• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of pnasPNASInfo for AuthorsSubscriptionsAboutThis Article
Proc Natl Acad Sci U S A. Apr 15, 2003; 100(8): 4568–4573.
Published online Apr 3, 2003. doi:  10.1073/pnas.0830998100
PMCID: PMC153596
Cell Biology

C-terminal-binding protein corepresses epithelial and proapoptotic gene expression programs

Abstract

The genesis of carcinoma cells often involves epithelial-to-mesenchymal transitions and the acquisition of apoptosis resistance, but it is unclear whether these alterations are controlled coordinately or independently. Our previously reported effects of adenovirus E1a in human tumor cells raised the possibility that the E1a-interacting corepressor protein C-terminal-binding protein (CtBP) might selectively repress epithelial cell adhesion and proapoptotic genes. Here, we report that CtBP-knockout cells were hypersensitive to apoptosis. Correspondingly, microarray analysis of CtBP-knockout vs. CtBP-rescued mouse embryo fibroblasts revealed that many epithelial-specific and proapoptotic genes were indeed regulated by CtBP. Neither the apoptosis nor the repression activities of CtBP required histidine-315, suggesting that the proposed dehydrogenase activity is not essential for CtBP function. The results presented herein establish two functional roles of CtBP: to corepress epithelial genes, thus permitting epithelial-to-mesenchymal transitions, and to modulate the cellular threshold for apoptotic responses.

Carcinomas, representing the majority of human cancers, develop from epithelial cells. This origin has motivated intensive investigation of the unique phenotypes of epithelial cells and how these are altered during carcinogenesis. In particular, epithelial cells possess prominent cell–cell and cell–matrix adhesions that partly regulate such critical functions as selectively permeable barrier assembly, cell polarity, developmental pattern formation, and epithelial gene expression. In addition, high-turnover epithelial cells of, for example, the gastrointestinal tract are programmed to undergo apoptosis on their release from the extracellular matrix (known as “anoikis;” ref. 1), preventing the colonization of mislocalized cells.

As they progress to malignancy, carcinoma cells lose the expression of certain epithelial-specific genes, a process called epithelial-to-mesenchymal transition (EMT), accompanied by a loss of apoptosis sensitivity (2, 3). The mechanistic basis for the down-regulation of epithelial genes during EMT is poorly understood, with the exception of the E-cadherin gene, where three types of repressor (Snail/Slug, ZEB-1/ZEB-2, and E2A) have been implicated in repressing the gene in mesenchymal and certain carcinoma cells (2, 4). However, studies of epithelial-specific gene promoters, in contrast to muscle-specific promoters, for example, have not revealed unifying factors for cell type-specific expression (5); these genes are therefore assumed to be regulated by diverse factors. Similarly, proapoptotic gene promoters are thought to be regulated by diverse factors. Thus, the possibility that a single gatekeeper transcription factor is required for the coordinate conversion of epithelial cells to both mesenchymal and apoptosis-resistant transcriptional programs by diverse oncogenes has not been widely considered.

The existence of such a factor has recently been inferred from the effects of adenovirus E1a protein in human tumor cells (reviewed in ref. 6). The 243-aa form of E1a binds to and affects a discrete set of cellular transcription factors. Through these factors, E1a up-regulates epithelial-specific gene expression in poorly differentiated carcinoma cells as well as in tumor cells of nonepithelial origin (7). Interestingly, E1a also sensitizes cells to anoikis (1) as well as other apoptotic scenarios (8, 9). These effects suggested that a single E1a-interacting protein might contribute to gene expression patterns underlying both EMT and cell-survival programs.

The E1a-interacting protein C-terminal-binding protein (CtBP; ref. 10) was an attractive candidate for several reasons. First, CtBP is a corepressor, implying that its inactivation by E1a would derepress (i.e., activate) target genes. Second, two of the five repressors proposed to repress the E-cadherin promoter in mesenchymal and some carcinoma cells (ZEB-1 and -2; refs. 1 and 21) use CtBP as their corepressor (11), implicating CtBP in EMT. Third, we reported recently that the knockout of CtBP genes 1 and 2 was embryonic lethal at day 8 of development (12). Embryos resulting from this compound homozygous knockout showed very limited cell differentiation, possibly due to a defect in EMT. Finally, the anoikis-sensitization effect of E1a was partially abrogated by mutations that prevented E1a–CtBP interaction (4).

In this paper, we report that CtBP-knockout cells were hypersensitive to apoptosis. Correspondingly, many epithelial-specific and proapoptotic genes were up-regulated in the knockout cells and were repressed by reexpression of CtBP. These data suggest that CtBP coordinately corepresses epithelial and proapoptotic gene expression programs, potentially contributing to EMT and tumor malignancy.

Materials and Methods

Cell Lines.

CtBP knockout mice and MEFs were described previously (12). For CtBP-rescued cell lines, the human CtBP1 or mouse CtBP2 gene in the retroviral vector MSCV-ires-zeo or MSCV-ires-puro, respectively, was packaged by transfection in [var phi]NX cells, and infected knockout cells were selected for zeocin or puromycin resistance. Expression of the CtBP genes in >90% of the drug-resistant cells was verified by immunofluorescence by using an anti-CtBP (human) or anti-CtBP2 (mouse) antibody. Ras-transformed cell lines were generated by transfection with HA-HrasV12 in the vector pcDNA3.1hyg, followed by Western blotting of hygromycin-resistant clones for HA-H-rasV12.

Antibodies and Western Blotting.

Western blotting was performed as described previously (7), by using horseradish peroxidase-labeled secondary antibodies from The Jackson Laboratory. Primary antibodies were from the following sources: Desmoglein-2, BioDesign (Kennebunk, ME) DG3, 10 mAb; E-cadherin, DECMA-1, Sigma; Occludin, Research Diagnostics (Flanders, NJ) mAb; Plakoglobin, Zymed PG-11E4 mAb; β-actin, ICN C4 mAb; keratin-8, R. Oshima, The Burnham Institute, TROMA-1 mAb; Bax, pAb, J. Reed, The Burnham Institute; Gravin, pAb, I. Gellman, Albert Einstein School of Medicine; P53, PharMingen pAb240; PERP and CtBP, coding sequences for mouse PERP and human CtBP were subcloned into pGEX4T, and recombinant proteins were purified from bacterial extracts on glutathione-Sepharose. Rabbit polyclonal antibodies were prepared by the Robert Sargeant Antibody Laboratory (Ramona, CA), and characterized for monospecificity on Western blots.

RT-PCR Analysis.

RNAs were purified by using the Qiagen (Chatsworth, CA) RNeasy kit, and 6 μg of RNA was reverse-transcribed by using SuperScript RT (Invitrogen). PCR primers, usually corresponding to 3′ untranslated regions, were chosen by using the program PRIMER 3 (www-genome.wi.mit.edu/cgi-bin/primer/primer3_www.cgi). Pilot PCR reactions were first carried out to ascertain cDNA amounts that generated linear dose response (i.e., product vs. cDNA input) at 25 cycles of amplification. Reactions were carried out by using at least two different RNA preparations per cell line.

Microarray Analysis.

Total cellular RNA was prepared from cultured cells by using the RNeasy Kit (Qiagen) and hybridized to oligonucleotide microarrays (U74v2 GeneChip; Affymetrix, Santa Clara, CA) as described (13). Duplicates were hybridized for all indicated samples. Scanned image files were quantitated with GENECHIP 3.2 (Affymetrix). Resulting Affymetrix text files were compared with genespring software by using the following criteria unless otherwise indicated: maximum hybridization intensity >200 in at least one of the comparison samples, a fold change between normalized data from compared samples of at least two, and statistical significance among replicates.

Apoptosis Assays.

Cells were assayed for apoptosis-related caspase activation as described (4). Unless otherwise specified, drug treatments were for 4 h before the assay, as was irradiation with 40 J/m2 UV light by using a Stratalinker (Strategene). FASL (Alexis, San Diego) was used at 100 ng/ml together with 1 μg/ml enhancer antibody on cells that were placed in suspension during the treatment.

Plasmid Constructs.

The PERP promoter (14) and pG13/pG15 constructs were kindly provided by T. Jacks (Massachusetts Institute of Technology) and B. Vogelstein (Johns Hopkins University), respectively. The MSCV-ires-zeo retroviral vector (15) was kindly provided by W. Sha (University of California, Berkeley, CA); MSCV-ires-puro was constructed by substituting the IRES-puro element from pIRESpuro2 (CLONTECH) into the EcoRI-SalI site of MSCV-ires-zeo. The H315Q and G183A mutants of CtBP were generated by using the Stratagene Chameleon system followed by sequence verification. CtBP sequences were subcloned into the BglII-NotI site of these vectors. The SV40 promoter-gal4-binding site-luciferase reporter was constructed by inserting 4× gal4-binding sites into the BglII site of pGL2promoter (Promega). The gal4-CtBP expression constructs were generated by ligating a HindIII–BglII fragment containing amino acids 1–147 of the gal4 protein with a BglII–SalI fragment containing the CtBP coding region into the HindIII-XhoI site of pcDNA3.1.

Transient Transfections.

Transient transfections in the CtBP-knockout cell line CtBP90 were performed by using Fugene-6 at a 3:1 (microliters per microgram) ratio with DNA. Luciferase activity was assayed 24–36 h after transfection.

In Vitro Protein Interaction.

GST and GST-E1a C terminus were expressed in bacteria and purified by glutathione-Sepharose affinity (Sigma). His-tagged CtBP and H315Q mutant were expressed in BL21 (DE3) and purified by Ni-NTA affinity (Qiagen). GST and GST-E1a were coupled to glutathione-Sepharose beads (Pharmacia) and blocked with BSA. Recombinant CtBP or H315Q was added to the binding buffer HEG300 (10 mM Hepes, pH 7.4/10% glycerol/300 mM NaC1/0.5 mM EDTA/0.1% Nonidet P-40/1 mM DTT/10 mM NaF/10 mM Na3VO4) plus protease inhibitors (Complete, Boehringer–Mannheim) for 1 h at 4°C. The beads were washed three times with HEG300, boiled in sample buffer, and electrophoresed on an SDS-polyacrylamide gel. After transfer to a poly(vinylidene difluoride) membrane, the bound fraction was detected by Western blotting by using an anti-His-tag antibody (Qiagen).

Results

CtBP Is Critical for the Repression of Epithelial Phenotype-Related Genes.

Embryos resulting from the homozygous compound knockout of CtBP genes 1 and 2 failed to survive beyond day 8 of development and showed little evidence of cell differentiation (12). This phenotype was consistent with a defect in EMT due to lack of gene repression normally carried out by CtBP.

To address this possibility, we first compared CtBP1,2−/− mouse embryo fibroblasts (MEFs) against either CtBP1,2−/+ control cells (derived from littermates) or mixed populations of cells in which CtBP1 was reexpressed by using a retroviral vector.

Western blotting showed that several selected epithelial-specific proteins, including E-cadherin, plakoglobin, occludin, and keratin-8, were overexpressed in the knockout cells relative to the control heterozygous cells, and that these genes could be repressed by rescuing CtBP1 expression in the knockout cells (Fig. (Fig.1).1). To extend this analysis, three microarray analyses were performed, comparing CtBP1,2−/− MEFs to heterozygous controls, CtBP1-rescued controls, and CtBP1,2-rescued cells. Inspection of the complete microarray results (which are published as supporting information on the PNAS web site, www.pnas.org) revealed a programmatic increase in the expression of epithelial genes and certain proapoptotic genes (discussed below) in cells lacking CtBP. Genes whose expression was consistently lower both in the CtBP-heterozygous (control) cells and in the CtBP1-rescued cells relative to the knockout cells are listed in Fig. Fig.1.1. These included cytokeratins, tight junction components, and laminins. (E-cadherin and the tight junction component JCAM-1 scored only in the knockout-vs.-heterozygous control comparison, presumably due to technical issues, and plakoglobin was absent from the microarray chip; nevertheless, their regulation was verified by using specific probes.) Although the microarray data contained sporadic genes of nonepithelial specificity, these did not cluster sufficiently to predict any cell type other than epithelial. These data suggested that CtBP corepresses epithelial genes by direct or indirect mechanisms.

Figure 1
Epithelial phenotype-related proteins are regulated by CtBP. Equal amounts of total cellular protein from CtBP1,2−/− cells (−/−), cells rescued with vector alone (vec−/−), cells rescued with ...

CtBP Confers Apoptosis Resistance and Represses Proapoptotic Gene Expression.

The acquisition of apoptosis resistance frequently accompanies EMT during tumor progression. We considered the possibility that these two phenotypic alterations were programmatically linked by the corepressor CtBP. To establish this connection, we asked whether CtBP regulates cell survival.

First, we assayed CtBP1,2−/−-, CtBP 1,2−/+-, or CtBP1-rescued MEFs for caspase activation and/or frequency of apoptotic nuclei in response to the stimuli FASL, etoposide, UV irradiation, staurosporine, or cell detachment from matrix (anoikis). In all cases, the CtBP-knockout cells were hypersensitive to apoptosis compared with either CtBP-rescued or control heterozygous cells (Fig. (Fig.22 a–e). The apoptosis hypersensitivity of the knockout cells was predicted to decrease their tumorigenic potential without affecting growth under normal culture conditions, which was confirmed in Fig. Fig.22f.

Figure 2
CtBP-knockout cells are hypersensitive to apoptosis and deficient in tumorigenicity after transformation by H-rasV12. (a) Anoikis. Cell lines described in Fig. Fig.11 were assayed for anoikis-related caspase activation (Upper) or for the percentage ...

To identify proapoptotic genes regulated by CtBP, we used a combination of microarray analysis, RT-PCR, and Western blotting (Fig. (Fig.3).3). PERP (p53-effector related to pmp-22), a highly proapoptotic transcriptional target of p53 of unknown biochemical function (14), was consistently and dramatically down-regulated by CtBP. Because several other p53-target genes [p21, PTEN, and several insulin-like growth factor-binding proteins (IGF-BPs); see supporting information on the PNAS web site and Fig. Fig.3)3) were also identified in the microarray analysis, we probed for the additional proapoptotic p53-target genes p21, Bax, Noxa, PERP, and Puma by RT-PCR and/or Western blotting. With the exception of Puma (data not shown), all these genes were expressed at significantly higher levels in cells lacking CtBP, suggesting that CtBP might be an antagonist of transactivation by p53 or p53-related proteins. However, the activity of a minimal p53-responsive promoter was not affected by CtBP (Fig. (Fig.33c), even when the reporter was stably integrated into CtBP−/− or CtBP-rescued MEFs (data not shown). Also, the PERP promoter was strongly repressed by CtBP in transient transfection assays, even when the three known p53-binding sites in the promoter (14) were mutated (data not shown), indicating that antagonism of p53 function was not likely to be the mechanism of CtBP's repression effect. Nevertheless, a striking pattern of up-regulation of proapoptotic genes occurred in the homozygous knockout cells, which could be repressed by reexpressing CtBP. These included well-established p53 target genes (Bax, Noxa, PERP), genes whose related family members are targets of p53 (IGF-BP-1,2,6,7), and genes that are not known to be regulated by p53 (Id-1, gravin, Forkhead box G1, J1, and M1). As expected from their apoptosis-hypersensitive phenotype, a systematic increase in cell survival genes was not seen in the knockout cells; however, two exceptions were the secreted factors, the CXC cytokine family member GRO1, and the protein midkine (although the presence or absence of the corresponding receptors on the knockout cells is unknown).

Figure 3
CtBP-knockout up-regulates proapoptotic genes. (a) RT-PCR analysis of the indicated genes was performed under conditions of linear response to cDNA input amount. RNAs were derived from cells treated with 50 μM etoposide for 4 h. Similar results ...

The Dehydrogenase-Catalytic Site Is Not Required for Transcriptional Repression.

In light of a previous report proposing that CtBP has a dehydrogenase activity that is required for gene repression (16), we investigated the effect of mutating histidine-315, the critical acid/base catalyst of the active site in homologous dehydrogenase enzymes (17). Stable mixed populations of knockout cells rescued with the histidine-315 mutated to glutamine form showed a similar degree of epithelial gene down-regulation, suggesting that this enzymatic activity of CtBP was not required for gene repression (Fig. (Fig.1).1). This mutant also inhibited apoptosis as efficiently as wild-type CtBP (Fig. (Fig.22 a and c).

To explore further whether the proposed dehydrogenase activity of CtBP was relevant for gene repression, we tested the effect of the H315Q CtBP mutant on the PERP promoter in transient transfections. The PERP promoter, in contrast with a β-actin control promoter, was strongly repressed by both the wild-type CtBP and H315Q mutant (Fig. (Fig.44a). Although the H315Q mutation was proposed (16) to prevent the interaction of CtBP with the E1a C terminus and presumably other repressor proteins containing CtBP-recruitment motifs, we found that this mutation did not prevent the interaction of GST-E1a with CtBP in vitro (Fig. (Fig.44b). Finally, we examined the ability of Gal4-CtBP(wt) and Gal4-CtBP(H315Q) to repress transcription from a reporter construct in which an SV40 early promoter is adjacent to gal4-binding sites, revealing the repressive potential of gal4 fusion proteins brought into proximity with this promoter. This experiment was performed in CtBP-knockout cells, ruling out the possibility of dimerization with the endogenous protein. As shown in Fig. Fig.44c, repression was equivalent for wild-type CtBP, the H315Q mutant of CtBP, or a mutant of CtBP incapable of binding NAD(H), G183A (18).

Figure 4
Dehydrogenase activity of CtBP is not required for transcriptional repression. (a) The PERP or β-actin control promoters linked to luciferase (1.5 μg) were cotransfected with the indicated CtBP mutants, and activity of the luciferase reporters ...

Discussion

The results presented herein establish two functional roles of CtBP: to corepress epithelial genes, thus permitting EMT, and to modulate the cellular threshold for apoptotic responses.

Embryos resulting from the compound knockout of CtBP1 and CtBP2 had a chaotic architecture and terminated at day 8 of development, demonstrating that these proteins are developmentally critical corepressors (12). Considering the complexity of the developmental effects, we chose to identify gene expression/apoptosis changes occurring on reexpression of CtBP. That CtBP reexpression by itself repressed many of the genes that were abnormally overexpressed in the knockout cells implies that the CtBP-knockout cells did not irreversibly lose some of the relevant CtBP-interactive repressor proteins (e.g., ZEB, Snail, Slug). Indeed, we found that the levels of ZEB-1, Snail, and Slug were not affected by CtBP (data not shown), consistent with the ability of restored CtBP to repress E-cadherin in stably or transiently transfected CtBP-knockout cells.

Although the microarray results revealed sporadic increases in the expression of certain nonepithelial genes, only epithelial genes (and not genes corresponding to other cell types) were systematically up-regulated. Yet, a diversity of repressor proteins use CtBP as a corepressor (10), some of which target nonepithelial genes. For example, ZEB-1 is not only a repressor for E-cadherin but also for certain muscle-specific and lymphoid-specific enhancers (11). This raises the question of why the genetic knockout of CtBP, or the functional knockout of CtBP with E1a, preferentially up-regulates certain epithelial-specific and proapoptotic genes. We propose that the selectivity arises because tissue-specific nonepithelial genes require cell-type-specific transactivators, but epithelial genes generally do not (5). For example, muscle genes that are repressed by ZEB-CtBP complexes require muscle-specific factors (e.g., myoD, mef2a, and myogenin) for expression and are therefore refractory to being activated by removal of CtBP alone. However, epithelial-specific genes such as E-cadherin (19) or keratin-18 (20) do not appear to require such factors; in fact, their promoters appear to contain only the binding sites for ubiquitous transactivators such as AP-1, AP-2, NF1/CTF, and Sp1. Accordingly, expression of these genes requires only derepression, which is afforded by removing CtBP. In the case of the E-cadherin promoter, this relieves repression by ZEB-1 (4) and ZEB-2 (21); the corresponding keratin-18 repressors remain to be identified. Combining these concepts, we propose that CtBP is required for a wide variety of cell differentiation events, consistent with the apparently ubiquitous expression of the protein that is nevertheless not required for cell proliferation. In mouse embryos lacking CtBP, the initial differentiation event, EMT, is blocked, resulting in the up-regulation of epithelial-specific genes. This is consistent with the idea of the epithelial phenotype as a “default phenotype” that occurs in the absence of EMT-promoting factors, as we hypothesized previously (5).

Early embryonic or normal epithelial cells are more sensitive to apoptosis than corresponding cells that have undergone EMT, such as late embryonic mesenchymal cells or poorly differentiated carcinoma cells, respectively (3, 22). That CtBP regulates both programs suggests two nonmutually exclusive hypotheses. The first is that a transcription factor or family of factors coordinates EMT with the expression of certain apoptotic-regulatory genes. The second is that epithelial cell–cell adhesions provide signals that regulate apoptosis-regulatory genes. In this case, the absence of CtBP would regulate apoptosis genes indirectly by promoting cell adhesion. Resolving these mechanisms will require more information about whether the repressor proteins that regulate apoptotic genes, such as PERP, Bax, and Noxa, directly recruit CtBP and determining whether cell–cell adhesions, acting, for example, through the E-cadherin-β-catenin-APC-axin-LEF-1 pathway, can affect the expression of proapoptotic genes.

The mechanism of repression by CtBP is somewhat controversial (10). There are varying reports on the participation of histone deacetylases in this repression process, and although an interaction between CtBP and polycomb-2 protein has been shown, its functional significance has not yet been established. Recently, the crystal structure of CtBP1 was reported and, as expected from the primary sequence, was found to be similar to that of 2-hydroxyacid dehydrogenases (16). Although it was reported that CtBP had a weak dehydrogenase activity and that mutations impairing this activity abrogated its repression effect, our assays indicate that mutants incapable of possessing dehydrogenase activity [due to catalytic or NAD(H)-binding defects] are capable of repressing transcription and/or conferring apoptosis resistance. Whether CtBP represses transcription by interaction with effector proteins or through its intrinsic enzymatic activity remains unknown.

Supplementary Material

Supporting Tables:

Acknowledgments

Caroline Ho provided technical assistance. This work was supported by a U.S. Department of Defense–Breast Cancer Research Program grant (to S.M.F.) and by National Institutes of Health Grants CA96561 and DK44239 (to R.H.G. and Q.Z.). M.G. is currently a Scientific Research Worker of the National Funds for Scientific Research (Belgium).

Abbreviations

EMT
epithelial-to-mesenchymal transition
CtBP
C-terminal-binding protein
MEF
mouse embryo fibroblast

References

1. Frisch S M, Francis H. J Cell Biol. 1994;124:619–626. [PMC free article] [PubMed]
2. Thierry J P. Nat Rev Cancer. 2002;2:442–454. [PubMed]
3. Green D R, Evan G I. Cancer Cell. 2002;1:19–30. [PubMed]
4. Grooteclaes M, Frisch S M. Oncogene. 2000;19:3823–3828. [PubMed]
5. Frisch S M. BioEssays. 1997;19:705–709. [PubMed]
6. Frisch S M, Mymryk J S. Nat Rev Mol Cell Biol. 2002;3:441–452. [PubMed]
7. Frisch S M. J Cell Biol. 1994;124:1085–1096. [PMC free article] [PubMed]
8. Frisch S M, Dolter K E. Cancer Res. 1995;55:5551–5555. [PubMed]
9. Nahle Z, Polakoff J, Davuluri R V, McCurrach M E, Jacobson M D, Narita M, Zhang M Q, Lazebnik Y, Bar-Sagi D, Lowe S W. Nat Cell Biol. 2002;4:859–864. [PubMed]
10. Chinnadurai G. Mol Cell. 2002;9:213–224. [PubMed]
11. Postigo A A, Dean D C. Proc Natl Acad Sci USA. 1999;96:6683–6688. [PMC free article] [PubMed]
12. Hildebrand J D, Soriano P. Mol Cell Biol. 2002;22:5296–5307. [PMC free article] [PubMed]
13. Su A I, Cooke M P, Ching K A, Hakak Y, Walker J R, Wiltshire T, Orth A P, Vega R G, Sapinoso L M, Moqrich A, et al. Proc Natl Acad Sci USA. 2002;99:4465–4470. [PMC free article] [PubMed]
14. Attardi L D, Reczek E E, Cosmas C, Demicco E G, McCurrach M E, Lowe S W, Jacks T. Genes Dev. 2000;14:704–718. [PMC free article] [PubMed]
15. Kuang A A, Diehl G E, Zhang J, Winoto A. J Biol Chem. 2000;275:25065–25068. [PubMed]
16. Kumar V, Carlson J E, Ohgi K A, Edwards T A, Rose D W, Escalante C R, Rosenfeld M G, Aggarwal A K. Mol Cell. 2002;10:857–869. [PubMed]
17. Adams M, Buehner M, Chandrasekhar K, Ford G, Hackert M, Liljas A, Rossmann M, Smiley I, Allison W, Everse J. Proc Natl Acad Sci USA. 1973;70:1968–1972. [PMC free article] [PubMed]
18. Zhang Q, Piston D, Goodman R. Science. 2002;295:1895–1897. [PubMed]
19. Giroldi L A, Bringuier P P, de Weijert M, Jansen C, van Bokhoven A, Schalken J A. Biochem Biophys Res Commun. 1997;241:453–458. [PubMed]
20. Pankov R, Neznanov N, Umezawa A, Oshima R G. Mol Cell Biol. 1994;14:7744–7757. [PMC free article] [PubMed]
21. Comijn J, Berx G, Vermassen P, Verschueren K, van Grunsven L, Bruyneel E, Mareel M, Huylebroeck D, van Roy F. Mol Cell. 2001;7:1267–1278. [PubMed]
22. Heyer B, MacAuley A, Behrendtsen O, Werb Z. Genes Dev. 2000;14:2072–2084. [PMC free article] [PubMed]
23. Burgering B M, Kops G J. Trends Biochem Sci. 2002;27:352–360. [PubMed]
24. Perks C M, Newcomb P V, Norman M R, Holly J M. J Mol Endocrinol. 1999;22:141–150. [PubMed]
25. Butt A J, Firth S M, Baxter R C. Immunol Cell Biol. 1999;77:256–262. [PubMed]
26. Sueoka N, Lee H Y, Wiehle S, Cristiano R J, Fang B, Ji L, Roth J A, Hong W K, Cohen P, Kurie J M. Oncogene. 2000;19:4432–4436. [PubMed]
27. Parrinello S, Lin C Q, Murata K, Itahana Y, Singh J, Krtolica A, Campisi J, Desprez P Y. J Biol Chem. 2001;276:39213–39219. [PubMed]

Articles from Proceedings of the National Academy of Sciences of the United States of America are provided here courtesy of National Academy of Sciences

Formats:

Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...

Links

  • Cited in Books
    Cited in Books
    PubMed Central articles cited in books
  • MedGen
    MedGen
    Related information in MedGen
  • PubMed
    PubMed
    PubMed citations for these articles
  • Substance
    Substance
    PubChem Substance links

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...