JH signaling is high during late embryogenesis in T. domestica

To investigate the potential role of JH signaling during embryogenesis of T. domestica, we initially elucidated the embryonic development pattern and established a detailed morphological description of the embryo using scanning electron microscopy (SEM) and DAPI staining. As shown in Fig. 1A–F, T. domestica follows a typical short-germ band type of embryonic development (also see Additional file 1: Fig. S1), wherein only the most anterior cephalic and thoracic segments of the embryo are specified during the early syncytial blastoderm stage (embryonic day 2, ED2; A and A’). The abdominal segments are subsequently added in sequence from a posteriorly located undifferentiated growth zone (GZ) during germ band elongation (B, B’). From germ band formation (ED3), morphogenesis of the tissues and organs of each embryonic layer takes place, with the germ band giving rise to a fully segmented embryo by ED7 when dorsal closure ends (C–E). Morphogenetic growth and segmentation of the legs occur simultaneously. Differentiation and maturation occur from ED7, during which embryonic legs become plump and stout when viewed from the outside (E, F). After completing embryogenesis, the embryo hatches as the first instar juvenile during ED11–12. Our study is consistent with a recent observation (Truman et al., [39]).

Fig. 1

Correlation between the morphological changes of embryo and JH signaling in T. domestica. A–F Morphological pattern of embryo during embryonic development in T. domestica, as observed through scanning electron microscopy (SEM). Ventral (A) and lateral (A’) views of the newly formed germ band containing only the most anterior cephalic and thoracic segments at day 2 after egg laying (ED2). Ventral (B) and lateral (B’) views of the germ band with completed segmentation observed at ED3. Dorsal closure, progressing from posterior to anterior, initiates at ED5 (C), advances at ED6 (D), and concludes at ED7 (E). By ED7, morphogenesis and segmentation of the legs are completed, followed by differentiation and maturation of tissues and organs, including legs, until hatching. The shedding of the first embryonic (E1) cuticle occurs at ED9 (F). G Transcriptional dynamics of key genes involved in JH biosynthesis and signaling in whole eggs of T. domestica. FPKM (Fragments Per Kilobase of exon model per Million mapped fragments, which indicates gene expression level in RNA-seq) values of JHAMT, CYP15A1, Met, and Kr-h1 were obtained from RNA-seq data generated from wild-type eggs collected at ED1, ED3, ED5, ED7.ED9, and ED11, respectively. H Effect of JH III application and knockout (KO) of JHAMT, CYP15A1, and Met on Kr-h1 expression in whole eggs of T. domestica. Relative expression levels of Kr-h1 were detected using qPCR, with the ribosomal protein S26 (RPS26) as the reference gene. JH III (100 ng/egg) was applied at ED6, and treated eggs were collected 12 h post-treatment for qPCR analysis (the acetone treated eggs as controls). Gene-specific gRNAs (400 ng/μL) and Cas9 (300 ng/μL) were injected into the freshly laid eggs within 3 h, and eggs were collected at ED10 for qPCR analysis (the eggs injected with Cas9 as controls). All RNA-seq and qPCR data are presented as the mean ± s.e.m. (n = 3)

Next, we examined the expression levels of genes involved in JH biosynthesis and signaling during embryonic development in T. domestica using RNA-seq to establish a correlation between the morphological changes of the embryo and the presence of JH signaling. T. domesticaalready possesses the ability to produce JH III [9]. The transcripts ofJHAMT and CYP15A1 remained low (FPKM < 3) throughout the entire period of embryogenesis, with small peaks observed at ED9 and ED7, respectively. The mRNA abundance of Met was relatively high during the morphogenesis and growth period (ED3–ED7), while the expression level of Kr-h1 was massively upregulated from ED7 onward and remained high during differentiation and maturation period (ED7–ED11) (Fig. 1G).

We further investigated whether the JH signaling pathway known in metamorphic insects had already existed during embryonic development in ametabolous firebrats. Kr-h1 gene expression was significantly upregulated in the embryos of T. domestica after JH III treatment at ED6 compared with the acetone treated control. Conversely, CRISPR/Cas9-meditated gene knockout (KO) of JHAMT, CYP15A1, and Met respectively resulted in downregulation of Kr-h1 transcription level at ED10 in the mosaic mutants of G0 (generation 0) once compared with that in the controls (Fig. 1H).

Collectively, the JH signaling pathway primitively exists in embryogenesis of ametabolous insects, and JH signaling is highly activated during late embryogenesis, suggesting that JH signaling is functionally important for embryonic differentiation and maturation in ametabolous insects.

JH signaling-deficient embryos exhibit wrinkled and feeble legs in T. domestica

Subsequently, we examined the embryonic phenotypes of JH-deficient (KO-JHAMT, KO-CYP15A1) and JH signaling-deficient (KO-Met, KO-Kr-h1) mutants; genomic DNA modifications were confirmed via Sanger sequencing (Additional file 1: Fig. S2A–D). We observed a significant reduction in embryo hatchability (Fig. 2A); however, the timing of hatching was not delayed in the remaining G0 mosaic mutants of JHAMT, CYP15A1, Met, or Kr-h1 compared to the control population (Fig. 2B). These findings indicate that the absence of JH or JH signaling results in significant lethality in embryos of the ametabolous T. domestica. We dissected the eggshells to observe the developmental progress of the unhatched mutant embryos on embryonic day 14 (ED14). The phenotypes observed in the unhatched eggs were classified into four classes (refer to Additional file 1: Fig. S3): class I (embryos displaying apparent completion of embryogenesis without extraembryonic fluid); class II (embryos that had completed dorsal closure but failed to resorb their extraembryonic fluid); class III (embryos that had not completed dorsal closure, characterized by “big heads”); class IV (cloudy liquid devoid of embryos, including the absence of germ bands). The statistical analysis results showed that class II comprise of the highest proportion, followed by class I, in the embryos of KO-JHAMT, CYP15A1, Met, and Kr-h1 G0 mutants, respectively (refer to Fig. 2C). Preliminary phenotype examinations suggest that JH signaling plays a crucial role in the resorption of extraembryonic fluid and hatching during late embryogenesis in ametabolous insects.

Fig. 2

Effects of JH signaling deprivation on embryogenesis of T. domestica. Hatchability (A) and embryogenesis period (B) by the G0 mosaic mutants (including KO-JHAMT, KO-CYP15A1, KO-Met, and KO-Kr-h1). Hatched juveniles were counted at ED14, and the numbers above bars indicate the number of hatched/total embryos. The embryogenesis period was recorded at 12-h interval. C Phenotypes of unhatched embryos counted at ED14 according to classification grades (see Additional file 1: Fig. S3). Bars indicate mean ± s.e.m. for percentage (A, C, n = 4 or 5 batches) or days of embryogenesis (B, n = 310 individuals for control; 42 for KO-JHAMT; 127 for KO-CYP15A1; 74 for KO-Met; 28 for KO-Kr-h1) for each population. D–H’ Embryo phenotypes resulting from KO-JHAMT, KO-CYP15A1, KO-Met, and KO-Kr-h1 in T. domestica. Photos collected by optical microscope (D–H) and scanning electron microscopy (SEM) (scale bar = 200 μm) (D’–H’) are shown. D, D’ Normal freshly eclosed first instar juvenile. Class II embryos of JHAMT (E, E’), CYP15A1 (F, F’), Met (G, G’), and Kr-h1 (H, H’) G0 mutants. The legs were highlighted by red dashed lines

Furthermore, we conducted SEM examinations to investigate the specific external phenotypes of the mutant embryos classified as class II and class I at ED14. Under a light microscope, control embryos exhibited a white body with elongated and robust appendages (Fig. 2D). However, in class II embryos of KO-JHAMT, CYP15A1, Met, and Kr-h1 G0 mutants, which appeared yellowish, unabsorbed extraembryonic fluid adhered to the body, resulting in appendages such as legs becoming stuck (Fig. 2E–H). SEM results showed that the class II mutant embryos displayed shrunken and feeble legs when compared to those of control embryos (Fig. 2D’–H’). A small portion of class II embryos displayed even more pronounced morphological abnormalities than previously described: body structures were difficult to discern, and the legs were notably shrunken (Additional file 1: Fig. S4A–E1’). Mutant embryos at class I displayed a dry body surface with stretched-out but wrinkled legs (see Additional file 1: Fig. S4A–E2’). Collectively, the morphological defects observed in CRISPR/Cas9-mediated KO-JHAMT, CYP15A1, Met, and Kr-h1 indicate the indispensable role of JH signaling in leg maturation during late embryogenesis in ametabolous insects.

JH promotes leg maturation by inducing energy supply and muscle development

We conducted RNA-seq experiments to further explore the molecular mechanisms underlying JH-dependent leg maturation during embryogenesis in T. domestica. Embryonic legs were dissected from control specimens, as well as KO-JHAMT, CYP15A1, Met, and Kr-h1 G0 mutants, at both ED10 and ED11 (mixed in a ratio of 6:4) for RNA-seq analysis. Through four sets of comparative transcriptome analyses, we identified differential expression patterns, with 2414 downregulated and 2722 upregulated genes in KO-JHAMT vs. control, 2160 downregulated and 2542 upregulated genes in KO-CYP15A1 vs. control, 2166 downregulated and 2358 upregulated genes in KO-Met vs. control, and 2576 downregulated and 2325 upregulated genes in KO-Kr-h1 vs. control individuals. Notably, mutations in JHAMT, CYP15A1, and Met both resulted in significant downregulation of Kr-h1 expression level in the legs (Fig. 3A, C, E, G). KEGG enrichment analysis revealed a high consistency in significantly downregulated pathways across all four comparative analysis mentioned above (the top 15 pathways are listed in Fig. 3B, D, F, H), including those associated with energy supply (such as mitochondrial biogenesis, thermogenesis, oxidative phosphorylation, the TCA (tricarboxylic acid) cycle, and glycolysis/gluconeogenesis), muscle contraction, as well as some protein synthesis and degradation pathways (including ribosome biogenesis, ribosome function, and proteasome activity). In addition, pathways related to neural signal transduction, such as retrograde endocannabinoid signaling and the synaptic vesicle cycle, showed significant downregulation in embryonic legs upon suppression of JH signaling. Notably, these two neural signal transduction pathways were not included among the top 15 downregulated pathways in the KO-CYP15A1 vs. control and KO-Met vs. control groups (refer to Additional file 1: Tables S1–4 for a complete list of significantly downregulated pathways, with adjusted P values < 0.05). Similarly, the phenotype analysis revealed that there were more severe developmental defects in the KO-JHAMT or KO-Kr-h1 G0 mutants compared to KO-CYP15A1 or KO-Met (Fig. 2C). Similarly, the significantly upregulated pathways identified through KEGG analysis showed high consistency (refer to Additional file 1: Fig. S5A–D).

Fig. 3

Effects of JH signaling deprivation on the transcriptional profiles in embryonic legs of T. domestica. Volcano plots (A, C, E, G) and KEGG enrichment bubble maps (B, D, F, H; top 15 downregulated pathways) are generated from comparative analysis of embryonic leg transcriptomes in T. domestica. The embryonic legs were sampled at ED10 and ED11 (mixed in a ratio of 6:4) from control, KO-JHAMT, KO-CYP15A1, KO-Met, and KO-Kr-h1 embryos of G0 for RNA-seq. Differential expression analyses were conducted according to the schemes of KO-JHAMT vs. control (A, B), KO-CYP15A1 vs. control (C, D), KO-Met vs. control (E, F), and KO-Kr-h1 vs. control (G, H), respectively; genes with an adjusted P value < 0.05 (with no threshold for fold change) are designated as differentially expressed. Each group contained three independent biological replicates, and at least 200 embryos classified as class II or class I were pooled to generate one biological replicate. CRISPR/Cas9-mediated knockout of JH biosynthesis genes (JHAMT, CYP15A1) or JH receptor genes (Met) significantly suppressed JH signaling, as indicated by the expression level of Kr-h1 on the volcano plots. Oxidative phosphorylation, TCA cycle, and muscle contraction pathways are highlighted in blue on the bubble maps. I Venn diagram analysis of commonly downregulated, top 20 (fold change) annotated genes from the four groups of comparative transcriptomes. Genes that simultaneously appeared within at least three groups are listed, with two muscle activity-related genes, sarcoplasmic calcium-binding protein 1 and PDZ and LIM domain protein Zasp, highlighted in blue. Histological changes of the leg muscles from control (J), KO-JHAMT (K), and KO-Kr-h1 (L) embryos at ED11, as observed through transmission electron microscopy (TEM). Detailed magnifications of muscle fibers in J–L (red boxes) are shown in J’–L’. The Z-disks, muscle fibers, and mitochondria were indicated with the red, blue, and green arrows, respectively

Additionally, we conducted Venn analysis on the four sets of significantly downregulated genes, revealing 1195 genes that are commonly downregulated among all four groups (Additional file 1: Fig. S6A). Subsequently, the top 20 annotated genes, sorted by fold change in ascending order, from each group were extracted from the commonly downregulated gene set (Additional file 1: Fig. S6B), and these genes were then subjected to another Venn analysis (Fig. 3I). The genes that appeared simultaneously in three or four groups were listed in the figure: lipid droplet associated hydrolase, urea amide hydrolase, and cytochrome P450 genes involved in energy supply and material metabolism; spaetzle involved in immunity; arrestin homolog involved in GPCR signaling; retinol binding protein involved in retinoic acid signaling; sarcoplasmic calcium binding protein and PDZ and LIM domain protein Zasp associated with muscle structure and contraction. Both of the aforementioned analysis schemes highlighted the importance of muscle development and activity. Examination of the leg muscles via transmission electron microscopy (TEM) at ED11 showed that both KO-JHAMT and KO-Kr-h1 severely halted the progression of muscle fiber development, specifically, there were abundant and integrated muscle fibers between Z-disks in the legs of control animals, however, the two mutant specimens exhibited obvious deficiencies of the muscle fibers between Z-disks, thereby hindering leg maturation compared to the control (Fig. 3J–L’). In summary, the transcriptome and TEM results collectively indicate that JH signaling plays a pivotal role in energy supply and muscle development in the legs during the late embryogenesis in T. domestica.

JH signaling promotes embryonic leg maturation in G. bimaculatus

Next, we aimed to determine whether the fundamental role of JH signaling observed in the embryonic legs of T. domestica extends to hemimetabolous insects, which evolved from ametabolous ancestors. To address this, we employed G. bimaculatus as a model representative of hemimetabolous insects. The embryogenesis of G. bimaculatus lasts approximately 11–12 days, showing a good correspondence in the pattern of embryonic development timing compared to T. domestica. Similarly, germ band formation and dorsal closure are completed by ED3 and ED7 in G. bimaculatus, respectively. Subsequently, differentiation and maturation occur from ED7 until hatching [15]. The investigation into the expression patterns of JHAMT, CYP15A1, Met, and Kr-h1 during embryonic development in G. bimaculatus, using RNA-seq, revealed that all four genes exhibit expression peaks during differentiation and maturation period (Fig. 4A). By using CRISPR/Cas9-mediated genome editing, we conducted Kr-h1 loss-of-function experiments in G. bimaculatus (Additional file 1: Fig. S2E) to evaluate the potential roles of JH signaling in regulating embryonic development in hemimetabolous insects. The Kr-h1 G0 mutants of G. bimaculatus were unable to hatch, with the majority (~ 60%) arrested at class I (Fig. 4B, counted at ED13), similar to observations in T. domestica. The unhatched embryos exhibited swollen (femur) and twisted (tibia) legs, particularly the jumping legs on the third thoracic segment. Additionally, they showed cuticle tanning within eggshells (Fig. 4B’, B’’).

Fig. 4

JH signaling is essential for leg maturation in embryos of G. bimaculatus. A Expression of JHAMT, CYP15A1, Met, and Kr-h1 in embryos of G. bimaculatus. The FPKM values in whole eggs at different ages from ED0 to ED11 are presented. RNA-seq data are represented as the mean ± s.e.m. for three biological replicates. B–B’’ Embryo phenotypes resulting from CRISPR/Cas9-mediated knockout of Kr-h1 (G0 mutants) in G. bimaculatus. Statistical analysis of the embryos of control and KO-Kr-h1 on ED13 (G0, n = 3 batches) is shown in B, following the classification criteria in T. domestica. Phenotype observation of the control (B, hatched) and KO-Kr-h1 (B’, class I) embryo at ED11.5. The legs were highlighted by red dashed lines. C Four genes, which are included in the gene set (screened through Venn analysis in T. domestica, containing 13 genes, refer to Fig. 3I), were also significantly downregulated in the embryonic legs of Kr-h1 G0 mutants in G. bimaculatus compared with those in control (transcriptome data). PDZ and LIM domain protein Zasp is highlighted in blue. D KEGG bubble map (top 15 downregulated pathways) derived from comparative transcriptomes (KO-Kr-h1 vs. control) of embryonic legs in G. bimaculatus. Embryonic legs were collected at ED10 and ED11 (mixed in a ratio of 6:4) from control and G0 mutants for RNA-seq. Genes with an adjusted P value < 0.05 (with no threshold for fold change) are considered differentially expressed. Each group contained three independent biological replicates, and at least 200 embryos classified as class I were pooled to generate one biological replicate. Oxidative phosphorylation, TCA cycle, and muscle contraction pathways are highlighted in blue on the diagram. E, F TEM observation of leg muscles in control and KO-Kr-h1 embryos at ED11. Detailed magnifications of muscle fibers in E and F (red boxes) are shown in E’ and F’, respectively

Comparative transcriptome experiments were also utilized to explore the molecular mechanisms by which JH regulates embryonic leg maturation in G. bimaculatus. Embryonic legs dissected at ED10 and ED11 (mixed in a 6:4 ratio) from KO-Kr-h1 and control embryos were subjected to RNA-seq. Notably, among the 2159 downregulated genes in G. bimaculatus, 4 of them, including lipid droplet-associated hydrolase, cytochrome P450 6k1, cytochrome P450 6j1, and PDZ and LIM domain protein Zasp, were also among the 13 genes screened through Venn analysis in T. domestica (Fig. 4C, related to Fig. 3I); the top 15 pathways included energy supply pathways (thermogenesis, oxidative phosphorylation, and TCA cycle), cardiac muscle contraction, and the neural transduction related retrograde endocannabinoid signaling, mirroring pathways identified in T. domestica. In addition, other pathways involved in muscle development and activity were significantly downregulated in the embryonic legs of G. bimaculatus when JH signaling was inhibited, compared to the control, including adrenergic signaling in cardiomyocytes, the cGMP-PKG signaling pathway, vascular muscle contraction, and the calcium signaling pathway (Fig. 4D; Additional file 1: Fig. S7A, Table S5). However, the pathways that showed upregulation in the embryonic legs between the two species exhibited inconsistency when JH signaling was blocked (Additional file 1: Fig. S7A, B; Fig. S5). Since muscles are an important component of the legs, we further investigated the downregulated genes within the cardiac muscle contraction pathway enriched in both T. domestica and G. bimaculatus. We discovered 12 genes that were consistently downregulated in both species, including calcium transport proteins such as voltage-dependent L-type Ca2+ channel, Na+/Ca2+ exchanger, and sarco/endoplasmic reticulum Ca2+-ATPase, as well as muscle structural proteins like tropomyosin and myosin heavy chain, alongside certain respiratory chain complex proteins (Additional file 1: Fig. S7C). Examination of the leg muscles using TEM showed that the 11-day-old control embryos displayed intact muscle fibers, whereas the KO-Kr-h1 embryos exhibited obvious defects in their muscle fibers (Fig. 4E–F’). These findings suggest that JH signaling plays a conserved role in promoting embryonic leg maturation across both ametabolous and hemimetabolous insects.