Introduction

Arthropods commonly carry microbial symbionts that are passed from mother to offspring (Hurst & Frost, 2015; Werren et al., 2008; Hurst, 2017). The maternally transmitted bacteria, belonging to the genus Wolbachia (Alphaproteobacteria), are estimated to be present in at least 40% of all insect species, making them one of the most widespread endosymbiont genera on the planet (Werren et al, 2008; Zug & Hammerstein, 2012). Wolbachia have achieved evolutionary success by manipulating host reproduction through various means that enhance endosymbiont transmission (Werren et al., 2008). Such manipulation of the host’s reproduction includes cytoplasmic incompatibility (CI), parthenogenesis, feminization, and male killing (MK), each of which seemingly affects the biological features, distribution, and evolution of the host. Among these strategies, MK directly skews the sex ratio of the host population toward females by killing male offspring of infected mothers during development. The lack of symbiont transmission through male hosts often leads to the evolution of MK. Wolbachia have been shown to induce MK in a wide range of insect taxa (e.g., dipterans, lepidopterans, and coleopterans) and other arthropods (Hurst et al., 1999; Kageyama & Traut, 2004; Hurst et al., 2000). Furthermore, various bacteria, viruses, and microsporidia induce MK phenotypes (Kageyama et al., 2012; Fujita et al., 2021; Kageyama et al., 2023; Nagamine et al., 2023), and recent studies have postulated that these microbes have evolved their MK ability independently (Harumoto & Lemaitre, 2018; Kageyama et al., 2023; Nagamine et al., 2023; Arai et al., 2023a).

The evolution and molecular mechanisms underlying Wolbachia-induced MK have attracted considerable attention for decades. It has been hypothesized that the mechanisms underlying MK are associated with sex determination cascades in insects (Hornett et al., 2022). Indeed, some MK-inducing Wolbachia strains have been shown to directly interact with the sex determination system in lepidopteran insects (Sugimoto and Ishikawa, 2012; Sugimoto et al., 2015; Fukui et al., 2015; Arai et al., 2023a). The Lepidopteran sex determination system consists of multiple transcriptional regulators, some of which exhibit sex-linked expression and/or splicing isoforms. Lepidoptera generally have a female heterogametic sex chromosome system (e.g., WZ/ZZ) and employ dosage compensation, which equalizes the sex-linked (Z-chromosome-linked) gene dose between males and females (Kiuchi et al., 2014). Dosage compensation is regulated by the masculinizer gene (masc), which is critical for male development. masc also controls the downstream master regulator of sex determination and differentiation, doublesex (dsx), which exhibits sex-dependent splicing isoforms (dsxF in females and dsxM in males) (Kiuchi et al., 2014). In Ostrinia and Homona male moths, MK-inducing Wolbachia strains impair dosage compensation by disrupting the expression of masc. Furthermore, they disrupt sex determination in male moths by inducing the “female” isoform of dsx (dsxF), leading to a “mismatch” between genetic sex (male: ZZ sex chromosome constitution) and phenotypic sex (female: based on dsxF) and consequently to male death (Arai et al., 2023a; Fukui et al., 2015; Sugimoto et al., 2015; Sugimoto and Ishikawa, 2012).

More recently, a protein produced in Wolbachia wFur strain named Oscar (Osugoroshi protein containing CifB C-terminus-like domain and many Ankyrin Repeats; Osugoroshi translates to male killing in Japanese) was shown to recapitulate the Wolbachia-induced MK phenotype in Ostrinia (Katsuma et al., 2022). The Oscar protein degrades and interacts with Masc protein, leading to the failure of dosage compensation and the production of female-type dsx isoforms in Ostrinia male moths (Fukui et al., 2024; Katsuma et al., 2022). Although Oscar homologs have been identified in several MK-inducing Wolbachia strains in Lepidoptera, some MK Wolbachia strains in Diptera (Drosophila) and Lepidoptera do not carry this protein (Arai et al., 2023b; 2024a). Furthermore, Oscar homologs are evolutionarily dynamic, with highly variable sequences and structures (classified as Type I and Type II Oscar) (Arai et al., 2024a), making it difficult to assess their functional relevance. In addition to Oscar, the helix-turn-helix domain-containing putative transcriptional regulator Wmk (approximately 300 aa) is widely conserved among Wolbachia strains and induces various toxicities in Drosophila flies (i.e., no effect, weak male lethality [30% of males die], strong male lethality [90%–100% of males die], and death of all males and females) (Perlmutter et al., 2019, 2020, 2021; Arai et al., 2023b). Although the mechanisms underlying the Wmk-induced toxicities and their connection to sex determination systems remain unclear, these findings suggest that Wolbachia strains carry multiple factors that cause male lethality. However, the diversity and commonality of these functions in insects remain largely unknown, partly due to the technical challenges in validating gene functions in non-model insects.

We recently discovered an MK-associated prophage region that underlies the evolutionary transition from a non-MK Wolbachia (wHm-c) to an MK Wolbachia (wHm-t) in the tea tortrix moth Homona magnanima (Tortricidae) (Arai et al., 2023b; Arai et al., 2024b). The MK-associated prophage element encodes four wmk homologs (wmk-1 to wmk-4) as well as an oscar homolog (Hm-Oscar, 1181 aa), which differs significantly in sequence length and structure from the wFur-encoded Oscar (1830 aa). In Drosophila, Wmk-1 and Wmk-3 have a lethality of almost 100%, while Wmk-2, Wmk-4, and Hm-Oscar induce no lethal effects when singly overexpressed. Furthermore, co-expression of the adjacent Wmk-3 and Wmk-4 has been shown to induce the death of 90% of male flies and restores female survival (Arai et al., 2023b), suggesting that their combined action is similar to that observed in Wolbachia-induced CI (CifA and CifB, Beckmann et al., 2017; LePage et al., 2017). However, the mechanistic links between prophage-encoded Wolbachia genes and MK in the native host Homona remain unclear.

In this study, we showed that the prophage-encoded Hm-oscar recapitulates Wolbachia-induced MK in H. magnanima. Furthermore, we clarified the mechanistic links to host sex-determination cascades both in vivo and in vitro and discussed the underlying mechanisms of MK in Lepidoptera, arguing for the diverse evolutionary origin of Wolbachia-induced MK.

Results

Hm-Oscar induces female-biased sex ratios

To achieve the transient overexpression of Hm-oscar and the four wmk genes (wmk-1, wmk-2, wmk-3, and wmk-4), constructed mRNA (cRNA) was injected into Wolbachia-free H. magnanima embryos. Subsequently, the adult moths that emerged from the cRNA-injected embryos were sexed based on their external morphology. When Hm-oscar was overexpressed, the sex ratio of adults was strongly female-biased (85.6% ± 11.1%; 42 males and 218 females in total, 15 replicates), which was in sharp contrast with the ratios observed in the GFP-injected (50.9% ± 2.08%; 126 females and 123 males in total, 10 replicates) and non-injected (NSR) (48.6% ± 3.08%; 201 females and 215 males in total, 9 replicates) groups (P = 0.001 and 0.002, respectively, Steel–Dwass test, Fig. 1). Compared with the GFP-injected group, the sex ratio was not biased by the overexpression of wmk-1 (48.9% ± 7.27%, 6 replicates, P = 0.999), wmk-2 (46.8% ± 4.77%, 6 replicates, P = 0.810), wmk-3 (46.7% ± 7.91%, 5 replicates, P = 0.999), and wmk-4 (47.1% ± 3.47%, 5 replicates, P = 0.773). Although the dual expression of wmk-3 and wmk-4 induces strong male lethality in Drosophila (Arai et al., 2023b), here, it did not skew the sex ratio of Homona compared with the ratio detected in the GFP-injected group (42.1% ± 7.67%, 5 replicates, P = 0.704). Similarly, the dual expression of the tandemly arrayed wmk-1 and wmk-2 did not bias the sex ratio (50.6% ± 2.82%, 5 replicates) compared with the value in the GFP-injected group (P = 1.000).

Transient expressions of the Hm-oscar gene resulted in female-biased sex ratios

Male ratio of adult progeny obtained from cRNA-injected groups (wmk-1, wmk-2, T2A-brigged wmk-1 and wmk-2, wmk-3, wmk-4, T2A-brigged wmk-3 and wmk-4, GFP, and Hm-oscar; n = 5–15 independent replicates using different egg masses), the non-injected NSR line, and the MK wHm-t-positive WT12 line. The total numbers of adult females and males are shown at the bottom. Different letters indicate significant differences (Steel–Dwass test, P < 0.05). The dot plots show all data points individually.

Males are killed mainly during the embryonic stage

The sex of unhatched embryos and hatched larvae (neonates) was determined by karyotyping for W chromatin, with the presence and absence of this substance indicating females and males, respectively. In the Hm-oscar-injected group, the sex ratio of the hatched larvae (neonates) was strongly female-biased (21 females and 3 males, P = 0.0001 in the binomial test) (Fig. 2a). In contrast, that of unhatched mature embryos was male-biased (18 females and 38 males, P = 0.005), suggesting that males died mainly during the embryonic stage. Likewise, the sex ratios of hatched larvae (34 females and 6 males) and unhatched embryos (27 females and 47 males) of the MK wHm-t infected line (WT12) were also female- and male-biased, respectively (P = 0.013 and 0, respectively). In the non-injected group (NSR), the sex ratios of hatched larvae (28 females and 28 males) and unhatched embryos (23 females and 25 males) were not biased (P = 0.553 and 0.443, respectively).

Hm-Oscar induced lethality in male embryos with the female-type sex determination

(a) Sex ratio of the hatched larvae and unhatched embryos in the Hm-oscar-expressed, wHm-t-infected, and non-infected/expressed groups. Females and males were discriminated based on the presence or absence of W chromatin. The number of individuals is indicated in brackets. (b) Splicing patterns of the downstream sex-determining gene dsx of H. magnanima embryos (5 days post oviposition). Abbreviations: HmOsc, Hm-oscar injected group; GFP, GFP-injected group; WT12, wHm-t-infected line; NSR, non-infected/injected line. M and F indicate W chromatin-negative (ZZ: male genotype) and W chromatin-positive (ZW: female genotype) mature embryos, respectively. dsx-F and dsx-M represent female and male-specific splicing variants, respectively.

Female-type sex determination in male embryos that are destined to die

Splicing patterns of the downstream sex determinant dsx were assessed in each of the H. magnanima embryos that were Hm-oscar-expressed, GFP-expressed, wHm-t-infected, or non-expressed (i.e., non-injected, NSR). In GFP-expressed and NSR embryos, females and males (which were identified by the presence and absence of W chromatin, respectively) exhibited a female- and male-type splicing variant of dsx, respectively. However, both wHm-t-infected and Hm-oscar-expressed embryos induced female-type dsx splicing regardless of the presence or absence of W chromatin (Fig. 2b).

Hm-Oscar impairs dosage compensation in male embryos

RNA-seq analysis revealed that, in Hm-oscar-injected embryos, Z-linked genes (homologs on the Bombyx mori chromosomes 1 and 15) were more expressed in males than in females (Fig. 3a), which was not observed in the GFP-injected group (Fig. 3b). Similarly, as previously reported by Arai et al. (20-23a), high levels of Z-linked gene expression were also observed in wHm-t-infected males, but not in NSR males (Figure 3c,d).

Hm-oscar-overexpressed male embryos showed higher levels of Z-linked gene expression

(a-d) Normalized expression levels (TPM) and chromosomal distributions of transcripts in H. magnanima embryos. RNA-seq data of embryos (108 hpo) were used to make the following comparisons: Hm-oscar-injected males versus Hm-oscar injected females (a), GFP-injected males versus GFP-injected females (b), WT12 males versus WT12 females (c), and NSR males versus NSR females (d). The chromosome number for each H. magnanima transcript-derived contig was assigned based on Bombyx mori gene models. The X axis represents the chromosome number of B. mori (shown as chr01 to chr28), and chr01 and chr15 (highlighted in orange) correspond to the Z chromosome of H. magnanima (Arai et al., 2023a).

Hm-Oscar suppresses the masculinizing functions of lepidopteran Masc

To confirm whether Hm-Oscar suppresses the functions of the upstream male sex determinant Masc, their interactions were assessed by transfection using BmN-4 cells, as described in Katsuma et al. (2022). BmN-4 cells, which are derived from the female ovaries of B. mori (Grace, 1967), exhibit a female-type default sex determination (Kiuchi et al., 2014). In contrast with the control groups, which were transfected with the non-inserted pIZ/V5-His vector and showed this default sex determination, the cells transfected with the Hm-masc-inserted pIZ/V5-His vector exhibited the male-type sex determination, as evidenced by the increased expression levels of the male-specific splicing of the sex-determining gene BmImpM (Fig. 4). Furthermore, the masculinizing function of Hm-masc was suppressed by co-transfecting the Hm-masc/Hm-oscar-inserted pIZ/V5-His vectors, as manifested by the low expression levels of BmImpM. While type I oscar of the MK wFur strain also suppressed Hm-masc, the function of this gene was less active than that of Hm-oscar. Hm-oscar also suppressed the masculinizing function of masc genes derived from various lepidopteran insects (i.e., Ostrinia furnacalis, Spodoptera frugiperda, Bombyx mori, Trichoplusia ni, and Papilio machaon), suggesting its broad spectrum of action in Lepidoptera.

Hm-Oscar suppressed the masculinizing function of lepidopteran Masc

Relative expression levels of the male-specific BmImpM variant in transfected BmN-4 cells. The expression of BmImpM was normalized using the housekeeping gene rp49. The expression levels of BmImpM in the Masc and Hm-Oscar/Oscar co-transfected cells were normalized by setting each Masc-transfected cell as 100. The dot plots show all data points individually. Each experimental condition was replicated three times. Abbreviations: Hm, Homona magnanima; Bm, Bombyx mori; Of, Ostrinia furnacalis; Pm, Papilio machaon; Sf, Spodoptra frugiperda; and Tv, Trilocha varians.

Discussion

In this study, we showed that the phage-encoded Hm-oscar, but not wmk, induced male lethality and a female-biased sex ratio in H. magnanima. Furthermore, the overexpressed Hm-oscar impaired male sex determination in Homona, recapitulating the wHm-t-induced phenotypes. Cell-based assays confirmed that Hm-Oscar suppressed the masculinizing functions of Masc. These results strongly suggested that Hm-Oscar underlies the wHm-t-induced MK function in H. magnanima. The Wolbachia-encoded Oscar homologs identified so far are classified as type I and type II (Arai et al., 2024a). Although long ankyrin repeat sequences are postulated to be critical for the function of wFur-encoded Oscar (1830 aa, type I) (Katsuma et al., 2022a), our study revealed that the functions of Hm-Oscar (1181 aa, type II), which encodes fewer ankyrin repeats, were comparable with those of Oscar carried by wFur. Interestingly, type II Oscar is also present in the feminizing Wolbachia wFem in butterflies belonging to genus Eurema (Arai et al., 2024a). Oscar homologs, which inhibit the masculinizing function and induce female sex determination, may have a conserved function in Wolbachia-induced MK and feminization in Lepidoptera.

In contrast to the results of this study, we have previously demonstrated that the phage-encoded wmk, but not Hm-oscar, induces male lethality in D. melanogaster (Arai et al., 2023b). Although the means of expression are different (i.e., transient in Homona and transgenic in Drosophila), this finding highlighted the differences in the mode/range of action of Wolbachia genes between insect species. It has been hypothesized that microbes induce MK in insects by targeting molecular mechanisms involved in sex determination and differentiation (Hornett et al., 2022b). Sex determination systems in insects are diverse; for example, Lepidoptera (including H. magnanima) and Diptera (including D. melanogaster) do not share any known sex-determining genes other than dsx (Suzuki, 2018). The different outcomes in Homona and Drosophila are probably due to their different sex determination systems. Because Oscar interacts with and suppresses Masc, Hm-Oscar could induce mortality in Homona males that possess Masc, but not in Drosophila males that lack it. Considering that Oscar homologs are not present in known MK Wolbachia strains in dipteran insects (Arai et al., 2024c; Katsuma et al., 2022), the mechanisms (i.e., causative genes) of Wolbachia-induced MK probably differ between insect taxa (e.g., between Lepidoptera and Diptera). While the mechanisms underlying Wmk-induced lethality remain unclear, the distinct effects associated with this gene between Homona and Drosophila may also reflect their genetic backgrounds [e.g., presence/absence of host factor(s) that interact with Wmk]. In addition, Katsuma et al. (2022) reported that the Wmk homologs encoded by wFur did not affect the masculinizing function of Masc in vitro, indicating that Wmk likely targets factors other than Masc. Our results strongly suggested that Wolbachia strains have acquired multiple MK genes through evolution. An intense evolutionary arms race between Wolbachia and their hosts may have increased the diversity of MK-inducing genes in the Wolbachia genome.

Wolbachia-induced phenotypes are known to be influenced by the genetic backgrounds of hosts (Hornett et al., 2006; Sasaki et al., 2002; Veneti et al., 2012). Our study showed that the wHm-t-encoded Hm-Oscar suppresses the function of Hm-Masc in H. magnanima more efficiently than the wFur-encoded Oscar, suggesting that this Wolbachia factor has undergone evolutionary tuning to adapt to its natural host. However, the mere presence of Oscar and Wmk homologs does not ensure the expression of the MK phenotype. For example, the type I Oscar-bearing wCauA strain did not induce MK in Ephestia (Cadra) cautella collected around 2000, although it did induce MK when transferred to the closely related host Ephestia kuehniella (Sasaki et al., 2002). However, an MK phenotype, presumably induced by wCauA, was recorded in E. cautella around the 1970s (Takahashi & Kuwahara, 1970). These findings suggested the emergence of suppressor(s) against MK induced by Oscar-bearing wCauA in E. cautella. Virulence genes often undergo duplication and substitution under strong selective pressure (Hill et al., 2022; Jones & Dangl, 2006). An intense evolutionary arms race between Wolbachia and their hosts could have increased and diversified the MK-associated Wolbachia genes. Conversely, natural selection favors the rescue of males by suppressing the Wolbachia-induced reproductive manipulations (Hornett et al., 2006, 2014), which may involve changes in the sex determination system because Wolbachia strains frequently hijack host reproduction systems. In this context, Wolbachia and other MK-inducing microbes may be a hidden driver for the diversification of complex insect sex determination systems.

In this study, we clarified the conserved roles of the Wolbachia-encoded Oscar homologs in Lepidoptera and demonstrated that Wolbachia have evolved distinct MK mechanisms (through causative genes) in insect taxa. The diversification of phenotype/virulence-associated genes and the rampant horizontal transmission of phages carrying virulence genes between Wolbachia strains may have contributed to the outstanding success of this bacterial genus. In addition to oscar and wmk, Wolbachia may retain other uncharacterized genes that induce male lethality, and further studies on diverse Wolbachia–host systems are highly warranted. Our findings provide insights into the molecular mechanisms and evolutionary relationships between endosymbionts and their hosts, which may also contribute to the design of pest management strategies.

Materials and methods

Experimental model and subject details

Homona magnanima

A laboratory-maintained H. magnanima line with a normal sex ratio (NSR) that was negative for Wolbachia and other endosymbionts, was used in our experiments. This line was initially collected in Tokyo, Japan, in 1999 and has been maintained inbred for over 250 generations in the laboratory. Larvae were reared using the artificial SilkMate 2S diet (Nosan Co., Yokohama, Japan) at 25°C under a long photoperiod (16L:8D) (Arai et al., 2019).

The laboratory-maintained all female H. magnanima WT12 line, which was initially collected in Taiwan (Tea Research Extension Station, Taoyuan city) in 2015 (Arai et al., 2020), was also used in this study. This line was maintained for over 50 generations by crossing it with the males of the NSR line.

BmN-4 cell line

BmN-4 cells (provided by Chisa Yasunaga-Aoki, Kyushu University, and maintained in our laboratory) were cultured at 26°C in IPL-41 Insect Medium (Applichem, Darmstadt, Germany) supplemented with 10% fetal bovine serum.

Transient expression of MK-associated phage genes

(i) mRNA synthesis

Codon-optimized wmk genes (wmk-1 to wmk-4), conjugated wmk pairs using a T2A peptide (i.e., wmk-1-T2A-wmk-2 and wmk-3-T2A-wmk-4), and Hm-oscar genes synthesized by Arai et al. (2023b) were used for mRNA synthesis. These synthetic genes were ligated into the plasmid pIZ/V5-His (Invitrogen, MA, USA) using the NEBuilder® HiFi DNA Assembly kit (New England Biolabs, MA, USA) following the manufacturer’s protocol. The inserts of the vector (i.e., wmk-1, wmk-2, wmk-3, wmk-4, Hm-oscar, wmk-1-T2A-wmk-2, wmk-3-T2A-wmk-4, and GFP as a control gene) were amplified using KOD-one (TOYOBO, Osaka, Japan) with the primer set containing the T7 promoter described in Fukui et al. (2015) (i.e., pIZ-F-T7: 5′-TAATACGACTCACTATAGGGAGACAGTTGAACAGCATCTGTTC-3′ and pIZ-R: 5′-GACAATACAAACTAAGATTTAGTCAG-3′) under the following PCR conditions: 20 cycles at 98°C for 10 s, 62°C for 5 s, and 68°C for 15 s. The amplicons were purified using the QIAquick PCR purification kit (Qiagen, Hilden, Germany), and 100–200 ng DNA was used for mRNA synthesis. The capped mRNA (cRNA) with poly(A) tail was synthesized using the mMESSAGE mMACHINE® T7 Ultra kit (Invitrogen, MA, USA) following the manufacturer’s protocol with some modifications. In brief, the assembled transcription reaction (10 µL of T7 2X NTP/ARCA, 2 µL of 10X T7 reaction buffer, 2 µL of T7 enzyme mix, and 6 µL of linear DNA template diluted with nuclease-free water) was incubated at 37°C for at least 10 h to maximize RNA yields. After poly(A) tailing, cRNA was purified using ISOGEN II (Nippongene, Tokyo, Japan) and dissolved in up to 10 µL of nuclease-free water to achieve an RNA concentration of approximately 1500–2000 ng/µL. The synthesized cRNA was preserved at −80°C until further use.

(ii) Preparation of H. magnanima eggs

A total of 15 males and 10 females of H. magnanima were mated in a plastic box (30 cm × 20 cm × 5 cm) for 3–4 days. The collection of egg masses began the day after oviposition was confirmed. In brief, newly oviposited egg masses were collected at 30-min intervals during the dark period using red light (which the moths cannot perceive). Females started to oviposit eggs at least 4 h into the dark period and, within less than 30 min post oviposition (mpo), the egg masses were collected. The collection lasted until the start of the light period. The egg masses were then subjected to microinjection assays.

(iii) Inoculum preparation, microinjection, and maintenance of the injected embryos

A glass needle for microinjection was prepared using glass capillary GD-1 (Narishige, Tokyo, Japan) with a PC-10 puller (Narishige). The glass capillaries were pulled at two temperatures (first stage: 75°C, second stage: 65°C) using two heavy weights and one light weight. The movement position during the second heating stage was set to 3 mm (range 1–10 mm).

The cRNA solution was diluted in a buffer (100 mM potassium acetate, 2 mM magnesium acetate, and 30 mM HEPES-KOH; at pH 7.4) containing 0.2% (W/V) Brilliant Blue FCF (Wako, Osaka, Japan) to obtain an RNA concentration of 1000 ng/µL. Approximately 1–4 µL of dye-containing mRNA solution was aspirated into the glass needles (capillaries), and the edge of each needle was ground using Micro Grinder EG-402 (Narishige) at an angle of 20 degrees for 2 s.

The fresh egg masses of H. magnanima (collected at less than 30 mpo as described above) were put on a double-sided sticky tape (15 mm × 5 m T4612, Nitoms, Tokyo, Japan) and placed on glass slides. Under a Nikon SMZ1270 microscope (Nicon, Tokyo, Japan), the RNA solution (30–100 fL) was injected into individual eggs using an microinjector IM-400 (Narishige) (balance pressure set at 0.010–0.030; injection pressure set at 0.050-0.100) and a QP-3JOY-2R electric micromanipulator (MicroSupport, Shizuoka, Shizuoka).

The injected egg masses were maintained in a 90-mm plastic Petri dish fitted with a slightly wet filter paper. As high humidity interfered with the development of the embryos, the injected eggs (egg mass) were first maintained at a relative humidity of 30%–50% (0–3 days post injection) and then at a higher humidity until hatching (4–6 dpo, 60% relative humidity). The hatched larvae were reared separately with 1/2 ounces of SilkMate 2S (Nosan Co.) until eclosion. The adult H. magnanima moths that emerged from the cRNA-injected embryos were sexed based on their external morphology.

Sexing of embryos/neonates and RNA extraction

To verify the effect of the transient expression of Hm-oscar on sex determination in H. magnanima, Hm-oscar/GFP-expressed, non-injected, and wHm-t-infected mature embryos showing black head capsule (1 day before hatching, 5–6 days post oviposition, dpo) were dissected on glass slides using forceps, as described in Arai et al. (2022). Their Malpighian tubules were fixed with methanol/acetic acid (50% v/v) and stained with lactic acetic orcein for W chromatin observations. The remaining tissues not used for sexing were stored in ISOGEN II (Nippon Gene) at −80°C until subsequent extraction. In total, 12 males or females (confirmed based on the presence or absence of W chromatin) were pooled and homogenized in ISOGEN II to extract RNA as described in Arai et al. (2023a). In brief, 0.4 times the volume of UltraPure distilled water (Invitrogen) was added to the ISOGEN II homogenates, which were centrifuged at 12000 × g for 15 min at 4°C to pellet proteins and DNAs. The resulting supernatant was mixed with the same volume of isopropanol to precipitate RNAs; then, the resulting solutions were transferred to EconoSpin columns (Epoch Life Science) and centrifuged at 17,900 × g and 4°C for 2 min. The RNAs captured in the column were washed twice with 80% ethanol and eluted in 20 µL of UltraPure distilled water (Invitrogen).

Hmdsx detection

Sex-specific dsx splicing variants of H. magnanima were detected as described in Arai et al. (2023a). In brief, total RNA (100–300 ng) extracted from sex-determined mature embryos was reverse-transcribed using PrimeScript™ II Reverse Transcriptase (TaKaRa, Shiga, Japan) at 30°C for 10 min, 45°C for 60 min, and 70°C for 15 min. Then, cDNA was amplified using KOD-FX Neo (Toyobo Co., Ltd.) with the following two primers: Hmdsx_long3F (5′-TGCCTAAAGTGAAAACGCCGAGGAGCC-3′) and Hmdsx_Mrev (5′-TGGAGGTCTCTTTTCATCCGG-3′). The PCR conditions used were as follows: 94°C for 2 min, followed by 45 cycles of 98°C for 10 s, 66°C for 30 s, and 68°C for 30 s. The amplicons were subjected to electrophoresis on 2.0% agarose Tris-borate-EDTA buffer (89 mM Tris-borate, 89 mM boric acid, 2 mM EDTA) gels.

RNA-seq-based quantification of Z chromosome-linked genes

The effects on dosage compensation were assessed by measuring differences in gene expression, as described in Fukui et al. (2015) and Arai et al. (2023a). A total of 1.0 µg of the total RNA extracted from Hm-oscar or GFP-overexpressed mature embryos (108 hpo) was used to prepare mRNA-seq libraries via the NEBNext Poly (A) mRNA Magnetic Isolation Module (New England Biolabs) and the NEBNext Ultra II RNA Library Prep kit for Illumina (New England Biolabs) following the manufacturer’s protocol. The adaptor sequences and low-quality reads (Qscore <20) were removed from the generated sequence data [150 bp paired-end (PE150)] using Trimmomatic (Bolger et al., 2014). The trimmed reads were mapped to the previously assembled transcriptome database for H. magnanima (Arai et al., 2023a) using Kallisto (Bray et al., 2016) to generate the normalized read count data (transcripts per million, TPM).

The binary logarithms of the TPM differences between males and females belonging to each H. magnanima group (i.e., Oscar/GFP-expressed) were calculated to assess the fold-changes in gene expression levels between sexes. As described in Arai et al. (2023a), the transcriptome data of H. magnanima were annotated using the B. mori gene sets obtained from KAIKOBASE (https://kaikobase.dna.affrc.go.jp). The binary logarithms of the TPM differences between males and females in the B. mori chromosomes 1 to 28 were plotted. The expression data of the wHm-t-infected and non-infected groups were also calculated based on the transcriptome data included in Arai et al. (2023a).

Transfection assays and quantification of BmIMP

The coding sequence of Hm-masc was amplified from cDNA derived from the RNA extracted from male embryos of the NSR line using KOD-one (TOYOBO) with the following primer set: HmMasc_CDS_HindIIIF: 5′-GCAAAGCTTCAACATGATCTCTCGCCAACCACAATCAACATCA-3′ and HmMasc_CDS_BamHI R: 5′-GCAGGATCCCAACCTACTGATAAGGAGGGAAGTAAGGCTGCTG-3′. The following

PCR conditions were applied: 20 cycles of 98°C for 10 s, 62°C for 5 s, and 68°C for 15 s. The amplicon was purified with the QIAquick PCR purification kit (Qiagen) and cloned into the pIZ/V5-His vector using the HindIII and BamHI enzymes (New England Biolabs). The codon-optimized oscar gene of wFur (Katsuma et al. 2022) was also cloned into the pIZ/V5-His vector using the KpnI and NotI enzymes (TaKaRa).

To verify the masculinizing function of Hm-masc, BmN-4 cells (4 × 105 cells per dish, diameter 35 mm) were transfected with 1 µg of plasmid DNA (pIZ/V5-His having Hm-masc) using FuGENE HD (Promega, WI, USA), as described in Katsuma et al. (2022). To clarify whether Hm-oscar suppressed the function of Hm-masc, 1 µg of plasmid DNA (pIZ/V5-His bearing either Hm-masc or Hm-oscar) was co-transfected to the BmN-4 cells using FuGENE HD (Promega). Three days after transfection, the cells were collected and subjected to RNA extraction via TRI REAGENT® (Molecular Research Center Inc., USA) and cDNA construction with AMV transcriptase (TaKaRa). The degree of masculinization in the BmN-4 cells (default female-type sex determination) was verified by quantifying the expression levels of BmImpM, which is involved in male-specific sex determination cascades, using primers rp49_F: 5′-CCCAACATTGGTTACGGTTC-3′ and rp49_R: 5′-GCTCTTTCCACGATCAGCTT-3′; BmIMP_F: 5′-ATGCGGGAAGAAGGTTTTATG-3′ and BmIMP_R: 5′-

TCATCCCGCCTCAGACGATTG-3′, as described in Fukui et al. (2015). Further interactions between Hm-Oscar and Masc proteins derived from lepidopteran insects [i.e., Trilocha varians Masc (TvMasc), Spodoptra frugiperda Masc (SfMasc), B. mori Masc (BmMasc), O. furnacalis Masc (OfMasc), and Papilio machaon Masc (PmMasc)] (Katsuma et al., 2022a) were assessed using the same procedures.

Quantification and statistical analysis

The number of surviving H. magnanima injected with either Hm-oscar, wmks, or GFP in each replicate were counted at the adult stage. The male ratios (number of adult males/numbers of all adults) under all conditions were compared using the Steel– Dwass test in R v4.0. The sex ratio bias was also assessed based on the total numbers of (i) male and female adults and (ii) male and female embryos under each condition using the binomial test in R v4.0.

Data and code availability

High-throughput sequencing data are available under accession numbers DRA018708 and PRJDB18169 (BioProject). All data generated during this study are included in the manuscript and supporting files. Further information and requests for resources and reagents are accessible from the lead contact, Hiroshi Arai (dazai39papilio@gmail.com/HArai@liverpool.ac.uk).

Acknowledgements

We thank Greg Hurst (Institute of Infection, Veterinary and Ecological Sciences, University of Liverpool, Liverpool, UK) and Takeshi Suzuki (Graduate School of Bio-Applications and Systems Engineering, Tokyo University of Agriculture and Technology, Tokyo, Japan) for their kind advice.

We acknowledge support from the Japan Society for the Promotion of Science (JSPS) Research Fellowships for Young Scientists [Grant Number 19J13123 and 21J00895 to H. Arai], JSPS Grant-in-Aid for Scientific Research [Grant Number 22K14902 to H. Arai, 23H02229 and 24H02293 to D. Kageyama, 22H00366 to S. Katsuma], JSPS Fund for the Promotion of Joint International Research (Fostering Joint International Research (B)) [Grant Number 21KK0105 to H. Arai and MN. Inoue].

Author contributions

H. Arai conducted gene functional validation assays, transcriptome assays, and data analysis; designed experiments; wrote the original manuscript; and revised the manuscript. S. Katsuma conducted transfection assays, quantified the gene expressions in the cells and presented the entire discussion. N. Matsuda-Imai constructed the pIZ/V5 vector having oscar gene derived from Ostrinia furnacalis. S-R. Lin revised the manuscript and contributed to the discussion. MN Inoue revised the manuscript and contributed to the discussion. D. Kageyama organized the project and revised the manuscript. Lastly, H. Arai and D. Kageyama took responsibility for the decision to submit the manuscript for publication and managed the experiments and discussion.

Declaration of interests

The authors declare no competing interests.

Inclusion and diversity

We support inclusive, diverse, and equitable conduct of research.

Benefit-Sharing

The wHm-t-infected H. magnanima was collected from Tea Research and Extension Station (Taoyuan City, Taiwan), and imported with permission from the Ministry of Agriculture, Forestry and Fisheries (No. 27 - Yokohama Shokubou 891 and No. 297 - Yokohama Shokubou 1326). All collaborators are presented as co-authors, and the results have been shared with the provider communities. Moreover, our group is committed to international scientific partnerships as well as institutional capacity building. The authors declare no competing interests.