The Drosophila Innexin7 Gap Junction Protein Is Required for Development of the Embryonic Nervous System

Abstract

The Drosophila genome encodes eight members of the innexin family of gap junction proteins. Most of the family members are expressed in complex and overlapping expression patterns during Drosophila development. Functional studies and mutant analysis have been performed for only few of the innexin genes. The authors generated an antibody against Innexin7 and studied its expression and functional role in embryonic development by using transgenic RNA interference (RNAi) lines. The authors found Innexin7 protein expression in all embryonic epithelia from early to late stages of development, including in the developing epidermis and the gastrointestinal tract. In early embryonic stages, the authors observed a nuclear localization of Innexin7, whereas Innexin7 was found in a punctuate pattern in the cytoplasm and at the membrane of most epithelial tissues at later stages of development. During central nervous system (CNS) development, Innexin7 was expressed in cells of the neuroectoderm and the mesectoderm and at later stages of embryogenesis, its expression was largely restricted to a segmental pattern of few glia and neuronal cells derived from the midline precursors. Coimmunostaining experiments showed that Innexin7 is expressed in midline glia, and in two different neuronal cells, the pCC and MP2 neurons, which are pioneer cells for axon guidance. RNAi-mediated knock down was used to gain insight into the embryonic function of innexin7. Down-regulation of innexin7 expression resulted in a severe disruption of embryonic nervous system development. Longitudinal, posterior, and anterior commissures were disrupted and the outgrowth of axon fibers of the ventral nerve cord was aberrant, causing peripheral nervous system defects. The results suggest an essential role for innexin7 for axon guidance and embryonic nervous system development in Drosophila.

Keywords:

• Drosophila
• gap junctions
• innexins
• nervous system
• RNAi

INTRODUCTION

Gap junctions consist of arrays of intercellular channels that permit the passage of ions and small signaling molecules between opposing cells, thereby allowing coordinated signalling of cells and tissue homeostasis (Goodenough et al. 1996; Phelan 2005; Bauer et al. 2005; Wei et al. 2004). Three gene families have evolved to construct gap junctions, the connexins and pannexins in deuterostomes and the innexins in protostomes including the fruit fly Drosophila (Panchin et al. 2000; Bauer et al. 2005; Phelan 2005; Barbe et al. 2006; Locovei et al. 2006). Members of all three gene families encode structurally very similar four-pass transmembrane proteins. Connexin proteins were shown to oligomerize into hexameric hemichannel subunits. For the formation of a functional gap junction channel, two hemichannels, one contributed by each of the opposing cells, dock head-to-head in the extracellular space to form a double membrane-spanning intercellular channel, thereby allowing the rapid exchange of ions and metabolites such as inositol phosphates and cyclic nucleotides (Segretain and Falk 2004; Martin and Evans 2004, for reviews). Whereas extensive studies in the last two decades have elucidated the role of connexin gap junction proteins in mammals (Evans and Martin 2002, for review), much less is known of the function and biochemistry of innexin gap junction proteins in protostomes (Bauer et al. 2005, for review).

The Drosophila innexin family of gap junction proteins consists of eight family members (Phelan 2005; Bauer et al. 2005). Drosophila gap junctions are established at very early stages of embryonic development, correlating with a maternal and zygotic expression of innexins (Bauer et al. 2004; Lehmann et al. 2006). Innexin1 (ogre) and innexin8 (shakingB) are functionally required in the giant fibre system and the visual system of adult flies (Watanabe and Kankel 1992; Curtin et al. 2002; Krishnan et al. 1993; Phelan et al. 1996; 1998; Shimohigashi and Meinertzhagen 1998; Zhang et al. 1999; Jacobs et al. 2000). innexin4 (zero population growth) was shown to control germ cell differentiation (Tazuke et al. 2002; Gilboa et al. 2003). Innexin2 is required for cell polarity and epithelial tissue organization in the Drosophila embryo (Bauer et al. 2002; 2004; Lehmann et al. 2006). Recently, an essential role of innexin2 could also be shown for the development of the proventriculus organ in the posterior foregut region of the Drosophila embryo (Lechner et al. 2007). It was demonstrated that during proventriculus development, innexin2 is a target gene of Hedgehog and Wingless signaling, thus connecting paracrine morphogen signaling with the regulation of gap junction communication (Pankratz and Hoch 1995; Bauer et al. 2002; Lechner et al. 2007). Furthermore, in a feedback loop, Innexin2-mediated gap junction communication was shown to be essential for the transcriptional activation of hedgehog, wingless, and Delta during foregut morphogenesis, thereby providing first evidence that the mutual transcriptional cross regulation of paracrine and gap junction-mediated signaling is essential for organogenesis in Drosophila (Lechner et al. 2007). These data pointed towards an essential requirement of gap junction communication for the transcriptional activation of morphogen-encoding genes activating evolutionary conserved signalling cascades essential for patterning in animals. For other innexin family members, including innexins 3, 5, 6 and 7, mutants are not available yet. However, the availability of the sequences of these genes allows RNAi knock down experiments to address their role during development. Using a transgenic line carrying an UAS/Gal4 RNAi construct for innexin3, we could recently show that innexin3 knock down causes innexin2-like mutant phenotypes and that heteromerization of Innexins 2 and 3 is crucial for epithelial organization and polarity of the embryonic epidermis (Lehmann et al. 2006). In the present study, we have analyzed the role of Innexin7 during embryonic development by using an anti-Innexin7 antibody and by RNAi-mediated knock down of the gene. We propose that Innexin7 plays an essential role for the development of the embryonic nervous system in Drosophila.

METHODS

Fly Strains

We used standard techniques for fly manipulation and stock keeping. As wild-type strain, the Oregon R strain was used. The UAS innexin7myc (UAS inx7myc) transgenic flies were generated by injection of the pUAST innexin7myc construct into w⁻1118 fly embryos and selected by eye color (Rubin and Spradling 1982; Brand and Perrimon 1993). UAS wizinx7 transgenic fly lines were generated by injecting pWizinx7 construct into w⁻1118 fly embryos (Rubin and Spradling 1982; Brand and Perrimon 1993). Ectopic overexpression was performed by using the UAS-Gal4 system (Brand and Perrimon 1993). As Gal4-drivers, the prd-Gal4 driver PGAL4-prd.FRG1/TM3, Sb¹ (Bloomington stock collection, no. 1947), the P{GAL4-Hsp70.PB} 31-1 driver (Bloomington stock collection, no. 1822), the w[*]; P{GAL4-sim.3.7}2/CyO; P{GAL4-sim.3.7}3 sim-Gal4 (Bloomington stock collection, no. 7150), and the Actin5c-Gal4 driver y[1] w[*]; P{w[+mC] = Act5C-GAL4}17bFO1 (Bloomington stock collection, no. 3954) were used. To verify the 31-1-Gal4 driver we used an UAS-GFP reporter line.

Homozygous UAS inx7myc (third chromosome) flies were crossed to prd-Gal4 flies. Progenies were collected after 12 h. Homozygous UAS wizinx7 flies were crossed to the 31-1-Gal4 neuronal or the sim-Gal4 midline cell driver (Brand et al. 1994; Scholz et al. 1997). The heat shock–inducible 31-1-Gal4 driver mediates expression in precursor cells of the developing central nervous system (CNS) and peripheral nervous system (PNS) (see Fig. 3E). Flies were allowed to lay eggs for 12 h on apple juice agar plates, followed by a heat shock for 1 h at 37°C. After the heat shock the embryos were aged for 3 additional hours at 25°C. The sim-Gal4 driver drives expression in all midline cells from stage 10 to stage 13 and is gradually restricted to midline glia (Scholz et al. 1997). Homozygous UAS wizinx7 flies were crossed to sim-Gal4 flies. Progenies were collected after 16 h. For real-time experiments, homozygous UAS inx7myc and wild-type flies were crossed individually to Actin5c-Gal4 flies. Progenies were collected after 12 h.

Plasmids

pUAST innexin7myc (UAS innexin7myc) was generated by cloning the open reading frame of inx7-RA (CG2977), c-terminally tagged with a myc-tag in the pUAST vector (Brand and Perrimon 1993). The insert was amplified by PCR using the following primers: GGA ATT CCA GTT ACG AGG AGC TAG CCG CGG CAA GC as a forward primer containing an EcoRI restriction site and GCT CTA GAG C TTA CAG ATC CTC TTC AGA GAT GAG TTT CTG CTC TAC AGG TAA CTT TGC CAT as a reverse primer containing an XbaI restriction site, stop codon, and myc-tag. Selected constructs were injected in fly embryos.

Cloning of recombinant plasmids for the generation of stable RNA interference (RNAi)-inducible fly strains was perfomed as described by Lee and Carthew (2003). A 480-bp fragment containing the forth exon of inx7-RA (bases 791 to 1271) was amplified by polymerase chain reaction (PCR), with PCR primers containing at their 5′ ends a XbaI restriction site, which is compatible with the AvrII and NheI sites. The described insert was cloned in two different orientations (“tail to tail”) into the AvrII and NheI site of pWiz (gift from R.W. Carthew), which was dephosphorylated with alkaline shrimp phosphatase prior to ligation. For transformation, competent SURE cells were used (Stratagene). Recombinants in “tail to tail” orientation were screened, selected and injected into fly embryos.

Antibody Generation

A polyclonal antiserum was generated in rabbits (Davids Biotechnology, Regensburg, Germany) against an Innexin7 (Inx7)-specific oligopeptide conjugated to KLH via a N-terminal cysteine. The peptide used was AKNRYPELSGLDTI, representing amino acids 316-331 of Innexin7 (see Fig. 1A). Immunoglobulin G's (IgGs) were purified from immune serum by affinity purification; antibody was used in a dilution of 1:100.

Figure 1 Innexin7 (inx7) antibody generation and Innexin7 protein expression pattern in Drosophila embryos. (A) Predicted structure of the Drosophila Innexin7 protein. Striking features, such as the putative PDZ domain, and phosphorylation sites are depicted using different symbols. The peptide derived from the C-terminal region, which was used for antibody generation is depicted in green. (B) Peptide competition experiment (Western blot). The specific Innexin7 signal (51.62 kDa, arrow) was competed with excess of the C-terminally derived Innexin7 peptide. (C) Ectopic expression of UASinx7myc in seven stripes using the paired-Gal4 driver. The anti-Innexin7 antibody (red) detects ectopic Innexin7 myc expression in the seven stripes of the paired pattern (see Methods) and thereby colocalizes with the anti-Myc antibody (green). (D–F) Lateral view of stage 5 (D, E) and stage 6 (F) embryos. Innexin7 (red) is localized within the nucleus and does not colocalize with armadillo/β-catenin (green), an apicolateral located core component of adherens junctions. (G) Dorsal view of a stage 7 embryo. Innexin7 staining is detected in the invaginating embryonic midline and in the developing neuroblasts, which delaminate from the epidermis. (H, I) Innexin7 is identified in the midgut (mg), hindgut (hg), and within a subset of cells within the ventral nerve cord (vnc) along the anterior-posterior axis. Innexin7 localization is found in a punctate pattern in the cytoplasm and at the membrane of epithelial cells (I, arrow). In contrast, nuclear staining is kept only in a few cell types including in the developing central nervous system (I, arrowhead; see also Figure 2). (J) Peptide competition experiment in Drosophila embryos. Dorsal view of a stage 15 embryo. Innexin7 staining is completely abolished upon competition with excess of Innexin7 peptide (compare to H and I).

Embryo Fixation and Immunohistochemistry

Embryos were staged (and heat shocked) as described previously. After embryo collection, the chorion was removed in of 4% sodium hypochlorite. After washing, the embryos were transferred to preheated (95°C) sodium chlorite/ TritonX100 (0.15 M NaCl, 0.1% Triton X-100) solution and fixed by shaking the embryo/solution mixture for 5 s at 95°C (water bath). After fixation the embryo/solution mixture was immediately chilled on ice. The solution was replaced by methanol/EGTA (0.005 M EGTA, pH 8.0 in methanol) and vortexed for 30 s. Methanol/EGTA was replaced by methanol. Fixed embryos were stored in methanol at −20°C.

Embryos were stained as described previously (see Fuss and Hoch 1998; Bauer et al. 2004). The following antibodies were used: 22C10 (1:10), anti-FasII (1:10), anti-Slit (1:5), and anti-BP102 (1:10) (all from Developmental Studies Hybridoma Bank). Additionally, we used mouse monoclonal anti-c-Myc 9E10 (c-Myc; Santa Cruz 1:20). As secondary antibodies we used anti-rabbit-Cy3 (1:200; Dianova) and anti-mouse-Alexa488 (1:200; Molecular Probes).

For peptide competition experiments in embryos the peptide AKNRYPELSGLDTI (see Antibody Generation) was used at a concentration of 1000:1 (peptide to antibody). The peptide was preincubated with the anti-Innexin7 and 22C10 antibody at room temperature for 3 h, spun down, and the supernatant then used for immunostaining. As a control, we used anti-Innexin7 and 22C10 antibody incubated in the absence of the peptide.

Microscopy and Image Processing

Fluorescent images were recorded using a Leica TSP2 confocal microscope (Leica, Wetzlar, Germany). Images of multi-labeled samples were acquired sequentially on separate channels. All images were processed with Adobe Photoshop software. Contrast and brightness of the pictures were changed in a linear fashion.

Immunoblotting and Biochemical Peptide Competition

For immunoblotting, wild-type embryos were homogenized in RIPA buffer (150 mM NaCl; 1% IGEPAL CA-630, 0.5% sodium deoxycholate, 0.1% sodium dodecyl sulfate [SDS], 50 mM Tris [pH 8.0]) with protease inhibitors (Roche). The embryo lysate was cleared by centrifugation (5 min, 4500 × g, 4°C).

The protein content of the supernatant was measured by a Bicinchoninic acid (BCA)-based standard assay. Thirty micrograms of protein was diluted in SDS–polyacrylamide gel electrophoresis (PAGE) sample buffer, and boiled for 3 min. Samples were run on a 12.5% SDS gel, and transferred to a polyvinylidene fluoride (PVDF) membrane (Immobilon P transfer membrane; Millipore). The immunoblot membrane was blocked with nonfat dry milk in Tris-buffered saline (TBS) plus 0.05% Tween (TBST) and then cut in stripes (lanewise). The stripes were incubated with preimmune serum (1:250, rabbit), anti-Inx7 (1:250, rabbit), with peptide preincubated anti-Inx7 antibody (1000:1, peptide to antibody, antibody dilution 1:250) (all in TBST). For peptide competition, peptide and antibody were diluted in TBST and preincubated for 30 min at room temperature. One strip was incubated with TBST alone (control secondary antibody). After washing, bound antibody was visualized with peroxidase-conjugated donkey α-rabbit IgG (1:15000) using the enhanced chemiluminescence (ECL) system (Amersham Pharmacia).

Reverse Transcriptase–Polymerase Chain Reaction (RT-PCR)

Total RNA was prepared from different embryo collections: UAS inx7myc/Actin5c Gal4; UAS wizinx7/ Actin5c Gal4; +/Actin 5c Gal4. RT-PCR was performed as described by Zinke et al. and Fuss et al. with minor modifications (Zinke et al. 2002; Fuss et al. 2006). Briefly, 0.5 μ g of total RNA was used for first strand cDNA reaction with oligo-d(T) primers and reverse transcriptase (both QuantiTect, Qiagen) in a final volume of 10 μ l at 42°C for 60 min. After inactivation of the reverse transcriptase, 40 μ l H₂O were added to a final volume of 50 μ l. innexin7 primer pairs were designed and were tested for primer-dimer artefacts and efficiencies using the iQ5 Optical System Software (BIO RAD). Primer efficiency determined for the primer pair was: 107.4 % (R² = 0.999).

PCR analysis with the innexin7 primer pair was performed on aliquots of the reverse transcribed total RNA (described above) according to the manufacturer's instructions using the iQ5 Real-Time PCR Detection System (BIO RAD). PCR reactions were set up with the iQ SYBR Green Supermix (BIO RAD), in a final volume of 25 μ l. For normalization the expression levels of the ribosomal protein L32 (RpL32, rp49) and actin5c (Act5c) were used (for details see Fuss et al. 2006). The following primers were used in the iQ5 Real-Time PCR reaction:

Table

inx7 RT NT-F:	GCT GCA CTA TCG ATG GAC CT
inx7 RT NT-R:	AGG TGT TGA TGA CCG GAG AC
act-Sy-F:	GTG CAC CGC AAG TGC TTC TAA
act-Sy-R1:	TGC TGC ACT CCA AAC TTC CAC
rp49-Real-F1:	GCT AAG CTG TCG CAC AAA TG
rp49-Real-R1:	GTT CGA TCC GTA ACC GAT GT

RESULTS

Innexin7 Is Expressed in Epithelia of the Drosophila Embryo

Members of the connexin, pannexin, and innexin gene families encode structurally similar four-pass transmembrane proteins with two extracellular loops, four hydrophobic membrane-spanning domains, and three cytoplasmic domains, including an intracellular loop domain and cytoplasmic N-and C-terminus (Bauer et al. 2005). Sizes of the Drosophila Innexin proteins range from 42 to 55 kDa. The carboxy-terminal tail of innexins show the highest variability in length. Innexin7, with 134 amino acids, contains one of the largest C-terminal domains (Fig. 1A). The C-terminus contains multiple putative phosphorylation sites and a putative PDZ binding domain (class II motif; Bauer et al. 2005). Functional data for these domains is still lacking. RNA in situ hybridisation experiments have previously shown that most of the Drosophila innexin genes are expressed in complex and overlapping expression patterns during development (Stebbings et al. 2000; Bauer et al. 2001). Innexins 1, 2, 3, and 7 show a high degree of overlapping expression during oogenesis and embryogenesis and are found in many epithelial tissues, including the epidermis, the gut, and the nervous system (Stebbings et al. 2000; Bauer et al. 2001; Ostrowski and Hoch, unpublished).

As a first step towards a functional analysis of Innexin7 during embryonic development, we generated an anti-Innexin7 antibody against a peptide containing the C-terminal amino acids 316 to 331 of Innexin7 (see Methods; Fig. 1A). This antibody recognized a single band at 52 kDa in embryonic extracts of wild-type embryos (Figure 1B, lane 2), which corresponds to the predicted size of the protein (51.62 kDa). No specific signal could be detected using the preimmune serum (Figure 1B, lane 1). Peptide competition experiments confirmed the specificity of the antibody (Figure 1B, lane 3). To further test the specificity in vivo, we overexpressed an innexin7-myc construct in seven stripes in the epidermis using paired-Gal4:UAS-innexin7-myc embryos and performed immunostaining using anti-Innexin7 antibodies. As shown in Figure 1C, the antibody detects the seven stripe pattern. In contrast, the seven-stripe pattern was lost, when these embryos were incubated with excess of Innexin7 peptide (data not shown; see also Figure 1H to Figure 1J), demonstrating the specificity of the antibody in vivo.

Immunostaining of wild-type embryos using anti-Innexin7 antibodies combined with confocal microscopy revealed that the protein is strongly expressed in all epithelia from early to late stages of development, including in the developing epidermis, the gastrointestinal tract, and the nervous system (Figure 1D to Figure 1I). In early embryonic stages 5 to 7, we find a nuclear localization of Innexin7 in all embryonic cells (Figure 1D to Figure 1G), as shown by coimmunostaining with anti-Armadillo/β-catenin antibodies. From stage 7/8 onwards until late stages, Innexin7 localization is found in a punctate pattern in the cytoplasm and at the membrane of epithelial cells (Figure 1H, Figure 1I). In contrast, nuclear staining is kept only in a few cell types including in the developing central nervous system (Figure 1I, see also Fig. 2). Innexin7 staining is abolished upon peptide competition at late stages (Figure 1J), again demonstrating the specificity of the antibody in vivo.

Figure 2 Innexin7 is dynamically expressed in the Drosophila central nervous system (CNS) during embryonic nervous system development. Innexin7 expression in a stage 7 (A) and a stage 9 (B) embryo. Innexin7 (red) is localized to the developing neuroectoderm and enriched within the embryonic midline (ml) (B). (C–G) During progression of the embryonic development, two different expression domains of Innexin7 can be detected: the midline (C) and neuronal cells next to the midline (E, F, G). During embryogenesis, Innexin7-expressing cells in the embryonic midline are restricted to only two types of midline glia (mlg), here indicated by colocalization with the marker slit (C, green), which is made by midline glia cells. Innexin7 is not coexpressed with the homeodomain protein repo (green), which is expressed in all CNS glia except the midline glia (D). (E, F, G) Double staining of Innexin7 with neuronal markers prospero (pros, green) and FasciclinII (FasII, green). The Innexin7-expressing cells next to the midline colocalize with the nuclear marker prospero (F) and are ensheathed by FasII-positive fascicles (G), which led to their identification as vMP2 and pCC cells. Additionally, three other Innexin7-positive neuronal cells could be detected (G, arrow), which are not yet identified.

Innexin7 Expression in the Embryonic Central Nervous System

The Drosophila central nervous system (CNS) derives from neural precursor cells, the neuroblasts, which are born from the neuroectoderm by the process of delamination (Hartenstein and Campos-Ortega 1984). Before gastrulation, the neuroectoderm is positioned in a ventrolateral position and separated from the ventral mesoderm cells by a single row of cells on either side of the embryo, the mesectodermal cells. Mesectodermal cells give rise to the midline from which glia cells and neurons, including pioneering neurons of the intersegmental connectives of the CNS, are derived (Klämbt et al. 1991).

Innexin7 expression in the developing nervous system is dynamic. Before gastrulation, it covers the neuroectoderm region and the mesectodermal cells (Figure 2A). Innexin7 expression is visibly lost in the neuroectoderm during germband extension and persists in the midline cells (Figure 2B). Glia cells in the embryonic CNS can be subdivided into two major classes, the midline glia and the lateral glia (Giesen et al. 1997). Midline glia act as pathfinders for neuronal cells to build longitudinal nerve cords and they act as spacers for the anterior and posterior commissures (Klämbt et al. 1991; Klämbt and Goodman 1991a, 1991b). Midline glia can be identified using slit, an evolutionarily conserved signalling molecule (Rothberg et al. 1988; Zou et al. 2000), as a marker. In contrast, repo (reversed polarity) is expressed in all CNS glia except the midline glia (Yuasa et al. 2003). Our coimmunostainings of Innexin7 with slit or repo indicate that Innexin7 is coexpressed with slit in the midline glia, whereas colocalization with repo was not observed (Figure 2C, Figure 2D).

To analyze whether Innexin7 is also expressed in neurons derived from the midline precursor cells, we performed costainings with the neuronal markers Prospero (pros) and FasciclinII (fasII), which are known to localize to the dorsal [pros] and ventral [pros and fasII] MP2 and pCC cells of the CNS. pCC and MP2 neurons are pioneer cells for axon guidance (Vaessin et al. 1991; Klämbt and Goodman 1991a, 1991b; Matsuzaki et al. 1992). Prospero is a nuclear protein with a putative homeodomain (Chu-LaGraff et al. 1991) and is also highly expressed in longitudinal and belt glia cells of the CNS (Doe et al. 1991). FasciclinII is expressed on the axons and glial cells of the MP1 and vMP2 fascicles (Grenniloh et al. 1991). We found colocalization of Innexin7 with both markers, which suggests that Innexin7 is expressed within the vMP2 and pCC neuron (Figure 2E to Figure 2G). It is of note that in the developing CNS, Innexin7 is colocalized with Prospero in the nucleus of vMP2 and pCC neurons (Figure 2F), whereas it is localized in a punctuate pattern in the cytoplasm and the membrane of the developing epidermis (Figure 2E, Figure 2F).

Innexin7 Is Required for Embryonic Nervous System Development

To analyze the role of innexin7 during CNS development, we used RNAi-mediated knock down of the gene. To this end, we generated a transgenic line carrying an RNAi knock down construct for innexin7 (UASwiz inx7), in which part of its coding regions was cloned into a “tail to tail” orientation (Figure 3A; Methods). We first used an actin-Gal4 driver line, which mediates ubiquitous expression, in combination with UASwiz inx7. To test the extent of the innexin7 knock down, we performed quantitative real-time (RT) PCR experiments on mRNA isolated from wild-type embryos (control) and from embryos in which innexin7 was either knocked down by RNAi or overexpressed by using UAS-innexin7 (Figure 3B). The expression levels of actin 5C (Act5C, act) and Ribosomal protein L32 (RpL32, rp49) were used as reference genes for normalization (see Methods). As shown in Figure 3B, we obtained an almost complete reduction (93%) of the innexin7 mRNA level in the knock down experiment, as compared to wild type.

Figure 3 RNAi-mediated knockdown of innexin7 in the nervous system causes severe defects in the Drosophila embryonic nervous system. (A) Scheme of the UASwizinx7 construct used for knockdown of innexin7. (B) UASwizinx7 can reduce the level of innexin7 mRNA expression. RT-PCR shows that the UASwizinx7 knockdown decreases innexin7 mRNA expression level down to 7% as compared to wild type. (C, D, F–H) Wild-type pattern and 31-1-Gal4: UASwizinx7 knockdown phenotypes stained with 22C10 (green). (C, D) Lateral view of wild-type embryos stained with 22C10. (E) 31-1-Gal4–mediated expression pattern monitored using UAS-GFP. (F–H) Lateral view of innexin7 RNAi knockdown embryos. By overexpressing the RNAi construct in the nervous system, a severe disorganisation of the vnc along the anterior posterior axis can be observed (arrows in G and H). (H) Magnification of the embryo depicted in (G). (I–M) Wild-type pattern and sim-Gal4: UASwizinx7 knockdown phenotypes stained with 22C10 (green) and BP102 (green). (I, K) Dorsal view of wild-type embryos stained with 22C10 (I) or BP102 (K). (J, L, M) Expression of UASwizinx7 in midline cells only results in intrasegmental fusion of the ventral nerve cord and disruption of longitudinal, posterior, and anterior commissures in the CNS.

To analyze the role of Innexin7 during CNS development, we first used the neural driver line 31-1-Gal4, which mediates ubiquitous expression in precursor cells of the CNS (Brand et al. 1994), in combination with the UASwizinx7 effector line. We analyzed the embryos by performing coimmunostainings with the neuronal cell markers 22C10 (futsch). The 22C10 antigen is expressed by some CNS neurons (motor neurons and others) as well as by all neurons in the PNS of Drosophila (Hummel 2000; Figure 3C, Figure 3F, Figure 3I). In this innexin7 knock down experiment, the longitudinal, posterior, and anterior commissures in the CNS were disrupted and we often found holes in the ventral nerve cord, indicating that cell death has occurred (Figure 3G, Figure 3H). In addition, the proper outgrowth of axon fibers of the ventral nerve cord did not occur properly, resulting also in defects in the peripheral nervous system (Figure 3D), suggesting that innexin7 may be required for axon guidance.

Because Innexin7 is specifically expressed in midline glia cells, which are involved in the formation of the anterior and posterior commissures of the CNS and in the vMP2 and pCC neurons, which are both derived from the midline precursor cells (Klämbt et al. 1991; Klämbt and Goodman 1991a, 1991b), we analyzed the knock down of innexin7 in these cells using the sim-Gal4 driver line in combination with the UASwiz inx7 effector line. The sim-Gal4 driver mediates expression in all midline cells from stages 10 to 13 (Scholz et al. 1997). We analyzed the embryos by performing coimmunostainings with the panaxonal BP102 antibody visualizing the major axon tracts within the CNS, the bilaterally symmetric longitudinal commissures, and the two commissures connecting each hemisegment (Figure 3K; Seeger et al. 1993, Fig. 3K, Fig. 3L). We found intrasegmental fusion of the ventral nerve cord and that the longitudinal, posterior, and anterior commissures in the CNS were disrupted (Figure 3L, Figure 3M), pointing to an important function of innexin7 in midline cells, which is consistent with the Innexin7 protein expression in these cells. In summary, these data indicate that establishing and maintaining usual Innexin7 levels is critical for the survival of neural cells and for normal development and wiring of the CNS.

DISCUSSION

We generated an antibody against Innexin7 and analyzed its role in embryonic development by using transgenic RNAi lines. Our study provides evidence for an essential function of innexin7 in the development of the embryonic nervous system of Drosophila.

Innexin7 Is Expressed in Epithelia Tissues and in the Developing Central Nervous System

Our immunohistochemical analysis indicates that innexin7 RNA and protein are already present at the earliest stages of embryonic development, suggesting a maternal component, which has also been observed for innexins 2 and 3 (Figure 1D to Figure 1G; Bauer et al. 2004; Lehmann et al. 2006). We find Innexin7 protein expression in all epithelia from early to late stages of development, including in the developing epidermis, the gastrointestinal tract, and the nervous system (Figure 1, Figure 2). In early embryonic stages, a nuclear localization of Innexin7 is observed in all embryonic cells (Figure 1D to Figure 1G), whereas at later stages, Innexin7 is found in a punctate pattern in the cytoplasm and at the membrane of epithelial cells (Figure 1H). During CNS development, Innexin7 is expressed in cells of the neuroectoderm and the mesectoderm and at later in stages of embryogenesis, its expression is largely restricted to a segmental pattern of few glia and neuronal cells derived from the midline precursors. Coimmunostainings with neuronal and glia markers indicate that Innexin7 is expressed in MP2 and pCC cells (Figure 2F). pCC and MP2 neurons are pioneer cells for axon guidance (Vaessin et al. 1991; Klämbt et al. 1991; Klämbt and Goodman 1991a, 1991b; Matsuzaki et al. 1992). Knock down of innexin7 in midline cells results in a severe disruption of the ventral nerve cord (Figure 3J, Figure 3L, Figure 3M), indicating that Innexin7 is required in midline cells and plays an essential role of CNS development.

Gap junctions have been implicated in many aspects of neural development. Innexin1 plays an essential role for optic lobe development of adult flies. innexin1 mutants (ogre) show a strongly reduced number of neurons within the optic lobe, indicating that it is required for cell survival (Watanabe and Kankel 1992). Gap junctions are also involved in neural connectivity. Transient expression of shakingB and ogre, possibly between the retina and lamina, may play a role in final target selection or chemical synapse formation or both in the Drosophila visual system (Curtin et al. 2002). Furthermore, transient gap junctions between pathfinding axonal growth cones and non-neuronal “guidepost cells” or between the growth cones of two extending axons may provide a means for positional information to pass between cells for the selection of targets during CNS development (Taghert et al. 1982; Wolszon et al. 1994). In C. elegans, it was recently shown that the innexin gap junction protein NSY-5 coordinates left-right asymmetry in the developing nervous system (Chuang et al. 2007). It was shown that Nsy-5–dependent gap junctions in the embryo transiently connect the AWC olfactory neurons with those of numerous other neurons, thereby establishing stochastic, asymmetric patterns of gene expression during embryogenesis required for left-right asymmetry. Temporal and spatial regulation of gap junctional coupling has been hypothesized to be one mechanism by which morphogens may be restricted and cell fate regulated during embryogenesis. The precise role of Innexin7 during CNS development will have to be addressed further by cell-specific knock down experiments.

Tissue-Specific Nuclear and Membrane Expression of Innexin7

In early embryonic stages and in few cells of the developing CNS, we observed a nuclear localization of Innexin7 (Figure 1D), whereas Innexin7 was found in a punctate pattern in the cytoplasm and at the membrane of most epithelial tissues at later stages of development (Figure 1H). The functional significance of these stage and tissue-specific subcellular localization patterns is not known. The membrane localization of Innexin7 is consistent with its putative role as a gap junction protein. Because the anti-Innexin7 antibody is made against a C-terminal peptide of the protein, the nuclear signal may reflect a tissue and stage-specific cleavage of the Innexin7 C-terminus and its subsequent nuclear localization. However, biochemical evidence for such a mechanism is still lacking. We recently demonstrated that Innexin2 is involved in the transcriptional regulation of the hedgehog, wingless, and Delta genes (Lechner et al. 2007). The C-terminal domain of Innexin2 may participate in transcriptional activation of these genes.

This work was supported by a DFG grant to M.H. (SFB 645).