Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Dev Biol. Author manuscript; available in PMC 2010 October 1.
Published in final edited form as:
PMCID: PMC2749522



Genetic studies of the fruit fly Drosophila have revealed a hierarchy of segmentation genes (maternal, gap, pair-rule and HOX) that subdivide the syncytial blastoderm into sequentially finer scale coordinates. Within this hierarchy, the pair-rule genes translate gradients of information into periodic stripes of expression. How pair-rule genes function during the progressive mode of segmentation seen in short and intermediate-germ insects is an ongoing question. Here we report that the nuclear receptor Of’E75A is expressed with double segment periodicity in the head and thorax. In the abdomen, Of’E75A is expressed in a unique pattern during posterior elongation, and briefly resembles a sequence that is typical of pair-rule genes. Depletion of Of’E75A mRNA caused loss of a subset of odd-numbered parasegments, as well as parasegment 6. Because these parasegments straddle segment boundaries, we observe fusions between adjacent segments. Finally, expression of Of’E75A in the blastoderm requires even-skipped, which is a gap gene in Oncopeltus. These data show that the function of Of’E75A during embryogenesis shares many properties with canonical pair-rule genes in other insects. They further suggest that parasegment specification may occur through irregular and episodic pair-rule-like activity.

Keywords: Pair-rule, Oncopeltus fasciatus, nuclear receptor, ecdysone-response, E75, intermediate-germ insect, segmentation, milkweed bug


The Drosophila paradigm for patterning the anterior-posterior (A-P) axis has provided a touchstone for comparative studies of segmentation in other insects and arthropods (Peel et al., 2005). This comprehensive model, revealed by saturation screens of the early embryo and oocyte, showed that the syncytial Drosophila blastoderm is progressively subdivided into finer scale coordinates by the maternal, gap and pair-rule genes (Nusslein-Volhard and Wieschaus, 1980; Schupbach and Wieschaus, 1986). Upon cellularization, these coordinates are integrated into finer-scale information by the segment polarity and homeotic (HOX) genes, which will specify the A-P segment compartment and segment identity, respectively (Lawrence, 1992). One surprising outcome of the Drosophila screens was the revelation that the blastoderm is transiently organized into parasegments by the pair-rule genes (Nusslein-Vollhard and Wieschaus, 1980; Martinez-Arias and Lawrence, 1985). For instance, the pair-rule gene fushi-tarazu (ftz) is expressed in odd-numbered parasegments that straddle segment boundaries (Hafen, et al., 1984). Ftz mutants have half the number of segments, which are enlarged through fusion (Wakimoto et al., 1984). The unique identities within a segment are then specified by staggered and overlapping pair-rule expression (Akam, 1987). Once they have specified the identities of alternating parasegments, a number of pair-rule genes switch to secondary segmental expression and activate segment polarity genes (Macdonald et al., 1986; Benedyk et al., 1994; DiNardo and O’Farrell, 1987).

How pair-rule patterning occurs during the progressive mode of segmentation seen in short and intermediate-germ insects remains an ongoing question. In these insects, segments appear sequentially in an anterior to posterior progression. Although expression patterns of orthologs of Drosophila pair-rule genes appear in a diverse assortment of patterns, they are frequently expressed with double-segment periodicity and produce secondary stripes in a stereotyped sequence (Patel et al., 1994; Davis et al., 2001; Choe et al., 2006). In this pattern, a broad stripe appears anterior to the posterior growth zone, and then resolves into two secondary stripes as expression is cleared from the center (Patel et al., 1994). Therefore, the resolution of pair-rule stripes into segmental stripes that occurs simultaneously along the A-P axis in the long germ Drosophila, probably occur step-wise in short- and intermediate-germ insects.

Functional studies of pair-rule gene orthologs in short and intermediate germ insects are limited to the beetle Tribolium, the cricket Gryllus, and the milkweed bug Oncopeltus. A comprehensive analysis of all 9 orthologs of the Drosophila pair-rule genes in Tribolium revealed that only two, paired (prd) and sloppy-paired (slp), produced a canonical pair-rule phenotype in RNAi experiments (Choe et al., 2006). Three others produced an asegmental phenotype, but these were required for prd and slp pair-rule expression. Loss of any of the four remaining Tribolium pair-rule orthologs did not produce segmentation phenotypes (Choe et al., 2006). In the cricket, loss of even-skipped (Gb’eve) produced a pair-rule phenotype in the gnathal segments and thorax, with fusions between adjacent segments. In the abdomen, Gb’eve depletion also caused fusions between adjacent segments, as well as growth and patterning defects (Mito et al., 2007). Interestingly, knock down of Oncopeltus eve (Of’eve) produced gap and segmental phenotypes, but no pair-rule phenotype (Liu and Kaufman, 2005).

The E75 locus was first identified as a primary response gene in studies of transcriptional activation in response to the steroid hormone, 20-hydroxyecdysone (20E), in in vitro cultured salivary glands (Ashburner, 1972, 1974). Subsequent molecular and genetic studies have revealed that the E75 gene produces four polypeptides, E75A-D that belong to the nuclear receptor superfamily (Feigl et al., 1989; Segraves and Hogness, 1990; Dubrovskaya et al 2004). Although E75 has long been considered an “orphan receptor” that binds DNA in an unliganded manner, it was recently shown that E75A contains heme, and that NO gas regulates E75 transcriptional activity (Reinking et al., 2005). Therefore, NO gas is probably the ligand for this orphan receptor. Here we show that Of’E75A is expressed with two-segment periodicity in the Oncopeltus blastoderm, and that it is expressed in a dynamic pattern in the extending germ band. We also show that Of’E75ARNAi produces a pair-rule phenotype in the gnathal segments and thorax, resulting from deletions of parasegments 3 and 5. In the abdomen, Of’E75A is required for specification of three odd-numbered parasegments, but it is also used in other growth and patterning processes in even-numbered parasegments. Our results show that Of’E75A has a pair-rule-like function in regions of the Oncopeltus embryo.


Animals and staging

The milkweed bugs were reared on sunflower seeds under long day (17L:7D) photoperiod at 26° C. Embryos from egg lays of 4 hours or less were incubated at 26° C, and the midpoint of the egg lay was taken as hour 0. For blastoderm-stage embryos, we used 30 hr old embryos. For segmented germ bands, we used 72 hr embryos. We found that developmental rates were very temperature sensitive and these times may not correspond with the developmental times of other studies of Oncopeltus.


The degenerate primers ZFII (5′-RCAYTTYTKIARICKRCA-3′) and ZF PBOX (5′-GARGGITGYAARGGITTYTT-3′) (Jindra et al., 1994) were used to amplify zinc finger regions of nuclear receptors from cDNA made from total RNA of 24-hr old embryos. We used 1.5 mM Mg, and 2 μM for each primer. The reaction was run for 40 cycles of 95° C for 30 sec, 50° C for 30 seconds, and 72° C for 15 seconds. Putative Of’E75 and Of’HR78 cDNAs were obtained.

The full Of’E75 transcripts were acquired by 5′ and 3′ RACE from 0-48 hr embryonic cDNA. Both 5′ and 3′ RACE PCRs were performed with gene-specific primers (GGAGGCTGCAGTACTGGCATCTGT and TGTAAGGGGTTCTTCAGGCGGT, respectively) targeting the DNA-binding domain (DBD). The initial 3′ RACE PCR produced transcripts extending into the ligand-binding domain (LBD). Two subsequent 3′ RACE PCRs using gene-specific primers CCCACCTTGGCTTGTCCGTTGAA within the LBD and GCGAGGGGGAGAATCTTGCTGGGGCCAA within the F-domain produced the complete 3′ end of the transcript.

in situ hybridization

Labeled probes were generated with the Ambion MAXIscript labeling kit (Austin, TX). We performed in situ hybridization according to the method of Liu and Kaufman (2004) except that we omitted the detergent step, and embryos of all stages were first boiled for 2-3 minutes Briefly, embryos were boiled in 4mL scintillation vials for 2-3 minutes, then chilled on ice for 5 minutes, and then gently shaken in a 1:1 heptane: 12% formaldehyde/1X phosphate-buffered saline for 20 minutes. After shaking, the aqueous phase was removed, and methanol was added to burst the chorions (Liu and Kaufman, 2004). One of our preparations involved cutting a blastoderm in half and staining either side with a different probe. This was performed after boiling, dechorionating, and rehydrating. We used a needle to perforate along the midline of a fixed blastoderm and carefully separated the lateral halves. cDNAs used to generate the Dfd, pb, Ubx, Kr, en and hb probes were a generous gift of P. Liu and T. Kaufman (Indiana U., Bloomington, IN).


Germ band stage embryos were mounted in Aqua Polymount (Warrington, PA) and photographed using IP Lab software (Rockville, MD). Images were processed with Adobe Photoshop software (San Jose, CA). In some cases, images were compiled from multiple Z-sections of the same field to bring all information into focus.

dsRNA interference

Double-stranded RNA (dsRNA) used for RNA-mediated interference (RNAi) was transcribed from plasmids using MEGAscript (Ambion, Austin, TX). The complementary strands were then annealed in a thermocycler according to Hughes and Kaufman (2000). Females were injected between mid-abdominal sternites with up to 10 μl of 2-5 μg/μl dsRNA in water.

Cuticle Analysis

Pharate first instar nymphs were removed from the chorion within 12 hours of hatching, and mounted in Hoyers’s according to Stern and Sucena (2000). Images were obtained using DIC optics on a Nikon Eclipse E1000 microscope (Mellville, NY) with BD Biosciences IPLab software (Rockville, MD).


Isolation of two E75 isoforms from Oncopeltus fasciatus

Nuclear receptors are characterized by a conserved DNA-binding domain (C) that contain two zinc finger motifs (Fig. 1; King-Jones and Thummel, 2005). We used degenerate primers to the P-Box region of zinc finger I and to zinc finger II of the DNA-binding domain (Fig. 1, Jindra et al., 1994) to amplify a 133 bp fragment from cDNA made from embryos 24 hr after egg lay (AEL). The common region of Of’E75 (domains D, E, F; Fig. 1) was then isolated by 3′ RACE as described in the Materials and Methods. A comparison with other known insect E75 common region sequences is shown in Table 1.

Figure 1
Diagram of E75 isoform structure showing the standard nuclear receptor domains (A-F; for review, see King-Jones and Thummel, 2005). The splice sites are based on E75 from Drosophila (Segraves and Hogness, 1990) and the location of the probes for the various ...
Table 1
Percent identity of deduced amino acid sequences of E75 from several different insects with the length of each domain given in parentheses. Comparison includes entire C domain (i.e., includes both zinc fingers). E75 sequences were all obtained from GenBank ...

Two Of’E75 transcripts were then isolated by 5′ RACE. One had a total length of 2651 bp encoding a polypeptide of 690 amino acids that contains both zinc fingers (Fig. 1). Three of the known E75 N-terminal isoforms (A/B domain in Fig. 1) in insects (E75A, C, and E) have two zinc fingers. These three isoform types formed distinct clusters in an unrooted, neighbor-joining tree. The longer Oncopeltus sequence clustered weakly with the E75A group (Fig. S1).

In addition to having overall similarity to E75A, the A-type isoform shared a conserved motif with E75A sequences from other insects. With the exception of the D isoforms, the isoform-specific regions of E75 are highly variable in length and sequence, between species and between isoforms. Each of these 3 isoforms has a different, highly conserved 6-10 amino acid sequence adjacent to the splice site with the exon containing the first zinc finger (Supplementary Fig. 2). The Oncopeltus Of’E75 transcript shows no similarity in this region with the E75Cs and E75Es, but shares 3 out of 6 residues with a majority of E75A isoforms, including a highly conserved leucine, and has one conservative substitution (Supplementary Fig. 2). Based on this motif and the tendency of this sequence to cluster with the A-isoforms from other species, we call this isoform Of’ E75A, while recognizing that it shows some divergence and might be a new isoform.

The other transcript is 2548 bp encoding a polypeptide of 636 amino acids and contains only the second zinc finger of the C domain (Fig. 1), characteristic of other known insect E75Bs (Supplementary Figs. 1 and 2). Complete transcripts for the two isoforms, Of’E75A and Of’E75B, are in GenBank (accession no. EF490808 and EF490809, respectively).

in situ expression of E75A during blastoderm and germ band stages

In Oncopeltus, the simultaneous and sequential aspects of intermediate-germ development coincide with the blastoderm and germ band stages, respectively. Once the segments of the prospective head and thorax are specified at the blastoderm stage, the germ band invaginates posteriorly into the yolk through an invagination pore (Fig. 2B,C). Upon invagination, the two separate lateral halves of head and thorax are joined within the yolk, and the abdominal primordium begins to produce segments in an anterior to posterior progression (Butt, 1949; Liu and Kaufman, 2004a).

Figure 2
In situ expression of Of’E75A in blastoderm-stage embryos. Each image is a lateral view with the anterior of the embryos to the left. Asterisks indicate Of’E75A stripes of expression. (A) Diffuse Of’E75A resolves into two transverse ...

Since Of’E75 mRNA is present on the first day of embryogenesis, we used in situ hybridization to examine its expression during this period. Although diffuse staining can be seen as early as blastoderm formation (~18hrs AEL), Of’E75A expression first resolves into two stripes 8-14 hours later (Fig. 2A). As the invagination pore appears at the posterior end of the embryo, a third stripe emerges de novo (Fig. 2B). A fourth stripe then appears at the onset of germ band invagination (~32-34 hours AEL, Fig. 2C). Expression of the anterior-most stripe was more diffuse and less intense compared to the posterior three stripes of the blastoderm. We believe that the stripes are appearing in an anterior- to posterior progression, but we cannot exclude the possibility that one or more stripes arise between two existing stripes. Identical in situ results were obtained with probes that recognize the E75 common regions and the E75A-specific probes, but no staining was seen with the E75B-specific probe, which is complementary to the 5′untranslated region of the transcript (probe locations are denoted in Fig. 1). Probes designed to complement untranslated regions are commonly used in in situ hybridization in order to avoid cross-hybridization between isoforms (Beach and Jeffery, 1992; Uttendaele et al, 1996; Vazdarjanova and Guzowski, 2004; Chang et al, 2007). However, a longer probe including some of the short E75B-specific coding sequence might show E75B expression

Liu and Kaufman (2005) showed that 6 stripes of Of’eve, which marks the posterior half of each segment, are present just prior to germ band invagination, and probably correspond to the mandibular, maxillary, labial, and three thoracic segments. Despite repeated attempts, we were unable to produce E75A/ eve double-stained embryos at the blastoderm stage. We therefore cut individual blastoderms in half, stained each half separately with either Of’eve or Of’E75A, and reunited the two halves of the same embryo (Fig. 2D). The three Of’E75A stripes of the Oncopeltus blastoderm roughly align with every other Of’eve stripe: the mandibular, labial and T2 segments (N=4). Although it is difficult to position the four stripes with confidence without double-stained embryos, the parallel staining with Of’eve does allow us to conclude that Of’E75A expression follows a double-segment periodicity during the blastoderm stage.

We co-stained developing germ band stage embryos with probes for Of’E75A and engrailed (Of’en), which marks the posterior 1/3 of each segment. Two stripes of Of’E75A are present during the earliest stages of germ band invagination (Fig. 3A, B). As the gnathal segments invaginate, most of the posterior proliferation zone expresses Of’E75A, with one stripe just anterior to it (data not shown). At the onset of abdominal segmentation, Of’E75A expression is cleared from the posterior end of the germ band (Fig. 3C, D). In summary, four Of’E75A stripes are present during the late blastoderm stage, but only two are present in the embryonic primordium after germ band invagination is completed.

Figure 3
In situ hybridization with probes for Of’E75A (blue) and Of’en (orange) during germ band invagination (A-C) and abdominal segmentation (D-F). The site of the invagination pore is indicated with an asterisk. The uninvaginated anterior tissues ...

Four stripes of Of’E75A are present and fade during progressive segmentation of the abdomen. Two stripes are present at the earliest stages of germ band invagination. The anterior-most stripe persists for three cycles of segmentation until it fades in the anterior half of abdominal segment 3 (A3, Fig. 3E, arrow). The broader posterior band divides into two thinner stripes after the first abdominal stripe of Of’en appears. The two narrower stripes persist for three cycles (Fig. 3E), then fade just posterior to the A4 and A5 Of’en stripes (data not shown). The width of this band and its subdivision into two smaller segmental bands resembles the pattern of other pair-rule orthologs during progressive segmentation (Patel et al., 1994; Davis et al., 2001; Choe et al., 2006; Mito et al., 2007). A final stripe of Of’E75A arises after formation of A6, and recedes without dividing after the 8th abdominal segment is specified. We do not detect Of’E75A expression after this point. The modal Of’E75A expression pattern for each cycle of segmentation is summarized in Fig. 3G.

Effect of E75ARNAi on embryonic development

To determine the function of Of’E75A, we used dsRNA to knock-down Of’E75A transcript levels. Parental E75 RNAi produced a range of defects that increased in intensity over a period of 2-3 weeks, then gradually subsided. In the mildest cases, the T2 and T3 segments and limbs were distorted or fused (Fig. 4B, G). In more severely affected embryos, the T2 and T3 limbs are lost altogether (Fig. 4C), and the labium is either fused with the first thoracic segment, or is completely lost (Fig. 4E). In these latter cases, the T1 limb is deformed (data not shown). E75ARNAi nymphs frequently had fewer abdominal segments and showed an overall shortening of the abdomen. In such cases, many of the remaining segments were wider than segments of control nymphs (Fig. 4C). In these embryos, a lateral spiracle was frequently found on the abdominal segment that is adjacent to an enlarged thoracic segment, although A1 does not normally have a lateral spiracle (data not shown). This suggests that A1 was either lost, or was fused with the thoracic segments. dsRNA made from the Of’E75 common regions and dsRNA made from Of’E75A (Fig. 1) had similar results except that those obtained with Of’E75A were not as severe. Of’E75BRNAi using two different constructs (one containing only the 5′UTR and one including the 5′UTR and part of the E75B-specific coding domain) (Fig. 1) had no effect on development. Of’E75 dsRNA significantly reduced the level of Of’E75 mRNA seen in the embryo at 24 and 30 hrs (Supplementary Figure S3A). Likewise, Of’E75A dsRNA reduced the level of Of’E75A mRNA at 30 hrs but did not suppress Of’E75B mRNA expression; the opposite was true for Of’E75B dsRNA (Supplementary Figure S3B). (Supplementary Figure S3B). Normal embryos were obtained after parental RNAi with the putative Of’HR78 sequence (data not shown).

Figure 4
Normal and E75A RNAi phenotypes of fully developed embryos and 1st instar nymphs. (A) Lateral view of normal nymph from an uninjected female just prior to hatching. T1-3 are the first through third thoracic segments. The abdominal segment numbers are ...

We next examined the effect of Of’E75ARNAi at the segmented germ band stage. (Supplementary Information: Figure S4). In the thorax, we frequently found loss of space or fusion between 1) the labial and first thoracic segment (T1, parasegment 3, ps3), 2) the T2 and T3 segments (ps5) and 3) the T3 and A1 segments (ps6). In extreme cases the labial, T2 and T3 appendages were lost entirely. In the abdomen, we found a smoothening or loss of the inter-segmental furrows between adjacent segments, with the most frequent defects at the A3/A4 (ps9), A5/A6 (ps11), and A7/A8 (ps13) boundaries. In addition, we observed defects in parasegments 12, 14 and 15 at a low frequency (Fig. 4H). In summary, Of’E75A knock-down primarily affected odd-numbered parasegments and ps6, with rare defects in ps12 and ps14.

Molecular analysis of the effects of E75ARNAi on thoracic segmentation

Our in situ analysis of blastoderm stage embryos indicated that Of’E75A stripes were roughly aligned with the mandibular, labial, and T2 segments. To determine the fate of thoracic segments after Of’E75A depletion, we stained with HOX gene markers. The gene Deformed (Of’Dfd) is expressed in the mandibular and maxillary segments (Hughes and Kaufman, 2000; Angelini et al., 2005). Dfd expression was not affected by Of’E75A depletion, even in strong Of’E75ARNAi knock-downs (data not shown). In contrast, proboscipedia (Of’pb), which is largely expressed in the labial segment (Fig. 5A; Hughes and Kaufman, 2000; Angelini et al., 2005), reveals that the labial appendage is fused posteriorly with T1 after Of’E75A-depletion (Fig. 5B). In more severely-affected embryos, Of’pb is found in an anterior patch of an enlarged thoracic limb (Fig. 5C). To determine the fate of the T3 segment, we stained germ band stage embryos with a probe for the homeotic gene Ultrabithorax (Of’Ubx), which is largely expressed in the anterior half of the first abdominal segment, with patches of expression in the posterior of T3 and in the T3 limb (Fig. 5D, Angelini et al., 2005). Of’Ubx expression was maintained in all Of’E75ARNAi embryos (N= 30), even in enlarged limbless segments that were fused with anterior thoracic segments (Fig. 5F). In summary, our analysis of HOX genes shows that 1) the mandibular and maxillary segments were not affected by Of’E75A depletion, 2) Of’E75A is required for specification of ps3, which straddles the labial and T1 segments, and 3) posterior T3 and anterior A1 segments contribute to an enlarged thoracic segment in the two-legged Of’E75ARNAi embryos.

Figure 5
HOX gene expression in control and Of’E75A -depleted germ bands. (A-C) (A) At the completion of segmentation, proboscipedia (Of’pb) is expressed in the labial appendage (La), with some faint expression in the mandibular segment. An asterisk ...

We next examined the expression of Ofen and Kruppel (Of’Kr) during late germ band invagination after E75ARNAi. At this stage of development, Of’Kr is expressed from the maxillary to the posterior of the third thoracic segment, and encompasses 5 stripes of en (Fig. 6A; Liu and Kaufman, 2004b). The number of Of’en stripes within the Kr domain was reduced from five to three in 13/14 E75A-depleted germ bands (Fig. 6B). This indicates that Of’E75A is required for the formation of the posterior portion of two of the 5 segments within the Kr domain.

Figure 6
Loss of alternating segments in Of’E75A-depleted embryos. Of’Kr expression (purple) and Of’en (orange) near the end of germ band invagination. Asterisks indicate the location of Of’en stripes.

Effects of E75ARNAi on abdominal segmentation

To determine if the reduction of segment number in the abdomen and loss of inter-segmental furrows also occurred through fusion in Of’E75A knock-downs, we made use of features on the first nymphal cuticle. The dorsal compartments of A2-A8 contain two dorso-ventral rows of bristles. Within each segment, the anterior-most row is interrupted by a spiracle (arrow, Fig. 7), while the posterior row is comprised of bristles only (arrowheads, Fig. 7A). A trio of large macrochaetes, just below the D/V midline, distinguishes A5 and A6 (asterisks, Fig.7A). A7 has a single large macrochaete in the posterior-ventral compartment (Fig. 7A, double arrowheads). After Of’E75A depletion, we frequently observed enlarged segments that contained 5-6 large macrochaetes in the ventral compartment, indicating a fusion between the A5 and A6 segments (Fig. 7B). The dorsal compartment of these fused segments contained only one spiracle-containing row in the anterior of the enlarged segment. Two rows of bristles appeared in the place of the A6 spiracle-containing row. The loss of one spiracle/bristle row in E75ARNAi cuticles suggests that fusion of A5 and A6 occurred as a region that spans from posterior A5 to the anterior A6 spiracle/bristle row was deleted. An identical arrangement was found in the dorsal compartment of enlarged segments that appeared anterior to the A5/A6 fusion, indicating fusion between A3 and A4 (Fig. 7C). We also detected loss of the ventral large macrochaete of A7 in a slightly enlarged segment posterior to A5/A6. In these situations, the segment posterior to A7 did not have a bristle row + spiracle, normally present on A8 (Fig. 7D). These data are consistent with fusion between A7 and A8. However, we also detected additional losses of anterior features of A7, suggesting that Of’E75A has an additional role in patterning this segment.

Figure 7
Fusion between adjacent segments in Of’E75ARNAi abdomens. (A) A control cuticle preparation of A4-A7 along the dorsal-ventral (D/V) boundary. Anterior is to the left. Vertical dashed lines indicate boundaries between segments. Arrows indicate ...

Regulation of Of’E75A by the gap gene, Of’eve

In Drosophila, proper pair-rule gene expression is determined through a combination of maternal, gap, and pair-rule gene expression (Lawrence, 1992). In contrast to its pair-rule function in Drosophila, Of’eve has gap and segmental functions in Oncopeltus and its expression is required for expression of the gap genes, Of’hb and Of’Kr (Liu and Kaufman, 2005). We therefore followed the expression of Of’E75A in Oncopeltus even-skipped (Of’eve)-depleted embryos. The three stripes of Of’E75A expression (Fig. 8A) were lost in 26/26 eve-depleted embryos from three separate clutches (Fig. 8B). In their place, low levels of unpatterned Of’E75A expression were found in the posterior 1/3-2/3 of the blastoderm. To determine whether Of’E75A expression might be regulated by Of’eve through Of’hb and Of’Kr, we used a mixture of Of’hb and Of’Kr dsRNA to knock down both transcripts. Although a majority of Of’hb/Of’KrRNAi blastoderms had poorly resolved Of’E75A stripes, depletion of Of’Kr and Of’hb together or individually did not affect the onset or placement of stripes (data not shown). Therefore, Of’eve is required for the expression of Of’E75A, but this requirement is not mediated by Of’Kr or Of’hb.

Figure 8
Of’E75A mRNA expression in Of’eve-depleted blastoderm stage embryos at the onset of germ band invagination. (A) In control embryos, just prior to formation of the invagination pore, Of’E75A mRNA is expressed in three broad bands ...


These studies have isolated two isoforms of the nuclear receptor E75, E75A and E75B, from the intermediate germ insect, Oncopeltus fasciatus. We show that Of’E75A is expressed with double-segment periodicity in the blastoderm. During progressive segmentation of the abdomen, Of’E75A is expressed in a dynamic pattern, which briefly resembles an intra-stripe repression sequence observed with pair-rule orthologs during progressive segmentation in short and intermediate-germ insects. Loss of Of’E75A produces an imperfect pair-rule-like phenotype, with the major effects occurring in ps 3, 5, 6, 9, 11 and 13, but defects were also observed in ps 12 and 14 as well. Finally, we show that Of’eve, which acts as a gap and segmentation gene in Oncopeltus, is required for striped expression of Of’E75A.

Regulation of Of’E75A in the blastoderm of Oncopeltus

Liu and Kaufman (2005) showed that Of’eve had both gap and segmentation gene activities. Initially, Of’eve expression encompasses the posterior two-thirds of the Oncopeltus blastoderm. It then resolves into segmental stripes just before the onset of germ band invagination. While it is possible that the segmental function of Of’eve regulates Of’E75A expression during the blastoderm stage, the sequence of expression makes this unlikely. Of’E75A stripes first emerge at 18 hours; the resolution of Of’eve stripes does not occur for another 14 hours (Fig. (Fig.2,2, ,8;8; Liu and Kaufman, 2005). Therefore, the early gap function of Of’eve is required for the appearance of Of’E75A stripes. Interestingly, Of’KrRNAi and Of’hbRNAi had no appreciable affect upon the appearance or placement of Of’E75A stripes. This shows that Of’eve acts in parallel pathways to regulate segmentation at both the gap and parasegmental levels.

The role of Of’E75A in parasegment specification

In general, we found that Of’E75A was required for specification of odd-numbered parasegments (Fig 4H). One exception to this, however, was ps6. In our analysis, we found that ps5, encompassing the T2-T3 boundary, was more sensitive to loss of Of’E75A than ps6, which straddles the T3-A1 boundary. In fact, although all 89 affected germ bands examined showed a loss or reduction of T2, we did not observe a single instance where the T3 appendage was lost or affected without a greater loss or reduction of T2 (Fig. 5E). Often the reduction of the T2 appendage was incomplete, and a vestige of T2 projected from a fairly complete T3 limb (Fig. 4E and data not shown). We suggest that this pattern reflects the temporal order of specification; Of’E75A is first used to specify a ps5/ps6 progenitor, and ps5 is subsequently specified from the progenitor parasegment (Fig. 9). Another deviation from the pair-rule pattern occurred at ps7, which encompasses the posterior half of A1 and the anterior half of A2, since we did not observe any deletions of this parasegment. Consistent with this, we did not observe any Of’E75A expression in A1 or A2 during segment specification (Fig. 3G). In contrast, in situ analysis of E75 expression in the short germ Tribolium castaneum (Tc’E75) during this stage reveals a stripe of Tc’E75 in the anterior half of the first abdominal segment, although all other aspects of expression are comparable to Of’E75 expression (M. Rynerson and LMR, unpublished data). We therefore suggest that the deviation from the odd-numbered parasegment pair-rule pattern that occurs at ps6/ps7 in Oncopeltus could be related to the transition from blastoderm to germ band modes of development.

Figure 9
The Of’E75A mutant series reflects the temporal order of parasegment specification. Head and thoracic segments are shown for control, moderate and severe germ band stage embryos. The gnathal and thoracic segments are labeled Mn = mandibular, Mx ...

In contrast to the pair-rule role of Of’E75A, we observed a second function for Of’E75A in the abdomen (Fig. 4H), where defects were observed within even numbered parasegments, (ps 12 and 14). This scenario, where odd and even numbered parasegments are both affected, is consistent with the activity of pair-rule genes in Drosophila. For instance, cloning and analysis of eve revealed that, in addition to the 7 pair-rule stripes that appear at the onset of gastrulation, 7 narrower “minor” stripes appear de novo during gastrulation (MacDonald et al., 1986). All 14 stripes were then required for expression of en in the posterior half of every segment, and loss of en resulted in a segment polarity phenotype (MacDonald et al., 1986). In the case of Of’E75ARNAi embryos, malformations in ps12 and ps14 could be due to a similar switch from pair-rule activity in odd-numbered parasegments to a segmental function within both even- and odd-numbered parasegments.

Comparisons with parasegment specification in other systems

Our results with Of’E75ARNAi in the milkweed bug are very comparable to the effects found with Gb’eve depletion in the cricket, which is also an intermediate-germ, hemimetabolous insect (Mito et al., 2007). Like Of’E75ARNAi in the milkweed bug, Gb’eveRNAi causes fusions between the labial and T1, and T2 and T3 segments in the anterior half of the cricket embryo. For both genes, stripes of expression were found in the mandibular and maxillary segments, although knock-down experiments did not result in defects in these segments. In addition, loss of Of’E75A in the milkweed bug and loss of Gb’eve in the cricket both show a gap-like phenotype in the thorax. In the abdomen, both Gb’eve and Of’E75A stripes appear in a variety of patterns. In the cricket abdomen, Gb’eve stripes A4/A5 and A8/A9 appear through a typical pair rule pattern, when a broad band subdivides into two segmental stripes (Mito et al., 2007). In the Oncopeltus abdomen, only the A5/A6 stripes of Of’E75A appear in a typical pair-rule pattern. By contrast, Gb’eve stripes in A2, A3, A6 and A7 of the cricket abdomen and the A3 stripe of Of’E75A in the milkweed bug all arise as segmental stripes from a domain of expression in the posterior growth zone. The final stripe seen in both expression patterns is a de novo segmental stripe; in the cricket this Gb’eve stripes is in A10, and a de novo segmental stripe of Of’E75A appears in A9 or A10 (the precise segment could not be determined, Fig. 3G). Depletion of both Of’E75A and Gb’eve cause fusions between adjacent abdominal segments, but these also occur with other effects of growth and patterning (Mito et al., 2007). Therefore, pair-rule patterning in the cricket and milkweed bug appears to occur piece-meal, through the action of gene products that are not expressed with perfect double-segment periodicity and are used to specify alternating parasegments as well as for non-pair-rule functions in different regions of the segmenting embryo.

Could E75A be a pair-rule gene in Drosophila that was simply overlooked in the initial embryonic screens? We believe this is unlikely. In Drosophila E75A mRNA is maternally inherited and evenly distributed (Berkeley Drosophila Genome Project). It is first transcribed between 6-8 hours after egg deposition (Sullivan and Thummel, 2003; Dubrovsky et al., 2004), while the pair-rule genes are expressed between 1.5 and 3 hours, from the onset of cellularization through gastrulation. E75 null embryos are missing the anterior midgut constriction and possess an ectopic posterior constriction; no pair-rule phenotype was reported for these animals (Bilder and Scott, 1995). Therefore, if E75A acted as a pair-rule gene in the last common ancestor of Oncopeltus and Drosophila, its function was lost somewhere along the line that led to the fruit fly.

Possible connections between the segmentation and molting gene networks

E75A expression in postembryonic development is initiated by ecdysteroid, and the appearance of the different isoforms depends on the changing ecdysteroid titer (Segraves and Hogness, 1990; Zhou et al., 1998; Bialecki et al., 2002; Sullivan and Thummel, 2003; Dubrovsky et al., 2004; Keshan et al., 2006). Maternally derived conjugated ecdysteroids are present in the freshly oviposited eggs of most insects, then are hydrolyzed and used to regulate production of both serosal cuticle in locusts (Lagueux et al., 1979) and the first embryonic cuticle in both locusts and cockroaches before the embryo’s prothoracic glands are formed (Hoffmann and Lagueux, 1985; Lanot et al., 1989; Sbrenna, 1991). In Oncopeltus, ecdysone conjugates are likewise maternally loaded; low amounts of ecdysone and to a lesser extent 20E are then found from blastoderm formation through germ band elongation (Dorn, 1983; 2001). This period coincides with the time that we find E75A mRNA.

Whether ecdysteroids might be responsible for initiating E75 expression in Oncopeltus embryos is unclear. In the cockroach, Blattella germanica, E75A mRNA was found in the early embryo, but was not correlated with an increase in ecdysteroid titer (Mané-Padrós et al., 2008). The only studies to suggest a possible role for ecdysteroids in early germ band development have been in Manduca sexta embryos. In this system, the appearance of 26-hydroxyecdysone from a maternally provided conjugate has been temporally correlated with gastrulation and segmentation (Lanot et al., 1989). When isolated germ bands were incubated in vitro, their elongation was retarded. The addition of various ecdysteroids (e.g., makisterone A, 20-hydroxyecdysone and ecdysone) restored the longitudinal growth, suggesting that they are required for proper germ band formation.

Several years ago, Patel (1994) noticed that a large number of Drosophila segmentation genes are also used during neurogenesis of Drosophila and other insects. For instance, the pair-rule genes eve (Doe et al., 1988a), ftz (Doe et al., 1988b), runt (Dormand and Brand, 1998), sloppy paired (Bhat et al., 2000) and tenascin major (Levine et al., 1994) all play a role in patterning the embryonic CNS of Drosophila, where they interact with gap, segment polarity and HOX genes (Doe and Scott, 1988). In addition, the early neurogenic expression patterns of genes are often conserved between insect species (Broadus and Doe, 1995; Liu and Kaufman, 2004a, 2004b). Patel suggested that nervous system patterning provides a large source of transcription factors that have established interactions. When one transcription factor is co-opted for a new function, he reasoned, the entire network would be co-opted with it to regulate the new function.

A similar phenomenon, where the ecdysone response cascade was co-opted from the molting cycle for segmentation purposes, may have occurred within more basal insects or hexapods, but was then subsequently lost along the evolution of Diptera. In addition to Of’E75, we have detected other ecdysone-regulated genes, HR78 and broad (br), in early embryonic development of the milkweed bug (DFE, HCK, LMR and J.Truman, unpublished data). Using pRNAi, we found that Of’br is also required for posterior growth and segmentation of the embryo (DFE, LMR and J.Truman, in preparation) but we could not detect a role for Of’HR78 (data not shown). In the cockroach, a more basal insect, Maestro et al (2005) report the presence of Ultraspiracle, the heterodimeric binding partner of the ecdysone receptor, during segmentation. Together with our data on the role of E75A in the embryonic milkweed bug, these disparate results support a greater role for this gene network during segmentation of hemimetabolous insects. Moreover, postembryonic addition of segments (anamorphosis) is a feature during development of many extant and fossil arthropods. These additions occur at molts, indicating that the molting hormones regulate segment addition. Perhaps the segmentation and molting gene networks may have once been more intertwined in an ancestor. In this scenario, the presence of a number of ecdysone response genes during segmentation of a more basal insect would be a vestige of an ancestral, anamorphic developmental mode.

Supplementary Material






We thank Drs. J. W. Truman, B. Wakimoto, and G. K. Davis for insightful comments on this manuscript. Dr. Takashi Koyama provided some assistance with molecular biology. Dr. Sergi Castellano provided expertise and assistance with the phylogenetic analysis. We thank Dr. Thom Kaufman for over 7 different cDNAs that were used in this study. This research was supported by NIH GM60122 to LMR and NIH GM063622 to David Stern.


Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.


  • Akam M. The molecular basis for metameric pattern in the Drosophila embryo. Development. 1987;101:1–22. [PubMed]
  • Angelini DR, Liu PZ, Hughes CL, Kaufman TC. Hox gene function and interaction in the milkweed bug Oncopeltus fasciatus (Hemiptera) Dev Biol. 2005;287:440–455. [PubMed]
  • Ashburner M. Patterns of puffing activity in the salivary gland chromosomes of Drosophila. VI. Induction by ecdysone in salivary glands of D. melanogaster cultured in vitro. Chromosoma. 1972;38:255–281. [PubMed]
  • Ashburner M. Sequential gene activation by ecdysone in polytene chromosomes of Drosophila melanogaster. II. Effects of inhibitors of protein synthesis. Dev. Biol. 1974;39:141–157. [PubMed]
  • Beach RL, Jeffery WR. Multiple actin genes encoding the same alpha- muscle isoform are expressed during ascidian development. Dev. Biol. 1992;151:55–66. [PubMed]
  • Benedyk MJ, Mullen JR, DiNardo S. odd-paired: a zinc finger pair-rule protein required for the timely activation of engrailed and wingless in Drosophila embryos. Genes Dev. 1994;8:105–117. [PubMed]
  • Bhat KM, vanBeers EH, Bhat P. Sloppy paired acts as the downstream target of Wingless in the Drosophila CNS and interaction between sloppy paired and gooseberry inhibits sloppy paired during neurogenesis. Development. 2000;127:655–665. [PubMed]
  • Bialecki M, Shilton A, Fichtenberg C, Segraves WA, Thummel CS. Loss of the ecdysteroid-inducible E75A orphan nuclear receptor uncouples molting from metamorphosis in Drosophila. Dev. Cell. 2002;3:209–220. [PubMed]
  • Bilder D, Scott MP. Genomic regions required for morphogenesis of the Drosophila embryonic midgut. Genetics. 1995;141:1087–1100. [PubMed]
  • Broadus J, Doe CQ. Evolution of neuroblast identity: seven-up and prospero expression reveal homologous and divergent neuroblast fates in Drosophila and Schistocerca. Development. 1995;121:3989–3996. [PubMed]
  • Butt FH. Embryology of the milkweed bug, Oncopeltus fasciatus (Hemiptera) Cornell Univ. Agr. Expt. Sta. Mem. 1949;283:1–43.
  • Chang C, Lin G, Cook CE, Horng S, Lee H, Huang T. Apvasa marks germ-cell migration in the parthenogenetic pea aphid Acyrthosiphon pisum (Hemiptera: Aphidoidea) Dev. Genes & Evol. 2007;217:275–287. [PubMed]
  • Choe PC, Miller SC, Brown SJ. A pair-rule gene circuit defines segments sequentially in the short germ insect Tribolium castaneum. Proc. Nat. Acad. Sci. USA. 2006;103:6560–6564. [PubMed]
  • Cruz J, Mane-Padros D, Belles X, Martin D. Functions of the ecdysone receptor isoform-A in the hemimetabolous insect Blattella germanica revealed by systemic RNAi in vivo. Dev. Biol. 2006;297:158–171. [PubMed]
  • Davis GK, Jaramillo CA, Patel NH. Pax group III genes and the evolution of insect pair-rule patterning. Development. 2001;128:3445–3458. [PubMed]
  • DiNardo S, O’Farrell PH. Establishment and refinement of segmental pattern in the Drosophila embryo: spatial control of engrailed expression by pair-rule genes. Genes Dev. 1987;1:1212–1225. [PubMed]
  • Doe CQ, Scott M. Segmentation and homeotic gene function in the developing nervous system in Drosophila. Trends Neurosci. 1988;11:101–106. [PubMed]
  • Doe CQ, Smouse D, Goodman CS. Control of neuronal fate by the Drosophila segmentation gene even-skipped. Nature. 1988;333:376–378. [PubMed]
  • Dormand EL, Brand AH. Runt determines cell fates in the Drosophila embryonic CNS. Development. 1998;125:1659–1667. [PubMed]
  • Dorn A. Hormones during embryogenesis in the milkweed bug, Oncopeltus fasciatus. Entomol. Gen. 1983;8:193–214.
  • Dorn A. Arthropoda-Insecta: embryology. In: Adiyodi KG, Adiyodi RG, Dorn A, editors. Reproductive Biology of Invertebrates: Progress in Developmental Endocrinology. 10B. John Wiley and Sons, Inc.; New York: 2001. pp. 71–116.
  • Dubrovskaya VA, Berger EM, Dubrovsky EB. Juvenile hormone regulation of the E75 nuclear receptor is conserved in Diptera and Lepidoptera. Gene. 2004;340:171–177. [PubMed]
  • Dubrovsky EB, Dubrovskaya VA, Berger EM. Hormonal regulation and functional role of Drosophila E75A orphan nuclear receptor in the juvenile hormone signaling pathway. Dev. Biol. 2004;268:258–270. [PubMed]
  • Feigl G, Gram M, Pongs O. A member of the steroid hormone receptor gene family is expressed in the 20-OH-ecdysone inducible puff 75B in Drosophila melanogaster. Nucl. A. Res. 1989;17:7167–7178. [PMC free article] [PubMed]
  • Hafen E, Kuriowa A, Gehring WJ. Spatial distribution of transcripts from the segmentation gene fushi tarazu during Drosophila embryonic development. Cell. 1984;37:833–841. [PubMed]
  • Hoffmann JA, Lagueux M. Endocrine aspects of embryonic development in insects. In: Kerkut GA, Gilbert LI, editors. Comprehensive Insect Physiology, Biochemistry and Pharmacology. Vol. 1. Pergamon Press; Oxford: 1985. pp. 435–460.
  • Hughes C, Kaufman TC. RNAi analysis of Deformed, proboscipedia and Sex combs reduced in the milkweed bug Oncopeltus fasciatus: novel roles for Hox genes in the hemipteran head. Development. 2000;127:3683–3694. [PubMed]
  • Jindra M, Sehnal F, Riddiford LM. Isolation and developmental expression of the ecdysteroid-induced GHR3 gene of the wax moth, Galleria mellonella. Insect Biochem. Mol. Biol. 1994;24:763–773. [PubMed]
  • Keshan B, Hiruma K, Riddiford LM. Developmental expression and hormonal regulation of different isoforms of the transcription factor, E75, in the tobacco hornworm, Manduca sexta. Dev. Biol. 2006;295:623–632. [PubMed]
  • King-Jones K, Thummel CS. Nuclear receptors-a perspective from Drosophila. Nature Rev. Genet. 2005;6:311–323. [PubMed]
  • Lagueux M, Hetru C, Goltzene F, Kappler C, Hoffmann JA. Ecdysone titer and metabolism in relation to cuticulogenesis in embryos of Locusta migratoria. J. Insect Physiol. 1979;25:709–723.
  • Lanot R, Dorn A, Gunster B, Thiebold J, Lagueux M, Hoffmann JA. Functions of ecdysteroids in oocyte maturation and embryonic development of insects. In: Koolman J, editor. Ecdysone. From Chemistry to Mode of Action. Georg Thieme Verlag; Stuttgart: 1989. pp. 262–269.
  • Lawrence P. The Making of a Fly. Blackwell Scientific Publications; Oxford: 1992.
  • Levine A, Bashem-Ahrend A, Budai-Hadrian O, Gartenberg D, Menasherow S, Wides R. Odd Oz: a novel Drosophila pair-rule gene. Cell. 1994;77:587–598. [PubMed]
  • Liu PZ, Kaufman TC. Hunchback is required for suppression of abdominal identity, and for proper germband growth and segmentation in the intermediate germband insect Oncopeltus fasciatus. Development. 2004a;131:1515–1527. [PubMed]
  • Liu PZ, Kaufman TC. Kruppel is a gap gene in the intermediate germband insect Oncopeltus fasciatus and is required for development of both blastoderm and germband-derived segments. Development. 2004b;131:4567–4579. [PubMed]
  • Liu PZ, Kaufman TC. even-skipped is not a pair-rule gene but has segmental and gap-like functions in Oncopeltus fasciatus, an intermediate germband insect. Development. 2005;132:2081–2092. [PubMed]
  • Macdonald PM, Ingham P, Struhl G. Isolation, structure and expression of even-skipped: a second pair-rule gene of Drosophila containing a homeobox. Cell. 1986;47:721–734. [PubMed]
  • Maestro D, Cruz J, Pascual N, Martin D, Belles X. Differential expression of two RXR/ultraspiracle isoforms during the life cycle of the hemimetabolous insect Blattella germanica (Dictyoptera, Blattellidae) Mol. Cell. Endocrinol. 2005;238:27–37. [PubMed]
  • Mané-Padrós, Cruz J, Vilaplana L, Pascual N, Bellés X, Martin D. The nuclear hormone receptor BgE75 links molting and developmental progression in the direct-developing insect Blattella germanica. Dev. Biol. 2008;315:147–160. [PubMed]
  • Martinez-Arias A, Lawrence PA. Parasegments and compartments in the Drosophila embryo. Nature. 1985;313:639–642. [PubMed]
  • Mito T, Kobayashi C, Sarashina I, Zhang H, Shinahara W, Miyawaki K, Shinmyo Y, Ohuchi H, Noji S. even-skipped has gap-like, pair-rule-like, and segmental functions in the cricket Gryllus bimaculatus, a basal, intermediate germ insect (Orthoptera) Dev. Biol. 2007;303:202–213. [PubMed]
  • Nusslein-Volhard C, Weischaus E. Mutations affecting segment number and polarity in Drosophila. Nature. 1980;287:795–801. [PubMed]
  • Patel NH. Developmental evolution: Insights from studies of insect segmentation. Science. 1994;266:581–590. [PubMed]
  • Patel NH, Condron BG, Zinn K. Pair-rule expression patterns of even-skipped are found in both short- and long-germ beetles. Nature. 1994;367:429–434. [PubMed]
  • Peel AD, Chipman AD, Akam M. Arthropod segmentation: beyond the Drosophila paradigm. Nat. Rev. Genet. 2005;6:905–916. [PubMed]
  • Reinking J, Lam MMS, Pardee K, Sampson HM, Liu S, Yang P, Williams S, White W, Lajoie G, Edwards A, Krause HM. The Drosophila nuclear receptor E75 contains a heme and is gas responsive. Cell. 2005;122:195–207. [PubMed]
  • Sbrenna G. Roles of morphogenetic hormones in embryonic cuticle deposition in arthropods. In: Gupta AP, editor. Morphogenetic Hormones of Arthropods. Roles in Histogenesis, Organogenesis and Morphogenesis. Vol. 3. Rutgers University Press; New Brunswick: 1991. pp. 44–80.
  • Schupbach T, Wieschaus E. Germline autonomy of maternal-effect mutations altering the embryonic body pattern of Drosophila. Dev. Biol. 1986;113:443–448. [PubMed]
  • Segraves WA, Hogness DS. The E75 ecdysone-inducible gene responsible for the 75B early puff in Drosophila encodes two new members of the steroid receptor superfamily. Genes Dev. 1990;4:204–219. [PubMed]
  • Stern DL, Sucena E. Preparation of Larval and Adult Cuticles for Light Microscopy. In: Sullivan W, Ashburner M, Hawley RS, editors. Drosophila Protocols. Cold Spring Harbor Laboratory Press; Cold Spring Harbor: 2000. pp. 601–615.
  • Sullivan AA, Thummel CS. Temporal profiles of nuclear receptor gene expression reveal coordinate transcriptional responses during Drosophila development. Mol. Endocrinol. 2003;17:2125–2137. [PubMed]
  • Uttendaele H, Marazzi G, Wu G, Yan Q, Sassoon D, Kitajewski J. Notch4/int-3, a mammary proto-oncogene, is an endothelial cell-specific mammalian Notch gene. Development. 1996;122:2251–2259. [PubMed]
  • Vazdar A, Guzowski JF. Differences in hippocampal neuronal population responses to modifications of an environmental context: evidence for distinct, yet complementary, functions of CA3 and CA1 ensembles. J. Neurosci. 2004;29:6489–6496. [PubMed]
  • Wakimoto BT, Turner FR, Kaufman TC. Defects in embryogenesis in mutants associated with the Antennapedia gene complex of Drosophila melanogaster. Dev. Biol. 1984;102:147–172. [PubMed]
  • Zhou B, Hiruma H, Shinoda T, Riddiford LM. Juvenile hormone prevents ecdysteroid-induced expression of Broad Complex RNAs in the epidermis of the tobacco hornworm, Manduca sexta. Dev. Biol. 1998;203:233–244. [PubMed]