Bone-derived and recombinant transforming growth factor beta's are potent inhibitors of tumor cell growth

Two naturally occurring chrondogenesis inducing peptides have been purified to homogeneity from demineralized bovine bone. Cartilage-inducing factors A and B are the bone-derived equivalents of transforming growth factor-beta types I and II. Both peptides exhibit identical biological activities in chondrogenesis assays and stimulate anchorage independent cell growth. In this study we show that both bone-derived factors are potent (ng/ml) inhibitors of both DNA synthesis and the anchorage independent growth of a variety of human and non-human tumor cells. Unique in this study is also a comparison of the activities of these polypeptide growth factors with recombinant transforming growth factor type I expressed in mammalian cells.     

Cloning of a rat glia maturation factor-gamma (rGMFG) cDNA and expression of its mRNA and protein in rat organs

We isolated a rat glia maturation factor-gamma(rGMFG) cDNA and examined the tissue distribution of GMFG in rat by Northern and Western blot analyses. Sequence analysis of the entire cDNA revealed an open reading frame of 426 nucleotides with a deduced protein of 142 amino acid residues. The deduced amino acid sequence of the putative product is highly homologous (78.9%) to rat glia maturation factor-beta (rGMFB). Northern blot analysis indicated that a 0.9-kb mRNA is predominantly expressed in rat thymus, testis, and spleen. GMFG showed a different tissue distribution from GMFB, being present predominantly in proliferative and differentiative organs.     

Genetic diversity in natural populations of mammalian reoviruses: tryptic peptide analysis of outer capsid polypeptides of murine, bovine, and human type 1 and 3 reovirus strains.

We have studied the structural relationships between the outer capsid polypeptides of eight murine, bovine, and human isolates of type 1 and 3 mammalian reoviruses. Our results show that the outer capsid polypeptides of reoviruses isolated from different mammalian species, in different years and different geographical areas, have both conserved and unique methionine-containing tryptic peptides. We found that tryptic peptides from mu 1C polypeptides of two human, one murine, and two bovine type 3 isolates and one human and two bovine type 1 reoviruses are highly conserved. Our data show that only one tryptic peptide pattern of the mu 1C polypeptide (encoded by the M2 gene) was present in reoviruses isolated from the three different mammalian species. The mu 1C polypeptide of the type 3 Dearing strain contained one tryptic peptide not found in any other reovirus isolate examined. In marked contrast to the mu 1C polypeptides, the sigma 3 polypeptides (encoded by the S4 gene) of three type 1 and three type 3 isolates were divided into two patterns based on significant differences in their tryptic peptides. In addition, at least seven tryptic peptides were conserved among the sigma 3 polypeptides of all virus strains examined. The sigma 3 polypeptide of the type 3 Dearing strain was distinguishable from the sigma 3 polypeptides of all other strains examined. The one mu 1C and two sigma 3 tryptic peptide patterns were found to occur interchangeably in isolates of type 1 or type 3. About 1/3 of the tyrosine-containing tryptic peptides of sigma 1 polypeptides of four type 3 isolates examined were conserved. Comparison of peptide differences in sigma 1 polypeptides of these isolates showed that each had one or more unique tryptic peptides, suggesting that the S1 genes coding for these polypeptides had undergone genetic drift or, alternatively, that there are at least two tryptic peptide patterns present among the sigma 1 polypeptides of these isolates. Our results suggest that genetic drift and reassortment are the most likely explanation for the extensive genetic diversity found in natural populations of mammalian reoviruses.     

Disulfide isoforms of recombinant glia maturation factor beta

Recombinant human glia maturation factor beta (r-hGMF-beta) is a single-chain polypeptide (141 amino acid residues) containing three cysteines, at positions 7, 86 and 95. Nascent r-hGMF-beta exists in the reduced state and has no biological activity. The protein can be activated through oxidative refolding by incubation with a mixture of reduced and oxidized glutathione. Reverse-phase HPLC analysis of the refolded r-hGMF-beta shows the presence of four peaks, corresponding to the reduced form plus three newly generated intrachain disulfide-containing isoforms predicted from the number of cysteine residues. Only one isoform shows biological activity when tested for growth suppression on C6 glioma cells. We infer from the HPLC elution pattern that the active form contains the disulfide bridge Cys86-Cys95.     

Revised structure of the active form of human deoxyribonuclease IIalpha

Deoxyribonuclease IIalpha (DNase IIalpha) is an acid endonuclease found in lysosomes, nuclei, and various secretions. Murine DNase IIalpha is required for digesting the DNA of apoptotic cells after phagocytosis and for correct development and viability. DNase IIalpha purified from porcine spleen was previously shown to contain three peptides, two of which were thiol crosslinked, all derived by processing of a single polypeptide. Commercial bovine protein is consistent with this structure. However, screening of 18 human cell lines failed to demonstrate this processing, rather a 45 kDa protein was consistently observed. Incubation of cells with the N-glycosylation inhibitor tunicamycin resulted in a 37 kDa protein, which is close to the predicted formula weight. The protein also contains at least one thiol crosslink. Similar results were obtained with overexpressed DNase IIalpha. These results suggest that active DNase IIalpha consists of one contiguous polypeptide. We suggest the previous structure reflects proteolysis during protein purification.     

Sequence of membrane-associated thyroid hormone binding protein from bovine liver: its identity with protein disulphide isomerase

The complementary DNAs of the bovine liver membrane-associated 3,5,3'-triiodo-L-thyronine binding protein with 55 k-dalton (T3BP) were cloned and the nucleotide sequences were determined. Monospecific antibodies against T3BP were used to screen a bovine liver cDNA library in lambda gtll. We analyzed the sequences of two cloned T3BP cDNAs. The clones encoded a polypeptide of 510 amino acid residues, including a signal peptide of 20 amino acid. Northern blot analysis of bovine and human RNA showed that the mRNAs encoding T3BP are 2.7 kilobase in length. Amino acid sequence of N-terminal and other three peptides isolated from purified T3BP were found in the predicted amino acid sequence from the cDNA sequence. The predicted amino acid sequence is closely homologous (93%) with that of rat protein disulphide isomerase (EC 5.3.4.1), which catalyzes the isomerization of the protein disulphide bonds and has been shown to play an important role in post-translational regulation. The results suggest that T3BP and protein disulphide isomerase should be the same protein.     

Purification of homologous protein carboxyl methyltransferase isozymes from human and bovine erythrocytes

We have purified the two major isozymes of the L-isoaspartyl/D-aspartyl protein methyltransferase from both human and bovine erythrocytes. These four enzymes all have polypeptide molecular weights of approximately 26,500 and appear to be monomers in solution. Each of these enzymes cross-reacts with antibodies directed against protein carboxyl methyltransferase I from bovine brain. Their structures also appear to be similar when analyzed by dodecyl sulfate gel electrophoresis for the large fragments produced by digestion with Staphylococcus aureus protease V8 or when analyzed by high-performance liquid chromatography (HPLC) for tryptic peptides. The structural relatedness of these enzymes was confirmed by sequence analysis of a total of 433 residues in 32 tryptic fragments of the human erythrocyte isozymes I and II and of the bovine erythrocyte isozyme II. We found sequence identify or probable identity in 111 out of 112 residues when we compared the human isozymes I and II and identities in 127 out of 134 residues when the human and bovine isozymes II were compared. These results suggest that the erythrocyte isozymes from both organisms may have nearly identical structures and confirm the similarities in the function of these methyltransferases that have been previously demonstrated.     

Purification of a protein from bovine plasma that binds to type 1 plasminogen activator inhibitor and prevents its interaction with extracellular matrix. Evidence that the protein is vitronectin

Type 1 plasminogen activator inhibitor (PAI-1) binds to the extracellular matrix of cultured bovine aortic endothelial cells. Bovine plasma and bovine lung extract contain protein(s) that bind to PAI-1 and prevent this interaction. One of these proteins was purified approximately 425-fold from ammonium sulfate-fractionated plasma using standard chromatographic procedures together with affinity chromatography on PAI-1-Sepharose. The final product consisted of a major polypeptide of Mr 65,000 and two minor polypeptides of Mr 80,000 and 57,000. NH2-terminal amino acid sequence analysis of the Mr 65,000 polypeptide revealed that it was homologous with vitronectin, and antiserum against this purified binding protein recognized vitronectin and vice versa. Immunological analysis using these antisera demonstrated that the three peptides were immunologically related, and that vitronectin was present in the extracellular matrix of bovine endothelial cells and also in bovine lung.     

Primary structure of human copper-zinc superoxide dismutase. Cysteine - and tryptophan - containing peptides

Analysis of soluble peptides derived from tryptic and chymotryptic digestions of carboxymethylated human superoxide dismutase gives primary structural information for approximately 67% of the protein. Regions so far elucidated appear to be highly homologous to the corresponding bovine enzyme; in particular Cys 6 and the two cysteine residues 55 and 144, which form the intrasubunit disulfide bond of the bovine enzyme, are conserved. A cluster of three substitutions including the fourth cysteine residue unique to the human enzyme has been found in positions 107–109 and may be related to the presence of persulfide groups in the human enzyme. The single tryptophan residue of the human protein is not homologous to the single tyrosine residue of the bovine protein.     

The effects of shRNA-mediated gene silencing of transcription factor SNAI1 on the biological phenotypes of breast cancer cell line MCF-7

Developing spermatozoa require a series of posttesticular modifications within the luminal environment of the epididymis to achieve maturation; this involves several surface modifications including changes in plasma membrane lipids, proteins, carbohydrates, and alterations in the outer acrosomal membrane. Epididymal maturation can therefore allow sperm to gain forward motility and fertilization capabilities. The objective of this study was to identify maturation-dependent protein(s) and to investigate their role with the production of functionally competent spermatozoa. Lectin blot analyses of caput and cauda sperm plasma membrane fractions identified a 17.5 kDa wheat germ agglutinin (WGA)-binding polypeptide present in the cauda sperm plasma membrane not in the caput sperm plasma membrane. Among the several WGA-stained bands, the presence of a 17.5 kDa WGA-binding polypeptide band was detected only in cauda epididymal fluid not in caput epididymal fluid suggesting that the 17.5 kDa WGA-binding polypeptide is secreted from the cauda epididymis and binds to the cauda sperm plasma membrane during epididymal transit. Proteomic identification of the 17.5 kDa polypeptide yielded 13 peptides that matched the sequence of peroxiredoxin-5 (PRDX5) protein (Bos Taurus). We propose that bovine cauda sperm PRDX5 acts as an antioxidant enzyme in the epididymal environment, which is crucial in protecting the viable sperm population against the damage caused by endogeneous or exogeneous peroxide.     

Voltage-dependent anion-selective channels VDAC2 and VDAC3 are abundant proteins in bovine outer dense fibers, a cytoskeletal component of the sperm flagellum

Outer dense fibers (ODF) are specific subcellular components of the sperm flagellum. The functions of ODF have not yet been clearly elucidated. We have investigated the protein composition of purified ODF from bovine spermatozoa and found that one of the most abundant proteins is a 30-32-kDa polypeptide. This protein was analyzed by sequencing peptides derived following limited proteolysis. Peptide sequences were found to match VDAC2 and VDAC3. VDACs (voltage-dependent, anion-selective channels) or eukaryotic porins are a group of proteins first identified in the mitochondrial outer membrane that are able to form hydrophilic pore structures in membranes. In mammals, three VDAC isoforms (VDAC1, -2, -3) have been identified by cDNA cloning and sequencing. Antibodies against synthetic peptides specific for the three mammal VDAC isoforms were generated in rabbits. Their specificity was demonstrated by immunoblotting using recombinant VDAC1, -2, and -3. In protein extracts of bovine spermatozoa, VDAC1, -2, and -3 were detected by specific antibodies, while only VDAC2 and -3 were found as solubilized proteins derived from purified bovine ODFs. Immunofluorescence microscopy of spermatozoa revealed that anti-VDAC2 and anti-VDAC3 antibodies clearly bound to the sperm flagellum, in particular to the ODF. Transmission electron immunomicroscopy supported the finding that VDAC2 protein is abundant in the ODF. Since the ODF does not have any known membranous structure, it is tempting to speculate that VDAC2 and VDAC3 might have an alternative structural organization and different functions in ODF than in mitochondria.     

Studies on cytochrome c oxidase, XII. Isolation and primary structure of polypeptide VIb from bovine heart

The isolation and complete sequence analysis of the cytoplasmically synthesized polypeptide VIb from bovine heart cytochrome c oxidase is described. The protein is a stoichiometric constituent of the respiratory complex IV. Its primary structure is deduced from N-terminal sequencing and overlapping peptides obtained from tryptic cleavage and specific cleavage at arginyl and tryptophyl peptide bonds. The polypeptide chain consists of 84 amino acids from which a Mr of 9419 is derived. It has a relatively high content of histidine and proline and contains a single cysteine. A hydrophobic sequence of 20 amino acids points to a membrane-penetrating structure similar to that found in polypeptides I, II, III, IV and VIIIa, VIIIb, VIIIc of the bovine oxidase. The sequence of VIb is tissue-specific, it contributes to the formation of nuclear coded isoenzymes of cytochrome c oxidase. The protein thus may be involved in a tissue-specific regulation of cellular respiration.     

Mitogenic properties of a new endothelial cell growth factor related to pleiotrophin.

A growth factor was isolated from a neutral pH extract of adult bovine brain. Purification of this polypeptide was achieved by a three step procedure including cationic exchange, heparin-Sepharose affinity and Mono S chromatography. This heparin binding protein had a molecular weight of 18,000 as assessed by silver-stained SDS-PAGE and was not immunologically and structurally related to acidic or basic FGF. Freshly purified protein had a maximal mitogenic effect on bovine brain capillary cells at a concentration of 100 pM. Microsequencing revealed an unique amino-terminal sequence homologous to heparin-binding growth-associated molecule (HB-GAM), a neuronal maturation protein, to pleiotrophin (PTN), a fibroblast cell growth factor and to one form of the putative protein product of the MK gene, a retinoic acid induced-gene.     

Three species of polyoma virus tumor antigens share common peptides probably near the amino termini of the proteins.

Detergent extracts of polyoma virus-infected mouse cells contain three major proteins of approximately 100,000--108,000 (100K), 55,000 (55K) and 21,500 (22K) daltons, which react with sera obtained from rats carrying tumors induced by the virus. A comparison of the 35S-methionine-, 3H-leucine- and 3H-proline-labeled tryptic peptides of each of these proteins by cation-exchange chromatography followed by descending paper chromatography has shown that: at least five peptides are shared by all three T-reactive proteins; at least three peptides are shared by the 55K and 22K proteins, but not by the 100K protein; at least three peptides are found only in the 22K protein; at least six peptides are found only in the 55K protein; and at least sixteen peptides are found only in the 100K protein. The results are consistent with the hypothesis that the polypeptide chains of the 100K, 55K and 22K dalton tumor antigens of polyoma virus share a common virus-coded amino terminal region. The data also suggest that there is a portion of the polypeptide chains (probably immediately adjacent to the common amino terminal region of the molecules) that is shared by the 55K and 22K proteins, but not by the 100K protein (perhaps because this portion of the genetic information is spliced out of the messenger RNA coding for the 100K protein). The facts that all the peptides common to the 100K and 55K proteins are also found in the 22K protein and are thus assigned to the common amino terminal region of the molecules, and that there are several peptides unique to the 100K protein, as well as several peptides unique to the 55K protein, suggest that the presumed carboxy terminal portion of the polypeptide chain of the 100K protein is considerably, if not entirely, different from that of the 55K protein.     

Cloning of cDNAs encoding human interphotoreceptor retinoid-binding protein (IRBP) and comparison with bovine IRBP sequences

We have determined the sequence of the human interphotoreceptor retinoid-binding protein mRNA from three separately isolated cDNAs. The sequence is 4.28 kb long and encodes a protein of 1247 amino acids (aa) including a putative signal peptide and propeptide. The sequence is shorter (by about 1.67 kb) than the bovine mRNA with the major difference in the lengths located in the 3'-untranslated region. We suggest that this resulted from an insertion in the bovine gene or a large deletion from the human gene. The insertion/deletion is flanked on either side by sequences that are similar in the bovine and human sequences. Like the bovine polypeptide, the deduced protein sequence from the human cDNA contains a fourfold repeat, with each repeat containing about 300 aa. Among the four repeats, the identity is about 30-40%. The identity between the complete bovine and human polypeptide sequences is 84%. The identity between the nucleotide sequences is 83% (excluding the major insertion/deletion). Comparison with the bovine gene indicates that the human sequence may lack about 5-10 bp at the 5' end of the cDNA; it, however, includes a poly(A) tail at the 3' end. Thus, the human sequence is virtually full length, is similar to the bovine sequence, and contains a striking fourfold repeat.     

Characterization of G25K, a GTP-binding protein containing a novel putative nucleotide binding domain

Amino acid sequences were obtained for four peptides (p1, -2, -3 and 4) generated by chemical or proteolytic cleavage of a 25 kDa GTP-binding protein purified from human placental and platelet membranes. The peptides shared sequence similarities with those contained in several of the ras-related GTP-binding proteins. Peptide p2, a 12-mer, was homologous with a region of the GTP-binding proteins that contains a structural motif proposed to contribute to the nucleotide binding site. However, whereas nearly all GTP-binding proteins exhibit the residues NKXD as this motif, p2 contains TQID. Antisera (Ap1 and Ap3) raised against synthetic peptides corresponding to p1 and p3 specifically reacted on Western blots with the 25 kDa GTP-binding protein purified from human placenta, human platelet and bovine brain as well as with a 25 kDa polypeptide in various cell lines. These results demonstrate the widespread existence of an abundant 25 kDa GTP-binding protein which contains a putative nucleotide binding domain that is chemically distinct from that described for all GTP-binding proteins of known primary structure.     

VGF metabolic-related gene: distribution of its derived peptides in mammalian pancreatic islets.

The vgf gene has been shown to be involved in several metabolic pathways. Because the pancreas is crucial to metabolism and food intake, we studied the VGF peptides in bovine, rat, and pig Langherans islets using antisera raised against specific sites along the primary sequence of the rat/mouse and human VGF protein precursor. Whereas almost all of the pancreatic endocrine cells expressed vgf mRNA, when using the VGF antisera a different staining pattern became apparent. VGF(556-565) and VGF(282-291) immunoreactivity were exclusively found in delta somatostatin-producing cells, whereas the human C-terminus antiserum selectively immunolabeled alpha glucagon and pancreatic polypeptide cells. The same cells were decorated with the VGF(443-588) antiserum, which also weakly labeled beta insulin-secreting cells. Finally, the VGF(298-306) peptide and the rat C terminus were found in virtually all pancreatic endocrine cells. Using bovine, swine, and rat pancreatic extracts, data from chromatography and ELISA assay showed the presence of a high molecular mass form compatible with the proVGF and lower molecular mass fractions corresponding to short VGF peptides. In conclusion, selective VGF distribution may suggest a multifaceted cell type-specific processing of proVGF, resulting in different peptides probably involved in neuroendocrine regulatory metabolic mechanisms.     

Identification of two calcineurin alpha isoforms in bovine brain by two different monoclonal antibodies.

Calcineurin (calcium- and calmodulin-stimulated phosphatase) alpha subunit purified from bovine brain was found to be composed of two polypeptides, 61 KDa (alpha 1) and 59 KDa (alpha 2). The two peptides were separated and extracted from polyacrylamide gel. The immuno-peptide mapping of the purified peptides by partial proteolysis showed that the 59-KDa polypeptide was not a degradative product of the 61-KDa polypeptide. The interaction of the enzyme with two monoclonal antibodies, Vj6 and Vd3, raised against bovine brain calcineurin revealed that the 61-KDa polypeptide was recognized by both Vj6 and Vd3, whereas the 59-KDa one was recognized only by Vj6. These results indicate that there are at least two isoforms of calcineurin alpha subunits in bovine brain.