A dominant repression domain in Tbx3 mediates transcriptional repression and cell immortalization: relevance to mutations in Tbx3 that cause ulnar-mammary syndrome.

Mutations in Tbx3 are responsible for ulnar-mammary syndrome (UMS), an autosomal dominant disorder affecting limb, tooth, hair, apocrine gland and genital development. Tbx3 is a member of a family of transcription factors that share a highly conserved DNA-binding domain known as the T-domain. UMS-causing mutations in Tbx3 have been found at numerous sites within the TBX3 gene, with many occurring downstream from the N-terminally located T-domain. The occurrence of mutations downstream of the DNA-binding domain raises the possibility that there exist important functional domains in C-terminal portions of the Tbx3 protein that affect its behavior as a transcription factor. To determine if and how such C-terminal mutations affect transcription we have mapped regions that confer transcriptional activity and nuclear localization and characterized the DNA binding properties of Tbx3. We find that Tbx3 binds the canonical Brachyury binding site as a monomer and represses transcription. We show that a key repression domain (RD1) resides in the Tbx3 C-terminus that can function as a portable repression domain. Most UMS-associated C-terminal mutants lack the RD1 and exhibit decreased or loss of transcriptional repression activity. In addition, we identify a domain responsible for nuclear localization of Tbx3 and show that two C-terminal mutants of Tbx3 have increased rates of protein decay. Finally, we show that Tbx3 can immortalize primary embryo fibroblasts and that the RD1 repression domain is required for this activity. Our results identify critical functional domains within the Tbx3 protein and facilitate interpretation of the functional consequences of present and future UMS mutations.

The interplay between Mad and Myc in proliferation and differentiation.

Members of the Myc family of transcription factors are key regulators of cell proliferation, and excessive levels of Myc lead to tumor formation. Mad family proteins are related to Myc, but they antagonize the oncogenic activity of Myc in cell-culture assays. Here, we examine current models of Mad function and the relationship between Mad and Myc in cell proliferation, differentiation and tumorigenesis.

Mga, a dual-specificity transcription factor that interacts with Max and contains a T-domain DNA-binding motif.

The basic-helix-loop-helix-leucine zipper (bHLHZip) proteins Myc, Mad and Mnt are part of a transcription activation/repression system involved in the regulation of cell proliferation. The function of these proteins as transcription factors is mediated by heterodimerization with the small bHLHZip protein Max, which is required for their specific DNA binding to E-box sequences. We have identified a novel Max-interacting protein, Mga, which contains a Myc-like bHLHZip motif, but otherwise shows no relationship with Myc or other Max-interacting proteins. Like Myc, Mad and Mnt proteins, Mga requires heterodimerization with Max for binding to the preferred Myc-Max-binding site CACGTG. In addition to the bHLHZip domain, Mga contains a second DNA-binding domain: the T-box or T-domain. The T-domain is a highly conserved DNA-binding motif originally defined in Brachyury and characteristic of the Tbx family of transcription factors. Mga binds the preferred Brachyury-binding sequence and represses transcription of reporter genes containing promoter-proximal Brachyury-binding sites. Surprisingly, Mga is converted to a transcription activator of both Myc-Max and Brachyury site-containing reporters in a Max-dependent manner. Our results suggest that Mga functions as a dual-specificity transcription factor that regulates the expression of both Max-network and T-box family target genes.

Analysis of the Max-binding protein MNT in human medulloblastomas.

Medulloblastomas (MBs) are the most frequent malignant brain tumors in children. The molecular pathogenesis of these tumors is still poorly understood. Microsatellite and restriction-fragment-length polymorphism studies have revealed allelic loss of genetic material on the short arm of chromosome 17 in the region 17p13 in approximately 50% of MBs, suggesting the presence of a tumor-suppressor gene in this region. A candidate for this putative tumor-suppressor is the MNT gene, located at 17p13.3 and encoding a Max-interacting nuclear protein with transcriptional-repressor activity. In this study, we analyzed MNT mRNA and protein expression in 44 MB samples, including 32 primary tumors, 3 recurrent tumors and 9 MB cell lines. Allelic loss at 17p13.3 was found in 49% of informative cases. RT-PCR showed MNT mRNA expression in all cases analyzed. Endogenous Mnt protein with an apparent molecular weight of 72 to 74 kDa was detected in lysates from MB cell lines. The presence and functional integrity of Mnt in MBs were tested in electrophoretic mobility-shift assays. These experiments demonstrated that Mnt interacts with Max, and that this heterodimer binds DNA specifically, suggesting a functional bHLHZip domain of MB-derived Mnt. In support, single-strand conformation-polymorphism (SSCP) analyses revealed no mutation in the bHLHZip region. Deletion of the Mnt Sin3 interaction domain was shown to convert Mnt from an inhibitor of myc/ras-co-transformation into a molecule capable of cooperating with Ras in transformation. This region therefore was screened for mutation by SSCP: again, no alterations were found. These findings indicate that the MNT gene located at 17p13.3 is not likely to be involved in the molecular pathogenesis of MBs.

Targeted disruption of the MYC antagonist MAD1 inhibits cell cycle exit during granulocyte differentiation.

The switch from transcriptionally activating MYC-MAX to transcriptionally repressing MAD1-MAX protein heterodimers has been correlated with the initiation of terminal differentiation in many cell types. To investigate the function of MAD1-MAX dimers during differentiation, we disrupted the Mad1 gene by homologous recombination in mice. Analysis of hematopoietic differentiation in homozygous mutant animals revealed that cell cycle exit of granulocytic precursors was inhibited following the colony-forming cell stage, resulting in increased proliferation and delayed terminal differentiation of low proliferative potential cluster-forming cells. Surprisingly, the numbers of terminally differentiated bone marrow and peripheral blood granulocytes were essentially unchanged in Mad1 null mice. This imbalance between the frequencies of precursor and mature granulocytes was correlated with a compensatory decrease in granulocytic cluster-forming cell survival under apoptosis-inducing conditions. In addition, recovery of the peripheral granulocyte compartment following bone marrow ablation was significantly enhanced in Mad1 knockout mice. Two Mad1-related genes, Mxi1 and Mad3, were found to be expressed ectopically in adult spleen, indicating that functional redundancy and cross-regulation between MAD family members may allow for apparently normal differentiation in the absence of MAD1. These findings demonstrate that MAD1 regulates cell cycle withdrawal during a late stage of granulocyte differentiation, and suggest that the relative levels of MYC versus MAD1 mediate a balance between cell proliferation and terminal differentiation.

Sequential expression of the MAD family of transcriptional repressors during differentiation and development.

Members of the Myc proto-oncogene family encode transcription factors that function in multiple aspects of cell behavior, including proliferation, differentiation, transformation and apoptosis. Recent studies have shown that MYC activities are modulated by a network of nuclear bHLH-Zip proteins. The MAX protein is at the center of this network in that it associates with MYC as well as with the family of MAD proteins: MAD1, MXI1, MAD3 and MAD4. Whereas MYC-MAX complexes activate transcription, MAD-MAX complexes repress transcription through identical E-box binding sites. MAD proteins therefore act as antagonists of MYC. Here we report the expression patterns of the Mad gene family in the adult and developing mouse. High level of Mad gene expression in the adult is limited to tissues that display constant renewal of differentiated cell populations. In embryos, Mad transcripts are widely distributed with expression peaking during organogenesis at the onset of differentiation. A detailed analysis of their pattern of expression during chrondrocyte and neuronal differentiation in vivo, and during neuronal differentiation of P19 cells in vitro, shows that Mad family genes are sequentially induced. Mad3 transcripts and proteins are detected in proliferating cells prior to differentiation. Mxi1 and Mad4 transcripts are most abundant in cells that have further advanced along the differentiation pathway, whereas Mad1 is primarily expressed late in differentiation. Taken together, our data suggest that the different members of the MAD protein family exert their functions at distinct steps during the transition between proliferation and differentiation.

Mnt: a novel Max-interacting protein and Myc antagonist.

We have identified a novel Max-binding protein, Mnt, which belongs to neither the Myc nor the Mad families (Hurlin et al. 1997). Mnt interacts with Max in vivo and functions as a transcriptional repressor of reporter genes containing promoter-proximal CACGTG sites. Mnt:Max complexes also efficiently suppress Myc-dependent activation from the same promoter. Transcription repression by Mnt maps to a 13 amino acid N-terminal region related to the Sin3 interaction domain (SID) of Mad proteins. This region of Mnt mediates interaction with mSin3 corepressor proteins and its deletion converts Mnt from a repressor to an activator and from a suppressor of Myc-dependent transformation to a cooperating oncogene. This latter result suggests that Mnt and Myc regulate an overlapping set of target genes in vivo. Expression of mnt RNA is observed in many tissues and in both proliferating and differentiating cells. Likewise, Mnt protein is expressed in many proliferating cell types in culture where both Myc:Max and Mnt:Max complexes are detected. An exception is P19 embryonal carcinoma cells, where Mnt is expressed and in a complex with Max, but Myc proteins are not detected. Mnt is likely to be a key regulator of Myc activities in vivo and, in addition, may possess Myc-independent functions.

Mnt, a novel Max-interacting protein is coexpressed with Myc in proliferating cells and mediates repression at Myc binding sites.

The small constitutively expressed bHLHZip protein Max is known to form sequence-specific DNA binding heterodimers with members of both the Myc and Mad families of bHLHZip proteins. Myc:Max complexes activate transcription, promote proliferation, and block terminal differentiation. In contrast, Mad:Max heterodimers act as transcriptional repressors, have an antiproliferative effect, and are induced upon differentiation in a wide variety of cell types. We have identified a novel bHLHZip Max-binding protein, Mnt, which belongs to neither the Myc nor the Mad families and which is coexpressed with Myc in a number of proliferating cell types. Mnt:Max heterodimers act as transcriptional repressors and efficiently suppress Myc-dependent activation from a promoter containing proximal CACGTG sites. Transcription repression by Mnt maps to a 13-amino-acid amino-terminal region related to the Sin3 interaction domain (SID) of Mad proteins. We show that this region of Mnt mediates interaction with mSin3 corepressor proteins and that its deletion converts Mnt from a repressor to an activator. Furthermore, wild-type Mnt suppresses Myc+Ras cotransformation of primary cells, whereas Mnt containing a SID deletion cooperates with Ras in the absence of Myc to transform cells. This suggests that Mnt and Myc regulate an overlapping set of target genes in vivo. When mnt is expressed as a transgene under control of the beta-actin promoter in mice the transgenic embryos exhibit a delay in development and die during mid-gestation, when c- and N-Myc functions are critical. We propose that Mnt:Max:Sin3 complexes normally function to restrict Myc:Max activities associated with cell proliferation.

Regulation of Myc and Mad during epidermal differentiation and HPV-associated tumorigenesis.

c-Myc and Mad each form heterodimers with Max that bind the same E-box related DNA sequences. Whereas Myc:Max complexes activate transcription and promote cell proliferation and transformation, Mad:Max complexes repress transcription and block c-Myc-mediated cell transformation. Here we examine these antagonistic transcription factors during epithelial differentiation and neoplastic progression. During differentiation of primary human keratinocytes, Mad is rapidly induced and c-Myc is downregulated, resulting in a switch from c-Myc:Max to Mad:Max heterodimers. In normal epidermis and colonic mucosa c-myc expression is restricted to proliferating cell layers, while mad expression is restricted to differentiating cell layers. Using HPV18 transformed keratinocytes that vary in their ability to differentiate in organotypic cultures, we find that Mad induction occurs only in those cells that retain a differentiation response. In the epidermis of transgenic mice in which expression of the HPV16 E6 and E7 oncogenes are targeted to basal keratinocytes, neoplastic progression occurs and is marked by an expansion of c-myc expressing basal-like cells. Expression of mad is found only in growth-arrested differentiating cells on the outer edges of preneoplastic lesions. The squamous cell carcinomas that arise evidence a variable number of sites within the tumor masses where mad expression and morphological differentiation coincide; increasing malignancy correlates with loss of both mad and capability to differentiate. These results indicate that c-Myc and Mad expression are tightly coupled to the transition from proliferation to differentiation of epithelial cells and that restriction of Mad expression may be associated with loss of normal differentiation capability and with tumorigenesis.

Mad3 and Mad4: novel Max-interacting transcriptional repressors that suppress c-myc dependent transformation and are expressed during neural and epidermal differentiation.

The basic helix-loop-helix-leucine zipper (bHLHZip) protein Max associates with members of the Myc family, as well as with the related proteins Mad (Mad1) and Mxi1. Whereas both Myc:Max and Mad:Max heterodimers bind related E-box sequences, Myc:Max activates transcription and promotes proliferation while Mad:Max represses transcription and suppresses Myc dependent transformation. Here we report the identification and characterization of two novel Mad1- and Mxi1-related proteins, Mad3 and Mad4. Mad3 and Mad4 interact with both Max and mSin3 and repress transcription from a promoter containing CACGTG binding sites. Using a rat embryo fibroblast transformation assay, we show that both Mad3 and Mad4 inhibit c-Myc dependent cell transformation. An examination of the expression patterns of all mad genes during murine embryogenesis reveals that mad1, mad3 and mad4 are expressed primarily in growth-arrested differentiating cells. mxi1 is also expressed in differentiating cells, but is co-expressed with either c-myc, N-myc, or both in proliferating cells of the developing central nervous system and the epidermis. In the developing central nervous system and epidermis, downregulation of myc genes occurs concomitant with upregulation of mad family genes. These expression patterns, together with the demonstrated ability of Mad family proteins to interfere with the proliferation promoting activities of Myc, suggest that the regulated expression of Myc and Mad family proteins function in a concerted fashion to regulate cell growth in differentiating tissues.

The Max transcription factor network: involvement of Mad in differentiation and an approach to identification of target genes.

The small bHLHZip protein, Max, was originally identified through its interaction with Myc family proteins and appears to be an obligate partner for Myc function. Max has now been found to interact with at least two other proteins, Mad and Mxi1. These also belong to the bHLHZip class but are otherwise unrelated to Myc. Mad has been shown to abrogate the positive transcriptional activity of Myc and to inhibit Myc in co-transformation assays. This suggests that Mad may antagonize Myc function. Mad is rapidly induced upon differentiation, a time when Myc is frequently down-regulated. We show here evidence for Mad expression upon differentiation of myeloblasts, monoblasts, and keratinocytes. Mad:Max complexes are detected during differentiation and appear to replace the Myc:Max complexes present in proliferating cell populations. Since these complexes appear to form even in the presence of Myc, there may exist mechanisms that act to inhibit Myc:Max, or to promote Mad:Max, complex formation. We speculate that Max complex switching causes a change in the transcriptional activity of groups of target genes. Mad is not induced in all differentiating cell types, suggesting that other, possibly tissue-restricted, proteins might act in similar switch mechanisms to effect changes in transcriptional programs. We have also developed an approach to identification of the gene targets for Myc:Max complexes. By employing an immunoisolation procedure, we have begun characterization of several clones whose expression levels correlate with those of c-myc. Further identification of Myc-regulated genes may allow us to determine the molecular mechanism by which Myc governs cell proliferation and differentiation.

Characteristics of an infinite life span diploid human fibroblast cell strain and a near-diploid strain arising from a clone of cells expressing a transfected v-myc oncogene.

Diploid human fibroblasts were transfected with a plasmid carrying a v-myc oncogene linked to the neo gene or with a vector control carrying a neo gene. Drug-resistant clones were isolated and subcultured as needed. All populations went into crisis and eventually senesced. But while they were senescing, viable-appearing clones were noted among the progeny of a transfected population that expressed the v-myc oncogene. After several months, these cells began replicating more rapidly. Karyotype analysis indicated that they were clonally derived since all of them had 45 chromosomes, including 2 marker chromosomes. This cell strain was designated MSU-1.1. Similar analysis showed that cells from an earlier passage were diploid. These cells were designated MSU-1.0. Both strains have undergone more than 200 population doublings since their siblings senesced, without any change in chromosome complement. Both strains express the v-myc protein and have the same integration site for the transfected v-myc and neo genes. The MSU-1.0 cells cannot grow without exogenously added growth factors. The MSU-1.1 cells grow moderately well under the same conditions and grow to a higher saturation density than MSU-1.0 cells. Since the chance of human cells acquiring an infinite life span in culture is very rare, the data suggest that MSU-1.1 cells are derived from MSU-1.0 cells. The expression of v-myc is probably required for acquisition of an infinite life span, since this phenotype did not develop in populations not expressing this oncogene. However, expression of v-myc is clearly not sufficient, since all of the progeny of the clone that gave rise to the MSU-1.0 cells expressed this oncogene, but the vast majority of them senesced.

Progression of human papillomavirus type 18-immortalized human keratinocytes to a malignant phenotype.

We have developed a model system for progression of human epithelial cells to malignancy, using a human papillomavirus type 18 (HPV-18)-immortalized human keratinocyte cell line. Cells of cell line FEP-1811 were nontumorigenic in athymic mice through at least 12 passages in culture, but after 32 passages were weakly tumorigenic, producing tumors that regressed. After 62 passages they produced invasive squamous cell carcinomas that grew progressively. The progression to malignancy was associated with an increase in the efficiency of forming colonies in soft agar and with altered differentiation properties. In an organotypic culture system, FEP-1811 cells at passages 12 and 32 exhibited features typical of premalignant intraepithelial neoplasia in vivo, and cells at passage 68 exhibited features consistent with squamous cell carcinomas. No change in copy number of the transfected HPV-18 genome or in the level of expression of the viral transforming gene products E6 and E7 was detected between tumorigenic and nontumorigenic cells. Cytogenetic analysis of cells at early, middle, and late passage levels and cells cultured from tumors revealed that several chromosomal abnormalities segregated with the tumorigenic cell populations.

Malignant transformation of human fibroblasts by oncogene transfection or carcinogen treatment.

Exposure to chemical carcinogens or radiation is considered to cause most human cancer, but human cells in culture have not been successfully transformed to malignancy by such agents. Malignant transformation is a multi-step process and one explanation for the failure to induce such transformation of human cells in culture could be inability to recognize the phenotypes of carcinogen-treated cells that have undergone intermediate changes, so that these cells can be isolated and exposed a second time to cause further changes. To identify possible intermediates, we transfected diploid human fibroblasts with oncogenes known to be active in cells derived from fibrosarcomas and determined the phenotypes produced. H- or N-ras oncogenes flanked by suitable enhancer and promoter sequences caused the cells to exhibit several characteristics of malignant cells, but not to acquire an infinite life span or form tumors. Transfection of these oncogenes in the same constructions, or a viral K-ras oncogene, into an infinite life span, near-diploid, non-tumorigenic cell strain developed in this laboratory (MSU-1.1 cells) resulted in distinct foci of morphologically-altered, anchorage independent, and growth factor independent cells that formed progressively-growing, invasive malignant sarcomas in athymic mice and expressed the p21s of the transfected ras genes. Transfection of two other infinite life span human cell lines with the H-ras oncogene in the same construction also yielded malignant cells. Recently, we succeeded in inducing the malignant state in MSU-1.1 cells using carcinogen identified that are one step removed from malignant transformation, others that are two-steps removed, etc. Furthermore, we know what new phenotypes these cells need to express to be malignantly transformed and which oncogenes can make such a change. If, as suggested above, proto-oncogenes are the cellular targets for carcinogen attack, it should be possible, by carcinogen treatment to bring about the malignant state. We have recently succeeded in achieving just such transformation by exposing MSU-1.1 cells to chemical carcinogens.

Malignant transformation of human fibroblasts caused by expression of a transfected T24 HRAS oncogene.

We showed previously that diploid human fibroblasts that express a transfected HRAS oncogene from the human bladder carcinoma cell line T24 exhibit several characteristics of transformed cells but do not acquire an infinite life-span and are not tumorigenic. To extend these studies of the T24 HRAS in human cells, we have utilized an infinite life-span, but otherwise phenotypically normal, human fibroblast cell strain, MSU-1.1, developed in this laboratory after transfection of diploid fibroblasts with a viral v-myc oncogene. Transfection of MSU-1.1 cells with the T24 HRAS flanked by two transcriptional enhancer elements (pHO6T1) yielded foci of morphologically transformed cells. No such transformation occurred if the plasmid containing T24 HRAS had only one enhancer or none at all or if the normal human HRAS gene was transfected in the pHO6 vector (pHO6N1). Cell strains derived from such foci expressed high levels of T24 HRAS product p21, formed colonies in soft agar at high frequency, proliferated rapidly in serum-free medium that does not support growth of the parental cell line, and formed progressively growing, invasive fibrosarcomas. These foci-derived T24 HRAS-transformed cell strains, as well as cells from the tumors derived from them, had the same near-diploid karyotype as that of the parental MSU-1.1 cells. Transfection of pHO6T1 into two other infinite life-span human fibroblast cell lines, cells that had not been transfected with v-myc, also resulted in malignant transformation, suggesting that the infinite life-span phenotype of MSU-1.1 cells, and not necessarily expression of the v-myc oncogene, was the factor that complemented T24 HRAS expression to cause malignant transformation.

Transformation of diploid human fibroblasts by transfection with the v-sis, PDGF2/c-sis, or T24 H-ras genes.

Gene transfection techniques have provided powerful methods to examine the roles of cellular and retroviral oncogenes in the transformation process in rodent fibroblasts. However, the use of such techniques with diploid human fibroblasts has been limited. We have developed transfection procedures to reproducibly transfect such cells with oncogenes, and methods for the biological characterization of the transformants. We have shown that the v-sis and T24 H-ras oncogenes, as well as the platelet-derived growth factor gene (PDGF2/c-sis), are capable of inducing a transformed phenotype in normal diploid human fibroblasts, but are not capable of conferring infinite lifespan or making such cells tumorigenic.

Morphological transformation, focus formation, and anchorage independence induced in diploid human fibroblasts by expression of a transfected H-ras oncogene.

In an attempt to determine how normal human fibroblasts respond to high expression of the T24 H-ras oncogene, we tranfected such cells with the plasmid vector pHO6T1 (D. A. Spandidos and N. M. Wilkie, Nature (Lond.), 310:469-475, 1984), containing the T24 H-ras oncogene with 5' and 3' enhancer sequences, and the aminoglycoside phosphotransferase gene which confers resistance to the drug, G418. Approximately 1.5% of the G418-resistant colonies obtained after transfection and selection consisted of cells exhibiting obvious morphological transformation; i.e., they were highly refractile and more rounded than normal fibroblasts. DNA hybridization analysis showed that the morphologically transformed cells contained the transfected T24 H-ras oncogene, and radioimmunoprecipitation analysis showed that they were expressing the T24 H-ras protein product, M, 21,000 protein. Morphologically transformed cells formed colonies in soft agar at a frequency at least 60 times higher than that of cells that had been transfected with the control plasmid containing the normal cellular H-ras gene. Cells transfected with plasmid pHO6T1 could also be identified by their ability to form distinct foci when grown to confluence in nonselective medium following transfection. This study demonstrates that normal diploid human fibroblasts in culture can be transformed by transfection with a H-ras oncogene, and that such transformation correlates with expression of the mutant Mr 21,000 protein.