Dual function of the transcription factor Ftz‐f1 on oviposition in the cockroach Blattella germanica

The transcription factor Ftz‐f1 has multiple functions in insect development in a spatial‐temporal line. One of these roles is in the insect ovaries, specifically in the regulation of steroidogenic enzymes production. We studied the function of F in Blattella germanica oogenesis, as it shows two moments of high expression in ovaries: before the imaginal moult, and just before ovulation in the adult. Injecting dsftz‐f1 into adult females, either just after the imaginal moult or just prior to choriogenesis, prevented oviposition, with differences between the two approaches. In 3‐day‐old adult females treated with dsftz‐f1 just after the emergence, the expression of ftz‐f1 was not modified, but the steroidogenic genes increased their expression. ftz‐f1 transcript levels in the ovaries of 5‐day‐old dsftz‐f1‐treated females were significantly depleted, and the expression levels of the same steroidogenic genes began to decrease. These results suggest that Ftz‐f1 regulates the expression of steroidogenic genes in B. germanica, with phm possibly being a key target. Ftz‐f1 has a different temporal function in the cytoskeleton of follicular cells of the basal ovarian follicles. Early in the gonadotrophic cycle, Ftz‐f1 promotes the expression of genes related to the cytoskeleton and muscle proteins, while at the end of the cycle it maintains the expression levels of these genes, thus ensuring correct ovulation.


INTRODUCTION
Insect oogenesis is a highly regulated process that involves the development and maturation of eggs within the female reproductive system.In insects it consists of multiple steps and stages, including germ cell proliferation, oocyte growth and maturation, vitellogenesis and eggshell formation.All of these processes are under the control of two hormones: the juvenile hormone (JH) and ecdysone.
Ecdysone biosynthesis is regulated by a series of enzymatic reactions which convert cholesterol to ecdysone (Niwa & Niwa, 2014).The first gene of the pathway is neverland (nvd) that encodes for an oxygenase-like protein, while the rest of the genes belong to the family of cytochrome P450 enzymes, including spook (spo), spookier (spok), spookiest (spot) shadow (sad), disembodied (dbd) and phantom (phm).These genes are involved in the synthesis of ecdysone, whereas shade (shd) is involved in its transformation into 20-hydroxyecdysone (20E), the bioactive form of the hormone.The precise role of each steroidogenic gene in insect oogenesis may vary depending on the species and context.However, their overall contribution lies in the biosynthesis of ecdysone, which is key for metamorphosis and egg formation.In nymphs, ecdysone is synthesised and secreted by the prothoracic glands, whereas in adults, after the degeneration of the prothoracic gland, the main source of ecdysteroids is the ovary (Ameku et al., 2017;Belles & Piulachs, 2015;Romaña et al., 1995).
In Drosophila melanogaster, the biosynthesis of ecdysteroids in the ovary is essential for germinal stem cell (GSC) proliferation and maintenance.Knockdown of nvd reduces the ovarian ecdysteroid levels in adult females, impairing GSC proliferation and maintenance, although the number of egg chambers is not affected (Ameku et al., 2017).In adult females of the locust Schistocerca gregaria, depletion of spo, phm (Marchal et al., 2011), dib, sad and shd (Schellens et al., 2022) expression lowers circulating ecdysteroid levels, which results in changes in the size and morphology of the eggs, reduction of number of oviposited eggs and number of emerging nymphs (Schellens et al., 2022).In the hemipteran Diaphorina citri, inhibition of sha and shd impairs ovarian development (Zhang et al., 2023).In general, knockdown of steroidogenic genes in adult females decreases the 20E pulse, which impairs ovarian development, and prevents oocyte maturation.
The ecdysone signalling begins when 20E binds to its receptor, a heterodimer formed by the ecdysone receptor (EcR) and ultraspiracle (USP) (Fahrbach et al., 2012;Niwa et al., 2010).This binding activates the early-genes in the 20E signalling cascade, which includes BrC, E74 and E75, which in turn activates the late-genes (like HR3, HR4 and ftz-f1).All these transcription factors activate the cascade and, at the same time, can inhibit their own expression (Fahrbach et al., 2012).
In insects, the transcription factors involved in JH and ecdysone signalling have been thoroughly studied in relation to metamorphosis.However, there is much less data available on its functions associated with oogenesis, even less in insects with panoistic ovaries, like Blattella germanica.One of these transcription factors is Fushi tarazu factor 1 (Ftz-f1), which plays essential roles in insect development, particularly in regulating the timing of moulting and metamorphosis (Lavorgna et al., 1993).The expression of ftz-f1 has been related with the JH levels in corpora allata cells from last instar nymphs of B. germanica, where Ftz-f1 is necessary to ensure the strong activation of JH synthesis in adults (Borras-Castells et al., 2017).In addition, it is known that in some insect species Ftz-f1 is involved in ecdysteroid synthesis, promoting the expression of steroidogenic genes (Gu et al., 2021;Parvy et al., 2005;Talamillo et al., 2013).Moreover, Ftz-f1 is required to complete oogenesis in different insects, like D. melanogaster (Beachum et al., 2021;Knapp et al., 2020), Aedes aegypti (Zhu et al., 2003), Apis mellifera (Mello et al., 2018) and Henosepilachna vigintioctopunctata (Liu et al., 2022).
Ftz-f1 has critical functions linked to the coordination of physiological processes involved in oogenesis, including hormone signalling, gene expression and nutrient acquisition, all of which are necessary for proper egg development.Finally, Ftz-f1 also plays a role in ovulation, that is, the release of mature eggs from the ovaries (Knapp et al., 2020).
Two isoforms of Ftz-f1 have been described in dipterans like D. melanogaster (Lavorgna et al., 1993) and A. aegypti (Cruz et al., 2009), as well as in some coleopterans and lepidopterans (Wu et al., 2022).However, a single ftz-f1 has been identified in the genome of the cockroach B. germanica (Harrison et al., 2018), which plays critical functions during the last nymphal moults (Cruz et al., 2008).Moreover, according to the transcriptomes available in our laboratory, ftz-f1 transcripts are maternally provided to the embryo, where they would play roles in early embryogenesis (Ylla et al., 2018).Recently, more data has been reported on the function of Ftz-f1 in the B. germanica embryo (Cruz et al., 2023;Wexler et al., 2023), on the basis of the small number of embryos that actually develop after maternal RNAi targeting ftz-f1 transcripts.Now, we report the function of Ftz-f1 in B. germanica oogenesis, in particular addressing the question of why Ftz-f1-depleted females do not oviposit, although the oocytes grow and mature.
The role of Ftz-f1 in the meroistic polytrophic ovary of D. melanogaster was described by Knapp et al. (2020), where ftz-f1 expression is induced at the end of the gonadotrophic cycle by ecdysone signalling.Importantly, a disruption of ftz-f1 expression impairs follicle cell differentiation, and prevents oviposition.To understand the function of Ftz-f1 in the panoistic ovary of B. germanica, and its possible relationship with 20E, we studied the function of this transcription factor in adults, since the ovary is responsible for synthesising the 20E in this stage (Romaña et al., 1995).
One characteristic of B. germanica oogenesis is that the basal oocyte is the only one that matures in each gonadotrophic cycle.
Once released from the germarium, the oocyte is surrounded by a monolayer of follicular cells, establishing an ovarian follicle (Figure 1a).
The ovarian follicles remain in the vitellarium waiting to take up the basal position through successive gonadotrophic cycles, when the basal one will be ovulated and oviposited.The follicular cells surrounding the oocytes mature and change its characteristics accompanying the oocyte growth (Figure 1b).It is in the sixth (last) nymphal instar, when the number of ovarian follicles in the ovarioles is established (Ramos et al., 2020), and basal ovarian follicles (BOFs) start to grow.During all the instar the follicular cells proliferate (Figure 1c), and the mitotic activity is maintained until the beginning of the vitellogenic period in adults (3-day-old females) (Irles & Piulachs, 2014).In these 3-day-old adult females, the JH titre in the hemolymph starts to increase, and the vitellogenin begins to be synthesised by the fat body.In the follicular cells of the BOF cytokinesis is arrested, thus they become binucleated (Figure 1c).Once binucleated, the cells contract the cytoplasm, leaving large intercellular spaces (Figure 1d, asterisks), which allows the storage proteins, mainly vitellogenin, to reach the oocyte membrane, a process known as patency (Davey, 1981;Davey & Huebner, 1974).In late vitellogenesis, the follicular cells enter to endocycle (Figure 1c).All of that results in a BOF that grows exponentially.At the end of vitellogenesis (Figure 1c) the follicular cells close the intercellular spaces, and there is a specific amplification of genes coding for chorion proteins (Irles & Piulachs, 2011;Irles et al., 2009;Irles, Bellés, & Piulachs, 2009).At this moment, the apical follicular cells in the BOF change their morphology and become specialised in building the sponge-like body, the structure that contains the micropyle which conveys oxygen to the future embryos (Irles, Belles, & Piulachs, 2009;Irles, Bellés, & Piulachs, 2009;Lawson, 1951;Wigglesworth & Beament, 1950).Chorion deposition is a process that takes a few hours with important and fast changes in the expression levels of the genes related to chorion synthesis.At the end of choriogenesis, there is a reorganisation of actin fibres in the follicular cells.
The actin fibres are arranged in parallel filaments in the surface of the basal pole of the follicular cells, as a muscle-like organisation with the actin bundles cross-striated (Zhang & Kunkel, 1992).A distribution appropriate to help in the ovulation of the BOF coordinated with the extracellular membrane (Zhang & Kunkel, 1992).The distribution of the cytoskeleton fibres and their role in the changes of the follicular cells during oogenesis has also been studied in insect species with meroistic ovaries (Knapp et al., 2020), but their function and regulation are not fully understood.
To elucidate the function of Ftz-f1 in B. germanica ovaries, we performed two types of treatments: in newly emerged adult females and in vitellogenic females.According to the respective results of these two treatments, we can define different functions of Ftz-f1 during oogenesis.We can conclude that Ftz-f1 regulates the expression of steroidogenic enzymes in the B. germanica panoistic ovaries, just as it happens in meroistic ovaries.However, Ftz-f1 can either stimulate or inhibit the expression of the steroidogenic enzymes, depending on the treatment time.Moreover, the cytoskeleton at the end of the cycle is dramatically affected, which prevents ovulation.Once again, the results are different depending on the moment of treatment, either increasing the expression of the genes involved in the formation of the cytoskeleton or decreasing them.

RESULTS
dsftz-f1 treatment in the early adult stage prevents oviposition in B. germanica In B. germanica ovaries, ftz-f1 is expressed through the gonadotrophic cycle (Figure 2a).In sixth (last) instar nymphs, ftz-f1 expression gradually increases until Day 4. Later on, and coinciding with the highest levels of ecdysteroids in the hemolymph, ftz-f1 expression in the ovaries decreases, increasing again when the ecdysteroid levels begin to drop, subsequently reaching the highest levels on Day 7, the last day of the instar, just before moulting to adult.In the adult, the expression of ftz-f1 in ovaries slowly declines, but increases suddenly and dramatically on Day 7, coinciding with the highest levels of ecdysone in the ovaries, and with choriogenesis in the BOF.
To discriminate the regulatory function of Ftz-f1 throughout the gonadotrophic cycle, we designed two types of treatments.Injecting newly emerged females (early-treated females) with dsftz-f1 in order to observe the early regulatory functions of this transcription factor, or injecting the dsftz-f1 to 5-day-old females (late-treated females) to affect more precisely the last peak of ftz-f1 expression.
In the first type of experiments, early-treated females were kept with males to promote mating.The dsPolyH-treated control females successfully oviposited 7 days after the treatment (Table 1), and carried the ootheca throughout the entire embryogenesis, as expected.
Moreover, the sponge-like body, the chorion respiratory device (Irles, Belles, & Piulachs, 2009) formed on 7-day-old females at the end of choriogenesis process, was evident in most of the ovarian follicles of the dsftz-f1-treated females (Figure 2b,c), although we do not know if it was functional.Nevertheless, the BOFs, which had a soft appearance, did not enter in the oviduct at the end of the gonadotrophic cycle, being finally reabsorbed.
We measured ftz-f1 mRNA levels in the ovaries of dsftz-f1-early treated adult females at different ages.The dsRNA treatment applied to 0-day-old adult females did not appear to affect ftz-f1 expression when measured 3 days later.However, a significant reduction in the ftz-f1 expression was observed 5 days after the treatment.The low levels were maintained until the end of the first gonadotrophic cycle (Figure 2d), when ftz-f1 expression was around 96% lower than in the F I G U R E 2 ftz-f1 in Blattella germanica ovaries.(a) ftz-f1 expression pattern in ovaries in sixth instar nymphs and adult females during the first gonadotrophic cycle; profiles of ecdysteroid titers in hemolymph (green dashed line) and ovarian ecdysteroid content (orange line) are indicated (data from Pascual et al., 1992;Romaña et al., 1995).(b-c) Apical pole in a basal ovarian follicle (BOF) from a 7-day-old adult, showing the sponge-like body (SLB) in dsPolyH (B) and dsftz-f1-treated (c) females, treated at Day 0, and dissected on Day 7. (d) ftz-f1 expression levels in ovaries of 3-day-, 5-day-and 7-day-old adult females treated with dsftz-f1 on the day of emergence.The asterisks indicate significant difference with respect to the controls: *p = 0.02 and **p = 0.004).(e) Expression levels of citrus, brownie (brw) and yellow-g (y-g) measured in ovaries from 7-dayold adult females, treated with dsftz-f1 on the day of emergence.In (a) and (d), the data represent copies of mRNA per 1000 copies of actin-5c; in (e) the data represent copies of mRNA per copy of actin-5c.Data are expressed as the mean ± S.E.M. (n = 3-5), and were evaluated for normality using the Shapiro-Wilk test.All datasets passed normality test.The statistical analysis was performed using Student's t-test.The asterisk indicates a significant difference with respect to the controls: *p = 0.03 and **p = 0.053.controls.Given that the highest levels of ftz-f1 expression in the adult ovaries were observed in Day 7 (Figure 2a), coinciding with choriogenesis, we conjectured that Ftz-f1 might be involved in the formation of the eggshell.For this reason, we measured three of the main chorion genes of B. germanica: yellow-g (y-g), which is expressed throughout choriogenesis (Irles, Bellés, & Piulachs, 2009); citrus, a mid-late chorion gene (Irles & Piulachs, 2011) and brownie (brw) a late chorion gene (Irles, Belles, & Piulachs, 2009).In dsftz-f1-early treated females, a clear depletion of the three chorion genes was observed (Figure 2e), although the reduction was not statistically significant in the case of brw, which was depleted at 68% as average.Conversely, the expressions of citrus and y-g were significantly reduced ( p = 0.03, 52% reduction, and p = 0.0003, 99%, respectively), (Figure 2e).These results support the idea that Ftz-f1 is generally involved in the formation of the eggshell in adult females of B. germanica.

dsftz-f1 treatment in early adults modifies the expression of steroidogenic genes in the ovaries
In adult B. germanica, JH contributes to vitellogenesis and follicular cells patterning (Irles & Piulachs, 2014), whereas 20E promotes choriogenesis (Pascual et al., 1992).This function of 20E, together with the previous observation that Ftz-f1 regulates steroidogenesis in the prothoracic gland of D. melanogaster (Parvy et al., 2005), led us to study the expression of steroidogenic genes in the ovaries of dsftz-f1early treated females.
We measured mRNA levels of nvd, spo, phm and shd, in ovaries of 3-day-old (Figure 3a), 5-day-old (Figure 3b) and 7-day-old (Figure 3c) dsftz-f1-treated females.In 3-day-old dsftz-f1-treated adult females, the expression of nvd and spo in the ovaries was not significantly different to the controls (Figure 3a).However, phm expression was significantly upregulated ( p = 0.001), and shd expression tended to increase.Later, in 5-day-old dsftz-f1-treated females, the mRNA levels of nvd and spo were similar to those of the controls, while the expression of phm was reduced ( p = 0.04) and that of shd tended to be reduced (Figure 3b).Just at the end of the gonadotrophic cycle, in 7day-old adult females, the significant reduction of ftz-f1 observed in the ovaries (Figure 2d) resulted in changes in the expression of the steroidogenic genes.The expression of nvd tended to increase (p = 0.05) coinciding with a significant depletion of spo (78%; p = 0.02), phm (64.95%; p = 0.047) and shd (81.64%; p = 0.02) (Figure 3c).The depletion of the expression of these three genes suggests that ecdysone synthesis was reduced at the end of the gonadotrophic cycle in the ovaries of dsftz-f1-early treated females.In particular, shd depletion could imply changes in ovarian 20E signalling at the end of the gonadotrophic cycle, as this enzyme transforms ecdysone into the biologically more active 20E form.
We then measured the expression of EcR in the ovaries of dsftz-f1-treated females (Figure 3d).In 3-day-old females the expression of EcR was not affected.However, in 5-day-old females, ftz-f1 depletion resulted in a significant reduction in EcR expression (43.17%; p = 0.007), which was even higher in ovaries from 7-dayold treated insects (83.48%; p = 0.02) (Figure 3d).Together, the decreased expression of shd and EcR contributed to prevent ecdysone signalling in the ovaries at the end of the cycle.Following this idea, we also measured the expression of E75A (Figure 3e), an early gene of the ecdysone cascade, also in B. germanica ovaries (Ramos et al., 2020).In 3-and 5-day-old dsftz-f1-treated females, the expression of E75A was low and not significantly different from that of the controls (Figure 3e), although in 3-day-old treated females the E75A levels were 50% higher as average than in controls.Later, in 7-day-old dsftz-f1-treated females, E75A was depleted around 96% ( p = 0.011) (Figure 3e), suggesting that the ecdysone signalling was impaired at the end of the gonadotrophic cycle.This result agrees with the reduced expression of chrorion genes in dsftz-f1 treated females (Figure 2e), as these genes are ecdysonedependent.
The increased expression of phm and shd in 3-day-old females, together with the absence of changes in EcR, and the tendency to increase E75A expression, may explain, at least in part, the significant increase in number of ovarian follicles observed in the ovarioles of dsftz-f1 treated females ( p = 0.03) (Figure 3f-h).This suggests that 20E production was increased in these females, since the number of ovarian follicles in the vitellaria of B. germanica is positively correlated with 20E levels (Ramos et al., 2020).Note: A dose of 1 μg of dsftz-f1 or dsPolyH was injected into the abdomen of newly ecdysed females.Females were maintained with males until oviposition is done, and the presence of spermatozoa in the spermatheca was assessed at the end of the experiment.The length of the basal ovarian follicle (BOF), the days that a female takes to oviposit (from the day of adult emergence), the number of mated females, the number of females that oviposit and form the ootheca, the days that the ootheca was transported and number of nymphs that hatch were recorded.Data were represented as mean ± S.M.E.dsftz-f1 treatment in early adult stage affects the development and structure of the follicular cells Although in dsftz-f1-early treated females the BOFs incorporated vitellogenin and grew, and at least some chorion layers were synthesised, they were fragile when manipulated, and were never oviposited.
Hence, we examined the follicular epithelium in the BOFs of dsftz-f1early treated females.In B. germanica adult females, the follicular cells of the BOFs are binucleated at the end of the gonadotrophic cycle (Figure 4a), while patency becomes closed.In dsftz-f1-treated females, the follicular cells in the BOFs were also binucleated, but the nucleus size was heterogeneous (Figure 4b), and most of them were smaller than in the controls (compare Figure 4a,b).
We also observed an effect of dsftz-f1 on the distribution of the F-actin fibres in the basal pole of the follicular cells (Figure 4c, arrow).
In the controls, and towards the end of the gonadotrophic cycle, the actin fibres cover the basal pole of the cells establishing intimate connections between them (Figure 4d), packing the entire ovarian follicle.
Just before ovulation, the actin fibres enwrap the cell like small balls, which give them a rounded shape (Figure 4e).At this moment, the eggs are ready to be ovulated while follicular cells enter in apoptosis.
The structure of the cytoskeleton in the follicular cells surrounding the egg appears essential to facilitate the egg drop through the oviduct.In dsftz-f1-treated females, the F-actin fibres also increased at the basal pole of follicular cells (Figure 4f), although the fibres between the cells were looser, and the general packaging was not as tight as in the controls (Figure 4g), which probably impairs ovulation.
We found that the actin-5c in ovaries from dsftz-f1-treated females was upregulated 2.9 times compared to the controls ( p = 0.0002) (Figure 5a).Accordingly, we estimated the mRNA expression of the different genes studied under this treatment using eif4aIII as a reference gene.
The expression of ftz-f1 in ovaries from 7-day-old females was significantly depleted (around 78%; p = 0.027) compared to the controls (Figure 5b).In these dsftz-f1 late-treated females, the expression of nvd, spo did not change significantly, while phm and shd expression tended to be downregulated (38% and 50%, respectively; Figure 5c).In agreement with this data, E75A expression was significantly reduced (66.66%;p = 0.006) (Figure 5d).Conversely, the depletion of ftz-f1 mRNA levels did not modify the expression of EcR (Figure 5e).We also measured brw, citrus and y-g expression.brw increased its expression with a threefold change, whereas citrus showed no differences with respect to the controls.In contrast, the expression of y-g was significantly reduced (97.9%;p = 0.0008) (Figure 5f).
In 5-day-old control females, that is, the moment when the females were treated with dsftz-f1, the follicular cells in the BOFs change their programme (Irles et al., 2016;Irles & Piulachs, 2014).The mitotic cycle finishes, all the follicular cells are binucleated (Figure 6a), and they start an endocycle with the endoreplication of some genome regions, which enhances the expression of chorion genes and the synthesis of the chorion layers that culminates in 7-day-old females.In dsftz-f1 late-treated 7-day-old females, the follicular cells did not present a uniform distribution.They were also binucleated, but the nuclei were bigger than in the controls, and presented odd morphologies (Figure 6b).This suggests that dsftz-f1-treated insects increased the endocycles in the follicular cells of the BOFs.
In addition, in these dsftz-f1 late-treated females, the distribution of F-actin fibres in the basal pole of the follicular cells presented diverse phenotypes.In some BOFs, the F-actin distribution was similar to that of dsPolyH-treated females (Figure 7a,b).However, the size of the follicular cells was not uniform (Figure 7c, asterisks) and the F-actin fibres were not tightly packaged on the cell surface (see Figure 7g, arrow a), leaving spaces between the cells (Figure 7c, arrows).Other BOFs were more heavily affected, and the F-actin fibres extended between the cells, forming a disorganised grid covering the ovarian follicle (Figure 7d).More internally (see Figure 7g, arrow b), in the follicular cells of dsPolyH-treated control females, the F-actin fibres were mainly distributed in the follicular cell membranes, being tightly packaged after patency had vanished (Figure 7e).Moreover, some F-actin labelling was observed in the cytoplasm of follicular cells (Figure 7e).In dsftz-f1-treated females, we noticed an increase in F-actin fibres in the follicular cells that showed a multilayered-like distribution, covering the cell walls (Figure 7f).Moreover, we detected bundles of F-actin fibres forming bridges between cells through the cytoplasm.
We also measured TpnI and Mhc expression (Figure 7h,i) in ovaries, observing that both tended to be upregulated (46% and 34.5%,  (j) Expression levels of SUMO in ovaries.The data represent copies of mRNA per copy of eukaryotic translation initiation factor 4aIII (eif4III), and are expressed as the mean ± S.E.M. (n = 3).The data in panels (h-j) were evaluated for normality using the Shapiro-Wilk test.All datasets passed normality test.The statistical analysis was performed using Student's t-test.Asterisks indicate statistically significant differences with respect to controls: *p = 0.04.
respectively).This suggests that an abnormal increase in the expression of genes related with muscle function contributes to impair ovulation.Finally, we quantified the expression of SUMO (small ubiquitin related modifier), a protein that in D. melanogaster is involved in the expression of ftz-f1, and at the same time Ftz-f1 is SUMOylated regulating a number of cellular processes in the ovary, including the control of ecdysteroid levels (Talamillo et al., 2013).We found that SUMO expression in B. germanica ovaries is significantly upregulated ( p = 0.048) in females late-treated with dsftz-1 (Figure 7j).

DISCUSSION
Ftz-f1 is a transcription factor that plays diverse roles in the insect ovary in different moments.Its function in the oogenesis of D. melanogaster has been thoroughly studied, especially its involvement in regulating the production of steroidogenic genes, since ftz-f1 is necessary to induce expression of dib and phm in this fly (Parvy et al., 2005).Moreover, ftz-f1 is required for egg chamber survival in the follicular cells.In the absence of Ftz-f1, the follicular cells do not exit the mitotic cycle, which prevents ovulation (Knapp et al., 2020).
Depletion of ftz-f1 in B. germanica impairs oviposition, and in the low number of eggs that were oviposited, embryo development was impaired (Cruz et al., 2023;Wexler et al., 2023).This suggests that maternal ftz-f1 transcripts previously reported in transcriptomic analysis (Ylla et al., 2018) have a key role in early embryogenesis.The expression profile of ftz-f1 in B. germanica ovaries might be equivalent to the ftz-f1 levels in D. melanogaster.In this fly ftz-f1 expression is upregulated after ecdysone pulses in the late embryo, larvae and pupae, and Ftz-f1 is considered a competence factor for stage-specific responses to ecdysone, driving the progression into the next developmental stage (Broadus et al., 1999;Cho et al., 2014;Lavorgna et al., 1993;Woodard et al., 1994).
The two types of treatment that we followed to unveil the function of ftz-f1 in B. germanica ovary, result in a prevention of ovulation.
However, some differences appear between the two treatments that helped to assess specific functions of ftz-f1 in the ovaries in different contexts.In the early treatment, there was not a fast depletion of ftz-f1 expression.However, we found an increase in the number of ovarian follicles in the ovarioles of dsftz-f1-treated females, which could result from the increase of phm and shd expression observed in ovaries from 3-day-old dsftz-f1-treated insects.The number of ovarian follicles in B. germanica ovarioles is under the control of the 20E (Ramos et al., 2020), and the increase of the expression of phm and shd in dsftz-f1-treated females suggest a pulse of 20E in the ovary that results in the differentiation of new ovarian follicles from the germarium (Ramos et al., 2020).Later, at the end of the cycle, the steroidogenic enzymes were depleted, with the exception of nvd, the first of the synthetic pathway, which was upregulated apparently due  et al., 2005) and Bombyx mori (Gu et al., 2021).This suggests that the function of Ftz-f1 in 20E biosynthesis is conserved throughout ovarian types, from panoistic to meroistic polytrophic.
The late treatment helped to define the function of Ftz-f1 in B. germanica ovaries in more detail.The effect of the dsftz-f1 treatment was faster, as the expression of ftz-f1 was dramatically downregulated only 2 days after the treatment (Figure 5b).Steroidogenic gene expression showed a clear tendency to reduction, and consequently E75 expression decreased.These changes impacted the expression of chorion genes, which behave differently compared to their expression in early-treated females.The late treatment takes place when the follicular cells already entered into endoreplication.
The decrease in the 20E signal affects the chorion genes differently, depending on whether they are early-or late-chorion genes.Observing the results of both treatments in ovaries of 7-day-old females it seems evident that Ftz-f1 induces the expression of steroidogenic genes, late in the gonadotrophic cycle, when the 20E is required to complete the chorion synthesis (Pascual et al., 1992).
Studies based on D. melanogaster (Knapp et al., 2020) and Tribolium castaneum (Xu et al., 2010) have shown that ftz-f1 knockdown impacts follicle cell development, impairing normal maturation, ultimately preventing ovulation and the development of the ovarian follicles.The follicular cells in B. germanica BOF were also affected after ftz-f1 transcripts were depleted.In early treated females, the cytokinesis was completed and, although most of the cells were binucleated, the nuclei were smaller, suggesting a failure in the endoreplication programme.Conversely, in late treatment, when cytokinesis has been completed, the nuclei of follicular cells become large showing a bizarre morphology, which suggests that the endoreplication process was enhanced, in line with the higher levels of citrus and brw, genes that are expressed in mid-late choriogenesis process.In addition, there is an increase of cytoskeleton fibres in these follicular cells, but they were arranged in a disorganised way, not following the parallel disposition observed in controls (Zhang & Kunkel, 1992).
The most evident phenotype resulting from ftz-f1 depletion in adult B. germanica is the absence of egg ovulation, which has been observed by previous authors (Cruz et al., 2023;Wexler et al., 2023).
Our results explain that the ftz-f1-depleted females cannot oviposit because the cytoskeleton of the follicular cells is disorganised, which impairs the follicle contractions required to expel the egg.The most dramatic changes occur when dsftz-f1 is injected late in the gonadotrophic cycle.The increased levels of genes related to the cytoskeleton and muscles, the perturbed distribution of F-actin fibres in the follicular cells of the BOFs and the possible failures of chorion due to the changes in 20E biosynthesis appear to be the main causes that prevent eggs from travelling through the oviduct and being finally ovulated.

EXPERIMENTAL PROCEDURES Cockroach colony and tissue sampling
Adult females of the cockroach B. germanica (L.) were obtained from a colony fed ad libitum on Panlab dog chow and water, and reared in the dark at 29 ± 1 C and 60%-70% relative humidity.Freshly ecdysed adult females were selected and used at appropriate ages.Mated females were used in all experiments (the presence of spermatozoa in the spermatheca was assessed at the end of the experiment to confirm mating).All dissections and tissue sampling were performed on carbon dioxide-anaesthetised specimens.

RNAi experiments
To knockdown the expression of ftz-f1, a dsRNA (dsftz-f1) was designed targeting the A/B domain of Ftz-f1 (498 bp; Cruz et al., 2008).A dsRNA (dsPolyH) corresponding to 307-bp of the Autographa californica nucleopoyhedrovirus sequence was used as control (dsPolyH).The dsRNA was synthesised using the kit MEGAscript RNAi (Thermo Fisher Scientific, Carlsbad CA, USA), and the dsRNA resultant was purified with phenol: chloroform: isoamyl alcohol (125:24:1).The dose used was 1 μg for either dsftz-f1 or dsPolyH in a volume of 1 μL, and was injected into the abdomen of 0-day-old and 5-day-old adult females, using a Hamilton microsyringe.

Tissue staining and microscopy
After dissection, ovaries were immediately fixed in paraformaldehyde (4% in PBS) for 2 h.Washing samples were performed as previously described (Irles & Piulachs, 2014).Ovaries were also incubated at room temperature for 20 min in 300 ng/mL phalloidin-TRITC (Sigma) and then for 5 min in 1 μg/mL DAPI (Sigma) PBT, to show the F-actin and nuclei, respectively.After three washes with PBT, ovaries were mounted in Mowiol (Calbiochem, Madison, WI, USA).Five females were treated with dsPolyH and five females were treated with dsftz-f1 on adult Day 0 (early treatment), and four females treated with dsPolyH, and five females treated with dsftz-f1on adult Day 5 (latetreatment), were processed.At least 20 ovarioles from each female were observed using a Zeiss AxioImager Z1 microscope (Apotome) (Carl Zeiss MicroImaging).
We considered that an ovarian follicle has been released from the germarium when the follicular cell layer is surrounding the oocyte.
The most basal follicle was excluded when quantifying ovarian follicles in the vitellarium.

Statistics
Data are expressed as mean ± standard error of the mean (S.E.M.) from at least three independent experiments.The data were evaluated for normality and homogeneity of variance using the Shapiro-Wilk test, which showed that no transformations were needed.All datasets passed normality test.Statistical analyses were performed employing Student's t-test.A p value <0.05 was considered statistically significant.Data were analysed using GraphPad, Prism 8 statistics software.
Xavier Belles for the critical reading of the manuscript.The funders had no role in the design of the study, data collection or analysis, the decision to publish or in the preparation of the manuscript.

F
I G U R E 1 Panoistic ovariole of Blattella germanica and dynamics of patency in follicular epithelia.(a) Panoistic ovariole from a 0-day-old adult female.Only the basal ovarian follicle (BOF) grows and matures in each gonadotrophic cycle.LOv, lateral oviduct; Pe, pedicel; OM, oocyte membrane; FE, follicular epithelium; sBOF, sub-basal ovarian follicle; G, germaria; TF, terminal filament; s, stalk; G, germarium; Vt, vitellarium.(b) Detail from panel A at higher magnification.FE, follicular epithelia; N, oocyte nucleus; n, nucleolus; Oo, oocyte.(d) Top view of follicular cells from a 5-day-old basal ovarian follicle, at the maximum of patency.The asterisks indicate the intercellular spaces.(c) Cartoon representing the change that follicular cells on BOF suffer through the first gonadotrophic cycle.The period when each process occurs is indicated by the age of the female (N6D0: nymph sixth Day 0; AdD represents adult stage followed by the number that indicates the female age.The hormonal signals in the adult, and the vitellogenic period are indicated.JH: Juvenile hormone).In B and D, DAPI (blue) was used for DNA staining and phalloidin-TRITC (green in (b), and red in (d)) was used to stain F-actin microfilaments.
T A B L E 1 Effect of dsftz-f1 treatment on B. germanica reproduction.

Ftz
-f1 AND OVIPOSITION IN BLATTELLA GERMANICA F I G U R E 3 Expression of ecdysone-related genes in the ovaries of Blattella germanica treated with dsftz-f1 in the freshly emerged adult stage.Adult females were treated with 1 μg of dsftz-f1 on the day of emergence and the expression of ecdysone-related genes was measured in ovaries of different days during the gonadotrophic cycle.neverland (ndv), spook (spo), phantom (phm) and shade (shd) were measured in ovaries from 3-day-old (a), 5-day-old (b) and 7-day-old (c) dsftz-f1-treated adults.Asterisks indicate statistically significant differences with respect to the controls: *p < 0.05; **p < 0.02 and ***p < 0.001.(d) Expression of EcR in ovaries from 3-day-, 5-day-and 7-day-old, dsftz-f1-treated adults.(e) Expression of E75A in ovaries from 3-day-, 5-day-and 7-day-old, dsftz-f1-treated adults.Asterisks indicate statistically significant differences with respect to the controls: *p = 0.02; **p = 0.01 and ***p = 0.007.The data represent copies of mRNA per 1000 copies of actin-5c and are expressed as the mean ± S.E.M. (n = 3-6).(f) Ovariole from a 7-day-old dsPolyH-treated female, showing the number of ovarian follicles in the vitellaria.(g) Ovariole from a 7-day-old dsftz-f1-treated female, showing a high number of ovarian follicles in the vitellaria, compared to the control females.In (f) and (g), DAPI (blue) was used for DNA staining and phalloidin-TRITC (red) was used to stain F-actin microfilaments.(h) Violin plots showing the number of differentiated ovarian follicles in dsPolyH-and dsftz-f1-treated females (n = 15 ovarioles, from three females from each treatment).Asterisks indicate statistically significant differences with respect to controls: *p = 0.03.The data in panels (a)-(b) and (e) were evaluated for normality using the Shapiro-Wilk test.All datasets passed normality test.The statistical analysis was performed using Student's t-test.
measured the expression of Troponin I (TpnI or WupA in D. melanogaster), which, among other cell functions, contributes to maintaining the apical-basal polarity of the cell and the nuclear divisions.We also measured the expression of Myosin (Myosin heavy chain, Mhc), which provides force for muscle contractions by interacting with actin in an ATP-dependent process.The expression of both genes was significantly downregulated in the ovaries of dsftz-f1treated females ( p = 0.024 and 0.04, respectively; Figure 4h,I) with respect to controls, suggesting that Ftz-f1 plays a role in B. germanica ovulation, acting at least on Tpnl and Mhc.dsftz-f1 treatment in late adult stage affects the development of follicular cells, the expression of steroidogenic genes and choriogenesis Finally, we also considered the possibility that the high levels of ftz-f1 mRNA in ovaries at the end of the gonadotrophic cycle have a differentiated function, contributing to successfully complete egg development.Thus, 5-day-old B. germanica adult females were treated with 1 μg of dsftz-f1, in order to observe the effects in the BOFs just before ovulation and oviposition.None of these late-treated females oviposited, thus showing the importance of Ftz-f1 at the end of the cycle to complete BOF development.When quantifying the expression of ftz-f1 by quantitative realtime PCR (qRT-PCR), we found that the actin-5c was overexpressed in the dsftz-f1-treated samples.Taking the eukaryotic translation initiation factor 4aIII (eif4aIII) as a reference gene(Irles & Piulachs, 2014), we

F
I G U R E 4 Depletion of ftz-f1 in the follicular epithelium of basal ovarian follicles in early-treated Blattella germanica adult females.Newly emerged adult females were treated with 1 μg of dsftz-f1 on the day of emergence and the follicular epithelium was observed on Day 7. (a) Follicular cells from a dsPolyH-treated basal ovarian follicle.All cells have similar size and nuclei are paired.(b) Follicular cells from a dsftz-f1-treated basal ovarian follicle.Cells are different in sizes and not all them are binucleated.The arrows indicate small cells with small nuclei size.DAPI was used for DNA staining.(c) Scheme of two follicular cells, indicating (arrow) the depth where the F-actins were observed in panels (d)-(g).(d) F-actin distribution in follicular cells from 7-day-old dsPolyH-treated females, the image corresponds to an early chorion 7-day-old female.(e) F-actins distribution in follicular cells in 7-day-old dsPolyH-treated females, in late choriogenesis.In (d) and (e), the F-actin fibres were well oriented on the cell surface forming parallel bundles.(f).F-actin distribution in follicular cells in dsftz-f1, 7-day-old females in late choriogenesis.(g).Detail of F, at a higher magnification.The fibres are not always oriented in the same direction in the follicular cell surface, and there are still spaces between the cells.In D-G phalloidin-TRITC was used to stain F-actin microfilaments.(h).Expression levels of Troponin I (TpnI) in ovaries from 7-day-old females.(i) Expression levels of Myosin heavy chain (Mhc) in ovaries from 7-day-old females.In H and I, the data represent copies of mRNA per 1000 copies of actin-5c and are expressed as the mean ± S.E.M. (n = 3).The data in panels H and I were evaluated for normality using the Shapiro-Wilk test.All datasets passed normality test.The statistical analysis was performed using Student's t-test.Asterisks indicate statistically significant differences with respect to the controls: *p = 0.04; **p = 0.024.F I G U R E 5 Expression of ecdysone-related genes in ovaries of 5-day-old adult Blattella germanica treated with dsftz-f1.Five-day-old adult females were treated with 1 μg of dsftz-f1 and the expression of actin-5c and the ecdysone-related genes in ovaries was measured 2 days later.(a) Expression levels of actin-5c.(b) Expression levels of ftz-f1.(c) Expression levels of steroidogenic genes: neverland (ndv), spook (spo), phantom ( phm) and shade (shd).(d) Expression levels of E75A.(e) Expression levels of EcR.(f) Expression levels of chorion genes: citrus, brownie (brw) and yellow-g (y-g).The data represent copies of mRNA per copy of eukaryotic translation initiation factor 4aIII (eif4III), and are expressed as the mean ± S.E.M. (n = 3).The data were evaluated for normality using the Shapiro-Wilk test.All datasets passed normality test.The statistical analysis was performed using Student's t-test.The asterisk indicates a significant difference with respect to the controls: *p = 0.02; **p = 0.006 ***p < 0.0008.F I G U R E 6 Depletion of ftz-f1 in the follicular cell nuclei from basal ovarian follicles in late-treated Blattella germanica adult females.Fiveday-old adult females were treated with 1 μg of dsftz-f1 and follicular cell nuclei in basal ovarian follicles were analysed in 7-day-old-females.(a) Nuclei of follicular cells from a dsPolyH basal ovarian follicle.All nuclei were a similar size, paired and some of them were starting to become pyknotic.(b) Nuclei of follicular cells from a dsftz-f1 basal ovarian follicle.The nuclei were paired, presenting a great diversity of shapes and sizes.All of them were larger than in the controls.DAPI was used for DNA staining.

F
I G U R E 7 Depletion of ftz-f1 in the follicular epithelium of basal ovarian follicles in late-treated Blattella germanica adult females.Fiveday-old adult females were treated with 1 μg of dsftz-f1, and F-actin distribution in basal ovarian follicles was analysed in 7-day-old-females.(a) F-actin distribution in follicular cells in 7-day-old dsPolyH-treated females.(b) F-actin distribution in follicular cells in 7-day-old dsPolyHtreated females, in late choriogenesis.(c) F-actin distribution in follicular cells in 7-day-old dsftz-f1-treated females.Asterisks indicate large cells, the arrowhead indicates small size cells and the arrows indicate spaces between the cells.(d) F-actin distribution in basal ovarian follicles highly affected by the dsftz-f1 treatment.F-actin fibres were extended between the cells forming a disorganised grid that covered the ovarian follicle.Images from (a-d) corresponds to the basal pole of follicular cells (see arrow a in panel (g)).(e) and (f).F-actin distribution in follicular cells, at a deeper level (see arrow b in panel (g)).(e) In dsPolyH-treated females, the F-actin develops around the cell membranes.(f) In dsftz-f1-treated females the F-actins presented a multilayered-like distribution, covering the cell walls.There were bundles of F-actin fibres forming bridges between the cells through the cytoplasm (arrows).(g) Scheme of two follicular cells, indicating (arrows) the deeper where the F-actins were observed in the former panels.(h) Expression levels of Troponin I (TpnI) in ovaries.(i) Expression levels of Myosin heavy chain (Mhc) in ovaries.
to a compensation mechanism triggered by the decrease in 20E determined by the downregulation of the other steroidogenic genes.The depletion of shd impairs the transformation of ecdysone to 20E, explaining that the levels of EcR and E75A crumbled (Figure 2d,e), suggesting a failure in 20E signalling that might affect the expression of chorion genes.At the beginning of the gonadotrophic cycle in B. germanica adults, Ftz-f1 is involved in regulating steroidogenic gene expression, especially phm, as occurs in D. melanogaster (Parvy RNA purification Kit (Canvax Biotech S.L., C ordoba, Spain), following the manufacturer's instructions.A total of 300 ng from each RNA extraction was treated with DNAse (Promega, Madison, WI, USA) and first-strand cDNA was synthesised by cDNA synthesis using Transcriptor First Strand cDNA Synthesis Kit (Roche Diagnostics GmbH, Germany).RNA quantity and quality were estimated by spectrophotometric absorption at 260/280 nm in a Nanodrop Spectrophotometer ND-1000 ® (NanoDrop Technologies, Wilmington, DE, USA).The expression pattern of the examined B. germanica genes was determined by qRT-PCR in ovaries from different adult stages.Gene expression in treated individuals was quantified individually, using the ovary pair.The PCR primers used in qRT-PCR expression studies were designed using the Primer3 v.0.4.0 (Rozen & Skaletsky, 2000).The actin-5c gene of B. germanica (accession number: AJ862721) or the eukaryotic translation initiation factor 4aIII (eif4III; accession number: CCX34983) were used as a reference gene.qRT-PCR reactions were made using the iTaq Universal SYBR Green Supermix (BioRad) containing 200 nM of each specific primer (performed in triplicate).Amplification reactions were carried out at 95 C for 3 min, and 40 cycles of 95 C for 10 s and 60 C for 1 min, using CFX Opus 384 Real-Time PCR System