Improved predictive diagnosis of diabetic macular edema based on hybrid models: An observational study.

Diabetic Macular Edema (DME) is the most common sight-threatening complication of type 2 diabetes. Optical Coherence Tomography (OCT) is the most useful imaging technique to diagnose, follow up, and evaluate treatments for DME. However, OCT exam and devices are expensive and unavailable in all clinics in low- and middle-income countries. Our primary goal was therefore to develop an alternative method to OCT for DME diagnosis by introducing spectral information derived from spontaneous electroretinogram (ERG) signals as a single input or combined with fundus that is much more widespread. Baseline ERGs were recorded in 233 patients and transformed into scalograms and spectrograms via Wavelet and Fourier transforms, respectively. Using transfer learning, distinct Convolutional Neural Networks (CNN) were trained as classifiers for DME using OCT, scalogram, spectrogram, and eye fundus images. Input data were randomly split into training and test sets with a proportion of 80 %-20 %, respectively. The top performers for each input type were selected, OpticNet-71 for OCT, DenseNet-201 for eye fundus, and non-evoked ERG-derived scalograms, to generate a combined model by assigning different weights for each of the selected models. Model validation was performed using a dataset alien to the training phase of the models. None of the models powered by mock ERG-derived input performed well. In contrast, hybrid models showed better results, in particular, the model powered by eye fundus combined with mock ERG-derived information with a 91 % AUC and 86 % F1-score, and the model powered by OCT and mock ERG-derived scalogram images with a 93 % AUC and 89 % F1-score. These data show that the spontaneous ERG-derived input adds predictive value to the fundus- and OCT-based models to diagnose DME, except for the sensitivity of the OCT model which remains the same. The inclusion of mock ERG signals, which have recently been shown to take only 5 min to record in daylight conditions, therefore represents a potential improvement over existing OCT-based models, as well as a reliable and cost-effective alternative when combined with the fundus, especially in underserved areas, to predict DME.

Influence of HCV genotype 1 subtypes on the virus response to PEG interferon alpha-2a plus ribavirin therapy.

New factors that influence the viral response in HCV non-genotype 2/3 patients must be identified in order to optimize anti-HCV treatment. This multicenter prospective study evaluates the influence of HCV variability and pharmacological parameters on the virological response of these patients to pegylated interferon α2a (peg-IFN-α2a: 180 µg/week) and ribavirin (RBV; 800-1,200 mg/day) for 48 weeks. HCV subtypes were identified by sequencing the NS5B region. Serum RBV and peg-IFN-α2a concentrations were measured at weeks 4 and 12. The 115 patients (67 men; median age = 49, range 31-76) included 64 who had never been treated and 27 co-infected with HIV. The mean baseline HCV RNA was 6.30 ± 0.06 log IU/ml and the HCV genotypes were: G1 (n = 93) with 1a (n = 37) and 1b (n = 50), G4 (n = 20) and G5 (n = 2). Most patients (79/108; 73%) had an early virological response. Independent predictors of an early virological response were interferon naive patients (OR= 2.98, 95% CI: 1.15-7.72) and RBV of >2,200 ng/ml at week 12 (OR = 3.41, 95% CI: 1.31-8.90). Forty of 104 patients (38%) had a sustained virological response. The only independent predictors of a sustained virological response were subtype 1b (OR = 6.82, 95% CI: 1.7-26.8), and HCV RNA <15 IU/ml at week 12 (OR = 25, 95% CI: 6.4-97.6). Thus a serum RBV concentration of >2,200 ng/ml was associated with an early virological response and patients infected with HCV subtype 1b had a better chance of a sustained virological response than did those infected with subtype 1a.

[Etiology and outcome of liver granulomatosis: a retrospective study of 21 cases]. / Etiologies et évolution des hépatites granulomateuses : étude rétrospective de 21 cas consécutifs.

PURPOSE: To assess the etiologies and outcome of liver granulomatosis. METHODS: We analyzed all consecutive liver granulomatosis diagnosed in our internal medicine department from 2000 to 2008. RESULTS: Among 471 liver biopsies, 21 disclosed evidence of liver granulomatosis (4.5%), in sixteen women (76%) and five men, with a median age of 41years. Thirteen were caucasians (62%). At the time of diagnosis, six (28.5%) had isolated abnormal liver function tests, and fifteen (71.4%) presented with clinical manifestations. The underlying cause was identified in 18 cases (85.7%). Eleven (52.3%) were systemic diseases: five (23.8%) primary biliary cirrhosis, two (9.5%) primary sclerosing cholangitis, two (9.5%) common variable immunodeficiency, one (4.7%) Sjögren's syndrome, and one (4.7%) Behçet's disease. Two (9.5%) patients had sarcoidosis. Three (14.3%) liver granulomatosis were of infectious origin (tuberculosis, schistosomiasis, and hepatitis C virus), two (9.5%) were neoplastic (Hodgkin's lymphoma and liver cell adenoma), and three (14.3%) were idiopathic. With a median of 38 months of follow-up, four patients (19%, two common variable immunodeficiency and two sarcoidosis) developed portal hypertension, independently of cirrhosis. One patient died of cryptococcosis. CONCLUSION: In accordance with other European studies, systemic diseases are the main causes of hepatic granulomas. Liver granulomatosis related to common variable immunodeficiency and sarcoidosis are at risk of portal hypertension.

Differential role of TRP channels in prostate cancer.

A major clinical problem with PC (prostate cancer) is the cell's ability to survive and proliferate upon androgen withdrawal. Indeed, deregulated cell differentiation and proliferation, together with the suppression of apoptosis, provides the condition for abnormal tissue growth. Here, we examine the differential role of TRP (transient receptor potential) channels in the control of Ca(2+) homoeostasis and growth of PC cells.

Pilot study of interferon-alpha-ribavirin-interleukin-2 for treatment of nonresponder patients with severe liver disease infected by hepatitis C virus genotype 1.

The aim of this randomized prospective study was to assess the efficacy and safety of a triple therapy with interferon-alpha (IFN-alpha)-ribavirin-interleukin-2 (IL-2) for the treatment of patients with genotype 1 infection and high viral load nonresponsive to primary IFN-ribavirin therapy. Twenty hepatitis C virus (HCV) genotype 1 patients with high viral load and Metavir fibrosis score >or=2 without HIV co-infection who were previously nonviral responders to standard treatment with IFN plus ribavirin were intensively re-treated with IFN-alpha2a (3 millions (M) IU every 2 days) combined with ribavirin (1000-1200 mg/day) for a 24-week period. Patients were randomized to receive four cycles of subcutaneous injection of IL-2 (3 MIU/day, 5 days a week every 3 weeks) during either the first 12 weeks (group 1, n = 10) or the last 12 weeks (group 2, n = 10) of combination therapy. At the end of triple therapy, six patients (30%) achieved a biochemical response and 4 (20%) a viral response followed by a relapse after triple therapy withdrawal. After 12 weeks of therapy, no difference in viral load was observed between the groups. The decrease in viral load in group 2 was not raised after the addition of IL-2 to IFN plus ribavirin combination therapy. No serious adverse effects were observed. In conclusion, in patients with poor predictive factors of response, the addition of IL-2 to IFN ribavirin combination therapy does not exert a favourable impact on HCV treatment.

Alterations in the regulatory volume decrease (RVD) and swelling-activated Cl- current associated with neuroendocrine differentiation of prostate cancer epithelial cells.

Neuroendocrine (NE) differentiation of prostate epithelial/basal cells is a hallmark of advanced, androgen-independent prostate cancer, for which there is no successful therapy. Here we report for the first time on alterations in regulatory volume decrease (RVD) and its key determinant, swelling-activated Cl- current (I(Cl,swell)), associated with NE differentiation of androgen-dependent LNCaP prostate cancer epithelial cells. NE-differentiating regimens, namely, chronic cAMP elevation or androgen deprivation, resulted in generally augmented I(Cl,swell) and enhanced RVD. This occurred as a result of both the increased endogenous expression of ClC-3, which is a volume-sensitive Cl- channel involved, as we show, in I(Cl,swell) in LNCaP (lymph-node carcinoma of the prostate) cells and the weaker negative I(Cl,swell) control from Ca2+ entering via store-dependent pathways. The changes in the RVD of NE-differentiated cells generally mimicked those reported for Bcl-2-conferred apoptotic resistance. Our results suggest that strengthening the mechanism that helps to maintain volume constancy may contribute to better survival rates of apoptosis-resistant NE cells.

Evidence for specific TRPM8 expression in human prostate secretory epithelial cells: functional androgen receptor requirement.

TRPM8 (melastatine-related transient receptor potential member 8), a member of the transient receptor potential (TRP) superfamily of cation channels, has been shown to be a calcium-channel protein. TRPM8 mRNA has also been shown to be overexpressed in prostate cancer and is considered to play an important role in prostate physiology. This study was designed to determine the androgen-regulation mechanisms for TRPM8 mRNA expression and to identify the phenotype of TRPM8-expressing cells in the human prostate. Our findings show that trpm8 gene expression requires a functional androgen receptor. Furthermore, this article argues strongly in favour of the fact that the trpm8 gene is a primary androgen-responsive gene. Single-cell reverse transcriptase PCR and immunohistochemical experiments also showed that the trpm8 gene was mainly expressed in the apical secretory epithelial cells of the human prostate and trpm8 down-regulation occurred during the loss of the apical differentiated phenotype of the primary cultured human prostate epithelial cells. The androgen-regulated trpm8 expression mechanisms are important in understanding the progression of prostate cancer to androgen-independence. These findings may contribute to design a strategy to predict prostate cancer status from the TRPM8 mRNA level. Furthermore, as the TRPM8 channel is localized in human prostate cells, it will be interesting to understand its physiological function in the normal prostate and its potential role in prostate cancer development.

Receptor-operated Ca2+ entry mediated by TRPC3/TRPC6 proteins in rat prostate smooth muscle (PS1) cell line.

Prostate smooth muscle cells predominantly express alpha1-adrenoceptors (alpha1-AR). alpha1-AR antagonists induce prostate smooth muscle relaxation and therefore they are useful therapeutic compounds for the treatment of benign prostatic hyperplasia symptoms. However, the Ca(2+) entry pathways associated with the activation of alpha1-AR in the prostate have yet to be elucidated. In many cell types, mammalian homologues of transient receptor potential (TRP) genes, first identified in Drosophila, encode TRPC (canonical TRP) proteins. They function as receptor-operated channels (ROCs) which are involved in various physiological processes such as contraction, proliferation, apoptosis, and differentiation. To date, the expression and function of TRPC channels have not been studied in prostate smooth muscle. In fura-2 loaded PS1 (a prostate smooth muscle cell line) which express endogenous alpha1A-ARs, alpha-agonists epinephrine (EPI), and phenylephrine (PHE) induced Ca(2+) influx which depended on the extracellular Ca(2+) and PLC activation but was independent of PKC activation. Thus, we have tested two membrane-permeable analogues of diacylglycerol (DAG), oleoyl-acyl-sn-glycerol (OAG) and 1,2-dioctanoyl-sn-glycerol (DOG). They initiated Ca(2+) influx whose properties were similar to those induced by the alpha-agonists. Sensitivity to 2-aminoethyl diphenylborate (2-APB), SKF-96365 and flufenamate implies that Ca(2+)-permeable channels mediated both alpha-agonist- and OAG-evoked Ca(2+) influx. Following the sarcoplasmic reticulum (SR) Ca(2+) store depletion by thapsigargin (Tg), a SERCA inhibitor, OAG and PHE were both still able to activate Ca(2+) influx. However, OAG failed to enhance Ca(2+) influx when added in the presence of an alpha-agonist. RT-PCR and Western blotting performed on PS1 cells revealed the presence of mRNAs and the corresponding TRPC3 and TRPC6 proteins. Experiments using an antisense strategy showed that both alpha-agonist- and OAG-induced Ca(2+) influx required TRPC3 and TRPC6, whereas the Tg-activated ("capacitative") Ca(2+) entry involved only TRPC3 encoded protein. It may be thus concluded that PS1 cells express TRPC3 and TRPC6 proteins which function as receptor- and store-operated Ca(2+) entry pathways.

Receptor-coupled, DAG-gated Ca2+-permeable cationic channels in LNCaP human prostate cancer epithelial cells.

Although the prostate gland is a rich source of alpha1-adreno- (alpha1-AR) and m1-cholino receptors (m1-AChR), the membrane processes associated with their activation in glandular epithelial cells is poorly understood. We used the whole-cell patch-clamp technique to show that the agonists of the respective receptors, phenylephrine (PHE) and carbachol (CCh), activate cationic membrane currents in lymph node carcinoma of the prostate (LNCaP) human prostate cancer epithelial cells, which are not dependent on the filling status of intracellular IP3-sensitive Ca2+ stores, but directly gated by diacylglycerol (DAG), as evidenced by the ability of its membrane permeable analogue, OAG, to mimic the effects of the agonists. The underlying cationic channels are characterized by the weak field-strength Eisenman IV permeability sequence for monovalent cations (PK(25) > PCs(4.6) > PLi(1.4) > PNa(1.0)), and the following permeability sequence for divalent cations: PCa(1.0) > PMg(0.74) > PBa(0.6) > PSr(0.36) > PMn(0.3). They are 4.3 times more permeable to Ca2+ than Na+ and more sensitive to the inhibitor 2-APB than SK&F 96365. RT-PCR analysis shows that DAG-gated members of the transient receptor potential (TRP) channel family, including TRPC1 and TRPC3, are present in LNCaP cells. We conclude that, in prostate cancer epithelial cells, alpha1-ARs and m1-AChRs are functionally coupled to Ca2+-permeable DAG-gated cationic channels, for which TRPC1 and TRPC3 are the most likely candidates.

The cAMP response element binding protein-2 (CREB-2) can interact with the C/EBP-homologous protein (CHOP).

cAMP response element binding protein-2 (CREB-2) is a basic leucine zipper (bZIP) factor that was originally described as a repressor of CRE-dependent transcription but that can also act as a transcriptional activator. Moreover, CREB-2 is able to function in association with the viral Tax protein as an activator of the human T-cell leukemia virus type I (HTLV-I) promoter. Here we show that CREB-2 is able to interact with C/EBP-homologous protein (CHOP), a bZIP transcription factor known to inhibit CAAT/enhancer-dependent transcription. Cotransfection of CHOP with CREB-2 results in decreased activation driven by the cellular CRE motif or the HTLV-I proximal Tax-responsive element, confirming that CREB-2 and CHOP can interact with each other in vivo.

How the sequestration of a protein interferes with its mechanism of action: example of a new family of proteins characterized by a particular cysteine-rich carboxy-terminal domain involved in gene expression regulation.

We describe here a new family of proteins characterized by a particular cysteine-rich carboxy-terminal domain and involved in gene expression regulation. This family presently includes three members: I-mfa (inhibitor of MyoD family), HIC p40 and HIC p32 (human I-mfa domain-containing protein). I-mfa, by interacting with MyoD family members, represses both transcriptional activation and myogenesis mediated by these factors. HIC two isoforms, HIC p40 and HIC p32, are involved in the positive regulation of Tax-mediated HTLV-I (human T-cell leukemia virus type 1) promoter activation and in the negative regulation of Tat-mediated HIV-1 (human immunodeficiency virus type 1) promoter transcription. The common carboxy-terminal region of HIC p40 and HIC p32, which is clearly involved in these regulations, shares 77% homology with the carboxy-terminal domain of I-mfa. This suggests that I-mfa, HIC p40 and HIC p32 are part of a new family of proteins involved in gene expression regulation and characterized by a specific cysteine-rich carboxy-terminal domain. Moreover, the three proteins present different subcellular localizations: I-mfa and HIC p32 are mainly cytoplasmic while HIC p40 is mainly nucleolar. The specific localization of each member of this new family will be discussed, possibly explaining how they work. Effectively, a mechanism of protein sequestration in a particular compartment, cytoplasm or nucleolus, could be involved in their function, as it is the case for many other proteins. This relationship between sequestration and function regulation will be exemplified for several cellular factors.

Critical residues of epitopes recognized by several anti-p53 monoclonal antibodies correspond to key residues of p53 involved in interactions with the mdm2 protein.

The aim of this work was to study the reactivity of antibodies directed against the N-terminus of p53 protein. First, we analysed the cross-reactivity of anti-p53 antibodies from human, mouse and rabbit sera with peptides derived from human, mouse and Xenopus p53. Next, we characterized more precisely a series of monoclonal antibodies directed against the N-terminal part of p53 and produced by immunizing mice with either full length human or Xenopus p53. For each of these mAbs we localized the epitope recognized on human p53 by the Spot method of multiple peptide synthesis, defined critical residues on p53 involved in the interaction by alanine scanning replacement experiments and determined kinetic parameters using real-time interaction analysis. These antibodies could be divided into two groups according to their epitopic and kinetic characteristics and their cross-reactivity with murine p53. Our results indicate that critical residues involved in the interaction of some of these mAbs with p53 correspond to key residues on p53 involved in its interaction with the mdm2 protein. These antibodies could, therefore, represent powerful tools for the study of p53 regulation.

Two-dimensional electrophoresis and mass spectrometry identification of proteins bound by a murine monoclonal anti-cardiolipin antibody: a powerful technique to characterize the cross-reactivity of a single autoantibody.

Antigenic cross-reactivity, i.e., the capacity of a single antibody to react with apparently dissimilar structures, is a common characteristic of autoantibodies produced during systemic lupus erythematosus (SLE), an autoimmune disease developed by humans and certain strains of mice. Characterization of the extent of cross-reactivity of SLE-related autoantibodies may help identify the immunogenic stimulus, or stimuli, of autoantibody-secreting B-lymphocytes. Two-dimensional polyacrylamide gel electrophoresis (2-D PAGE) was combined with mass spectrometry (MS) to identify cell proteins recognized by a single monoclonal autoantibody (mAb 4B7), derived from an (NZW x BXSB)F1 mouse and selected based on its capacity to react with cardiolipin, that binds to elements in the cytoplasm and nucleoli of HEp-2 cells as assessed by indirect immunofluorescence assay. Proteins from HL-60 extract were separated by 1-D and 2-D PAGE. Western blotting with mAb 4B7 after SDS-PAGE revealed four bands, two intensely labeled at 35 and 32 kDa, and two weaker ones at 20 and 60 kDa; three spots were detected after 2-D PAGE. After trypsin in-gel digestion of the three protein spots, MS yielded representative matrix assisted laser desorption/ionization-time of flight (MALDI-TOF) Reflector or quadrupole-time of flight (Q-TOF) spectra. The three corresponding proteins were identified as the nucleolar phosphoprotein B23 (nucleophosmin), heterogeneous nuclear ribonucleoprotein A2 (hnRNP A2) and the 60 kDa Ro/SS-A RNP. Thus, these results showed that 2-D PAGE combined with MS constitutes a sensitive and powerful technique to characterize the full extent of cross-reactivity of a single mAb and may constitute a new approach to further characterize the immunogenic cellular components involved in the breakage of B-cell tolerance observed in SLE.

Molecular interactions involved in the transactivation of the human T-cell leukemia virus type 1 promoter mediated by Tax and CREB-2 (ATF-4).

The human T-cell leukemia virus type 1 (HTLV-1) Tax protein activates viral transcription through three 21-bp repeats located in the U3 region of the HTLV-1 long terminal repeat and called Tax-responsive elements (TxREs). Each TxRE contains nucleotide sequences corresponding to imperfect cyclic AMP response elements (CRE). In this study, we demonstrate that the bZIP transcriptional factor CREB-2 is able to bind in vitro to the TxREs and that CREB-2 binding to each of the 21-bp motifs is enhanced by Tax. We also demonstrate that Tax can weakly interact with CREB-2 bound to a cellular palindromic CRE motif such as that found in the somatostatin promoter. Mutagenesis of Tax and CREB-2 demonstrates that both N- and C-terminal domains of Tax and the C-terminal region of CREB-2 are required for direct interaction between the two proteins. In addition, the Tax mutant M47, defective for HTLV-1 activation, is unable to form in vitro a ternary complex with CREB-2 and TxRE. In agreement with recent results suggesting that Tax can recruit the coactivator CREB-binding protein (CBP) on the HTLV-1 promoter, we provide evidence that Tax, CREB-2, and CBP are capable of cooperating to stimulate viral transcription. Taken together, our data highlight the major role played by CREB-2 in Tax-mediated transactivation.

Molecular cloning of a novel human I-mfa domain-containing protein that differently regulates human T-cell leukemia virus type I and HIV-1 expression.

Regulation of viral genome expression is the result of complex cooperation between viral proteins and host cell factors. We report here the characterization of a novel cellular factor sharing homology with the specific cysteine-rich C-terminal domain of the basic helix-loop-helix repressor protein I-mfa. The synthesis of this new factor, called HIC for Human I-mfa domain-Containing protein, is controlled at the translational level by two different codons, an ATG and an upstream non-ATG translational initiator, allowing the production of two protein isoforms, p32 and p40, respectively. We show that the HIC protein isoforms present different subcellular localizations, p32 being mainly distributed throughout the cytoplasm, whereas p40 is targeted to the nucleolus. Moreover, in trying to understand the function of HIC, we have found that both isoforms stimulate in T-cells the expression of a luciferase reporter gene driven by the human T-cell leukemia virus type I-long terminal repeat in the presence of the viral transactivator Tax. We demonstrate by mutagenesis that the I-mfa-like domain of HIC is involved in this regulation. Finally, we also show that HIC is able to down-regulate the luciferase expression from the human immunodeficiency virus type 1-long terminal repeat induced by the viral transactivator Tat. From these results, we propose that HIC and I-mfa represent two members of a new family of proteins regulating gene expression and characterized by a particular cysteine-rich C-terminal domain.

Sequence requirement for the nucleolar localization of human I-mfa domain-containing protein (HIC p40).

The human I-mfa domain-containing protein (HIC) mRNA produces two protein isoforms, HIC p32 and p40, synthesized from alternative translational initiations. p32 translation is initiated from a standard AUG codon and p40 is an N-terminal extension of p32 generated from an upstream GUG codon. The two isoforms show different subcellular localization: p32 is distributed throughout the cytoplasm whereas p40 can be found both in the cytoplasm and the nucleolus. To investigate the possibility that p40 contains a nucleolus targeting sequence in its N-terminal region, COS cells were transfected with an eukaryotic expression vector coding for green fluorescent protein (GFP) fused to the p40 N terminus. The localization of this fusion protein in the nucleolus indicated that the N-terminal amino acids of p40 probably contain a nucleolar localization signal (NoLS). To find the structural motifs required for nucleolar localization of p40, deletion mutants were expressed in COS cells as fusion polypeptides with GFP. We defined a domain of 19 amino acids near the N terminus that contains an arginine-rich subdomain that conforms to other known NoLS. To demonstrate that this sequence is an authentic NoLS, the sequence was fused to GFP. This fusion protein was observed to migrate into the nucleolus. Taken together, our studies demonstrate that p40 contains a NoLS.

CREB-2, a cellular CRE-dependent transcription repressor, functions in association with Tax as an activator of the human T-cell leukemia virus type 1 promoter.

The Tax protein of the human T-cell leukemia virus type 1 (HTLV-1) has been implicated in human T-cell immortalization. The primary function of Tax is to transcriptionally activate the HTLV-1 promoter, but Tax is also known to stimulate expression of cellular genes. It has been reported to associate with several transcription factors, as well as proteins not involved in transcription. To better characterize potential cellular targets of Tax present in infected cells, a Saccharomyces cerevisiae two-hybrid screening was performed with a cDNA library constructed from the HTLV-1-infected MT2 cell line. From this study, we found 158 positive clones representing seven different cDNAs. We focused our attention on the cDNA encoding the transcription factor CREB-2. CREB-2 is an unconventional member of the ATF/CREB family in that it lacks a protein kinase A (PKA) phosphorylation site and has been reported to negatively regulate transcription from the cyclic AMP response element of the human enkephalin promoter. In this study, we demonstrate that CREB-2 cooperates with Tax to enhance viral transcription and that its basic-leucine zipper C-terminal domain is required for both in vitro and in vivo interactions with Tax. Our results confirm that the activation of the HTLV-1 promoter through Tax and factors of the ATF/CREB family is PKA independent.

Activation of E2F-mediated transcription by human T-cell leukemia virus type I Tax protein in a p16(INK4A)-negative T-cell line.

The human T-cell leukemia virus type I (HTLV-I) is a causative agent of adult T-cell leukemia. Although the exact mechanism by which HTLV-I contributes to leukemogenesis is still unclear, the Tax protein is thought to play a major role in this process. This 40-kDa polypeptide is able to interact with the tumor suppressor p16(INK4A). Consequently, Tax can activate the signaling pathway that lead to the release of E2F that in turn induces expression of factors required for cell cycle progression. In this paper, we demonstrate that Tax can also activate E2F-mediated transcription independently of p16(INK4A). Indeed, when Tax is coexpressed with the E2F-1 transcription factor in CEM T-cells, which lack expression of p16(INK4A), it strongly potentiates the E2F-dependent activation of a reporter construct driven by a promoter containing E2F binding sites. This stimulation is abrogated by mutations affecting the E2F-binding sites. In addition, Tax also stimulates the transcription of the E2F-1 gene itself. Using Tax mutants that fail to activate either ATF- or NF-kappaB-dependent promoters and different 5' truncation mutants of the E2F-1 promoter, we show that the Tax-dependent transcriptional control of the E2F1 gene involves, at least in part, the ATF binding site located in the E2F-1 promoter.