*Author to whom all correspondence should be addressed. Email: email@example.com
In many animal groups, left–right (LR) asymmetry within the body is observed. The left and right sides of the body are generally defined with reference to the anterior–posterior (AP) and dorsal–ventral (DV) axes. In this study, we investigated whether LR asymmetry is solely dependent on the AP and DV polarities in Drosophila embryos. We focused on the proventriculus, a posterior part of the foregut, and the hindgut because LR asymmetry in these body parts is highly stable in normal embryos. In embryos with a fully reversed AP polarity, LR asymmetry in both the proventriculus and the hindgut was re-oriented in relation to the reversed AP polarity. This demonstrates that inversion of AP polarity does not affect LR asymmetry of these tissues, and implies that LR asymmetry is specified in relation to the AP and DV polarities. Our findings were not consistent with the alternative hypothesis that LR asymmetry is predetermined by maternal signals that localize asymmetrically along the LR axis in the oocyte and/or early embryo.
The left–right (LR) axis can be defined with reference to the other two embryonic axes, the anterior–posterior (AP) and dorsal–ventral (DV) axes. Genetic and molecular studies, mainly in vertebrates, have so far identified many genes that are expressed asymmetrically along the LR axis and/or are required for LR asymmetry (reviewed by Burdine & Schier 2000; Levin 2005). Two general models have been proposed to explain the initial assignment of such ‘handedness’ (reviewed by Brown & Wolpert 1990; Burdine & Schier 2000; Levin 2005). One is that the LR axis is determined within the embryo with respect to the AP and the DV axes. This model predicts that the reversal of the AP axis would not affect LR asymmetry, and the LR axis would simply be re-oriented with respect to the new AP orientation. The other model is that LR asymmetry is predetermined during oogenesis, and that maternally supplied positional information specifies LR asymmetry in the embryo. According to this model, reversal of the AP axis would result in an inversion of the LR axis with respect to the new AP orientation.
The stereotyped LR asymmetric pattern of the embryonic gut in Drosophila is known to be affected by several mutations. Among them, mutations in the maternal dicephalic (dic) and wunen (wun) genes cause a reversal of the LR asymmetry of the proventriculus, as well as of the AP polarity in the oocyte (Ligoxygakis et al. 2001), suggesting that maternal genes specify some aspects of the LR asymmetry of the foregut. In contrast, reversed AP polarity in the duplicated hindgut of bicoid (bcd) mutant embryos, with respect to the AP axis in the oocytes, retains normal LR asymmetry (Hayashi & Murakami 2001). In this study, we report LR asymmetry of the proventriculus and the hindgut in embryos with a fully reversed AP axis.
Materials and methods
st osk150 e/TM3 Sb Ser females were mated with P[ry+ osk-bcd3′UTR]st osk54 ry ss/ry[osk-bcd3′UTR (III)] males to obtain P[ry+ osk-bcd3′UTR]st osk54 ry ss/st osk150 e females. PY282, an enhancer trap line which expresses β-galactosidase (β-gal) in the embryonic hindgut (Murakami et al. 1994; Hayashi & Murakami 2001), was used to observe LR asymmetry of the hindgut. P[ry+ osk-bcd3′UTR]st osk54 ry ss/st osk150 e females mated with PY282 males produced OBO embryos, which were used as recipients for bcd mRNA injection to obtain OBOb embryos.
The open reading frame (ORF) in bcd cDNA was amplified by PCR reaction with a primer pair: T7-bcd5′ (5′-AAATTAATACGACTCACTATAGGGAAGCTACTTGTT CTTTTTGCAGGATCCATGGCGCAACCGCCGCC-3′) and pA-bcd3′ (5′-TTTTTTTTTTTTTTTTTTTTTTTTTTTTTT-CTCTAACACGCCTCTCGTCC-3′). The nucleotide sequences complementary with 5′-terminal regions of the ORF and 3′UTR are underlined in T7-bcd5′ and pA-bcd3′ respectively. Since the T7-bcd5′ primer contains the promoter sequence for T7-RNA polymerase, bcd mRNA was transcribed in vitro from the amplified DNA by using a mMESSAGE mMACHINE Kit (Ambion, Austin, TX, USA). 0.1 nL of bcd mRNA dissolved in distilled water (1 µg/µL) was injected into the posterior pole region of early embryos at 15 ± 15 min after egg laying (25°C) according to the protocol previously described by Okada et al. (1974). The injected embryos were allowed to develop in silicon oil (FL-100 450 CS, Shin-Etsu Silicon) at 25°C until stage 15–16. The embryos were then fixed in heptane/2.5% glutaraldehyde in PBS (130 mm NaCl, 7 mm Na2HPO4, 3 mm NaH2PO4) for 20 min with shaking and stained for β-gal as previously described by Sano et al. (2002). Fixed and stained embryos were mounted in 50% glycerol without removal of vitelline membranes.
Results and discussion
We wished to examine whether LR asymmetry is specified in relation to the AP and DV polarities in the Drosophila embryo, and to test this we set about reversing the AP polarity in embryos. Genetic screens have so far identified several maternal genes that are required for the establishment of AP patterning in Drosophila (Nüsslein-Volhard et al. 1987; St Johnston & Nüsslein-Volhard 1992) and among these, bcd and oskar (osk) play central roles. Bcd mRNA becomes localized to the anterior end of the oocyte during oogenesis and is the source of the morphogen gradient of Bcd protein in the embryo. In the absence of bcd activity, the embryo fails to form head and thoracic structures and develops an ectopic hindgut in place of the foregut (Frohnhöfer & Nüsslein-Volhard 1986; Hayashi & Murakami 2001). When injected into the posterior region of early embryos, bcd mRNA can induce head, thoracic and foregut structures posteriorly, resulting in the formation of dicephalic embryos (Driever et al. 1990).
In contrast, osk mRNA and its protein product are normally localized to the posterior pole of the oocyte to direct nanos (nos) RNA localization (Ephrussi et al. 1991). In the embryo, nos mRNA is translated to form a gradient of Nos protein with the highest concentration in the posterior, which in turn functions in the development of the posterior (abdominal) structures (Lehmann & Nüsslein-Volhard 1991; Wang & Lehmann 1991; Wang et al. 1994). Because Nos represses bcd activity, mislocalization of osk mRNA to the anterior pole results in the formation of posterior structures instead of the head and thorax (Ephrussi & Lehmann 1992). They substituted the 3′UTR of the osk gene with bcd 3′UTR that includes a signal for anterior localization, and have shown that embryos from osk-bcd 3′UTR females exhibit a bicaudal phenotype. The osk-bcd 3′UTR females that lack osk activity produce embryos (OBO embryos) with a single abdomen of reversed polarity in their anterior region. Hence, to obtain embryos with a perfect reversal of the AP polarity, we injected bcd mRNA into the posterior pole of OBO embryos. Approximately 37% of these bcd-injected OBO embryos (called OBOb embryos; N = 47) exhibited a fully reversed AP polarity and hatched normally. DV polarity was unaffected (data not shown).
After obtaining individual embryos with reversed AP axes, we first focused on LR asymmetry of the proventriculus, a distinctive organ located at the posterior end of the foregut (Skaer 1993; Campos-Ortega & Hartenstein 1997). It has been reported that the posterior of the proventriculus is oriented to the right side of the embryo after stage 15, and that this asymmetry is extremely stable in normal embryos (Ligoxygakis et al. 2001). We found that the LR polarity of the proventriculus was re-oriented in relation to the reversed AP polarity in OBOb embryos (Fig. 1). In 97.3% of the AP-reversed OBOb embryos (N = 37) that developed normally to stage 15/16, the posterior of the proventriculus was oriented to the right. This was not significantly different to the percentage of wild-type embryos with normally oriented proventriculi (100%, N = 50; P = 0.24; χ2-test). These observations clearly show that the inversion of the AP polarity does not alter the LR polarity of the proventriculus.
The hindgut is also known to exhibit LR asymmetry in Drosophila embryos (Hayashi & Murakami 2001). At late stage 13, the ventrally bending hindgut tube twists 90° to the left, so that the anterior end of the hindgut points to the right side of the embryo (Hayashi & Murakami 2001). OBOb embryos formed a single hindgut in their anterior portion and in 100% of embryos (N = 10), LR polarity of the hindgut was re-oriented in relation to the new AP axis (Fig. 1). Similarly, 83.9% of OBO embryos (N = 31) formed the hindgut with normal LR polarity in their anterior. In the remaining embryos, the anterior hindgut showed either reversed (3.2%) or intermediate (12.9%) LR polarity. The above observations indicate that a fully reversed AP polarity does not affect LR polarity of the proventriculus or the hindgut in the Drosophila embryo. This implies that LR polarity is specified in relation to the AP and DV polarities in the embryo, and it is therefore unlikely that LR polarity is predetermined during oogenesis.
Contrary to our observations, Ligoxygakis et al. (2001) have demonstrated that LR asymmetry of the proventriculus can be reversed by mutations in dic and wun, which also cause reversal of AP polarity in the embryo relative to its mother. Indeed, in the ovaries of these mutant females, a fraction of egg chambers shows a reversal of AP polarity (González-Reyes & St Jonston 1998; Ligoxygakis et al. 2001). However, it remains elusive whether LR reversal of the proventriculus is directly correlated with the reversal of AP polarity, because the embryos derived from the ‘reversed’ and ‘normal’ egg chambers are morphologically indistinguishable. Thus, it is possible that dic and wun have a direct effect on LR polarity of the proventriculus, independent of their effect on AP polarity. We demonstrate that, in embryos with a fully reversed AP polarity, LR asymmetry in both the proventriculus and the hindgut is re-oriented in relation to the reversed AP polarity.
There is a further question of whether the reversal of AP polarity in the embryo affects LR asymmetry in other tissues that show LR asymmetry, such as the testes in adult males which turn in a left-handed orientation (Miller 1950). Furthermore, the male spermiduct and genitalia also exhibit LR asymmetries (Adam et al. 2003) and it would be interesting to know whether LR asymmetry of these tissues is altered in adult flies derived from OBOb embryos. Since OBOb embryos are able to develop to adulthood (data not shown), they will provide a simple system with which to investigate the link between LR asymmetry in the embryonic and adult body forms.
We thank Dr R. Murakami and Dr K. Matsuno for enhancer trap lines and many helpful suggestions for this study. This work was supported by the Super Science High school (SSH) program (the Japan Society for the Promotion of Science).