SEARCH

SEARCH BY CITATION

Keywords:

  • Drosophila;
  • dorsal closure;
  • morphogenesis;
  • Sec61α;
  • translocon;
  • Dpp;
  • thick veins;
  • JNK;
  • cuticle

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  9. Supporting Information

During dorsal closure in Drosophila, signaling events in the dorsalmost row of epidermal cells (DME cells) direct the migration of lateral epidermal sheets towards the dorsal midline where they fuse to enclose the embryo. A Jun amino-terminal kinase (JNK) cascade in the DME cells induces the expression of Decapentaplegic (Dpp). Dpp signaling then regulates the cytoskeleton in the DME cells and amnioserosa to affect the cell shape changes necessary to complete dorsal closure. We identified a mutation in Sec61α that specifically perturbs dorsal closure. Sec61α encodes the main subunit of the translocon complex for co-translational import of proteins into the ER. JNK signaling is normal in Sec61α mutant embryos, but Dpp signaling is attenuated and the DME cells fail to maintain an actinomyosin cable as epithelial migration fails. Consistent with this model, dorsal closure is rescued in Sec61α mutant embryos by an activated form of the Dpp receptor Thick veins. Developmental Dynamics 239:784–797, 2010. © 2010 Wiley-Liss, Inc.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  9. Supporting Information

Dorsal closure is one of the final morphogenetic events necessary for elaborating the body plan of the Drosophila larva. Near the midpoint of embryogenesis, immediately after germ band retraction, epidermal tissue covers the ventral and lateral regions of the embryo leaving a large dorsal hole covered only by a squamous extraembryonic epithelium known as the amnioserosa. Coordinated cell shape changes in the absence of cell division in the epidermal cells, coupled with cell shape changes and cell death in the amnioserosa, drive the elongation of the epidermal cell sheets dorsalward, where they meet at the dorsal midline and thereby enclose the embryo.

Dorsal closure in wild-type animals involves three distinct stages. Prior to the start of dorsal closure, the cells of the lateral epidermis are polygonal in shape. During the first phase of dorsal closure, known as initiation, the dorsalmost epithelial (DME) cells of the epidermis (also referred to as leading edge cells) elongate in the dorsal-ventral (D-V) axis, whereas the more ventral cells remain polygonal. These DME cells provide an organizing center for the events of dorsal closure. During the initiation phase, the DME cells accumulate actin and myosin in a contractile ring at the level of the adherens junction, which eventually link across these cells to create a continuous actin cable that coordinates the migration of the leading edge of the epithelium (Young et al.,1993). During the second phase of dorsal closure, known as epithelial migration, the more ventral epidermal cells begin to elongate in the D-V axis as the epidermal sheets migrate towards the dorsal midline. This epidermal migration results from the contraction of the actinomyosin cable in the DME cells through a purse string mechanism (Young et al.,1993; Kiehart et al.,2000; Hutson et al.,2003). The actin cable is also necessary to maintain an organized leading front of the DME cells during the migration process (Bloor and Kiehart,2002; Jacinto et al.,2002). Over the last several years, it has become clear that additional forces are contributed by the amnioserosa, which undergoes coordinated cell-shape changes, including contractions perpendicular to the anterior-posterior axis, as well as apical constrictions that eventually lead to those cells being extruded from the epithelium (Kiehart et al.,2000; Harden et al.,2002; Franke et al.,2005; Fernandez et al.,2007). The yolk sac also plays an essential role in these processes as it serves as an attachment substrate for the amnioserosal cells as they contract (Narasimha and Brown,2004; Reed et al.,2004). Finally, during the completion or zippering phase of dorsal closure, the DME cells meet at the dorsal midline and fuse with DME cells from the contralateral side starting at the anterior and posterior ends and gradually suturing the epidermis towards the center. Again the DME cells show organizing activity by sprouting filopodia and lamellapodia that aid in the alignment and fusion of the two epidermal sheets (Jacinto et al.,2000).

Nearly a hundred genes have been identified whose mutant phenotype includes some defect in dorsal closure. Since it is only the epidermal cells that secrete cuticle late during embryogenesis, a failure to complete dorsal closure produces a characteristic dorsal open phenotype in the resulting dead embryos. Genes that produce this phenotype when perturbed have typically been grouped into two categories, those encoding signaling molecules that are likely required to regulate dorsal closure and those that encode the cellular effectors of the process. This latter class includes cytoskeletal genes such as zipper, encoding non-muscle myosin (Young et al.,1993), myosin phophatase (Mizuno et al.,2002), lethal(2)giant larvae, (Manfruelli et al.,1996), and chickadee (Jasper et al.,2001), encoding profilin. Additionally, extracellular matrix genes including the integrin subunits lethal (1)myospheroid (Leptin et al.,1989) and scab (Stark et al.,1997), also play roles in dorsal closure, as do genes whose products are components of cellular junctions including armadillo (Peifer and Wieschaus,1990), canoe (Miyamoto et al.,1995), coracle (Fehon et al.,1994), discs large (Perrimon,1988), and Neurexin (Baumgartner et al.,1996).

Genetic and biochemical analyses of the genes predicted to encode signaling molecules reveal that although they fall into many classes, they generally impinge on two different conserved signaling pathways, a Jun amino-terminal kinase (JNK) pathway and a transforming growth factor β (TGF-β) pathway (reviewed in Harden,2002). The JNK pathway is a conserved mitogen-activated protein kinase (MAPK) cascade consisting of sequentially acting serine/threonine kinases that lead to the phosphorylation of the transcription factor DJun. DJun then complexes with DFos (encoded by kayak), which together make up the adapter protein 1 (AP-1) transcriptional complex (Riesgo-Escovar and Hafen,1997a). Although JNK signaling is active in both the amnioserosa and the dorsal epidermis during germ band retraction, it is restricted to the DME cells during the initiation of dorsal closure (Reed et al.,2001).

A critical transcriptional target of the JNK pathway in the DME cell is the TGF-β ligand decapentaplegic (dpp; Glise and Noselli,1997; Hou et al.,1997; Riesgo-Escovar and Hafen,1997b). Dpp binds to its receptor, which is a heterotetramer of two type I receptors encoded by thick veins (tkv) and two type II receptors encoded by punt (reviewed in O'Connor et al.,2006). Upon binding of Dpp to its receptor complex, the receptor phosphorylates the SMAD protein Mothers Against Dpp (Mad) that then binds to a co-SMAD named Medea and translocates to the nucleus where it regulates the transcriptional response to the Dpp signal. Mutations in tkv and punt display dorsal open phenotypes (Affolter et al.,1994; Nellen et al.,1994; Arora et al.,1995). Detailed analysis of the cellular events occurring in tkv mutant embryos indicates that Dpp signaling is required for the DME cells to maintain an actinomyosin contractile cable and drive the dorsalward movement of the lateral epidermis, as well as for the cell-shape changes in the amnioserosa (Fernandez et al.,2007).

Here we report on the isolation and characterization of a strong loss of function mutation in Sec61α, the major subunit of the protein translocon complex for co-translational insertion into the endoplasmic reticulum. Surprisingly, zygotic loss of Sec61α results in specific defects in dorsal closure during mid-embryogenesis. Sec61α mutant embryos initiate dorsal closure correctly, as indicated by normal JNK signaling in the DME cells, but cannot generate or maintain a robust contractile actin/myosin cable, and thus fail during the epithelial migration stage of dorsal closure. Strong reductions in the levels of phosphorylated Mad in the lateral epidermis indicate that Dpp signaling is strongly reduced in Sec61α mutant embryos. In support of this idea, an activated Tkv receptor provides a highly penetrant, partial rescue of the dorsal closure defects in Sec61α mutant animals.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  9. Supporting Information

Identification of Sec61α Alleles

While mapping an EMS-induced mutation recovered from a screen for dominant modifiers of the transcription factor broad (Ward et al.,2003), we identified a mutation that failed to complement Df(2L)BSC6 (cytology 26D3;26F7). We tested loss-of-function mutations for genes that map to this interval and found that the mutation failed to complement P{lacW} Sec61αk04917 (Table 1). P{lacW}Sec61αk04917 (hereafter referred to as Sec61αk04917) has a P-element inserted into the second intron of the Sec61α gene (Fig. 1A). Sec61α encodes the major subunit of the tripartite translocon complex required for co-translational insertion of proteins into the endoplasmic reticulum (Johnson and van Waes,1999). Sequence analysis of genomic DNA isolated from mutant embryos revealed a nonsense mutation in Sec61α (G/C to A/T transition at nucleotide 683 of GenBank sequence AY069569 generating a Trp193 to stop mutation; Fig. 1A). We have, therefore, named this mutation Sec61α1. Sec61α1 is predicted to encode a protein that is prematurely truncated at the end of the fifth of ten transmembrane domains, almost certainly creating a nonfunctional protein (Fig. 1A).

thumbnail image

Figure 1. Molecular organization of the Sec61α locus, and embryonic cuticle defects associated with zygotic loss-of-function alleles of Sec61α. A: Schematic diagrams of the genomic locus (top), transcript (middle), and protein of Sec61α (bottom). The centromere is to the right of this genomic fragment. The transcript is drawn to scale with the genomic diagram, with exons depicted as rectangles and coding regions filled with black. Introns are depicted as bent lines below the exons. Black bars in the protein schematic represent the location of the ten transmembrane domains. The genomic region used as a rescue construct is indicated by the bracketed line. P{lacW} Sec61αk04917 in inserted into the second intron as shown. The location of the Sec61α1 mutation is indicated by an arrow in the transcript and an asterisk in the protein. The nucleotide and protein sequences of the Sec61α1 mutation are shown below the protein. Nucleotide sequence number is based on GenBank sequence AY069569. B–E: Brightfield photomicrographs of cuticle preparations from wild-type and Sec61α mutant animals. B: A w1118 late stage-17 embryo showing normal cuticular features including ventral denticle belts (arrowheads) and head skeleton (arrow). C: A Sec61α1 mutant embryo showing no cuticular features. D: A Sec61α1/Df(2L)BSC6 mutant embryo also showing no cuticle. E: A Sec61α1 late stage-17 mutant embryo containing two copies of the Sec61α genomic rescue fragment inserted on the third chromosome shows normal cuticular features. All embryos are depicted with anterior to the left and the dorsal surface facing up. The embryos in B and E were taken prior to hatching, but would have hatched given the chance.

Download figure to PowerPoint

Table 1. Lethal-Phase of Sec61α Mutant and Rescued Animalsa
Genotype% Embryonic lethality (n)% Larval lethality (n)% Pupal lethality (n)
  • a

    Values represent the mean ± sd from three independent experiments. Numbers in parentheses indicate total number of animals of indicated genotype that were scored. N/A, not applicable.

  • b

    Genomic rescue construct line CM15-1.

Sec61α1100 (400)N/AN/A
Sec61α1/Df(2L)BSC6100 (299)N/AN/A
Sec61α1/Sec61αk0491746 ± 6 (448)100 (250)N/A
Sec61αk0491758 ± 8 (582)100 (250)N/A
Sec61α1;P{Sec61α}b5 ± 1 (397)37 ± 4 (376)100 (237)
Sec61α1/Sec61αk04917; P{Sec61α}b2 ± 3 (298)22 ± 8 (292)4 ± 3 (230)

Lethal phase analyses of Sec61α1 support the molecular analysis and indicate that it is an amorphic allele. Specifically, Sec61α1, Df(2L)BSC6 and Sec61α1/Df(2L)BSC6 mutant animals are completely embryonic lethal, and cuticle preparations of the dead embryos for all of these genotypes showed no ventral denticle belts, differentiated head skeleton, or any recognizable cuticle (Table 1 and Fig. 1C, D; and data not shown). In contrast, Sec61αk04917 and Sec61α1/Sec61αk04917 mutant animals showed only ∼50% embryonic lethality (Table 1), with the majority of the animals dying in early first instar, suggesting that Sec61αk04917 is a hypomorphic allele.

Sec61α Mutant Animals Show Specific Defects in Dorsal Closure During Embryogenesis

Since Sec61α1 mutant animals have little to no epidermal cuticle, we were able to examine the terminal phenotype of these animals by indirect immunofluorescence using antibodies against Coracle (Cor). Cor is expressed in all ectodermally derived epithelial cells and localizes to septate junctions (Fehon et al.,1994). All late stage 17 Sec61α1 mutant embryos have no epidermis over the dorsal third of the embryo with the brain and gut extruded through the dorsal hole (Fig. 2C), indicating a completely penetrant defect in dorsal closure. There is little variability in this terminal phenotype as we never observed dorsally closed or partially closed mutant embryos. Sec61α1/Df(2L)BSC6 mutant embryos showed identical phenotypes in both penetrance and expressivity (Fig. 2D). Interestingly, Df(2L)BSC6 mutant embryos also survived to late embryogenesis and showed identical defects in dorsal closure (data not shown).

thumbnail image

Figure 2. Sec61α mutant animals show defects in dorsal closure. A–D: Confocal optical sections of (A) w1118 stage-16 embryo, (B) Sec61α1 late stage-17 embryo, (C) Sec61α1/Df(2L)BSC6 late stage-17 embryo, and (D) Df(2L)BSC6 late stage-17 embryo all stained with antibodies against Coracle to visualize cell outlines in the epidermis. Anterior is to the left and dorsal is up in all cases. Note that in all the mutant animals, there are no epidermal cells covering the dorsal surface, and the brain (arrows) and guts (arrowheads) are extruded from the dorsal hole. Note that although dorsal closure has failed, head involution has occurred (asterisk in B indicates the dorsal anterior region of the epidermis after head involution had occurred). Scale bar = 20 μm.

Download figure to PowerPoint

To demonstrate that the dorsal closure defects of Sec61α1 were due to the loss of Sec61α, we generated a genomic rescue construct consisting of the entire Sec61α gene and extending ∼100 bp into the 5′ UTR of the upstream gene DLP and ∼350 bp into the 5′ UTR of the downstream gene CG9536 (Fig. 1A; note the genes are oriented in opposite directions). Four independent insertions of this construct rescued the embryonic lethality, and thus the dorsal closure defect of Sec61α1 mutant animals (Fig. 1E and Table 1). The rescue, however, was not complete with Sec61α1. For example, although 95 (± 1)% of Sec61α1;P{Sec61α line CM15-1} animals hatched as larvae, none of them eclosed and most died as pharate adults (Table 1). Rescue of Sec61α1/Sec61αk04917 mutant animals with this Sec61α genomic construct was more complete, resulting in >70% viable adults (Table 1). Together these results suggest that either there is a very strong requirement for Sec61α in late pupae that the rescue constructs cannot produce (but is enough in conjunction with Sec61αk04917), or that the Sec61α1 chromosome possesses a second-site pupal lethal mutation.

Although Sec61α1 mutants showed a completely penetrant defect in dorsal closure, most other morphogenetic events occurring during mid-embryogenesis continued to completion. Most notably, 100% of Sec61α1 mutant animals have normally involuted heads, even though dorsal closure has completely failed (n > 500 mutant embryos observed; Fig. 2B). We, therefore, assessed other embryonic morphogenetic events in Sec61α1 mutant embryos. To assess the morphogenesis of the peripheral nervous system, we stained wild-type and Sec61α1 mutant stage-14 embryos with an antibody that recognizes Neuroglian, and found that peripheral nervous system development was unaffected by lack of zygotic Sec61α expression (see Supp. Fig. S1A, B, which is available online). Similarly we assessed the morphogenesis of the salivary glands by staining stage-14 wild-type and Sec61α1 mutants with antibodies against Cor, and again found no differences (Supp. Fig. S1C, D). We next used a chitin-binding probe, wheat germ agglutinin, and antibodies against Uinflatable (a protein that localizes to the apical plasma membrane in tracheal cells; Zhang and Ward,2009) to assess tracheal morphogenesis. Although the patterning of the tracheal system was unaffected in Sec61α1 mutant embryos, the diameter of the tracheal tubes was smaller (Supp. Fig. S1E, F and data not shown). Tracheal morphogenesis requires the secretion of chitin into the lumen of the tracheae starting at stage 14. This chitin forms a cylinder in the center of the lumen that is required for proper radial expansion of the tracheae (Araujo et al.,2005; Devine et al.,2005; Tonning et al.,2005). Later in development, the expression and secretion of putative chitin modifying enzymes encoded by serpentine and vermiform are required for tracheal length control (Luschnig et al.,2006; Wang et al.,2006). It appears that chitin secretion into the tracheal lumen is reduced in Sec61α1 mutant embryos resulting in the diametric defects. However, since the tracheae are not convoluted nor show other tracheal length defects, it is likely that the secretion of the putative chitin modifying enzymes occurred normally. One additional tracheal defect that was observed was a failure in the fusion of some tracheal metameres. Approximately 50% of Sec61α1 mutant embryos had at least one discontinuity (n = 98, data not shown). Finally, just to exclude the possibility that Sec61α mutant embryos were dying about the time of dorsal closure, thus rendering these defects nonspecific, we generated time-lapse movies of late stage-17 Sec61α1 and Sec61α1/CyO, Dfd-EYFP embryos and observed that the Sec61α1 mutant embryos were alive and moving through the end of embryogenesis (data not shown). Taken together, these observations suggest that zygotic loss of Sec61α has very specific effects on the morphogenetic process of dorsal closure during mid-embryogenesis.

Dorsal Closure Initiates But Fails During the Epithelial Migration Phase in Sec61α Mutant Embryos

In order to determine how zygotic loss of Sec61α might result in such specific defects in dorsal closure, we first carefully examined the dorsal closure defects in Sec61α1 mutant embryos through a time-course analysis. We collected wild-type and mutant embryos at 1-hr intervals, aged them to 12–16 hr after egg laying, and stained them with antibodies against Cor to visualize the epidermis. Overall, Sec61α1 mutant embryos showed a 1–2-hr developmental delay and were going through dorsal closure when their heterozygous siblings had already completed the process. We, therefore, used gut and head morphology to accurately stage the animals. At the end of germ band retraction in mutant and wild-type animals, the cells of the lateral epidermis are polygonal in shape. During the initiation phase of dorsal closure in stage-13 wild-type embryos, the DME cells elongate in the dorsal-ventral axis (Fig. 3A), whereas the more ventral cells remain polygonal. DME cells in stage-14 Sec61α1 mutant embryos also elongate in the dorsal-ventral axis, suggesting that initiation of dorsal closure had begun normally (Fig. 3C). Subsequently, during the epithelial migration phase of dorsal closure in wild-type embryos, the more ventral lateral epidermal cells elongate dorsally as the epidermal sheets are coordinately pulled towards the dorsal midline (Fig. 3B, D). During this process, the leading edge of the epidermis maintains an organized appearance showing a continuous curve along the anterior-posterior axis (Fig. 3B). In contrast, the leading edge of the epidermis appears disorganized and the lateral epidermal cells fail to fully elongate in stage-15 Sec61α1 mutant embryos (Fig. 3E). The DME cells eventually revert to a non-elongated state (Fig. 3E, G). Finally, during the zippering phase of dorsal closure in wild-type embryos (stage 15), the DME cells meet at the dorsal midline and fuse with the contralateral leading edge cells starting at the anterior and posterior ends and gradually suturing the epidermis towards the center (Fig. 3D, F). In Sec61α1 mutant embryos, although there is some zippering observed at the anterior and posterior ends of the epidermis, the majority of the DME cells never come in close apposition and, therefore, cannot fuse. Subsequently, either the amnioserosa tears away or is degraded (presumably through apoptosis), allowing the underlying brain and guts to protrude through the epidermal hole (Fig. 3G). Although we noted some variability in the extent to which Sec61α1 mutant embryos completed the epithelial migration phase, we never observed mutant embryos that were completely closed (n > 500 mutant embryos observed). Taken together, these observations suggest that initiation occurs normally in Sec61α1 mutant embryos, but that epithelial migration is severely affected.

thumbnail image

Figure 3. Sec61α mutant animals initiate dorsal closure but fail during the epithelial migration phase. A–G: Confocal optical sections of w1118 (A, B, D, F) and Sec61α1 (C, E, G) embryos stained with antibodies against Coracle (Cor) to visualize epidermal cells during dorsal closure. Anterior is to the left in all cases and lateral or dorsal views are shown. A′–G′: Higher magnification views of the DME cells and lateral epidermis of the embryos shown in A–G. A, A′: Stage-13 wild-type embryo showing the initiation of dorsal closure. Note that the DME cells are elongated in the dorsal-ventral (D-V) axis (arrowheads in A′), whereas the lateral epidermal cells remain polygonal. B, B′: Stage-14 wild-type embryo showing D-V elongation of both the DME cells and the lateral epidermis. Note also the uniform curvature of the leading edge of the epithelium (arrowheads in B). C, C′: Stage-14 Sec61α1 mutant embryo showing elongated DME cells and mostly elongated lateral epidermal cells. The leading edge is less organized and maintains a scalloped appearance. D,D': Stage-15 wild-type embryo near the completion of dorsal closure. The DME cells are still elongated and the contraleteral DME cells are zippering together starting at the anterior and posterior ends. E, E′: Stage-15 Sec61α1 mutant embryo. The DME cells and lateral epidermis are not strongly elongated in the D-V axis, and the leading edge does not present a uniformly curved appearance. The amnioserosa is still intact in this embryo as the brain and guts have not broken through the dorsal surface. F, F′: Stage-16 wild-type embryo after the completion of dorsal closure. G, G′: Stage-16 Sec61α1 mutant embryo after dorsal closure has failed. In this case, all of the epidermal cells are polygonal, the leading edge of the epidermis is irregular, and the amnioserosa has clearly degraded or been torn away as the hindgut is protruding through the dorsal surface (arrow). Note also that head involution has occurred normally (arrowheads). H–K: Confocal optical sections of w1118 (H, J) and Sec61α1 (I, K) embryos stained with antibodies against Phospho-tyrosine (PY) to visualize the amnioserosa and epidermis during dorsal closure. The amnioserosa is intact in w1118 stage-14 (H) and stage-15 (J) embryos, as well as in the stage-14 Sec61α1 mutant embryo (I), but has torn away from the epidermis of the stage-15 Sec61α1 mutant embryo (arrows in K). Scale bars = 20 μm.

Download figure to PowerPoint

Since the guts protrude from the dorsal surface in late stage-16 Sec61α1 mutant embryos (Fig. 3G), we wanted to address when the amnioserosa was breaking down in these animals. We, therefore, stained wild-type and Sec61α1 mutant embryos with antibodies against phospho-tyrosine to visualize amnioserosal cells in addition to the epidermis. We found that the amnioserosa was completely intact through stage 14 in the mutant embryos, just as in wild-type (Fig. 3 H, I), but started to tear apart from the epidermis in stage-15 mutant embryos (Fig. 3K).

One of the hallmarks of the specification of the DME cells is the execution of a Jun amino terminal kinase (JNK) signaling pathway. At the end of this signaling pathway, Jun and Fos induce the expression of puckered (puc), which encodes a phosphatase that negatively regulates the JNK pathway (Ring and Martinez Arias,1993; Hou et al.,1997; Riesgo-Escovar and Hafen,1997b). We, therefore, examined embryos produced by Sec61α1/Cyo, Dfd-EYFP;pucE69/+ parents for the expression of puc (as determined by β-gal expression of the puc enhancer trap). In two independent experiments, we examined 285 stage-14–16 embryos that we could unambiguously identify as either Sec61α1 or one of the two EYFP genotypes (Sec61α1/Cyo, Dfd-EYFP or Cyo, Dfd-EYFP/Cyo, Dfd-EYFP). We could also distinguish each embryo as pucE69/ pucE69, pucE69/+, or +/+. The same percentage of Sec61α1 embryos expressed puc, as did their non-mutant siblings (67% in both cases, with a Mendelian expectation of 75%). When we examined individual Sec61α1;pucE69/+ embryos, we found that puc was clearly expressed in the DME cells. In fact, to the extent that we could tell by confocal microscopy, each DME cell expressed puc (Fig. 4A, B). Thus, loss of Sec61α did not appear to affect the JNK signaling pathway in the DME cells associated with the initiation of dorsal closure.

thumbnail image

Figure 4. JNK signaling occurs normally in the DME cells of Sec61α mutant embryos. Confocal optical sections of a Sec61α1/CyO, P{Dfd-YFP};pucE69/+ stage-14 embryo (A) and a Sec61α1;pucE69/+ stage-14 embryo (B) stained with antibodies against Cor (in green) to outline epidermal cells and against β-Gal to recognize pucE69 expression (red) in the DME cells. Note that each DME cell expresses puc in both mutant and heterozygous embryos. Anterior is to the left and dorsal is up in all these images.

Download figure to PowerPoint

Zygotic Loss of Sec61α Attenuates Dpp Signaling During Dorsal Closure

Another major consequence of JNK signaling in the DME cells is the expression and secretion of Dpp (Ring and Martinez Arias,1993; Hou et al.,1997; Riesgo-Escovar and Hafen,1997b). Dpp is a TGF-β/BMP ligand that binds to a receptor complex consisting of a Type-I receptor (encoded by thick veins, tkv) and a type-II receptor (encoded by punt) in the epidermal cells and in the amnioserosa (Affolter et al.,1994; Nellen et al.,1994; Simin et al.,1998). All three of these proteins go through the secretory pathway and thus must enter the endoplasmic reticulum through Sec61α. Mutations in tkv and put show similar dorsal closure phenotypes including initial elongation of the DME cells and subsequent failure to maintain a uniform migration front at the leading edge with non-elongated ventral epidermal cells (Riesgo-Escovar and Hafen,1997a; Simin et al.,1998; Ricos et al.,1999). These defects in the organization of the DME cells and inability to drive the epidermal migration in tkv mutant embryos result in part from a loss of an actinomyosin cable in the DME cells (Zahedi et al.,2008). We, therefore, examined the expression of the myosin regulatory light chain (encoded by spaghetti squash or sqh; Karess et al.,1991) in the DME cells in Sec61α1 mutant embryos. Notably, Sqh was absent or very strongly reduced at the leading edge of the DME cells of stage-14 mutant embryos compared to wild-type embryos (Fig. 5A, B). Sqh expression in the lateral epidermis was similar, however, between wild-type and Sec61α1 mutant embryos, indicating that this is a specific loss of the myosin cable in the DME cells of Sec61α1 mutant embryos.

thumbnail image

Figure 5. The actinomyosin cable is strongly reduced or eliminated in the DME cells of Sec61α mutant embryos, and Dpp signaling is attenuated. A, B: Confocal optical sections of stage-14 w1118 (A) and Sec61α1 (B) mutant embryos stained for antibodies against Spaghetti squash (Sqh, red) and Coracle (Cor, green). A′, B′: The Sqh channel alone. Cor staining outlines the epidermal cells and clearly shows the interface between the DME cells and the amnioserosa. Note the thick cable of Sqh running through the leading edge of the wild-type DME cells that is nearly absent in the Sec61α1 mutant embryo (arrowheads). C, D: Confocal optical sections of stage-14 w1118 (C) and Sec61α1 (D) mutant embryos stained for antibodies against phosphorylated Mad (pMad, red) and Coracle (Cor, green). C′, D′: The pMad channel alone. In the w1118 embryo, pMad is strongly expressed in the nuclei of lateral epidermal cells in a band that extends 4–5 cells ventral to the DME cells. In the Sec61α1 mutant embryo, pMad staining is strongly reduced.

Download figure to PowerPoint

In order to more directly determine whether lack of zygotic Sec61α expression affected Dpp signaling, we examined the levels of phosphorylated Mothers Against Dpp (Mad) in wild-type and Sec61α1 mutant embryos. The expression of phosphorylated Mad (pMad) has been extensively used as a readout of Dpp signaling at various stages of Drosophila development including during dorsal closure (e.g., Dorfman and Shilo,2001; Fernandez et al.,2007). In early stage-14 wild-type embryos, pMad is found at high levels in 4–5 rows of lateral epidermal cells ventral to the DME cells (Fig. 5C). In contrast, in four independent collections of Sec61α1 mutant embryos, we consistently observed strongly reduced or absent pMad staining in early stage-14 mutant embryos (Fig. 5D; n = 51 early stage-14 Sec61α1 mutant embryos).

If a mutation in Sec61α affects dorsal closure through the expression of Dpp or one of its receptors, we reasoned that it should be possible to rescue that defect by specifically activating the Dpp signaling pathway during mid-embryogenesis. We, therefore, used the UAS-GAL4 system (Brand and Perrimon, 1993) to express a constitutively activated form of the Tkv receptor (Nellen et al.,1996), tkvQ253D, in the epidermis (using e22c- and T80-Gal4; Jacinto et al.,2000) or in the amnioserosa (332.3-Gal4; Harden et al.,2002) of Sec61α1 mutant embryos. Since the expression of tkvQ253D did not rescue the cuticle defect of Sec61α1 mutant embryos (data not shown), we assessed the terminal phenotype of the mutant embryos by indirect immunofluorescence of 16–20-hr-old mutant embryos stained with Cor. In control experiments, we first determined that 100% (n = 206) of 16–20-hr AEL Sec61α1 mutant embryos robustly stain with Cor antibodies, whereas only 1% (n = 150) of Sec61α1/+ embryos stain well with Cor. We could, thus, identify the Sec61α1 mutant embryos in these experiments with near certainty. In addition, we performed control experiments to examine the phenotypes of Sec61α1,*-Gal4/Sec61α1 (where * indicates e22c or 332.3), and found that 100% of the embryos that stained with Cor at 16–20 hr AEL were completely dorsally open with the brains and guts extruded (Fig. 6A and data not shown). Experimentally, we crossed *-Gal4, Sec61α1/CyO flies to Sec61α1/+;UAS-tkvQ253D/+ flies. Complete rescue of dorsal closure would result in 50% of the Cor staining embryos displaying a completely open dorsal surface and 50% showing a closed dorsal surface. Animals expressing UAS-tkvQ253D using all three Gal4 lines showed nearly completely penetrant rescue (Table 2). The degree of rescue was slightly variable, ranging from animals possessing a small dorsal hole, but with the brain and guts completely contained within the embryo (Fig. 6B), to animals with the lateral epidermal sheets completely apposed at the dorsal midline, but not necessarily fused (Fig. 6C). In general, e22c-Gal4 produced a stronger degree of rescue than T80-Gal4 or 332.3-Gal4. We conducted an identical set of experiments using UAS-Dpp instead of UAS-tkvQ253D. We observed ∼20% of the expected rescue using e22c-Gal4 to drive the expression of Dpp, with less penetrant rescue resulting from expression driven by T80-Gal4 and 332.3-Gal4 (Table 2). The degree of rescue was similarly variable with the UAS-tkvQ253D experiments (data not shown). Taken together, these results strongly indicate that zygotic loss of Sec61α results in specific defects in dorsal closure through a reduction in Dpp signaling during mid-embryogenesis.

thumbnail image

Figure 6. An activated Thick veins receptor can partially rescue the dorsal closure defects of Sec61α1. A: Confocal optical section of a non-rescued, Sec61α1/e22c-Gal4, Sec61α1 stage-17 embryo. Note that the dorsal surface is completely open and the brain (asterisk) and guts (arrow) are extruded. Arrowheads (in A–C) mark the anterior and posterior extent of the open dorsal surface. B, C: Confocal optical sections of partially rescued, presumably Sec61α1/e22c-Gal4, Sec61α1;tkvQD/+ embryos, showing the range of phenotypes observed. The weakest rescued embryos (B) showed a small dorsal hole with the brain and guts confined to the interior. Stronger rescue (C) resulted in a nearly completely enclosed embryo, with only a small gap where the lateral epidermal sheets were apposed but had not fused.

Download figure to PowerPoint

Table 2. tkvQ253D Rescue of Sec61α1 Mutant Animals
Cross% Sec61α1 embryos showing DC rescue (n)a
  • a

    Progeny (16-20 hr (AEL)) of the indicated crosses were fixed and stained with antibodies against Coracle. Only Sec61α1 mutant animals stain at this stage. % DC rescue reflects the mean ± sd from three independent experiments. Note that maximal expected rescue is 50%. Numbers in parentheses represent the total number of embryos that stained with Cor.

e22c-Gal4, Sec61α1/CyOXSec61α1/+; UAS-tkvQ253D/+67 ± 6 (55)
T80-Gal4, Sec61α1/CyOXSec61α1/+; UAS-tkvQ253D/+49 ± 5 (63)
332.3-Gal4, Sec61α1/CyOXSec61α1/+; UAS-tkvQ253D/+45 ± 2 (329)
e22c-Gal4, Sec61α1/CyOXSec61α1/+; UAS-Dpp/+10 ± 3 (106)
T80-Gal4, Sec61α1/CyOXSec61α1/+; UAS-Dpp/+3 ± 3 (255)
332.3-Gal4, Sec61α1/CyOXSec61α1/+; UAS-Dpp/+5 ± 3 (105)

As a further test of the model that zygotic loss of Sec61α results in reduced Dpp signaling during mid-embryogenesis, we analyzed midgut morphogenesis in Sec61α1 mutant embryos since Dpp is known to be required for this process (Immergluck et al.,1990; Panganiban et al.,1990). During midgut morphogenesis, Dpp is secreted from the visceral mesoderm starting at stage 14 and is required in the underlying endoderm for the second midgut constriction. By stage 16, a four-chambered midgut is clearly apparent in wild-type animals (Fig. 7A). Consistent with a reduction in Dpp signaling, we never observed a four-chambered midgut in stage-16 Sec61α1 mutant embryos (n > 100 mutant embryos). Although the guts were sometimes more extremely affected, the most obvious defect observed was a lack of the second midgut constriction (Fig. 7B). Since Sec61α1 mutant embryos eventually display extruded guts due to defective dorsal closure, and we observed a developmental delay in these animals, we wanted to be sure that we were not simply observing mutant animals prior to the formation of this constriction. We, therefore, examined the midguts in early stage-17 Sec61α1 mutant embryos that were expressing tkvQ253D in the epidermis (using e22c-Gal4). Epidermal expression of TkvQ253D was sufficient to rescue dorsal closure to the point of preventing gut extrusion, but is not predicted to rescue any Dpp signaling defects in the mesoderm and underlying endoderm. We never observed a four-chambered midgut in these animals (Fig. 7C), consistent with the notion that Dpp signaling is particularly sensitive to the zygotic loss of Sec61α during mid-embryogenesis.

thumbnail image

Figure 7. Midgut morphogenesis is defective in Sec61α mutant embryos. Confocal optical sections of a stage-16 w1118 embryo (A), a stage-16 Sec61α1 mutant embryo (B), and a stage-17 Sec61α1/e22c-Gal4, Sec61α1;tkvQD/+ embryo (C) stained with antibodies against Fasciclin III (Fas III). Note the four chambers of the midgut in the w1118 embryo (asterisks) and the disorganized midguts in the Sec61α1 mutant embryos. The second midgut constriction is indicated by an arrow in A and is missing in Sec61α1 mutant embryos.

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  9. Supporting Information

Zygotic Loss of Sec61α Has a Specific Effect on Dorsal Closure

Sec61α encodes the major subunit of the translocon complex in the endoplasmic reticulum (reviewed in Johnson and van Waes,1999). The translocon complex consists of three subunits: Sec61α, β, and γ, with the 10-transmembrane domain α subunit forming the major pore through the membrane. The truncated product of the Sec61α1 mutation, if expressed at all, would almost certainly produce a nonfunctional protein, and therefore prevent proteins from entering the ER and secretory pathway. The identical phenotypes observed in Sec61α1, Sec61α1/Df(2L)BSC6, and even Df(2L)BSC6 homozygous mutant embryos demonstrate that Sec61α1 is a nonfunctional, amorphic allele.

Sec61α is produced maternally and the maternal complement is likely partially functional at least through mid-embryogenesis. The Berkeley Drosophila Genome Project gene expression pattern database shows that Sec61α is strongly expressed in stage-1–3 embryos (Tomancak et al.,2002). In addition, a Sec61α:GFP protein trap transgenic line is expressed in germline cysts (Snapp et al.,2004). Evidence that the maternal contribution still has function in mid-embryogenesis includes direct and indirect measures of protein expression in Sec61α1 mutant embryos (that have a complete zygotic loss of Sec61α). For example, the transmembrane protein Neuroglian is expressed in Sec61α1 mutant embryos at levels indistinguishable to those of wild-type embryos that were fixed and stained in parallel and imaged using the same parameters (Supp. Fig. S1). We similarly observed only a slight reduction in the expression of the transmembrane proteins E-cadherin and α-integrin in mutant embryos as compared to wild-type (data not shown). Since these proteins could have been secreted hours previously, however, we wanted to assess a secreted protein that is not expressed until later in development. We can observe this indirectly through the function of Serpentine and Vermiform, two putative chitin deacetylases that are first synthesized in stage-12 embryos and subsequently secreted into the tracheal lumen at stage 14. These enzymes are required for tube length control in tracheae (Luschnig et al.,2006; Wang et al.,2006), and although there is reduced chitin secreted into the tracheae of Sec61α1 mutant embryos, tracheal length is not affected, suggesting that these enzymes were secreted and functioned properly (Supp. Fig. S1). Taken together, these observations suggest that the maternal contribution of Sec61α still has some function (and perhaps fairly good function) at mid-embryogenesis. By the end of embryogenesis, however, secretion appears to be more strongly affected as the mutant embryos fail to secrete chitin and thus display a naked cuticle phenotype.

At mid-embryogenesis, the zygotic loss of Sec61α produces a highly specific and fully penetrant defect in dorsal closure. This phenotype is observed in Sec61α1, Sec61α1/Df(2L)BSC6, and even Df(2L)BSC6 mutant animals. We were intrigued by the fact that although dorsal closure was completely disrupted in these embryos, head involution occurred normally, and therefore examined other developmental processes occurring roughly concurrently with dorsal closure that required secreted or transmembrane proteins for proper execution. For example, tracheal patterning and morphogenesis are dependent on several signaling pathways, most notably FGF signaling, which requires the transmembrane receptor Breathless in the tracheal cells to receive a secreted FGF signal encoded by branchless (reviewed in Affolter and Caussinus,2008). Although the early events of patterning occur prior to dorsal closure, the formation and extension of the dorsal branch occur during stages 12–15 of embryogenesis and require both the aforementioned FGF signaling pathway as well as Notch-Delta signaling to inhibit the cells ventral to the dorsal tip from becoming migratory. In addition, connecting the adjacent metamers to form a complete tracheal network requires cell adhesion and specialized vesicle trafficking in the fusion cells that occurs during stages 14–16 (Jiang et al.,2007). Although one or two of metameres in about half of the embryos are not completely fused, the vast majority of these fusion events occurs normally, and thus it appears that tracheal morphogenesis is largely occurring normally in Sec61α1 mutant embryos (Supp. Fig. 1 and data not shown). Similarly, the migration of the neurons and glia of the peripheral nervous system require guidance from secreted ligands and transmembrane receptors (reviewed in Cooper,2002). We found that the peripheral neurons are correctly positioned in the Sec61α1 mutant embryos, again suggesting that the secretion of all these factors had occurred normally. Finally, the salivary gland is specified during stage 11 of embryogenesis, and invaginates and executes a posterior migration over the visceral mesoderm during stage 14 (reviewed in Myat,2005). This process requires the apposition of the salivary gland epithelium with the visceral mesoderm in a process requiring integrins. As shown in Supp. Figure S1, posterior migration of the salivary gland occurred normally in Sec61α1 mutant embryos. Thus, of all the morphogenetic events occurring simultaneously during mid-embryogenesis, zygotic loss of Sec61α specifically perturbs dorsal closure.

It should be noted that zygotic loss of Sec61β does not show a similar defect in dorsal closure, but rather only shows defects in cuticle secretion late in embryogenesis (Valcarcel et al.,1999). Maternal loss of Sec61β, however, results in earlier embryonic lethality with defects in D-V patterning resulting from improper Gurken signaling (Valcarcel et al.,1999; Kelkar and Dobberstein,2009). Sec61β encodes a peripheral component of the translocon complex that is not essential for protein translocation across the ER or for cell viability in yeast (Toikkanen et al.,1996; Leroux and Rokeach,2008). Recent work in yeast has rather demonstrated a unique role for Sec61β in post-ER trafficking to the plasma membrane via the exocyst (Toikkanen et al.,2003). In support of this, Kelkar and Dobberstein (2009) showed that Gurken is normally inserted into the ER in Sec61β germ line clones in Drosophila, but fails to traffic to the plasma membrane. Together these observations suggest that Sec61β does not have a strong effect on protein translocation into the ER, and thus it is not surprising that Sec61β does not show identical phenotypes to Sec61α.

Sec61α Dorsal Closure Defects Result From Attenuated Dpp Signaling

Dorsal closure is regulated by the complex interplay of several well-characterized signaling pathways including wingless, Notch, JNK, and Dpp (Harden,2002), most of which have components that are transmembrane or secreted (all of the JNK signaling proteins are cytoplasmic, although it has often been assumed that there is a membrane receptor necessary to initiate the cascade). The terminal phenotype of Sec61α1 and Sec61α1/Df(2L) BSC6 mutant embryos is informative as the dorsal surface is completely exposed with extruded brain and guts, whereas the head involutes correctly. Mutations in JNK signaling component genes show similarly strong defects in dorsal closure, but additionally fail in head involution producing terminal phenotypes that are fully exposed from near the posterior pole through the head (see, for example, Riesgo-Escovar and Hafen,1997b). Thus, any potential secreted upstream signals in the JNK signaling pathway appear to be unaffected by zygotic loss of Sec61α. Closer examination of the events of dorsal closure support this notion. The DME cells in Sec61α1 mutant embryos elongate in the D-V axis indicating that these cells had been correctly specified, a process requiring JNK signaling. Most telling, puc is expressed normally in the DME cells of mutant embryos (Fig. 4). Although observing the β-gal signal from the pucE69 enhancer trap line does not allow for a quantitative measure of the strength of JNK signaling, it appeared that JNK signaling is active in every DME cell in Sec61α1 mutant embryos (Fig. 4B).

Our results strongly suggest that zygotic loss of Sec61α attenuates either the secretion of Dpp from the DME cells, or its reception in the epidermis and/or amnioserosa. Four pieces of evidence support this conclusion. First, the phenotypes associated with loss of Sec61α overlap those of mutations in the Dpp receptors tkv and punt. In all three mutations, the DME cells elongate in the D-V axis normally, whereas the lateral epidermal cells only partially elongate (Fig. 3; Riesgo-Escovar and Hafen,1997a; Simin et al.,1998; Ricos et al.,1999). In addition, during the epithelial migration phase of dorsal closure the leading edge is irregular in both tkv and Sec61α mutant embryos (Fig. 3; Fernandez et al.,2007). Furthermore, the amnioserosa tears away from the epidermis in stage-15 Sec61α1 mutant embryos, just as it does in tkv mutant embryos (Fig. 3; Fernandez et al.,2007). Finally, the terminal phenotype observed in strong loss of function mutations in Sec61α, tkv, and punt is a large square dorsal hole accompanied by normal head involution (Fig. 2; Affolter et al.,1994; Nellen et al.,1994; Arora et al.,1995). The second piece of evidence in support of a loss of Dpp signaling in Sec61α mutant embryos is a strong reduction in the actinomyosin cable in the DME cells. Dpp regulates the expression of zipper (encoding nonmuscle myosin) in the DME cells (Arquier et al.,2001). We observed a strong and specific reduction in the levels of the myosin regulatory light chain (Sqh) in the DME cells, although we could clearly observe wild-type levels of Sqh in other epidermal cells in the Sec61α1 mutant (Fig. 5). The third piece of evidence in support of reduced Dpp signaling is the elimination or strong reduction of pMad staining in the lateral epidermis in early stage-14 Sec61α mutant embryos (Fig. 5). The final and strongest piece of evidence implicating loss of Dpp signaling in Sec61α1 mutant embryos was the highly penetrant and very strong rescue of dorsal closure by tkvQ253D (Table 2 and Fig. 6), and the less penetrant, yet equally strong, rescue of Sec61α1 mutant embryos by epidermal expression of Dpp (Table 2).

It is noteworthy that although the penetrance of the rescue was similar if tkvQ253D was expressed in the epidermis (e22c-Gal4 or T80-Gal4) or the amnioserosa (332.3-Gal4), the overall strength of the rescue (degree of closure) was more pronounced with the epidermal expression. These results are consistent with the observations of Fernandez et al. (2007) that the expression of tkvQ253D in the epidermis gave stronger rescue of the dorsal closure defects in strong loss-of-function tkv mutants (in terms of penetrance in this case) than expression in the amnioserosa. Our results further support their conclusion that Dpp signaling is required in both the epidermis and amnioserosa for normal dorsal closure. Finally, it is not surprising that the rescue of Sec61α1 by tkvQ253D was not complete since the TkvQ253D protein would have had to go through the secretory pathway, and thus may not have been expressed at comparable levels to the experiments aimed at rescuing a tkv mutation (Fernandez et al.,2007).

Our results also suggest that Dpp signaling is attenuated but likely not eliminated by zygotic loss of Sec61α. This notion is supported by the observation that although the phenotypes are similar between Sec61α1 and tkv mutants, there are minor differences. Specifically, in tkv mutant embryos, the DME cells bunch together at segmental boundaries (Ricos et al.,1999), and although we observed this phenotype in Sec61α1 embryos, the penetrance was substantially reduced (data not shown). In addition, although we found strong reductions in the actinomyosin cable in the leading edge of Sec61α1 mutant embryos, we still could observe a faint, often discontinuous, cable linking DME cells (Fig. 5). We also occasionally observed pMad staining in the lateral epidermis of Sec61α1 mutant embryos, although it was always clearly attenuated compared to wild-type embryos (Fig. 5).

Taken together, these results suggest a model in which zygotic loss of Sec61α gradually reduces traffic through the secretory pathway as the maternal stores are turned over, eventually resulting in a situation in which the expression of a dose-sensitive secreted protein falls below a critical threshold necessary to support dorsal closure. Our results indicate that the protein in question functions in the Dpp signaling pathway, and thus is most likely Dpp or one of its receptors. The observation of defective midgut constrictions in Sec61α mutant animals (Fig. 7) further supports the notion that Dpp signaling is specifically sensitive to loss of Sec61α at this stage of development, since Dpp signaling during mid-embryogenesis is necessary for the formation of the second midgut constriction (Immergluck et al.,1990; Panganiban et al.,1990). Our experiments, however, could not definitively distinguish whether the dose-sensitive factor was Dpp itself or one of its receptors, but the tkvQ253D rescue experiments argue that it has to be at the level of one of the receptors or upstream in the pathway. In addition, the rescue of dorsal closure in Sec61α embryos by epidermal expression of Dpp indicates that there is not a complete absence of Tkv or Punt at the plasma membrane. We suggest that Dpp is the most likely candidate, as dpp is one of the few haploinsufficient genes encoded by the Drosophila genome (Spencer et al.,1982). dpp expression must be finely controlled in early embryogenesis due to its role as a morphogen for initial D-V patterning of the embryo (reviewed in O'Connor et al.,2006). Perhaps the dose-sensitivity of Dpp or its receptors during dorsal closure is a consequence of this earlier requirement to maintain precise levels of Dpp, although it is also possible that Dpp signaling during dorsal closure requires similar levels of tight control to precisely coordinate the cellular behaviors of at least two distinct tissues, the lateral epidermis and the amnioserosa.

EXPERIMENTAL PROCEDURES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  9. Supporting Information

Drosophila Strains

Sec61α1 is an EMS-induced mutation on the E(br)165 chromosome reported in Ward et al. (2003). P{lacW}Sec61αk04917, Df(2L)BSC6, 332.3-Gal4, e22c-Gal4, T80-Gal4, UAS-Dpp, pucE69, and w1118 were obtained from the Bloomington Drosophila Stock Center (Bloomington, IN). UAS-TkvQ253D (Nellen et al.,1996) was obtained from Matt Gibson (Stowers Institute, Kansas City, MO). w1118 was used as the wild-type strain for all the experiments reported here. All Drosophila stocks were maintained on media consisting of corn meal, sugar, yeast, and agar in incubators maintained at a constant temperature of 21°C or in a room that typically fluctuated between 21° and 22.5°C. Genetic experiments were conducted in incubators controlled at a constant temperature of 25°C.

Molecular Analysis of Sec61α1

Complementation analyses confirmed a lethal mutation in the region uncovered by Df(2L)BSC6 (breakpoints 26D3; 26F7) that also failed to complement the P-element allele P{lacW}Sec61αk04917. Sequencing of genomic DNA isolated from homozygous mutant late embryos (by the DNA Facility of the Iowa State University Office of Biotechnology in Ames, IA) confirmed that the chromosome had a mutation in Sec61α. Both DNA strands were sequenced to confirm this result.

To generate a Sec61α genomic rescue construct, we PCR amplified genomic DNA isolated from w1118 embryos using the following primers: 5′-TCGGCACTAGCACTGAATACT-3′ and 5′-ATAGCAGGTAGCCGAACGTG-3′. This generated a 4.4-kb genomic fragment containing all of Sec61α and extending ∼100 bp into the upstream gene DLP and ∼350 bp into the downstream gene CG9536 (note that these genes are oriented in opposite directions, and thus the genomic rescue construct only extends into the 5′ UTR of each gene). We blunt-end ligated this fragment into a StuI-cut pCasper 4 vector (Thummel et al.,1988). We checked orientation by restriction digest and sequenced the Sec61α portion of the plasmid (Iowa State University DNA Facility). The plasmid was injected into w1118 embryos at the Duke University Model System Genomics facility (Duke University, Durham, NC). We crossed four independent third chromosome inserts into our Sec61α1 stock, and all four rescued the embryonic lethality of Sec61α1.

Lethal Phase and Phenotypic Analyses

Sec61α1, P{lacW}Sec61αk04917, and Df (2L)BSC6 were balanced with CyO, P{w+, Dfd-EYFP} (Le et al.,2006) in order that mutant embryos and larvae could be unambiguously identified by the absence of YFP. In order to identify mutant animals at earlier stages of development, we also balanced Sec61α1 with CyO, P{ftz/lacB}. Embryos were collected for 4 hr at 25°C, aged for 12 hr and then selected based upon the absence of YFP. Embryonic lethality was determined as the percentage of unhatched embryos 48 hr after selecting non-YFP-expressing mutant embryos produced through a 4-hr egg collection. Larval lethality was determined as the percentage of non-pupariating mutant larvae 7 days after selecting newly hatched mutant larvae. Pupal lethality was determined as the percentage of non-eclosing mutant pupae ∼7 days after pupariating. All experiments were performed in triplicate and means with standard deviations were determined. Non-hatched embryos were dechorionated in 6% sodium hypochloride, mounted on microscope slides in Hoyer's medium, and subsequently examined for cuticular phenotypes on a Nikon Eclipse 80i compound microscope. All cuticular phenotypes were documented on a Nikon Eclipse 80i compound microscope equipped with a Photometrics CoolSNAP ES high-performance digital CCD camera. Photomicrographs were cropped and adjusted for brightness and contrast with Adobe Photoshop (version CS3, San Jose, CA), and figures were compiled in Adobe Illustrator (version CS3, San Jose, CA).

Time-lapse movies were generated by mounting stage-17 Sec61α1/CyO, P{w+, Dfd-EYFP} and Sec61α1 mutant embryos on apple juice plates. We first collected a fluorescence image to distinguish the mutant and heterozygous embryos by YFP expression, and then automatically collected brightfield images every minute for 140 min using MetaMorph (Molecular Devices, Sunnyvale, CA) software on a Nikon Eclipse 80i compound microscope equipped with a Photometrics CoolSNAP ES high-performance digital CCD camera using a 4× Nikon Plan Apo lens.

Antibody Production and Immunostaining

A Spaghetti squash (Sqh)-GST fusion protein was generated by blunt-end ligating nucleotides 217 to 811 of GenBank nucleotide sequence AY122159 (generated by PCR and verified by sequence analysis) into a PGEX2TK vector (GE Healthcare Bio-Sciences AB, Uppsala, Sweden) digested with SmaI. This fragment represents the entire coding sequence of Sqh. The protein was overexpressed in Escherichia coli BL21 (DE3) cells, and purified on glutathione sepharose 4 fast flow (GE Healthcare) according to standard procedures (Rebay and Fehon,2000). Purified protein was used for antibody generation in mice at the Pocono Rabbit Farm and Laboratory Inc. (PRF&L, Canadensis, PA).

Embryos were fixed and processed for antibody staining as described (Fehon et al.,1991). Embryonic staging was determined by gut morphology. The following primary antibodies were used at the given dilutions: mouse anti-Neuroglian 1:100 (clone BP104 from the Developmental Studies Hybridoma Bank at the University of Iowa, Iowa City, IA), mouse anti-β-galactosidase 1:100 (clone 40-1a, DSHB), mouse anti-Fasciclin III 1:100 (clone 7G10, DSHB), mouse anti-phosphoMAD 1:250 (from P. ten Dijke), mouse anti-Sqh 1:1,000, guinea pig anti-Uninflatable 1:20 (Zhang and Ward,2009), rabbit anti-GFP 1:250 (Clontech, Mountain View, CA), mouse anti-Coracle 1:400 and guinea pig anti-Coracle 1:10,000 (gifts from Richard Fehon, University of Chicago; Fehon et al.,1994), and FITC conjugated monoclonal anti-phosphotyrosine (Sigma, St. Louis, MO) 1:50. Secondary antibodies were obtained from Jackson ImmunoResearch Laboratories (West Grove, PA) and used at 1:800. Alexa Fluor 555 conjugated wheat germ agglutinin (WGA) was used at a concentration of 5 μg/ml (Invitrogen, Carlsbad, CA). Rhodamine conjugated Chitin Binding Probe (CBP) was used at 1:500 according to the manufacturer's recommendation (New England Biolabs Inc., Beverly, MA). Confocal images were acquired on a Zeiss LSM510 Meta Laser Scanning Confocal Microscope (Carl Zeiss Inc., Thornwood, NY). Photomicrographs were cropped and adjusted for brightness and contrast with Adobe Photoshop, and figures were compiled in Adobe Illustrator. The photomicrographs used to compile Figure 4 were additionally modified with an unsharp mask filter (50%, 3 pixel radius, 5 level threshold) in Adobe Photoshop.

Dpp and tkvQ253D Rescue Experiments

We first individually recombined T80-Gal4, e22c-Gal4, and 332.2-Gal4 onto Sec61α1 chromosomes, and balanced then over CyO, P{w+, Dfd-EYFP}, creating *-Gal4, Sec61α1/CyO, P{w+, Dfd-EYFP}. We then crossed UAS-tkvQ253D and UAS-Dpp into Sec61α1/CyO, P{w+, Dfd-EYFP}, creating Sec61α1/+; UAS-tkvQ253D/+ and Sec61α1/+; UAS-Dpp/+. Finally, we crossed these Gal4 and UAS-tkvQ253D lines together and collected embryos from a 4-hr collection. We aged the embryos for 16 hr such that all the embryos would be 16–20 hr after egg laying. At this point, all Sec61α1/+ embryos would have completed dorsal closure and secreted cuticle that would prevent the embryos from being stained with antibodies. We fixed and stained the embryos as described above, and examined all stained embryos for dorsal open phenotypes. We considered an embryo to have been rescued if the dorsal hole was small and the brain and guts were completely confined to the interior of the animal. Since only half of the Sec61α1 mutant embryos would contain either UAS-tkvQ253D/+ or UAS-Dpp/+, the % rescue reported in Table 2 would have a theoretical maximum value of 50%. The experiment was done in triplicate and mean with standard deviations were calculated.

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  9. Supporting Information

We thank Elspeth Pearce for help with construction of the Sec61α rescue construct, and members of the lab and anonymous reviewers for helpful comments about the manuscript. We thank Greg Beitel, Matt Gibson, and the Bloomington Drosophila Stock Center for fly stocks used in this study. We thank Richard Fehon, Rebecca Hays, and the developmental studies hybridoma bank for the antibodies used in this study. The project was supported by NIH Grant Number P20 RR15563 and NIH Grant Number R01HD047570 from the National Center for Research Resources.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  9. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  9. Supporting Information

Additional Supporting Information may be found in the online version of this article.

FilenameFormatSizeDescription
DVDY_22219_sm_suppinfofigS1.tif3196KSupp. Fig. S1. Most mid-embryonic morphogenetic events occur normally in Sec61α mutant embryos. Confocal optical sections of w1118 (A, C, E) and Sec61α1 mutant (B, D, F) embryos stained with antibodies against Neuroglian (A, B), Coracle (C, D), or Uninflatable (E, F). In all cases, anterior is to the left. Wild-type and mutant animals for each experiment were fixed and stained in parallel and imaged using identical settings. A, B: The peripheral nervous system (a segmentally repeated unit is indicated by the arrow) is normally patterned in stage-14 Sec61α1 mutant embryos. C, D: Salivary gland morphogenesis (dorsal and posterior migration of the gland) has occurred normally in stage-14 Sec61α1 mutant embryos as judged by Cor staining to visualize the gland. The salivary gland (arrow) occupies its normal position and has fully elongated towards the posterior. Note also that Cor is correctly localized to the apical lateral membrane (the region of the septate junction) in the mutant embryos. E, F: Tracheal patterning is normal in stage-15 Sec61α1 mutant embryos, but the diameter of the tracheal tubes is smaller. The dorsal trunk is indicated by an arrow. The dorsal branches are notably present in Sec61α1 mutant embryos (indicated by arrowheads).

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.