Two-dimensional gel electrophoretic analysis of cytoplasmic proteins from Friend erythroleukemia cells chemically induced to undergo terminal erythroid differentiation.

The proerythroblastoid Friend erythroleukemia cell (FELC) line, clone TR 19-9, was treated with 4 mM hexamethylene bisacetamide (HMBA) over a 6-day period. Greater than 94% of the FELC reacted positively to the benzidine assay for hemoglobin by Day 4 of treatment. Protein accumulation during the final 4 days of treatment (from Days 2 to 6) was monitored by labeling for 24-h periods with a 14C-labeled amino acid mixture. At the end of each radiolabeling time point, cells were harvested and cytoplasmic proteins were isolated and subjected to two-dimensional gel electrophoresis in triplicate. Short-term fluorographic exposures were made in the linear X-ray film response range to monitor those polypeptides which were most rapidly accumulated. Fluorographs were digitized for computer image analysis and gel data comparison rationales were used to combine the polypeptides contained on the replicate fluorographs into a single cytoplasmic polypeptide profile or Master Image for each of the two experimental conditions, control and HMBA-treated FELC. These two images were merged into a single Master Composite Image containing a total of 211 polypeptides so that those polypeptides common to both and/or unique to each of the experimental conditions could be viewed graphically in the same plane. A total of 98 polypeptides in HMBA-treated FELC were shown to have large accumulation rate differences from the control FELC;32 of these polypeptides were present in the HMBA Master Image which were not detected in the Control Master Image and 66 polypeptides were present in the Control Master Image but not detected in the HMBA Master Image. Five polypeptides, found in both Master Images, were shown to vary quantitatively in the HMBA-treated FELC from the corresponding polypeptides in the control. These quantitative data measurements on the rates of accumulation of various common polypeptides offer a mode for simultaneously monitoring the kinetics of induction and repression of many gene products throughout an experimental time course.

Binding of benzo(a)pyrene derivatives to specific proteins in nuclei of intact hamster embryo cells.

In hamster embryo cells incubated for 24 hr with 4 microM [3H]benzo(a)pyrene, a major portion of the nonextractable radioactivity in nuclear preparations copurifies with the protein fraction. When these proteins are analyzed by sodium dodecyl sulfate:polyacrylamide gel electrophoresis, significant variations in the labeling intensities of the various proteins are seen. Control experiments demonstrate that the labeling is due to covalent binding to protein. Histones H3 an H2A are heavily labeled while the other histones of the nucleosome core, H2B and H4, are devoid of radioactivity. Large amounts of label are associated with proteins with mobilities similar to the very lysine-rich histones H1. However, the results of differential extraction experiments suggest that the labeled proteins do not belong to either the H1 or the high-mobility-group class of chromosomal proteins. During 6 hr of inhibition of protein synthesis by cycloheximide, the metabolism of [3H]benzo(a)pyrene, as monitored by high-pressure liquid chromatography, remained normal. Patterns of labelling of nuclear proteins after 3 or 6 hr were identical in the presence and absence of cycloheximide. This finding strongly suggests that binding of benzo(a)pyrene derivatives to nuclear proteins occurs in situ.

Benzo[a]pyrene metabolism in human T 47D mammary tumor cells: evidence for sulfate conjugation and translocation of reactive metabolites across cell membranes.

Human mammary tumor T 47D cells were examined for their capacity to metabolize benzo[a]pyrene (BaP) to gain insight into potential metabolic pathways for BaP in human epithelial tissue. Confluent cultures metabolized 95% of 4 microM BaP after 24 h incubation. BaP metabolites were analyzed from culture medium since only residual amounts remained in cells. Tetraols/triols, dihydrodiols, quinones and phenols were either unconjugated or existed as sulfate conjugates. Glucuronide conjugation was minor. Remaining water-soluble (WS) metabolites could be extracted with butanol or removed from culture medium protein with methanol/water and suggestive evidence indicates these may be glutathione conjugates. A portion of BaP-WS metabolites were covalently bound to medium protein. This latter phenomenon is attributed to translocation of reactive BaP metabolites across cell membranes which could potentially occur in vivo during cellular processing of BaP.

Time course of metabolism of benzo[e]pyrene by hamster embryo cells and the effect of chemical modifiers.

In cultures of hamster embryo cells, benzo[a]pyrene (B[a]P) is metabolized primarily in the bay region. In contrast, little or no bay region metabolism of the noncarcinogenic isomer benzo[e]pyrene (B[e]P) could be detected during 12--96-h incubations of hamster embryo cells with 4 microM [3H]B[e]P. The upper limit to 9,10-dihydro-9,10-dihydroxy-B[e]P formation is about 0.2% of the ethyl acetate-soluble metabolites (less than 0.1% of the total metabolites). The major identified metabolites of B[e]P were 4,5-dihydro-4,5-dihydroxy B[e]P and the glucuronide conjugates of 3-OH-B[e]P and 4,5-dihydro-4,5-dihydroxy B[e]P. Simultaneous treatment of cells with either B[a]P or 7,8-benzoflavone (BF) did not induce bay region metabolism of [3H]B[e]P.

Benzo(a)pyrene-binding proteins of hamster embryo cell nuclei: comparison of nuclear isolation procedures.

Hamster embryo cells metabolize benzo(a)pyrene to derivatives that covalently modify the nuclear macromolecules including proteins. Not all proteins are modified to the same extent nor by the same metabolites. In particular, a protein of apparent molecular weight 32,000 is highly modified by derivatives of trans-9,10-dihydro-9,10-dihydroxy B(a)P. This protein is shown here to be preferentially lost from nuclei during purification by centrifugation through high molarity sucrose solutions followed by osmotic shock. It does not appear to be a cytoplasmic contaminant, but shares many properties of an abundant protein from Xenopus laevis oocytes, nucleoplasmin.

Two-dimensional gel electrophoretic comparison of cytoplasmic proteins in C3H10T1/2 cells, a chemically transformed 10T1/2 line, and a transformation resistant C3H mouse ventral prostate cell line.

The cytoplasmic proteins from three cell lines derived from the C3H mouse were compared after treatment with benzo(a)pyrene by two-dimensional gel electrophoresis and computerized image analysis. The three C3H lines were the transformable 10T1/2, the transformation resistant CVP and the methylcholanthrene transformed 10T1/2Cl line. Specialized algorithms were used to analyze gel images to record and compare individual proteins in the three cytoplasmic preparations. Multiple replicate gels were used to construct a master image to insure all the proteins were represented in the analytical system. In studies designed to examine the efficacy of utilizing image analysis, benzo(a)pyrene treatment caused significant alteration in the expression of numerous proteins in all three cell lines. Comparative analysis between cell types also showed most of the induced and repressed proteins were unique to a given cell line and did not match the induced or repressed proteins in either of the other cell lines. These results suggest that the response to chemical carcinogen treatment may be somewhat cell-specific and result in a variable response to the cells ability to respond to the chemical insult.

Specificity in interaction of benzo[a]pyrene with nuclear macromolecules: implication of derivatives of two dihydrodiols in protein binding.

Benzo[a]pyrene (B[a]P), 7,8-dihydroxy-7,8-dihydro-B[a]P, and 9,10-dihydro-B[a]P are metabolized by hamster embryo cells to derivatives that bind to nuclear macromolecules. The selectivity for different classes of macromolecules varies depending on the compound analyzed. The ratio of DNA specific activity to protein specific activity (pmol bound/mg of macromolecules) is high (1.51) for 7,8-dihydroxy-7,8-dihydro-B[a]P, extremely low (0.03) for 9,10-dihydroxy-9,10-dihydro-B[a]P, and intermediate (0.26) for B[a]P. Histones H3 and H2A are the major targets of 7,8-dihydroxy-7,8-dihydro-B[a]P; a protein(s) with a mobility similar to that of histone H1 is heavily labeled by 9,10-dihydroxy-9,10-dihydro-B[a]P, with minor labeling of other (nonhistone) bands. The labeling pattern seen with B[a]P is a combination of the patterns seen with the two dihydrodiol metabolites studied. Analysis of the ethyl acetate-soluble metabolites suggests that hamster embryo cells produce 9,10-dihydroxy-7,8-oxy-7,8,9,10-tetrahydro-B[a]P from 9,10-dihydroxy-9,10-dihydro-B[a]P and raise the possibility that this vicinal diol epoxide is an intermediate in the binding of 9,10-dihydroxy-9,10-dihydro-B[a]P to nuclear proteins. The differences seen suggest that factors other than the intrinsic chemical reactivity of the epoxide group are extremely important in the interaction of potential ultimate carcinogens with biological systems.

Comparison of 14C-amino acid mixture and [35S]methionine labeling of cellular proteins from mouse fibroblast C3H10T1/2 cells by two-dimensional gel electrophoresis.

Total cellular proteins from mouse C3H10T1/2 fibroblasts were compared by two-dimensional (2-D) gel electrophoresis after radiolabeling with [35S]methionine (35S-Met) or 14C-amino acids (14C-AA). 35S-Met labeling of protein was three to four times greater than 14C-AA incorporation over a 24 h period. Automated comparative analysis of replicate fluorographs after 6, 12, and 24 h of labeling showed considerable homology between radiolabeling methods. More than 88% percent of 35S-Met and 14C-AA-labeled proteins were common at each time point. However, the total number of 35S-Met-labeled proteins dropped from 6 to 24 h while the number of 14C-AA-labeled proteins increased. Additionally, twenty-one proteins were uniquely labeled by 14C-AA that were not detectable by 35S-Met over the labeling period. Densitometric analysis showed that several 35S-Met and 14C-AA-labeled proteins exhibited time-related differences in radiolabel incorporation while most proteins remained relatively constant. Protein patterns of silver-stained gels from 6 to 24 h were highly registered and showed few qualitative differences. Proteins detected in radiolabeled gels were generally, but not always, found in silver-stained gels. Thus, 35S-Met appears better suited for short-term radiolabeling of cellular protein while more comprehensive labeling of protein occurs with 14C-AA during prolonged incubation of cell cultures under present experimental conditions.