Partial Peptide Mapping Research Articles

Human factor VIII from heparinized plasma. Purification and characterization of a single-chain form.

Human factor VIII was purified from heparinized blood by cryoprecipitation, poly(ethyleneglycol) precipitation, Affi-Gel blue, aminohexyl, polyelectrolyte E5 and immunoaffinity chromatography. A purification of 280,000-fold over plasma with a specific activity over 5300 units/mg was achieved. Analyses of factor VIII using HPLC indicated a molecular mass of 280-340 kDa. Variation in the native mass may reflect heterogeneity of the protein due to associated lipid since structural analysis confirmed that factor VIII contained variable amounts of free fatty acids and diglycerides and triglycerides, but no phospholipids. Additional characterization by denaturing polyacrylamide gel electrophoresis under reducing conditions, followed by silver staining, showed a major single-chain polypeptide of factor VIII with a mass of approximately 260 kDa. To determine whether proteolyzed forms of factor VIII were present during fractionation, we analysed earlier steps in purification. This revealed additional species of factor VIII eluting faster than the single-chain form during chromatography on polyelectrolyte E5. Gel electrophoresis showed that these species of factor VIII consisted of multiple polypeptide chains, and partial peptide mapping using Staphylococcus aureus V8 protease indicated that they were structurally related. Monoclonal and hemophilic antibodies were used in immunoadsorption experiments to demonstrate that the purified factor VIII was composed predominantly of the 260-kDa factor VIII chain.

Proteins associated with rabbit reticulocyte mRNA caps during translation as investigated by photocrosslinking.

This laboratory previously detected by UV crosslinking a number of proteins associated with cytoplasmic mRNA in mammalian cells, and the data suggested that they are involved in translation. To find out which proteins are associated with caps we made use of reticulocyte mRNA specifically labeled in the cap with 32P together with a cell-free translation system and UV crosslinking. Approximately 8 bands corresponding to proteins crosslinked to the cap itself have been detected by polyacrylamide gel electrophoresis after UV crosslinking and digestion with RNases or tobacco pyrophosphatase. All but one were specific for methylated caps. One was similar in size and partial peptide map to a cap-binding protein, CBP I, previously identified in other laboratories, and most of the others corresponded to proteins previously known to be associated with mRNA but not known to be associated with caps. The results suggest that most mRNA-associated proteins are associated with caps or poly(A). Also, the number of cap-associated proteins may be greater than previously suspected.

Identification and partial characterization of the major secretory protein of the long hyaline gland in the male grasshopper, Melanoplus sanguinipes

In the adult male migratory grasshopper, Melanoplus sanguinipes, the major secretory product of the long hyaline gland (a distinct tubule type of the accessory reproductive gland complex) is an abundant glycoprotein (LHPI) with a native molecular weight of about 72,000 Da. Sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE) of this protein gives rise to a 36,000 Da subunit. Partial peptide mapping of the 36,000 Da subunit, using the agents N- chlorosuccinimide , papain, and staphylococcal V8 protease, shows that this polypeptide is found in a viscous secretion and, to a much lesser extent, in empty spermatophores, both of which are recoverable from between the female's ovipositor valves after mating. Polyclonal antibodies raised against the LHPI 36,000 Da subunit electrophoretically purified from long hyaline glands, cross react with similar molecular weight counterparts in viscous secretion and spermatophore extracts. LHPI is present in 1-day-old adults and accumulates up to levels of about 1 mg in 17-day-old virgins. LHPI is the single most abundant protein in the long hyaline gland accounting for 37–51% of the total protein in virgin males 5–17 days old. An estimated 160 μg was lost from the gland in a single mating on day 7 of which 110 μg was recovered at the completion of mating. Mating on day 5, 6, or 7 resulted in a 1.6–1.8 fold increase in rate of in vivo incorporation of [ 3H]leucine into LHPI measured on day 7. This apparent increase in LHPI synthesis is paralleled by increased accumulation of protein in the gland in mated males.

Eukaryotic elongation factor Tu is present in mRNA-protein complexes

By two-dimensional gel electrophoresis, partial peptide mapping, and antibody binding we have shown that eukaryotic elongation factor Tu is in close contact with mRNA in rabbit reticulocytes. It can be crosslinked to mRNA by irradiating both polysomes and 40–80 S mRNA-protein complexes with short-wave UV light. To our knowledge this is the first case in which a known translation factor has been shown to be associated with mRNA in native ribonucleoproteins.

Monoclonal antibodies that recognize lacto-N-fucopentaose III (CD15) react with the adhesion-promoting glycoprotein family (LFA-1/HMac-1/gp 150,95) and CR1 on human neutrophils.

A variety of monoclonal antibodies has been used to study the roles of surface proteins in neutrophil function. Many monoclonal antibodies that bind to human neutrophils react with the oligosaccharide lacto-N-fucopentaose III. Sequential immunoprecipitation of radiolabeled proteins from extracts of neutrophils labeled at the cell surface with 125I, and partial proteolysis peptide mapping studies were used to compare the proteins recognized by several widely used monoclonal antibodies that react with human neutrophils. The monoclonal antibodies that react with lacto-N-fucopentaose III (CD15) immunoprecipitated five distinct neutrophil surface proteins. The data indicate that CD15 monoclonal antibodies react with a subset of the LFA-1/HMac-1/gp 150,95 glycoprotein family as well as with CR1 on human neutrophils. The CD15 antibodies studied differed in their avidities for these proteins. The molecules immunoprecipitated by the CD15 antibodies tested were more resistant to proteolysis than the homologous proteins immunoprecipitated by the other monoclonal antibodies studied that react directly with the alpha M (CD11) or beta (CD18) chains of the LFA-1/HMac-1/gp 150,95 glycoprotein family. Some of the differences in antibody reactivity and protease sensitivity of the membrane proteins recognized by these antibodies may be due to differences in glycosylation. The data suggest that the antibodies studied can detect differences in post-translational modification among copies of certain surface proteins.

The photochemical reaction center of Chloroflexus aurantiacus is composed of two structurally similar polypeptides.

A method has been devised which allowed the isolation of highly purified reaction center from the thermophilic green bacterium, Chloroflexus aurantiacus. The procedure consisted of three chromatography steps. The final step was fast protein liquid chromatography on Mono Q in the presence of nonanoyl-N-methylglucamide (Mega-9). The purified reaction center complex was photochemically active and had an A280/A813 of 1.4 or less. Under non-denaturing conditions, a pigmented protein band having a Mr of 52,000-55,000 was observed in sodium dodecyl sulfate gels. When the isolated complex was heat-dissociated in the presence of sodium dodecyl sulfate, just two polypeptides having very similar Mr (24,000 and 24,500) were observed. Two protein bands were also observed in two-dimensional isoelectric focusing/sodium-dodecyl-sulfate polyacrylamide gel electrophoresis; the PI values of the two polypeptides were 6.5 and 6.7. Partial peptide mapping of the two isolated subunits, using both enzymatic and chemical cleavage techniques, yielded almost identical patterns which indicated a high degree of sequence homology between the two polypeptides. The N-terminal amino acid sequences of the two polypeptides were identical and did not exhibit any homology to reaction center subunits of purple sulfur bacteria. The Chloroflexus reaction center is believed to be composed of one molecule of each polypeptide, the photoactive bacteriochlorophyll a dimer and, as accessory pigments, an additional bacteriochlorophyll a and three bacteriopheophytins. Hence, it appears to be the smallest photochemically active reaction center isolated to date.

Analysis of a cDNA clone expressing a human autoimmune antigen: full-length sequence of the U2 small nuclear RNA-associated B" antigen.

A U2 small nuclear RNA-associated protein, designated B'', was recently identified as the target antigen for autoimmune sera from certain patients with systemic lupus erythematosus and other rheumatic diseases. Such antibodies enabled us to isolate cDNA clone lambda HB''-1 from a phage lambda gt11 expression library. This clone appeared to code for the B'' protein as established by in vitro translation of hybrid-selected mRNA. The identity of clone lambda HB''-1 was further confirmed by partial peptide mapping and analysis of the reactivity of the recombinant antigen with monospecific and monoclonal antibodies. Analysis of the nucleotide sequence of the 1015-base-pair cDNA insert of clone lambda HB''-1 revealed a large open reading frame of 800 nucleotides containing the coding sequence for a polypeptide of 25,457 daltons. In vitro transcription of the lambda HB''-1 cDNA insert and subsequent translation resulted in a protein product with the molecular size of the B'' protein. These data demonstrate that clone lambda HB''-1 contains the complete coding sequence of this antigen. The deduced polypeptide sequence contains three very hydrophilic regions that might constitute RNA binding sites and/or antigenic determinants. These findings might have implications both for the understanding of the pathogenesis of rheumatic diseases as well as for the elucidation of the biological function of autoimmune antigens.

Maleyl-BSA and fucoidan induce expression of a set of early proteins in murine mononuclear phagocytes.

We examined the effect of maleyl-BSA on specific protein expression in murine peritoneal macrophages by radiolabeling treated macrophages with [35S]methionine followed by SDS-polyacrylamide gel electrophoresis. Such treatment induces the expression of a set of at least seven proteins (38, 42, 57, 65, 75, 80, and 85 kD). A similar set of proteins is also induced by treatment of macrophages with the algal polysaccharide fucoidan. The proteins resemble those induced in response to treatment of this same cell population with bacterial lipopolysaccharide (LPS), as judged by co-migration in both one- and two-dimensional electrophoresis. Two proteins induced by either LPS or maleyl-BSA (e.g., p57 and p85) show similar primary structure, as assessed by partial proteolytic peptide mapping confirming their identity. The induction of these proteins by maleyl-BSA is a transient phenomenon, being expressed as early as 1 hr after treatment and declining after 8 hr even in the continuous presence of the stimulus. The dose of maleyl-BSA required to induce the response varies to some extent with the protein in question, but agrees with the Kd for ligand-receptor binding. Chloroquine, which blocks the degradation of ligand, does not inhibit the induction of early protein synthesis. Whereas the induction of these proteins is blocked by inhibition of RNA synthesis with actinomycin D, the reversible inhibition of protein synthesis with cycloheximide during the induction phase does not prevent their expression. LPS, maleyl-BSA, and fucoidan previously have been shown to stimulate protease secretion and tumoricidal function in appropriately primed macrophages. The present findings now demonstrate that all three agents can also mediate the expression of early genes which may participate in the acquisition of functional competence.

The polypeptide P16 is a carboxy-terminal cleavage product of rat growth hormone in anterior pituitary and GH3 pituitary tumor cells.

P16 is a small polypeptide originally found in GH3 rat pituitary tumor cells whose expression is tightly linked to the expression of rat GH (rGH) at a genetic level. It is estimated to be 3-5 kilodaltons smaller than rGH and exhibits the same complex response to T3, dexamethasone, and insulin in GH3 cells as does rGH. P16 was also found in high but variable abundance in anterior but not posterior pituitary. To approach the question of whether it arises from a unique gene or derives instead from the rGH gene by a posttranscriptional mechanism, we have measured its structural relatedness to rGH by peptide mapping techniques. From partial peptide maps of rGH and P16 by V8 protease, it appeared that the two proteins were related by loss of a common, small peptide. Both proteins also contained many tryptic peptides in common. Cleavage by N-chlorosuccinimide at tryptophan residues showed that rGH and P16 both contained the same N-terminal peptide but differed in their C-termini. Hence, P16 differs from rGH by loss of an amino acid segment somewhere in the C-terminus. Charge calibration of two-dimensional gels indicated that P16 was more acidic than rGH by at least five negative charges. These observations taken together imply that rGH gives rise to P16 by a highly specific cleavage in the C-terminus mostly likely between residues 152 and 156. This region also harbors an alanine-leucine at which pro-rGH is cleaved to remove the 26 amino acid signal peptide.

Characterization of rabbit uterine estrogen receptor proteins by radioiodination and partial peptide mapping.

Two estrogen binding proteins (Mr = 50,000 and 65,000) were purified from rabbit uterine cytosol using an improved procedure for affinity chromatography on diethylstilbestrol-agarose. The estrogen receptors were radioiodinated while adsorbed to the resin using the lactoperoxidase or Bolton-Hunter techniques. After elution, the labeled receptors were utilized for peptide mapping studies and investigations of receptor function. Partial peptide mapping revealed strong homology between the Mr 50,000 and 65,000 proteins suggesting common structural features. Estrogen receptors labeled by the lactoperoxidase procedure were rendered unable to bind immobilized heparin or hormone; in contrast, the Bolton-Hunter labeling technique yields proteins that retain both their ability to bind hormone and to absorb on heparin-agarose. The development of these iodination methodologies appears useful for the investigation of both the structure and functional properties of the receptor proteins.

Translational strategies in potexviruses: Products encoded by clover yellow mosaic virus, foxtail mosaic virus, and viola mottle virus RNAs in vitro

We have examined the template properties of the genomic RNAs from three potexviruses: clover yellow mosaic virus (CYMV), foxtail mosaic virus (FMV), and viola mottle virus (VMV). All three RNAs encode a large (160,000 to 182,000 molecular weight) nonstructural protein when translated in the rabbit reticulocyte cell-free system. Only CYMV RNA is able to direct the synthesis of significant quantities of a product whose electrophoretic mobility, antigenic determinants, and partial peptide map resemble those of authentic coat protein. All three viral RNAs also encode a number of discrete polypeptide products whose molecular weights are intermediate between those of the large nonstructural proteins and those of their respective coat proteins. We have examined the kinetics of synthesis in vitro of these intermediate products. For each of the three viral templates, as well as for papaya mosaic virus RNA, pulse-chase experiments suggest that most, if not all, the intermediate polypeptides are incomplete (“paused”) or prematurely terminated precursors to the corresponding large nonstructural protein. A partial proteolytic map of the in vitro products encoded by CYMV RNA supports this interpretation. Our data, and those of others, are consistent with a model in which all potexviruses would encode a large nonstructural protein.

Synthesis and processing of bovine herpesvirus 1 glycoproteins

Four unique glycoproteins or glycoprotein complexes were recognized by a panel of monoclonal antibodies to bovine herpesvirus 1 (BHV-1), i.e., GVP 6/11a/16 (130,000-molecular-weight glycoprotein [130K glycoprotein]/74K/55K), GVP 7 (108K), GVP 3/9 (180K/91K), and GVP 11b (71K). The absence of any antigenic or structural relationship between GVP 11a and GVP 11b, which were previously identified as one glycoprotein, GVP 11, demonstrated that these two GVP 11 species are unique glycoproteins. GVP 3 and GVP 9 showed complete sequence homology, as shown by the identity of their antigenic determinants and by partial peptide mapping. This observation, as well as the ratio of their apparent molecular weights, indicated that GVP 3 (180K) is a dimeric form of GVP 9 (91K). GVP 6 and GVP 11a, as well as GVP 6 and GVP 16, showed at least partial sequence homology, since they shared several antigenic determinants and peptides. In addition, GVP 6, GVP 11a, and GVP 16 were derived from one primary precursor. These results, as well as the ratio of their apparent molecular weights, indicated that the GVP 6/11a/16 complex consists of two forms: one in which GVP 6 (130K) is uncleaved and the other one in which GVP 6 is cleaved and composed of GVP 11a (74K) and GVP 16 (55K), linked by disulfide bridges. An antigenically distinct precursor to each of the four BHV-1 glycoproteins or glycoprotein complexes was identified by monoclonal antibodies. These precursors, pGVP 6 (117K), pGVP 11a (62K), pGVP 7 (100K), pGVP 9 (69K), and pGVP 11b (63K) were sensitive to endo-beta-N-acetylglucosaminidase H treatment, indicating that they represent the partially glycosylated high-mannose-type intermediate forms generated by cotranslational glycosylation of the primary, unglycosylated precursors to GVP 6/11a/16, GVP 7, GVP 3/9, and GVP 11b, which were identified as having apparent molecular weights of 105,000, 90,000, 61,000, and 58,000, respectively. A new nomenclature for the BHV-1 glycoproteins, based on roman numerals, is proposed.

Cellular localization of Drosophila 83-kilodalton heat shock protein in normal, heat-shocked, and recovering cultured cells with a specific antibody

To gain insight on the possible functions of heat shock proteins (hsp's) in Drosophila, we have purified the 83-kilodalton hsp (hsp 83) from cultured cells and studied its intracellular localization by immunofluorescence in normal, heat-shocked, and recovering cells. The specificity of the antibody was assessed by one- and two-dimensional gel immunoblotting and by partial proteolytic digestion. The anti-hsp 83 antibody does not show any significant cross-reactivity with hsp's of different avian or mammalian cell lines, but cross-reacts with hsp's of similar molecular masses in other dipteran insects. The partial proteolytic peptide maps of Drosophila hsp 83 differ from those of mouse hsp 89 and chicken hsp 84. Immunoblotting of Drosophila Kc cells heat shocked at different temperatures indicates a maximal expression of hsp 83 at 33 degrees C. By immunofluorescence, hsp 83 is shown to have a strictly cytoplasmic localization. In unstressed cells, it is distributed in the entire cytoplasm with a slight enrichment in the perinuclear region. After heat shock, it seems to concentrate at the cell periphery close to the plasma membrane and it gradually redistributes to the whole cytoplasm during cellular recovery at normal temperatures.

The chlorophyll a/b-proteins of PS I and PS II are immunologically related

Polyclonal antisera were raised against three apoproteins of the light-harvesting chlorophyll a/b-protein complex of photosystem I (LHC I) from spinach. These antisera were found to cross-react with all four LHC I and both LHC II apoproteins on immunoblots. Affinity purification of the cross-reactive antibodies on either the ‘native’ LHC II complex or on the 28 kDa apoprotein of LHC II immobilized on nitrocellulose yielded antibodies which recognize the same apoproteins as the crude sera. Partial peptide mapping of LHC I and LHC II apoproteins with Staphylococcus aureus V8 protease and N-chlorosuccinimide showed different patterns for each polypeptide. These data demonstrate that the four polypeptides of LHC I are distinct from each other and from the two major polypeptides of LHC II but all six are immunologically related.

Purification of a 190 kDa protein from smooth muscle: relationship to talin

Several studies of vinculin-binding proteins have described a 190 kDa protein in chicken gizzard smooth muscle which binds radioiodinated vinculin. We have purified and studied the 190 kDa protein from chicken gizzard smooth muscle. By indirect immunofluorescence, an antiserum raised against the 190 kDa protein stains adhesion plaques (focal contacts), ruffling membranes, and fibrillar streaks on the dorsal and ventral surfaces of fibroblasts. Both the binding to vinculin and the location of the protein in fibroblasts are properties shared with talin, a 215 kDa protein in smooth muscle and fibroblasts. Because antisera against talin and the 190 kDa cross-react the relationship of these two proteins has been investigated further. Upon prolonged storage at 4°C, purified talin degrades into a 190 kDa fragment. A 190 kDa fragment is also generated from talin by the Staphylococcus aureus V-8 proteinase and by trypsin. Comparison of partial peptide maps of talin and the 190 kDa protein reveal that the proteins are very similar and when the 190 kDa fragment of talin is compared with the purified 190 kDa protein by partial proteolytic digestion no differences are found in the pattern of peptides generated. In addition, the amount of 190 kDa protein detected in muscle tissues excised from chick embryos can be drastically reduced if proteinase inhibitors are added to the tissue homogenates. We conclude that the purified 190 kDa dalton protein is a proteolytic fragment of talin. Although markedly reduced by proteinase inhibitors, detection of the 190 kDa protein is not completely abolished, suggesting that some talin may already be cleaved within living cells.

Isolation and characterization of three major larval serum proteins of the mediterranean fruit fly Ceratitis capitata (Diptera)

Three major larval serum proteins (MLSP-1,2 and 3) of the Dipteran species Ceratitis capitata have been isolated and characterized. The structure of these proteins was found to depend on the pH. At acidic pH, they form hexamers which dissociate above pH 6.5. Their dissociation pattern in the pH range 6.5–8.5 was studied by gel filtration analysis. MLSP-3 was found to be the most readily dissociated protein followed by MLSP-1 and 2. Our data suggest that, in vivo, these proteins associate randomly to both homo- and heterohexameric forms. Amino acid analysis and partial peptide mapping, indicated the high degree of homology in the primary structure of these proteins, especially between MLSP-1 and 2. Partial homology of these three proteins with MLSP-4, another major larval serum protein of C. capitata which has been isolated previously in our laboratory (Mintzas and Rebutsicas, 1984) was also found. The amino acid analysis suggested the presence of glucosamine in MLSP-1, 2 and 3 while neutral sugars were identified only in MLSP-1 and MLSP-4.

Demonstration of a relationship between talin and P235, a major substrate of the calcium-dependent protease in platelets.

Talin is a 225,000-Dalton protein we have purified from smooth muscle. In chick embryo fibroblasts talin is found in adhesion plaques (focal contacts), areas where the cell is closely apposed to the substratum. In comparison with other cytoskeletal proteins, we found talin to be unusually susceptible to proteolysis and have identified a 190,000-Dalton proteolytic fragment of talin in the immunoblots of many tissues. These observations raised the possibility that the cleavage of talin to this fragment has physiological relevance. One system that we have investigated in which significant proteolysis occurs is platelets. During platelet activation several high-molecular-weight proteins are cleaved to lower-molecular-weight forms. Here we demonstrate that talin is closely related to one of these platelet high-molecular-weight proteins, P235. The purification of talin is comparable to that developed for P235, and the two proteins have similar biophysical properties. In addition, antibodies raised against chicken gizzard talin recognize P235 in purified form as well as in crude platelet extracts. The platelet protein also resembles smooth-muscle talin in its susceptibility to endogenous proteolysis: P235 is rapidly cleaved to a 190-200 kD polypeptide by a calcium-activated protease found in platelet extracts. Moreover, partial proteolysis of P235 and talin with chymotrypsin, elastase, or trypsin also generates remarkably similar one-dimensional peptide maps. Because of their similar biophysical properties, immunological crossreactivity, and similar one-dimensional partial peptide maps, we conclude that P235 is the platelet form of talin.

Expression of High Molecular Weight Polypeptides by Carnation Mottle Virus RNA

SUMMARY In vitro translation of virion RNA from carnation mottle virus (CarMV) in a messenger RNA-dependent rabbit reticulocyte lysate (MDL) resulted in the synthesis of four virus-specific polypeptides of apparent M r 100000 (P100), 77000 (P77), 38000 (P38) and 30000 (P30). Partial peptide mapping experiments and in vitro translation in the presence of partially purified calf liver amber suppressor tRNA demonstrated that P30, P77 and P100 are a series of overlapping polypeptides generated by a double readthrough mechanism. In addition we report the infection of Chenopodium quinoa protoplasts with CarMV. The viral coat protein was detected in virus-infected protoplasts. Only one other infection-specific protein with an apparent M r of 100000 corresponded in size to any of the other in vitro translation products.

Identification and Analysis of Keratin Polypeptides from Isolated Rat Vaginal Epithelium*

Under the influence of estradiol, the vaginal epithelium undergoes a process of growth and differentiation known as cornification. To analyze the biosynthetic changes that underlie this phenomenon in rats, it was necessary to first isolate the vaginal epithelium from hormonally stimulated animals. This was accomplished by two different methods: enzymatic digestion with a trypsin/pancreatin solution, which produced a disaggregated population of cells from the living epithelial layers; or disruption of basement membrane adhesion with a chaotropic salt solution (2 M NaSCN), which produced intact, full thickness tissue sheets. Extended extraction of the cells and sheets with an alkaline urea buffer followed by analysis of these extracts on denaturing polyacrylamide gels revealed that both preparations contained six major polypeptides in the mol wt range of 50,000-60,000 daltons. Differential solubilization indicated that these proteins are keratin subunits, and amino acid analysis confirmed this for the 53,000-dalton polypeptide (53K) purified from preparative sodium dodecyl sulfate gels. On two-dimensional gel electrophoresis, 53K produced a group of three acidic spots (alpha, beta, and gamma), with isoelectric points of 4.6, 5.5, and 5.4, respectively. Partial peptide maps revealed that these spots are isoelectric variants of the same protein. Immunization of rabbits with 53K produced a keratin antiserum that was primarily directed against the immunogen and cross-reacted weakly with a 50,000- and a 57,000-dalton band. This work demonstrated the feasibility of obtaining well defined molecular probes from a hormone-responsive epithelial system.

Proteins Associated with Adaptation of Cultured Tobacco Cells to NaCl

Cultured tobacco cells (Nicotiana tabacum L. cv Wisconsin 38) adapted to grow in medium containing high levels of NaCl or polyethylene glycol (PEG) produce several new or enhanced polypeptide bands on sodium dodecyl sulfate-polyacrylamide gel electrophoresis. The intensities of some of the polypeptide bands (molecular weights of 58, 37, 35.5, 34, 26, 21, 19.5, and 18 kilodaltons) increase with increasing levels of NaCl adaptation, while the intensities of other polypeptide bands (54, 52, 17.5, and 16.5 kilodaltons) are reduced. Enhanced levels of 43- and 26-kilodalton polypeptides are present in both NaCl and PEG-induced water stress adapted cells but are not detectable in unadapted cells. In addition, PEG adapted cells have enhanced levels of 29-, 17.5-, 16.5-, and 11-kilodalton polypeptides and reduced levels of 58-, 54-, 52-, 37-, 35.5-, 34-, 21-, 19.5-, and 18-kilodalton polypeptide bands.Synthesis of 26-kilodalton polypeptide(s) occurs at two different periods during culture growth of NaCl adapted cells. Unadapted cells also incorporate (35)S into a 26-kilodalton polypeptide during the later stage of culture growth beginning at midlog phase. The 26-kilodalton polypeptides from adapted and unadapted cells have similar partial proteolysis peptide maps and are immunologically cross-reactive. During adaptation to NaCl, unadapted cells synthesize and accumulate a major 26-kilodalton polypeptide, and the beginning of synthesis corresponds to the period of osmotic adjustment and culture growth. From our results, we suggest an involvement of the 26-kilodalton polypeptide in the adaptation of cultured tobacco cells to NaCl and water stress.