and the presence of complex patterns of infections with four of the five known wCer European strains (1, 2, 4, and 5) and the possible existence of new Wolbachia strains for the southernmost European R. cerasi population (i.e., Crete; Augustinos et al., 2014) and from Iran (wCer6) (Karimi and Darsouei, 2014). Similarly, strain wCin2 (which is identical to wCer2 based on loci examined to date) is fixed in all populations of R. cingulata; a species native to the United States, but introduced into Europe at the end of the 20th century. Invasive populations in Europe harbor wCin1 (identical to wCer1 based on loci examined to date) at frequencies that vary over space and time (up to 61.5%), as a result of horizontal transfer (multiple events) from R. cerasi (Schuler et al., 2016b). The above studies indicate that the prevalence of Wolbachia types in R. cerasi and R. cingulata is highly dynamic.

Phenotypic Effects of Wolbachia in Tephritids
Despite the numerous reports of Wolbachia in tephritids, the fitness consequences of such associations remain mostly unknown. The studies reporting phenotypic effects of Wolbachia have relied on transinfection and on antibiotic-curing; only two species of tephritids have been successfully transinfected with Wolbachia ( Table 1). Evidence of Wolbachia-induced CI has been detected in four species of tephritids. Early studies (Boller and Bush, 1974;Boller et al., 1976) identified reproductive incompatibilities in R. cerasi that were later attributed to the Wolbachia strain wCer2 (100% embryonic mortality in the CI cross; Riegler and Stauffer, 2002). Artificially transferred Wolbachia (strains wCer2 and wCer4) originally from R. cerasi to C. capitata also resulted in strong CI (100% embryonic mortality). wCer2 in two genetic backgrounds of B. oleae resulted in strong CI as well (Apostolaki et al., 2011). In addition, wCer2 and wCer4 are bi-directionally incompatible in C. capitata (Zabalou et al., 2004b(Zabalou et al., , 2009).
In addition to CI, Wolbachia-infected C. capitata females (Benakeio strain) exhibit higher embryonic mortality (17-32% in crosses with Wolbachia-free males and 65-67% in crosses with Wolbachia-infected males) than their Wolbachiafree counterparts crossed with Wolbachia-free males (12% embryonic mortality). Therefore, it appears that wCer2 and wCer4 have additional fertility effects on medfly females, other than CI. It is also possible that these Wolbachia strains can only partially rescue the modification that they induce in sperm (Zabalou et al., 2004b). A similar pattern is reported in the Vienna 8 genetic sexing strain (GSS) infected with wCer2 (Zabalou et al., 2009). The wCer2 strain also causes increased embryo death in non-CI crosses of B. oleae (Apostolaki et al., 2011). In D. simulans, wCer2 causes fecundity costs, moderate levels of CI, and incomplete rescue of its own CI modification (Riegler et al., 2004). Interestingly, a recent study examined the genome of wCer2 and revealed the presence of three pairs of Type I cif genes and one Type IV cifB gene without a cifA complement, which might explain its idiosyncratic expression of CI (Morrow et al., 2020).
Two studies conducted several years apart (Sarakatsanou et al., 2011;Kyritsis, 2016;Kyritsis et al., 2019) examined the effects of a single Wolbachia strain (wCer2) on fitness components of two C. capitata genotypes (i.e., Benakeio and Vienna 8 GSS laboratory lines), as well as the effects of two different Wolbachia strains (wCer2 and wCer4) on a single medfly genotype (Benakeio). The following general patterns emerged (exceptions noted): (a) Wolbachia causes higher eggto-larva mortality; (b) Wolbachia causes higher egg-to-adult mortality (exception: Vienna 8 GSS + wCer2 in Sarakatsanou et al., 2011); (c) Wolbachia shortens egg-to-adult development time (exception: Benakeio + wCer2 in Kyritsis, 2016;Kyritsis et al., 2019). In addition, Sarakatsanou et al. (2011) found that Wolbachia shortens both male and female adult lifespan (exception: males of Vienna 8 GSS and wCer2), and reduces life time female net fecundity. However, Kyritsis (2016) and Kyritsis et al. (2019) reported no effects of Wolbachia infection on adult lifespan, and a reduced fecundity in the case of wCer4 infection only. Even though wCer2 and wCer4 in general tended to have consistent effects on medfly, the magnitude of their effects differed. Collectively, the results from these studies indicate that the effect of Wolbachia infection on life history traits depends both on the C. capitata genetic background and on the Wolbachia strain. Furthermore, inconsistencies between the two studies might be indicative of evolution of the host and/or Wolbachia strain during that period. Evidence of Wolbachia evolving reduced fitness costs has been reported in D. simulans (Weeks et al., 2007). Adult flight ability and longevity under stress conditions also appear to be determined by the interaction of Wolbachia strain and medfly genotype (Kyritsis, 2016;Kyritsis et al., 2019). A more recent study (Dionysopoulou et al., 2020) demonstrated Wolbachia effects on medfly reared in natural host fruits and at different temperatures. Medlfies infected with wCer4 had low survival rates in both apples and bitter oranges, whereas those infected with wCer2 were less vulnerable in apples than in bitter oranges. In addition, wCer4 infected flies were particularly susceptible to high temperatures.
A recent study by Conte et al. (2019) examined the phenotypic effects induced by two Wolbachia strains native to A. fraterculus (sp1). No evidence of bidirectional cytoplasmic incompatibility was detected in reciprocal crosses among singly infected laboratory strains. However, the same work described the presence of slightly detrimental effects on larval survival and a female-biased sex ratio, suggesting the induction of malekilling phenomena. Moreover, Devescovi et al. (2019) found that Wolbachia reduced the embryo hatching in crosses involving cured females and infected males (uni-directional CI) within two morphotypes of this cryptic species complex; stronger CI was detected within the Peruvian morphotype than the Brazilian-1 morphotype (also referred as to "A. fraterculus sp. 1"). No evidence of bidirectional CI was detected in the crosses between the two morphotypes, leading Devescovi et al. (2019) to conclude that Wolbachia is not directly involved in the speciation process of these morphotypes. Ribeiro (2009) reported evidence consistent with CI caused by Wolbachia in A. obliqua and in "A. fraterculus sp. 1, " which according to wsp sequences, are identical. Nonetheless, confounding effects of the treatment to remove Wolbachia (removed by exposure of pupae to 30 • C) or other potential biases cannot be ruled out, as all intraspecific crosses involving at least one cured parent resulted in much lower (<30%) embryo hatching than the intraspecific crosses involving both infected parents (66 and 81% embryo hatching).
Recent work demonstrates that Wolbachia infection can affect male sexual competitiveness of C. capitata. Different Wolbachia strains (wCer2 and wCer4) exerted differential impact on males mating competitiveness, and a single strain (wCer2) had different impact on different medfly genotypes (Benakeio and Vienna 8 GSS laboratory lines) (Kyritsis, 2016;Kyritsis et al., 2019).

Modes of Horizontal Transmission of Wolbachia Between Tephritid Hosts
Considering the dynamics of Wolbachia associated with arthropods in general, at the population level Wolbachia appears to be predominantly maintained by vertical transmission. Above the species level, however, the lack of congruence between the host and symbiont phylogenetic trees implies that Wolbachia horizontal transfers and extinctions are common and underlie its widespread taxonomic and geographic distribution (Bailly-Bechet et al., 2017). The possible routes by which Wolbachia may be horizontally acquired by a new host can generally be classified as via ingestion or via a vector. In both cases, to become established as a stable cytoplasmically inherited infection, Wolbachia must cross one or more cell types or tissues. For example, if Wolbachia invaded the host hemolymph directly as a result of a vector (e.g., parasitoid wasp or ectoparasitic mite), it would have to invade the egg during oogenesis. Similarly, if Wolbachia were acquired via ingestion (e.g., as a result of scavenging), it would have to cross the gut into the hemolymph, before it invaded the egg. Support for the above routes comes from studies reporting: (a) that Wolbachia can retain viability outside cells and infect mosquito cell lines, as well as ovaries and testes that are maintained ex vivo (Rasgon et al., 2006;Hughes et al., 2012a); (b) that Wolbachia cells injected into Drosophila hemolymph reach the germline after crossing multiple somatic tissues (Frydman et al., 2006); (c) that Wolbachia can move between parasitic wasp larvae (Trichogramma) sharing the same host egg, and achieve vertical transmission (Huigens et al., 2004); and (d) that parasitic wasps of the white fly, Bemisia tabaci (Gennadius), can transfer Wolbachia from an infected to a naïve host, as a result of nonlethal probing (i.e., probing without oviposition), whereby the parasitoid ovipositor or mouthparts function as a "dirty needle" .
No direct evidence of Wolbachia transmission via parasitoids exists in tephritids, but sharing of Wolbachia strains between a parasitoid and several sympatric tephritids (Morrow et al., 2014;Mascarenhas et al., 2016) is consistent with parasitoid-mediated transmission, or transmission from tephritid host to parasitoid (Johannesen, 2017). The potential for horizontal transfer of Wolbachia among tephritids via parasitoids is high, due to the multiple instances where a single parasitoid utilizes several different tephritid host species (Quilici and Rousse, 2012;Murillo et al., 2016;Schuler et al., 2016a), and the high frequency of superparasitism by some fruit fly parasitoids (Tormos et al., 2012;Devescovi et al., 2017).
Wolbachia may invade a new host species via introgressive hybridization between two host species. This mechanism would also transfer mitochondria from the infected to the uninfected species nuclear background, akin to the artificial backcrossing approach described above (Figure 3). Ability of tephritids to hybridize in the lab has been reported in numerous species (Table 2), and hybridization in nature has been documented in B. dorsalis/B. carambolae (Wee and Tan, 2005), members of the Ceratitis FAR complex (Virgilio et al., 2013), and R. cingulata/R. cerasi in Europe (Johannesen et al., 2013). Thus, there is potential for wild tephritid populations to acquire Wolbachia infections via hybridization.

Wolbachia-BASED IIT IN TEPHRITIDS
There are two main approaches for implementing IIT, which depend on whether uni-or bi-directional CI will be used. If the target pest population lacks Wolbachia, such as the tephritids C. capitata, B. oleae (Gmelin), and A. ludens [and the mosquito Aedes aegypti (L.)], only unidirectional CI is feasible. In target populations that harbor one (or more) CI-inducing Wolbachia strain(s) (i.e., native strain; yellow in Figure 4), bi-directional CI can be achieved by releasing males that lack the native strain(s) and harbor one (or more) "foreign" Wolbachia strain(s) (blue in Figure 4) that is incompatible with the native strain. In contrast, if the released males are doubly infected with the native and foreign strains, the CI pattern employed for population suppresion is uni-directional (Figure 4).
In the case of Uni-CI patterns, the accidental release of Wolbachia-infected females, which would be reproductively compatible with wild and released males, may result in the replacement of the target pest population with a population harboring the Wolbachia infection of the released males, leading to failure of the IIT-based suppression program (Bourtzis, 2008). As described in Section "The Influence of Wolbachia on Host Ecology, " under certain conditions, a Wolbachia infection with frequency close to zero might be able to rapidly spread through a host population. Thus, without efficient sex separation mechanisms (outlined in section below), it is desirable to ensure that accidentally released females are sterile. In several tephritid systems, female sterility is achieved at a lower irradiation dose than male sterility, such as A. ludens, A. obliqua, Anastrepha suspensa, A. serpentina, B. tryoni, and Z. curcubitae (Toledo et al., 2004;Bakri et al., 2005;Rull et al., 2007;Collins et al., 2009). Therefore, an IIT program that used radiation at doses to ensure female sterility without compromising male quality (e.g., male competitiveness) could be effective (e.g., Drosophila suzukii based on results to date; Nikolouli et al., 2020).
In an IIT program based on bi-CI pattern (e.g., the recent field study of Aedes albopictus; Caputo et al., 2019), accidental release of fertile transinfected females, which would only be  (Drew and Romig, 2013;Schutze et al., 2015). Bold-face names are species where at least one report of Wolbachia infection exists (see Supplementary File S1).
compatible with the released males, would not necessarily lead to population replacement and program failure. This is due to the generally higher threshold density required to achieve invasion (theoretically above 50% when the two incompatible Wolbachia strains exert equivalent fitness costs/benefits; see Section "The Influence of Wolbachia on Host Ecology"). Nonetheless, the actual outcome is strongly dependent in multiple factors (see Dobson et al., 2002;Moretti et al., 2018a). Therefore, for both uni-CI-and bi-CI-based IIT programs, as pointed out by Bourtzis et al. (2014), the outcome of accidental releases of infected females must be thoroughly evaluated via modeling and/or semifield assays prior to field applications.

The Advantage of Genetic Sexing Strains (GSS)
In general, SIT and IIT are most effective when only males are produced and released (Kerremans and Franz, 1995;Rendón et al., 2004). The release of only males in a large-scale operation can be accomplished by either killing female zygotes during development or by selectively removing them from the massreared population prior to release (Robinson, 2002a;Gilles et al., 2014;Lutrat et al., 2019). Genetic sexing strains (GSS) are those in which individuals can be separated by sex prior to the adult stage on the basis of a sex-linked phenotype (Franz, 2005). The earlier in development the females are removed, the most cost-effective the mass rearing operation will be, as investment in growth of females would be null or minimized. In most tephritids, male sex is determined by the presence of the maleness factor on the Y chromosome (Pane et al., 2002). GSS based on male-linked [e.g., Y chromosome -autosome (Y;A)] translocations have been developed in a few species to produce conditional female lethality (e.g., temperature sensitive lethality during embryonic development) or a visual sex marker (e.g., pupal color). Examples of tephritid species for which GSS are available include C. capitata (Franz, 2005), A. ludens (Zepeda-Cisneros et al., 2014), Z. cucurbitae , B. dorsalis (Isasawin et al., 2012), and B. carambolae (Isasawin et al., 2014). Unfortunately, despite substantial efforts, GSS are still lacking for most tephritid pests. The recent development of CRISPR/Cas9-mediated mutagenesis in tephritids, however, might enable a faster development of tephritid GSS (Reid and O'Brochta, 2016;Choo et al., 2017; see reviews by Ogaugwu and Durvasula, 2017).

Choice and Evaluation of Wolbachia Strains
The target population and the donor colony should be thoroughly screened for Wolbachia, ideally with the higher sensitivity methods described in Section "Methods to Assess Wolbachia Infection Status, " to detect low-titer and multistrain infections. The Wolbachia strain selected should cause strong uni-CI with a Wolbachia-free, or strong bi-CI with Wolbachia-infected, target population. The selected Wolbachia strain should be artificially transferred to one or more lab colonies, representative of the genetic background of the target pest population. Most cases of successful establishment of stable transinfected insect lines have relied on embryonic microinjection . Introgressive backcrossing might be feasible in scenarios where geographically isolated populations of the same target species harbor distinct Wolbachia strains (e.g., A. striata in Mexico vs. A. striata in Brazil; Supplementary File S1).
A thorough biological characterization of the artificial host-Wolbachia association should be conducted, as both host background and Wolbachia strain are important determinants of CI expression and other relevant fitness parameters (Bourtzis and Robinson, 2006;reviewed in Bourtzis, 2008; see also Kyritsis et al., 2019). The main desired characteristics of the association are: strong induction of CI; no rescue by Wolbachia strain(s) present in the target population; small or no fitness cost for FIGURE 4 | Use of bi-directional CI in IIT-based population suppression programs. (A) Patterns of compatibility with two bi-directionally incompatible strains, when doubly infected hosts occur. Empty male and female symbols signify absence of Wolbachia. Blue and yellow ovals represent distinct (mutually incompatible) Wolbachia strains. Green tick marks = Successful offspring production. Red crosses = no offspring production. (B) Options for implementing IIT-based population suppression when the target wild population harbors a CI inducing strain (yellow = "native"), according to the patterns of compatibility depicted in panel a. Bi-directional CI is achieved when the released males only harbor a strain (blue = "foreign") that is incompatible with the native strain. Additional options exist, including double infections of both target and released insects with different Wolbachia strains (not shown), such as in Aedes albopictus (Moretti et al., 2018b).
parameters relevant to the program. These fitness parameters can be classified into those related to a cost-effective mass production (e.g., female fecundity including embryo hatching success) and those related to the success of released males (e.g., mating and sperm competitive ability, as well as dispersal/flight ability). Some host-Wolbachia combinations result in higher female fecundity, such as D. simulans after many generations (Weeks et al., 2007) and Drosophila mauritiana Tsacas and David (Fast et al., 2011). In contrast, other host-Wolbachia combinations result in lower fertility (e.g., low embryo success in C. capitata and B. oleae; Zabalou et al., 2004bZabalou et al., , 2009Apostolaki et al., 2011). Wolbachia could affect male mating success by influencing assortative mating; a phenomenon detected in some studies of Drosophila (e.g., Koukou et al., 2006;Miller et al., 2010), but not others (e.g., Champion de Crespigny and Wedell, 2007;Arbuthnott et al., 2016;Cooper et al., 2017). Such influence of Wolbachia on mating preferences was questioned (Sharon et al., 2010) on the basis of evidence that gut microbiota influence assortative mating in Drosophila (Sharon et al., 2010;Ringo et al., 2011;Arbuthnott et al., 2016), a finding that itself has been questioned recently (Leftwich et al., 2017(Leftwich et al., , 2018. In addition, at least one case has been reported where sperm from Wolbachiainfected males was less competitive (Champion de Crespigny and Wedell, 2006). Similarly, Wolbachia-infected D. simulans produce fewer sperm (Snook et al., 2000). All of the above parameters should be evaluated under relevant conditions known to interact with Wolbachia, such as temperature and nutrition (reviewed in Bourtzis and Robinson, 2006;e.g., Serbus et al., 2015;Corbin et al., 2017;Ross et al., 2019a), interaction with other microorganisms (e.g., Ye et al., 2017), as well as male age, paternal grandmother age, and mating status (e.g., Karr et al., 1998;Awrahman et al., 2014;Layton et al., 2019).

Species Recalcitrant to Wolbachia?
Certain species or clades appear to be "resistant" to Wolbachia infection, based on their lack of infection in nature and the failure to achieve stable transfections. The reasons are unknown, but could involve host and/or bacterial factors. For example, none of the members of the diverse repleta species group of Drosophila, comprised mostly of cactophilic flies (Markow and O'Grady, 2005), has ever been found to harbor Wolbachia (Mateos et al., 2006). Similarly, due to numerous failed transinfection attempts, and the lack of natural infection in wild Anopheles mosquitoes, this genus was regarded impervious to Wolbachia (reviewed in Hughes and Rasgon, 2014). This view has been challenged by the successful establishment of Wolbachia-transfected Anopheles stephensi Liston (Bian et al., 2013), and the recent discovery of a natural stable Wolbachia infection in Anopheles coluzzii Coetzee & Wilkerson (Shaw et al., 2016). Nonetheless, reports of Wolbachia in other species of Anopheles have been called into question (Sicard et al., 2019). The lack of natural infections and transinfection failure in A. ludens may reflect a general refractoriness to Wolbachia. Nonetheless, initial attempts to transinfect C. capitata also failed and transfection with Wolbachia was attained subsequently with different Wolbachia strains (Zabalou et al., 2004b). Hence, transinfection attempts with additional Wolbachia strains may result in successful and stable infection in A. ludens as well.

Potential for Target Populations to Become Resistant to Sterile Males
There are two ways in which a target population may become resistant to the effects of released Wolbachia-infected males. The first is endosymbiont-based, whereby the target population may acquire (e.g., via horizontal transmission) a Wolbachia strain that can rescue the modification (sterility) induced by the strain present in the released males. Generally, such acquisition of a Wolbachia strain during the relatively short lifespan of a release program seems unlikely. Nonetheless, knowledge on the Wolbachia infection status and strain identity of interacting species, such as other fruit flies sharing the same host plant and parasitoids, might aid in the selection of Wolbachia strains that are unlikely to be compatible with strains that can potentially be horizontally acquired by the target population. Permanent screening of wild flies from the target population could provide valuable information in order to foresee potential lack of effectiveness of the method. Laboratory experiments in which the conditions for horizontal transmission are favored (or even forced) might also help to determine the probability of such phenomena to occur in nature.
The second mechanism is host-based, whereby pre-or post-mating selection on wild females to avoid or reduce fertilization by incompatible sperm (reviewed by Wedell, 2013), acts on standing (or de novo) genetic variation. Evidence consistent with the influence of Wolbachia on premating mechanisms comes from the observation that females and males of Drosophila paulistorum Dobzhansky and Pavan exhibit assortative mating according to the Wolbachia strain they harbor (Miller et al., 2010;Schneider et al., 2019). In addition, treatment with antibiotic (which removed Wolbachia) decreases mate discrimination in D. melanogaster (Koukou et al., 2006). The evolution of resistance to mating with mass-reared males by wild females can be potentially minimized by frequently refreshing the genetic background of the mass-reared strain, with or without artificial selection (McInnis et al., 2002;Gilchrist et al., 2012;Zygouridis et al., 2014;Quintero-Fong et al., 2016;Sánchez-Rosario et al., 2017), which is a routine process in mass-rearing programs aimed at countering inbreeding and adaptation to mass rearing that is detrimental the success of released males . Nonetheless, if the basis for mate discrimination were solely determined by Wolbachia infection state (e.g., if females could distinguish Wolbachia-infected vs. Wolbachia-uninfected males solely on the basis of a Wolbachia-encoded factor), refreshing the fly genetic background of mass-reared strain is unlikely to slow down the evolution of resistance to released males in the target population.
Several lines of evidence are consistent with the influence of Wolbachia infection on post-mating mechanisms. The existence of genetic incompatibility is predicted to favor polyandry (multiple mating by females) as a female strategy to minimize the probability of her eggs being fertilized by sperm from incompatible males (Zeh and Zeh, 1996). Consistent with this prediction, uninfected D. simulans females remate sooner than Wolbachia-infected females (Champion de Crespigny et al., 2008). Furthermore, Wolbachia modifies the length of the spermathecal duct of females of the cricket Allonemobius socius Scudder (Marshall, 2007), which in turn may afford the female greater control on the outcome of sperm competition (e.g., D. melanogaster; Miller and Pitnick, 2002). Finally, the fact that host background can influence the CI phenotype (reviewed by Bourtzis and Robinson, 2006), suggests that target populations may have genetic variants that are more resistant to CI, which could increase in frequency as a result of the strong selection exerted by the massive release of Wolbachiainfected males.

Wolbachia-Based Approaches
The recent identification of Wolbachia "CI genes" offers potential alternative ways of harnessing reproductive incompatibility in control of pest tephritids. First, to identify strains with the desired characteristics, at least ability to induce CI, a productive endeavor might be to search for CI loci in the genomes of candidate strains being considered for IIT, prior to artificial transfer efforts. A candidate Wolbachia strain that lacks CI loci homologs, or that contains CI loci homologs that are highly similar to (and thus potentially compatible with) strains present in target population, should be avoided. Secondly, it may be possible in the future to genetically engineer Wolbachia strains with the desired characteristics (e.g., one or more specific CI operons) for IIT programs, or to replace strains used previously in a control program, as a means of addressing resistance phenomena (Sullivan and O'Neill, 2017). Finally, a thorough understanding of the CI mechanism might enable the development of IIT based on Wolbachia transgenes, rather than Wolbachia infection. This might be particularly helpful in the control of species that are resistant to Wolbachia infection. Nonetheless, the release of such genetically modified insects might not be feasible due to regulatory hurdles and lack of public acceptance.
It has recently been shown that some Wolbachia strains can provide protection against major pathogens and parasites of insects, including RNA viruses and bacteria (Hedges et al., 2008;Teixeira et al., 2008;Ye et al., 2013;Martinez et al., 2014). It is very common for pathogens to appear in rearing facilities. Thus, if a Wolbachia strain could simultaneously cause strong CI and protect against one or more pathogens (e.g., RNA virus), this would have multiple benefits in an operational Wolbachia-based population suppression program. Furthermore, a Wolbachia strain that does not induce (strong) CI, but protects against pathogens might be desirable in a program that does not rely on CI (e.g., SIT) for population suppression.
Wolbachia-mediated pathogen protection would enable high production and quality levels, thereby contributing to a cost-effective and sustainable insect pest management program.

Potential Influence of Other Symbionts
Multiple studies have revealed that although Wolbachia appears to be the dominant facultative heritable symbiont of arthropods, numerous other diverse bacteria (e.g., Spiroplasma, Arsenophonus, Rickettsia, and Cardinium) form such associations with insects, causing a diversity of reproductive and nonreproductive phenotypes (reviewed in Zchori-Fein and Hurst and Frost, 2015;McLean et al., 2018). Despite the long-standing recognition that "Wolbachia do not walk alone" (Duron et al., 2008), many studies of Wolbachia fail to rule out the association of their study organism with other facultative heritable symbionts. Even intensely studied groups in terms of heritable symbionts, such as tsetse flies (genus Glossina), can yield surprises of bacterial associates (e.g., the recent discovery of Spiroplasma in two species of Glossina; Doudoumis et al., 2017). With few exceptions (Martínez et al., 2012;Augustinos et al., 2015;Asimakis et al., 2019;Conte et al., 2019;Devescovi et al., 2019), research on tephritid facultative heritable bacteria has not examined the possibility of players other than Wolbachia. Therefore, we urge that such research include screens for other symbionts, including viruses, protozoans, and fungi.
Tephritids are hosts to non-heritable bacteria, generally harbored in their gut (for recent reviews see Noman et al., 2019;Raza et al., 2020). Whether Wolbachia influences tephritid interactions with other microbes, has not been evaluated, but evidence for such interactions exists for other systems (reviewed in Brinker et al., 2019). For example, in Drosophila neotestacea Grimaldi, James, and Jaenike, the presence of Wolbachia promotes the abundance of Spiroplasma, and is positively correlated with abundance of Bacteroidales and Lactobacillales (Fromont et al., 2019). Similarly, Wolbachia influences the microbiome of D. melanogaster (Simhadri et al., 2017) and Armadillidium vulgare (Latreille) (Dittmer and Bouchon, 2018). It is therefore important to evaluate interactions between Wolbachia and the microbiome that influence negatively or positively aspects of mass-reared tephritids used in IIT or SIT.

CONCLUSION
Given the widespread occurrence of Wolbachia in tephritids and its known fitness consequences in this group of dipterans and in other host taxa, Wolbachia is likely an influential component of tephritid ecology and evolution. Further exploration of Wolbachia-tephritid associations is expected to reveal a diversity of effects, including interactions with other microbial partners, as seen in more extensively studied systems such as Drosophila and mosquitoes. The recent exciting progress in understanding the basis of CI, and many other aspects of Wolbachia biology, should accelerate progress in the development of Wolbachia-based IIT for tephritid species, particularly with the aid of comparative Wolbachia genomics to identify potential CI patterns on the basis of CI gene composition. We consider that one of the major obstacles to effectively implementing IIT will be to avoid population replacement due to accidental release of Wolbachiainfected females. The threshold number of accidentally released females, which is generally much higher in systems that employ bidirectional-CI compared to unidirectional-CI, must be thoroughly investigated prior to any field implementation. Where an unacceptable risk of population replacement exists, we recommend that SIT be explored as a complementary strategy to support IIT.

AUTHOR'S NOTE
An earlier version of manuscript has been released as a Pre-Print at https://www.biorxiv.org/content/10.1101/358333v1.

ACKNOWLEDGMENTS
K. Bourtzis provided valuable suggestions on an earlier version of this manuscript. We thank two reviewers for their thorough and helpful feedback.