43K Protein Research Articles

The relationship of the postsynaptic 43K protein to acetylcholine receptors in receptor clusters isolated from cultured rat myotubes.

We have examined the relationship of acetylcholine receptors (AChR) to the Mr 43,000 receptor-associated protein (43K) in the AChR clusters of cultured rat myotubes. Indirect immunofluorescence revealed that the 43K protein was concentrated at the AChR domains of the receptor clusters in intact rat myotubes, in myotube fragments, and in clusters that had been purified approximately 100-fold by extraction with saponin. The association of the 43K protein with clustered AChR was not affected by buffers of high or low ionic strength, by alkaline pHs up to 10, or by chymotrypsin at 10 micrograms/ml. However, the 43K protein was removed from clusters with lithium diiodosalicylate or at alkaline pH (greater than 10). Upon extraction of 43K, several changes were observed in the AChR population. Receptors redistributed in the plane of the muscle membrane in alkali-extracted samples. The number of binding sites accessible to an anti-AChR monoclonal antibody directed against cytoplasmic epitopes (88B) doubled. Receptors became more susceptible to digestion by chymotrypsin, which destroyed the binding sites for the 88B antibody only after 43K was extracted. These results suggest that in isolated AChR clusters the 43K protein covers part of the cytoplasmic domain of AChR and may contribute to the unique distribution of this membrane protein.

Inhibition of Cleavage of a Plant Viral Polyprotein by an Inhibitor Activity Present in Wheat Germ and Cowpea Embryos

In rabbit reticulocyte lysate, the bottom component RNA of cowpea mosaic virus directs the synthesis of a 200,000-molecular-weight precursor protein (200K protein) that is cleaved during synthesis by a reticulocyte enzyme to form a 32K protein and a 170K protein. Cleavage of the 200K protein was found to be effectively inhibited by inhibitor activity in wheat germ and cowpea embryo extracts. The inhibitor was nondialyzable, precipitatable by ammonium sulfate, and partially stable at high temperatures. The activity appeared to be specific in that it caused no inhibition of the secondary cleavage reactions (cleavage of the 170K protein) at concentrations that were sufficient to cause complete inhibition of the primary cleavage reaction (cleavage of the 200K protein).

Medium tumor antigen of polyomavirus transformation-defective mutant NG59 is associated with 73-kilodalton heat shock protein.

Affinity-purified medium T antigen encoded by NG59, a nontransforming mutant of polyomavirus, is specifically associated with a protein of 72,000 daltons (72K protein). Medium T antigens of wild-type polyomavirus and the transformation-competent mutant dl8 are not associated with the 72K protein. Instead, they form a complex with another protein of 61,000 daltons. Several lines of evidence suggest that the medium T antigen-associated 72K protein is equivalent to the abundant and constitutive 73K heat shock protein. First, on two-dimensional polyacrylamide gels the 72K protein migrated with the same pI (5.6) as did the 73K heat shock protein. Second, the 72K protein was immunoprecipitable with antibodies against heat shock proteins. Third, when digested with V8 protease, the 72K protein gave rise to the same pattern of fragments as did the 73K heat shock protein.

A 14,000-Mr envelope protein of vaccinia virus is involved in cell fusion and forms covalently linked trimers.

A monoclonal antibody, MAbC3, that reacts with a 14,000-molecular-weight envelope protein (14K protein) of vaccinia virus completely inhibited virus-induced cell fusion during infection. Immunoblot and immunofluorescence studies revealed that the 14K protein was synthesized at about 6 to 7 h postinfection and transported from the cytoplasm to the cell surface. Synthesis and transport of the 14K protein during infection occurred in the presence of rifampin, an inhibitor of virus maturation. One- and two-dimensional gel electrophoretic analyses demonstrated that the 14K protein forms largely trimers (42K) that are covalently linked by disulfide bonds. The facts that MAbC3 prevents virus uncoating and blocks virus-induced cell fusion but does not prevent virus attachment to cells and the 14K envelope protein forms trimers all suggest that this protein plays major role in virus penetration.

Analysis of adenovirus early region 4-encoded polypeptides synthesized in productively infected cells.

Peptide-specific antisera were developed to analyze the products encoded by adenovirus type 5 early region 4 (E4) open reading frames 6 and 7. Reading frame 6 previously was shown to encode a 34-kilodalton polypeptide (34K polypeptide) that forms a complex with the early region 1B (E1B)-55K antigen and is required for efficient viral growth in lytic infection. Antisera that were generated recognized the E4-34K protein as well as a family of related polypeptides generated by the fusion of open reading frames 6 and 7. These polypeptides shared amino-terminal sequences with the 34K protein. Short-pulse analysis suggested that the heterogeneity observed with the 6/7 fusion products resulted from differential splicing patterns of related E4 mRNAs. An antiserum directed against the amino terminus of reading frame 6 recognized only the free form of the 34K antigen that was not associated with the E1B-55K protein. This observation allowed the determination of the stability of the free and complexed form of this polypeptide. Pulse-chase analyses demonstrated that both forms of the 34K protein had half-lives greater than 24 h, suggesting that complex formation did not result in stabilization of this gene product. The half-lives of the 6/7 fusion products were approximately 4 h. The 34K protein also was shown to have a nuclear localization within infected cells. Finally, analysis of a mutant carrying deletions in both the E4-34K and E1B-55K polypeptides indicated that the complex formed between these two proteins was a functional unit in lytic infection.

Phosphorylation of ribosomal protein S6 during lens cell differentiation: correlation with translational efficiency.

In vitro differentiation of embryonic chicken lens epithelial explants to form lens fiber cells is accompanied by an increase in protein synthesis without a corresponding increase in mRNA levels. This apparent increase in translational efficiency is correlated with a specific enhancement of phosphorylation of a 32K protein, which we identify as ribosomal protein S6 by two dimensional gel electrophoresis of purified ribosomal proteins. Serum, insulin, and chicken vitreous humor, three agents known to initiate differentiation in this system, all lead to enhanced S6 phosphorylation. Maximal enhancement of phosphorylation is reached within the first hour after the onset of differentiation, and is not blocked by inhibitors of RNA and protein synthesis.

A microtubule-binding protein associated with membranes of the Golgi apparatus.

A monoclonal antibody (M3A5), raised against microtubule-associated protein 2 (MAP-2), recognized an antigen associated with the Golgi complex in a variety of non-neuronal tissue culture cells. In double immunofluorescence studies M3A5 staining was very similar to that of specific Golgi markers, even after disruption of the Golgi apparatus organization with monensin or nocodazole. M3A5 recognized one band of Mr approximately 110,000 in immunoblots of culture cell extracts; this protein, designated 110K, was enriched in Golgi stack fractions prepared from rat liver. The 110K protein has been shown to partition into the aqueous phase by Triton X-114 extraction of a Golgi-enriched fraction and was eluted after pH 11.0 carbonate washing. It is therefore likely to be a peripheral membrane protein. Proteinase K treatment of an isolated Golgi stack fraction resulted in complete digestion of the 110K protein, both in the presence and absence of Triton X-100. A the 110K protein is accessible to protease in intact vesicles in vitro, it is presumably located on the cytoplasmic face of the Golgi membrane in vivo. The 110K protein was able to interact specifically with taxol-polymerized microtubules in vitro. These results suggest that the 110K protein may serve to link the Golgi apparatus to the microtubule network and so may belong to a novel class of proteins: the microtubule-binding proteins.

Processing and estrogen regulation of the 52-kilodalton protein inside MCF7 breast cancer cells.

A 52K glycoprotein is secreted by human breast cancer cells in culture after estrogen stimulation. Using monoclonal antibodies, we have quantitated and characterized the corresponding proteins of the cell compartment. Using pulse-chase experiments, we have shown that about 40% of the 52K protein is secreted, the majority being successively processed into a 48K and a 34K protein. This last protein is very stable. The processing is inhibited by lysosomotropic agents and leupeptin, suggesting that it occurs in acidic vesicles, such as lysosomes or endosomes. Estradiol increased the intracellular level of immunoreactive 52K related proteins by 4-fold. Its effect is, however, more obvious in the medium, since there is a constitutive level in the cell. The stimulatory effects of estradiol on [3H]mannose and [35S]methionine incorporation into these proteins were similar and the endoglycosydase H sensitivity of the proteins was not altered, suggesting that estradiol did not modulate the glycosylation step. Antiestrogens did not stimulate synthesis and glycosylation of the 52K related proteins. Estradiol also increased the stability of the 52K precursor as well as that of total proteins. We conclude that the secreted 52K protein is the precursor of two cellular proteins of 48K and 34K. Estradiol stimulates both the intracellular accumulation of these proteins and the secretion of the precursor.

The distribution of nuclear proteins and transcriptionally-active sequences in rat liver chromatin fractions

Chromatin fractions from rat liver nuclei digested by nucleases were separated by differential solubility into several fractions. Material solubilized during digestion (predominantly monomer nucleosomes and polynucleosomes) had the highest HMG14+17 DNA ratios but were not enriched in active gene sequences (albumin and c-Ha- ras1 genes). Material soluble in a low ionic strength buffer containing 0.2 mM MgCl 2 (monomer nucleosomes and polynucleosomes) contained in addition to the histones, HMG14 and 17 plus a 41K non-histone protein. This fraction was depleted in active gene sequences and enriched in inactive sequences. The insoluble material was highly enriched in active sequences and had the lowest HMG14+17 DNA ratio. This fraction could be further fractionated into a histone-containing 2 M NaCl-soluble fraction and a 2 M NaCl-insoluble matrix-bound fraction, both of which were enriched in active sequences. The results show that the HMG proteins do not partition with active sequences during fractionation of chromatin. The 41K protein may be associated with inactive chromatin fraction.

Preparation of antibodies that react with the large non-structural proteins of tobacco mosaic virus by usingEscherichia coli expressed fragments

We have establishedEscherichia coli strains that overproduce two regions of the large non-structural proteins of tobacco mosaic virus, the 126K and 183K proteins, as fusion proteins with β-galactosidase. The two fusion proteins included respectively 514 amino acids common to both the 126K and 183K proteins, and 432 amino acids specific to the 183K protein. The synthesis of the fusion proteins inE. coli was controlled by the lipoprotein promoter andlac promoter-operator systems. After induction, the fusion proteins that were synthesized aggregated and formed inclusion bodies. Antisera raised against the purified fusion proteins reacted specifically with both the 126K and 183K proteins or with only the 183K protein in TMV-infected tobacco protoplasts. The 54K protein corresponding to the C-terminus of the 183K protein, which has been suggested to be synthesized from a third subgenomic mRNA, could not be detected by this method.

Effect of tunicamycin on the structural proteins and infectivity of budded Autographa californica nuclear polyhedrosis virus

The budded phenotype of Autograha californica nuclear polyhedrosis virus (AcNPV BV) has a 64K glycoprotein in its envelope which is important in the ability of the virus to functionally enter host cells. Tunicamycin (TM), an inhibitor of N-linked glycosylation, was used to determine whether the sugar moiety was essential for the normal function of the 64K protein. At 10 μg/ml TM, glycosylation was completely inhibited. Infectivity of AcNPV BV was reduced to 2% of control while extracellular particles formed were 28% of control. Particles produced in the presence of 10 μg/ml TM were neutralizable by a monoclonal antibody reactive with the 64K envelope protein and their infectivity was sensitive to chloroquine. These results indicated that the major envelope protein, though unglycosylated, was still able to function normally in viral entry. Particles grown in the presence of 10 μg/ml TM however, contained on average, only 30% of the normal concentration of this major envelope protein. The combination of reduced number of extracellular particles and reduced amount of major envelope protein on these particles could account for the observed reduction in infectivity.

Elevated synthesis of an axonally transported protein correlates with axon outgrowth in normal and injured pyramidal tracts

Axons of the adult mammalian CNS typically fail to regenerate after injury. Among the hypotheses to account for this failure is the proposition that certain axonal proteins necessary for axon growth are expressed in much greater abundance in developing than in mature neurons, and that these proteins are not reinduced after injury to mature axons (Skene and Willard, 1981b). In the present experiments, we have found that hamster pyramidal tract neurons synthesize an acidic, 43K protein that is transported into growing axons during the first 2 weeks of postnatal development, and then declines at least an order of magnitude by the fourth postnatal week. The decline in synthesis of the 43K protein coincides with the cessation of pyramidal tract axon elongation. This protein resembles a "growth-associated protein," GAP-43, which is induced during regeneration of CNS axons in lower vertebrates. The 43K protein in hamster pyramidal tract neurons is not reinduced after axotomy in adult animals, which correlates with the failure of the injured axons to regenerate. Injury to neonatal pyramidal tract axons does not reverse or delay the decline in 43K protein synthesis. This is consistent with previous findings (Kalil and Reh, 1982) that pyramidal tract axons regrow for only a brief period after neonatal injury. Taken together, these results lend support to the hypothesis that synthesis of GAP-43 is important for axon growth in development and regeneration.

Component Proteins and Structure of Rice Ragged Stunt Virus

Summary Rice ragged stunt virus (RRSV) particles consist of a polyhedral core particle approx. 50 nm in diameter to which are attached flat spikes about 20 nm wide and 10 nm high, giving a total diameter of about 70 nm. Polyacrylamide gel electrophoresis of disrupted particles of RRSV gave five major polypeptides with mol. wt. of 145000 (145K), 137K, 72K, 47K and 37K. Core particles prepared by suspending particles in 0.5 m-MgCl2 contained the 145K, 137K, 72K proteins and relatively small amounts of the 37K protein, whereas the supernatant fluid recovered after centrifugation of the core particles contained the 47K and 37K proteins. The results suggest that the 145K, 137K and 72K proteins are core proteins and one or both of the 47K and 37K proteins are components of the spike. Immunoblotting experiments indicated that all the proteins are recognized as antigens, suggesting that RRSV particles are less stable than those of the phytoreoviruses, rice dwarf and rice gall dwarf viruses.

MRNA polyadenylate-binding protein: gene isolation and sequencing and identification of a ribonucleoprotein consensus sequence.

We identified and produced antibodies to the major proteins that interact with poly(A)+ RNAs in the yeast Saccharomyces cerevisiae. The major proteins which were cross-linked by UV light to poly(A)+ RNA in intact yeast cells had apparent molecular weights of 72,000, 60,000, and 50,000. The poly(A) segment of the RNA was selectively cross-linked to the 72,000-molecular-weight protein (72K protein). Mice immunized with purified UV-cross-linked RNA-protein (RNP) complexes produced antibodies to the three major RNP proteins. A yeast genomic DNA library constructed in the lambda gt11 expression vector was screened with the anti-RNP serum, and recombinant bacteriophage clones were isolated. One recombinant phage, lambda YPA72.1, bearing a 2.5-kilobase insert, produced a large beta-galactosidase-RNP fusion protein. Affinity-selected antibodies from the anti-RNP serum on this fusion protein recognized a single 72K protein which was cross-linked to the poly(A) segment of RNA in the intact cell. Furthermore, the fusion protein of lambda YPA72.1 had specific poly(A)-binding activity. Therefore, lambda YPA72.1 encodes the 72K poly(A)-binding protein. Immunofluorescence microscopy showed that this protein was localized in the cytoplasm. Hybrid-selected mRNA translated in vitro produced the 72K poly(A)-binding protein, and mRNA blot analysis detected a single 2.1-kilobase mRNA. DNA blot analysis suggested a single gene for the poly(A)-binding protein. DNA sequence analysis of genomic clones spanning the entire gene revealed a long open reading frame encoding a 64,272-molecular-weight protein with several distinct domains and repeating structural elements. A sequence of 11 to 13 amino acids is repeated three times in this protein. Strikingly, this repeated sequence (RNP consensus sequence) is highly homologous to a sequence that is repeated twice in a major mammalian heterogeneous nuclear RNP protein, A1. The conservation of the repetitive RNP consensus sequence suggests an important function and a common evolutionary origin for messenger RNP and heterogeneous nuclear RNP proteins.

MRNA polyadenylate-binding protein: gene isolation and sequencing and identification of a ribonucleoprotein consensus sequence.

We identified and produced antibodies to the major proteins that interact with poly(A)+ RNAs in the yeast Saccharomyces cerevisiae. The major proteins which were cross-linked by UV light to poly(A)+ RNA in intact yeast cells had apparent molecular weights of 72,000, 60,000, and 50,000. The poly(A) segment of the RNA was selectively cross-linked to the 72,000-molecular-weight protein (72K protein). Mice immunized with purified UV-cross-linked RNA-protein (RNP) complexes produced antibodies to the three major RNP proteins. A yeast genomic DNA library constructed in the lambda gt11 expression vector was screened with the anti-RNP serum, and recombinant bacteriophage clones were isolated. One recombinant phage, lambda YPA72.1, bearing a 2.5-kilobase insert, produced a large beta-galactosidase-RNP fusion protein. Affinity-selected antibodies from the anti-RNP serum on this fusion protein recognized a single 72K protein which was cross-linked to the poly(A) segment of RNA in the intact cell. Furthermore, the fusion protein of lambda YPA72.1 had specific poly(A)-binding activity. Therefore, lambda YPA72.1 encodes the 72K poly(A)-binding protein. Immunofluorescence microscopy showed that this protein was localized in the cytoplasm. Hybrid-selected mRNA translated in vitro produced the 72K poly(A)-binding protein, and mRNA blot analysis detected a single 2.1-kilobase mRNA. DNA blot analysis suggested a single gene for the poly(A)-binding protein. DNA sequence analysis of genomic clones spanning the entire gene revealed a long open reading frame encoding a 64,272-molecular-weight protein with several distinct domains and repeating structural elements. A sequence of 11 to 13 amino acids is repeated three times in this protein. Strikingly, this repeated sequence (RNP consensus sequence) is highly homologous to a sequence that is repeated twice in a major mammalian heterogeneous nuclear RNP protein, A1. The conservation of the repetitive RNP consensus sequence suggests an important function and a common evolutionary origin for messenger RNP and heterogeneous nuclear RNP proteins.

Detection and characterization of some new basic proteins in chicken postsynaptic densities.

Chicken brain postsynaptic density (PSD) polypeptides, obtained by treating synaptosomes with 0.5% Triton X-100 and then further purified on a sucrose gradient, are demonstrated to contain four basic proteins of 76K (pI greater than 9.2), 58K (pI 8.1-8.8, heterogeneous), 40K (pI 9.0), and 24K (pI 8.9). Nonequilibrium pH gradient-sodium dodecyl sulfate two-dimensional gels further reveal six more basic proteins with pI values higher than 9.2: 76K, 52K, 47K, 45K, 36K, and 34K. These basic proteins are a major part of the total chicken PSD polypeptides appearing on the gels. Some of these basic proteins (58K, 52K, 47K, 36K, 24K, and two at 76K) are distinguishable from those of brain mitochondria, the major contaminant. The 40K and 34K proteins may be common mitochondrial polypeptides. The 45K protein is probably a mitochondrial contaminant. A number of proteins including 76K (synapsin I-like protein) and 58K, along with some other minor ones, can be phosphorylated by endogenous protein kinase(s) in the presence of Ca2+, Mg2+, and [gamma-32P]ATP. No PSD basic proteins bind Ca2+.

Determination of the Proteolytic Processing Sites in the Polyprotein Encoded by the Bottom-Component RNA of Cowpea Mosaic Virus

The bottom-component RNA (B-RNA) of cowpea mosaic virus is expressed by the production of a approximately 200,000-dalton polyprotein (200K polyprotein), from which the functional proteins are formed by specific proteolytic cleavages. Partial amino-terminal sequences of the various B-RNA-encoded proteins have now been determined. Comparison of the information obtained with the B-RNA sequence allowed the localization of the coding regions for these proteins on B-RNA, the calculation of their precise molecular weights, and the determination of the cleavage sites at which they are released from the polyprotein precursor. Sequence analysis of the 32K protein, which is derived from the amino-terminal end of the 200K polyprotein, indicated that the AUG codon at nucleotide position 207 of the RNA sequence is the translation initiation codon. Sequence analysis of the 170K, 110K, 87K, 84K, 60K, and 58K proteins revealed the existence of three types of cleavage site in the 200K polyprotein: glutamine-serine (two sites), glutamine-methionine (one site), and glutamine-glycine (one site) amino acid pairs. The nature of these cleavage sites suggested that two different viral proteases are involved in the processing of the B-RNA-encoded polyprotein.

Botulinum C2 toxin ADP-ribosylates actin.

ADP-ribosylation of regulatory proteins is an important pathological mechanism by which various bacterial toxins affect eukaryotic cell functions. While diphtheria toxin catalyses the ADP-ribosylation of elongation factor 2, which results in inhibition of protein synthesis, cholera toxin and pertussis toxin ADP-ribosylate Ns and Ni, respectively, the GTP-binding regulatory components of the adenylate cyclase system, thereby modulating the bidirectional hormonal regulation of the adenylate cyclase. Botulinum C2 toxin is another toxin which has been reported to possess ADP-ribosyltransferase activity. This extremely toxic agent is produced by certain strains of Clostridium botulinum and induces hypotension, an increase in intestinal secretion, vascular permeability and haemorrhaging in the lungs. In contrast to botulinum neurotoxins, the botulinum C2 toxin apparently lacks any neurotoxic effects. Here we report that botulinum C2 toxin ADP-ribosylates a protein of relative molecular mass 43,000 (43K) in intact cells and in cell-free preparations. We present evidence that the 43K protein substrate is actin, which is apparently mono-ADP-ribosylated by the toxin. Botulinum C2 toxin also ADP-ribosylated purified liver G-actin, whereas liver F-actin was only poorly ADP-ribosylated and skeletal muscle actin was not ADP-ribosylated in either its G form or its F form. ADP-ribosylation of liver G-actin by botulinum C2 toxin resulted in a drastic reduction in viscosity of actin polymerized in vitro.

Purification and partial characterization of a novel thyroxine-binding protein (27K protein) from human plasma.

T4-binding globulin (TBG) prepared from human plasma by the standard three-step procedure (T4-agarose affinity chromatography, anion exchange chromatography, and gel filtration) often shows in sodium dodecyl sulfate-polyacrylamide gel electrophoresis in addition to the expected 54K band, another with a mol wt of 27,000 (27K protein). The two proteins can be separated after the three-step procedure by chromatofocusing (because of different isoelectric points, 4.2-4.8 for TBG and 5.0-5.2 for 27K protein) or by T4-aragose chromatography eluting with a linear gradient of T4 (TBG is eluted between 10(-10) and 10(-9) M T4, 27K protein between 10(-8) and 10(-7) M T4). The 27K protein does not appear to be a fragment of TBG since 1) it does not displace [125I]TBG bound to anti-TBG monoclonal antibodies; and 2) absorption of polyclonal antibody reacting with both TBG and 27K protein with sera from TBG-deficient patients completely prevents [125I]27K protein binding, while only slightly affecting [125I]TBG binding. On the other hand, 27K protein is not simply a contaminant devoid of biological activity, but is a T4-binding protein, as supported by the following findings: 1) it covalently binds [125I]T4 by photoaffinity labeling, and this binding can be almost completely prevented by excess T4; 2) equilibrium dialysis shows two equivalent T4-binding sites per 66K, with an association constant of 0.85 X 10(7) M-1, intermediate between albumin and prealbumin; and 3) tryptophanyl fluorescence analysis shows quenching of 37% of the fluorescence when the protein is titrated with T4. The 27K protein appears as a single 27K band in sodium dodecyl sulfate-polyacrylamide gel electrophoresis, pH 8.8, but under nondenaturing nonreducing conditions mostly remains at the origin of the gel; a fraction enters the gel and migrates slightly ahead of albumin. This electrophoretic pattern is distinct from those of albumin, prealbumin, and TBG. In immunoelectrophoresis in agar at pH 8.6, 27K protein moves slightly faster than TBG. The results of equilibrium sedimentation indicate a mol wt of 66,000, suggesting that the 27K protein might exist as a dimer. These data indicate that the 27K protein is a previously unrecognized T4-binding protein with a low affinity for the hormone. Further studies are required to clarify its physiological role in the transport of circulating thyroid hormones.

Endothelial cell injury in vitro is associated with increased secretion of an Mr 43,000 glycoprotein ligand

A novel, serum albumin-binding glycoprotein of molecular weight (mw) 43,000 (43K protein) was initially purified from the culture medium of bovine aortic endothelial (BAE) cells (Sage, H., Johnson, C., and Bornstein, P., J. Biol. Chem. 259:3993-4007, 1984). Its secretion by normal mesenchymal cells and by transformed cells of both ectodermal and endodermal origin suggested a general role in cellular function. To examine the effect of sublethal injury in vitro on the biosynthesis of 43K protein, BAE cells were exposed to endotoxin. At concentrations which produced minimal cell detachment and lysis, the cells secreted 70-100% more protein compared to control cultures, and the relative increase in 43K protein over total protein was approximately three-fold. A second type of cellular injury, manifested by rapid cellular proliferation and migration in response to sparse plating density (a condition that we have termed 'culture shock'), was also accompanied by a significant increase in the secretion of 43K protein. Pulse-chase studies revealed that the initial product secreted within 1.5 h was of Mr 38,000, and that between 6 and 21 h this molecule was converted to the final form of Mr 43,000. The 43K protein was not associated with RNA or glycosaminoglycan, but appeared to be linked to complex oligosaccharides containing peripheral sialosyl residues. Treatment with tunicamycin produced lower mw forms that displayed reduced affinity for albumin. By immunologic criteria, peptide mapping, and amino acid analysis, the 43K protein was shown to be structurally distinct from several proteins of Mr 40,000-50,000 associated with endothelium or with serum, including tissue factor, a plasminogen anti-activator, and several apolipoproteins. In addition, the 43K protein was not present in the extracellular matrices of endothelial, fibroblastic, or smooth muscle cells, nor was it found in plasma, serum, platelet releasate, or alveolar lavage fluids. These studies identify a unique Mr 43,000 glycoprotein that is associated with cellular stress or injury in vitro. As a secreted but nonmatrix macromolecule, this protein may be part of a 'survival kit' used by the endothelium to cope with cellular injury.