Search

Collagen IV comprises the predominant protein network of basement membranes, a specialized extracellular matrix, which underlie epithelia and endothelia. These networks assemble through oligomerization and covalent crosslinking to endow mechanical strength and shape cell behavior through interactions with cell-surface receptors. A recently discovered sulfilimine (S=N) bond between a methionine sulfur and hydroxylysine nitrogen reinforces the collagen IV network. We demonstrate that peroxidasin, an enzyme found in basement membranes, catalyzes formation of the sulfilimine bond. Drosophila peroxidasin mutants have disorganized collagen IV networks and torn visceral muscle basement membranes, pointing to a critical role for the enzyme in tissue biogenesis. Peroxidasin generates hypohalous acids as reaction intermediates, suggesting a paradoxically anabolic role for these usually destructive oxidants. This work highlights sulfilimine bond formation as what is to our knowledge the first known physiologic function for peroxidasin, a role for hypohalous oxidants in tissue biogenesis, and a possible role for peroxidasin in inflammatory diseases.

Papilins are homologous, secreted extracellular matrix proteins which share a common order of protein domains. They occur widely, from nematodes to man, and can differ in the number of repeats of a given type of domain. Within one species the number of repeats can vary by differential RNA splicing. A distinctly conserved cassette of domains at the amino-end of papilins is homologous with a cassette of protein domains at the carboxyl-end of the ADAMTS subgroup of secreted, matrix-associated metalloproteases. Papilins primarily occur in basement membranes. Papilins interact with several extracellular matrix components and ADAMTS enzymes. Papilins are essential for embryonic development of Drosophila melanogaster and Caenorhabditis elegans.

We show that the Drosophila gene rhea , isolated because its wing blister phenotype is typical of mutants affecting integrin function, encodes talin. Embryos deficient in talin have very similar phenotypes to integrin (βPS) null embryos, including failure in germ band retraction and muscle detachment. We demonstrate that talin is not required for the presence of integrins on the cell surface or their localization at muscle termini. However, talin is required for formation of focal adhesion-like clusters of integrins on the basal surface of imaginal disc epithelia and junctional plaques between muscle and tendon cells. These results indicate that talin is essential for integrin function and acts by stably linking clusters of ECM-linked integrins to the cytoskeleton.

Papilin is an extracellular matrix glycoprotein that we have found to be involved in, (1) thin matrix layers during gastrulation, (2) matrix associated with wandering, phagocytic hemocytes, (3) basement membranes and (4) space-filling matrix during Drosophila development. Determination of its cDNA sequence led to the identification of Caenorhabditis and mammalian papilins. A distinctly conserved ‘papilin cassette’ of domains at the amino-end of papilins is also the carboxyl-end of the ADAMTS subgroup of secreted, matrix-associated metalloproteinases; this cassette contains one thrombospondin type 1 (TSR) domain, a specific cysteine-rich domain and several partial TSR domains. In vitro, papilin non-competitively inhibits procollagen N-proteinase, an ADAMTS metalloproteinase. Inhibiting papilin synthesis in Drosophila or Caenorhabditis causes defective cell arrangements and embryonic death. Ectopic expression of papilin in Drosophila causes lethal abnormalities in muscle, Malpighian tubule and trachea formation. We suggest that papilin influences cell rearrangements and may modulate metalloproteinases during organogenesis.

UDP-glucose:glycoprotein glucosyltransferase (UGT) is a soluble protein of the endoplasmic reticulum (ER) that operates as a gatekeeper for quality control by preventing transport of improperly folded glycoproteins out of the ER. We report the isolation of two cDNAs encoding human UDP-glucose:glycoprotein glucosyltransferase homologues. HUGT1 encodes a 1555 amino acid polypeptide that, upon cleavage of an N-terminal signal peptide, is predicted to produce a soluble 173 kDa protein with the ER retrieval signal REEL. HUGT2 encodes a 1516 amino acid polypeptide that also contains a signal peptide and the ER retrieval signal HDEL. HUGT1 shares 55% identity with HUGT2 and 31-45% identity with Drosophila, Caenorhabditis elegans, and Schizosaccharomyces pombe homologues, with most extensive conservation of residues in the carboxy-terminal fifth of the protein, the proposed catalytic domain. HUGT1 is expressed as multiple mRNA species that are induced to similar extents upon disruption of protein folding in the ER. In contrast, HUGT2 is transcribed as a single mRNA species that is not induced under similar conditions. HUGT1 and HUGT2 mRNAs are broadly expressed in multiple tissues and differ slightly in their tissue distribution. The HUGT1 and HUGT2 cDNAs were expressed by transient transfection in COS-1 monkey cells to obtain similar levels of protein localized to the ER. Extracts from HUGT1-transfected cells displayed a 27-fold increase in the transfer of [(14)C]glucose from UDP-[(14)C]glucose to denatured substrates. Despite its high degree of sequence identity with HUGT1, the expressed recombinant HUGT2 protein was not functional under the conditions optimized for HUGT1. Site-directed alanine mutagenesis within a highly conserved region of HUGT1 identified four residues that are essential for catalytic function.

The mammalian focal adhesion kinase (FAK) family of nonreceptor protein-tyrosine kinases have been implicated in controlling a multitude of cellular responses to the engagement of cell surface integrins and G protein-coupled receptors. We describe here aDrosophila melanogaster FAK homologue, DFak56, which maps to band 56D on the right arm of the second chromosome. Full-length DFak56 cDNA encodes a phosphoprotein of 140 kDa, which shares strong sequence similarity not only with mammalian p125FAKbut also with the more recently described mammalian Pyk2 (also known as CAKβ, RAFTK, FAK2, and CADTK) FAK family member. DFak56 has intrinsic tyrosine kinase activity and is phosphorylated on tyrosine in vivo. As is the case for FAK, tyrosine phosphorylation of DFak56 is increased upon plating Drosophila embryo cells on extracellular matrix proteins. In situ hybridization and immunofluorescence staining analysis showed that DFak56 is ubiquitously expressed with particularly high levels within the developing central nervous system. We utilized the UAS-GAL4 expression system to express DFak56 and analyze its function in vivo. Overexpression of DFak56 in the wing imaginal disc results in wing blistering in adults, a phenotype also observed with both position-specific integrin loss of function and position-specific integrin overexpression. Our results imply a role for DFak56 in adhesion-dependent signaling pathways in vivo during D. melanogasterdevelopment.

Tiggrin is a novel extracellular matrix ligand for the Drosophila PS2 integrins. We have used flanking P elements to generate a precise deletion of tiggrin. Most flies lacking tiggrin die as larvae or pupae. A few adults do emerge and these appear to be relatively normal, displaying only misshapen abdomens and a low frequency of wing defects. Examination of larvae shows that muscle connections, function and morphology are defective in tiggrin mutants. Muscle contraction waves that extend the length of the larvae are much slower in tiggrin mutants. Direct examination of bodywall muscles shows defects in muscle attachment sites, where tiggrin is specifically localized, and muscles appear thinner. Transgenes expressing tiggrin are capable of rescuing tiggrin mutant phenotypes. Transgenes expressing a mutant tiggrin, whose Arg-Gly-Asp (RGD) integrin recognition sequence has been mutated to Leu-Gly-Ala (LGA) show much reduced, but significant, rescuing ability. Cell spreading assays detect no interactions of this mutant tiggrin with PS2 integrins. Therefore, while the RGD sequence is critical for PS2 interactions and full activity in the whole fly, the mutant tiggrin retains some function(s) that are probably mediated by interactions with other ECM molecules or cell surface receptors

We have expressed Drosophila position-specific (PS) integrins on the surfaces of Schneider S2 cells and tested for adhesion and spreading on various matrix molecules. We report that PS1 integrin is a laminin receptor and that PS1 and PS2 integrins promote cell spreading on two different Drosophila extracellular matrix molecules, laminin and tiggrin, respectively. The differing ligand specificities of these two integrins, combined with data on the in vivo expression patterns of the integrins and their ligands, lead to a model for the structure of integrin-dependent attachments in the pupal wings and embryonic muscles of Drosophila.

We propose that integrins help to coordinate the differentiation of the internal, sarcomeric cytoarchitecture of a muscle fiber with its immediate environment and are essential for correct integration of muscle cells into tissue. We found that integrin α PS2 β PS accumulated at contact regions of Drosophila embryo cells cultured in D-22 medium on Drosophila laminin. Myotubes formed, but subsequent addition of serum or fibronectin was needed for sarcomere formation: integrin and actin became concentrated at Z-bands; myosin and actin occurred between the Z-bands. This change falled to occur in the multinucleate myotubes derived from integrin β PS null myospheroid mutants. In normal embryos/early larvae, integrin was located at Z-bands and at muscle insertions. Myogenesis and Z-bands were defective in myospheroid embryos. Attachment, spreading, and growth of myoblasts and neurons on the laminin substrate utilized different binding proteins and were independent of integrin.

A Drosophila laminin that has the chain composition alpha5 beta1 gamma1, relative to mammalian laminins, bound human and mouse nidogen almost as strongly as mouse laminin-1 (alpha1 beta1 gamma1) in solid-phase assays, and had only a fourfold lower affinity in a radioligand competition test. This is due to a short, highly conserved sequence that occurs in both laminin gamma1 chains and which binds nidogen. When the single conservative amino acid difference between the two sequences (Tyr--His) was introduced into the mouse laminin binding module gamma1 III4 it failed to cause any change of binding. A high affinity between Drosophila laminin and mouse nidogen resulted in the formation of a stable complex in solution. Drosophila laminin also bound to the mouse heparan sulfate proteoglycan perlecan and the formation of this complex was inhibited by heparin, but not by chondroitin sulfate. In addition, a weaker connection between the core protein of mouse perlecan and Drosophila laminin can be mediated through nidogen. Elastase and other proteases degraded Drosophila laminin to a restricted number of larger fragments (40-300 kDa), almost all of which were bound to a heparin affinity column. Three fragments could be displaced at low salt concentration and were derived from the short arms of the Drosophila laminin, as shown by sequence analysis. A more strongly bound 50-kDa fragment apparently comprised the globular domains LG2 and LG3 derived from the C-terminal part of its alpha chain. Therefore, Drosophila laminin and mouse laminin-1 differ in certain aspects of protease stability and heparin-binding sites that, in part, can be attributed to their different alpha chains. The data suggest the existence of a nidogen analog and heparan sulfate proteoglycans in Drosophila, which remain to be identified.

A Drosophila UDP-glucose:glycoprotein glucosyltransferase was isolated, cloned and characterized. Its 1548 amino acid sequence begins with a signal peptide, lacks any putative transmembrane domains and terminates in a potential endoplasmic reticulum retrieval signal, HGEL. The soluble, 170 kDa glycoprotein occurs throughout Drosophila embryos, in microsomes of highly secretory Drosophila Kc cells and in small amounts in cell culture media. The isolated enzyme transfers [14C]glucose from UDP-[14C]Glc to several purified extracellular matrix glycoproteins (laminin, peroxidasin and glutactin) made by these cells, and to bovine thyroglobulin. These proteins must be denatured to accept glucose, which is bound at endoglycosidase H-sensitive sites. The unusual ability to discriminate between malfolded and native glycoproteins is shared by the rat liver homologue, previously described by A.J.Parodi and coworkers. The amino acid sequence presented differs from most glycosyltransferases. There is weak, though significant, similarity with a few bacterial lipopolysaccharide glycotransferases and a yeast protein Kre5p. In contrast, the 56-68% amino acid identities with partial sequences from genome projects of Caenorhabditis elegans, rice and Arabidopsis suggest widespread homologues of the enzyme. This glucosyltransferase fits previously proposed hypotheses for an endoplasmic reticular sensor of the state of folding of newly made glycoproteins.

A new monoclonal antibody, specific to an epitope in the carboxyl terminus of the Drosophila collagen IV molecule (basement membrane collagen) was identified. The distributions of collagen IV, laminin, and an additional extracellular molecule, the 2G2 antigen (2G2-Ag), were followed immunocytochemically during early wing development. In late third instar larvae, collagen IV and laminin surround the entire wing disc, whereas the 2G2-Ag is limited to the region of the future wing pouch. For the first few hours following eversion of the disc, all three ECM components line the basal surfaces of all epithelial cells in the wing pouch, both those destined to line the wing veins and those destined to become tightly apposed in the large intervein regions. Collagen IV and laminin persist on these cells during the two initial rounds of apposition of dorsal and ventral wing surfaces; later, they become restricted to the cells lining the veins. The 2G2-Ag disappears completely quite early in the pupal period. Collagen IV appears to be synthesized at least twice, once in the larva and a second time in the pupa; in between it is enzymatically cleaved and may be eliminated, probably by hemocytes. In an extreme allele of blistered the wing is ballooned to form a single internal space. Collagen IV and laminin line all basal wing cell surfaces early in pupal development as they do in the wild type. Later, however, they continue to line the entire cavity of the mutant wing rather than assuming a restricted distribution. In a completely veinless wing (rhomboidveinlet vein), collagen IV and laminin are also present generally on basal surfaces at early times, but are completely absent between the tightly apposed wing layers later. The ECM distributions both in wild type wings and in mutants suggest that the matrix plays a role in the establishment of the wing variation pattern. One possibility, strengthened by recent findings regarding ECM receptors in Drosophila, is their involvement in dorsal-ventral wing layer adhesion. Our findings also lead us to suggest that certain sets of features which distinguish vein from intervein cells may be linked during cell differentiation and thus help to define these cell phenotypes. The features include cytoskeletal specializations and certain cell surface and ECM molecules.

Peroxidasin is a novel protein combining peroxidase and extracellular matrix motifs. Hemocytes differentiate early from head mesoderm, make peroxidasin and later phagocytose apoptotic cells. As hemocytes spread throughout the embryo, they synthesize extracellular matrix and peroxidasin, incorporating it into completed basement membranes. Cultured cells secrete peroxidasin; it occurs in larvae and adults. Each 1512 residue chain of the three-armed, disulfide-linked homotrimer combines a peroxidase domain with six leucine-rich regions, four Ig loops, a thrombospondin/procollagen homology and an amphipathic alpha-helix. The peroxidase domain is homologous with human myeloperoxidase and eosinophil peroxidase. This heme protein catalyzes H2O2-driven radioiodinations, oxidations and formation of dityrosine. We propose that peroxidasin functions uniquely in extracellular matrix consolidation, phagocytosis and defense.

Genetic and other studies of Drosophila integrins have implicated these extracellular matrix receptors in various morphogenetic events, but identification of their endogenous ligands has been elusive. We report the biochemical purification and cloning of tiggrin, a novel extracellular matrix protein from Drosophila. This 255 ×103Mr polypeptide contains the potential integrin recognition sequence Arg-Gly-Asp (RGD) and 16 repeats of a novel 73-77 amino acid motif. The tiggrin gene is at chromosome locus 26D12 and is expressed by embryonic hemocytes and fat body cells. Tiggrin protein is detected in matrices, especially at muscle attachment sites that also strongly express integrins. Tiggrin-coated surfaces support primary embryo cell culture and provide excellent substrates for αPS2βPS integrin-mediated cell spreading. Soluble RGD-peptides inhibit this cell spreading.

Two contrasting substrates, Drosophila laminin and human vitronectin, caused determined primary Drosophila embryo cells to follow alternate intermediate differentiation steps without affecting the final outcome of differentiation. Integrin alpha PS2 beta PS3 was essential for the initial spreading of myocytes on vitronectin: focal contacts rich in beta PS3 integrins formed and were connected by actin- and myosin-containing stress fibers. While alpha PS2 beta PS3 was unnecessary for myotube formation on laminin, it was required for the subsequent change to a sarcomeric cytoarchitecture. The differentiating primary cultures synthesized integrins and assembled them into detergent-insoluble, cytoskeleton-associated complexes. Collagen IV, laminin, glutactin, papilin, and other extracellular matrix proteins were made primarily by hemocytes and were secreted into the medium. Further differentiation within the cultures was influenced by secreted components and by later addition of vitronectin or bovine serum. Comparison of the differentiation of various cell types on the two substrates showed that vitronectin provided a selective advantage for the differentiation of myocytes, with enrichment over epithelia, epidermal cells, and neurites.

Publisher Summary Research on Drosophila extracellular matrix (ECM) has mostly concerned the pericellular matrix and its relation to cells during development. This chapter focuses on the preparation and properties of individual Drosophila ECM proteins. Homologies with the corresponding vertebrate proteins are notable for at least two major Drosophila ECM components, basement membrane collagen IV and laminin. They secrete ECM components into their culture media. The chapter describes how biochemical amounts of several ECM proteins are isolated from the conditioned media. Some of the progenitor cells from which these cell lines were derived may have been related to hemocytes or cells of the fat body. The relative electrophoretic mobilities of the nonreduced and reduced ECM proteins are also discussed in the chapter. The mobilities of ECM components may vary significantly from those of standard, commercial “marker proteins” for various reasons—the average mass residue of collagens is significantly less than that of marker proteins and highly charged proteins, such as glutactin, bind sodium dodecyl sulfate (SDS) abnormally.

Previous work from our laboratories has demonstrated that: (a) the striated collagen fibrils of the corneal stroma are heterotypic structures composed of type V collagen molecules coassembled along with those of type I collagen, (b) the high content of type V collagen within the corneal collagen fibrils is one factor responsible for the small, uniform fibrillar diameter (25 nm) characteristic of this tissue, (c) the completely processed form of type V collagen found within tissues retains a large noncollagenous region, termed the NH2-terminal domain, at the amino end of its alpha 1 chain, and (d) the NH2-terminal domain may contain at least some of the information for the observed regulation of fibril diameters. In the present investigation we have employed polyclonal antibodies against the retained NH2-terminal domain of the alpha 1(V) chain for immunohistochemical studies of embryonic avian corneas and for immunoscreening a chicken cDNA library. When combined with cDNA sequencing and molecular rotary shadowing, these approaches provide information on the molecular structure of the retained NH2-terminal domain as well as how this domain might function in the regulation of fibrillar structure. In immunofluorescence and immunoelectron microscopy analyses, the antibodies against the NH2-terminal domain react with type V molecules present within mature heterotypic fibrils of the corneal stroma. Thus, epitopes within at least a portion of this domain are exposed on the fibril surface. This is in marked contrast to mAbs which we have previously characterized as being directed against epitopes located in the major triple helical domain of the type V molecule. The helical epitopes recognized by these antibodies are antigenically masked on type V molecules that have been assembled into fibrils. Sequencing of the isolated cDNA clones has provided the conceptual amino acid sequence of the entire amino end of the alpha 1(V) procollagen chain. The sequence shows the location of what appear to be potential propeptidase cleavage sites. One of these, if preferentially used during processing of the type V procollagen molecule, can provide an explanation for the retention of the NH2-terminal domain in the completely processed molecule. The sequencing data also suggest that the NH2-terminal domain consists of several regions, providing a structure which fits well with that of the completely processed type V molecule as visualized by rotary shadowing.

A Drosophila laminin A chain gene was characterized as a 14 kb genomic nucleotide sequence which encodes an open reading frame of 3712 amino acids in 15 exons. Overall, this A chain is similar to its vertebrate counterparts, especially in its N- and C-terminal globular domains, but the sequence that forms the laminin A short arm is quite different and larger. Laminin messages appear in newly formed mesoderm and are later prominently expressed in hemocytes, which also synthesize basement membrane collagen IV. The composition of Drosophila basement membranes changes with development. A novel method of tandemly fused RNA probes showed that developmental increases of laminin mRNAs were primarily associated with periods of morphogenesis, and preceded those of collagen IV, a protein strongly expressed during growth. The ratio of A:B1:B2 mRNAs varied little during embryogenesis, with less mRNA for A than B chains. Staining of embryos with antibodies confirmed and extended the information provided by in situ hybridization. Homologs of the G-subdomains of this A chain, which occur in interacting regions of agrin, perlecan, laminin and sex steroid binding protein, may be involved in protein associations.

Glutactin, a new acidic sulfated glycoprotein, was isolated from Drosophila Kc cell culture media. Immunofluorescence microscopy located it to embryonic basement membranes, particularly to the sequentially invaginated envelope of the central nervous system, muscle apodemes and dorsal median cell processes. Its chromosome locus is 29D. The nucleic acid sequence coding for the 1023 residue long polypeptide contains one intron and was confirmed by partial amino acid sequencing. Glutactin has a signal peptide and an amino domain of greater than 500 residues that strongly resembles acetylcholine esterases and other serine esterases, but lacks the catalytically critical serine residue. The amino and carboxyl domains of glutactin are separated by 13 contiguous threonine residues. Glutamine and glutamic acid make up 44% of glutactin's very acidic carboxyl domain. Glutactin preferentially binds Ca2+ in the presence of excess Mg2+ and four of its tyrosines are O-sulfated. Several similarities with mammalian entactin caused our previous, preliminary mention of glutactin as a putative Drosophila entactin, but sequence comparison now shows them to be different proteins.

Several tyrosine residues of the extracellular p-collagens V and collagens V are sulfated [Fessler, L. I., Brosh, S., Chapin, S. and Fessler, J. H. (1986) J. Biol. Chem. 261, 5034-5040]. Here, the sulfation of their intracellular precursors, the procollagens V, was studied. A Golgi-enriched subcellular fraction of chick embryo tendon catalyzed the sulfation of tyrosine residues in both endogenous and added, unsulfated procollagens V with the sulfate donor 3'-phosphoadenosine 5'-[35S]phosphosulfate. Intracellular tyrosine sulfation of procollagen V occurred at a point distal to the cis Golgi compartment as judged by change of the N-linked carbohydrate of procollagen V from being endoglycosidase-H-sensitive to being resistant. The time course of the intracellular modifications of procollagen V was determined by incubating tendons with 3H-labeled amino acids and with [35S]sulfate. The pro alpha(V) chains were synthesised in about 10 min and then assembled into unsulfated procollagen V molecules. Tyrosine sulfation occurred 50 min after completion of polypeptide synthesis and the molecules were successively sulfated in the order in which they had been synthesized. The antimicrotubular drug Nocodazole, which disrupts the spatial organization of the Golgi, decreased the time interval between synthesis of procollagens V and sulfation. The sulfated procollagens V were soon secreted and cut to sulfated p-collagens V. Sulfated pro alpha 1(V) chains were cleaved faster than sulfated pro alpha 1'(V) chains. The relationship of sequential protein modification to spatial cellular organization is discussed.

The three polypeptide chains of a precursor of collagen, procollagen, were found to join their NH2-terminal regions in disulfide linkage independently of the state of folding of the remainder of the polypeptide chains. This folding becomes altered from random coil to collagen helix by secondary modification, hydroxylation, of the chains. When hydroxylation was inhibited with α,α'-dipyridyl, chick embryo skull bones synthesized and accumulated such disulfide-linked procollagen. Ultracentrifugal velocity sedimentation analysis of this radioactively labeled material at different temperatures showed that it was in a random coil configuration at chick body temperature: Tm for denaturation was 25°. If hydroxylation was restored, by adding ferric ions, then the previously synthesized procollagen became folded. This was deduced from velocity sedimentation analyses in different concentrations of the denaturant, urea, at constant temperature. Subsequently to hydroxylation the procollagen was proteolytically cleaved, producing collagen in the same sequence as material which was synthesized in the absence of α,α'-dipyridyl. The Tm of normally hydroxylated procollagen was found to be 42°. Therefore hydroxylation of procollagen leads to collagen helix formation at chick body temperature. A technique of differential sedimentation analysis of radioactive macromolecules is described which showed a consistent difference in the sedimentation of hydroxylated and nonhydroxylated procollagens, whether as native triple-stranded molecules, or as denatured triplet chains held together by disulfide linkage, or as single, denatured chains after reduction. This difference approximately corresponds to the difference in mass which accompanies hydroxylation and glycosylation. The disulfide-linked triplet procollagen which was synthesized while hydroxylation was inhibited with α,α'-dipyridyl was shown to contain two polypeptide chains of type pro-α1 and one pro-α2 per molecule. Therefore the portions of each procollagen chain which are additional to the collagen sequence, and which have been called "registration peptides," are able to associate selectively to triplets under conditions which preclude collagen helix formation.

Drosophila laminin was isolated from the medium of Drosophila Kc cell cultures. It was purified by velocity sedimentation, gel filtration, and chromatography. Drosophila laminin is a disulfide-linked molecule consisting of three chains with apparent molecular masses of 400, 215, and 185 kD. In electron micrographs, it has the cross-shaped appearance with globular domains characteristic of vertebrate laminin with closely similar dimensions. The amino acid composition and lectin-binding properties of Drosophila laminin are given. Polyclonal antibodies to Drosophila laminin were prepared and their specificity was established. In developing embryos immunofluorescence staining was detected between 6 and 8 h of development; and in sections of 8-9-h and older embryos immunostaining was seen at sites where basement membranes are present surrounding internal organs, muscles, underlying the hypodermal epithelium, and in the nervous system. Basement membrane staining was also seen in larva and adults. Cells from Drosophila embryos dissociated at the cellular blastoderm stage were grown in culture and some specific, differentiated cells synthesized laminin after several hours of culture as shown by immunofluorescence. The significance of the evolutionary conservation of the structure of this basement membrane component is discussed.

Tendons from 14--17-day-old chick embryos contain predominantly type I collagen and about 5% AB2 collagen; type III collagen is not detectable by biochemical methods, such as sodium dodecyl sulfate/polyacrylamide gel electrophoresis or cyanogen bromide pattern, but can be visualized by immunofluorescence staining with collagen-type-specific antibodies. Similarly, freshly dissociated tendon cells secrete only type I collagen into the culture medium but no significant amounts of type III collagen [Uitto, J., Lichtenstein, J. R., and Bauer, E. A. (1976) Biochemistry, 15, 4935--4942]. Transfer of tendon cells from chick embryos to monolayer conditions, however, initiated synthesis of type III collagen in about 10% of the cells within three days, as visualized by immunofluorescence staining. Secretion of type III collagen into the culture medium can also be detected by sodium dodecyl sulfate/polyacrylamide gel electrophoresis. With increasing number of passages the number of cells producing type III collagen reached levels of about 80% after the third passage, while 90% of all cells stained positively for type I collagen. This is reflected by an increase of production of type III collagen as determined by CM-cellulose chromatography. Using velocity sedimentation, the secretion of type III procollagen and of pN-collagen (carrying the amino-terminal extension only), into the culture medium of a second-passage tendon cell culture was detected. This study provides new evidence that the phenotype of cells may alter during transfer from the environment in vivo to conditions in vitro and that additional changes may occur with time in culture.

The processing of [3H]proline-labeled procollagen III in excised chick embryo blood vessels was found to differ significantly from that of procollagen I in the same tissue. While first the amino propeptides and then the carboxyl propeptides were fairly rapidly cleaved from procollagen I, only the carboxyl propeptides were split off procollagen III, leaving pN-collagen III. This intermediate, which is only slowly converted to collagen III by loss of amino propeptides, was characterized by its sedimentation properties, isolation of the amino propeptide, and reaction with purified antibodies that are specific against bovine amino propeptide III. It is interchain disulfide-linked, both through the amino propeptide and the carboxyl ends of the collagen chains. The conversion of procollagen III to pN-collagen III either in blood vessels, or after isolation by a carboxyl procollagen peptidase obtained from chick tendon fibroblast cultures, is inhibited by 50 mM arginine. Underhydroxylated procollagen III was isolated from blood vessels treated with alpha, alpha'-dipyridyl. Its amino propeptides reacted with the above antibodies but were not linked to each other. In contrast, its carboxyl propeptides were interchain disulfide-bridged, supporting previous suggestions that the carboxyl propeptides play a role in the assembly of procollagen trimer.

Les interactions cellule-matrice et l'adhesion cellulaire dans le developpement de la drosophile, leur role potentiel dans le developpement neuronal est passe en revue

The rate of biosynthesis of procollagen IV, the principal collagen of basement membranes, and the concentration of specific RNAs coding for procollagen IV were measured in neonatal rat lungs. Both decreased sharply at birth and then recovered again a few days later. The supramolecular assembly of procollagen IV was followed in neonatal rat, mouse, and chick lungs, which actively elaborate endothelial and alveolar basement membranes, and in chick embryo gizzard which is rich in smooth muscle. The tetramer of four procollagen IV molecules linked covalently through their amino ends was isolated as an assembly intermediate from all these tissues. While noncovalent association of the carboxyl ends of two procollagen IV molecules occurred readily, the subsequent establishment of covalent cross-links was substantially slower in the junctional complexes of the carboxyl ends than of the amino ends. Both disulfide bonds and other, unidentified covalent links formed. The six component carboxyl peptides of a junctional complex became progressively covalently linked into two kinds of carboxyl peptide pairs. We conclude that both amino-linked tetramers and carboxyl-linked dimers of procollagen IV molecules are intermediates in the biological assembly of the collagen networks of these basement membranes.

A sulfated glycoprotein was isolated from the culture media of Drosophila Kc cells and named papilin. Affinity purified antibodies against this protein localized it primarily to the basement membranes of embryos. The antibodies cross-reacted with another material which was not sulfated and appeared to be the core protein of papilin, which is proteoglycan-like. After reduction, papilin electrophoresed in sodium dodecyl sulfate-polyacrylamide gel electrophoresis as a broad band of about 900,000 apparent molecular weight and the core protein as a narrow band of approximately 400,000. The core protein was formed by some cell lines and by other cells on incubation with 1 mM 4-methylumbelliferyl xyloside, which inhibited formation of the proteoglycan-like form. The buoyant density of papilin in CsCl/4 M guanidine hydrochloride is 1.4 g/ml, that of the core protein is much less. Papilin forms oligomers linked by disulfide bridges, as shown by sodium dodecyl sulfate-agarose gel electrophoresis and electron microscopy. The protomer is a 225 +/- 15-nm thread which is disulfide-linked into a loop with fine, protruding thread ends. Oligomers form clover-leaf-like structures. The protein contains 22% combined serine and threonine residues and 25% combined aspartic and glutamic residues. 10 g of polypeptide has attached 6.4 g of glucosamine, 3.1 g of galactosamine, 6.1 g of uronic acid, and 2.7 g of neutral sugars. There are about 80 O-linked carbohydrate chains/core protein molecule. Sulfate is attached to these chains. The O-linkage is through an unidentified neutral sugar. Papilin is largely resistant to common glycosidases and several proteases. The degree of sulfation varies with the sulfate concentration of the incubation medium. This proteoglycan-like glycoprotein differs substantially from corresponding proteoglycans found in vertebrate basement membranes, in contrast to Drosophila basement membrane laminin and collagen IV which have been conserved evolutionarily.

Procollagen IV was isolated from culture media of the mouse endodermal cell line PF-HR9. Some of the triple helical procollagen IV molecules were associated at their NH2 ends to tetramers which were identified by electron microscopy, velocity sedimentation, and electrophoresis. The formation of these tetramers in cell cultures and from isolated procollagen IV molecules was investigated. After an initial noncovalent association, which is reversible, disulfide bonds form between molecules. Even alkylated molecules form disulfide-linked tetramers when exposed to a mixture of reduced and oxidized glutathione. This reaction requires an adequate concentration of procollagen and is not facilitated by added laminin, Ca2+, or Mg2+ ions. Cystine, as a normal constituent of cell culture media, interferes in tetramer assembly, presumably by forming mixed disulfides. Tetramers formed normally and under the influence of glutathione are similar, but probably not identical, and resemble those isolated from fragmented basement membranes. We conclude that the NH2 ends of procollagen IV molecules can associate into tetramers without the help of other molecules and that disulfide bridges subsequently stabilize the association in various ways.

Whole tendons of chick embryos synthesize procollagens V which consist of three pro-alpha chains: pro-alpha 1'(V), pro-alpha 1(V) and pro-alpha 2(V). This report shows that while the pro-alpha 1'(V) chain is similar to the pro-alpha 1(V) chain in many respects, such as similar but not identical peptide maps, it also distinctly differs from it in size and in other ways. The new chain is denoted as pro-alpha 1' to indicate the relationship. We have failed to see conversion of one chain into the other and they are regarded as variants, although we do not know whether they are different transcripts of one gene or products of two closely related genes. The pro-alpha(V) chains are assembled into the disulfide-linked homotrimer (pro-alpha 1'(V))3 and the heterotrimer [(pro-alpha 1'(V)S-S-pro-alpha 2(V))pro-alpha 1(V)] and a smaller amount of [(pro-alpha 1(V)2pro-alpha 2(V)]. The pro-alpha 1'(V) chains are processed similarly to the pro-alpha 1(V) by the initial removal of the presumed carboxyl propeptide yielding p-alpha 1'(V) and then by reduction in the size of the noncollagenous, presumed amino propeptide to yield alpha 1'(V). A size difference between the alpha 1'(V) and alpha 1(V) series of molecules is demonstrated by velocity sedimentation and by electrophoretic mobility of the reduced molecules. This difference is ascribed to a difference in the size of the propeptides because after pepsin digestion the products of both series of molecules are the same size and electrophorese like alpha 1(V)(pepsin). The carboxyl propeptides of pro-alpha 1(V) and pro-alpha 1'(V) are the same size, but the amino propeptide of pro-alpha 1'(V) is smaller than that of pro-alpha 1(V). The amino propeptide of pro-alpha 1'(V) and p-alpha 1'(V) also lacks asparagine-linked complex carbohydrate, which is linked to propeptides of the p-alpha 1(V) and p-alpha 2(V) chains. Differences between the alpha 1(V) and alpha 1'(V) series of molecules remain in material synthesized in the presence of tunicamycin. Primary cultures of tendon cells synthesize procollagen V consisting of the above three chains, but the procollagen V made by cultured tendon sheath synovial cells predominantly contains [(pro-alpha 1(V))2pro-alpha 2(V)].

The endodermal cell line PF-HR9, derived from the murine teratocarcinoma cell line PCC4-F, was grown as monolayers and as cell clusters called embryoid bodies. Procollagen IV and laminin were isolated from both kinds of culture media. Antibodies specific to collagen IV and to laminin demonstrated these materials in association with the cells and in the culture media. The procollagen IV consisted of pro alpha 1 IV and pro alpha 2 IV chains and gave a circular dichroic spectrum characteristic for collagen helices, with thermal transitions at 40, 44, and 51 degrees C. The molecules were visualized electron microscopically after rotary shadowing. Laminin showed the characteristic beaded cross-appearance, and procollagen IV was a 434 +/- 12-nm long linear thread containing a 17-nm carboxyl-terminal knob. The 7% of collagen helix with Tm = 51 degrees C corresponds to about a 30-nm length of the molecule and is probably that section of the amino end through which several procollagen IV molecules form a junctional complex. Several noncovalent associations of procollagen IV molecules were demonstrated by velocity sedimentation and electron microscopy of concentrated culture media, specifically associations of two and four procollagen IV molecules through their amino ends and dimers linked at their carboxyl ends. The results show that procollagen IV molecules associate noncovalently into the components which others have isolated from basement membranes and strongly support a network model of these supramolecular assemblies.

A general mechanism for the assembly of procollagens is proposed from a biosynthetic study of procollagen III. This was shown to proceed by a stepwise process punctuated by disulfide bond formation and an assembly intermediate was recovered. The biosynthesis of type III procollagen in excised chick embryo blood vessels was studied by radioactive labeling for 30 min. Velocity sedimentation under denaturing conditions and purified antibodies specific against bovine amino propeptide III were used to identify and characterize monomeric pro alpha 1 III chains and a type III procollagen intermediate which is interchain disulfide-linked only at the carboxyl end but not at the amino end. The monomeric chains presumably have intrachain disulfide bonds within the propeptides. The monomeric pro alpha 1 III chains were also found when alpha, alpha'-dipyridyl was present during incubation. Pulse-chase experiments show that the monomeric chains and the intermediate are biosynthetic precursors of type III procollagen. Furthermore, it is shown that monomeric pro alpha 1 chains are not triple helical when extracted under nondenaturing conditions. The results indicate that the assembly of pro alpha 1 III chains into type III procollagen starts with the association of the folded carboxyl propeptides and is followed by formation of disulfide bonds between carboxyl propeptides, folding of the triple helix, and formation of disulfide bonds between amino propeptides. All procollagens may follow a similar assembly sequence.

Chick embryo skull bones incorporated radioactive proline and cystein into procollagen in short term organ culture. Pulse-chase experiments showed that individual precursor chains (pro-alpha1 and pro-alpha2) were formed first and that these were subsequently linked by disulfide bounds into trimers. Radioautography showed that labeled material was secreted 30 min after adding label to the cells, and electrophoretic analyses showed that after this time completed labeled collagen molecules appeared. Conversion from disulfide-linked procollagen to collagen proceeded in more than one step. An intermediate form consisting of shorter chains, which were still trimerically linked, was found.

Radioactive labeling of p-collagens V, collagens V, and, to a small extent, of procollagen V occurred when [35S]sulfate was incubated with tendons or primary tendon cell cultures, or blood vessels and crops of 17- to 19-day-old chick embryos, or with lung slices from neonatal rats. Most or all of this label is in the form of 1 or more sulfated tyrosine residues/chain of p alpha 1(V), alpha 1(V), p alpha 1'(V), alpha 1'(V), p alpha 2(V), and alpha 2(V), and it remains attached through purification by dialysis, ammonium sulfate precipitation, CsCl-GdnCl2 equilibrium buoyant density and velocity sedimentations, ion-exchange chromatography, and sodium dodecyl sulfate gel electrophoresis. Radioactive tyrosine sulfate was identified in alkaline hydrolysates of these collagen V chains, after labeling the tissues with either [35S]sulfate or [3H]tyrosine, by electrophoretic and chromatographic comigration with a tyrosine sulfate standard. Tunicamycin A1, which inhibits the attachment of N-linked complex carbohydrate, did not interfere with the sulfation process. The tyrosine sulfate is located in a noncollagenous domain, which is probably adjacent to the amino end of the collagen helix, and is retained throughout the physiological proteolytic processing of procollagens V. After digestion with Staphylococcus aureus V8 protease, 35S-labeled p alpha 1(V) and alpha 1(V) chains gave the same map of labeled peptides, and this differed from the map given by p alpha 1'(V) and alpha 1'(V) chains. Little sulfation of p alpha 2(V) and alpha 2(V) chains occurs. The implications of these observations for the structure and properties of procollagens V and their derivatives are considered.

Chick calvaria labeled in vitro contain a triple-stranded collagen precursor that contains two pro α1 and one pro α2 chains. All chains contain cystine and are linked by S-S bonds. Ultracentrifugal analysis was used to measure the relative size of the disulfide-linked units. It is proposed that the S-S links help to determine the correct alignment of the triple-stranded structure.

Chick embryo procollagen V, procollagen I, and pC-collagen I were sprayed, rotary shadowed, and compared electronmicroscopically with mouse procollagen IV and Drosophila procollagen produced by cell cultures and prepared in the same way. All the molecules appeared as threads and had a prominent knob protrusion at one end. For procollagen I and pC-collagen I this must correspond to the carboxyl propeptides, and most likely corresponds to the similarly sized carboxyl propeptides of procollagen IV. Procollagen V appeared as a thread of nearly the same length as procollagen I but with substantial knobs at both ends, corresponding to its known propeptides. The length of procollagen V and the known, successive processing of its propeptides make procollagen V more similar to the interstitial collagens than to the basement membrane procollagen IV, which is 1.4 times as long and has so far been found not to be processed. Drosophila procollagen resembles procollagen IV in length and appearance.

Antibodies have been raised against a fusion protein containing the 3' region of the coding sequence of the Antennapedia (Antp) gene fused to β-galactosidase. The distribution of the protein on whole mount embryos and imaginal discs of third instar larvae was examined by immunofluorescence. In young embryos, expression of the Antp protein was limited to the thoracic segments in the epidermis, whereas it was found in all neuromeres of head, thorax and abdomen. At the end of embryogenesis, the Antp protein mainly accumulated in the ventral nervous system in certain parts of the thoracic neuromeres, from posterior T1 to anterior T3, with a gap in posterior T2. Comparison of Antp protein distribution in nervous systems from wild-type and Df P9 embryos, lacking the genes of the Bithorax-complex (BX-C), revealed a pattern of expression which indicated that the BX-C represses Antp in the posterior segments with the exception of the last abdominal neuromeres (A8-9) which are regulated independently. The protein pattern in nervous systems from Sex combs reduced(ScrxF9) mutant embryos was indistinguishable from that found in wild-type embryos; thus, neurogenic expression of Antp in T1 and the more anterior segments does not appear to be under the control of Scr+. All imaginal discs derived from the three thoracic segments express Antp protein. The distribution was distinct in each disc; strongest expression was observed in the proximal parts of the discs. In the leg discs the protein distribution seemed to be compartmentally restricted, whereas in the wing disc this was not the case. Antp protein was not detected in the eye-antennal disc. In embryos, as well as in imaginal discs, the protein is localized in the nucleus.

An underlying theme of this meeting is the search for potential biophysical relationships between the extracellular matrix and the cytoskeleton. We found that the fruit fly, Drosophila melanoqaster. shares with vertebrates some major extra-cellular matrix components that are adjacent to cell surfaces, and also a transmembrane receptor of the integrin type. Actin filaments are probably indirectly linked to the cytoplasmic portion of such receptors in vertebrates and very similar actins exist in Drosophila. This suggests that parts of the molecular relationships between the extracellular matrix and the cytoplasm have been conserved during evolution, and that these mechanisms are likely to have important roles in cell biology.

Disulfide-linked, triple-stranded procollagen (pro γ 112) was isolated from chick embryo skull bones. It was reversibly denatured and renatured as judged by sedimentation properties and susceptibility to digestion with pepsin. Refolding was an intramolecular process and aggregation between molecules did not occur.

During the biosynthesis and assembly of collagen structures, disulfide links can serve several functions. During biosynthesis they successively stabilize intra-peptide folding and associations of three chains into one molecule. Studies on the refolding and reassociation of reduced and denatured carboxyl propeptides of procollagen I showed that successive interactions of folding and assembly are successively weaker. Disulfide bridges were reestablished within correctly refolded carboxyl propeptides. Rearrangements of disulfide bridges may occur during the processing of type V procollagen molecules as these collagens become incorporated into extracellular matrix. The basement membrane procollagen IV molecules become disulfide linked at each end into networks, and there are indications that further rearrangements of disulfide links may allow additional modulation.

The specific mammalian collagenase isolated from cultures of metastatic mouse PMT sarcoma cells cleaves murine procollagen IV into two segments, of approximate mass ratio 3:1. These fragments were separated by velocity sedimentation, visualized by electron microscopy, and analyzed. The longer COOH-terminal procollagen segment has a 270-nm collagenous portion with a knob at one end. This knob consists of the three previously identified, noncollagenous carboxyl propeptides, of approximately 30,000 daltons each. These carboxyl propeptides are chain-specific, and the three chains of each segment have the same amino to carboxyl orientation. The collagenase cuts through all three chains at one site, and the three-component chains of both the longer COOH-terminal procollagen segment and the shorter NH2-terminal procollagen segment are linked by interchain disulfide bridges. The enzyme cuts off the same COOH-terminal procollagen segment from procollagen IV monomers and tetramers, and the flexibility of this segment is similar to that of interstitial collagen helices. The amino ends of the NH2-terminal procollagen segments derived from tetramers remain joined as the 32-nm long "7 S collagen" junctional complex, and the remaining 89-nm long projecting threads are significantly more flexible than the COOH-terminal procollagen segment. The electrophoretic mobilities of the enzyme cleavage products are consistent with a heterotrimeric composition of this procollagen IV.