D. suzukii were raised on cornmeal-yeast-agar diet at room temperature (approx. 20–22°C) in the open laboratory. The relative humidity in the lab was between 20% and 50% and the lights were on for about 12 h on most days. The original wild-type colony was established from infested fruit collected from a field in North Carolina in 2011 (Burrack et al., 2013) and the initial FL19 strain was previously made by piggyBac-mediated germline transformation of this wild type strain (Li et al., 2021). The wild type colony was periodically genetically augmented with flies collected in North Carolina (Elsensohn et al., 2021). The newly refreshed and the original 2011wild type colonies are maintained separately by our lab. To remobilize the FL19 transgene, ten FL19 virgin females were crossed with five males from the H7 piggyBac jumpstarter strain (Chu et al., 2018) (Figure 1). From the offspring of the cross, ten virgin females were crossed to five wild type males. The male offspring were screened for bright red fluorescence using a M205FA microscope (Leica Microsystems, Buffalo Grove, IL) with the DsRed filter [ex 545/25, em 595/50 nm]. Individual candidate males were then each crossed with five wild type virgin females. If the FL19 transgene had transposed to the X chromosome, then only the female offspring would show red fluorescence. If both sexes showed bright red fluorescence, this could indicate that the flies carried two autosomal copies of FL19. Homozygous lines for each putative transposition event were established by crossing and selecting for particularly high levels of red fluorescence.

For recombination mapping of X-linked FL19 transgenes, six crosses were set for each combination FL19(X) lines. The double heterozygous virgin female offspring were collected and crossed with wild type males. Recombinant male offspring that showed no red fluorescence was identified and counted. A map distance of the two FL19(X) sites were calculated by; map distance (cM) = 100 x (2 x number of non-fluorescent males)/total males.

To assess the level of female lethality in the homozygous strains, three vials were set on the same day with five pairs on a diet that either contained tetracycline (40 μg/ml) or lacked tetracycline. Parents were transferred three times to new cultures every 3 days to create four cultures. To determine if the strains showed dominant female lethality, five transgenic males were crossed with five wild type virgin females. Parents were transferred to new cultures every 3 days to create four cultures. Two replicates were set each on tetracycline and non-tetracycline foods. The number of male and female offspring from each cross were counted daily until 20 days after setting the cross.

Cut vials, which were regular fly vials cut at the middle to create 7.2 cm-long tubes and 2.3 cm-deep cups, were used for easy handling of eggs. A tube and a cup were taped together to create a vial for productivity tests. The cup contained about 10 ml of culture medium. Flies that were raised under non-crowded conditions were collected from several vials 3 to 14 days after eclosion and then allowed to lay eggs in the cut vials for about 20 h. The cut vials were then separated, eggs were picked with a needle and transferred to a new vial with about 10 ml of medium. Typically, 50 eggs were transferred to each vial. After 2 days unhatched eggs were counted. Emerging adults were sexed and counted up to 3 weeks after the egg-picking. At the end of emergence of adult flies, the number of pupal cases was counted. The egg survival ratio is the number of hatched eggs divided by the total number of eggs. The larval survival ratio is the number of pupae divided by the number of hatched eggs. The pupal survival ratio is the number of adults divided by the number of pupae. The egg to adult survival ratio is the number of adults divided by the total number of eggs. This ratio was multiplied by two for the transgenic lines on diet without tetracycline.

Ten transgenic males from strains reared without tetracycline in the diet and 10 wild type males were introduced into an 8 oz bottle with diet and left undisrupted for 1 hour. Ten wild type virgin females were added to the bottle which was kept undisrupted at room temperature (∼22°C) for approximately 24 h. All flies were four to 6 days old. Females were transferred individually to fresh vials with diet. The offspring of each female were counted, sexed and examined for fluorescence status (presence/absence) to determine whether the female had mated with an FL19(3 + 3) or wild type male. The presence of both fluorescent and non-fluorescent offspring would indicate remating during the 24 h period when males were with females. Four to eight bottles were set for each line. Mate success ratio of each line was calculated by; success ratio = (number of females with fluorescent sons only + a half number of females with both fluorescent and non-fluorescent sons)/number of fertile females.

The genomic location of the transgenes was determined using inverse PCR with primers for the piggyBac left and right ends as previously described (Li et al., 2001 and see Supplementary Table S1). If inverse PCR was only successful for one end, confirmation of the transgene location was obtained by PCR using one primer for the piggyBac end and one primer based on the flanking genome sequence of the insertion site.

For the female lethality tests (Table 4), contingency analysis (fit Y by X) was performed using JMP Pro 15 (SAS Institute). The egg to adult ratio (Table 5) was analyzed in SAS (Version 9.4, Cary, NC) on a diet with tetracycline, with the total number of adults divided by the number of eggs as the response variable in PROC LOGISTIC, and both line, tetracycline (+/-), and their interaction as predictor variables. All effects were statistically significant (p < 0.0001). For diet without tetracycline, the number of male eggs was estimated as the floor of the number of eggs divided by two. Least-squares means were obtained for each treatment combination, and the specific differences of interest were calculated. The log-odds ratio of the two treatments (each line compared to each wild type for both tetracycline + and tetracycline -) were obtained. To control for multiple tests, a Bonferroni correction was used where the adjusted p-value to determine statistical significance was set at 0.05/30 = 0.0016 for a diet with tetracycline and 0.05/15 = 0.0033 for a diet without tetracycline (only males produced from transgenic lines). The male competitiveness data (Table 6) were also analyzed in SAS, Version 9.4 (Cary, NC). The mating competitiveness index (MCI) was recalculated as the ceiling of the number mated with transgenic plus half the number that remated divided by the total number mated. This allowed the count to remain an integer. A z-test for one proportion was run for each line comparing the MCI to the null hypothesized proportion of 0.5.

The original FL19 male-only strain was made by piggyBac transposase-mediated germline transformation (Li et al., 2021). The transgene was located on chromosome 3 between the DsShal and DsCG9231 genes. The piggyBac H7 jumpstarter strain efficiently mediates remobilization of piggyBac transgenes (Chu et al., 2018). To remobilize the FL19 transgene, FL19 virgin females were crossed with H7 males and virgin female offspring collected (Figure 1). As X-linked transgenes are generally dosage compensated in male Drosophila melanogaster (Spradling and Rubin, 1983; Fitzsimons et al., 1999), we reasoned that G3 males carrying a single X-linked FL19 transgene would show an increased expression of the red fluorescent protein marker gene (Figure 1). Males that carried FL19 at the original location and at a second autosomal location would also show brighter red fluorescence. Such transpositions tend to be predominately local (Daniels and Chovnick, 1993), so the expectation was that males would carry FL19 at the original location and at a closely linked site on chromosome 3. X-linked transposition events were identified through crossing putative males with wild type virgin females (Figure 1). From approximately 25,000 G3 males, we derived four X-linked lines and eleven lines with two copies of FL19 on chromosome 3 (Table 1). Of the third chromosome lines, one was homozygous lethal and two lines were homozygous sterile (Table 1). Of the remaining eight homozygous viable lines, five (8, 36, 40, 70 and 75) that were vigorous on diet with tetracycline were selected for further study.

The chromosomal locations of the X-linked and most of the chromosome 3 lines were determined by inverse PCR and blast searches of the assembled D. suzukii genomes (Chiu et al., 2013; Paris et al., 2020) (Table 2). Some locations could not be determined as the transgene appeared to be located within a repetitive sequence. PCR analysis confirmed that all the chromosome 3 lines carried two copies of FL19, with one copy at the original location near the DsCG9231 gene (Figure 2). The other copies of FL19 were found to have inserted no more than 68 kb from the original location (Table 5; Figure 2). In line 36, the additional FL19 transgene is less than 3 kb from the original and is also within the intergenic region between the DsCG9231 and DsSha1 genes (Figure 2). In line 70, the additional FL19 transgene is also in an intergenic region, between the Dswnd and DsRnf146 genes (Table 2; Figure 2). In the other lines, 8 and 75, the transgene is located within genes. In line 8, the transgene is within an intron of the DsRnf146 gene. In the line 75 the transgene is within an exon of the DsCG14100 gene and would likely disrupt gene function.

In the X-linked lines, three of the four FL19 transgenes appear to be found at widely separated locations as their flanking sequences each align to a different contig of the genome assembly (Paris et al., 2020) (Table 2). The location of the fourth transgene (#79) could not be determined by inverse PCR. In line 46 the transgene is in a gene-poor region of the X chromosome with the nearest genes more than 20 kb distant. We next used recombination mapping to determine the genetic distance between each of the transgenes. The transgene in line 46 was found to be 45 cM from the transgene in line 79 (Table 3). The transgenes in lines 77 and 79 appear to be very closely linked locations as we did not recover any recombinants (Table 3). Lastly, in line 7 the transgene mapped to a region between the 46 and 79 transgenes (Table 3).

All the lines can be readily maintained on diet supplemented with tetracycline. When raised on diet that lacked tetracycline all lines produced 99–100% males except for the X-linked line 46, which gave 73.5% males (Table 1). In a future field release, flies would be raised on diet without tetracycline and the released fertile males would mate with wild type females. Ideally, all the female offspring would die. Therefore, we next collected males from the lines raised on diet without tetracycline and crossed to wild type virgin females. For the chromosome 3 lines that carry two copies of the FL19 transgene, all lines showed dominant female lethality (Table 4). However, none of the X-linked single copy lines showed dominant female lethality. Two lines, 7 and 77, produced significantly more male than female offspring (Pearson’s Chi-squared test, p < 0.0001). On a diet with tetracycline, approximately an equal number of males and females were produced from the crosses of transgenic males with wild type females.

One of the fitness measurements that is important in a mass rearing facility is the percentage of eggs that produce adults (Concha et al., 2016; Concha et al., 2020). We firstly measured the proportion of eggs that produced adults in our two wild type North Carolina colonies. One colony has been periodically refreshed with flies collected in North Carolina and the other has been continuously maintained in the laboratory since 2011. The 2011 wild type strain showed significantly reduced adult production compared to the newly refreshed strain on diet with or without tetracycline (p < 0.0001) (Table 5). Similarly, on diet with tetracycline, the adult production for all of the transgenic lines was significantly less than the newly refreshed wild type strain (Table 5). However, since the original FL19 line was by injecting embryos from the 2011 wild type strain, we also compared each of the transgenic lines to that strain. Compared to the 2011 wild type strain, adult production on a diet with tetracycline and male production on a diet without tetracycline was not significantly different in five of the seven lines (FL19, 8, 36, 40, 7(X) + FL19). Two of the transgenic lines, 70 and 75, did show significantly reduced adult production on both diets compared to the 2011 wild type strain (Table 5). Several of the transgenic lines (e.g., 8, 40, 70) showed a wider variation between replicates for pupal and adult survival on a diet without tetracycline than on a diet with tetracycline (see standard deviation (SD) values in Table 5). This variability in male survival suggests there may be some leaky expression of the lethal genes, particularly in some lines.

The sexual competitiveness of the males from the lines was assessed by presenting virgin wild type females with equal numbers of transgenic and wild type males as done previously with the C. hominivorax male-only strains (Concha et al., 2016). As previously, the MCI was calculated where an index of 0.5 indicates the transgenic and wild type males are equally competitive. Males from the original FL19 strain and from most of the new male-only lines were significantly out-competed by the wild type males (Table 6). However, males from line 36 and the combined 7 (X-linked) with the original FL19 strain competed effectively with the wild type males, with MCI values not significantly different than the expectation of 0.5 if fully competitive (Table 6).