Search

Human carcinoembryonic antigen (CEA) is a cell surface adhesion molecule member of the Immunoglobulin Superfamily (IgSF). Aberrant upregulation of CEA is a common feature found in a wide variety of human cancers such as colon, breast and lung. Previous in vitro and in vivo results have demonstrated that CEA can have tumorigenic effects including the inhibition of cell differentiation and anoikis, a specific type of apoptosis triggered by the absence of extracellular matrix-cell contacts. In the present work, we investigate the involvement of the caspase cascade in CEA-mediated inhibition of anoikis and the structural requirements for this signal. Expression of CEA and/or a chimeric protein consisting of the NCAM extracellular domain attached to the CEA-GPI anchor correlates with an early inactivation of caspase-9 and activation of the PI3-K/Akt survival pathway, and at later times, inactivation of caspase-8. The CEA-mediated caspase inactivation as well as activation of Akt was not observed by expression of a CEA molecule incapable of self-binding (DeltaNCEA). These results suggest that the intrinsic caspase pathway is involved in the inhibitory effects of anoikis by CEA and this signal is dependent on the presence of self-adhesive extracellular domains and a CEA-GPI anchor.

CEACAM1 (biliary glycoprotein or CD66a) is a member of the carcinoembryonic antigen (CEA) subgroup of the CEA family. Eleven RNA isoforms derived from the splicing of a single CEACAM1 gene have been described. Some of the CEACAM1 isoforms have been recognized by the CD66 antibodies in T and B lymphocytes, natural killer cells, granulocytes and epithelial cells in several human tissues. Although it is also present in soluble form in bile and serum, and elevated levels have been found in the serum of patients with liver diseases, it is not known which isoforms are primarily involved. In order to learn more about the distribution and properties of particular CEACAM1 isoforms, we have prepared a monoclonal antibody specific for the A2 domain of CEACAM1, designated TEC-11. This antibody does not cross-react with other members of the CEA family. Immunoblotting analysis revealed that the TEC-11 epitope was present in all cell types expressing CEACAM1 containing the A2 domain [CEACAM1(A2)], including granulocytes (160 000 MW isoform) and sperm cells (140 000 MW isoform). A 115 000 MW isoform of CEACAM1(A2) was present in human serum, bile, saliva and seminal fluid. Human bile, saliva and seminal fluid also contained the 160 000 MW CEACAM1(A2) isoform. Significantly higher serum levels of the 115 000 MW CEACAM1(A2) isoform were detected in patients with obstructive jaundice. The 160 000 MW isoform of CEACAM1(A2) in bile, but not a 115 000 MW isoform in serum and bile, carried the 3-fucosyl-N-acetyl-lactosamine moiety. The combined data indicate that various isoforms of CEACAM1(A2) are present in different body fluids where they could take part in different CEACAM1-mediated functions.

Human carcinoembryonic antigen (CEA), a widely used tumor marker, and CEACAM6 [formerly nonspecific cross-reacting antigen (NCA)] are up-regulated in many types of human cancers, whereas family member CEACAM1 [formerly biliary glycoprotein (BGP)] is usually down-regulated. Deregulated overexpression of CEA/CEACAM6 but not CEACAM1 can inhibit the differentiation and disrupt the polarization and tissue architecture of many different types of cells. In this report, we show that CEA and CEACAM6, but not CEACAM1, markedly inhibit the apoptosis of cells when deprived of their anchorage to the extracellular matrix, a process known as anoikis. By blocking this tissue architecture surveillance mechanism, the architectural perturbation initiated by CEA/CEACAM6 can thus be maintained.

Human carcinoembryonic antigen (CEA) is overexpressed in a wide variety of epithelial malignancies including colon cancer. CEA can function in vitro as a homotypic intercellular adhesion molecule and can block the terminal differentiation of rodent myoblasts, thus raising the possibility that deregulated expression of CEA might directly contribute to malignant progression. To address this question, the expression pattern and cell-surface levels of CEA were studied during malignant transformation of the colonic epithelium in sporadic and familial adenomatous polyposis-related neoplasms. The level of immunohistochemically detected CEA was higher in 30% to 62% of microadenomas and small adenomas from familial adenomatous polyposis patients compared with adjacent normal mucosa, and this proportion was positively correlated with lesion size and degree of dysplasia. Cytofluorometric analysis of highly purified single epithelial cell suspensions from freshly excised carcinomas versus adjacent normal tissue demonstrated up to a 20-fold increase of mean cell-surface CEA in a group of colon carcinomas representative of the overall majority of such tumors--from Dukes' stages A to D and ranging mainly from well to moderately differentiated, the degree of overproduction was inversely correlated with tumor differentiation and directly correlated with stage. A marked tendency toward nonpolarized versus apical cell-surface expression with progression was also noted. Nonspecific cross-reacting antigen (NCA), a CEA family member, is also a homotypic adhesion molecule and blocks terminal myogenic differentiation, whereas biliary glycoprotein is a CEA family adhesion molecule that does not. Cell-surface NCA showed even greater overexpression (up to 70-told) in dedifferentiated tumors, whereas total-cell biliary glycoprotein showed approximately 2-fold lower levels than was normal in more differentiated tumors and approximately 2-fold higher levels than in further progressed tumors. These results therefore support the suggestion that CEA and NCA can directly contribute to colon carcinogenesis by inhibiting colonocyte differentiation.

The immunoglobulin superfamily represents an ancient, highly diversified group of cell surface and extracellular molecules responsible for a wide range of molecular and cellular recognition functions. The human carcinoembryonic antigen (CEA) subfamily of the immunoglobulin superfamily presents evidence of continuing diversification of the immunoglobulin family, in that some of its members, including CEA itself and nonspecific cross-reacting antigen (NCA), are expressed only in primates and not in rodents. These "new" members are glycophosphatidylinositol linked to the external cell membrane and are up-regulated in cancer, unlike members present in both rodents and primates, i.e., biliary glycoprotein (BGP), which are transmembrane linked and down-regulated in cancer. CEA, NCA, and BGP have all been shown to function in vitro as intercellular adhesion molecules. We show here that the properties of adhesion are radically different, in that BGP-mediated adhesion is reversibly Ca2+ and Mg2+ dependent, temperature dependent, and ATP inhibitable, whereas CEA- and NCA-mediated adhesion is the opposite in all aspects. Also, the novel double-reciprocal, antiparallel binding observed for CEA-CEA interactions is not seen for BGP. Finally, the myogenic differentiation block demonstrated for the ectopic expression of CEA in myoblasts was also observed for NCA but not for BGP, which is consistent with the changes in expression seen in cancer. The appearance of new CEA family members with such different properties is discussed in the context of evolution and cancer.

Human carcinoembryonic antigen (CEA) belongs to a family of membrane glycoproteins that are overexpressed in many carcinomas; CEA functions in vitro as a homotypic intercellular adhesion molecule and can inhibit differentiation when expressed ectopically in myoblasts. The regulation of expression of CEA is therefore of considerable interest. The CEA gene promoter region between -403 and -124 base pairs upstream of the translation initiation site directed high levels of expression in CEA-expression SW403 cells and was 3 times more active in differentiated than in undifferentiated Caco-2 cells, correlating exactly with the 3-fold increase in CEA mRNA seen in differentiated Caco-2 cells. Inclusion of additional upstream sequences between -1098 and -403 base pairs repressed all activity. By in vitro footprinting and deletion analyses, four cis-acting elements were mapped within the positive regulatory region, and one element within the silencing region. Several nuclear factors binding to these domains were identified: USF, Sp1, and an Sp1-like factor. By co-transfection, USF directly activated the CEA gene promoter in vivo in both SW403 and Caco-2 cells. In addition, the levels of factors binding to each positively acting element increased dramatically with differentiation in Caco-2 cells. Thus the transcriptional control of the CEA gene depends on the interaction of several regulatory elements that bind multiple specific factors.

The levels of control involved in the regulation of expression of the carcinoembryonic antigen (CEA) gene family members were investigated using cultured Caco-2 cells that differentiated after confluence as characterized by the production of polarized monolayers and the subsequent appearance of domes. Three transcripts representing CEA (3.0 and 3.5 kilobases) and nonspecific cross-reacting antigen (NCA) (2.6 kilobases) were detected in Northern analyses of mRNA preparations of such cells when probed with human CEA cDNA, albeit at different levels. The major CEA 3.0-kilobase transcript increased 3-fold over an 11-day culture period after confluence, whereas the NCA transcript increased 25-fold over the same time period. gamma-interferon treatment enhanced CEA mRNA levels 32-fold and NCA mRNA levels 53-fold in Caco-2 monolayers 11 days after confluence. The NCA gene thus appears to be regulated by a mechanism different from that of CEA. During gamma-interferon treatment, the normal increase in Caco-2 dome formation with time in culture was increased further by a factor of 2. Over the 13-day time span for Caco-2 cultures, CEA protein levels increased 7-fold, NCA (Mr 48,000) protein 5-fold, while gamma-interferon treatment augmented CEA 18.5-fold further and NCA 20-fold. In 2 other colon carcinoma cell lines, SW1222 and T84, which are differentiated in culture to varying degrees, little if any changes were seen in CEA and NCA mRNA and protein levels in pre- versus postconfluent cultures. Both cell lines, however, responded to gamma-interferon treatment by increases in CEA and NCA mRNA levels and, in some cases, disproportionate increases in the corresponding proteins. The lack of direct proportionality between mRNA and protein expression suggests that, as observed in human colon carcinoma and adjacent normal tissue, both transcriptional and post-transcriptional control mechanisms regulate CEA gene family member expression in colon carcinoma cells.

A repetitive mammalian genetic element, HSAG-1, has been shown to promote the amplification of the vector, pSV2-DHFR, containing the functional cDNA for dihydrofolate reductase (DHFR). LR-73 cells, a line of Chinese hamster ovary cells, were transfected with this recombinant construct or with the parent vector, then subjected to culture in selective medium containing steadily increasing concentrations of methotrexate, a drug which specifically inhibits DHFR. Cultures transfected with the HSAG-1-containing construct acquired drug resistance faster than those transfected with the parent vector. This acceleration of acquisition of drug resistance was due to an increased probability of the generation and subsequent selection of cellular variants with increased copy numbers of the vector. The effect has also been observed in CHO(DHFR-) and HeLa cell lines. Possible mechanisms for the effect of the HSAG-1 element on gene amplification are discussed.

Carcinoembryonic antigen (CEA) is a heavily glycosylated protein used clinically as a tumor marker to detect recurrences of many types of tumors. This glycoprotein belongs to the immunoglobulin superfamily and is the prototype of the large CEA family of proteins. In a concerted effort to determine the function(s) of this family, we have been investigating a similar family of proteins in the mouse. In this paper, we report the characterization of a new mouse family member named mmCGM2; this gene product is highly homologous to the human biliary glycoprotein of the CEA gene family and to a rat hepatocyte ecto-ATPase. In vitro transcription, translation, and glycosylation experiments have revealed that the mmCGM2 cDNA encodes a glycoprotein of 42 kDA with a putative extracellular N-terminal domain and a C2-set type immunoglobulin domain. We have used this cDNA as a probe to detect many different transcripts (1.5-4.6 kilobases) in mouse adult tissues, some of which are specific to particular tissues, while others are expressed ubiquitously. After transfection of a plasmid bearing the mmCGM2 cDNA into mouse fibroblasts known to lack CEA-related gene expression, transfectant cell clones were chosen and used to investigate the adhesion properties conferred onto the cells. Cells expressing the mmCGM2 cDNA in a sense orientation aggregated in a calcium- and temperature-dependent fashion. Together with human biliary glycoprotein, the mmCGM2 gene product is the first member of the immunoglobulin superfamily to exhibit calcium-dependent adhesion. The constant tissue reorganization necessary to the differentiation of precise structures in tissues which express these gene family members (colon, liver, and uterus) implies the necessity of a variety of specific cell-cell contacts which could utilize the cell adhesion properties that we have demonstrated.

Intercellular adhesion molecules can be classified as Ca2+ dependent or Ca2+ independent. This classification has significant functional implications regarding cellular interactions. The best characterized Ca2(+)-dependent adhesion molecules, such as L-CAM or E-cadherin, belong to the family of closely related cell surface molecules called cadherins. On the other hand, those immunoglobulin supergene family members which function as adhesion molecules, such as neural cell adhesion molecule, have been found to be Ca2+ independent. In agreement with this generalization, we have recently shown that carcinoembryonic antigen (CEA) and nonspecific cross-reacting antigen (NCA), two closely related members of the CEA family, a subset of the immunoglobulin supergene family, function in vitro as Ca2(+)-independent adhesion molecules. In contrast, we show here that transfectants of a third member of the CEA family, biliary glycoprotein (BGP), also aggregate homotypically in suspension but require Ca2+ for aggregation. In addition, like the cadherins and unlike CEA or NCA or other adhesion molecules of the immunoglobulin supergene family, BGP transfectant aggregation requires physiological temperatures. Two forms of BGP, with three and two immunoglobulin C2-set domains, show Ca2(+)- and temperature-dependent adhesion, so that these properties do not reside in the third C2-set domain. The significance of this expression in the range of functional properties of the immunoglobulin supergene family and its CEA subset is discussed.

The immunoglobulin supergene family members have been shown to be involved in cell-cell recognition and interaction during cell growth and differentiation. Neural cell adhesion molecule, myelin-associated glycoprotein, and carcinoembryonic antigen (CEA) are immunoglobulin supergene family members which can mediate cell adhesion. We show here that nonspecific cross-reacting antigen (NCA), a closely related CEA family member, is found on the surface of rodent cells transfected with functional NCA complementary DNA in different glycosylated forms, all of which can be deglycosylated to an Mr 35,000 core protein. Furthermore, NCA can mediate Ca2(+)-independent, homotypic aggregation of these NCA-producing transfectant cells. Since CEA has three internal repeated C2-set, immunoglobulin-like domains, whereas NCA has one, only one such domain is required for the intercellular adhesive function. We also demonstrate that NCA- and CEA-producing transfectants can form heterotypic aggregates, whereas mixtures of CEA or NCA transfectants and neural cell adhesion molecule or long form-myelin-associated glycoprotein transfectants sort themselves out into homotypic aggregates. The results suggest that subsets of the immunoglobulin superfamily, such as the CEA family, can be used in both homotypic and heterotypic cellular interactions, whereas less closely related members of the family can be used to separate different cell types by strictly homotypic interactions.

A key attribute of the stringent response of bacteria is the rapid inhibition of ribosomal RNA synthesis mediated by unusual nucleotides in respnse to uncharged tRNA. The question as to whether mammalian cells show a stringent response analogous to that of bacteria was critically tested by the effective rapid amino acid starvation of both normal and transformed cells. Rapid starvation giving a high proportion of uncharged tRNA for leucine was produced within 7 minutes of expression of a nonleaky ts leucyl tRNA synthetase mutation in transformed CHO cells (tsH1) and in its normal growth control revertant (L-73). To control for the effect of temperature alone, ts revertants of tsH1 and L-73 were included in the study, and to control for effects due simply to the inhibition of protein synthesis, the translational elongation inhibitor cycloheximide was used. In addition, rapid starvation for histidine was effected by incubation of both the CHO cell lines and of freshly explanted normal Chinese hamster embryo fibroblasts in histidine-free medium containing high concentrations of histidinol. The rate of preribosomal RNA synthesis and the extent of its maturation to mature rRNA was measured using (3H-methyl) methionine as a donor of methyl groups during synthesis and methylation of pre-rRNA. There was no effect on pre-rRNA synthesis of the rapid generation of uncharged tRNA for 45 minutes for any of the cell types tested. A nonspecific inhibition of maturation of 18S rRNA and late (3 hour) inhibition of pre-rRNA synthesis was observed, but could be mimicked by the inhibition of protein synthesis to comparable levels with cycloheximide. Less severe amino acid starvation resulting in a more physiological inhibition of protein synthesis to 30% also had no specific effect on pre-rRNA synthesis and maturation. Intracellular nucleotide pools were also examined for the appearance of unusual nucleotides such as guanosine tetraphosphate or pentaphosphate and for changes in the levels of normal nucleotides after severe amino acid starvation. No such changes could be detected. We conclude that although mammalian cells may have some biochemical reactions which respond to uncharged tRNA, they do not possess a macromolecular control system analogous to the stringent response of bacteria.

A cell-cell hybrid of a Chinese hamster ovary (CHO) cell line and human peripheral blood lymphoblasts from a patient with B-cell chronic lymphocytic leukemia produces a surface antigen detectable by a monoclonal antibody. This surface antigen can be detected at significant levels on a fraction of chronic lymphocytic leukemia cells, but not on normal human lymphocytes from peripheral blood or bone marrow. Two different clones capable of transforming mouse cells to produce the surface antigen were isolated from a gene library of the hybrid cell by identification of small portions of the library containing functional genes, followed by detection of clones with human-specific reiterated sequences. Both clones also contain CHO-specific reiterated sequences and are therefore human-CHO recombinants. From the frequency of antigen-producing clones in the library, we estimate that they were present in the hybrid cell genome at a copy number of 100 to 1000.

A new dominant amplifiable selective system for use in bacterium-animal cell shuttle vectors was developed by the insertion of a 2-kilobase genomic fragment containing the cloned Escherichia coli gene for asparagine synthetase (AS) into the pBR322-simian virus 40 recombinant vector pSV2 so as to place the translational initiator codon for the bacterial AS about 1,000 base pairs downstream from the simian virus 40 early promoter. This new construct, pSV2-AS, retains bacterial sequences for transcriptional and translational initiation and so can express AS in bacteria. The construct can also complement AS- mutants of mammalian cells, giving AS+ transfectants capable of growth in medium lacking asparagine, with relatively high efficiency (about 300 colonies per microgram of DNA per 10(6) cells exposed). The vector can be amplified up to 100-fold in such AS+ transfectants by selection in asparagine-free medium containing increasing concentrations of the AS inhibitor beta-aspartyl hydroxamate. AS+ transfectants were found to be much more resistant to a second AS inhibitor, Albizziin, than were normal AS+ animal cell lines. This difference, which may indicate a strong resistance of the bacterial AS enzyme to Albizziin, was exploited to develop an effective selection for bacterial AS transfectants of a number of wild-type AS+ cell lines of rat, Chinese hamster, mouse, and human origin. LR-73 cells, a Chinese hamster AS+ cell line, were transfected with pSV2-AS with an efficiency of about 1,000 colonies per 0.5 microgram of DNA per 10(6) cells. The integrated construct in these cells was amplified by incubation of the transfectants in increasing concentrations of beta-aspartyl hydroxamate. Advantages and disadvantages of this new dominant, selectable, and amplifiable marker over markers commonly used in shuttle vectors are discussed.

T1026, a ts mutant of VSV which is much less cytopathogenic than its parent, HR, and which can establish persistent infection under certain conditions, is a double mutant. In addition to its ts mutation in the virion RNA polymerase, T1026 has a second non-ts mutation in a viral function termed "P". This function is responsible for the inhibition of total protein synthesis in infected cells and acts chiefly at the level of translational initiation. In some cell systems, the inhibition of protein synthesis produced by P appears to be selective for cellular protein synthesis, whereas in other cell systems, both cellular and viral protein synthesis are inhibited. T1026 and its ts revertants are phenotypically P- -that is, cells infected with them show total protein synthesis rates equal to or greater than uninfected cells, while synthesizing viral proteins at the same or even greater rates than HR-infected cells. The P- mutation is correlated with failure to increase plaque size after 2-3 days of incubation. Since viral mutants obtained from persistently infected cultures in a variety of systems appear to be double mutants with a ts mutation in the virion RNA polymerase and a small plaque marker, we suggest that T1026 could represent a model for such mutants.

A genetic system comprised of mammalian cell mutants which demonstrate concomitant resistance to a number of unrelated drugs has been described previously. The resistance is due to reduced cell membrane permeability and is correlated with the presence of large amounts of a plasma membrane glycoprotein termed P-glycoprotein. This system could represent a model for multiple drug resistance which develops in cancer patients treated with chemotherapeutic drugs. We demonstrate here that the multiple drug resistance phenotype can be transferred to mouse cells with DNA from a drug-resistant mutant and then amplified quantitatively by culture in media containing increasing concentrations of drug. The amount of P-glycoprotein was correlated directly with the degree of drug resistance in the transformants and amplified transformants. In addition, the drug resistance and expression of P-glycoprotein of the transformants were unstable and associated quantitatively with the number of double minute chromosomes. We suggest that the gene for multiple drug resistance and P-glycoprotein is contained in these extrachromosomal particles and is amplified by increases in double minute chromosome number. The potential use of this system for manipulation of mammalian genes in general is discussed.

To test the error catastrophe theory of aging we determined the error frequency of protein synthesis in several strains of cultured human fibroblasts at early and late passage. Error rates were calculated from analysis of native and substituted actins on two-dimensional gels of cellular proteins after induction of mistranslation by histidine starvation in the presence of histidinol. Early-passage cells from fetal, young, and old donors and cells from subjects with the Hutchinson-Gilford and Werner syndromes of accelerated aging had similar error frequencies. Late-passage cells from fetal, young, and old normal donors had similar or lower error frequencies than corresponding early-passage cells. No correlation was observed between error frequency, donor age, or maximal life span in vitro. We also examined an immortal cell line, simian virus 40-transformed W138 fibroblasts. These cells had a significantly elevated rate of mistranslation (2.8 +/- 0.2 x 10(-4))(+/- SEM) compared to their untransformed counterpart WI38 (0.6 +/- 0.1 X 10(-4)) or all diploid cells combined (1.1 +/- 0.1 x 10(-4)). Taken together, the data fail to support the error catastrophe theory of aging.

HSAG-1 is a cloned member of a heterogeneous middle repetitive family of genetic elements which is capable of eliciting a leukemia-related surface antigen detected with a monoclonal antibody after DNA transformation of mouse cells. HSAG-1 was originally isolated from a Chinese hamster-human leukemia hybrid cell gene library both by sib-selection for antigen producing activity and by hybridization with labelled human genomic human DNA. We show here that the human labelled site is at the right hand end of the insert, while the antigen-eliciting portion is included in a 1450 bp fragment at the left hand end of the insert. We also present the complete nucleotide sequence of the 3369 bp insert. The sequence contains 12 elements which bear a significant resemblance to accepted consensus sequences for Alu repetitive elements. The right hand end contains adjacent elements with close sequence similarity to portions of the human and hamster type I and type II Alu consensus sequences. All of the other Alu-related elements have diverged relative to the Alu consensus sequences by additions, long deletions and substitutions. The left hand portion of the insert which has the antigen-producing activity contains four of these diverged elements representing a relatively high proportion (26%) of the nucleotide sequence. The sequence is thus consistent with our previous observations of a repetitive family with biological function.

The antibiotic cycloheximide strongly inhibits protein synthesis in cell free systems (Bennett et al. , 1965) , in cultured cells (Ennis and Lubin, 1964) and in whole animals (Trakatellis et al. , 1965) . In general, the polysome-ribosome profiles obtained by velocity centrifugation analysis of cellular extracts show little or no change after drug treatment, possibly because the drug simply stops the attachment and movement of ribosomes along the messenger RNA (mRNA) in the polysomes (Wettstein et al. , 1964 , Colombo et al. , 1965 , and Trakatellis et al. , 1965) . In these studies, highly inhibitory drug concentrations were employed. This report describes the effect of less inhibitory doses of cycloheximide on the polysomes of hamster embryo cells grown in vitro . Under such conditions the free ribosomes in the cells disappear completely and become attached to mRNA in polysomes. The results are consistent with the hypothesis that low concentrations of the drug cause the ribosomes to move more slowly along the mRNA, but have less effect on ribosome attachment.

Differences in properties between normal hamster embryo cells and clonal lines of hamster embryo cells transformed with polyoma virus provided means of selecting for the growth of colonies of transformed cells in the presence of many normal cells. In reconstruction experiments where selective techniques were applied to artificial mixtures of small numbers of transformed cells and large numbers of normal cells, transformed cells could be detected at frequencies of 1 cell in 104 normal cells to 1 cell in 105 normal cells, depending upon the particular transformed line of cells used. Cells of one line, termed line Li, of particular interest because its properties were between those of normal cells and cells of the other transformed lines, could not be satisfactorily detected by any of the selective techniques employed. By means of reconstruction experiments, techniques were developed which succeeded in detecting cells of this line at a frequency of one cell in 104 normal cells. When the selective techniques were applied to cultures immediately after adsorption of virus, transformed cells were detected and selected for by some of the techniques, particularly those that were most successful for line Li. Thus new properties, such as a high relative plating efficiency in media containing low concentrations of serum, on which the techniques depend, are conferred upon cells very soon after infection with virus. The selective techniques succeeded in measuring frequencies of transformation too low to be measured by a nonselective colony assay, and thus provide the basis for an assay for transformation of considerably increased sensitivity.

Electrophoretic patterns of newly synthesized proteins have been compared for hamster embryo fibroblasts in asynchronous cultures, mitotically synchronized cultures, and stationary phase cultures. Only proteins with molecular weight between 30,000 and 150,000, comprising 60–70% of the total cell proteins and excluding histones and collagen were included in the comparison. Although no significant differences could be detected between such patterns for cells at different stages of the cell cycle, significant differences were detected between patterns for cells in stationary phase and for proliferating asynchronous or synchronous cells in any stage of the cell cycle. These differences amounted to at least 5% of the newly synthesized cellular proteins. Much larger differences were detected between patterns from a nuclear fraction of proliferating and resting cells. These results indicate that normal cells in stationary phase are arrested in a state distinct from any phase of the normal cell cycle and may provide a biochemical marker for resting cells.

Acute starvation of mammalian cells for amino acids results in translational errors that may be detected by two-dimensional polyacrylamide gel electrophoresis. Using this as an assay for error frequency in mammalian cells, we investigated the hypothesis that neoplastic transformation was associated with an increased error frequency which in turn leads to an increased mutation rate and a decreased efficiency of regulatory controls (phenomena of tumor progression). Although we found that transformation was not always associated with an increased level of mistranslation we showed that SV40 transformation increased the level of translational errors in all cell types tested.

Chinese hamster ovary (CHO) cells were subjected to severe amino acid starvation for histidine, leucine, methionine, asparagine, tyrosine, glutamine, valine, and lysine, using amino acid analogs or mutations in specific aminoacyl-tRNA synthetases. At protein synthetic rates of less than 5%, in all cases, the newly synthesized proteins were found on two-dimensional electrophoretic gels to consist of a few intensely labeled spots, with the exception of lysine. This pattern could also be produced by strong inhibition of cytoplasmic protein synthesis with cycloheximide, and was abolished by preincubation with the mitochondrial protein synthesis inhibitor chloramphenicol. It appears therefore that the spots represent mitochondrial protein synthesis and that animal cells must have separate aminoacyl-tRNA synthetases for mitochondrial tRNAs corresponding to all these amino acids except, possibly, for lysine.

ONLY mouse 28S ribosomal RNA (rRNA) can be detected in mouse-human hybrid cells1. Because these cells contain as many as thirty-five human chromosomes, including those believed to have ribosomal RNA genes, it has been suggested that the transcription of human ribosomal RNA genes may be repressed. Similar studies have not been performed with mouse-hamster hybrid cells because hamster and mouse 28S rRNAs have only a very small difference in their electro-phoretic mobilities2. We have now made use of the observation that free ribosomes (not engaged in translation) from hamster cells form dimers in conditions where mouse free ribosomes remain as monomers3 to determine whether mouse-hamster hybrid cells contain free ribosomes as monomers, dimers or as both.

When CHO cells are incubated under conditions of extreme amino acid starvation, effected by withdrawal of an amino acid from the medium together with genetic or chemical interference with the activity of the corresponding aminoacyl-tRNA synthetase, there is a rapid and profound decline in the functional capacity of the protein synthetic machinery. The effect was observed for all amino acids tested including leucine, asparagine, histidine, methionine and glutamine. This decline in protein synthetic potential appears to be due to a progressive permanent inactivation of the specific aminoacyl-tRNA synthetase concerned, as shown by a decline in the amount of cellular, specific aminoacyl-tRNA and a decline in the cell-free enzyme activity, measured after reversal of the starvation conditions. When cells are left for more than several hours under these starvation conditions, they shrink in size, lose viability and eventually disintegrate, with anomalous rapidity. We suggest that the progressive loss of protein synthetic capacity of the cells is the prime cause of these subsequent events. If the starvation conditions are reversed before cell death, regeneration of the protein synthetic potential occurs rapidly but requires protein synthesis itself, implying the existence of strong control mechanisms for cellular aminoacyl-tRNA synthetase activities.

The virulence of temperature-sensitive mutants of vesicular stomatitis virus (VSV) injected subcutaneously into newborn hamsters was positively correlated with their tendency to generate revertants and with their leakiness in cultured hamster embryo fibroblasts maintained at 37 degrees C, the measured body temperature of the animals under our experimental conditions. The complementation group of the mutants seemed important only in that it tended to determine reversion frequency and leakiness. One non-reverting group I mutant (T1026), however, was much less virulent than would be expected from its extreme leakiness at body temperature. The disease produced by the less virulent mutants was characterized by neurological symptoms and led to delayed death, unlike the rapid deatth produced by virulent mutants. Infectious virus could be found in higher titres in the brains than in peripheral organs of such animals (with ratios as high as 10(8)). This neurotropism was not correlated with the complementation group of the mutant but was shown to be the consequence of survival for more than 3 days after injection. Age was not responsible for the effect. Animals injected at birth with T1026 were completely resistant to subcutaneous superinfection with the highly virulent wildtype virus HR at 3 to 4 days, though non-T1026-protected animals were completely sensitive. When HR was injected intracerebrally at 3 to 4 days, the T1026-protected animals allowed replication to high titres in the brain but not in peripheral organs, whereas non-T1026-protected animals allowed replication to high titres in both brain and in peripheral organs. We suggest from these results that the observed neurotropism is produced by a resistance mechanism operative in peripheral organs but not in the brain; this resistance develops rapidly in newborn animals on exposure to virus and clears virus from the peripheral organs leaving it in the brain. It is possible that our effect represents a controlled and accelerated induction of the classical peripheral resistance of animals to various viruses which normally develops with age.

Acute starvation of mammalian cells for amino acids results in translational errors that may be detected by two-dimensional polyacrylamide gel electrophoresis. Using this as an assay for error frequency in mammalian cells, we investigated the hypothesis that neoplastic transformation was associated with an increased error frequency which in turn leads to an increased mutation rate and a decreased efficiency of regulatory controls (phenomena of tumor progression). Although we found that transformation was not always associated with an increased level of mistranslation we showed that SV40 transformation increased the level of translational errors in all cell types tested.

Using specific mutants as a means of identification, the bacterial protein for asparagine synthetase (Asn Syn) was shown to be antigenically and electrophoretically similar to its mammalian counterpart. This observation prompted us to attempt direct transfer of the cloned bacterial gene for the enzyme to mammalian cells. DNA from the replicative form of clone M13 OriC, containing the bacterial gene for Asn Syn, was shown to be capable of causing transformation of Jensen rat Asn Syn- cells to cells capable of growth in Asn-free medium; no prior modification of the bacterial gene was required. This relatively inefficient transformation (20 colonies/micrograms DNA/10(6) cells) was sensitive or insensitive to restriction enzyme digestion of the M13 OriC DNA in complete agreement with the known restriction map of the bacterial gene. Clones of transformed rat cells contained the bacterial DNA, which was amplified if increased levels of the enzyme were demanded and lost if selection was removed. The clones also contained polysomal bacterial RNA and a new protein with properties similar but not identical to those of the bacterial enzyme. The biological significance of this unusual degree of compatibility between the prokaryotic and eukaryotic Asn Syn gene systems is discussed.

Infectious particle production by temperature-sensitive (ts) mutants of vesicular stomatitis virus (VSV) was measured in a variety of different host cell types maintained in a state of quiescence or stimulated to proliferate. At permissive temperatures, all ts mutants and the wild-type virus replicated equally well and with the same kinetics in both quiescent and proliferating cells. At semi-permissive temperatures, however, Lts mutants, with temperature-sensitive virion polymerases, showed a delay of about 6 h in infectious particle production relative to wild-type virus in proliferating cells and greater than 16 h in quiescent cells. The effect was specific for the Lts class of mutants and was not seen for representative mutants in any of the four other complementation groups of VSV. Regarding cellular determinants, the effect was correlated only with the growth phase and not with the species of origin, interferon inducibility or with malignant transformation.

Summary Intracellular K+ ion concentration and transport were measured in L cells infected with wild type VSV, which rapidly inhibits total protein synthesis in infected cells, and with a mutant, R1, defective in this function. No alterations in intracellular K+ ion concentration or transport were observed until late in infection and the late changes seen were similar for both viruses. Thus the inhibition of total protein synthesis by VSV cannot be ascribed to virus induced changes in host K+ ion concentration.

The role of protein degradation in cellular proliferation was investigated by measurements of the rates of degradation of labile and stable proteins for a number of cell types under various growth conditions. The rate of protein degradation was found to be a relatively invariant parameter in that it did not change after strong inhibition of protein synthesis with cycloheximide or histidinol, it was the same in both exponential and stationary phase, and it did not correlate with the presence or absence of malignant tranformation. Using three different cell types with widely differing division rates, the rate of cell division and DNA synthesis (in %/hr) was found to be precisely equal to the rate of protein accumulation (in %/hr), i.e., to the rate of protein synthesis minus the rate of protein degradation. Division rates between the different cell types appeared to be determined chiefly by the rate of protein synthesis though, especially at low division rates, the rate of protein degradation could represent a large component of the protein accumulation rate.

Efficient selection procedures, using [3H]amino acids as the selecting agent, were developed for isolating temperature-sensitive (TS) mutations in CHO cells affecting protein synthesis. After chemical mutagenesis, leucyl-tRNA synthetase mutants were obtained when [3H]leucine was used as the selecting agent in two independent experiments. These mutations seem to involve the same genetic locus as the TSH1 mutant described previously (1). A selection with [3H]valine, in which all amino acids except leucine were at low concentration in the selective medium, resulted in a new class of mutants with reduced asparagyl-tRNA synthetase activity. These results were consistent with the finding that all mutants were phenotypically dependent on the concentration of amino acid, specific to the altered synthetase, in the medium. Our observations suggest that although leucyl synthetase mutations are a relatively common class of TS mutations in CHO cells, the spectrum of mutants obtained can be at least partially manipulated through concentrations of amino acids in selective media. The asparagyl-synthetase mutation was shown to be recessive and to complement the leucyl-synthetase mutation in cell-cell hybrids.

The control of expression of genes of the carcinoembryonic antigen family was investigated in 22 specimens of malignant and nonmalignant human colonic tissues. These surgical specimens included seven colonic adenocarcinomas that were compared with normal adjacent colonic mucosal tissues from the same individual. mRNA preparations from all colonic tissues expressed three bands of 3.5, 3.0, and 2.6 kilobases on Northern blots probed with carcinoembryonic antigen (CEA) complementary DNA probe while normal liver and spleen were negative. The major band of 3.0 kilobases was 6 to 10 times more intense in the colon tumor specimens than in the matched normal mucosa. However, the tumor/normal ratios of immunoreactive CEA in these pairs varied from 2- to greater than 100-fold. Furthermore, there was no direct proportionality between mRNA levels and gene product expression, suggesting that the known variations in CEA expression in human colonic tissues result from both transcriptional and posttranscriptional control mechanisms. Southern blots of DNA from these specimens did not reveal any gene rearrangements or amplifications accompanying expression. Finally, Southern blots of DNA digested with methylation-sensitive endonucleases and probed with a genomic DNA fragment upstream of CEA gene coding regions demonstrated that CEA expression is correlated with a decreased level of methylation in the 5' region of the CEA gene.

Functional human carcinoembryonic antigen (CEA)-like genes have been shown to be present in the mouse. Southern analyses of murine DNA using both human and murine CEA complementary DNA probes have revealed the presence of multiple CEA-like genes, while analyses of RNA from different mouse tissues showed CEA-like transcripts in adult colon and liver. Furthermore, a CEA-like protein, immunoprecipitable with a rabbit polyclonal serum raised against human CEA, has been detected in adult murine colon tissue. Several murine CEA complementary DNA clones have been isolated from a murine colon complementary DNA library, and characterization of one such clone demonstrates that both the N-terminal and the internal domains have been conserved between the two species. The existence of a murine counterpart of CEA strengthens the case for an essential function for this human tumor marker and provides an experimentally amenable system for elucidation of its biological properties.

Normal and diseased human tissues were analyzed for the transcription of genes of the carcinoembryonic (CEA) family. Epithelial tissues of colonic origin, whether malignant or normal, all express two closely related mRNA species of 3.0- and 3.5-kilobase mRNA which code for CEA. Only tissues of colonic origin were found to express these CEA-specific transcripts. Colon carcinomas consistently express a 2.6-kilobase mRNA species as well which codes for nonspecific cross-reacting antigen. Nonneoplastic colon mucosas, on the other hand, express lower or nondetectable levels of this transcript. Most breast carcinomas produce only the nonspecific cross-reacting antigen mRNA, whereas leukocytes of chronic myelogeneous leukemia express both nonspecific cross-reacting antigen mRNA and a 2.3-kilobase mRNA corresponding to a yet undefined gene of the CEA family. Thus the multiple CEA-like products reported to be produced by these tissues correspond to only four different mRNA species coding for three different peptides. These data suggest a less complex organization of the CEA family than was previously suspected and point to posttranscriptional modifications, such as variable patterns of glycosylation, as the likely reason for much of the observed complexity in CEA-like glycoproteins.

The fidelity of protein synthesis was measured in human diploid skin fibroblasts as a function of passage level (“aging in vitro”) and physiological age of tissue donor (“aging in vivo”) using two different test systems. First, in cell-free extracts the ratio of ∆ leu/∆ phe incorporation into peptide linkage following in the latter case using cells derived from elderly normal donors and from subjects with the premature aging disorders of Hutchinson-Gilford progeria and the Werner syndrome.Similar results were obtained using a second system of intact cells whereby histidine starvation induces quantifiable satellite spots resolved by two dimensional electrophoresis on polyacrylamide gels on the acidic side of the native actin species due to substitution of the neutral amino acid glutamine for the basic histidine. In fact, error frequencies appeared to decrease during aging in vitro, likely due to selection for clonal subpopulations with the highest fidelity of protein synthesis. The only increases were seen in the intact cell system where SV40-transformed cells showed a three-to-five fold greater error frequency compared to nontransformed fibroblasts. In total, these data fail to support the error catastrophe theory of cellular aging.

Publisher Summary This chapter presents a study in which DNA from a clone M13oriC81 containing bacterial asparagine synthetase is used to transform directly a rat cell line, Jensen sarcoma, to grow in asparagine-free medium containing B-aspartylhydroxamate. The efficiency of DNA transfer was about 2 to 20 colonies per 106 cells exposed per μg of DNA Southern blot analysis of the transformants showed that the M13oriC81 DNA was integrated into high molecular weight DNA. Unstable transformants that had lost their ability to grow in asparagine-free medium containing 3-aspartylhydroxamate also lost this M13oriC81 DNA. Transformants grown in increased amounts of 3-asparatylhydroxamate had increased amounts of M13oriC81 DNA. The resistance phenotype could be transferred using DNA from transformants and these second-step transformants also contained M13oriC81 DNA. Northern blot analysis showed that the M13oriC81 DNA, which was transcribed in RNA hybridized with M13oriC81 DNA, was detected in a polysome fraction obtained from transformants. Western blot analysis of two-dimensional gels of proteins of transformants using anti-asparagine synthetase antibody showed new spots with some of the properties of the bacterial enzyme. These results indicated that a bacterial gene can be introduced and expressed directly, if inefficiently, in mammalian cells.

The genetic approach to the problem of cellular growth control is limited by the availability of recessive mutations in cell lines which are capable of growth control in vitro. The CHO cell line has yielded many recessive mutations including, for example, tsH1, a temperature sensitive leucyl-tRNA synthetase mutant, which under non-permissive conditions rapidly shuts down protein synthesis and generates uncharged tRNA. Both CHO and tsH1 are transformed, however, and do not respond to environmental stimuli with the coordinated regulation of macromolecular processes observed in normal diploid fibroblasts. We describe here the isolation and characterization of growth control revertants obtained from both CHOwt and tsH1. The best of these GRC+L-73, isolated from tsH1, had 20 chromosomes, one less than tsH1, had normal fibroblastic morphology, would not grow in suspension, required high serum concentrations for growth, grew to relatively low cell densities at saturation in monolayer culture and showed a stationary phase characterized by arrest in a G1-like state with maintenance of high viability for several weeks. It is expected that this line as well as a ts revertant GRC+LR-73 will greatly facilitate the genetic investigation of growth control and, in particular, will help to elucidate the role of uncharged tRNA in the regulation of macromolecular synthesis in mammalian cells.

The cell-killing activity of asparaginase on three classes of Chinese hamster ovary cell mutants was examined: a mutant which overproduces asparagine synthetase (AH5); mutants defective in asparagine synthetase (N3 and N4); and mutants conditionally defective in asparagyl-transfer RNA synthetase (Asn 3, Asn 7, and Asn 9). The overproducer was more resistant to the cell-killing activity of asparaginase than wild-type Chinese hamster ovary cells, while mutants defective in asparagine synthetase were more sensitive. Surprisingly, the asparagyl-transfer RNA synthetase mutants were even more sensitive to asparaginase than the asparagine synthetase mutants. In a preliminary survey of four human lymphoid cell lines (RPMI 8402, RPMI 8392, B46M, and Molt-4F) which showed dramatically different asparaginase sensitivity, however, sensitivity to the cell-killing activity of asparaginase was correlated with reduced levels of asparagine synthetase and not with reduced levels of asparagyl-transfer RNA synthetase.

Plaques produced by our P − mutants of vesicular stomatitis virus (VSV), which are defective in the inhibition of total protein synthesis in infected cells, stop increasing in size after several days of incubation under conditions where those produced by P + mutants increase linearly in size. The basis for the arrest in size of P − plaques has been shown to be due to the induction of interferon, and the phenotype is termed PIF + for “plaque interferon positive.” Thus P − plaques can inhibit the increase in size of adjacent P + plaques and the factor responsible has the biological and physical properties of interferon. Also P − mutants, when plaqued on VERO cells which cannot be induced to produce interferon, produce plaques which increase linearly in size like P + plaques. Finally, the inclusion of anti-interferon antibody in the overlay medium also causes P − mutants to produce plaques like P + mutants. VERO cells were found to be useful to separate the effects of is mutations on plaque size from the interferon effect. Using other cell types the latter effect (PIF assay) can be used as an assay for the ability of viruses to induce interferon, for the isolation of PIF + mutants from PIF − virus, and as a test for the ability of cells to respond to interferon induction by PIF + viruses. The assay can be increased in sensitivity through the use of specific cell types and of cell cultures preincubated for several days in the stationary phase of growth. In its most sensitive form, the assay could detect PIF + behavior in certain ts mutants of VSV at permissive temperatures and in VSV mutants emerging from persistent infection. The assay has also been used to isolate novel mutants of VSV which show alterations in the viral P function.

The fidelity of protein synthesis was measured in human diploid skin fibroblasts as a function of passage level ("aging in vitro") and physiological age of tissue donor ("aging in vivo") using two different test systems. First, in cell-free extracts the ratio of delta leu/delta phe incorporation into peptide linkage following in the latter case using cells derived from elderly normal donors and from subjects with the premature aging disorders of Hutchinson-Gilford progeria and the Werner syndrome. Similar results were obtained using a second system of intact cells whereby histidine starvation induces quantifiable satellite spots resolved by two dimensional electrophoresis on polyacrylamide gels on the acidic side of the native actin species due to substitution of the neutral amino acid glutamine for the basic histidine. In fact, error frequencies appeared to decrease during aging in vitro, likely due to selection for clonal subpopulations with the highest fidelity of protein synthesis. The only increases were seen in the intact cell system where SV40-transformed cells showed a three-to-five fold greater error frequency compared to nontransformed fibroblasts. In total, these data fail to support the error catastrophe theory of cellular aging.

Previous studies on the synthesis and function of the protein synthetic machinery through the growth cycle of normal cultured hamster embryo fibroblasts (HA) were extended here to a series of four different clonal lines of polyoma virus-transformed HA cells. Under our culture conditions, these transformed cells could enter a stationary phase characterized by no mitotic cells, very low rates of DNA synthesis, and arrest in post-mitotic pre-DNA synthetic state. Cellular viability was initially high in stationary phase but, unlike normal cells, transformed cells slowly lost viability. The rate of protein synthesis in the stationary phase of the transformed cells fell to 25-30% of the exponential rate. Though this reduction was similar to that seen in normal cells, it was accomplished by different means. The specific reduction in the ribosome complement per cell to values below that of any cycling cell seen in normal cells, was not seen in any of the transformed lines. This observation, which implies a loss of normal control of ribosome synthesis through the growth cycle after transformation, was confirmed in normal Chinese hamster embryo fibroblasts and transformed CHO cell lines. Normal control of ribosome synthesis was restored in L-73 and LR-73, growth control revertants of one of the transformed CHO lines. The transformed lines reduced their protein synthetic rates in stationary phase either by a greater reduction in the proportion of functioning ribosomes than that seen in normal cells or by a decrease in the elongation rate of functioning ribosomes; the latter effect was not seen in the normal cells. A model for growth control of normal cells and its derangement in transformed cells is presented.