Bicaudal C is required for the function of the follicular epithelium during oogenesis in Rhodnius prolixus

The morphology and physiology of the oogenesis have been well studied in the vector of Chagas disease Rhodnius prolixus. However, the molecular interactions that regulate the process of egg formation, key for the reproductive cycle of the vector, is still largely unknown. In order to understand the molecular and cellular basis of the oogenesis, we examined the function of the gene Bicaudal C (BicC) during oogenesis and early development of R. prolixus. We show that R. prolixus BicC (Rp-BicC) gene is expressed in the germarium, with cytoplasmic distribution, as well as in the follicular epithelium of the developing oocytes. RNAi silencing of Rp-BicC resulted in sterile females that lay few, small, non-viable eggs. The ovaries are reduced in size and show a disarray of the follicular epithelium. This indicates that Rp-BicC has a central role in the regulation of oogenesis. Although the follicular cells are able to form the chorion, the uptake of vitelline by the oocytes is compromised. We show evidence that the polarity of the follicular epithelium and the endocytic pathway, which are crucial for the proper yolk deposition, are affected. This study provides insights into the molecular mechanisms underlying oocyte development and show that Rp-BicC is important for de developmental of the egg and, therefore, a key player in the reproduction of this insect.


Introduction
Rhodnius prolixus is a hematophagous insect and, like other triatomines, is the vector of Trypanosoma cruzi, the causative agent of Chagas disease (Coura and Borges-Pereira 2012;Chagas 1909). In addition to the sanitary relevance, R. prolixus has been a classical model of physiology since the pioneer studies of Sir Vincent Wigglesworth (Wigglesworth 1934(Wigglesworth , 1939(Wigglesworth , 1953(Wigglesworth , 1964 and, to some extent, an emerging model for developmental biology (Berni et al. 2014;Lavore et al. 2014;Lavore et al. 2012). The genome of R. prolixus has been sequenced (Mesquita et al. 2015), several tissue-specific transcriptomes have been analyzed (Medeiros et al. 2011a;Ribeiro et al. 2014), as well as developmental studies of ovaries and embryos (Brito et al. 2018;Lavore et al. 2015) providing a solid foundation for gene identification. The development of molecular tools such as parental RNA interference (RNAi) (Lavore et al. 2012) set the ground for functional analysis.
Oogenesis is a period of rapid cellular growth and differentiation, implying the differentiation of the oocyte from a niche of stem cells in the germ line, the accumulation of yolk, formation of the chorion, and the establishment of the future embryo axes. These events are entangled in a regulated manner by means of the communication between the different cell types that compose the ovary, hormonal signaling, and gene expression (Lynch and Roth 2011). Insect ovaries are classified into three distinct types: panoistic, polytrophic, and telotrophic, based on the morphology of germ cells in the mature ovary (Bonhag 1958;Davey 1965;de Cuevas 2005;Snodgrass 1935;Urbani 1970). The morphology and architecture of the ovaries have been studied in a variety of insects  (Büning 1994;de Cuevas 2005;Fortes et al. 2011;Huebner and Anderson 1972a, 1972b, 1972cHuebner and Injeyan 1981;Lutz and Huebner 1980), but a detailed understanding of the regulatory gene network have only been determined in Drosophila melanogaster Nusslein-Volhard 1988a, 1988b;Nusslein-Volhard and Roth 1989;Roth and Schupbach 1994;Schupbach and Roth 1994;Schupbach and Wieschaus 1986). The adult females of R. prolixus have two telotrophic type ovaries, each one made up of seven ovarioles (Huebner 1981a). Each ovariole is composed with nurse cells confined in a distal chamber referred to as the trophic chamber or tropharium, separated from the vitellarium, in which oocytes go through the different stages of oogenesis, previtellogenesis, vitellogenesis, and choriogenesis, accompanied by the follicular epithelium (Huebner 1981b;Huebner and Anderson 1972b). Oogenesis is triggered by feeding (Buxton 1930). The trophic chamber produces maternal RNAs and nutrients, which are transported to the developing oocyte through tubular bridges-the trophic cords, in a directional transport mediated by a network of microtubules (Harrison and Huebner 1997;Valdimarsson and Huebner 1989). The accompanying follicle cells shows dramatic morphological and physiological changes during the different stages of oogenesis, but they always keep an organized pattern (Huebner and Anderson 1972b). Together with the fat body, the follicular cells are responsible for the synthesis of vitellogenin (Vg), precursor of vitellin (Vn), the main component of egg yolk (Melo et al. 2000;Oliveira et al. 1986;Raikhel 2005;Raikhel and Dhadialla 1992). Later on, follicle cells produce the outermost layer of the egg, the chorion, which protects egg from dehydration and regulates oxygen intake and fertilization (Beament 1946). Despite the extensive morphological and biochemical studies (Atella et al. 2005;Harrison and Huebner 1997;Huebner and Anderson 1972a, 1972b, 1972cLutz and Huebner 1981;Oliveira et al. 1986). Fundamental evidence on the physiological network based on the transcriptome analysis of fat body and ovary has just emerged in R. prolixus through the pioneer work of Leyria et al. (2020). However, the regulatory gene networks required for proper egg formation and patterning remains to be studied.
Many orthologues of the genes involved in D. melanogaster oogenesis have been identified in R. prolixus (Medeiros et al. 2011a;Mesquita et al. 2015;Pascual 2019), one of this was Bicaudal C (BicC). BicC was originally identified in a D. melanogaster maternal mutagenesis screen (Mohler and Wieschaus 1986;Nusslein-Volhard 1977;Schupbach and Wieschaus 1991). Females heterozygous for BicC mutations produce embryos of several different phenotypic classes, including bicaudal embryos that consist only of a mirror-image duplication of 2-4 posterior segments. Homozygous BicC females are sterile because the centripetal follicle cells fail to migrate over the anterior surface of the oocyte at stage 10 during D. melanogaster oogenesis (Mahone et al. 1995;Nusslein-Volhard 1977;Schupbach and Wieschaus 1991). BicC encodes a protein with hnRNP K homology (KH) and sterile alpha motif (SAM) domains (Mahone et al. 1995), both RNA-binding motifs (Aviv et al. 2006;Chmiel et al. 2006;Johnson and Donaldson 2006) that interact with other proteins related to RNA metabolism and targets mRNAs to form regulatory ribonucleoprotein complexes (Chicoine et al. 2007). Also, it has been reported to be involved in the function of Malpighian tubules in the adults (Gamberi et al. 2017).
Here, we report the function of Bicaudal C (Rp-BicC) during oogenesis of R. prolixus. We identified the expression of BicC gene and carried out parental RNAi experiments. Our results show that Rp-BicC is required for the proper follicular epithelium function, affecting yolk uptake and chorion patterning.

Insect husbandry
A colony of Rhodnius prolixus was maintained in our insectarium of the Centro de Bioinvestigaciones (CeBio) in plastic jars containing strips of paper at 28°C and 80% relative humidity in controlled environment incubators with a 12h light/dark cycle. In this condition, embryogenesis takes 14 ± 1 days. Insects were regularly fed on chicken, ad libitum, which were housed, cared, fed, and handled in accordance with resolution 1047/2005 (National Council of Scientific and Technical Research, CONICET) regarding the national reference ethical framework for biomedical research with laboratory, farm, and nature collected animals, which is in accordance with the international standard procedures of the Office for Laboratory Animal Welfare, Department of Health and Human Services, NIH, and the recommendations established by the 2010/63/EU Directive of the European Parliament, related to the protection of animals used for scientific purposes. Biosecurity rules fulfill CONICET resolution 1619/2008, which is in accordance with the WHO Biosecurity Handbook (ISBN 92 4 354 6503).

Identification of the BicC transcript and cDNA synthesis
The transcript was identified by local BLASTX search (Altschul et al. 1990) using the D. melanogaster orthologue gene as query on transcriptomes from ovary and different early embryonic stages, assembled using the annotated genome of R. prolixus as reference (VectorBase, RproC3 version; Pascual and Rivera Pomar unpublished data). The predicted sequence was aligned with orthologues from other species with ClustalW (Larkin et al. 2007). Phylogenetic tree was inferred with the software BEAST v1.10.4 (Drummond and Rambaut 2007) in the CIPRES Science Gateway (Miller et al. 2010). The result was visualized with iTol online tool (Letunic and Bork 2016).
Fat body and ovaries were dissected, in phosphate-buffered saline (PBS 1×), from adult females in which oogenesis was induced (144 h post-feeding). Each tissue sample was grouped in pools of three organs, placed in TRIzol reagent, and kept frozen until use. cDNA synthesis and PCR amplification was performed in the same conditions mentioned previously.

Real-time quantitative PCR
Adult mated insects (6 days after feeding) were used to collect eggs at different developmental times-unfertilized, 0, 12, 24, and 48 h post-egg laying (hPL), and ovaries in vitellogenic stage, were immediately frozen and stored in liquid nitrogen until use. For each sample, 75 embryos in each stage and 10 vitellogenic ovaries were used to extract RNA for qRT-PCR.
Total RNA was isolated using TRIzol reagent (Invitrogen) and treated with DNAse (QIAGEN). cDNA was synthesized using SuperScript™ VILO™ MasterMix kit (Invitrogen) following the manufacturer's instructions. qRT-PCR was performed on three independent biological replicates, in triplicate (3 wells/cDNA sample), in a 10 μl final volume. Gene expression levels were quantified using SsoAdvanced Universal SYBR Green Supermix (Bio-Rad) in an Applied Biosystems 7.500 Real-Time PCR System (Thermo Fisher, Scientific). The amplification protocol was as follows: 95°C for 10 min followed by 35 cycles (95°C 15 s/55°C 30 s/72°C 45 s). A control without a template was included in all experiments. α-Tubulin were used as reference gene (Rp-αTub (220 bp) sense: 5′-CAAATAATTACGCCCGAGGA-3′ and antisense: 5′-TTGAGGAGCTGGGTAAATGG-3′). All primer pairs used for qRT-PCRs were tested for dimerization, efficiency, and amplification of a single product. The Ct value was averaged for the technical triplicate experiments and subtracted from the average Ct of the reference gene, to yield the expression difference (dCt) for each biological replicate. The results were analyzed according to (Ginzburg et al. 2017). To test whether the expression of Rp-BicC was significantly different between the different stages, one-way ANOVA was carried out followed by post hoc test using the GraphPad Prism v6.0 software (GraphPad Software, CA, USA, www.graphpad. com).

Ovary and embryo manipulation
Control and silenced adult females were fed to induce oogenesis, and 5 days later, the ovaries were dissected in PBS 1×. Ovaries were fixed in different ways depending on the subsequent analysis. For confocal microscopy, the fixation was performed on ice in 4% paraformaldehyde (PFA) in PBT (PBS 1× + 0.1% Tween-20) for 30 min, then washed three times in PBT and stored at 4°C until staining the nuclei with Hoechst (Sigma-Aldrich, USA, 1 μg/ml). Images were acquired with the Zeiss LSM 800 confocal microscope. For light microscopy, ovaries were fixed in formaldehyde 4%, washed with Millonig's buffer, dehydrated in graded series of ethanol (70%, 96%, 100%) and xylene (100%) and embedded in paraffin (Wanderley-Teixeira et al. 2006). Five-micrometer-thick sections were cut in a rotary microtome (Leica) and stained using standard hematoxylin-eosin procedure, mounted and photographed using an A1 ZEISS microscope. For transmission electron microscopy (TEM), the protocol was modified from Huebner and Anderson (1970). Ovaries were fixed in glutaraldehyde 2.5% and post-fixed in 1% osmium tetroxide in Millonig's buffer at pH 7.4. This was followed by dehydration in a graded series of ethanol (25%, 30%, 50%, 80%, 90%) and acetone (100%), after which the ovaries were infiltrated and embedded in epoxy resin (Durcupan ACM, Fluka AG, Switzerland). Ultrathin sections (~60 nm) were cut with a diamond knife, stained with aqueous uranyl acetate and Reynold's lead citrate (Reynolds 1963), and examined at 80 kV in a MET JEOL 1200 EXII transmission electron microscope.
For the analysis of lipids and membranes distribution, lipophilic styryl dye FM 4-64FX (Thermo Fisher Scientific) was injected in the body cavity of females in a 1:500 dilution (3 μg/μl) in PBS 1× and let to diffuse for 20 min. FM 4-64FX targets plasma membrane and marks exo/endocytosis hot spots in the cells (Smith and Betz 1996). For the dextran oocyte uptake analysis, 2 μl of Texas Red-conjugated dextran (10,000 MW; Molecular probes, Thermo Fisher Scientific) was injected between abdominal tergites in the hemocoel of females 4 days after blood meal and incubated 24 h. After the corresponding time, ovaries were dissected and fixed as described above for confocal microscopy and images were acquired in a Zeiss LSM 800 confocal microscope.
Eggs collected from individual females were used for scanning electron microscopy (SEM), fixed in glutaraldehyde 2.5%, washed with Millonig's buffer, dehydrated in a graded series of ethanol (70%, 96%, 100%), mounted with doublesided adhesive carbon tape on metallic stubs, metallized with gold and observed under a SEM Quanta 250 (FEI) operated at 20 kV (Sorrivas de Lozano et al. 2014).

Fluorescent immunohistochemistry
Ovaries were fixed for confocal microscopy, then washed in PBS (0.1% Triton X-100 in PBS), blocked with 5% normal goat serum for 2 h, and incubated overnight at 4°C with 1:200 dilution of rabbit polyclonal anti-vitellin antibody, which reacted with Vg/Vn (Aguirre et al. 2008;Blariza et al. 2016). After extensive washing, the ovaries were incubated for 2 h at room temperature with secondary Alexa 568conjugated anti-rabbit IgG (1:500 in PBS; Invitrogen, Life Technologies), washed and counterstained with Hoechst (Sigma-Aldrich, USA, 1 μg/ml) before image acquisition in a ZEISS LSM 800 confocal microscope.

RNA in situ hybridization
Digoxigenin-labeled antisense Rp-BicC RNA probes were synthesized using the RNA-Dig Labeling kit (SP6/T7) (Roche). In situ hybridization was carried out in 4% PFA fixed ovaries stored in PBT at 4°C. The ovaries were post-fixed in PBT + fixative solution (10% PFA in PBS + EGTA-Na 2 ) for 20 min on a rocking platform at room temperature. The ovaries were washed three times with PBT and digested with proteinase K (10 mg/ml) for 15 min and post-fixed as before, following three PBT washes. A pre-hybridization step was performed for 2 h at 60°C in Hybe (50 % formamide, 5x SSC, 0.2 mg/ml Sonicated salmon testes DNA, 0.1 mg/ml tRNA, 0.05 mg/ml heparin, 0.1 % Tween-20) before the addition of the probe and further incubated overnight at 60°C. The ovaries were rinsed three times with Hybe-B (50% formamide, 5× SSC, 0.1 % Tween-20) and then washed in Hybe-C (50% formamide, 2× SSC, 0.1 % Tween-20) during 2 h at 60°C , and further washed three times with PBT. The hybridized samples were blocked with antibody-hybridization solution (0.2% Tween-20, 1 mg/ml bovine serum albumin, 5% normal goat serum) for 3 h at room temperature and then incubated overnight with alkaline phosphatase-conjugated anti-DIG Fab fragments (Roche, 1:2000) at 4°C on a shaking platform. The antibody was washed away three times with PBT and one time with alkaline staining buffer (100 mM TRIS, 100 mM NaCl, 0.1% Tween-20). The enzymatic activity revealed with NBT/ BCIP (Roche). When staining was evident, the ovaries were washed in PBT three times to stop the reaction, dehydrated in a graded series of ethanol, and mounted in glycerol for observation and image acquisition using A1 ZEISS microscope.

Parental RNA interference
Double-stranded RNA (dsRNA) was produced by simultaneous transcription with T7 RNA polymerase (New England Biolabs) on PCR products containing T7 promoter sequences (CGACTCACTATAGGG) at both ends. Two independent templates, dsRNA BicC1 and dsRNA BicC2 , were used for independent experiments to evaluate potential off-target effects. dsRNA was quantitated and injected into virgin females, using different concentrations, as described in Lavore et al. (2012). Two days after injection, the females were fed to induce oogenesis and mated with males. After mating, eggs were collected and ovaries fixed as indicated above. A negative control was performed injecting virgin females with dsRNA corresponding to the βlactamase gene (dsRNA βlac ) of Escherichia coli gene (Lavore et al. 2012).

Results
The Rp-BicC transcript is expressed in ovaries and early embryos Rp-BicC was identified in transcriptomes derived from ovaries, unfertilized eggs, and early embryos of R. prolixus (Pascual and Rivera-Pomar, unpublished data) by sequence similarity search against D. melanogaster orthologue. The assembled Rp-BicC transcript from these RNA-seq dataset corresponds to the ab initio annotated transcriptional units RPRC0001612 and RPRC001613 within the supercontig KQ034133, indicating that the two different predictions in Vector Base (https://www.vectorbase.org/) were erroneous, and that they correspond to a single transcriptional unit (Fig.  1a). The transcript (1986 bp) derives from 14 exons and encodes a predicted polypeptide of 662 amino acids. Multiple alignment of BicC orthologous sequences showed that Rp-BicC conserve the typical KH and SAM domains as other species (Fig. 1b and Supplementary Fig. 1).
Reads from the Rp-BicC transcript were identified in all of the transcriptomes corresponding to the different stages come from ovaries, unfertilized eggs, and early embryos at 0, 12, 24, and 48 h post-egg laying (hPL; Pascual and Rivera-Pomar, unpublished data), indicating that the transcript is maternally contributed, although the zygotic expression cannot be ruled out. The relative expression level of Rp-BicC was assessed by qPCR. Rp-BicC was detected in all of the analyzed stages, showing a higher expression from unfertilized eggs up to 12hPL, with a significant decrease in the expression levels after on (Fig. 1c). In situ hybridization validated the expression of Rp-BicC transcript in the ovaries, showing cytoplasmic distribution in both, the germarium (Fig. 1d, arrowhead) and the follicular epithelium of previtellogenic and vitellogenic oocytes (Fig. 1d, e).

Rp-BicC is required for proper egg formation
To determine the role of Rp-BicC, we injected non-fed virgin females (n = 17) with different concentrations (ranging from 0.7 to 2.5 μg/female) of two independent dsRNA corresponding to different regions of the transcript (dsRNA BicC1 , 237 bp and dsRNA BicC2 , 291 bp). As control, we used dsRNA corresponding to the β-lactamase gene of E. coli (dsRNA βlac ). After feeding and mating, dsRNA Rp-BicC1 , dsRNA BicC2 , and dsRNA βlac injected females were evaluated for egg deposition and morphology, ovary phenotype, and embryo lethality. The number of eggs laid by the interfered females is significantly less than the control (Mann-Whitney statistical test for dsRNA Rp-BicC1 P < 0.01; dsRNA BicC2 P < 0.05) indicating that Rp-BicC is required for fertility (Table 1). To determine lethality, the eggs were incubated to develop for the expected time of embryogenesis to finish (> 14 days). None of the eggs from interfered females resulted in hatchlings, indicating that the embryogenesis was affected. The dissection of the eggs did not show any distinguishable embryonic structure, suggesting that BicC might act at very early stages of development. The eggs laid by the silenced females, as opposed to the control ones, were smaller, with irregular shape, and showed white coloration instead the characteristic pink (Fig. 2a), indicating the absence or significant reduction of the Rhodnius heme-binding protein (RHBP, an important component of the yolk). The Rp-BicC interfered eggs showed an irregular surface. To determine if there is a structural alteration in the chorion morphology we performed scanning electron microscopy. Compared to the regular hexagonal pattern of the chorion observed in the control (Fig. 2b), the eggs derived from the silenced females showed defects in the chorion structure, showing an irregular pattern, prominences, and a shrunken surface (Fig. 2c). The operculum is deformed, although it has a similar size as the control ones. This results indicates that the chorion and chorionic structures are formed, but the patterning is dramatically affected.

Rp-BicC is required for the development of follicular epithelium during oogenesis
To further investigate the effect of Rp-BicC, we studied the morphology of the ovary. The ovaries of the Rp-BicC silenced females have the same number of ovarioles as the control, but they are reduced in size (Fig. 3a-f). We analyzed the morphology under DIC optics (Fig. 3b, e) and by staining the nuclei to determine cell distribution (Fig. 3c, f). Compared to the control (Fig. 3b), the follicular epithelium of the Rp-BicC silenced females was folded and wrinkled (Fig. 3c) and both, previtellogenic and vitellogenic oocytes were smaller (Fig.  3b, c). The ovaries of Rp-BicC silenced females did not evidence to show significant morphological differences in the germarium, but displayed the absence of the large nucleoli characteristic of the trophic chamber. From previtellogenic stages on, we observed that the regular organization of the follicular cells is lost (Fig. 3e, f). Thin sections of the ovary stained with hematoxylin/eosin showed that, as compared to the control ones (Fig. 3d), silenced females displayed oocytes with irregular yolk distribution, accompanied by diminished number of yolk granules and presence of empty spaces in the cytoplasm. The follicular cells appear detached one from each other and the irregular columnar epithelium showed increased intercellular space (Fig. 3h). Transmission electron microscopy analysis indicates that, compared to the control (Fig. 3i), follicular cells of the Rp-BicC silenced females lack their contact with the basal membrane and tunica propria, reduction of the contacts that keep them together in a regular manner, and show vesiculated cytoplasm and less dense nucleoli (Fig. 3j). This results agrees with the observed phenotype of the chorion and indicate that the follicular cells are able to form the chorion, despite the disarray of the cells.

Rp-BicC affects the vesicle trafficking and polarity of follicle cells
In order to address the functional characterization of the morphological changes observed in the follicular epithelium, we analyzed if the transport of molecules, from the hemolymph to the oocyte was affected. We first analyzed if the yolk uptake by the oocyte was affected; for this, we used an antibody to localize the presence of vitellin/vitellogenin in the developing oocytes. In control females, we observed that vitellogenic oocytes accumulates vitellogenin within the cytoplasm of follicle cells (Fig. 4a-c). The ovaries of silenced females, compared to the control, showed a dramatic decrease of the antivitellogenin signal (Fig. 4d-f). A closer look showed that vitellogenin was concentrated in granules in the apical region of the follicle cells of the control females (Fig. 4g), while in the silenced females, we observed a dramatic decrease in the number of granules and intensity of fluorescent signal, although we cannot rule out a dispersed localization of the signal (Fig. 4h). However, to rule out an extra ovarian effect of Rp-BicC in the uptake of vitellogenin, we evaluated if the expression of Rp-BicC in ovaries and fat body, where circulating vitellogenin is synthesized, of the silenced females. Only the ovaries showed a reduced expression of Rp-BicC ( Supplementary Fig. 2). This suggests a link between Rp-BicC and the amount of vitellogenin in the follicular cells. We could hypothesize that the disordered follicular epithelium affects the vitellogenin diffusion through the extracellular space (patency) to the oocyte surface. To evaluate whether the uptake of macromolecules is affected, we injected fluorescent dextran (MW 10 kDa) in the abdominal cavity of both, control and silenced females, and analyzed the differences in the uptake of the fluorescent dextran into the oocyte. In control females, the vitellogenic oocytes accumulate fluorescent dextran ( Fig. 5a; the general morphology of the ovariole is shown by Hoechst staining in Fig. 5b), while the vitellogenic oocytes of the Rp-BicC silenced females shows a dramatic reduction of fluorescent dextran intake ( Fig. 5c; morphology in Fig. 5d). Taken together, these results support the notion that the transport of macromolecules requires an organized follicular epithelium, for which Rp-BicC seems to have a key role.
To study if the disordered epithelium has affected the polarity of the follicular cells, we used a lipophilic styryl dye to mark cell membrane and nascent endosomes that spread into the cytoplasm. In the ovaries of control females, we observed fluorescent signal in the apical and basal poles of the follicle cells. This fluorescent signal indicates bona fide regions of endo and exocytosis (Fig. 6a-c). In the ovaries of Rp-BicC silenced females, the fluorescent dye could only be detected in the follicular membrane in the apical pole (Fig. 6d-j). This suggests that the polarity of the follicular epithelium might be compromised, and therefore, it might affect the communication of the follicle cells with the developing oocyte. Furthermore, the absence of signal in the oocyte membrane agrees with the diminished accumulation of yolk as we showed in the ovaries of Rp-BicC silenced females. Therefore, we conclude that if the endo/exocytic pathway is affected the transport of vitellogenin to the oocyte might be impaired.

Discussion
R. prolixus has been a model system for many essential issues in biology, but the understanding of the molecular mechanisms had to wait until the sequencing of the genome (Mesquita et al. 2015). The oogenesis of R. prolixus is one of the best studied among insects, from the morphological work of Huebner (Huebner 1981a;Huebner and Anderson 1972a, 1972b, 1972cHuebner 1980, 1981) to the biochemistry studies of Masuda (Bouts et al. 2007;Braz et al. 2001;Coelho et al. 1997;Costa-Filho et al. 2001;Gondim et al. 1987;Machado et al. 1996;Machado et al. 1998). Studies on the cellular biology of the oogenesis have recently emerged (Pereira et al. 2020;Vieira et al. 2018); however, we still lack genetic information to understand the molecular basis of egg formation and patterning in R. prolixus. A transcriptome analysis of the different phases of oogenesis has been reported (Brito et al. 2018) where Rp-BicC was not assigned although one annotated gene was identified (RPRC001612). In this respect, this report on Rp-BicC is the first contribution on a maternal gene required for oogenesis and to provide a bona fide mechanism for egg development and patterning of the chorion not affecting the polarity of the egg. In D. melanogaster, BicC is a maternal gene affecting embryonic anterior-posterior polarity, with a wide range of defects in segmentation (Mohler and Wieschaus 1986;Nusslein-Volhard 1977). We could not identify any embryonic structures in the eggs derived from Rp-BicC silenced females, as we have observed also for other maternal genes (Pagola, Pascual, and Rivera Pomar, unpublished data). This indicates that the role for Rp-BicC in embryogenesis has to be prior to germ band formation and deserves a detailed study in the future. Cell-to-cell interactions are crucial for the development of oogenesis. It affects both yolk deposition and the signaling from the follicle cells to the oocyte and the later embryogenesis. Based on the expression of Rp-BicC in follicle cells and the distribution of vitellogenin and membrane markers in the silenced females, we support the idea that Rp-BicC affects both the morphology and polarity of the follicular epithelium and, likely, the oocyte-follicle cell interaction. This suggests a conserved role for BicC. Bicc1 (the mouse homologue of BicC) is required for E-cadherin-based cell-cell adhesion, indicating that that lack of Bicc1 disrupts normal cell-cell junctions, and, in consequence, alter epithelial polarity (Park et al. 2016 Fig. 4 Silencing of Rp-BicC alter vitellin uptake in vitellogenic oocytes. a-c Immunostaining using anti-vitellin antibody to determine distribution of vitellin (visualized in red) in the oocytes of control females (n = 7). d-f Immunostaining using antivitellin antibody to determine distribution of vitellin (visualized in red) in the oocytes of RNAi BicC females (n = 6). All samples were counterstained with Hoechst (blue). Scale bar, 50 μm.  (Mahone et al. 1995). We observed that the primary effect of silencing Rp-BicC is a disorganized pattern of the follicular epithelium from the early previtellogenic stages until the end of vitellogenesis. The cells look loose and lacking the association among each other. At a first glance, the phenotype might be related to atresia, characterized by a degenerative process of the follicular epithelium, in which chorion deposition does not occur yielding non-viable oocytes (Huebner 1981b;Medeiros et al. 2011b). However, in the silenced Rp-BicC ovaries the follicular cells are able to synthesize the chorion.

Rp-BicC RNAi Control
Our results agrees with the BicC phenotype described for the hemipteran Nilaparvata lugens, which seems to affect yolk loading in the egg (Zhang et al. 2015), but we provide a plausible hypothesis to explain the mechanism that causes the phenotype. Rp-BicC is required for epithelial integrity and cell polarity as evidenced by the changes of distribution of endocytic pathway markers (Gaffield et al. 2009), suggesting that the apical-basal polarity of the follicular cells is a crucial requirement for vitellogenesis. We propose that transcytosis plays a role in the uptake of yolk precursors into the oocyte. This has been suggested for insects of different orders, such as Carausius morosus (Phamatodea), Podisus nigrispinus (Hemiptera), and Apis mellifera, Melipona quadrifasciata, Mischocyttarus cassununga, and Pachycondyla curvinodis (Hymenoptera) (Assis et al. 2019;Dohanik et al. 2018;Fausto et al. 2005;Giorgi et al. 1998;Ronnau et al. 2016). Up to now, vitellogenin uptake by the oocyte had exclusively been attributed to the phenomenon of patency (Davey 1981;Huebner and Anderson 1972b;Huebner and Injeyan 1980). Our results provide evidence of a transcellular route for vitellogenin transport through the follicular epithelium during oogenesis in R. prolixus.
It has been recently demonstrated that Rp-ATG6 and Rp-ATG8, part of PI3P-kinase complex that regulates the endocytic and autophagy machinery are essential for yolk accumulation (Pereira et al. 2020;Vieira et al. 2018). The phenotype of silenced Rp-ATG6 females shares similarities with the one of Rp-BicC-unviable, small and white eggs that accumulate a minor fraction of yolk. Rp-ATG8 is required for the maternal biogenesis of autophagosomes and has a role, although not exclusive, in follicular atresia. This indicates that Rp-ATG6 seems to only affect the oocyte uptake of yolk rather than the integrity of the follicular epithelium and the patterning of the chorion. This led us to propose that autophagy is involved in a process that requires cell polarity and that Rp-BicC might be an upstream gene that contributes to polarization.
The sequencing of many insect genomes poses a challenge ahead for comparative and evolutive studies of the cellular and molecular mechanisms of development, in particular in hemimetabolous insects. Studies on the classical model R. prolixus, as the one presented here, will open new and exciting paths to understand the genetic interactions that regulate the formation of the egg and its subsequent developmental processes.

RNAi BicC
Control d e f Fig. 6 Silencing of Rp-BicC alter lipid distribution in vitellogenic oocytes. a-c Distribution of the FM 4-64FX probe in oocytes after injection of control females (n = 7). d-f Distribution of the FM4-64FX probe in oocytes after injection of silenced (RNAi BicC ) females (n = 6). Hoechst (a, d, in blue) visualizes DNA; FM 4-64FX (b, e, in red) stain membranes and endocytic vesicles. Scale bar, 50 μm