Tumor suppressor genes in the 9p21 gene cluster are selective targets of inactivation in neuroendocrine gastroenteropancreatic tumors.

Functional inactivation of the Rb and p53 pathways appears to be a rite of passage for all cancerous cells. However, p53 and Rb alterations are rare events in neuroendocrine gastroenteropancreatic (GEP) tumors. The CDKN2 locus on chromosome 9p21 sits at the nexus of both pathways harboring tumor suppressor genes, which restrain cell growth by affecting the function of pRb and p53. Therefore, we analyzed the implication of their inactivation in 37 primary neuroendocrine GEP tumors and two cell culture models. RT-PCR analysis revealed loss of expression of at least one of the tumor suppressor genes CDKN2A/p16, CDKN2B/p15, and CDKN2D/p14 with distinct genetic profiles, most frequently in nonfunctional pancreatic tumors (57%) and small intestinal carcinoids (44%), and less commonly in insulinomas (30%) and gastrinomas (22%). DNA analysis and methylation-specific PCR attributed loss of expression to either homozygous deletion or 5'CpG island hypermethylation. 5-Aza-2-deoxycytidine treatment reversed CDKN2A/p16 and CDKN2B/p15 silencing with concurrent growth restraint. Thus, tumor suppressor genes localized in the 9p21 gene cluster are specific targets of inactivation in neuroendocrine GEP tumors, and demethylating agents might hold promise for selective therapy.

Mutagenesis of the DNA contact site in Fos protein: compatibility with the scissors grip model and requirement for transformation.

To elucidate the mechanisms involved in the transformation by fos we have initiated a study pertaining to the identification of molecular functions of Fos protein that are crucial for transformation. We have previously reported that the presence of an intact leucine zipper in Fos is an absolute requirement for the induction of transformation, but that the autorepression function of Fos is dispensable. We now show that Fos protein also needs an intact DNA (TRE)-binding site to be able to transform. Amino acid substitutions in this domain of Fos which impair DNA binding also destroy the transforming potential of Fos, suggesting that the interaction of Fos-Jun complexes with TREs may be a crucial part of Fos-induced transformation. This hypothesis is further strengthened by our observation that Fos and Jun can cooperate in the induction of transformation. We show that a Fos protein which contains a Jun leucine zipper and is thus capable of dimerization is still dependent on the presence of exogenous Jun to induce transformation. The critical positions in the Fos DNA-binding site include those which the 'scissors grip' model predicts to be crucial, although the DNA-binding site in Fos seems to extend beyond the basic region into an adjacent cluster of acidic amino acids.

fosB is a transforming gene encoding a transcriptional activator.

The fosB gene encodes a nuclear protein that shows a high degree of homology with c-Fos in several of the known functionally crucial domains, e.g., the leucine zipper and the DNA-binding site, but shows considerable divergence in other regions. Here, we report that FosB, when placed under the control of a constitutive promoter, exhibits clear transforming properties in focus assays using mouse NIH3T3 or rat 208F fibroblasts. The transforming potential of FosB is considerably stronger than that of a corresponding c-fos construct and resembles that of viral fos genes. Using chimeric fos/fosB constructs we show that the C-terminal half of FosB is responsible for these stronger transforming properties, apparently by giving rise to significantly higher levels of protein as compared with the corresponding c-fos sequence. Surprisingly, substitution of the N-terminus of Fos with that of FosB decreases its transforming potential. These differences in the transforming potential are not related to DNA or protein expression, but rather seem to reflect differences in the molecular function(s) encoded in the N-terminal halves of Fos and FosB protein. Both, fosB- and v-fos transformed cells show increased expression of a number of endogenous genes, including c-jun, transin, alpha 1(III) collagen and tissue plasminogen activator. Transactivation by FosB and v-fos of the c-jun and alpha 1(III) collagen gene promoters and of a 3 x TRE-tk chimeric promoter could be shown in transient CAT assays. v-Fos, but not FosB-transformed cells, also show elevated levels of urokinase and plasminogen activator inhibitor mRNAs, pointing to potential differences in the gene regulatory properties of the two Fos family members.

Transcriptional activation and transformation by chimaeric Fos-estrogen receptor proteins: altered properties as a consequence of gene fusion.

We have generated a series of conditionally active Fos and FosB proteins by fusion with a C-terminal fragment of the human estrogen receptor (ER) which harbours the ligand binding site and the overlapping hormone-inducible transactivation domain TAF-2. The chimaeric Fos-ER proteins showed estrogen-inducible activation of TRE (TPA-responsive element)-directed transcription and hormone-dependent transformation of fibroblasts. These properties of the fusion proteins were independent of the transregulatory and transforming properties of their normal non-fused counterparts c-Fos, v-Fos, FosB-L and FosB-S. Thus c-Fos-ER and FosB-S-ER were strong transforming proteins in the presence of hormone, although c-Fos and FosB-S possess only marginal oncogenic properties. In addition, all fusion proteins showed increased transactivation in the presence of estrogen, again most noticeable in the case of c-Fos-ER and FosB-S-ER. The ER-fusion thus basically eliminated the differences in the transactivating potential seen among the various native Fos proteins. Our data therefore provide evidence: (i) that the hormone binding domain of the human estrogen receptor, apart from delivering hormonal control to a heterologous protein, can have profound effects on the activity of certain transcription factors, particularly on proteins with weak oncogenic and/or transregulatory potential, and (ii) that the transforming potential of c-Fos and FosB-S can be dramatically elevated by increasing their transactivating properties.

Transition from SCLC to NSCLC phenotype is accompanied by an increased TRE-binding activity and recruitment of specific AP-1 proteins.

Transitions from small cell (SCLC) to non-small cell lung cancer (NSCLC) cells have been documented both in vitro and in vivo and are thought to be an important step during tumor progression of human small cell lung cancer towards a treatment-resistant tumor state. We have screened NSCLC and SCLC cell lines for differences in the composition of nuclear transcription factors using consensus oligonucleotide sequences (SRE, Ets, TRE, CRE, B-motif, GAS, E-box). We found NSCLC cells to exhibit significantly higher AP-1 binding activity than SCLC cells consistent with the increased expression of CD44, an AP-1 target gene. To gain more insight into the molecular mechanisms underlying these differences, we analysed SCLC cell lines (NCI-N592 and NCI-H69) which were phenotypically transformed into NSCLC-type cells by transfection with activated H-ras and c-myc oncogenes. In these cells, ras-induced transition is accompanied by a strong induction of AP-1-binding activity along with increased expression of CD44 mRNA and protein. When analysing the composition of the AP-1 complex in more detail and comparing ras-induced versus phorbol ester-induced changes, we found Fra-1 to be the major component induced in ras-transfected but not in phorbol-ester treated or non-treated parental SCLC cells. This finding is paralleled by the observation that among the various members of the Fos and Jun family analysed (c-Fos, FosB, Fra-1, Fra-2, c-Jun, JunD, JunB) fra-1 is the only gene to be exclusively expressed in NSCLC cells but not in cells of SCLC origin. Our data, thus, point to a histiotype-related mechanism of recruitment among AP-1 proteins which may have bearings on the fate of lung cancer development.

Differential frequencies of p16(INK4a) promoter hypermethylation, p53 mutation, and K-ras mutation in exfoliative material mark the development of lung cancer in symptomatic chronic smokers.

PURPOSE: The aim of this study was to investigate the frequency of three (epi)genetic alterations (p53 and K-ras mutations and p16(INK4a) promoter hypermethylation) in symptomatic chronic smokers compared with patients with lung cancer and to evaluate the use of exfoliative material for such analyses. PATIENTS AND METHODS: Fifty-one patients with histologically confirmed lung cancer and 25 chronic smokers (> 20 pack-years) were investigated for mutations in the K-ras (codon 12) and p53 (codons 248, 249, and 273) genes and for allelic hypermethylation of the p16(INK4a) gene. DNA was isolated from sputum and bilateral bronchial lavage, and brushings were taken at bronchoscopy. RESULTS: Forty-one genetic lesions were detected within exfoliative material from the group of 51 patients with lung cancer and 10 lesions in the chronic smoker group. K-ras mutations occurred exclusively in the lung cancer group, whereas p53 mutations and p16(INK4a) promoter hypermethylation were also found in chronic smokers. Three of eight chronic smokers who harbored an (epi)genetic alteration were subsequently diagnosed with lung cancer. Analysis of sputum yielded information equivalent to that of samples obtained during bronchoscopy. CONCLUSION: p16(INK4a) promoter hypermethylation and p53 mutations can occur in chronic smokers before any clinical evidence of neoplasia and may be indicative of an increased risk of developing lung cancer or of early disease. K-ras mutations occur exclusively in the presence of clinically detectable neoplastic transformation. Molecular analysis of sputum for such markers may provide an effective means of screening chronic smokers to enable earlier detection and therapeutic intervention of lung cancer.

Frequent detection of ras and p53 mutations in brush cytology samples from lung cancer patients by a restriction fragment length polymorphism-based "enriched PCR" technique.

RFLP-mediated PCR has been successfully applied as a reliable tool in the detection of ras mutations in many cancers and provides a basis for "mutant-enriched PCR" protocols. We have, therefore, modified this technique to the sensitive detection of K-ras codon 12 and also p53 "hot spot" mutations, which, frequently in lung cancer, affect codons at the positions 157, 175, 245, 248, 249, and 273. With a high sensitivity of 1 mutant allele in 10(4) normal alleles, our enrichment assay allows the detection of oncogene alleles when only a few tumor cells are present within a normal cell population. Brush cytology material obtained from the tumor site of 20 patients with endoscopically apparent bronchial carcinoma was compared to macroscopically normal mucosa taken from the contralateral bronchus ("control" cytology). We found K-ras codon 12 mutations in 5 cases (25%) and p53 mutations in 13 cases (65%) in the tumor-derived cell material but, with the exception of two cases, not in cell material taken from the control cytology. Seventy-five % of the samples analyzed showed that at least one of the two oncogenes was affected. In several cases, two p53 lesions were detected concomitantly. The majority of the mutations could be reconfirmed by an alternative approach exploiting changes in the genomic RFLP pattern induced by these mutations and were also demonstrated in separate diagnostic biopsies taken. Thus, we conclude that the established enriched PCR protocol ensures a high sensitivity and preserved specificity for the diagnosis of oncogene lesions associated with lung cancer. Because conventional techniques normally yield a lower incidence of corresponding ras and p53 mutations, we think that both the high rate and the heterogeneity of p53 mutations found in some cases are, indeed, related to the increased sensitivity of this new enriched PCR technique.

Axl receptor tyrosine kinase expression in human lung cancer cell lines correlates with cellular adhesion.

Axl is a receptor tyrosine kinase (RTK) with oncogenic potential and transforming activity. Since Axl bears structural similarities to cell adhesion molecules such as neural cell adhesion molecule (NCAM) (FNIII domains), it is thought that Axl might play a role in adhesion. In this study, we have analysed the expression of the Axl protein and its ligand, Gas6, in human lung cancer cell lines of different histological origin. Axl expression occurred in approximately 60% of non-small cell lung cancer (NSCLC) cell lines, which grow adherently, and in normal bronchial epithelial cells (NHBE), but not in cell lines of small cell lung cancer origin (SCLC), which grow in suspension. A number of SCLC sublines, which could be selected spontaneously or after oncogene transfection for adherent growth, all expressed Axl protein. Overexpression of Axl per se, however, did not induce any change in the adhesion phenotype. All Axl-expressing cell lines demonstrated a membrane-bound 140 kD form, as well as a soluble 85 kD form, detectable in supernatant, of Axl-RTK. Expression of the Axl ligand Gas6 was detected in approximately 80% of all cell lines investigated. We conclude from these data that loss of Axl expression is a feature of SCLC tumour cells. Axl expression appears to be a consequence of cellular adhesion and possibly influences differentiation in human lung cancers.

Simple and reliable factor V genotyping by PNA-mediated PCR clamping.

Resistance to activated protein C (APC resistance) is the most common cause of thrombophilia and linked to a single point mutation in the factor V gene (G-->A transition at nucleotide 1691). In the past, several PCR based methods have been proposed to determine the allelostatus of individual patients from small amounts of blood DNA including PCR followed by restriction fragment length polymorphism detection (PCR-RFLP), PCR using sequence-specific primers (PCR-SSP) and oligonucleotide ligation assay (OLA). Here, we present a novel approach based on the method of peptide nucleic acid(PNA)-mediated PCR clamping which is extremely sensitive to base pair mismatches. If PNAs specific for the two allelic variants are applied separately in each case a clear discrimination between a heterozygous or homozygous normal or homozygous Factor V Leiden status is possible and no further confirmation step is required. In a prospective study, 60 patients with suspected venous thrombosis events were tested and compared to the conventional PCR-RFLP technique. The concordance between both methods was 100%. PNA-based factor V genotyping, therefore, should be considered for large scale screening of those patients considered to be at risk for deep venous thrombosis.

Preferential histiotypic expression of CD44-isoforms in human lung cancer.

The CD44 transmembrane glycoprotein is expressed in most adult tissues and in the majority of neoplasias. Due to alternative splicing, this cell adhesion molecule exists in multiple isoforms some of which have been associated with specific types of tumours as well as with increased tumour metastasis. In this study, we have looked at the level and type of CD44 expression in lung cancer which represents a histologically heterogenous form of cancer composed of small cell lung carcinoma (SCLC) and non-small cell lung carcinoma (NSCLC), the latter subgroup comprising adenocarcinoma (ADC), bronchio-alveolar carcinoma (BAC), large cell carcinoma (LCC), and squamous cell carcinoma (SCC). We analysed 20 lung cancer cell lines and 64 primary tumours by RT-PCR and immunohistochemical detection of the CD44 standard and variant protein isoforms. Our results suggest that (i) CD44 is expressed in all histologically distinct subsets of lung cancer with a tendency SCC > BAC > ADC > LCC > SCLC, (ii) expression of the CD44 isoforms v5, v7, v8, and, most notably that of CD44 exon v6, strongly correlates with tumours of squamous cell and bronchio-alveolar carcinoma origin, tumours which commonly exhibit a comparatively low metastasizing potential, and (iii) the expression of CD44 isoforms is independent from the tumour size and lymph node status at surgery, the proliferative status of the tumour cell population (Ki67 antigen expression) and the histopathological grading (G1 to G3). Only non-differentiated tumours (G4), which were restricted to SCLC and LCC samples revealed markedly reduced CD44 standard and isoform antigen. In conclusion, our data point to a clear histiotype-related pattern of CD44 variant expression preferentially that of CD44v6 in SCC and BAC.

Transition from suspension to adherent growth is accompanied by tissue factor expression and matrix metalloproteinase secretion in a small cell lung cancer cell line.

Small cell lung cancer (SCLC) is a very malignant tumor known to grow aggressively and to metastasize early. It is well established that metastasis generally involves both tumor cell adhesion and proteolytic degradation of the extracellular matrix. However, SCLC cells cultured in vitro, such as the classic SCLC cell line NCI-H69, grow in floating aggregates and express only negligible proteolytic activity. In this report, we show that NCI-H69 cells can be selected for adherent growth. In contrast to parental suspension cells, the adherent cells were found to express tissue factor as well as gelatinolytic activity, attributable to matrix metalloproteinases 2 and 9. Such a switch of tumor cell characteristics, if it could occur in SCLC patients, might add to the understanding of the steps involved in the spreading of this highly metastatic type of lung cancer.

p16(INK4a) alterations in chronic pancreatitis-indicator for high-risk lesions for pancreatic cancer.

BACKGROUND: p16(INK4a) alterations are considered to be an early event in pancreatic tumorigenesis and have been described in duct lesions adjacent to pancreatic cancers. This study evaluates whether duct lesions in chronic pancreatitis tissues of patients without pancreatic cancer also harbor genetic alterations in the p16(INK4a) tumor-suppressor gene, and thus represent high-risk precursors for pancreatic cancer. METHODS: Tissues were obtained from 20 pancreatic specimens taken from patients operated on for histologically verified chronic pancreatitis. Pancreatic intraductal neoplasias (PanIN) were identified in hematoxylin-and-eosin-stained slides. p16 protein expression was investigated immunohistochemically in all specimens. DNA from PanIN and non-PanIN tissue was analyzed genetically for p16(INK4a) mutations by single-strand conformation variation analysis and direct sequencing of the encoding region. Additionally, p16(INK4a) promoter methylation was analyzed by a methylation specific polymerase test. RESULTS: PanIN-1a lesions were identified in 10 of the 20 chronic pancreatitis specimens. Four of these 10 PanIN specimens (40%), but none of the 20 non-PanIN tissues, revealed a loss of p16 expression in immunohistochemistry. The mutational analysis of the p16(INK4a) gene showed 1 known polymorphism (c.442G > A; A148T) but no mutations. Two of the 10 specimens with PanIN revealed an inactivating hypermethylation of the p16(INK4a) promoter. CONCLUSIONS: This study shows for the first time that p16(INK4a) alterations can be observed in a considerable number of PanIN1 in chronic pancreatitis tissues not associated with pancreatic cancer. Therefore, p16(INK4a) alterations, especially promoter methylation, might indicate high-risk precursors in chronic pancreatitis that might progress to cancer.

Multiple primary tumors as an indicator for p16INK4a germline mutations in pancreatic cancer patients?

Multiple primary tumors in pancreatic cancer patients might indicate a genetic predisposition to the development of malignancies. In this study we evaluated whether the mutation rate of the TP53 and p16INK4a genes of pancreatic cancers differs in pancreatic cancer patients with and without multiple primaries. Furthermore, we investigated whether pancreatic cancer patients with multiple primaries carry germline mutations in either p16INK4a, TP53, or BRCA2 tumor suppressor genes to detect a genetic alteration that predisposes to the development of different primaries. Fourteen (23%) of 60 pancreatic cancer patients developed histologically verified additional primaries during their lifetimes. Normal constitutional and tumor DNA of the 14 patients with a positive cancer history, but negative family history, were analyzed for p16INK4a, TP53, and BRCA2 mutations by single-strand conformational variant (SSCV) analysis and direct sequencing. Hypermethylation of the p16INK4a promoter region in pancreatic cancers was identified by methylation-specific polymerase chain reaction (PCR; MSP). Four of 14 pancreatic carcinomas carried somatic intragenic p16INK4a mutations, and another four tumors revealed hypermethylation of the p16INK4a promoter region. Somatic intragenic TP53 mutations were identified in six of 14 tumors. None of the pancreatic cancer patients carried TP53 or BRCA2 germline mutations. In contrast, one of 14 pancreatic cancer patients with multiple primaries carried the p16INK4a mutation A68V in his germline. This mutation was localized in the conserved second ankyrin repeat of p16INK4a and did not occur in 100 control patients. The frequency of somatic TP53 and p16INK4a mutations in pancreatic cancer is similar in patients with and without multiple primaries. TP53 and BRCA2 germline mutations seem not to be significantly associated with the occurrence of multiple primaries in pancreatic cancer patients. However, p16INK4a germline mutations might be causative for tumor development in some pancreatic cancer patients with multiple primaries. The genetic investigation of patients with accumulation of different cancers even without a positive family history may be a new approach for the understanding of the relation of different cancers.

A rare common integration site of proviruses of the mouse mammary tumor virus in P-type mammary tumors of mouse strain GR.

The mouse mammary tumor virus (MMTV) can induce mammary tumors in mice by proviral activation of the cellular oncogenes int-1 or int-2. Activation of these genes, however, is observed in only a few hormone- and pregnancy-dependent mammary tumors of the mouse strain GR. To study the possible involvement of other oncogenes we cloned three MMTV proviral-host fragments (MT 40, 42, and 53) from different mammary tumors of GR with a single acquired MMTV provirus. From a genomic library of normal mouse DNA we isolated phages with insert DNAs that covered 20-30 kb of the uninterrupted regions. Suitable probes devoid of repetitive DNA sequences were isolated in order to screen other mammary tumors for MMTV proviral integrations in these regions. Only two mammary tumors, MT 40 and 42, showed integration of extra MMTV proviruses within the same region. The integrations occurred only 60 bp apart. The other mammary tumors, however, did not contain MMTV proviral integrations in this region, nor in the MT 53 region. Using mouse-hamster somatic cell hybrid DNA, the MT 40/42 integration region was assigned to mouse chromosome 7, and the second region, MT 53, to chromosome 16. The two regions bear no homology to known cellular oncogenes. We did not observe any mRNA being expressed from these cloned segments either in tumors or in normal mammary glands. These findings indicate that plaque(P)-type mammary tumors in mouse strain GR do not originate from MMTV provirus insertions in a particularly favored integration region, but that there may be a variety of integration sites in these tumors.

Sensitive detection of p53 gene mutations by a 'mutant enriched' PCR-SSCP technique.

For the rapid and sensitive detection of p53 'hot spot' mutations, we combined polymerase chain reaction based single-strand conformational polymorphism (PCR-SSCP) analysis with sequence specific-clamping by peptide nucleic acids (PNAs) in a one-step reaction tube protocol. For this purpose, we designed two PNA molecules comprising aa 246-250 of exon 7 and aa 270-275 of exon 8, respectively, to suppress the amplification of wild-type p53 allelic variants during PCR amplification. Using this method in a survey of 20 brush cytology samples from lung cancer patients, we were able to detect five p53 point mutations occurring in codons 248, 249 and 273 which could not be retrieved by conventional PCR-SSCP. Thus, allelic suppression by PNA molecules opens a way to largely improve the sensitivity of existing PCR-SSCP protocols (approximately 10-50-fold) and could be useful in the detection of 'hot spot' oncogene lesions in histological samples containing only a small number of cancer cells.

Two functionally different regions in Fos are required for the sequence-specific DNA interaction of the Fos/Jun protein complex.

THE products of the cellular and retroviral fos genes associate with other nuclear proteins, among them the transcription factor AP1/Jun (see ref. 3 for a review). The Fos/Jun complex binds to a specific symmetrical DNA recognition sequence (termed TRE), thus stimulating transcription of the respective gene. Here, we show that two distinct regions in Fos are required for the formation of a Fos/Jun/TRE complex. These are the leucine zipper, involved in the association with Jun, and a directly adjacent basic region. Specific amino-acid substitutions in this basic, presumably alpha-helical, region abolish the interaction of Fos/Jun with the TRE but not the association of the two proteins. The functionally crucial amino acids are located in a region of Fos which is structurally similar to the putative DNA-binding sites in Jun and in the yeast transcriptional activator GCN4 (refs 15 and 16).

APC resistance as an additional thrombotic risk factor in a patient suffering from polycythemia vera and recurrent thrombosis.

Patients with polycythemia vera (PV) have a high risk of thrombosis. However, thrombosis is not sufficiently predictable with standard diagnostic procedures. We report on a patient with PV and recurrent thrombosis who nevertheless had a low platelet count while under therapy with hydroxyurea. As a result of duodenal ulcer and gastrointestinal bleeding, treatment with phenprocoumon was stopped years ago. Recently, heterozygosity for the factor V gene defect was diagnosed and anticoagulation therapy was reconsidered. In conclusion, the presence of resistance to activated protein C was an additional thrombotic risk factor that was important for our decision to change the treatment strategy in our patient.

T1 is a c-Fos- and FosB-responsive gene which is induced by growth factors through multiple signal transduction pathways.

Stimulation of quiescent cells with growth factors triggers changes in gene expression through multiple signal transduction pathways. One of these changes in Swiss 3T3 cells is the strong accumulation of T1 mRNA which encodes a secreted glycoprotein of the immunoglobulin superfamily. Proliferating cells continued to express T1 mRNA at a lower level, whereas growth arrest induced either by serum deprivation or by contact inhibition was paralleled by the disappearance of the T1 mRNA. T1 mRNA synthesis in response to serum and platelet-derived growth factor stimulation is mediated through protein kinase C-dependent and protein kinase C-independent pathways. Activation of protein kinase A also led to T1 gene expression. Ongoing protein synthesis is a prerequisite for T1 gene induction by growth factors which defines T1 as a delayed early serum-responsive gene. The ability of the immediate early transcription factors c-Fos and FosB to directly induce the T1 gene was demonstrated in a conditional expression system in the absence of protein synthesis. Furthermore, all known inducers of the T1 gene also lead to c-fos gene activation. Thus we show that the T1 gene is regulated by signals which are transduced through multiple pathways and provide evidence that the Fos proteins play an important role in the integration of these pathways.

Alternative splicing of fosB transcripts results in differentially expressed mRNAs encoding functionally antagonistic proteins.

We show that serum-stimulated fibroblasts transiently express two different forms of fosB mRNA, which are generated by alternative splicing of the transcript from a single gene. In addition to the known long form (fosB-L), encoding a protein of 338 amino acids (FosB-L), a second shorter form (fosB-S) with a deletion of 140 bp was detected. This deletion creates a stop codon 3' to the leucine repeat, giving rise to a protein of 237 amino acids (FosB-S) lacking the carboxyl terminus of FosB-L. Only the long FosB form efficiently induces transformation in mouse and rat fibroblast cell lines and trans-represses the c-fos promoter. Both of these functions are suppressed by coexpressed FosB-S. Upon serum stimulation, maximum expression of the oncogenic fosB-L form precedes the expression of the antagonistic fosB-S form, indicating a new mechanisms regulating the action of members of the Fos family. However, FosB-L and FosB-S do not differ in all trans-regulatory properties: Trans-activation of a 5x TRE-CAT reporter construct in HeLa and NIH-3T3 cells was found with both FosB forms. These observations suggest a correlation between fosB-induced transformation and trans-repression, thus pointing to different mechanisms involved in transformation by fosB and c-fos/v-fos.

Non-leucine residues in the leucine repeats of Fos and Jun contribute to the stability and determine the specificity of dimerization.

Various transcription factors, including C/EBP, GCN4 and members of the Fos, Jun and Myc families have been shown to form highly specific complexes via alpha-helical structures referred to as leucine zippers. Experimental evidence has suggested that dimerization involves the formation of hydrophobic bonds between leucine residues in laterally aligned coiled coil structures. However, the specificity of interaction between leucine zipper proteins is not understood. In this study, we show that amino acids, which are located in positions a, e, and g are instrumental in the formation of Fos/Jun heterodimers, presumably by establishing intermolecular electrostatic and hydrophobic interactions. These residues are highly conserved in proteins of the Fos or Jun families but completely different between Fos and Jun, suggesting that these residues determine the specificity of interaction. This conclusion is supported by the observation that the substitution of amino acids in position a or g in Fos with the corresponding Jun amino acids facilitates the association of two Fos leucine repeats. In addition, we show that a conserved histidine residue, located 7 amino acids (i.e., two alpha-helical turns) C-terminally to the 5th leucine in Fos and Jun, is also important for complex formation.