Permissive and Instructive Anterior Patterning Rely on mRNA Localization in the Wasp Embryo

See allHide authors and affiliations

Science  30 Mar 2007:
Vol. 315, Issue 5820, pp. 1841-1843
DOI: 10.1126/science.1137528


The long-germ mode of embryogenesis, in which segments arise simultaneously along the anteriorposterior axis, has evolved several times in different lineages of the holometabolous, or fully metamorphosing, insects. Drosophila's long-germ fate map is established largely by the activity of the dipteran-specific Bicoid (Bcd) morphogen gradient, which operates both instructively and permissively to accomplish anterior patterning. By contrast, all nondipteran long-germ insects must achieve anterior patterning independently of bcd. We show that bcd's permissive function is mimicked in the wasp by a maternal repression system in which anterior localization of the wasp ortholog of giant represses anterior expression of the trunk gap genes so that head and thorax can properly form.

The highly conserved segmented insect body plan is achieved by great flexibility in developmental mechanisms. In the ancestral short-germ mode of embryogenesis, head and thorax arise from the egg's posterior, and abdominal segments emerge progressively from a posterior growth zone. By contrast, in the more derived long-germ mode, all segments form simultaneously in a syncytial environment, with head and thorax at the egg'santerior. Long-germ development has evolved several times in different holometabolous insect lineages (1), including the Diptera, of which the most extensively studied member is Drosophila melanogaster (Dm). A morphogen gradient of Bicoid (Bcd) protein, formed by translation from a maternal, anteriorly localized mRNA source, establishes the Drosophila body plan. Embryos derived from bcd mutant mothers lack head, thorax, and some abdominal segments (2, 3).

Despite its critical role in patterning the Drosophila long-germ embryo, bcd is distinctive to the higher Diptera (4). Thus, all other insects, including long-germ nondipterans, must employ a bcd-independent mechanism to accomplish segmentation. To identify such a mechanism, we investigated anterior patterning in the hymenopteran parasitoid wasp, Nasonia vitripennis (Nv) (5). The embryonic fate map of this independently evolved (6) long-germ insect is essentially identical to that of Drosophila, except that it is formed in the absence of bcd. We previously showed that in the early Nasonia embryo, bcd's morphogenetic activity is performed by orthodenticle1 (Nv-otd1), the ortholog of the Drosophila bcd target gene, Dm-otd (5). Although strictly zygotic in Drosophila, Nv-otd1 mRNA is maternally provided and localized to both oocyte poles, resulting in bipolar protein gradients. The anterior Nv-otd1 gradient regulates expression of zygotic head and thoracic gap genes, including the Nasonia orthologs of the bcd targets, giant (gt), and hunchback (hb) (5).

In addition to instructively activating the genes that pattern the head and thorax, bcd also functions permissively in Drosophila to indirectly repress posteriorly acting genes, such as the trunk gap genes, that would otherwise inhibit anterior development (7). We show here that Nasonia accomplishes this task by further employing maternal mRNA localization to position a repressor of trunk development at the anterior, thereby allowing formation of the head and thorax.

In the Drosophila embryo, the gap gene Krüppel (Dm-Kr) is expressed in a broad central stripe (Fig. 1A) and is required for formation of thoracic segment 1 (T1) through abdominal segment 5 (A5) (8). The positioning of Kr and, hence, of the trunk, is established by bcd and the terminal system; in embryos derived from bcd mutant mothers, the Dm-Kr domain broadens and shifts anteriorly (Fig. 1B) (7, 9). bcd's zygotic targets, Dm-hb and Dm-gt, mediate this regulation; in single Dm-hb (7, 9) or Dm-gt mutant embryos (Fig. 1C), Dm-Kr shows slight anterior expansion (10), and in embryos mutant for both, Dm-Kr's anterior shift is comparable to that seen from loss of bcd alone (11). However, Dm-Kr does not reach the anterior tip of the embryo because of additional repression by the terminal system; in embryos lacking bcd and torsolike (Dm-tsl), Dm-Kr expands throughout the anterior (Fig. 1D), resulting in embryos with as few as four abdominal segments (Fig. 1, E and F) (7).

Fig. 1.

Regulation of Kr in Drosophila and Nasonia. Dm-Kr expression in embryos that are wild-type (A), lacking Dm-bcd (B), mutant for Dm-gt (C), or lacking Dm-bcd and Dm-tsl (D). (E) Wild-type or (F) Dm-bcd;tsl cuticle. Cellular blastoderm expression of Nv-Kr in wild-type embryos (G) or after Nv-otd1 RNAi (H).

To determine how the Nasonia embryo's trunk is positioned, we isolated the Nasonia ortholog of Kr and found that it was expressed in a central gap–like domain (Fig. 1G). To confirm that Nv-Kr was acting as a gap gene, we knocked down Nv-Kr using parental RNA interference (RNAi) (5, 12) and observed a gap phenotype: loss of T3 and A1 to A4. We hypothesized that because Nv-otd1 performs other bcd functions, the otd1 anterior morphogenetic gradient might be functioning to position Nv-Kr. However, after Nv-otd1 RNAi, we found that although Nv-Kr expanded posteriorly, the anterior boundary was unaffected (Fig. 1H).

We next asked whether one of bcd's zygotic gap gene targets could be implicated in Nv-Kr anterior boundary formation. Our analysis of Nv-gt revealed an essential role in repressing Nv-Kr. In Drosophila, Dm-gt is expressed in two regions, a posterior stripe and a bcd-activated anterior domain (13). Nv-gt expression in cellularized embryos was similar (Fig. 2E); however, we found Nv-gt expressed maternally at earlier stages. In freshly laid embryos, Nv-gt mRNA occupied a broad anterior gradient, with highest levels at the anterior pole (Fig. 2A). This domain persisted through pole cell formation (Fig. 2B) and syncytial blastoderm stages (Fig. 2C), intensifying at the onset of cellularization simultaneously with the appearance of a posterior cap, marking commencement of zygotic expression (Fig. 2D). During cellularization, Nv-gt expression resolved into zygotic anterior and posterior gap domains (Fig. 2E), which are regulated by Nv-otd1 (5). To confirm that the early Nv-gt expression we had observed was maternal, we examined egg chambers and detected Nv-gt in the nurse cells and in the oocyte, where it accumulated around the nucleus (Fig. 2F). In mature oocytes, Nv-gt was localized to the anterior pole (Fig. 2G), ultimately assuming a tight anterior localization pattern (Fig. 2H).

Fig. 2.

Expression of maternal and zygotic Nv-gt during Nasonia oogenesis and embryogenesis. Maternal Nv-gt mRNA in freshly laid embryos (A), during pole cell formation (B), and in the syncytial blastoderm (C). Zygotic Nv-gt expression at onset of cellularization (D) and in the cellular blastoderm (E). Maternal anterior localization during oocyte development (F to H).

Having established that Nv-gt is maternally provided and anteriorly localized, we used RNAi to determine its role during segmentation. Knock-down of Nv-gt resulted in unhatched larvae, all arrested with the same phenotype: complete loss of head and thoracic (T1-T3) segments (Fig. 3, A and B). In addition, A6 and A7 showed fusions and deletions due to loss of posterior zygotic Nv-gt (Fig. 3B). Anterior defects were also visible at earlier stages in knockdown embryos; analysis of engrailed (Nv-en) revealed anterior truncations (Fig. 3, C and D), whereas Ultrabithorax/AbdominalA (Ubx/AbdA) expression showed that all remaining tissue was abdominal (Fig. 3, E and F). The Nv-gt RNAi anterior phenotype was more severe than that of Dm-gt null mutations, in which anterior defects are limited to loss of labial and labral structures (13). By contrast, knockdown of Nv-gt resembled the all-abdominal phenotype observed in Drosophila embryos lacking both bcd and Dm-tsl (Fig. 1F).

Fig. 3.

Knockdown of Nv-gt in Nasonia results in anterior deletions. Wild-type (A) and Nv-gt RNAi (B) cuticles. Yellow arrows indicate spiracles on T2, A1 to A3. Expression of Nv-en (C and D), Nv-Kr (G and H), and Nv-hbzyg (I and J) in wild-type [(C), (G), and (I)] and Nv-gt RNAi [(D), (H), and (J)] embryos. Ubx/AbdA protein expression in wild-type (E) and Nv-gt RNAi (F) embryos.

In Drosophila, it has been shown that Dm-gt and Dm-Kr mutually repress each other, a relationship that defines the embryo's center (10, 11). To determine whether Nv-gt also regulates Nv-Kr, we examined Nv-Kr expression in Nv-gt RNAi embryos and found dramatic expansion of Nv-Kr to the anterior pole (Fig. 3, G and H), indicating a major role for Nv-gt in setting the Nv-Kr anterior boundary. Moreover, loss of zygotic Nv-gt alone did not result in anterior Nv-Kr expansion. As noted earlier, the anterior Nv-Kr border was unaffected by Nv-otd1 RNAi (Fig. 1H), whereas zygotic Nv-gt was eliminated (5). Thus, maternal Nv-gt appears to be the main repressor of Nv-Kr, mimicking both bcd and the terminal system in Drosophila.

The absence of head and thorax after Nv-gt RNAi was also similar to the phenotype of the Nasonia zygotic hb mutant headless (14). To investigate this correlation, we examined zygotic Nv-hb expression after Nv-gt RNAi and observed loss of the anterior Nv-hb domain (Fig. 3, I and J), suggesting that the anterior deletions seen in Nv-gt RNAi embryos were caused by loss of zygotic Nv-hb. Our observations generated two models for Nv-gt function: Either maternally localized Nv-gt activates expression of anterior zygotic Nv-hb (Fig. 4A), or Nv-gt negatively regulates a hb-repressor (Fig. 4A). Because Dm-Gt principally acts as a repressor (15), we favored the second model, hypothesizing that Nv-Kr might be the hb-repressor (Fig. 4A). In support of this model, we observed that the anterior and posterior zygotic Nv-hb domains expanded toward the center in Nv-Kr RNAi embryos (Figs. 3I and 4C). This double-repressor model for head and thorax patterning circumscribes maternally localized Nv-gt function solely to the repression of Nv-Kr.

Fig. 4.

Repression of Nv-Kr by maternal Nv-gt is required for head and thorax formation in Nasonia. (A) Two models for maternal Nv-gt function. Cuticular analysis (B, D, and F) and Nv-hbzyg expression (C, E, and G) after knockdown for Nv-Kr [(B) and (C)], Nv-gt+gfp [(D) and (E)], and Nv-gt+Kr [(F) and (G)].

To test this model, we asked whether expansion of Nv-Kr into the developing head and thoracic region causes deletion of those segments. If so, we would expect to see anterior patterning restored in embryos lacking Nv-gt and Nv-Kr. As a control for double RNAi, we examined embryos from females knocked down for Nv-gt and green fluorescent protein (gfp) and observed the expected Nv-gt phenotype: deletion of head and thorax, as well as loss of anterior Nv-hb expression (Fig. 4, D and E). Knockdown of Nv-gt and Nv-Kr yielded striking results. In 92% of examined embryos, the head and thorax (T1/T2) were restored (Fig. 4F), and the resulting cuticular phenotypes were essentially identical to those after Nv-Kr RNAi alone (Fig. 4B). Consistent with rescued head and thorax development, anterior zygotic Nv-hb was also restored, although not to wild-type levels (Fig. 4G). Nonetheless, the amount of Nv-hb present in Nv-gt+Kr RNAi embryos was sufficient to direct head and thorax development, demonstrating that Nv-Kr expansion impedes anterior patterning and that maternally localized Nv-gt confines Nv-Kr to the embryo's center. Thus, whereas in Drosophila, bcd-activated Dm-gt plays only a moderate role in positioning Nv-Kr (Fig. 1C), in Nasonia, maternal Nv-gt is sufficient to perform this function. This distinction led us to consider whether Dm-gt'srole in Drosophila would be enhanced if the Drosophila embryo were reengineered to develop like Nasonia—with Dm-gt maternally provided and anteriorly localized. We found that, whereas Dm-gt was sufficient to repress Dm-Kr anteriorly in the absence of bcd (fig. S1B), head and thoracic structures were not rescued (fig. S1C)—an unsurprising result given that, in addition to permitting anterior development by regulating Kr-repressing gap genes, bcd also functions instructively to activate genes required for head and thorax formation. In Nasonia, by contrast, the instructive and permissive anterior patterning functions are discrete. Head- and thorax-specific genes are triggered by an instructive anterior determinant, maternal Nv-otd1, which is localized independently of the permissively acting maternal repression system, Nv-gt.

A comparison of the molecular mechanisms employed by two independently evolved (6) long-germ insects not only uncovers those features essential to this developmental mode but also sheds light on how the bcd-dependent anterior patterning program might have evolved. Through analysis of the regulation of the trunk gap gene Kr in Drosophila and Nasonia, we have been able to demonstrate that anterior repression of Kr is essential for head and thorax formation and is a common feature of long-germ patterning. Both insects accomplish this task through maternal, anteriorly localized factors that either indirectly (Drosophila) or directly (Nasonia) repress Kr and, hence, trunk fates. In Drosophila, the terminal system and bcd regulate expression of gap genes, including Dm-gt, that repress Dm-Kr. Nasonia's bcd-independent long-germ embryos must solve the same problem, but they employ a maternally localized repression system in which maternal Nv-gt is localized to the oocyte's anterior, where it represses Nv-Kr. In the dipteran lineage, whereas gt retained the ability to repress Kr, maternal regulation of Kr'sposition was taken over by two novel features—bcd, a specific dipteran innovation, and the terminal pathway, which, although present ancestrally, appears to function less extensively in the anterior of nondipteran insects (16, 17). In addition to activating anterior patterning genes such as otd and hb, bcd also acquired regulation of gt, which became a strictly zygotic gene with a reduced role in repressing Kr. Our findings thus identify two independent mechanisms for long-germ anterior patterning—one using two maternally localized genes, otd1 and gt, that respectively activate anterior zygotic patterning genes and repress trunk fates, and a second using bcd for these same functions, thereby demoting otd and gt to zygotic gap genes. Interestingly, it appears that long-germ embryos use RNA localization for a number of different developmental processes (5, 18, 19). By contrast, in short-germ insects, although some localized RNAs have been identified, there is as yet no evidence of their contribution to anterior-posterior patterning (20). mRNA localization indeed appears to be an important component of long-germ embryogenesis, perhaps even playing a role in the transition from the ancestral short-germ to the derived long-germ fate.

Supporting Online Material

Materials and Methods

SOM Text

Fig. S1


References and Notes

View Abstract

Stay Connected to Science

Navigate This Article