We have identified a novel a integrin subunit in Drosophila, that associates with bPS integrin. We report the temporal expression of the gene encoding this integrin subunit, which we have called aPS3, throughout development and the localization of its expression during embryogenesis. aPS3 RNA was localized to tissues undergoing invagination, tissue movement and morphogenesis such as salivary gland, trachea, midgut, dorsal vessel, midline of the ventral nerve cord, amnioserosa and the amnioproctodeal invagination. aPS3 DNA localized to the chromosomal vicinity of scab (scb), previously identified by a failure of dorsal closure. Embryos homozygous for the l19 allele of scb had no detectable aPS3 RNA and the 1035 allele of scb contains a P element inserted just 5¢ of the coding region for the shorter of the gene’s two transcripts. Furthermore, mutations in the scb locus exhibit additional defects corresponding to sites of aPS3 transcription, including abnormal salivary glands, mislocalization of the pericardial cells and interrupted trachea. Removal of both maternal and zygotic bPS produced similar defects, indicating that these two integrin subunits associate in vivo and function in the movement and morphogenesis of tissues during development in Drosophila. Phenotypic similarities suggest that laminin A is a potential ligand for this integrin, at least in some tissues.

Integrins are heterodimeric transmembrane receptors for the extracellular matrix or for surface molecules on other cells (Hynes, 1992). Through their cytoplasmic domains, they can interact with cytoskeletal elements and cause changes in cell shape in response to ligand binding. Their emerging role in signaling expands the repertoire of cellular responses that may be triggered by integrins (Damsky and Werb, 1992; Adams and Watt, 1993; Clark and Brugge, 1995; Schwartz et al., 1995). This ability to ‘read’ the external environment and produce changes in cell shape or to initiate signaling cascades make these molecules of particular interest in the study of developmental mechanisms.

In Drosophila, members of this family have been identified (reviews, Brown, 1993; Gotwals et al., 1994). The known integrin molecules in flies are fewer than those known in vertebrates (8 β subunits and 16 α subunits). Only two β integrin subunits have been identified in flies, β PS and βν. Two α integrin subunits, α PS1 and α PS2 are known, both of which associate with β PS; these three molecules have been designated the PS integrins.

The study of integrins in Drosophila has led to an understanding of some of the functions of these molecules during development. Previous work has focused primarily on the roles of the PS integrin molecules in adhesive events during development, for example: the attachment of embryonic muscle to epidermis, the attachment of photoreceptors to the basement membrane of the retina and the maintenance of close apposition of wing surfaces (Wright, 1960; Brower and Jaffe, 1989; Leptin et al., 1989; Wilcox et al., 1989; Zusman et al., 1990, 1993; Brabant and Brower, 1993).

Some of the phenotypes observed in embryos mutant for β PS are not observed in embryos null for either of its two known associated α subunits, or in embryos doubly mutant for these genes (Brower et al., 1995; Roote and Zusman, 1995). It therefore seems likely that β PS has additional partners.

The identification of new α integrins by methods that rely upon sequence similarity such as polymerase chain reaction or low stringency hybridizations is challenging because of the high degree of sequence divergence. We have therefore exploited the heterodimeric nature of integrin receptors to screen for molecules that interact with β PS using co-immunoprecipitation, and we report herein the identification and characterization of a novel α integrin subunit, which we have designated α PS3.

We find that α PS3 is transcribed in tissues undergoing morphogenetic movements. The gene encoding α PS3 localizes to the site of the previously described scab (scb) locus (Nusslein Volhard et al., 1984). No detectable α PS3 transcript is found in late embryos homozygous for the l19 allele of scb. The 1035 allele of scb has a P element inserted just 5′ of the coding region for the shorter of two transcripts of α PS3. We find that mutations in scb, originally identified by a failure of dorsal closure, cause additional defects that occur at sites of α PS3 RNA expression, and we see similar defects in embryos null for β PS. Some of these defects also occur in embryos mutant for laminin A, suggesting a possible ligand interaction.

Surface labeling and immunoprecipitation

Schneider L2 cells were maintained at room temperature in Schneider’s M13 medium (Sigma) with 5% fetal bovine serum (Hazelton). Embryos were prepared as described (Leptin et al., 1987). Surface labeling was performed using 125I and lactoperoxidase (Hynes, 1973) and cells were lysed in buffer consisting of 50 mM Tris-HCl, pH 8.0, 150 mM NaCl, 1 mM CaCl2, 1 mM MgCl2, 0.5% NP-40, and 2 mM phenylmethanesulfonylfluoride (PMSF). Samples were pre-cleared by spinning for 10 minutes at 10,000 g. For immunoprecipitation, supernatant was incubated with #185 rabbit antiserum raised against a peptide comprising the 21 amino acid carboxyl terminus from β PS, rotated overnight at 4°C with Protein-A Sepharose, washed and resuspended in gel sample buffer. In some experiments, the samples were denatured prior to immunoprecipitation by the addition of 1/20 volume of 20% SDS, boiling and the addition of Triton X-100 to 5%. In some experiments, NaCl at 0.25, 0.5 or 1 M, EDTA or EGTA at 10 mM, and GRGDSP or GRGESP peptides at 1 mg/ml were added prior to addition of antibody.

Samples were prepared in buffer (4% SDS, 100 mM Tris HCl, pH 6.8, 10 mM EDTA, 10% glycerol and bromphenol blue) with or without 50 mM dithiothreitol (DTT), boiled, separated on 7% SDSPAGE gels (Laemmli, 1970) and subjected to autoradiography.

For two-dimensional gels, samples were first run under non-reducing conditions. The lane was excised and subjected to reducing SDS-PAGE in the second dimension and autoradiography.

Characterization of protein interactions

Sucrose density gradients (3-25%) were performed as described (Zhou and Standring, 1992). Fractions were analyzed by immunoprecipitation with 185 antibody and SDS-PAGE.

To assay lectin binding, immunoprecipitates were washed and the protein A-Sepharose/antibody/antigen complexes incubated in 50 mM Tris-HCl, pH 8.0, 150 mM NaCl, 1 mM CaCl2, 0.5% NP-40, 10 mM EDTA to elute the 90 kDa protein. Supernatant was brought to 15 mM CaCl2 and MgCl2, and incubated with either lentil lectin or Concanavalin A complexed to Sepharose 4B (Phamacia). After incubation, non-bound supernatant was collected and the beads resuspended in gel sample buffer after washing. Both fractions were analyzed by electrophoresis.

Protein purification and peptide sequencing

Preparation of a 185 antibody column: a 10 ml aliquot of 185 serum was dialyzed against 10 mM phosphate buffer, pH 7.2 and passed through a DEAE column (Whatman DE 52) in dialysis buffer (Catty and Raykundalia, 1988). IgG-containing flow-through fractions were collected and coupled to CNBr-activated Sepharose 4B at a concentration of 10 mg/ml resin following manufacturer’s instructions.

Confluent Schneider L2 cells were lysed in 50 mM Tris-HCl, pH 8.0, 150 mM NaCl, 1 mM CaCl2, 1 mM MgCl2, 0.5% NP-40, 2 mM PMSF then centrifuged at 12,000 g. The supernatant was precleared on a Sepharose 4B column and then passed onto the 185 antibody column overnight. The column was washed and eluted with 3 volumes of lysis buffer containing 10 mM EDTA. Pooled eluates were brought to 15 mM each of CaCl2 and MgCl2 and incubated with lentil lectin Sepharose 4B (Pharmacia). The lentil lectin beads were washed, resuspended in reducing gel sample buffer and subjected to SDSPAGE. The separated proteins were electroblotted onto either PVDFP (Millipore) or nitrocellulose membranes and stained with Coomassie blue (PVDF) or Ponceau S (nitrocellulose). The 90 kDa band (∼30 pmol) was excised from PVDF membrane for N-terminal sequence determination, while protein on nitrocellulose (>50 pmol) was used for tryptic digestion and internal peptide sequence determination. Tryptic digests, peptide purification and peptide sequencing were performed by the MIT/CCR/HHMI Biopolymers Laboratory.

PCR, cDNA cloning and sequencing

cDNA was isolated by polymerase chain reaction (PCR) using degenerate oligonucleotide primers based on reverse translation of the tryptic peptide sequences. A few inosines were included to reduce redundancy. The 64-fold redundant N-terminal primer GATCGAATTCAAT/CGAT/CTAT/CACICCIGAA/GCAT/CTAT/CGC based upon amino acids NDYTPEHYA, and the 16-fold redundant reverse primer GATCAAGCTTC/TTGC/TTCICCA/GAAA/GTTIGCA/GTT based upon NANFGEQ, contained EcoRI and HindIII sites, respectively, to facilitate cloning. Drosophila 0-24 hour embryonic cDNA was used as template. Reactions were run on a Perkin-Elmer Cetus DNA Thermal Cycler. 40 cycles of denaturation at 94°C for 1 minute, annealing at 42°C for 2 minutes 30 seconds and extension at 72°C for 4 minutes 15 seconds were followed by final extension at 72°C for 8 minutes. The resultant product was digested with EcoRI and HindIII, subcloned into Bluescript SK (Stratagene) and designated BS-PCR.

cDNA clones were obtained by screening 3.5× 105 plaques from a λ gt11 size-selected, EcoRI-linkered, cDNA library (Zinn et al., 1988) from 9-12 hour embryos with 32P-labelled BS-PCR. After rescreening, positive plaques were selected, DNA was extracted from plate lysates (Sambrook et al., 1989), subcloned into the EcoRI site of Blue-script SK (Stratagene) and transformed into E. coli strain DH5α.

Sequencing of the ends of BS-PCR and the cDNA subclones was performed by chain termination (Sanger et al., 1988) using Sequenase (United States Biochemical Corporation). Sequences were analyzed using the GCG sequence analysis software package version 7.2 (1993). A similarity tree was plotted comparing peptide sequence of the long splice variant of α PS3 with representatives for each α using the Pileup function of GCG. Database accession numbers used were: α 2, M28249; α 1, X68742; α X, Y00093; α M, M18044; α L, Y00796; α V, P06756; α 8, P26009; α 5, P08648; α IIb, A34269; C. elegans F54F2.1, P34446; α PS2, P12080; α 9, L24158; α 4, S06046; α 6, X53586; α 7, L23423; α 3, M59911; α PS1, X73975; C. elegans F54G8.3, Q03600.

Chromosome in situ hybridizations

BS-PCR was linearized with HincII and labeled with biotinylated dUTP (Boehringer Mannheim) by random priming. Salivary glands from Canton-Special climbing third instar larvae were prepared and hybridized according to standard methods (Pardue, 1986), incubated with horseradish peroxidase (HRP) conjugated to streptavidin using a DETEK-HRP kit (Enzo Diagnostics). HRP was visualized using Sigmafast DAB tablets and squashes counterstained with orcein or Giemsa. The polytene band was identified by reference to maps (Lindsley and Zimm, 1992).

Northern blot analysis

Wild-type RNAs were prepared from timed egg lays, staged larvae and adults of the Canton-Special strain according to standard methods (Ayme and Tissieres, 1985). Mutant embryos were picked on the basis of their dorsal closure phenotype and their RNA extracted (Ashburner, 1989). Following agarose gel electrophoresis, RNA was transferred to Biotrans membranes (ICN), using standard methods (Sambrook et al., 1989).

The complete α probe consisted of the 5′-most HindIII fragment of cDNA 11 ligated to a fragment obtained by partial digestion of cDNA 10, containing all of the remaining open reading frame. This subcloned insert was digested with HindIII to give fragments of 500 and three of ∼1000 base pairs that were labeled by random priming with 32P-dCTP. The 5′-most EcoRI to XhoI fragment of cDNA 11, which contains no open reading frame, was used as the 5′ probe. A 3′ untranslated probe was obtained by HincII digestion of cDNA 9. A probe for the alternate 5′ coding region was created by PCR from adult cDNA. A subcloned 1.1 kb PstI fragment from cDNA 10 was also used, as well as BS-PCR. For quantitation, a control probe was made from rp49 DNA (a ribosomal protein gene).

Filters were prehybridized for 1 hour at 37°C, followed by overnight hybridization in Garber’s hybridization solution (0.75 M NaCl, 0.15 M trizma base, 10 mM EDTA, 10 mM NaOH, 0.1325 M HCl, 50 mM sodium phosphate pH 6.5, 7% dextran sulfate, 0.5% SDS, 10× Denhardt’s solution, 50 μg/ml denatured salmon sperm DNA and 50% formamide) at 42°C. Membranes were washed four times in 5% SDS, 30 mM NaCl, 40 mM Trizma base, 2 mM EDTA, 2 mM NaOH, 3 mM HCl at 65°C and subjected to autoradiography.

Whole-mount in situ hybridizations and plastic sectioning

Embryos were collected on apple juice agar plates from cages of Canton-Special flies and staged (Campos-Ortega and Hartenstein, 1985). Whole-mount in situ hybridizations were performed essentially according to the methods of Tautz and Pfeifle (1989) using the Genius kit (Boehringer Mannheim). Proteinase K digestion was varied; longer times gave better internal signal. Sense and antisense RNA probes were synthesized using the Dig RNA labeling kit (Boehringer Mannheim), and hydrolyzed to an average size of 200-300 base pairs (Cox et al., 1984). Anti-digoxigenin antibody was preadsorbed against embryos before use. Color detection was stopped after 30 minutes to 1 hour. Since the complete α sense probe gave some background during stage 17, the sense control and antisense probes from BS-PCR were used for this stage; the complete α sense and antisense were used at all other stages. Some embryos were subsequently embedded in JB4 plastic resin (Polysciences) and 8 μm sections were made.

Drosophila strains

mysXG43,scb2, Df(2R)XTE18 and Df(2R)JP1 that uncover scb, and balancers FM7a and CyO have been previously described (Lindsley and Zimm, 1992). Chromosomes containing mysXG43, which behaves as a null mutation in β PS, sometimes contained FRT101 to generate germline clones using the FLP recombinase system (Chou and Perrimon, 1992). Resultant embryos lacking both zygotic and maternal mys are designated mys. scb alleles X5 and X6 were isolated by DEB mutagenesis and l19 by gamma irradiation in the laboratory of Eileen Underwood (Bowling Green State University). The P-element insertion line 1035 fails to complement scb and was obtained from the Bloomington Stock Center. lamA9− 32 (Henchcliffe et al., 1993) and ifK27e (Wilcox et al., 1989) act as loss-of-function mutations in laminin A and α PS2.

Videomicroscopy

Time-lapse videomicrosopy was performed as previously described (Roote and Zusman, 1995). All embryos were derived from mothers carrying the klarsicht mutation to clear the yolk. scb embryos were identified based upon the previously described dorsal hole phenotype; scbX5, scbX6, scb2 and Df(2R)XTE18 were examined.

Immunostaining

Embryos of wild-type, mys, scbX5, scbX6, scb2, and both deficiencies for scb, were collected on apple juice agar plates, prepared for immunostaining as previously described (Zusman et al., 1990) and stained with either an antibody to pericardial protein (Mab#3, (Yarnitzky and Volk, 1995)) or tracheal protein (2A12, (Samakovlis et al., 1996)). lamA9− 32 embryos were included when staining for trachea. To visualize salivary glands, wild-type, scbX5, scbX6 and scb2 embryos containing a transgene (fkhd360-505:lacZ, (Kuo et al., 1996)) were stained with anti-β-galactosidase antibody (Promega) as described (Blair, 1992). Embryos were dehydrated, mounted in a 3:1 solution of methyl salicylate and Canada balsam and examined under bright-field illumination. For examination of muscles, wild-type, scbX5, scbX6, scb2, ifK27e and Df(2R)XTE18 embryos were prepared and viewed under polarized light optics (Brabant and Brower, 1993). Late-stage embryos were identified by dorsal hole phenotype. At earlier stages, the identity of scb embryos was based on percentage of the population exhibiting defects and the similarity of these defects to those observed after the dorsal hole was evident. mys could be identified at earlier stages based upon its abnormal gut.

Insertion site cloning

The site of the PZ element insertion in scb1035 was obtained by plasmid rescue (Mlodzik and Hiromi, 1992) utilizing the bacterial origin in the element. Heterozygous genomic DNA from scb1035 was digested with Xba, ligated and plated on kanamycin-resistant media. The primer GTATACTTCGGTAAGCTTCGGCTTTC was used to sequence out of the PZ element into the recovered flanking DNA. To identify untranslated 5′ sequences from the short transcript, the 5′ RACE System kit (Life Technologies) was used according to manufacturers instructions using a GATCACTAGTCCGGGAACTGGGAGCGTAGA primer and adult RNA to prime first-strand cDNA. This was then used as template in the RACE PCR using the CGATCTGCAGAACCGAGGAGCACACACGAGGAGC nested primer and amplified using the GATCACTAGTCCATTGGGTCGCAGGCTCATC primer specific to the shorter transcript. The resulting product was subcloned using the pCR-Script Cloning Kit (Stratagene) according to manufacturers’ instructions and sequenced (MIT/ CCR/HHMI Biopolymers Laboratory).

Association of a 90 kDa protein with β PS

We used anti-β PS serum to immunoprecipitate β PS with associated proteins from non-ionic detergent lysates of surface-labeled cells. With non-reduced samples, a broad band of ∼95 kDa was seen in gels (Fig. 1A). On reduction, this resolved into a 110 kDa and a 90 kDa band (Fig. 1A) and a small band was also observed at or close to the dye front. Resolution of the 95 kDa band into the 110 kDa and the 90 kDa bands was confirmed by two-dimensional non-reducing/reducing gel electrophoresis (data not shown). A similar pattern was observed when cells from dissociated 10-15 hour Drosophila embryos were used (data not shown). In all cases, if the lysate was denatured by boiling in 4% SDS to disrupt noncovalent protein-protein interactions, only the 110 kDa band was observed (Fig. 1A) showing it to be β PS.

Fig. 1.

Association of β PS and the 90 kDa protein. (A,B) Immunoprecipitations from Schneider L2 cells using antibody to β PS. pre, preimmune; Ab, immune. (A) Cells were immunoprecipitated either directly, n, or after denaturation by boiling in SDS, d. Non-reducing gels show a broad band at ∼95 kDa; under reducing conditions, this resolves into two bands of 110 kDa and 90 kDa. Denatured lysates yield only the 110 kDa band corresponding to β PS. (B) The effects of divalent cation chelators, high salt and peptides on the association of the 90 kDa protein with β PS. Prior to immunoprecipitation the samples were supplemented with EDTA, EGTA, NaCl or peptides. The 90 kDa protein is removed by EDTA and greatly reduced in EGTA. (C) Sucrose density gradient analysis of the β PS subunit and the 90 kDa protein. Detergent lysates of 125I surface-labeled cells were fractionated on density gradients in the presence or absence of 10 mM EDTA. The gradient fractions were immunoprecipitated using antibody 185. β PS sediments to a lower position in the gradient (higher apparent molecular weight) in conjunction with the 90 kDa protein when EDTA is absent. A third band runs at the dye front when the 90 kDa protein is present.

Fig. 1.

Association of β PS and the 90 kDa protein. (A,B) Immunoprecipitations from Schneider L2 cells using antibody to β PS. pre, preimmune; Ab, immune. (A) Cells were immunoprecipitated either directly, n, or after denaturation by boiling in SDS, d. Non-reducing gels show a broad band at ∼95 kDa; under reducing conditions, this resolves into two bands of 110 kDa and 90 kDa. Denatured lysates yield only the 110 kDa band corresponding to β PS. (B) The effects of divalent cation chelators, high salt and peptides on the association of the 90 kDa protein with β PS. Prior to immunoprecipitation the samples were supplemented with EDTA, EGTA, NaCl or peptides. The 90 kDa protein is removed by EDTA and greatly reduced in EGTA. (C) Sucrose density gradient analysis of the β PS subunit and the 90 kDa protein. Detergent lysates of 125I surface-labeled cells were fractionated on density gradients in the presence or absence of 10 mM EDTA. The gradient fractions were immunoprecipitated using antibody 185. β PS sediments to a lower position in the gradient (higher apparent molecular weight) in conjunction with the 90 kDa protein when EDTA is absent. A third band runs at the dye front when the 90 kDa protein is present.

To characterize the nature of the interaction of the 90 kDa protein with β PS, the immunoprecipitation experiments were performed under a variety of conditions (Fig. 1B). The 90 kDa protein was absent from precipitates after lysis in 10 mM EDTA and was reduced if 10mM EGTA was added, consistent with a divalent Ca2+/Mg2+ requirement for the interaction. The 90 kDa protein was retained in the presence of 1 M NaCl in the buffers, and inclusion of RGE or RGD peptides had no effect, indicating the association was not an RGD-dependent ligand-receptor interaction. After overnight incubation of intact surface-labeled cells with EDTA, the 90 kDa protein was retained, indicating that it was not a divalent cation-dependent associated peripheral protein (data not shown).

To confirm that the 90 kDa protein formed a non-covalent, divalent cation-dependent complex with β PS and that its appearance was not due to spurious cross reactivity, we fractionated these two proteins using sucrose density gradient centrifugation, in the presence or absence of 10 mM EDTA. Fractions were collected and subjected to immunoprecipitation with the anti-β PS antibody (Fig. 1C). In the presence of EDTA, the 90 kDa protein was absent from the fractions containing β PS. In the presence of divalent cations, β PS and the 90 kDa protein sedimented together at a higher apparent molecular mass than when EDTA was present. A small protein running with the dye front in 7% gels also sedimented in the same fractions as β PS and the 90 kDa protein (Fig. 1C). These results confirmed the association of the 90 kDa protein, β PS and a small protein in a non-covalent, divalent cation-dependent complex.

To examine whether the 90 kDa band was a glycoprotein, immunoprecipitated complexes of β PS and 90 kDa were treated with EDTA to release the 90 kDa protein into the supernatant. The supernatant was incubated with lentil lectin or Concanavalin A beads. The 90 kDa protein bound well to both, indicating that it was a glycoprotein (data not shown).

Cloning and identification of 90 kDa

The 90 kDa protein was purified by immunoprecipitation followed by lectin binding, elution and then SDS gel electrophoresis. The purified protein was subjected to tryptic digestion and the resulting peptides were sequenced. Degenerate primers based upon the amino acid sequences were used to identify the cDNA coding for the 90 kDa protein by PCR. An N-terminal primer and a reverse primer produced a DNA fragment of approximately 1 kb using Drosophila embryonic cDNA as template (data not shown). Upon subcloning and sequencing, the PCR product showed homology to an α integrin repeat. By in situ hybridization, the PCR product clearly localized to 51E7-11 on polytene chromosome arm 2R (Fig. 2B).

Fig. 2.

(A) Schematic of the full-length protein. Y’s indicate potential glycosylation sites; I’s are cysteine residues. A large extracellular domain at the left contains seven α repeats (shaded), three divalent cation-binding domains (striped) within the last three α repeats, a putative cobalt- or zinc-binding domain (striped) overlapping the start of repeat three, a consensus cleavage site (lightly shaded), a short transmembrane domain (black bar) and a short cytoplasmic tail at the far right. The seven overlapping cDNAs and the probes are drawn to scale. The alternate 5′ end of cDNA 2C is shown in gray and two small retained introns are indicated by triangles. (B) A polytene chromosome in situ hybridization localizes the PCR fragment (arrow) to 51E7-11 on chromosome arm 2R.

Fig. 2.

(A) Schematic of the full-length protein. Y’s indicate potential glycosylation sites; I’s are cysteine residues. A large extracellular domain at the left contains seven α repeats (shaded), three divalent cation-binding domains (striped) within the last three α repeats, a putative cobalt- or zinc-binding domain (striped) overlapping the start of repeat three, a consensus cleavage site (lightly shaded), a short transmembrane domain (black bar) and a short cytoplasmic tail at the far right. The seven overlapping cDNAs and the probes are drawn to scale. The alternate 5′ end of cDNA 2C is shown in gray and two small retained introns are indicated by triangles. (B) A polytene chromosome in situ hybridization localizes the PCR fragment (arrow) to 51E7-11 on chromosome arm 2R.

A 9-12 hour embryonic cDNA library was screened with the PCR product to obtain cDNA for the entire protein coding region. Six overlapping cDNAs were obtained as well as one splicing variant (Fig. 2A). The longest of the cDNAs (10B) was sequenced and proved to be nearly full length, lacking only 105 base pairs of the coding region at the 5′ end. The remaining sequence for this variant was obtained from cDNA 11. The most 3′ cDNA appears to contain a poly-adenylation signal, but not the poly(A) tail itself. The sequence contains a 1115 amino acid open reading frame starting with a methionine codon (Fig. 3). The splicing variant, 2C, contained an alternate 5′ end. An alignment of this end with the remaining sequence is given in the first line of Fig. 3.

Fig. 3.

Amino acid sequence of the open reading frame. The alternate 5′ ends are aligned; !, : and. indicate decreasing levels of similarity. The signal sequences, transmembrane domain and putative cleavage site are boxed. Each α repeat is shaded and within them the divalent cation binding motifs have a striped bar above, as does the putative cobalt- or zinc-binding motif at the start of repeat three. The potential glycosylation sites are marked by an asterisk above and cysteines by a plus sign. Amino acid sequences obtained from tryptic digest fragments are underlined; the first residue in each of these was inferred to be an arginine or lysine; amino acids 309, 310, 662, 666, 845 and 813 could not be assigned during peptide sequencing. The tryptic fragments contained an A at position 598 and a T in position 696 that are presumed to be errors in peptide analysis; the putative amino terminus from Schneider L2 cell lysates is marked by a # above. The conserved KFGFFNR is found just inside the transmembrane domain at the beginning of the cytoplasmic tail. The nucleotide sequence has been submitted to GenBank: accession number U76605.

Fig. 3.

Amino acid sequence of the open reading frame. The alternate 5′ ends are aligned; !, : and. indicate decreasing levels of similarity. The signal sequences, transmembrane domain and putative cleavage site are boxed. Each α repeat is shaded and within them the divalent cation binding motifs have a striped bar above, as does the putative cobalt- or zinc-binding motif at the start of repeat three. The potential glycosylation sites are marked by an asterisk above and cysteines by a plus sign. Amino acid sequences obtained from tryptic digest fragments are underlined; the first residue in each of these was inferred to be an arginine or lysine; amino acids 309, 310, 662, 666, 845 and 813 could not be assigned during peptide sequencing. The tryptic fragments contained an A at position 598 and a T in position 696 that are presumed to be errors in peptide analysis; the putative amino terminus from Schneider L2 cell lysates is marked by a # above. The conserved KFGFFNR is found just inside the transmembrane domain at the beginning of the cytoplasmic tail. The nucleotide sequence has been submitted to GenBank: accession number U76605.

The sequence revealed an α integrin subunit that we have designated α PS3, in accord with the nomenclature of the α PS1 and α PS2 subunits which also associate with β PS. In early papers, β PS was sometimes called PS3, but this term is no longer used. It seems likely that a 92 kDa protein previously noted (Leptin et al, 1987) corresponds with the 90 kDa fragment of α PS3 reported here.

The predicted protein structure is shown schematically in Fig. 2A. A 24 amino acid signal sequence (von Heijne, 1986) was followed by a large (1033 amino acid) extracellular domain, a single transmembrane domain (23 amino acid) and a short (35 amino acid) cytoplasmic tail. The extracellular domain contained seven repeats common to all α integrin subunits (Pytela, 1988). The alternate 5′ splice had a 25 amino acid signal sequence followed by 38 amino acids and terminated in the first repeat. The last three repeats each contained a consensus divalent cation-binding motif Dx(D/N)x(D/N)GxxD for α integrins. A degenerate site was seen at position 262 between α repeats 3 and 4, having the Gxx deleted. This site occurred exactly at the N terminus of the 90 kDa protein purified from Schneider L2 cells. A protein lacking the predicted structures from the first 261 amino acids would be unlikely to function as a normal integrin since a number of the conserved repeats would be eliminated. This may have been an artefact of the protein purification process, or may reflect some unusual protein variant. The 20 amino acids starting at position 195 (Fig. 3) matched a motif found in a series of zinc- or cobalt-dependent enzymes. The histidine in this motif has been hypothesized to function in metal ion binding (motifs database, GCG). Although functional evidence for this motif is not available, it may indicate novel divalent cation-dependence for α PS3.

The immunoprecipitations and sucrose density gradients (Fig. 1) showed that this integrin belongs to the group of α integrins that are cleaved external to the transmembrane domain into a light and heavy chain. The greatest homology to the consensus cleavage site (K/R)R(E/D)hydrophobic was the RRDL at position 928. A consensus cleavage site was seen in this position in alignments with α 6, α 7, α 3, α PS3 and C. elegans F54G8.3 and nearby in α IIb (Loftus et al., 1987). This site occurred between the conserved cysteines implicated in the disulfide bonding of the heavy and light chains (Calvete et al., 1989), and was subsequent to the last of the tryptic sequences obtained from the putative heavy chain in Schneider L2 cells, all of which fell between amino acids 262 and 848. These lines of evidence were consistent with cleavage occurring at this site in α PS3.

14 cysteine residues were present; 12 were in conserved locations, two were in novel sites and two residues usually found were absent. The cysteine preceding the cleavage site was conserved in all but one integrin α subunit. The first cysteine following the cleavage site was also conserved, but was one of a cluster of three in most α subunits; here it was found with only one other cysteine. Nine potential N-linked glycosylation sites (NxT/S) were found in the extracellular domain (Figs 2A, 3). The location of the first of these is shifted in the alternate 5′ end. The highly conserved KxGFF(K/N)R was found just inside the putative transmembrane domain. The remainder of the cytoplasmic domain contained no striking similarity to any known protein.

A comparison using representatives of the α integrins was made (Fig. 4). The invertebrate integrins interspersed with the vertebrate integrins, rather than clustering separately, indicating an early divergence for this family. The sequence of α PS3 did not have a close vertebrate homologue.

Fig. 4.

A comparison of amino acid sequences for representatives of the α integrins. Sequences were human wherever possible. Note that the invertebrate integrins are interspersed among the vertebrate integrins. Integrins can be grouped into those that are cleaved or those that contain an I domain; and further characterized by affinity for β 2, laminin or ligands containing an RGD.

Fig. 4.

A comparison of amino acid sequences for representatives of the α integrins. Sequences were human wherever possible. Note that the invertebrate integrins are interspersed among the vertebrate integrins. Integrins can be grouped into those that are cleaved or those that contain an I domain; and further characterized by affinity for β 2, laminin or ligands containing an RGD.

Temporal regulation during development

Northern blot analysis revealed two transcripts of approximately 4.5 and 4.9 kilobases (Fig. 5). The larger transcript predominated in embryos beginning with a trace amount at 0-2 hours, a peak at 8-11 hours, then a decrease through 16 hours of embryogenesis, was strongly re-expressed during the pupal period and had some expression in the adult. A trace of the smaller transcript became evident by 6-8 hours of embryogenesis, increased in level from 8-16 hours and was the major form from 18 hours of embryogenesis through larval life. A detailed examination of the larval stages (Fig. 5C) revealed some increase in the expression of the larger transcript in late third instar larvae. The transcripts were approximately equal during pupal life and the smaller predominated in the adult (Fig. 5A). Adult males had a greater quantity of the larger transcript than did adult females, although the smaller transcript was more abundant in both (Fig. 5C).

Fig. 5.

Northern blot analysis of α PS3 expression. (A) Embryonic through adult total RNAs probed with the complete α probe. Two transcripts are observed of approximately 4.5 and 4.9 kb. (B) Quantitation control with RP49 probe shows pupal stages are underrepresented. (C) Larval stages emphasized, complete α probe. (D) Quantitation control using RP49 for larval blot. (E) Probe containing only 5′ sequences from cDNA 11 detects only the larger of the two transcripts. Note that adult males have more of the larger transcript than do females. (F) Probe containing the alternate coding region from cDNA 2C hybridizes only to the smaller of the two transcripts. (G) Complete α probe reveals the absence of α PS3 transcription in scbl19 embryos and 1st instar larvae. (H) RP49 control.

Fig. 5.

Northern blot analysis of α PS3 expression. (A) Embryonic through adult total RNAs probed with the complete α probe. Two transcripts are observed of approximately 4.5 and 4.9 kb. (B) Quantitation control with RP49 probe shows pupal stages are underrepresented. (C) Larval stages emphasized, complete α probe. (D) Quantitation control using RP49 for larval blot. (E) Probe containing only 5′ sequences from cDNA 11 detects only the larger of the two transcripts. Note that adult males have more of the larger transcript than do females. (F) Probe containing the alternate coding region from cDNA 2C hybridizes only to the smaller of the two transcripts. (G) Complete α probe reveals the absence of α PS3 transcription in scbl19 embryos and 1st instar larvae. (H) RP49 control.

A short probe containing only 5′ untranslated sequences from cDNA11 detected only the larger transcript (Fig. 5E). Four of the cDNAs identified that contain non-coding 5′ sequences, were identical to each other within these sequences. All of these represent the larger transcript, probably as a result of screening a 9-12 hour embryonic library, at which time the larger transcript is clearly predominant. The alt probe containing sequence for the splicing variant found in cDNA 2C hybridized specifically to the smaller transcript (Fig. 5F). All other probes tested hybridized to both transcripts.

Spatial distribution of α PS3 expression

We examined the spatial distribution of α PS3 transcripts by in situ hybridization to whole-mount Drosophila embryos (Fig. 6) and subsequently in plastic sections (Fig. 7) to obtain greater detail. During earliest embryogenesis, a low and uniform level of α PS3 RNA expression was seen; specific expression began at the onset of gastrulation and continued throughout embryogenesis. The first discrete group of cells stained constituted the amnioproctodeal invagination, both prior to and during the invagination itself (Fig. 6A).

Fig. 6.

Expression pattern of α PS3 RNA during embryogenesis shown by whole-mount in situ hybridization. Anterior at left, dorsal view unless noted. (A) Lateral view, dorsal at top shows strong expression in cells of the amnioproctodeum just prior to invagination; arrows indicate the cephalic furrow. (B) An early germ-band-extended embryo has staining in portions of the visceral mesoderm (m), some cells in the invaginating tracheal pits (t) and two anterior patches (a), below the plane of focus. (C) In late germ-bandextended embryos, expression is in a pair of anterior spots (a), in the amnioserosa (as), especially at the edges, and in the anterior (am) and posterior (pm) midgut primordia. (D) Lateral view, dorsal at top, of a germband-retracted embryo shows expression in an anterior spot (a) at the level of the foregut, in the midgut (m), the invaginating salivary gland (s), the amnioserosa (am) and the forming tracheal tree (t). (E) Lateral view, dorsal at top, of a stage 15 embryo just after dorsal closure. Expression is seen in the dorsal vessel (v), between the posterior spiracles (p), in the antenno-maxillary complex (am) and in some branches of the forming tracheal tree (t). (F) At stage 16, expression is seen in the salivary gland (s), its ducts and pore, the antenno-maxillary complex (am), the midgut constrictions (m), especially the second, and a subset of the tracheal tree (t). (G) Detail of a stage 16 embryo, ventral view, showing expression in some midline cells (ml) of the ventral nerve cord, and lesser expression in some cells scattered throughout the nerve cord. (H) A stage 17 embryo showing expression in the dorsal tracheal trunks (dt), the transverse connectives of the trachea (t), the anterior spiracles (sp) and the midgut (m). (I) A representative field of mixed stage embryos probed with complete α sense RNA show no detectable signal. (J) A representative stage 17 embryo probed with the sense PCR fragment shows no detectable signal. (K) Detail of stage 14 ventral nerve cord showing midline cells (ml), and fainter signal in other cells of the nerve cord (c). The midgut staining is becoming concentrated at the sites that will constrict (m). (L) Detail of the central nervous system of a partially dissected stage 16 embryo, showing staining in the brain periphery (b) and a lateral view of the midline cells (ml).

Fig. 6.

Expression pattern of α PS3 RNA during embryogenesis shown by whole-mount in situ hybridization. Anterior at left, dorsal view unless noted. (A) Lateral view, dorsal at top shows strong expression in cells of the amnioproctodeum just prior to invagination; arrows indicate the cephalic furrow. (B) An early germ-band-extended embryo has staining in portions of the visceral mesoderm (m), some cells in the invaginating tracheal pits (t) and two anterior patches (a), below the plane of focus. (C) In late germ-bandextended embryos, expression is in a pair of anterior spots (a), in the amnioserosa (as), especially at the edges, and in the anterior (am) and posterior (pm) midgut primordia. (D) Lateral view, dorsal at top, of a germband-retracted embryo shows expression in an anterior spot (a) at the level of the foregut, in the midgut (m), the invaginating salivary gland (s), the amnioserosa (am) and the forming tracheal tree (t). (E) Lateral view, dorsal at top, of a stage 15 embryo just after dorsal closure. Expression is seen in the dorsal vessel (v), between the posterior spiracles (p), in the antenno-maxillary complex (am) and in some branches of the forming tracheal tree (t). (F) At stage 16, expression is seen in the salivary gland (s), its ducts and pore, the antenno-maxillary complex (am), the midgut constrictions (m), especially the second, and a subset of the tracheal tree (t). (G) Detail of a stage 16 embryo, ventral view, showing expression in some midline cells (ml) of the ventral nerve cord, and lesser expression in some cells scattered throughout the nerve cord. (H) A stage 17 embryo showing expression in the dorsal tracheal trunks (dt), the transverse connectives of the trachea (t), the anterior spiracles (sp) and the midgut (m). (I) A representative field of mixed stage embryos probed with complete α sense RNA show no detectable signal. (J) A representative stage 17 embryo probed with the sense PCR fragment shows no detectable signal. (K) Detail of stage 14 ventral nerve cord showing midline cells (ml), and fainter signal in other cells of the nerve cord (c). The midgut staining is becoming concentrated at the sites that will constrict (m). (L) Detail of the central nervous system of a partially dissected stage 16 embryo, showing staining in the brain periphery (b) and a lateral view of the midline cells (ml).

Fig. 7.

Tissue-specific expression of α PS3 RNA in plastic sections of whole-mount in situ preparations viewed in bright field (A,C,E,G,I,K) and the same sections using Nomarski (B,D,F,H,J,L, respectively). Anterior is at the top except for E, F dorsal at top. (A) An early germ-band-extended embryo showing domains of repeated expression. (B) The staining is revealed to be in the mesoderm (m). (C) A late germ-band-extended embryo, shows laterally repeated patches and staining above the ventral midline. (D) The lateral staining is in some cells of the invaginating tracheal pits (t). (E) Cross section of a germ-band-retracted embryo shows staining in the midgut (m) and amnioserosa (a). (F) Midgut staining appears to be confined to the mesoderm (m). (G,H) Detail of the lumen of the proventriculus (p). (I,J) Sagittal section of a stage 17 embryo reveals prominent staining in midgut (m), gastric caeca (c) and dorsal tracheal through late trunk (t). (K,L) Glancing sagittal section of a germ-band-extended embryo, dorsal at right, shows strong staining in the amnioserosa (as).

Fig. 7.

Tissue-specific expression of α PS3 RNA in plastic sections of whole-mount in situ preparations viewed in bright field (A,C,E,G,I,K) and the same sections using Nomarski (B,D,F,H,J,L, respectively). Anterior is at the top except for E, F dorsal at top. (A) An early germ-band-extended embryo showing domains of repeated expression. (B) The staining is revealed to be in the mesoderm (m). (C) A late germ-band-extended embryo, shows laterally repeated patches and staining above the ventral midline. (D) The lateral staining is in some cells of the invaginating tracheal pits (t). (E) Cross section of a germ-band-retracted embryo shows staining in the midgut (m) and amnioserosa (a). (F) Midgut staining appears to be confined to the mesoderm (m). (G,H) Detail of the lumen of the proventriculus (p). (I,J) Sagittal section of a stage 17 embryo reveals prominent staining in midgut (m), gastric caeca (c) and dorsal tracheal through late trunk (t). (K,L) Glancing sagittal section of a germ-band-extended embryo, dorsal at right, shows strong staining in the amnioserosa (as).

In germ-band-extended embryos, signal was evident throughout the amnioserosa, strongest at the junction with the epidermis (Figs 6C, 7K,L). This continued through germ-band retraction and during dorsal closure (Fig. 6D). Subsequent to dorsal closure, strong expression was seen in the cells of the dorsal vessel (Fig. 6E).

Two regions of expression were seen in the anterior of early The germ-band-extended embryos. These loose accumulations of cells appeared to aggregate into two rounded clusters later (Fig. 6B,C). They were seen at either side of the foregut in stage 12 after germband retraction (Fig. 6D).

The developing tracheal system showed α PS3 RNA expression beginning in some of the cells of the invaginating tracheal pits (Figs 6B, 7C,B). Expression continued in subsets of the tracheal system throughout embryogenesis, particularly in the transverse connectives and anterior spiracles (Figs 6D-F, 7C,D). During stage 17, prominent expression was seen in the dorsal trunks as well as the transverse connectives and anterior spiracles (Figs 6H, 7I,J).

Portions of the midgut expressed α PS3 throughout development. Expression in early germ-bandextended embryos was in the presumptive visceral mesoderm (Figs 6B, 7A,B). Both the anterior and posterior midgut primordia had strong expression that continued through germ-band retraction and dorsal closure of the midgut (Fig. 6C,D). At these stages, the midgut expression appeared restricted to the mesoderm (Fig. 7E,F). Subsequently, expression was strongest in the forming midgut constrictions (Fig. 6F) and prefigured the appearance of any evident constriction (Fig. 6K). Some weak signal was seen in the forming proventriculus (Fig. 7G,H). By late in embryogenesis, signal was seen throughout the midgut including the gastric caeca (Figs 6H, 7I,J).

Expression of α PS3 was seen in the midline cells of the ventral nerve cord at about stages 15 and 16 (Fig. the midline cells of the ventral nerve cord at about stages 15 and 16 (Fig. 6G,K,L). Lower level expression was seen in some cells scattered throughout the ventral nerve cord (Fig. 6G,K). RNA was also seen in some cells of the brain periphery at this time (Fig. 6L). Subsequent to dorsal closure, signal was evident in the antenno-maxillary complex (Fig. 6E,F), and this continued through late embryogenesis.

A high level of signal was seen in the salivary glands beginning during their invagination and continuing throughout embryogenesis (Fig. 6D,F). Staining was present in both salivary ducts and the pore.

Defects during embryonic development

The chromosomal location of α PS3 DNA placed it at 51E7-11, in the vicinity of the previously described scab (scb) locus (Nusslein-Volhard et al., 1984). The defect in dorsal closure described for scb is similar to that seen in myospheroid (mys) embryos (mutant for β PS). This defect is not seen in nulls for either α PS1 or α PS2, nor in embryos doubly mutant for these genes (Brower et al., 1995), indicating that some other α subunit must be paired with β PS during dorsal closure. Taken together, these data suggested that scb might be the gene encoding α PS3. Southern blot analysis of heterozygous scb adults using the complete α probe revealed DNA polymorphisms in 6 alleles including l19 and 1035 (data not shown). Mutant embryos homozygous for the l19 allele of scb were picked by the presence of a dorsal hole; similar embryos homozygous for the XG43 allele of mys were picked as controls. Northern blot analysis showed the presence of α PS3 transcript in the mys embryos; none was detected in l19 embryos (Fig. 5F). The insertion site of the P element in scb allele 1035 was identified by plasmid rescue. Sequencing out of the PZ element into the flanking genomic DNA revealed the site of insertion to be 170 base pairs upstream of the start methionine for the alternate 5′ end encoding the shorter mRNA. This insertion site is 3 bases upstream of the 5′ end of the short transcript obtained by RACE PCR and 32 bases upstream of the 5′ end of cDNA 2C.

Mutant embryos homozygous for l19 or 1035, as well as other alleles of scb, like mys, sometimes undergo a secondary closure after the initial failure of dorsal closure. Unlike mys mutant embryos, whose muscles detach from the body wall, scb embryos show vigorous muscular movements of the cuticle, and some of those undergoing secondary dorsal closure actually hatch. Little or no expression of α PS3 RNA was seen in the body wall muscles of embryos and examination of scbX5, scbX6, scb2 and Df(2R)XTE18 embryos shows that muscle attachment sites appear normal (Fig. 8L).

Fig. 8.

Comparison of trachea, dorsal vessel, salivary gland, muscle, and germ-band defects in mutant embryos of scb, mys and lamA. Wild-type (A,B,H) and abnormal embryos from scb (C,D,I), mys (E,F,J) and lam A (G) producing crosses shown with anterior at left. A, C, E and G have been stained with 2A12 antibody showing gaps in the dorsal tracheal trunks of scb2, mysXG43 and lamA9− 32 embryos (arrows); lateral view, dorsal at top. Arrowhead in E indicates undigested yolk in the abnormal mys gut; scale bar 160 μm. (B,D,F) Mab#3 antibody staining of embryos in dorsal view shows mislocalization (arrows) and gaps (arrowheads) in the pericardial cells of the developing heart and dorsal vessel of a scbX5 and a mys embryo (scale bar, 160 μm). (H-J) The reporter gene shows salivary gland development (dorsal view). scb and mys embryos show one misshapen and smaller gland (arrows). (K,L) Somatic muscles are birefringent and appear as black and white bands in wild-type stage 17 embryos viewed under polarized light. (K) Lateral view, anterior at top, of a typical inflated (α PS2 mutant) embryo showing muscles rounded into black and white balls. (L) An embryo homozygous for Df(2R)XTE18 (noncomplementing to scb) has normal muscle attachments. Scale bar, 100 μm. (M,N) Dorsal views, anterior at top, of germ-band extension examined by videomicroscopy. (N) scbX5 shows the germ band abnormally twisted to the side (arrow), not centered on the midline as in wild type (M).

Fig. 8.

Comparison of trachea, dorsal vessel, salivary gland, muscle, and germ-band defects in mutant embryos of scb, mys and lamA. Wild-type (A,B,H) and abnormal embryos from scb (C,D,I), mys (E,F,J) and lam A (G) producing crosses shown with anterior at left. A, C, E and G have been stained with 2A12 antibody showing gaps in the dorsal tracheal trunks of scb2, mysXG43 and lamA9− 32 embryos (arrows); lateral view, dorsal at top. Arrowhead in E indicates undigested yolk in the abnormal mys gut; scale bar 160 μm. (B,D,F) Mab#3 antibody staining of embryos in dorsal view shows mislocalization (arrows) and gaps (arrowheads) in the pericardial cells of the developing heart and dorsal vessel of a scbX5 and a mys embryo (scale bar, 160 μm). (H-J) The reporter gene shows salivary gland development (dorsal view). scb and mys embryos show one misshapen and smaller gland (arrows). (K,L) Somatic muscles are birefringent and appear as black and white bands in wild-type stage 17 embryos viewed under polarized light. (K) Lateral view, anterior at top, of a typical inflated (α PS2 mutant) embryo showing muscles rounded into black and white balls. (L) An embryo homozygous for Df(2R)XTE18 (noncomplementing to scb) has normal muscle attachments. Scale bar, 100 μm. (M,N) Dorsal views, anterior at top, of germ-band extension examined by videomicroscopy. (N) scbX5 shows the germ band abnormally twisted to the side (arrow), not centered on the midline as in wild type (M).

In addition to dorsal closure, another phenotype observed in embryos lacking β PS, but not observed in embryos null for α PS1, α PS2 or both, was twisting of the germ band (Roote and Zusman, 1995). Time-lapse videomicroscopy reveals that, during germ-band extension in scbX5, scbX6, scb2 embryos and in embryos deficient for scb, the germ band twists laterally rather than extending dorsally as in wild-type embryos so that the ventral side of the posterior midgut is visible from the lateral side of the embryo (Fig. 8N). As in mys embryos, proper orientation of the germ band is recovered by the completion of germ-band extension.

The strong expression of α PS3 in the dorsal vessel (Fig. 6E) led us to examine scb embryos for defects in this tissue. Defects in the dorsal vessel have not been reported for mys embryos, although a detachment of alary muscles from the heart (posterior dorsal vessel) and its failure to mature at late stages has been identified (Wright, 1960) in embryos lacking zygotic, but not maternal β PS. We therefore examined earlier mys embryos lacking both maternal and zygotic β PS to preclude the possibility of maternal rescue. The heart and dorsal vessel form from two types of cells, the external pericardial cells and the internal cardioblasts. We stained mys, scbX5, scbX6, scb2 and scb-deficient embryos with antibodies that recognize pericardial cells. Embryos from the scb and deficiency lines show mislocalization of the pericardial cells, which normally organize in a line along the edge of the dorsal cuticle, and appear to have fewer of these cells in this area than wild type at the same stage (Fig. 8B,D). We were able to confirm a similar, although more severe defect in mys embryos; the pericardial cells appear to dissociate, migrate randomly and are sparse (Fig. 8F). The increase in severity of the defect suggests the possibility that either α PS1 or α PS2 may function in this process as well. The defect in mys is also strikingly similar to that published for laminin A mutations (Yarnitzky and Volk, 1995). Laminin is a common ligand for integrins in vertebrates.

Since α PS3 RNA is expressed in parts of the trachea throughout embryogenesis, we used antibodies to examine the trachea in wild-type, mys, scbX5, scbX6, scb2 and both scb deficiencies. To explore further the potential of laminin as a ligand, we examined lamA9-32 mutant embryos for tracheal defects also. Embryos from each of the mutant stocks have significant gaps in the dorsal trunk of the trachea, which are not present in wild-type embryos (Fig. 8A,C,E,G).

Due to the strong expression of α PS3 RNA in the salivary glands throughout development, we also examined the glands of wild-type, mys, scbX5, scbX6 and scb2 embryos using a reporter gene. mys and scb embryos frequently have one gland misshapen and smaller than the other (Fig. 8I,J). The salivary gland is sometimes shifted closer to the midline.

In conclusion, several defects are seen in scb mutant embryos, all of which are shared with mys and some with lamA; among them are defects previously reported in mys, but not in mew (α PS1) or if (α PS2) mutations. Furthermore, these defects arise in areas where aPS3 is strongly expressed.

The tissues in which α PS3 is expressed suggest a role for this integrin in morphogenesis and cell and tissue movement, as well as in simple adhesion. The Drosophila gene scb was identified based on a failure of dorsal closure (Nusslein-Volhard et al., 1984), and maps to the same chromosomal area to which α PS3 localized. Six alleles of scb show DNA polymorphisms in the vicinity of α PS3 and two of these have been further characterized. The 1035 allele has a P element inserted just 5′ of the coding region for the shorter transcript of α PS3. The 5′most sequence obtained by RACE PCR upon adult cDNA suggest this site is three bases from the transcription start site.

Homozygous scbl19 embryos fail to produce measurable levels of α PS3 transcript and embryos from scbl19, scb1035,scbX5 and scbX6 show the defect in dorsal closure reported for the original scb2 allele. This failure of dorsal closure has also been reported for mys mutants, which lack the β PS subunit, but it has not been observed in embryos lacking either α PS1 or α PS2, or in embryos doubly mutant for these α subunits (Wright, 1960; Brower et al., 1995). We have identified additional defects in scb embryos corresponding to sites of α PS3 expression. Similar defects have either been previously reported for mys embryos, or are reported here, confirming the evidence obtained biochemically that α PS3 and β PS together form a receptor. Moreover, these defects appear to be related to the movement and morphogenesis of tissues. Additional sites of α PS3 RNA expression are also suggestive of a role in tissue movements.

One example of a tissue having a complex morphogenesis is the trachea (Manning and Krasnow, 1993). Little or no division occurs in this tissue after stage 11 and development proceeds by tissue migration from a series of invaginating pits to a highly branched network. α PS3 expression is seen beginning in the invaginating tracheal pits and continues in subsets of the tracheal network throughout embryogenesis. The defect in tracheal development in scb embryos, and the corresponding defects in embryos lacking both maternal and zygotic β PS indicate that α PS3 with β PS plays a role during the migration and fusion of tissue in tracheal development.

α PS3 expression is also seen in the salivary glands beginning as they invaginate from the surface of the embryo and continuing as they form a tubular structure surrounding a central lumen. The defects observed in both scb and mys embryos in the morphology of the salivary glands show that α PS3 and β PS are required for the proper formation of this structure.

Embryos lacking both maternal and zygotic β PS have a defect in the extension of the germ band that is not observed in embryos doubly mutant for α PS1 and α PS2 (Roote and Zusman, 1995). This defect is seen in scb embryos, demonstrating a role for α PS3 and β PS in the movement of the germ band during embryogenesis.

We observed expression of α PS3 RNA in the amnioserosa throughout development and especially at its junction with the epidermis. After dorsal closure, expression is clearly present in some cells of the dorsal vessel. We therefore examined the pericardial cells, a population of cells that lie at the edge of the epidermis during dorsal closure and later contribute to the formation of the dorsal vessel. We have observed defects in the localization of the pericardial cells in scb mutant embryos, and have been able to identify a similar, although more severe defect in embryos lacking both maternal and zygotic β PS. The defect in mys embryos is strikingly similar to that reported for embryos missing the laminin A chain (Yarnitzky and Volk, 1995). This supports the hypothesis that α PS3 and β PS play a role in the movement and migration of the pericardial cells during dorsal vessel formation and suggests that laminin may be a ligand during this process.

Many additional tissues involved in movement during development show transcription of α PS3. Expression of α PS3 is seen in the amnioserosa throughout development; this is a tissue that expands and retracts and over which the dorsal epidermis closes. α PS3 transcription is seen in a group of anterior cells that migrate and then cluster. The expression of α PS3 in the visceral mesoderm and the midgut as it forms, occurs during a complex series of morphogenetic movements including invaginations and fusion of primordia. α PS3 expression prefigures and continues in the midgut constrictions. The cells of the amnioproctodeal invagination express α PS3 just before and during their invagination. An examination of the failure of migration in the proventriculus of β PS null embryos (Pankratz and Hoch, 1995) showed no corresponding failure in the proventriculus of α PS2, suggesting that either α PS1 or α PS3 may account for this defect. α PS3 does show some transcription in this tissue. All of these sites of expression are consistent with a role for α PS3 in movement and morphogenesis of tissues.

The temporal distribution of the two α PS3 transcripts, which differ in their 5′ ends, suggests that the larger transcript may be more responsive to rapid morphogenetic change and that this protein functions in these dynamic processes, while the regulation of the smaller transcript may permit more stable expression reflecting an ongoing requirement for this variant in established structures. The mature proteins, after signal sequence cleavage, produced by these transcripts differ from the 5′ end into the first α repeat; 39 amino acid from the larger transcript and 38 amino acid from the smaller. These residues are 37% identical and 61% similar in this region.

Previous studies have not demonstrated expression of the PS integrins in the central nervous system, although unexplained phenotypes have been reported for mutants in β PS. The embryonic ventral nerve cord in mys embryos fails to condense, has an overall slightly sloppy appearance and has occasional aberrant axon tracts (Wright, 1960; de la Pompa et al., 1989). α PS3 is clearly expressed in midline cells of the nerve cord, as well as in cells scattered throughout. β PS protein has recently been detected in the midline cells (Stephenie Paine-Saunders and R. O. Hynes, unpublished data), indicating that the neural defects in mys may be due to a requirement for integrins in this tissue rather than a secondary effect. α PS3 expression also occurs in some cells of the brain periphery as well as in the antenno-maxillary complex. The failure of condensation in the nerve cord has not been reported for mutants in either of the other two known α subunits.

Some sites of α PS3 expression, such as the gut, do not give rise to severe defects in scb mutant embryos, and some of the other defects seen are milder than we might anticipate from the extent and duration of the RNA expression. Several explanations are possible for this. We have not removed any possible maternal α PS3 contribution in these experiments. Only a small amount of α PS3 RNA is present in 0-2 hour embryos but, since we do also see α PS3 expression in the adult ovary, a small contribution of maternal RNA might rescue some phenotypes. It may be, however, that α PS3 protein is not essential in all tissues where its RNA is expressed, or that its functions are partially overlapping with other molecules. α PS1 and α PS2 are expressed in some of the same tissues as α PS3, and multiply mutant embryos might show an increase in severity of the pericardial cell phenotype or demonstrate a redundant function for α PS3 in the gut. Other non-integrin receptor molecules may also have overlapping or compensating functions.

Several lines of evidence suggest that α PS3 may be a laminin-binding integrin. α PS3 does share a conserved cleavage site with members of the laminin-binding integrins. Although in a comparison of amino acid sequences α PS3 does not cluster with the laminin-binding integrins (Fig. 4A), if DNA sequence is compared, then α PS3 has regions of greatest similarity to α 6, a laminin receptor. Furthermore, mutants in laminin A have been reported to have defects in dorsal vessel development (Yarnitzky and Volk, 1995) similar to those seen in scb and mys, and we have shown that they also exhibit similar defects in the trachea. Laminin is a good candidate for a ligand for this integrin, at least during dorsal vessel and tracheal development.

This new member of the PS integrin family expands the known functions of the integrins during development and presents the possibility for unraveling the mechanisms through which integrins function in movement and morphogenesis.

We wish to thank Ramila Patel-King, Stephenie Paine-Saunders and Joy Yang for RNAs, Grant Wheeler for cDNA, Kai Zinn for his cDNA library, Don Rio for Schneider L2 cells, Kim Mercer for plastic sectioning and the MIT Biopolymers facility for sequencing and primers. We thank Eileen Underwood for her generous contribution of scab alleles, as well as Norbert Perrimon, Eric Wieschaus, Ross MacIntyre, Kathy Matthews and the Bloomington Stock Center for fly stocks. We also wish to thank Talila Volk, Mark Krasnow and Steven Beckendorf for antibodies and transposon-containing Drosophila lines. We appreciate the technical assistance of Pallavi Narendranath, Sraddha Prativadi and Kassandra Gorham. We thank Laird Bloom and Michael DiPersio for critical review of the manuscript. R. O. Hynes is an Investigator of the Howard Hughes Medical Institute; Karen Stark was supported by the Howard Hughes Medical Institute and by a grant from the Markey Foundation. This work was supported by the Howard Hughes Medical Institute to R. O. Hynes and by NSF grants 9404055, IBN-9630783 and IBN-9722893 to Susan Zusman.

Adams
,
J. C.
and
Watt
,
F. M.
(
1993
).
Regulation of development and differentiation by the extracellular matrix
.
Development
117
,
1183
98
.
Ashburner
,
M.
(
1989
).
Drosophila A Laboratory Manual
.
Cold Spring Harbor
:
Cold Spring Harbor Laboratory Press
.
Ayme
,
A.
and
Tissieres
,
A.
(
1985
).
Locus 67B of Drosophila melanogaster contains seven, not four closely related heat shock genes
.
EMBO J
.
4
,
29492954
.
Blair
,
S. S.
(
1992
).
engrailed expression in the anterior lineage compartment of the developing wing blade of Drosophila
.
Development
115
,
21
34
.
Brabant
,
M. C.
and
Brower
,
D. L.
(
1993
).
PS2 integrin requirements in Drosophila embryo and wing morphogenesis
.
Dev. Biol
.
157
,
49
59
.
Brower
,
D. L.
,
Bunch
,
T. A.
,
Mukai
,
L.
,
Adamson
,
T. E.
,
Wehrli
,
M.
,
Lam
,
S.
,
Friedlander
,
E.
,
Roote
,
C.
and
Zusman
,
S.
(
1995
).
Nonequivalent requirements for PS1 and PS2 integrin at cell attachments in Drosophila: genetic analysis of the α PS1 integrin subunit
.
Development
121
,
1311
1320
.
Brower
,
D. L.
and
Jaffe
,
S. M.
(
1989
).
Requirement for integrins during Drosophila wing development
.
Nature
342
,
285
287
.
Brown
,
N. H.
(
1993
).
Integrins Hold Drosophila Together
.
BioEssays
15
,
383
390
.
Calvete
,
J.
,
Alvarez
,
M. V.
,
Rivas
,
G.
,
Hew
,
C.-L.
,
Henschen
,
A.
and
Gonzalez-Rodriguez
,
J.
(
1989
).
Interchain and intrachain disulphide bonds in human platelet glycoprotein IIb
.
J. Biochemi
.
261
,
551
560
.
Campos-Ortega
,
J. A.
and
Hartenstein
,
V.
(
1985
).
The Embryonic Development of Drosophila melanogaster
.
Berlin
:
Springer-Verlag
.
Catty
,
D.
and
Raykundalia
,
C.
(
1988
).
Production and quality control of polyclonal antibodies
.
In Antibodies-A Practical Approach (D. Catty, Ed
.), pp.
19
79
.
Oxford
:
IRL Press
.
Chou
,
T. B.
and
Perrimon
,
N.
(
1992
).
Use of a yeast site-specific recombinase to produce female germline chimeras in Drosophila
.
Genetics
131
,
643
653
.
Clark
,
E. A.
and
Brugge
,
J. S.
(
1995
).
Integrins and signal transduction pathways: the road taken
.
Science
268
,
233
239
.
Cox
,
K. H.
,
DeLeon
,
D. V.
,
Angerer
,
L. M.
and
Angerer
,
R. C.
(
1984
).
Detection of mRNAs in Sea Urchin Embryos in Situ Hybridization Using Asymmetric RNA probes
.
Dev. Biol
.
101
,
485
502
.
Damsky
,
C. H.
and
Werb
,
Z.
(
1992
).
Signal transduction by integrin receptors for extracellular matrix: cooperative processing of extracellular information
.
Curr. Opin. Cell Biol
.
4
,
772
81
.
de la Pompa
,
J. L.
,
Garcia
,
J. R.
and
Ferrus
,
A.
(
1989
).
Genetic Analysis of Muscle Development in Drosophila melanogaster
.
Dev. Biol
.
131
,
439
454
.
Gotwals
,
P. J.
,
Paine-Sauders
,
S. E.
,
Stark
,
K. A.
and
Hynes
,
R. O.
(
1994
).
Drosophila integrins and their ligands
.
Curr. Opin. Cell Biol
.
6
,
734
739
.
Henchcliffe
,
C.
,
Garcia-Alonso
,
L.
,
Tang
,
J.
and
Goodman
,
C. S.
(
1993
).
Genetic analysis of laminin A reveals diverse functions during morphogenesis in Drosophila
.
Development
118
,
325
337
.
Hynes
,
R. O.
(
1973
).
Alteration of cell-suface proteins by viral transformation and by proteolysis
.
Proc. Natl. Acad. Sci.,USA
70
,
3170
3174
.
Hynes
,
R. O.
(
1992
).
Integrins: versatility, modulation and signaling in cell adhesion
.
Cell
69
,
11
25
.
Kuo
,
Y. M.
,
Jones
,
N.
,
Zhou
,
B.
,
Panzer
,
S.
,
Larson
,
B.
and
Beckendorf
,
S. K.
(
1996
).
Salivary duct determination in Drosophila: roles of the EGF receptor signaling pathway and the transcription factors Fork head and Trachealess
.
Development
122
,
1909
1917
.
Laemmli
,
U. K.
(
1970
).
Cleavage of structural proteins during the assembly of the head of the bacteriophage T4
.
Nature
227
,
680
685
.
Leptin
,
M.
,
Aebersold
,
R.
and
Wilcox
,
M.
(
1987
).
Drosophila position-specific antigens resemble the vertebrate fibronectin-receptor family
.
EMBO J
.
6
,
1037
1043
.
Leptin
,
M.
,
Bogaert
,
T.
,
Lehmann
,
R.
and
Wilcox
,
M.
(
1989
).
The function of PS Integrins during Drosophila Embryogenesis
.
Cell
56
,
401
408
.
Lindsley
,
D. L.
and
Zimm
,
G. G.
(
1992
).
The Genome of Drosophila melanogaster
.
San Diego
:
Academic Press, Inc
.
Loftus
,
J. C.
,
Plow
,
E. F.
,
Jennings
,
L. K.
and
Ginsberg
,
M. H.
(
1987
).
Alternative Proteolytic Processing of Platelet Membrane Glyprotein IIb
.
J. Biol. Chem
.
263
,
11025
11028
.
Manning
,
G.
and
Krasnow
,
M. A.
(
1993
).
Development of the Drosophila Tracheal System
.
In The Development of Drosophila melanogaster
(ed.
M.
Bate
and
A.
Martinez-Arias
), Vol.
I
, pp.
609
685
.
Cold Spring Harbor
:
Cold Spring Harbor Laboratory Press
.
Mlodzik
,
M.
and
Hiromi
,
Y.
(
1992
).
The enhancer trap method in Drosophila: its applicatio to neurobioogy
.
Methods Neurosci
.
9
,
397
414
.
Nusslein-Volhard
,
C.
,
Wieschaus
,
E.
and
Kluding
,
H.
(
1984
).
Mutations affecting the pattern of the larval cuticle in Drosophila melanogaster I. Zygotic loci on the second chromosome
.
Wilh. Roux’s Arch. Dev. Biol
.
193
,
267
282
.
Pankratz
,
M. J.
and
Hoch
,
M.
(
1995
).
Control of epithelial morphogenesis by cell signaling and integrin molecules in the Drosophila foregut
.
Development
121
,
1885
1898
.
Pardue
,
M. L.
(
1986
).
In situ hybridization to DNA of chromosomes and nuclei
.
In Drosophila a Practical Approach
(ed.
D. B.
Roberts
), pp.
111
137
.
Washington, DC
:
IRL Press
.
Pytela
,
R.
(
1988
).
Amino acid sequence of the murine Mac-1 a chain reveals homology with the integrin family and an additional domain related to von Willebrand factor
.
EMBO J
.
7
,
1371
1378
.
Roote
,
C. E.
and
Zusman
,
S.
(
1995
).
Functions for PS integrins in tissue adhesion, migration and shape changes during early embryonic development in Drosophila
.
Dev. Biol
.
169
,
322
336
.
Samakovlis
,
C.
,
Hacohen
,
N.
,
Manning
,
G.
,
Sutherland
,
D. C.
,
Guillemin
,
K.
and
Krasnow
,
M. A.
(
1996
).
Development of the Drosophila tracheal system occurs by a series of morphologically distinct but genetically coupled branching events
.
Development
122
,
1395
1407
.
Sambrook
,
J.
,
Fritsch
,
E. F.
and
Maniatis
,
T.
(
1989
).
Molecular Cloning A Laboratory Manual
.
Cold Spring Harbor: Cold Spring Harbor Press
.
Sanger
,
F.
,
Nicklen
,
S.
and
Coulson
,
A. R.
(
1988
).
DNA sequencing with chain-terminating ihibitors
.
Proc. Natl. Acad. Sci., USA
74
,
5463
5467
.
Schwartz
,
M. A.
,
Schaller
,
M. D.
and
Ginsberg
,
M. H.
(
1995
).
Integrins: emerging paradigms of signal transduction
.
Annual Review of Cell And Developmental Biology
11
,
549
99
.
Tautz
,
D.
and
Pfeifle
,
C.
(
1989
).
A non-radioactive in situ hybridization method for the localization of specific RNAs in Drosophila embryos reveals a translational control of segmentation gene hunchback
.
Chromosoma
98
,
81
85
.
von Heijne
,
G.
(
1986
).
A new method for predicting signal sequence cleavage sites
.
Nucl. Acids Res
.
14
,
4683
4690
.
Wilcox
,
M.
,
DiAntonio
,
A.
and
Leptin
,
M.
(
1989
).
The function of PS integrins in Drosophila wing morphogenesis
.
Development
107
,
891
897
.
Wright
,
T. R. F.
(
1960
).
The Phenogenetics of the Embryonic Mutant, lethal myospheroid, in Drosophila melanogaster
.
J. Exp. Zool
.
143
,
77
99
.
Yarnitzky
,
T.
and
Volk
,
T.
(
1995
).
Laminin is required for heart, somatic muscles and gut development in the Drosophila embryo
.
Dev. Biol
.
169
,
609
618
.
Zhou
,
S.
and
Standring
,
D. N.
(
1992
).
Hepatitis B virus capsid particles are assembled from core-protein dimer precursors
.
Proc. Natl. Acad. Sci., USA
89
,
10046
10050
.
Zinn
,
K.
,
McAllister
,
L.
and
Goodman
,
C. S.
(
1988
).
Sequence analysis and neuronal expression of fasciclin I in grasshopper and Drosophila
.
Cell
53
,
577
587
.
Zusman
,
S.
,
Grinblat
,
Y.
,
Yee
,
G.
,
Kafatos
,
F. C.
and
Hynes
,
R. O.
(
1993
).
Analyses of PS integrin functions during Drosophila development
.
Development
118
,
737
750
.
Zusman
,
S.
,
Patel-King
,
R. S.
,
Ffrench-Constant
,
C.
and
Hynes
,
R. O.
(
1990
).
Requirements for integrins during Drosophila development
.
Development
108
,
391
402
.

A paper reporting on another aspect of the same novel integrin subunit is in press: Groteweil, M. S., Beck, C. D. O., Wu, K. H., Zhu, X. R. and Davis, R. L. (1997). Integrin-mediated short-term memory in Drosophila. Nature (In Press).