![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
|
|
|
|
|
||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Unidad de Hepatología Experimental, Centro de Investigación, Hospital Universitario La Fe, Valencia, Spain (M.P., M.J.G.-L., J.V.C., R.J.); CIBERehd: Centro de Investigación Biomédica en Red en el Área Temática de Enfermedades Hepáticas y Digestivas, Fondo de Investigación Sanitaria, Barcelona, Spain (M.P., M.J.G.-L., J.V.C., R.J.); and Departamento de Bioquímica y Biología Molecular, Facultad de Medicina, Universidad de Valencia, Valencia, Spain (J.V.C., R.J.)
(Received October 23, 2007; accepted March 5, 2008)
 |
Abstract |
|---|
 Top Abstract Materials and MethodsResultsDiscussionReferences |
|---|
). ATFs diverge considerably outside their bZIP DNA-binding regions, and it is the nonconserved portions that contain the diverse transcriptional regulatory domains. More than 20 different mammalian cDNAs with the prefix ATF or CREB have been described, and a large functional diversity has been found (Hai and Hartman, 2001). A comparison of the primary structures of bZIP transcription factors revealed that ATF5 is closely related to ATF4, which is involved in the integration of stress responses, such as the unfolded protein response, or those induced by amino acid deprivation and oxidants (Harding et al., 2003).
ATF5 has been associated with multiple distinct processes such as the differentiation of neural progenitor cells (Angelastro et al., 2003), the repression of cAMP-induced transcription in JEG3 choriocarcinoma cells (Pati et al., 1999), the inhibition or promotion of apoptosis (Persengiev et al., 2002; Wei et al., 2006), the response to amino acid limitation in HeLa cells (Watatani et al., 2007), and the circadian rhythm in adrenal medulla chromaffin cells (Lemos et al., 2007). ATF5 has been extensively investigated in the brain (Angelastro et al., 2005, 2006; Mason et al., 2005), but few studies have investigated the functionality of ATF5 in other tissues and cell types (Pati et al., 1999; Lemos et al., 2007; Wang et al., 2007; Watatani et al., 2007). Noteworthy, tissue expression analyses by Northern blot suggest that ATF5 is efficiently expressed in the adult liver (Peters et al., 2001; Hansen et al., 2002; Forgacs et al., 2005). Despite this, however, the role of ATF5 in human hepatic cells has not yet been investigated.
Cytochromes P450 (P450) are a superfamily of monooxygenases, playing a key role in the oxidative metabolism of biological signaling molecules such as steroids, and xenochemicals including pharmaceutical drugs and environmental contaminants (Nelson et al., 1996). P450 enzymes generally play a protective role against these xenochemicals and also in the stress responses elicited by bile acids in cholestasis and by hyperbilirubinemia (Goodwin and Moore, 2004; Qatanani and Moore, 2005). Many P450s are highly expressed in the liver, and recent studies have shown that their hepatic-specific expression is primarily governed at the transcriptional level by the concerted action of both liver-enriched and ubiquitous transcription factors (Akiyama and Gonzalez, 2003). In addition, the P450 expression can be transcriptionally induced by exposure of cells to xenochemicals as an adaptive response to increase the removal of potentially toxic xenobiotics and endobiotics. In this context, constitutive androstane receptor (CAR), a nuclear receptor with a high expression level in the liver, mediates the induction of P450s and other genes involved in the humoral response to both endobiotic and xenobiotic harmful stress (Qatanani and Moore, 2005). A prototypic CAR target gene in the human liver is CYP2B6, which shows an induced expression after exposure to phenobarbital (PB), and to agonist ligands such as CITCO (Honkakoski et al., 1998; Wang and Negishi, 2003). However, CAR is also an apparently constitutive transactivator, whose constitutive activity is inhibited by the inverse agonist ligands androstanol and androstenol (Qatanani and Moore, 2005).
In the present study, we show that ATF5 is an unexpectedly abundant liver-enriched transcription factor (LETF) whose expression declines in dedifferentiated hepatoma cells and in cultured hepatocytes. Among the several potential roles in the liver we have found ATF5 activates the human CYP2B6 and cooperates with CAR in sustaining the hepatic-specific expression of this P450 in human hepatocytes and hepatoma cells. We have also found that ATF5 and its target gene CYP2B6 are induced under endoplasmic reticulum (ER) stress conditions, pointing to an alternative mechanism to adapt P450-mediated detoxification to specific harmful stress conditions.
|
Materials and Methods |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
Cell Culture. Human hepatoma cells (HepG2, Hep3B, and Mz) were plated in Ham's F-12/Leibovitz L-15 (1:1, v/v) supplemented with 6% fetal calf serum and cultured to 60 to 70% confluence. Human hepatoma BC2 cells were cultured in a mixture of 75% minimal essential medium and 25% Medium 199, supplemented with 10% fetal bovine serum, 1 mg/ml bovine serum albumin, 0.7 µM insulin, and hydrocortisone hemisuccinate, and maintained at confluence for 3 weeks. HeLa (human cervix carcinoma) and human embryonic kidney (HEK) 293 (AdE1A-transformed HEK) were grown in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum and maintained as monolayer cultures. The culture medium for HEK293 cells was also supplemented with 3.5 g/l glucose.
Human hepatocytes were either isolated from liver biopsies (1–3 g) of patients undergoing liver surgery after informed consent or from liver organ donations for transplantation. All the liver samples were from healthy areas of the liver and from donors who were not suspected of harboring any infectious disease and tested negative for human immunodeficiency virus and hepatitis. None of the patients or donors was a habitual consumer of alcohol or other drugs. Samples were obtained in conformance to the rules of the Hospital's Ethics Committee. Ten liver biopsies (four male and six female, aged between 2 and 79 years) were used. Hepatocytes were isolated using a two-step perfusion technique (Gomez-Lechon and Castell, 2000) and then seeded onto fibronectin-coated plastic dishes (3.5 µg/cm2) at a density of 8 x 104 viable cells/cm2. Hepatocytes were cultured at 37°C in an atmosphere of 5% CO2 in air using Ham's F-12/Williams' E medium (1:1) supplemented with 2% newborn calf serum, 50 mU/ml penicillin, 50 µg/ml streptomycin, 0.1% bovine serum albumin, 10–8 M insulin, 25 µg/ml transferrin, 0.1 µM sodium selenite, 65.5 µM ethanolamine, 7.2 µM linoleic acid, 17.5 mM glucose, 6.14 mM ascorbic acid, and 0.64 mM N-omega-nitro-l-arginine methyl ester. The medium was changed 1 h later to remove unattached hepatocytes. At 24 h, the cells were shifted to serum-free hormone-supplemented medium (10 nM dexamethasone and insulin).
Adenoviral Vectors and Infection. A recombinant adenovirus encoding human ATF5 was produced as follows: ATF5 mRNA was amplified from a human liver cDNA pool (n = 12) with the Expand High Fidelity PCR System (Roche, Barcelona, Spain). Primers for ATF5, forward: 5'-GCA CGA ATT CTA CAG CCATGT CAC T-3' and reverse: 5'-CAG AAG CTT CAC CCC TGC CCT TCT A-3', amplified a predicted 879-base pair DNA fragment. In these sequences, bold letters indicate the start and stop translation codons in the forward and reverse primers, respectively. The forward primer includes an EcoRI site, whereas the reverse primer includes a HindIII site (italicized). The PCR product was purified by agarose gel electrophoresis, double-digested with EcoRI/HindIII, and directionally ligated into the adenoviral shuttle vector pAC/CMVpLpA, which was previously digested with the same restriction enzymes. The presence of the insert in the correct orientation was confirmed by restriction enzyme digestion. Sequence analysis of the ATF5 insert revealed a single base change with regard to published human ATF5 sequences. This mutation results in a single amino acid substitution (P159L; NP_036200.2
[GenBank]
) located out of the known functional domains of the protein. The plasmid construct was cotransfected with the vector pJM17, containing the full-length E1 defective adenovirus-5 genome (dl309), into HEK293 cells by calcium phosphate/DNA coprecipitation. A homologous recombination between adenovirus sequences in both the shuttle vector pAC/CMV-pLpA and the pJM17 plasmid generates a genome of a packable size in which most of the adenovirus early region 1 is lacking, thus rendering the recombinant virus replication defective (Jover et al., 2001). The resulting virus [named adenovirus (Ad)-ATF5] was plaque-purified, expanded into a high-concentration stock, and titrated by plaque assay as previously described.
To generate a recombinant adenovirus for the expression of human CAR, a similar strategy was used. The primers used to amplify CAR cDNA from human liver were forward: 5'-CCA CCC CAA CAG TCG ACG TCA TG-3' and reverse: 5'-GGT CCA AGC TTT TTC CCA CTC C-3' (SalI and HindIII italicized). This primer pair amplified a predicted 1139-base pair fragment. The presence of the insert in the correct orientation was confirmed by restriction enzyme digestion and sequence analysis.
Cultured human hepatocytes (24 h) and HepG2 cells were infected with recombinant adenoviruses for 120 min at a multiplicity of infection (m.o.i.) ranging from 6 to 60 plaque-forming units/cell. Thereafter, cells were washed and fresh medium was added. At 48 h post-transfection, cells were analyzed or directly frozen in liquid N2.
ER Stress Induction. For ER stress, HepG2 cells were incubated with either 5 µg/ml tunicamycin or vehicle, DMSO, in Dulbecco's modified Eagle's medium supplemented with 6% fetal calf serum. Alternatively, HepG2 cells and human hepatocytes were incubated in amino acid–free Krebs-Ringer bicarbonate (KRB) buffer (117.8 mM NaCl, 4.6 mM KCl, 1.2 mM KHPO4, 1.2 mM Mg2SO4·7H2O, 25 mM NaHCO3, 10 mM glucose, and 1.25 mM CaCl2). At each time, cells were collected and washed once with phosphate-buffered saline (PBS) one time and then processed for mRNA and/or protein extraction as described below.
Purification and Quantification of mRNA Levels. Total cellular RNA was extracted using TRIzol reagent (Invitrogen). Total RNAs from the 18 human tissues analyzed in Fig. 1A were purchased from Ambion (Austin, TX) (FirstChoice Human Total RNA Survey Panel). Reverse transcription was done using the Moloney murine leukemia virus reverse transcriptase and 1 µg of total RNA following the manufacturer's instructions (Invitrogen). cDNA was diluted at 1/20, and 3 µl was amplified with the rapid thermal cycler "Lightcycler" from Roche in 15 µl of LightCycler FastStart DNA Master SYBR Green I (Roche), 0.3 µM each primer, and with the optimum amount of MgCl2 (3–5 mM) (Table 1). In parallel, we analyzed the mRNA concentration of several human housekeeping genes: porphobilinogen deaminase, β-actin, and TATA-box binding protein as internal controls for normalization. After initial denaturing for 8 min at 95°C, amplification was performed using optimized conditions for each primer set (Table 1). PCR amplicons were confirmed to be specific both by size (agarose gel electrophoresis) and melting curve analysis. The real-time monitoring of the PCR reaction and the precise quantification of the products in the exponential phase of the amplification were performed with the LightCycler quantification software according to the manufacturer's recommendations (Roche).
|
|
Extraction of Protein and Immunoblotting. Nuclear extracts from cultured cells were prepared as described (Andrews and Faller, 1991). Protein concentrations were measured by the Bradford assay (Bio-Rad, Madrid, Spain). Cell proteins were resolved by SDS-polyacrylamide gel electrophoresis on a 12% gel, and the separated proteins were transferred electrophoretically to polyvinylidene difluoride membranes (Millipore, Madrid, Spain) using the Miniprotean II system (Bio-Rad). Membranes were blocked overnight in PBS containing 5% skimmed milk and were immunolabeled for 1 h at room temperature with ATF5 antipeptide antiserum provided by Dr. Angelastro (Angelastro et al., 2003) or from Imgenex (IMG-3026, Imgenex, San Diego, CA) and incubated overnight at 4°C. Dilution of both antibodies was 1/500 in PBS containing 5% skimmed milk. For detection, blots were washed and probed with secondary horseradish peroxidase-conjugated antibody (Santa Cruz Biotechnology, Santa Cruz, CA) and were visualized on film via an enhanced chemiluminescence detection kit (Amersham, Barcelona, Spain). Equal loading was verified by Coomassie staining of the membranes.
Statistical Analysis. Each experiment was performed in several independent cell cultures as indicated. Each quantitative determination was done at least in duplicate. The results are expressed as the mean value ± S.D. Statistical significance was calculated by the Student's t test.
|
Results |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
, HNF4
, and CCAAT/enhancer-binding protein (C/EBP) .
The results depicted in Fig. 1A showed that ATF5 mRNA is detectable in all the tissues, but the highest expression is found in the liver, where the mRNA level is at least 7-fold higher than in any other tissue examined. The expression profile of ATF5 was more hepatic-specific than other well recognized LETFs, such as HNF3, HNF4, or C/EBP. ATF4, the closest ATF5 homolog, shows a ubiquitous expression profile that is very different from the typical liver-enriched regulators in general and from ATF5 in particular (Fig. 1A). On the other hand, ATF5 mRNA showed a higher concentration than the other LETFs in absolute terms (Fig. 1B), indicating that ATF5 is an unusually abundant LETF.
Second, we investigated the expression of ATF5 in different hepatic and nonhepatic cell models. Preliminary evidence of a down-regulation of ATF5 in in vitro cell models was obtained from a whole genome expression analysis. Human liver tissue, human primary cultured hepatocytes, and HepG2 cells were analyzed and compared by using Affymetrix (Santa Clara, CA) human expression arrays (HG-U133). The results showed that ATF5 mRNA was down-regulated in human in vitro hepatic cell systems. The ATF5 mRNA level in cultured human hepatocytes (48 h) was only 20 to 30% of the tissue level. More strikingly, HepG2 cells only displayed 2% of the liver content (Table 2). The expression of other ATF factors (ATF2, ATF3, and ATF4) did not show a similar profile and were not repressed in human in vitro hepatic cell systems (Table 2).
|
To further show the ATF5 differential expression, we performed a Q-RT-PCR analysis in the human liver and in several hepatic and nonhepatic cell models. The results confirmed previous evidence from the expression arrays and showed that ATF5 is greatly down-regulated in human hepatoma cells to levels as low as those found in nonhepatic cell lines (Fig. 2). Cultured human hepatocytes (48 h) also presented a significantly decreased level, although not as pronounced as in cell lines (Fig. 2). These data collectively show that a high ATF5 expression is associated with the adult hepatic phenotype and also support the hypothesis that ATF5 may be a key transcription factor for the maintenance of hepatocyte differentiation and/or maturation.
|
|
To further investigate the potential role of ATF5 in the liver function, HepG2 cells were transfected with increasing doses of Ad-ATF5, and the impact on the expression of characteristic hepatic genes was assessed by Q-RT-PCR. As a member of the ATF/CREB family, ATF5 could play a role in the regulation of genes involved in the response to cAMP signaling, such as PEPCK or AldoB, two typical hepatic genes for glucose/fructose metabolism. The expression analysis showed that ATF5 was able to significantly induce these genes. Nonetheless, their response to ATF5 was rather weak (Fig. 4), and the dose-response profile did not change when cells were treated with either glucagon or 8-Br-cAMP (data not shown).
|
The battery of characteristic hepatic genes analyzed also included eight drug-metabolizing P450s. The results showed that CYP2B6 was specifically and robustly up-regulated by ATF5 (Fig. 4), suggesting that this transcription factor could be involved in the expression and regulation of the drug-metabolizing CYP2B6 in the liver. We also found a weak activation of CYP2C9, CYP2C19, and CYP1A2 (up to 2-fold increase), but CYP2E1, CYP2D6, CYP3A4, and CYP3A5 were not activated by ATF5 (data not shown).
ATF5 Cooperates with CAR in the Transcriptional Activation of CYP2B6 in Human Hepatoma Cells. The nuclear receptor CAR is one of the best characterized transcriptional activators of CYP2B6 (Kawamoto et al., 1999). Therefore, we investigated a potential cooperation between ATF5 and CAR in the regulation of this gene.
HepG2 cells were cotransfected with recombinant adenoviruses for ATF5 and CAR, and their effect on CYP2B6 mRNA levels was assessed by Q-RT-PCR. Transduction with Ad-ATF5 caused an important dose-dependent activation, but transduction with Ad-CAR had a more limited activating effect on CYP2B6. However, the cotransfection of ATF5 and CAR elicited an important additive induction, leading to a 15- to 20-fold increase in the mRNA levels of CYP2B6 (Fig. 5).
|
ATF5 and CAR Promote Higher CYP2B6 Basal Expression and Induction by Xenochemicals in Cultured Human Hepatocytes. Our findings in HepG2 cells suggest that ATF5 and CAR can cooperate in CYP2B6 transcription activation. However, the mechanism of nuclear translocation of CAR in response to inducers is not functional in HepG2 cells, where CAR is already present in the nuclei of noninduced cells (Kawamoto et al., 1999). Therefore, we investigated the effects of ATF5 and CAR on CYP2B6 in cultured human hepatocytes.
First, we analyzed the time course expression of these genes during the initial stages of culture and found that the level of ATF5 and CAR progressively decreased from the early beginning of the culture (i.e., isolation) to up to 10 to 15% of the initial content at 23 h (Fig. 6). This down-regulation, which is associated with the adaptation of hepatocytes to tissue disaggregation and culture conditions, is a common feature of many typical LETFs. CYP2B6 mRNA was also decreased along with culture time (Fig. 6), but the down-regulation of ATF5 preceded the decrease in CYP2B6 mRNA (Fig. 6), which further supports the notion that ATF5 is associated with the differentiated hepatic phenotype and is a key activator of CYP2B6 basal expression in the human liver.
|
|
|
We next analyzed whether the combination of ATF5 and CAR also has a significant role in the induction of CYP2B6 by xenochemicals. Our results, shown in Fig. 8, demonstrate that the response of human hepatocytes to prototypical inducers such as PB or CITCO can be enhanced by the transfection of ATF5 and CAR. However, the relative -fold increase of CYP2B6 mRNA after PB or CITCO when ATF5 and CAR are overexpressed is not larger than in the nontransfected human hepatocytes (Fig. 8). Therefore, we conclude that ATF5 and CAR play a role in sustaining a high basal CYP2B6 expression, which translates in higher induced levels when CAR is further activated by xenochemicals.
|
(Rutkowski and Kaufman, 2003). ATF4 induction leads, in turn, to the regulation of many target genes involved in both amino acid metabolism and transport, and redox chemistry to cope with the ER stress. Translational induction of ATF4 involves the differential contribution of two upstream open reading frames (uORFs) in the 5' leader of the ATF4 mRNA (Vattem and Wek, 2004). We wondered whether ATF5 had also conserved uORFs in the 5' leader of its mRNA. Sequence analysis showed that, similarly to ATF4, ATF5 had two conserved uORFs with very similar coding lengths and number of spacing nucleotides (Fig. 9A), suggesting a potential similar translational mechanism for both ATF factors.
This finding led us to assess the possibility of post-transcriptional induction of ATF5 under stress stimuli. We analyzed ATF5 mRNA and protein levels in HepG2 cells deprived of nutrients by incubation in amino acid-free KRB buffer. Figure 9B shows evidence that, despite no significant ATF5 mRNA changes, there was an important induction of ATF5 protein in human hepatoma cells. Our results also indicate that ATF5 may be involved in ER stress responses in the liver, which, in turn, should involve an induction of CYP2B6, its target gene.
We investigated the effect of several stress stimuli in the CYP2B6 expression level and found that the incubation of HepG2 cells with amino acid–free KRB buffer and tunicamycin caused a time-dependent increase of the CYP2B6 mRNA level (Fig. 9B), whereas other ER stress-inducing compounds, such as glucosamine, triggered a more modest activation (data not shown). Tunicamycin causes ER stress by inhibiting N-glycosylation of newly synthesized proteins. As a positive control, we measured the mRNA level of C/EBP homologous protein (CHOP), a well characterized ER stress response gene. The up-regulation of CHOP was also observed under the different stress conditions but with a different time course profile (Fig. 9B). This result shows that HepG2 cells develop an effective stress response that involves CYP2B6 induction. To extrapolate these results into a more physiological condition, we performed a pilot experiment in cultured human hepatocytes. The results in Fig. 9C show that amino acid deprivation (48 h) can also trigger a stress response involving CYP2B6 induction in human hepatocytes.
|
Discussion |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
). However, most studies on CYP2B6 regulation have focused on the mechanism responsible for the transcriptional induction mediated by xenochemicals, where the role of the nuclear receptor CAR has been very intensely investigated (Wang and Negishi, 2003). However, CAR does not operate alone in the transactivation of P450 genes. Several LETFs, nuclear receptors, and coactivators (e.g., HNF4, pregnane X receptor, retinoid X receptor, glucocorticoid receptor, thyroid hormone receptor, vitamin D receptor, liver X receptor, steroid receptor coactivator 1, and peroxisome proliferative activated receptor, gamma, coactivator 1, alpha) collaborate with CAR and establish an extensive cross-talk, controlling signaling pathways that regulate the homeostasis of bile acids, lipids, hormones, glucose, inflammation, and vitamins, among others (Pascussi et al., 2008). Despite that, the relevance of many of these transcriptional regulators in the control of human CYP2B6 by CAR remains to be investigated, and very few studies have dealt with the tissue-specific regulation of CYP2B6 in the liver.
Previous evidence from our laboratory suggests that the CYP2B6 expression is controlled by HNF4 and C/EBP (Jover et al., 1998, 2001), two important LETFs. In the present study, we have shown for the first time that ATF5, another LETF, is able to activate CYP2B6 in cooperation with CAR. Moreover, we have also shown that ATF5 could play a role in modulating the CYP2B6 expression levels under specific stress conditions, such as amino acid limitation or chemical-induced stress.
The potentiation of CYP2B6 expression by ATF5 and CAR was shown both in the absence and presence of CYP2B6 inducers. CAR is mainly located in the cytoplasm, and chemical inducers facilitate its translocation into the nuclear compartment where CAR shows constitutive activity and binds to responsive elements in DNA (Baes et al., 1994; Honkakoski et al., 1998). In our experimental conditions and in absence of chemical inducers, the transfection of CAR activated CYP2B6 in both HepG2 cells and cultured human hepatocytes. The mechanism of nuclear translocation of CAR in response to inducers is not functional in HepG2 cells, where CAR is already present in the nuclei of noninduced cells (Kawamoto et al., 1999). However, our results also support the notion that in the absence of added inducers a fraction of human CAR traffics and reaches the nucleus in cultured human hepatocytes, even though cytoplasmic retention is operative in this cell model (Pascussi et al., 2000). Thus, our results suggest a basal constitutive activity of human CAR on CYP2B6, which is in agreement with two recent studies in which CAR was also expressed in human hepatocytes by means of viral vectors (Kamiyama et al., 2007; Stoner et al., 2007).
CYP2B6 induction by xenochemicals (PB and CITCO) was also potentiated by the coinfection of human hepatocytes with Ad-ATF5 and Ad-CAR, although the relative -fold increase was not augmented with respect to the -fold increase attained in nontransfected cells. Thus, ATF5 and CAR play a role in sustaining a high basal CYP2B6 expression, which turns out into higher induced levels when CAR is further activated by xenochemicals.
The molecular mechanism by which ATF5 causes an increase in the CYP2B6 mRNA expression remains to be elucidated. There is no previous report showing functional ATF/CREB-like elements in the CYP2B6 regulatory sequences. However, in a recent study on CYP2B6 promoter polymorphism, a putative ATF binding site at about –1.85 kilobase was predicted by in silico analysis (Zukunft et al., 2005), which is in the vicinity of the –1.7 kilobase phenobarbital-responsive enhancer module for CAR. A direct effect of ATF5 on the CYP2B6 gene could also take place through C/EBP response elements. ATF5 has been seen to interact with C/EBPβ (Nishizawa and Nagata, 1992), and ATF4 with C/EBP and C/EBP (Gombart et al., 2007). Moreover, several C/EBP family proteins regulate CYP2B genes (Luc et al., 1996; Jover et al., 1998; Cassel et al., 2000) and play important roles in the differentiation and development of a number of tissues, including the liver and intestine. A hallmark of bZIP proteins is that they extensively heterodimerize with each other both inside and outside individual families; therefore, one manner in which ATF5 could activate CYP2B6 is through the formation of ATF5-C/EBP heterodimers, as well as through the binding to either C/EBP sites or ATF-C/EBP composite sites.
An alternative indirect mechanism for the cooperation between ATF5 and CAR in the transactivation of CYP2B6 could involve an increase in CAR nuclear translocation promoted by ATF5. In this regard, it has been shown that transfection of the coactivator glucocorticoid receptor interacting protein 1 mediates ligand-independent nuclear translocation and activation of CAR (Min et al., 2002). However, three independent experiments using Ad-green fluorescent protein (GFP)-CAR in human hepatocytes have shown that ATF5 does not visibly potentiate CAR translocation to nucleus (data not shown).
In this study, we have found that ATF5 could up-regulate the CYP2B6 expression after particular stress stimuli, such as amino acid limitation and ER stress induced by tunicamycin. This is a new finding that points to a potential role of CYP2B6 in stress conditions that could require detoxification of potentially harmful compounds, such as bile acids or bilirubin, or in chemical-induced liver stress. In this respect, it is important to emphasize that CAR activity has been shown to ameliorate the effects of hyperbilirubinemia, caloric restriction, and toxic bile acids (Goodwin and Moore, 2004). Our results suggest that ATF5 could also participate in these processes through its cooperation with CAR and the transactivation of CYP2B6 to respond to the harmful stress induced by endobiotics/xenobiotics with the induction of detoxification genes.
We have shown that stress conditions cause an up-regulation of ATF5 protein and a concomitant increase in CYP2B6 mRNA. Therefore, it is feasible that the increase in CYP2B6 is mediated, at least partially, by ATF5. However, we cannot rule out that other stress-activated transcription factors of the ATF/CREB family are also involved in this response. One of the most likely candidates is ATF4, which is also translationally activated in stress responses (Rutkowski and Kaufman, 2003). We have measured ATF4 expression and induction by stress stimuli and found that they are operative in our hepatic cell models (data not shown). However, it is important to remark that contrary to ATF5, ATF4 is neither an LETF (Fig. 1) nor a transcriptional regulator associated with the hepatic phenotype (Table 2), which is evidence against a specific role of ATF4 in supporting the transcription of CYP2B6 in hepatocytes. On the other hand, ATF2, ATF3, and ATF6 are also involved in stress responses and can likely bind to the same DNA response elements in target genes (Clerk and Sugden, 1997; Hai and Hartman, 2001). In summary, ATF5 may not be the only factor involved in CYP2B6 activation in stress responses, and a comprehensive study by gain/loss-of-function experiments to ascertain the specific role of the different ATF/CREB candidates will be needed.
Tissue-expression analysis of ATF5 revealed that the liver is the organ with the highest ATF5 mRNA level (10-fold higher than any other tissue), which indicates that ATF5 should be an LETF. However, ATF5 mRNA is also detected in most of the tissues examined, which suggests ATF5 may have both ubiquitous and liver-specific roles. In consonance, ATF/CREB proteins are involved in the control of development and homeostasis in different cell types (Pati et al., 1999; Persengiev and Green, 2003; Al Sarraj et al., 2005; Angelastro et al., 2005). ATF5 shares the ability with other ATF/CREB family members to regulate the transcriptional response to intracellular cAMP through CRE, and a limited number of studies have suggested this possibility. For example, ATF5 has been shown to be able to bind to CRE sites (Peters et al., 2001; Forgacs et al., 2005) and to repress cAMP-induced transcription and CRE-mediated expression (Pati et al., 1999; Angelastro et al., 2003; Forgacs et al., 2005). However, the role of ATF5 in the regulation of cAMP-mediated responses may also be tissue-dependent because ATF5 did not affect CRE-containing reporter genes in human hepatoma cells (Al Sarraj et al., 2005). We have observed a weak transactivation of CRE-containing genes (AldoB and PEPCK) after Ad-ATF5 transfection, suggesting that this transcription factor only plays a minor role in the control of cAMP-responsive genes in the liver.
Our experimental evidence supports the notion that ATF5 is associated with hepatocyte differentiation and/or the maintenance of the adult hepatic phenotype. This observation contrasts with the described function of ATF5 in brain development where it blocks the differentiation of neuroprogenitor cells into neurons and glia and must be down-regulated to allow this process to occur (Angelastro et al., 2005; Mason et al., 2005). The different function of ATF5 in both the liver and nervous system could be explained by its different embryonic origin. Moreover, in Caco-2 cells, a model for enterocyte differentiation, ATF5 also shows a significant up-regulation when cells reach the differentiated condition (Peters et al., 2001); and in ATDC5 cells, a model for chondrocyte differentiation, ATF5 is once more associated with the differentiated state (Shinomura et al., 2006). Further investigation is needed to understand the specific transcriptional environment that determines the role of ATF5 in tissue development and differentiation.
In conclusion, our findings uncover a new abundant LETF that is associated not only with the regulation of drug-metabolizing CYP2B6 in the liver but also with the differentiated adult hepatic phenotype. Moreover, this transcription factor can cooperate with CAR and participate in the adaptation of the liver to harmful stress responses through the up-regulation of detoxifying CYP2B6. A potential contribution of ATF5 to the variability of the CYP2B6 expression in the human population remains to be investigated.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Article, publication date, and citation information can be found at http://dmd.aspetjournals.org.
ABBREVIATIONS: ATF, activating transcription factor; CREB, cAMP response element-binding protein; bZIP, basic region leucine zipper; CRE, cAMP-responsive element; P450, cytochrome P450; CAR, constitutive androstane receptor; PB, phenobarbital; CITCO, 6-(4-chloropheny-l)-imidazo[2,1-b][1,3]thiazole-5-carbaldehyde O-(3,4-dichlorobenzyl)oxime; LETF, liver-enriched transcription factor; ER, endoplasmic reticulum; DMSO, dimethyl sulfoxide; PCR, polymerase chain reaction; HEK, human embryonic kidney; Ad, adenovirus; m.o.i., multiplicity of infection; KRB, Krebs-Ringer bicarbonate; PBS, phosphate-buffered saline; Q-RT-PCR, quantitative real-time polymerase chain reaction; HNF, hepatic nuclear factor; C/EBP, CCAAT/enhancer-binding protein; uORF, upstream open reading frame; CHOP, C/EBP homologous protein; GFP, green fluorescent protein.
Address correspondence to: Ramiro Jover, Unidad Hepatología Experimental Centro Investigación H.U. La Fe, Avenida Campanar 21, E-46009, Valencia, Spain. E-mail: ramiro.jover{at}uv.es
|
References |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
C) leading to enhanced transcription and relocation of the transcriptional start site. Mol Pharmacol 67: 1772–1782.This article has been cited by other articles:
|
|
 |
|
  G. Li, W. Li, J. M. Angelastro, L. A. Greene, and D. X. Liu Identification of a Novel DNA Binding Site and a Transcriptional Target for Activating Transcription Factor 5 in C6 Glioma and MCF-7 Breast Cancer Cells Mol. Cancer Res., June 1, 2009; 7(6): 933 - 943. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V. Besnard, S. E. Wert, M. T. Stahlman, A. D. Postle, Y. Xu, M. Ikegami, and J. A. Whitsett Deletion of Scap in Alveolar Type II Cells Influences Lung Lipid Homeostasis and Identifies a Compensatory Role for Pulmonary Lipofibroblasts J. Biol. Chem., February 6, 2009; 284(6): 4018 - 4030. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
 |
 |
|
 |
 |
 |
36-kDa protein. Fifty micrograms of nuclear protein extracts was loaded per lane. Coomassie blue staining of membranes was performed to check for equal loading (not shown).
