Research | Open | Published:
Posterior localization of ApVas1 positions the preformed germ plasm in the sexual oviparous pea aphid Acyrthosiphon pisum
EvoDevovolume 5, Article number: 18 (2014)
Germline specification in some animals is driven by the maternally inherited germ plasm during early embryogenesis (inheritance mode), whereas in others it is induced by signals from neighboring cells in mid or late development (induction mode). In the Metazoa, the induction mode appears as a more prevalent and ancestral condition; the inheritance mode is therefore derived. However, regarding germline specification in organisms with asexual and sexual reproduction it has not been clear whether both strategies are used, one for each reproductive phase, or if just one strategy is used for both phases. Previously we have demonstrated that specification of germ cells in the asexual viviparous pea aphid depends on a preformed germ plasm. In this study, we extended this work to investigate how germ cells were specified in the sexual oviparous embryos, aiming to understand whether or not developmental plasticity of germline specification exists in the pea aphid.
We employed Apvas1, a Drosophila vasa ortholog in the pea aphid, as a germline marker to examine whether germ plasm is preformed during oviparous development, as has already been seen in the viviparous embryos. During oogenesis, Apvas1 mRNA and ApVas1 protein were both evenly distributed. After fertilization, uniform expression of Apvas1 remained in the egg but posterior localization of ApVas1 occurred from the fifth nuclear cycle onward. Posterior co-localization of Apvas1/ApVas1 was first identified in the syncytial blastoderm undergoing cellularization, and later we could detect specific expression of Apvas1/ApVas1 in the morphologically identifiable germ cells of mature embryos. This suggests that Apvas1/ApVas1-positive cells are primordial germ cells and posterior localization of ApVas1 prior to cellularization positions the preformed germ plasm.
We conclude that both asexual and sexual pea aphids rely on the preformed germ plasm to specify germ cells and that developmental plasticity of germline specification, unlike axis patterning, occurs in neither of the two aphid reproductive phases. Consequently, the maternal inheritance mode implicated by a preformed germ plasm in the oviparous pea aphid becomes a non-canonical case in the Hemimetabola, where so far the zygotic induction mode prevails in most other studied insects.
Studying model organisms can expand our knowledge in developmental biology and enrich our understanding of developmental diversity. One issue, developmental plasticity, has long been a difficult subject to access in animal models displaying little phenotypic plasticity. However, the pea aphid Acyrthosiphon pisum, a hemimetabolous hemipteran insect with abundant adaptive capacity in response to specific environmental cues, has proven an excellent model for study, and the recently published genome has made research on this species more accessible [1, 2]. With a sequenced genome, scientists can start exploring - at molecular and systematic levels - how and why a genome can direct the alteration of phenotypic and reproductive traits by stimuli from outside. For example, environmental stresses such as high population density or predation trigger unwinged aphid females to produce winged offspring for dispersal [3, 4], and aphid females also sense changing photoperiods to alter reproductive modes [5–7]. However, how the environmental signals are received by the genome and where the sequences required for polyphenic switches are located in the genome remain largely unknown.
Embryogenesis in aphid life cycles can be categorized as two types: asexual viviparous (bringing forth living young without fertilization) and sexual oviparous (producing fertilized eggs). In the remaining paragraphs of this article we abbreviate ‘asexual viviparous’ and ‘sexual oviparous’ to ‘viviparous’ and ‘oviparous’ , respectively. Recent studies suggest that there are two developmental programs underlying early development of the viviparous and oviparous embryos in the pea aphid. For example, the expressions of axis patterning genes (hunchback, orthodenticle, caudal, nanos) and the terminal gene torso-like display distinctly different patterns during early development in viviparous compared to oviparous embryos [8–11]. Comparison of transcriptomes of synchronized embryos in asexual and sexual phases shows that more than 30 genes are differentially transcribed, further supporting the existence of divergent programs of development underlying these two reproductive cycles . It appears that the change of developmental programs occurs during the transition period from asexual to sexual phases - our previous work indicates that the germline-specific identity of Api-piwi6 and Api-ago3a, both of which encode components of the Piwi-interacting RNA (piRNA) pathway, is lost from viviparity to oviparity . Accordingly, we would like to know whether embryonic development is globally changed in distinct reproductive phases or whether some developmental mechanisms remain conserved.
Previously we identified posterior localization of Vasa (Vas) and Nanos (Nos), two of the most conserved germline markers in animals, in the syncytial blastoderm of the asexual aphid . After cellularization, signals of Vas and Nos detected using cross-reacting antibodies were specifically recruited into the primordial germ cells (PGCs), suggesting that germline specification in the asexual aphid relies on a germ plasm preformed within the egg prior to cellularization . However, we have not found any published descriptions of the assembly of maternal germ plasm in oocytes or early embryos in the sexual phase. To date, early segregation of germ cells in oviparous embryos has been reported in just two species: the black willow aphid Melanoxanthus spp.  and the spring grain aphid Toxoptera graminum. In the black willow aphid, newly segregated ‘primitive germ cells’ were identified in the posterior region of the invaginating blastoderm. In the spring grain aphid, germline segregation was observed slightly later during gastrulation and segregated germ cells were located aside the posterior abdomen of the elongating germ band (embryo proper). In both aphid species, the presence of germ (polar) granules - an ultrastructural feature of the preformed germ plasm - was not described. Taken together, this suggests that germline specification in oviparous aphids is not germ-plasm dependent and, if this is the case, aphids may adopt two programs to specify germ cells in asexual and sexual phases.
In some existing animal models such as Drosophila melanogaster (fly), Caenorhabditis elegans (nematode), Danio rerio (zebrafish), and Xenopus laevis (frog), it is clear that specification of germ cells is driven by the maternal germ plasm assembled within the developing oocytes (reviewed in [17, 18]). Removal of maternal germ plasm leads to the loss of PGCs or sterility in adults [19–21]. In Mus musculus (mouse), however, germline specification occurs much later during gastrulation and there is no maternally inherited germ plasm. Instead, specification of germ cells is triggered via induction of signals (the ‘induction mode’): Bmp4, Bmp8b, and Bmp2, members of the bone morphogenetic protein family, are secreted from the extra-embryonic ectoderm to induce certain cells in the neighboring proximal epiblast - the primary ectoderm - to form the precursors of PGCs [22–24]. After PGC precursors migrate to the posterior primitive streak, germline fate is further specified through the function of Blimp1, a key transcription factor known to repress the somatic program in the PGCs . Although cytoplasmic inheritance of maternal germ plasm, the ‘inheritance mode’, is prevalent in most current animal models, the induction mode actually appears more common, and is likely ancestral, when other published data about germline development in emerging and non-model animals are collated (reviewed in ).
Among insect species, the inheritance mode for germline specification is mostly confined to Holometabola such as flies, mosquitoes, and wasps [27, 28], all of which undergo complete metamorphosis. The lack of evidence to support the existence of maternal germ plasm in some holometabolous species like Bombyx mori (silkworm) [29, 30], Apis mellifera (honeybee) [31, 32], and Tribolium castaneum (beetle) [33, 34], however, indicates that the inheritance mode is not a universal mechanism in Holometabola. By contrast, evidence from classical histological studies argues that insects adopting the induction mode are the majority in Hemimetabola, which in contrast to Holometabola undergo incomplete metamorphosis and are less derived. Ubiquitous expression of conserved germline genes such as vas, piwi, and nos in oocytes and early embryos of Hemimetabola further supports the above conclusion [35–37]. A recent report concerning the abolishment of PGC formation via knockdown of the mesodermal marker twist in the cricket Gryllus bimaculatus - a model for the basally branching insects - suggests the presence of signal induction for germline specification in hemimetabolous insects , though the actual signaling molecules are not yet identified.
Current data show that germline specification in Hemiptera is diverse. As described above, we have identified a presumptive germ plasm assembled in the posterior syncytial blastoderm of the viviparous pea aphid - a hallmark of the inheritance mode . In the milkweed bug Oncopeltus fasciatus (Hemiptera: Lygaeidae), however, asymmetric localization of germline genes is not detected in oocytes and early embryos, and the segregation of PGCs is not identified until the late blastoderm stage. Moreover, knockdown of the conserved germline markers vas and tudor does not affect the formation of PGCs. All of this evidence implies the absence of a maternal germ plasm and the existence of an inductive mechanism in Oncopeltus. Similar conclusions have also been described in germline development in other hemipteran species such as the assassin bug Rhodnius prolixus (Hemiptera: Reduviidae)  and several species of scale insects [39, 40]. Accordingly, the germ plasm-dependent mechanism found in the viviparous pea aphid so far appears to be a non-canonical case in the Hemiptera in contrast to most other hemipterans. Recent studies show that the oviparous embryos adopt conserved mechanisms to pattern early embryogenesis while the viviparous embryos use a distinct and less conserved mechanism. Given the gross morphological similarities in oviparous aphid development with other hemipterans [38–41], it would be interesting to know if germline specification is also similar.
In order to explore how germ cells are specified during oviparous development in the pea aphid, we employed whole-mount in situ hybridization of Apvas1 (previously known as Apvas, a Drosophila vas ortholog in the pea aphid), which we successfully used previously to label germline cell fate in viviparous embryos . However, asymmetric localization of Apvas1 mRNA, unlike that of the Vas signals, was not identified in the viviparous pea aphid. In order not to miss detecting Vas protein restricted to the potential germ plasm, we produced an antibody against ApVas1 for immunostaining. Through comparing germline specification in the viviparous and oviparous embryos, we aimed to identify similarities and differences between the development of germ cells in both of the reproductive cycles. Most importantly, this would reveal whether aphids adopt one or two versions of mechanisms for specifying germ cells in their life.
Pea aphid culture
The pea aphid strain ApL, formerly known as Sap05Ms2, was collected in Hokkaido, Japan and reared on broad beans (Vicia faba) in growth chambers under a long-day photoperiod (16 h light/8 h dark) at 20°C . Induction of the sexual phase was carried out under a short-day photoperiod (8 h light/16 h dark) at 15°C. Details of induction were described in Ishikawa et al. [44, 45]. Fertilized eggs for embryonic assays were maintained under short-day photoperiod at 15°C for 42 days. Embryonic staging for oviparous and viviparous embryogeneses followed Miura et al. .
Production and purification of antibody against ApVas1
In order to avoid the conserved motifs of the DEAD (Glu-Asp-Ala-Glu)-box protein family, highly divergent sequences in the N-terminus of ApVas1 were selected and synthesized to induce the ApVas1 antibody (Additional file 1: Figure S1A). This ApVas1 antigen contains 451 amino acids (aa) in the N-terminal region (aa 4-454). A translation template of Apvas1 was cloned using a forward primer (5'-GATCAGATCTGGGTGGTTGGGATGATGAATCTGG-3', encoding GGWDDESG) and a reverse primer (5'-AATTGAATTCCTATTCTCTCTCTGGTTGTTCACGATCAC-3', encoding GDREQPERE). Products of the polymerase chain reaction (PCR) were then subcloned into the pET32a(+) vector (Novagen, San Diego, CA, USA) to produce a His-tagged fusion protein in Escherichia coli C41(DE3) cells. Affinity purified His-tagged ApVas1 protein in homogenized polyacrylamide gel slices was used to immunize rabbits. For antibody purification, antisera were passed through a column packed with the cyanogen bromide (CNBr)-activated Sepharose 4B (GE Healthcare, Uppsala, Sweden) coupled with a polypeptide containing 186 aa in the N-terminus of ApVas1 (aa 4-189) (Additional file 1: Figure S1B). After non-specific bindings were washed off, the ApVas1 antibody was eluted with 0.1 M Glycine at pH 2.5 and immediately neutralized to pH 7.0 using 1 M Tris pH 8.0. The neutralized ApVas1 antibody was dialyzed against 1× phosphate buffered saline (PBS) and then stored at -80°C in 50% glycerol.
Fixation and dissection of oviparous eggs
Aphid eggs were dechorionated in a 1:1 solution of methanol and heptane for 1 min with vigorous shaking. The heptane-methanol solution was removed without disturbing the dechorionated eggs after they had sunk to the bottom of the 1.5 mL Eppendorf tube. Eggs were then sequentially washed with mixture of methanol and 4% paraformaldehyde using ratios of 3:1, 1:1, and 1:3. After that, they were fixed in 4% paraformaldehyde for 20 min with mild shaking on a rotator. Fixed embryos collected before the end of 2 dAEL (days after egg laying) could be directly stained with riboprobe or antibody. From 3 dAEL onward, however, the serosal membrane that would turn black during staining had to be removed using a tungsten needle before staining. For embryos collected by the end of 4 dAEL we did not separate them from the yolk, but we remove yolk for those collected from 5 dAEL to increase the accessibility of riboprobe or antibody to the embryos.
In situ hybridization, immunostaining, and imaging
Whole-mount in situ hybridization (WISH) on oviparous embryos was performed according to the protocol previously applied to the viviparous ovarioles . Before application of the riboprobe, further fixation of dechorionated eggs or dissected embryos using 4% paraformaldehyde for 4 h at 25°C was required. In order to increase the probe specificity, we synthesized a Digoxigenin (DIG)-labeled antisense riboprobe complementary to the 149 nucleotides located in the 5’ untranslated region (UTR) of Apvas1, assuming that cross-hybridization would not occur to Apvas2-4 mRNAs. In situ hybridization showed that the expression pattern detected using the 149-bp Apvas1 antisense riboprobe differed from the expressions of Apvas2-4 (Additional file 2: Figure S2). This indicates that the 149-bp probe is Apvas1-specific and we have used it for all detections of Apvas1 expression in this study. Sequences of the forward and reverse primers for amplifying the 5’ UTR of Apvas1 are 5'- TTTGGCGGTGATGATGGAGAAG-3' and 5'-AATCGAGTCTAGGTGGCAACG-3', respectively. WISH was carried out at 66°C for 36 h.
Immunostaining followed the protocol of Chang et al., where viviparous ovarioles were stained using cross-reacting antibodies against Vas and Nos. For oviparous embryos, an additional fixation using 4% paraformaldehyde for 30 min at 25°C was required before application of the primary antibody. Antibodies and their dilution ratios were: (1) primary antibodies: ApVas1 antibody (1:20), ApVas2-4 antibodies (1:20), mouse anti-actin antibody (1:500) (Sigma, St Louis, MO, USA); and (2) secondary antibodies: anti-rabbit IgG conjugated with Cy5 (Invitrogen, Paisley, UK), Alexa Fluor 488 goat anti-mouse (Invitrogen), biotinylated goat anti-rabbit IgG (Vector labs, Burlingame, CA, USA) (all diluted 1:500). Nuclear staining of embryos older than 2 dAEL used 4',6-diamidino-2-phenylindole (DAPI) (2 μg/mL) (Sigma) at 37°C for 2 h, but those collected within 48 hours after egg laying (hAEL) were stained using propidium iodide (PI) (10 μg/mL) (Invitrogen) at 37°C for 2 h to avoid interference by autofluorescence from the yolk. F-actin staining on oviparous ovarioles was performed with Rhodamine Phalloidin (1:40) (Invitrogen); however, this was carried out using the anti-actin antibody (Sigma) in embryos prefixed with solutions containing methanol.
Bright-field images were photographed using a BX51 microscope (Olympus, Melville, NY, USA) connected with a DP50 CCD camera (Olympus); fluorescence images were taken by the laser-scanning microscope Zeiss LSM510 META (Carl Zeiss, Jena, Germany).
About 50 ng of each ApVas fusion protein was loaded for protein electrophoresis (Figure 1A). Unfertilized eggs after oviposition were collected from females without mating; fertilized eggs were collected in parallel within 16 hAEL. Protein extracted from 10 unfertilized/fertilized eggs was dissolved in 100 μL of 2× sample buffer (57.7 mM Tris, 10% Glycine, 2% SDS, 10% glycerol, 12.5 mM EDTA, 0.02% bromophenol blue 0.02%, 1% beta-mercaptoethanol) and 6 μL of extracts was loaded onto each lane (Figure 1A). Primary antibody: anti-ApVas1 antibody (1:1,000); secondary antibody: goat peroxidase-labeled antibody against rabbit IgG (1:2,000) (Kirkegaard and Perry Laboratories (KPL), Gaithersburg, MD, USA). Signals were developed using the VisGlow™ Chemiluminescent Substrate, Horseradish Peroxidase (HRP) system (Visual Protein, Taipei, Taiwan).
Maternal expression of Apvas genes in ovarioles of the sexual female
The pea aphid has four vasa homologs (Apvas1-4), all of which encode an ATP-dependent RNA helicase of the DEAD-box protein family . For detecting the expressions of Apvas mRNA and protein, we synthesized antisense riboprobes of Apvas1-4 and antibodies against ApVas1-4, respectively. Among the four Apvas genes, Apvas1 is the sole germline marker throughout viviparous development while expressions of Apvas2-4 are not germline specific (Additional file 2: Figure S2) [2, 42]. We therefore examined the expression of Apvas1 in the sexual phase, assuming that it was also a germline marker in the oviparous embryos.
In situ hybridization results showed that none of the four Apvas mRNAs, including Apvas1, was asymmetrically localized in the oocytes (Additional file 3: Figure S3). In order to examine whether ApVas1 was asymmetrically localized or uniformly distributed, we stained the ovarioles with the affinity-purified antibody against ApVas1. Purification of this antibody was specifically carried out using an immobilized antigen composed of 186 amino acids in the N-terminal region of ApVas1 to avoid antisera cross-reacting with downstream sequences that contain the eight conserved motifs of the DEAD-box protein family that is in all ApVas proteins (Additional file 1: Figure S1B). Western blots showed that the ApVas1 antibody could specifically detect ApVas1 and a major protein with expected size of the full-length ApVas1 (62.8 kD) from the unfertilized and fertilized eggs collected within 16 hAEL (Figure 1A). This suggests: (1) the affinity-purified antibody is specific to ApVas1; and (2) the newly laid eggs inherit maternal ApVas1 from the ovaries.
Immunostaining was performed on dissected ovarian tubules (ovarioles) encompassing the developing oocytes. In order to cover all developmental stages of oogenesis, we stained ovarioles dissected from both nymphs and adults. The nymphal aphids contain a higher proportion of previtellogenic oocytes whilst many more vitellogenic oocytes can be obtained from the adult aphids. Staining results showed that ApVas1 was uniformly expressed in the cytoplasm of nurse cells and that of the previtellogenic oocytes (Figure 1B, C). In the vitellogenic oocytes, we detected transient localization of ApVas1 in both anterior and posterior regions (Figure 1D). We conclude that localized signals observed in the vitellogenic oocytes were ApVas1: (1) localization of ApVas1 was not detected in oocytes that were not stained using the ApVas1 antibody (negative control, Figure 1F); and (2) expression patterns of ApVas2-4 (Additional file 4: Figure S4) were different from those of ApVas1 (Figure 1B-E). In mature oocytes subjected to oviposition, ApVas1 expression became undetectable (Figure 1E). We infer that the enlargement of egg chambers dilutes the intensity of the ApVas1 signal, because a major band corresponding to the size of ApVas1 could still be detected on the western blot (Figure 1A).
Posterior localization of ApVas1/Apvas1 in the syncytial and cellularized blastoderm
In the newly laid eggs, we did not identify localized signals of ApVas1 before the completion of the fourth cycle of nuclear division (Figure 2A, B). At the time that we first observed localized ApVas1 (fifth nuclear cycle, Figure 2C), energid nuclei had not yet reached the embryonic surface, suggesting that ApVas1 is not zygotically synthesized and is more likely maternally provided. By the end of the eighth cycle, we did not identify any energid nuclei on the cortex (Figure 2D, E). Migration of nuclei to the inner periphery of the egg - a hallmark of the formation of the syncytial blastoderm - was first identified from the ninth nuclear division onward (Figure 2F). Hence, we could find dividing nuclei within the ApVas1 stripe (Figure 2F, G). As previously described by Miura et al., we also identified a mitotic wave along the anteroposterior axis from the ninth to tenth nuclear cycles, which displayed asynchronous nuclear divisions in different regions of the egg. However, localization of ApVas1 remained unaffected adjacent to the endosymbiotic bacteria (bacteria) at the posterior end of the egg, regardless of the status of nuclear divisions within the ApVas1 stripe (Figure 2F, G, G’).
In order to monitor the status of cellularization, we co-stained embryos using an antibody against the filamentous actin (F-actin). In insects such as Drosophila, crickets, and grasshoppers, formation of the ‘actin cap’ that surrounds each superficial energid nucleus has been regarded as a sign of cellularization and after cellularization F-actin is located to the inner periphery of the cell membrane [49–51]. We adopted these conserved features of F-actin polymerization to monitor cellularization in the pea aphid. During 16 to 24 hAEL, morphology of the actin caps was visualized and meanwhile we identified the expression of ApVas1 associated with the cytoplasm surrounding the energid nuclei (Figure 3A-A”). In the time window between 24 and 28 hAEL, we found that ApVas1 was incorporated by the forming cell membrane whose inner periphery was enriched with F-actin (Figure 3B, B’ , C’ , C”). We note that cellularization may not have been completed because within the interior region of the ApVas1-positive energids signals of F-actin staining are not detected (Figure 3C).
Apart from the ApVas1 protein, we also monitored the expression of Apvas1 mRNA in parallel. In situ results showed that Apvas1 transcripts were ubiquitously distributed in germaria, oocytes (Additional file 3: Figure S3A-C, F), and eggs collected within 24 hAEL (Additional file 5: Figure S5A). Posterior localization of Apvas1 was not identified until 28 hAEL (Figure 3D). Like posterior localization of ApVas1 in embryos collected within the same developmental period (24 to 28 hAEL; Figure 3C, C’), Apvas1 mRNA was enriched to a stripe area within the egg posterior (Figure 3D). Although we have not found optimal conditions to perform dual immunostaining and in situ hybridization on pea aphid embryos, the similar locations of the ApVas1 and Apvas1 stripes (Figure 3C’ , D) and the synchronous migration of ApVas1/Apvas1-positive cells in later embryogenesis (Figure 4) suggest that ApVas1 and Apvas1 are co-localized.
Migration of cells expressing ApVas1/Apvas1 during germ band formation
Prior to invagination of the blastoderm, we observed the increase of ApVas1-positive cells. Immunostaining results showed that the ApVas1 stripe expanded from two to three cells in width, and the total cell numbers increased from about 40 to 160 within the stripe (Figure 4A, A’ , B, B’). Initiation of invagination could be visualized by the appearance of invaginating furrows in the posterior region of the eggs (Figure 4E, F). According to our observations, invagination of the blastoderm formed the blastopore, pulled bacteria into the interior region of the egg, and meanwhile dragged the ApVas1 stripe on the cortex to migrate further posteriorly toward the posterior pole of the egg (Figure 4E, F, H). When bacteria became fully immersed within the yolk, the ApVas1 stripe located to the posterior cortex followed the movement of bacteria, migrating backward to the inner region of the egg (Figure 4I-L). Later, bacteria and ApVas1-positive cells further migrated toward the egg center (Figure 4N-P). Here, we confirm that PGCs travel over, rather than through, the bacteria. Taking embryos presented in Figure 4I and J as an example, these two embryos belong to the same preparation but are shown at different focal planes (4I, I’: cortex; 4 J, J’: interior region). If PGCs really traveled through the bacteria, we would not be able to observe an integrative stripe composed of ApVas1 positive cells without the interference of bacteria (Figure 4I) or the absence of ApVas1 positive cells within the bacteria (Figure 4J).
Like ApVas1, corresponding patterns of migration were also identified in the migration of Apvas1-positive cells during blastoderm invagination (Figure 4C, G, K, O). The chromogenic in situ results, moreover, allowed us to visualize the morphology of the invaginating blastoderm more clearly (Figure 4O), because under a bright-field microscope the shape of the blastoderm was not masked by the adjacent bacterial mass that emitted strong nuclear signals under a confocal microscope (Figure 4M, N). Identical patterns of blastoderm invagination and germ band formation have been described in embryogenesis of the black willow aphid Melanoxanthus spp.  and the spring grain aphid T. graminum (Additional file 6: Figure S6C) , suggesting that movement of Vas/vas-positive cells is likely conserved among the aphid species.
Migration of cells expressing ApVas1/Apvas1 during germ band extension, katatrepsis, and late embryogenesis
Although bacterial cells were migrating ahead of the ApVas1/Apvas1-positive cells during germ band formation (Figure 4E-P), these two populations of cells reached the posterior end of the germ band that was about to elongate almost at the same time (Figure 5A). However, their co-localization broke up while the germ band continued to extend. The ApVas1-positive cells, as a consequence, became located between the bacteria and the elongating germ band (Figure 5B, C). We infer that further migration of bacteria and ApVas1-positive cells toward the egg anterior was driven by the extension of the germ band.
During germ band extension, the morphology of the cephalic lobe, thorax, and abdomen became distinct. Meanwhile, we found that the morphology of the ApVas1 stripe was also changed: the stripe was transformed into a globular structure, extra-embryonically located adjacent to the elongating abdomen (Figure 5D-F). When embryos were fully segmented and the limb buds became visible ApVas1-positive cells aggregated to the posterior region, forming a U shaped ring (Figure 5G, G’). Prior to katatrepsis (embryo flip), we found that ApVas1-positive cells, together with the bacteria, were separated into clusters within the dorsal side of the embryo (Figure 5H, H’; Additional file 7: Figure S7A, B). After katatrepsis, the ApVas1-positive cells were located bilaterally in the dorsal region of the abdomen (Figure 5I, I’) [46, 52]. In parallel, we also examined the expression of Apvas1 in the elongating germ band (Additional file 8: Figure S8A, B), finding that it was restricted to the same places where ApVas1 was expressed (Additional file 8: Figure S8C, D). This implies that in later stages of development Apvas1 mRNA and ApVas1 protein are co-localized in the germ cells.
Apvas1 gene as a germline marker in both sexual and asexual females
The ultimate goal of this study is to understand how germ cells are specified in the oviparous pea aphid. In particular, we are interested in whether or not germline specification relies on a preformed germ plasm. Here, we employed vas as a germline marker to monitor the assembly of germ plasm and migration of germ cells. Although vas has been regarded as the most conserved germ-cell specific gene in animals, in some species vas mRNA and Vas protein are not concurrently localized to the preformed germ plasm. For example, in Drosophila it is Vas protein, rather than vas mRNA, that is localized to the germ (pole) plasm [53, 54]. By contrast, in zebrafish preformed germ plasm within the cleavage planes is enriched in vas mRNA, rather than the uniformly expressed Vas protein, in two cell-stage embryos [55–58]. In order to maximize the opportunities for identifying the germ plasm in the oviparous pea aphid, we accordingly explored the expression of both Apvas1 (mRNA) and ApVas1 (protein).
Among previous studies of embryogenesis in sexual females, germline development has been best characterized in T. graminum (spring grain aphid). According to Webster and Phillips , segregation of the ‘primitive germ cells’ - an old term for the primordial germ cells (PGCs) - in T. graminum was first detected extra-embryonically adjacent to the posterior abdomen of the elongated germ band. This location corresponds to where Apvas1/ApVas1 was expressed in embryos at equivalent developmental stage in the pea aphid (Figure 5D, E; Additional file 8: Figure S8). Specific expression of Apvas1/ApVas1, moreover, could be identified in germ cells that were morphologically identifiable in the dorsal region of embryos after katatrepsis (Apvas1: data not shown; ApVas1: Figure 5I, Additional file 7: Figure S7A). The above evidence suggests that Apvas1/ApVas1-positive cells are PGCs in the oviparous embryo. Taken together with our previous work showing that Apvas1 and ApVas1  are also germline-specific in the viviparous embryo, we conclude that the Apvas1 gene is a germline marker of pea aphids in both sexual and asexual phases.
Identification and assembly of the preformed germ plasm in the sexual female
The preformed germ plasm is usually assembled in a subcellular area within the egg prior to cellularization. Therefore, asymmetric localization of germline markers that will be inherited by the PGCs has become a geographical signature of the preformed germ plasm (reviewed in ). In the oviparous embryo, we could detect the asymmetric localization of ApVas1 in the eggs prior to cellularization (Figure 2C-G) and monitor the incorporation of ApVas1 into presumptive PGCs during cellularization (Figure 3B’). After katatrepsis, locations of the ApVas1-positive PGCs (Figure 5I, I’) in the dorsal region of the embryos correspond to the morphologically identifiable germ cells in the viviparous aphids at equivalent developmental stages [42, 46]. Accordingly, we conclude that the subcellular cytoplasm restricting ApVas1 expression in the posterior cortex of the early syncytial blastoderm is the preformed germ plasm (Figure 2C).
In early embryos of Drosophila and the pea aphid, Vas protein is localized to the germ plasm whereas vas mRNA is uniformly distributed (Figure 2C-G, Additional file 5: Figure S5A) [14, 42, 53, 54]. This suggests that Vas protein, rather than vas mRNA, is involved in the assembly of germ plasm in both species. However, the timings of Vas localization are different: in Drosophila posterior localization of Vas to the germ plasm occurs during mid oogenesis, but in the oviparous embryos a subcellular cytoplasm containing ApVas1 expression - the presumptive germ plasm - cannot be identified through oogenesis (Figure 1B, C). Although in some vitellogenic oocytes posterior localization of ApVas1 is identified (Figure 1D), we do not consider this the onset of germ-plasm assembly because later ApVas1 localization cannot be continuously detected in the posterior region of mature oocytes subjected to oviposition (Figure 1E) and newly laid eggs (Figure 2A, B). Besides, locations of ApVas1 localization in the oocytes and embryos do not exactly correspond to each other - ApVas1 is localized to the posteriormost end of the oocyte (Figure 1D) while in the embryo the ‘ApVas1 stripe’ is already located anteriorly to the bacteria (Figure 2C).
Compared with Drosophila, the oviparous pea aphid segregates germ plasm much later: not until the fifth nuclear division after fertilization (Figure 2C). In other insects that adopt the inheritance mode, the assembly of germ plasm also takes place during oogenesis [27, 28, 60]. Thus the ‘post-fertilization’ assembly of germ plasm in the oviparous aphid is so far unusual. Nonetheless, a similar situation can be found in the nematode C. elegans, where fertilization triggers the posterior localization of maternal P granules (germ granules) . At present we do not have experimental data to show whether the assembly of germ plasm in the oviparous aphid is fertilization dependent or independent. However, the segregation of germ plasm in the viviparous aphid does not depend on sperm entry . If aphids of both reproductive phases share the same mechanism of germline specification then fertilization may not play a role in the assembly of germ plasm in the oviparous embryo. On the other hand, unless we can examine the distribution of all transcripts specific to germ cells, we cannot exclude the possibility that the germ plasm has already been assembled during oogenesis via depositing germline determinants to localize ApVas1, or locally translate Apvas1, after fertilization.
Comparison of germline specification and migration in oviparous and viviparous pea aphids
Although distinct developmental programs have been identified in oviparous and viviparous embryogenesis [8, 10, 11], our data show that the germ-plasm dependent specification of germ cells is conserved in both modes of aphid development. The early segregation of germ plasm, we infer, allows germ cells to locate in an extra-embryonic position so that the germline fate can be isolated from the somatic body plans. At a molecular level, differential expression of duplicated germline genes (for example, vas, nos, piwi, and ago3) during the switch of reproductive cycles and the establishment of sexual/asexual-specific germline transcriptomes may be required for germline reprogramming , but not for the specification of germ cells. Consequently, germline specification - at the morphological level - does not display developmental plasticity in the pea aphid and perhaps in other aphid species as well.
Additionally, we find that segregation of germ plasm and migration of germ cells in both oviparous and viviparous embryos follows a very similar pattern (Figure 6) [14, 42]: (1) the preformed germ plasm expressing Vas protein is not identified during oogenesis (S1/A1; S2/A2); (2) segregation of the preformed germ plasm occurs in the posterior region of syncytial blastoderm (S3/A3); (3) PGCs form in the posterior blastoderm after cellularization (S4/A4); (4) after germ band formation, PGCs are located posteriorly to the germ band regardless of its orientation in the egg chamber (S5/A5); (5) during germ band elongation, PGCs are aggregated and located extra-embryonically aside the elongating abdomen (S6/A6); and (6) from katatrepsis onward, PGCs migrate toward the anteriormost region of egg chambers and afterwards they migrate posteriorly (S7/A7). However, we find that migrating PGCs are already grouped into clusters prior to katatrepsis in the oviparous embryos (Figure 5G, H) but in the viviparous embryos PGCs remain unseparated at the equivalent stage of development .
Evolution and development of germline specification in aphids and other hemimetabolous insects
The preformed germ plasm, a morphological feature of the inheritance mode, has not been reported in any hemipterans apart from the aphids. For example, Pseudococcus mcdanieli (mealybug), Lecanodiaspis pruinosa (false pit scale), and Icerya purchasi (cottony cushion scale) are scale insects that all belong to the same suborder Sternorrhyncha as the aphids; however, in these species germ plasm was undetectable using classical histological approaches and segregated germ cells were first identified adjacent to the posterior blastoderm (Figure 7) . In Oncopeltus fasciatus (milkweed bug), a true bug belonging to the suborder Heteroptera, how Vasa protein is distributed during early development has not been reported; nevertheless, mRNAs of 19 homologues to Drosophila germline genes are not asymmetrically localized during early development, and formation of germ cells is not affected by the knockdown of conserved germline markers .
Taken together, the above cases suggest that absence of the preformed germ plasm is the norm in the Hemiptera although how germ cells are specified in the other two hemipteran suborders Auchenorrhyncha and Coleorrhyncha remains unclear (Figure 7). Accordingly, the common ancestor of hemipterans likely employed the induction mode to specify germ cells and the presence of germ plasm in aphids might be the result of independent evolution within the Hemiptera. Nonetheless, we cannot exclude the possibility that the gain of a germ plasm in aphids occurred after the parthenogenetic phase was added to the sexual life cycle more than 200 million years ago (reviewed in ), but how the inheritance mode was adopted in the sexual phase is still an unanswered question.
Though the inheritance mode is common neither in Hemiptera nor in other orders of the Hemimetabola, this mechanism has also been reported in the mullein thrips Haplothrips verbasci, a member of the order Thysanoptera . According to Heming, a putative germ (pole) plasm could be identified in the posterior end of the unfertilized egg and polar granules in the germ plasm can be inherited by the newly segregated germ cells after fertilization in H. verbasci. Taking together with the germ plasm identified in the pea aphid, we find that both cases - one in the Hemiptera (aphid) and one in the Thysanoptera (thrips) - occur within the Paraneoptera, a monophyletic superorder of insects and a sister group to the Holometabola (Figure 7) (reviewed in ). This implies that the inheritance mode can be identified in derived orders of hemimetabolous insects and the gain of a preformed germ plasm is achieved via independent evolution.
Identification of the preformed germ plasm in both oviparous and viviparous embryos of the pea aphid suggests that developmental plasticity of the germline-specification mode does not occur in the alternate reproductive cycles in the aphids. This differs from the distinct developmental programs known to regulate axis patterning in the asexual and sexual phases. The preformed germ plasm in the pea aphid so far is the first maternal germ plasm identified among hemimetabolous insects using molecular approaches, and thus provides an opportunity for studying how a germ plasm is assembled in the Hemimetabola. It has been clear that a homolog of oskar (osk), a molecular anchor known to localize Vas protein in the germ plasm of Drosophila, does not exist in the pea aphid genome . Finding molecular anchors for ApVas1 localization in the germ plasm will thus enable us to understand how an ‘osk-independent’ molecular networking for germ plasm assembly functions in less derived insects.
day after egg laying
differential interference contrast
hour after egg laying
phosphate buffered saline
polymerase chain reaction
primordial germ cells
whole-mount in situ hybridization
The International Aphid Genomics Consortium: Genome sequence of the pea aphid Acyrthosiphon pisum. PLoS Biol. 2010, 8: e1000313. 10.1371/journal.pbio.1000313.
Shigenobu S, Bickel RD, Brisson JA, Butts T, Chang C-c, Christiaens O, Davis GK, Duncan EJ, Ferrier DEK, Iga M, Janssen R, Lin GW, Lu HL, McGregor AP, Miura T, Smagghe G, Smith JM, van der Zee M, Velarde RA, Wilson MJ, Dearden PK, Stern DL: Comprehensive survey of developmental genes in the pea aphid, Acyrthosiphon pisum: frequent lineage-specific duplications and losses of developmental genes. Insect Mol Biol. 2010, 19: 47-62.
Dixon AFG: Aphid Ecology. 1998, London: Chapman & Hall
Sutherland ORW: The role of crowding in the production of winged forms by two strains of pea aphid, Acyrthosiphon pisum. J Insect Physiol. 1969, 15: 1385-1410. 10.1016/0022-1910(69)90199-1.
Lamb RJ, Pointing PJ: Sexual morph determination in the aphid, Acyrthosiphon pisum. J Insect Physiol. 1972, 18: 2029-2042. 10.1016/0022-1910(72)90170-9.
MacKay PA: Production of sexual and asexual morphs and changes in reproductive sequence associated with photoperiod in the pea aphid, Acyrthosiphon pisum (Harris). Can J Zool. 1987, 65: 2602-2606. 10.1139/z87-394.
Via S: Inducing the sexual forms and hatching the eggs of pea aphids. Entomol Exp Appl. 1992, 65: 119-127. 10.1111/j.1570-7458.1992.tb01635.x.
Bickel RD, Cleveland HC, Barkas J, Jeschke CC, Raz AA, Stern DL, Davis GK: The pea aphid uses a version of the terminal system during oviparous, but not viviparous, development. EvoDevo. 2013, 4: 10. 10.1186/2041-9139-4-10.
Chang C-c, Hsiao YM, Huang TY, Cook CE, Shigenobu S, Chang TH: Noncanonical expression of caudal during early embryogenesis in the pea aphid Acyrthosiphon pisum: maternal cad-driven posterior development is not conserved. Insect Mol Biol. 2013, 22: 442-455. 10.1111/imb.12035.
Duncan EJ, Benton MA, Dearden PK: Canonical terminal patterning is an evolutionary novelty. Dev Biol. 2013, 377: 245-261. 10.1016/j.ydbio.2013.02.010.
Duncan EJ, Leask MP, Dearden PK: The pea aphid (Acyrthosiphon pisum) genome encodes two divergent early developmental programs. Dev Biol. 2013, 377: 262-274. 10.1016/j.ydbio.2013.01.036.
Gallot A, Shigenobu S, Hashiyama T, Jaubert-Possamai S, Tagu D: Sexual and asexual oogenesis require the expression of unique and shared sets of genes in the insect Acyrthosiphon pisum. BMC Genomics. 2012, 13: 76. 10.1186/1471-2164-13-76.
Lu HL, Tanguy S, Rispe C, Gauthier JP, Walsh T, Gordon K, Edwards O, Tagu D, Chang C-c, Jaubert-Possamai S: Expansion of genes encoding piRNA-associated argonaute proteins in the pea aphid: diversification of expression profiles in different plastic morphs. PLoS One. 2011, 6: e28051. 10.1371/journal.pone.0028051.
Chang C-c, Lee WC, Cook CE, Lin GW, Chang T: Germ-plasm specification and germline development in the parthenogenetic pea aphid Acyrthosiphon pisum: Vasa and Nanos as markers. Int J Dev Biol. 2006, 50: 413-421. 10.1387/ijdb.052100cc.
Tannreuther GW: History of the Germ Cells and Early Embryology of Certain Aphids, PhD Thesis. 1908, Chicago, IL: The University of Chicago, Department of Zoölogy
Webster FM, Phillips WJ: The Spring Grain-Aphis or "Green Bug". 1912, Washington, DC: Department of Agriculture, Bureau of Entomology
Ikenishi K: Germ plasm in Caenorhabditis elegans, Drosophila and Xenopus. Dev Growth Differ. 1998, 40: 1-10. 10.1046/j.1440-169X.1998.t01-4-00001.x.
Kosaka K, Kawakami K, Sakamoto H, Inoue K: Spatiotemporal localization of germ plasm RNAs during zebrafish oogenesis. Mech Dev. 2007, 124: 279-289. 10.1016/j.mod.2007.01.003.
Illmensee K, Mahowald AP: Transplantation of posterior polar plasm in Drosophila: induction of germ cells at the anterior pole of the egg. Proc Natl Acad Sci U S A. 1974, 71: 1016-1020. 10.1073/pnas.71.4.1016.
Hashimoto Y, Maegawa S, Nagai T, Yamaha E, Suzuki H, Yasuda K, Inoue K: Localized maternal factors are required for zebrafish germ cell formation. Dev Biol. 2004, 268: 152-161. 10.1016/j.ydbio.2003.12.013.
Tada H, Mochii M, Orii H, Watanabe K: Ectopic formation of primordial germ cells by transplantation of the germ plasm: direct evidence for germ cell determinant in Xenopus. Dev Biol. 2012, 371: 86-93. 10.1016/j.ydbio.2012.08.014.
Lawson KA, Dunn NR, Roelen BA, Zeinstra LM, Davis AM, Wright CV, Korving JP, Hogan BL: Bmp4 is required for the generation of primordial germ cells in the mouse embryo. Genes Dev. 1999, 13: 424-436. 10.1101/gad.13.4.424.
Ying Y, Liu XM, Marble A, Lawson KA, Zhao GQ: Requirement of Bmp8b for the generation of primordial germ cells in the mouse. Mol Endocrinol. 2000, 14: 1053-1063. 10.1210/mend.14.7.0479.
Ying Y, Zhao GQ: Cooperation of endoderm-derived BMP2 and extraembryonic ectoderm-derived BMP4 in primordial germ cell generation in the mouse. Dev Biol. 2001, 232: 484-492. 10.1006/dbio.2001.0173.
Ohinata Y, Payer B, O’Carroll D, Ancelin K, Ono Y, Sano M, Barton SC, Obukhanych T, Nussenzweig M, Tarakhovsky A, Saitou M, Surani MA: Blimp1 is a critical determinant of the germ cell lineage in mice. Nature. 2005, 436: 207-213. 10.1038/nature03813.
Extavour CG, Akam M: Mechanisms of germ cell specification across the metazoans: epigenesis and preformation. Development. 2003, 130: 5869-5884. 10.1242/dev.00804.
Juhn J, James AA: Oskar gene expression in the vector mosquitoes, Anopheles gambiae and Aedes aegypti. Insect Mol Biol. 2006, 15: 363-372. 10.1111/j.1365-2583.2006.00655.x.
Lynch JA, Ozuak O, Khila A, Abouheif E, Desplan C, Roth S: The phylogenetic origin of oskar coincided with the origin of maternally provisioned germ plasm and pole cells at the base of the Holometabola. PLoS Genet. 2011, 7: e1002029. 10.1371/journal.pgen.1002029.
Nakao H: Isolation and characterization of a Bombyx vasa-like gene. Dev Genes Evol. 1999, 209: 312-316. 10.1007/s004270050257.
Nakao H, Hatakeyama M, Lee JM, Shimoda M, Kanda T: Expression pattern of Bombyx vasa-like (BmVLG) protein and its implications in germ cell development. Dev Genes Evol. 2006, 216: 94-99. 10.1007/s00427-005-0033-8.
Nelson JA: The Embryology of the Honey Bee. 1915, Princeton, NJ: Princeton University Press
Dearden PK: Germ cell development in the Honeybee (Apis mellifera); Vasa and Nanos expression. BMC Dev Biol. 2006, 6: 6. 10.1186/1471-213X-6-6.
Lynch JA, El-Sherif E, Brown SJ: Comparisons of the embryonic development of Drosophila, Nasonia, and Tribolium. WIREs Dev Biol. 2012, 1: 16-39. 10.1002/wdev.3.
Schröder R: vasa mRNA accumulates at the posterior pole during blastoderm formation in the flour beetle Tribolium castaneum. Dev Genes Evol. 2006, 216: 277-283. 10.1007/s00427-005-0054-3.
Ewen-Campen B, Donoughe S, Clarke DN, Extavour CG: Germ cell specification requires zygotic mechanisms rather than germ plasm in a basally branching insect. Curr Biol. 2013, 23: 835-842. 10.1016/j.cub.2013.03.063.
Ewen-Campen B, Jones TE, Extavour CG: Evidence against a germ plasm in the milkweed bug Oncopeltus fasciatus, a hemimetabolous insect. Biol Open. 2013, 2: 556-568. 10.1242/bio.20134390.
Chang C-c, Dearden P, Akam M: Germ line development in the grasshopper Schistocerca gregaria: vasa as a marker. Dev Biol. 2002, 252: 100-118. 10.1006/dbio.2002.0840.
Heming BS, Huebner E: Development of the germ-cells and reproductive primordia in male and female embryos of Rhodnius prolixus Stål (Hemiptera, Reduviidae). Can J Zool. 1994, 72: 1100-1119. 10.1139/z94-148.
Shinji GO: Embryology of coccids, with especial reference to the formation of the ovary, origin and differentiation of the germ cells, germ layers, rudiments of the midgut, and the intracellular symbiotic organisms. J Morphol. 1919, 33: 73-168.
Schrader F: The sex ratio and oogenesis of Pseudococcus citri. Z Indukt Abstamm Vererbungsl. 1922, 30: 163-182.
Mellanby H: The early embryonic development of Rhodnius prolixus (Hemiptera, Heteroptera). Q J Microsc Sci. 1935, 78: 71-90.
Chang C-c, Lin GW, Cook CE, Horng SB, Lee HJ, Huang TY: Apvasa marks germ-cell migration in the parthenogenetic pea aphid Acyrthosiphon pisum (Hemiptera: Aphidoidea). Dev Genes Evol. 2007, 217: 275-287. 10.1007/s00427-007-0142-7.
Kanbe T, Akimoto S: Allelic and genotypic diversity in long-term asexual populations of the pea aphid, Acyrthosiphon pisum in comparison with sexual populations. Mol Ecol. 2009, 18: 801-816. 10.1111/j.1365-294X.2008.04077.x.
Ishikawa A, Hongo S, Miura T: Morphological and histological examination of polyphenic wing formation in the pea aphid Acyrthosiphon pisum (Hemiptera, Hexapoda). Zoomorphology. 2008, 127: 121-133. 10.1007/s00435-008-0057-5.
Ishikawa A, Ogawa K, Gotoh H, Walsh TK, Tagu D, Brisson JA, Rispe C, Jaubert-Possamai S, Kanbe T, Tsubota T, Shiotsuki T, Miura T: Juvenile hormone titre and related gene expression during the change of reproductive modes in the pea aphid. Insect Mol Biol. 2011, 21: 49-60.
Miura T, Braendle C, Shingleton A, Sisk G, Kambhampati S, Stern DL: A comparison of parthenogenetic and sexual embryogenesis of the pea aphid Acyrthosiphon pisum (Hemiptera: Aphidoidea). J Exp Zool B Mol Dev Evol. 2003, 295: 59-81.
Chang C-c, Huang TY, Shih CL, Lin GW, Chang TP, Chiu H, Chang WC: Whole-mount identification of gene transcripts in aphids: protocols and evaluation of probe accessibility. Arch Insect Biochem. 2008, 68: 186-196. 10.1002/arch.20243.
Linder P, Lasko PF, Ashburner M, Leroy P, Nielsen PJ, Nishi K, Schnier J, Slonimski PP: Birth of the D-E-A-D box. Nature. 1989, 337: 121-122. 10.1038/337121a0.
Postner MA, Miller KG, Wieschaus EF: Maternal effect mutations of the sponge locus affect actin cytoskeletal rearrangements in Drosophila melanogaster embryos. J Cell Biol. 1992, 119: 1205-1218. 10.1083/jcb.119.5.1205.
Sarashina I, Mito T, Saito M, Uneme H, Miyawaki K, Shinmyo Y, Ohuchi H, Noji S: Location of micropyles and early embryonic development of the two-spotted cricket Gryllus binaculatus (Insecta, Orthoptera). Dev Growth Differ. 2005, 47: 99-108. 10.1111/j.1440-169x.2005.00786.x.
Ho K, Dunin-Borkowski OM, Akam M: Cellularization in locust embryos occurs before blastoderm formation. Development. 1997, 124: 2761-2768.
Shingleton AW, Sisk GC, Stern DL: Diapause in the pea aphid (Acyrthosiphon pisum) is a slowing but not a cessation of development. BMC Dev Biol. 2003, 3: 7. 10.1186/1471-213X-3-7.
Hay B, Jan LY, Jan YN: A protein component of Drosophila polar granules is encoded by vasa and has extensive sequence similarity to ATP-dependent helicases. Cell. 1988, 55: 577-587. 10.1016/0092-8674(88)90216-4.
Lasko PF, Ashburner M: The product of the Drosophila gene vasa is very similar to eukaryotic initiation factor-4a. Nature. 1988, 335: 611-617. 10.1038/335611a0.
Yoon C, Kawakami K, Hopkins N: Zebrafish vasa homologue RNA is localized to the cleavage planes of 2- and 4-cell-stage embryos and is expressed in the primordial germ cells. Development. 1997, 124: 3157-3165.
Olsen LC, Aasland R, Fjose A: A vasa-like gene in zebrafish identifies putative primordial germ cells. Mech Dev. 1997, 66: 95-105. 10.1016/S0925-4773(97)00099-3.
Braat AK, Zandbergen T, van de Water S, Goos HJ, Zivkovic D: Characterization of zebrafish primordial germ cells: morphology and early distribution of vasa RNA. Dev Dyn. 1999, 216: 153-167. 10.1002/(SICI)1097-0177(199910)216:2<153::AID-DVDY6>3.0.CO;2-1.
Knaut H, Pelegri F, Bohmann K, Schwarz H, Nüsslein-Volhard C: Zebrafish vasa RNA but not its protein is a component of the germ plasm and segregates asymmetrically before germline specification. J Cell Biol. 2000, 149: 875-888. 10.1083/jcb.149.4.875.
Eddy EM: Germ plasm and the differentiation of the germ cell line. Int Rev Cytol. 1975, 43: 229-280.
Lynch JA, Desplan C: Novel modes of localization and function of nanos in the wasp Nasonia. Development. 2010, 137: 3813-3821. 10.1242/dev.054213.
Gruidl ME, Smith PA, Kuznicki KA, McCrone JS, Kirchner J, Roussell DL, Strome S, Bennett KL: Multiple potential germ-line helicases are components of the germ-line-specific P granules of Caenorhabditis elegans. Proc Natl Acad Sci U S A. 1996, 93: 13837-13842. 10.1073/pnas.93.24.13837.
Goss RJ: The early embryology of the book louse, Liposcelis divergens Badonnel (Psocoptera, Liposcelidae). J Morphol. 1952, 91: 135-167. 10.1002/jmor.1050910107.
Heming BS: Origin and fate of germ cells in male and female embryos of Haplothrips verbasci (Osborn) (Insecta, Thysanoptera, Phlaeothripidae). J Morphol. 1979, 160: 323-343. 10.1002/jmor.1051600305.
Butt FH: Embryology of the Milkweed Bug: Oncopeltus fasciatus (Hemiptera). 1949, New York, NY: Cornell University Agricultural Experiment Station
Grimaldi D, Engel MS: Evolution of the Insect. 2005, New York, NY: Cambridge University Press
Wheeler WC, Bang R, Schuh RT: Cladistic relationships among higher groups of Heteroptera: congruence between morphological and molecular data sets. Ent Scand. 1993, 24: 121-138. 10.1163/187631293X00235.
Cryan JR, Urban JM: Higher-level phylogeny of the insect order Hemiptera: is Auchenorrhyncha really paraphyletic?. Syst Entomol. 2012, 37: 7-21. 10.1111/j.1365-3113.2011.00611.x.
Davis GK: Cyclical parthenogenesis and viviparity in aphids as evolutionary novelties. J Exp Zool. 2012, 318B: 448-459.
We thank Hitoshi Miyakawa for technical support on confocal microscopy; Asano Ishikawa and Kota Ogawa for technical support on induction of sexual phase of pea aphids; Hsiao-Ling Lu, Yi-min Hsiao, and Chen-yo Chung for critical reading of the manuscript. This work was supported by the National Science Council (NSC98-2917-I-002-142 for GWL; 101-2313-B-002-059-MY3 and 101-2321-B-002-090-MY2 for GWL and CC), BAPHIQ of the Agricultural Council (22.214.171.124-B3(4)), and National Taiwan University (NTU-CESRP 101R4602D3).
The authors declare that they have no competing interests.
GWL: collection and analysis of data, experimental design, manuscript writing, critical revision of the manuscript. CEC: data analysis, manuscript writing, and critical revision of the manuscript. TM and CC: conception and design, data analysis, training and financial support, critical revision of the manuscript. All authors read and approved the final manuscript.