Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

Inhibition of adipogenesis by the stress‐induced protein CHOP (Gadd153).

Inhibition of adipogenesis by the stress‐induced protein CHOP (Gadd153). The EMBO Journal vol.14 no.19 pp.4654-4661, Inhibition of adipogenesis by the stress-induced protein CHOP (Gaddl53) in lead to a decrease in the of the N.Batchvarova, X.-Z.Wang and D.Ron' vitro, expression protein (Ron et al., 1992; Williams et al., 1992). Interfering with of Biomolecular of Medicine Skirball Institute Medicine, Departments the of C/EBPa antisense-RNA inhibits expression by New York and Cell Biology and the Kaplan Cancer Center, University et Lin and adipogenesis (Samuelsson al., 1991; Lane, New NY USA Medical Center, York, 10016, of C/EBPa in 1992) and ectopic expression pre-adipocytic 'Corresponding author fibroblasts converts them to et adipocytes (Freytag al., Lin Detailed of the 1994; and Lane, 1994). analysis cells is on Adipocytic conversion of 3T3-L1 dependent of C/EBP isoforms expression patterns during adipocytic induction of transcription factors from the C/EBP has led to a model the differentiation whereby early family that activate promoters of adipogenic genes. We the C/EBP6 and induces the expression of C/EBPa nuclear that find that expression of CHOP, a protein et Yeh et This model is gene (Cao al., 1991; al., 1995). dimerizes avidly with C/EBP isoforms a and and the existence of functional supported by C/EBP-binding directs the resulting heterodimer away from classic et sites in the promoter of the C/EBPa gene (Christy al., this differentia- C/EBP-binding sites, markedly inhibits the observation that of an 1991), and by over-expression tion process. Surprisingly, the presence of CHOP early form of blocks the inactive, dominant-negative C/EBPP in the differentiation process inhibits C/EBPa and differentia- expression of C/EBPa and inhibits adipocytic gene expression. Ectopic expression of C/EBPa et tion (Yeh al., 1995). the effect of CHOP on differentia- bypasses inhibitory with CHOP is a nuclear protein that dimerizes avidly further evidence that CHOP action is tion, providing C/EBP isoforms a and (Ron and Habener, 1992). mediated by inhibition of C/EBPa gene expression the CHOP Dormant under normal growth conditions, rather than the encoded merely inhibiting protein's is induced to level gene, also known as GADD153, high A similar of attenuated DNA-binding activity. pattern metabolic during the cellular stress caused by toxins, of and J is also observed in cells expression C/EBPa et inhibitors and nutrient deprivation (Fomace al., 1989; induced to differentiate in media with low glucose Price and Luethy and Holbrook, 1992; Calderwood, 1992; concentration. This stressed culture condition is Because of its Ron and Habener, 1992). unusually-struc- associated with induction of CHOP and endogenous bZIP CHOP does not homodimerize and tured domain, marked attenuation of the differentiation Our process. heterodimers bind to a subset CHOP-C/EBP only special data suggest that CHOP functions as an inducible X.-Z. of C/EBP-binding sites (M.Ubeda, H.Zinszner, in to inhibitor of adipocytic differentiation response and in For Wang, I.Wu, J.F.Habener D.Ron, preparation). with the metabolic stress. It does so by interfering some that are many C/EBP target genes, including adipose- accumulation of adipogenic C/EBP isoforms. serves as a specific, CHOP transcriptional inhibitor (Ron Keywords: adipose/C/EBP/CHOP/GADD153/glucose in and Habener, 1992). CHOP plays a special role the of cellular in the in vivo. control growth adipocytic lineage for this comes from the association Evidence very strong Introduction between the of an form of the presence oncogenic protein an in and myxoid liposarcoma, adipose tissue tumor man to mature Differentiation of pre-adipocytic fibroblasts et Rabbitts et Barone (Crozat al., 1993; al., 1993; et al., constitutes a well-established model for the adipocytes 1994; Zinszner et al., 1994). Despite being an inhibitor of tissue in vitro and development adipose (Green Kehinde, of CHOP is induced differentia- C/EBPs, during adipocytic 1974; Ailhaud et al., 1992). This differentiation process tion and Carlson et Recent (Ron Habener, 1992; al., 1993). is associated with the coordinate expression of a group of evidence that this is linked to low suggests glucose related transcription factors, the C/EBP proteins (Christy concentrations and cellular stress under culture conditions et Herrera et al., 1989; al., 1989; Cao et al., 1991). These in vitro (Carlson et al., 1993). We undertook this study to nuclear proteins contain a highly similar DNA-binding further explore the relationship between cellular stress, and dimerization bZIP domain, and are capable of binding CHOP and adipocytic differentiation. and activating the promoters of adipose-specific genes (Christy et al., 1989; Herrera et al., 1989; Kaestner et al., 1990). Results The C/EBPax isoform appears to play a particularly important role in the development and maintenance of Low glucose concentration induces CHOP and adipocytic differentiation. Expression of this isoform cor- inhibits adipocytic differentiation relates temporally with the acquisition of an adipocytic Low glucose concentration in the media has been found phenotype (Birkenmeier et al., 1989) and metabolic pro- to induce the CHOP gene in HeLa and cells 3T3-LI cesses that antagonize adipogenesis, such as the acute- (Carlson et al., 1993). In an attempt to study the process phase response in vivo and tumor-necrosis factor treatment of adipocytic differentiation under conditions in which the 44 Oxford University Press CHOP and adipogenesis A B C D)av: 2 3 4 6 8 o6 1 Lipid (onilent: - . (G t __: FXp C/EBHPI~ I15.4 -. - C H()P (!1i:HI>ti C/1FBl'f1 E =2 I1) I _ _ 14mInM - 3.8 L__Ll_ (Il ()HP ,_,_ ___Z E w C(TB!lo', C. EB1'iR 4.4 -a1mM - r c ti()ll C(; VBl't ,BI ('HO!'>1 rn,\1~~~ -~~~~~ ._ CVEBPII, 10.1 /BION Da':-l Ii1--- 4-I 6 7 C-HOP Fig. 1. Low glucose concentration in the culture media inhibits adipogenesis and induces CHOP Adipocytic differentiation of confluent cultures of 3T3-Ll cells was induced by 2 day exposure to an adipogenic cocktail (Serum, Insulin, IBMX and Dex). Cells cultured at low glucose (2 mM) were compared with those at high glucose concentration (14 mM). (A) Graphic display of experimental design: the glucose concentration is plotted against time. The stippled bar indicates treatment with the adipogenic cocktail. (B) Western blot analysis of whole cell lysates harvested at the indicated time points and probed with antisera specific to the C/EBP isoforms and CHOP. (C) Photograph of plates at day 8 of the differentiation process stained with the lipophilic dye Oil-Red-Oil. Lipid content was estimated by measuring the intensity of the red channel in the scanned photomicrograph. The experimental design, Western blot and stained plates are aligned in the horizontal plane of the figure. CHOP is gene activated, 3T3-Ll cells were subjected to glucose media, the cells committed to differentiation were the standard differentiation procedure while cultured in deprived of glucose at the time of maximal lipid accumula- media that contains a low concentration of glucose (2 mM, tion (days 3-8). Such cells differentiate almost normally see IA Figure for experimental design). When compared and accumulate a significant amount of lipid (Figure 1, with cells undergoing differentiation in media with high experiment 4). Conversely, cells deprived of glucose during glucose concentration (14 mM), the low glucose cells the early phase of differentiation, at a time when lipid accu- to This is were found differentiate much more slowly. mulation is absent, and then switched to high glucose media in reflected a 4-fold reduced accumulation of lipid, accumulate lipid very slowly (Figure 1, experiment 3). We exerts effect on measured by the intensity of staining of the cells with the conclude that low glucose its major adipog- in failure the time of maximal lipophilic dye Oil-Red-Oil (Figure IC) and their enesis, not as a substrate, during lipid rather as an inhibitor of com- to attain a cuboidal morphology, typical of adipocytes (not accumulation (days 3-8), but A 2 mM was chosen in the shown). concentration of glucose mitment to differentiation, early process (days 0-3). a modest decrease in Further evidence that low levels influence the because it is associated with only glucose no in cell internal of the differentiation comes rate of cell growth and change viability (data dynamics process of isoforms. The from of the C/EBP not shown). analysis expression in onset and attenuated as an source of in cul- induction of C/EBPa is delayed Glucose serves important energy cells when its are substrates for the in magnitude in the low glucose compared tured cells and glycolytic products concentration I The of acids. We therefore determined whether with the high glucose (Figure B). synthesis fatty which under our standard culture inhibition of differentiation in low induction of C/EBP,B, the observed adipocyte is sustained off concentration reflects substrate conditions throughout day 8, rapidly tapers glucose deprivation. By in to low concentration cells that had the initial of differ- cells exposed glucose (Figure 1B). switching undergone stages of CHOP reveals that to the in concentration to low Analysis protein exposure entiation high glucose (days 0-3) 4655 N.Batchvarova, X.-Z.Wang and D.Ron the . differentiation in i'~H t procedure high glucose concentration. A .( F with clones transfected Compared with the retro- empty the cells virus, wild-type CHOP-expressing exhibited markedly differentiation delayed (Figure 3A). Deletion of %:. ji. .... ..... :_ _ the basic or I DNA-binding region leucine dimeriza- zipper i, 1. R 6"mummm _0 _ as _ O .. . ~ ... ~ - tion domain abolished the anti-adipogenic effect of CHOP All CHOP derivatives (Figure 3A). were expressed at detectable as levels, though noted previously (Zinszner et the basic al., 1994), mutant accumulates to lower region levels than the other forms Over 50 clones (Figure 3B). 2. IBMX is Fig. required for the early induction of CHOP (day 2) by of cells CHOP-expressing and controls were evaluated in low glucose. 3T3-L1 pre-adipocytes were maintained in the indicated the course of four independent transfection experiments concentration of glucose from -1. Dexamethasone day (<0.25 ,uM, and the inhibitory effect was observed each time. Figure Dex), 3-isobutyl-l-methyl-xanthine both or (0.5 mM, IBMX), neither 3C a were added provides of the clonal variation in to the cultures from day 0-2. Whole cell obtained sample lysates adipogenic on days 1, 2 and 3 were analyzed Western blots differentiation by probed with of wild-type CHOP-expressing cells and, antibodies to CHOP (upper and CREB panel) (lower panel). for leucine minus comparison, zipper CHOP-expressing cells. adipogenic cocktail induces To (on days 0-2) a state of determine whether the inhibitory effect of CHOP increased responsiveness to the effect of 2 mM on glucose. adipogenesis is observable at physiologically-attainable The transient peak in CHOP level of expression noted on day 2 protein, CHOP expression in the transfected in cells differentiating in low is clones glucose highly reproducible was with the level compared of the protein in (see below). A second in CHOP peak expression, and one parental 3T3-L1 cells culture in low glucose or arginine which corresponds to the observed deficient previously induction media (another inducer of CHOP). CHOP levels of CHOP and in the six during adipogenesis (Ron wild-type clones (photographed in Habener, 1992; Figure 3C) Carlson et al., around 6. were no 1993), begins day The greater than those attained in the intensity parental cells of this later peak correlates with the positively to exposed to low glucose or arginine depleted degree media (Figure which cells have undergone commitment to differentiation 3D). In this small there is sample no correlation between (Figure 1, compare 1 the experiments and 3) and is level of CHOP and the strongly degree to which adipogenesis accentuated low by glucose concentration at these later is inhibited. as most However, clones expressed significant time points (Figure 1, compare experiments 1 and 3 with levels of the protein and most exhibited marked attenuation 2 and 4). Regardless of where in the differentiation of process adipocytic differentiation, it is likely that beyond a it occurs, the induction of CHOP is associated with an certain threshold level of CHOP other elements that vary attenuation in the level of and C/EBPa and correlates stochastically in the 3T3-LI population account for the with an inhibition in the process of differentiation. clonal variation in adipocyitc differentiation. We conclude The transient peak in CHOP observed in low glucose on that a physiologically-attainable level of CHOP is capable day 2 of the differentiation protocol suggests a of inhibiting adipogenesis in 3T3-LI synergism cells. between low and the glucose adipogenic cocktail. The effects of individual of the cocktail components on CHOP CHOP interferes with the normal differentiation- inducibility by low glucose were therefore In studied. the associated induction of C/EBPa and /3 absence of any of the 2 mM components cocktail, CHOP inhibits C/EBPa and , glucose DNA-binding activity to a did not induce CHOP to a detectable degree (Figure subclass of C/EBP target genes 2A). (Ron and Habener, 1992; in However, the presence of added 3-isobutyl-1-methyl- M.Ubeda et al., in preparation). Therefore, the observed xanthine (IBMX, a 3 ',5' nucleotide phosphodiesterase effect of CHOP on adipogenesis could be due to simple low inhibitor) glucose led to the brisk induction of CHOP stochiometric interaction with these C/EBP isoforms and (Figure 2B), comparable in level with that observed with inhibition of their binding to adipogenic target genes. the full differentiation cocktail (Figure 2D). Dexametha- However, analysis of cells in differentiating low glucose sone, the other essential ingredient in the differentiation concentrations suggested that inhibition of differentiation cocktail, was without effect (Figure 2C) and IMBX did and induction of CHOP are also associated with decreased not induce CHOP in the presence of normal expression glucose of C/EBPa and f8 (Figure iB). To determine concentration (Figure 2E). The effect of IBMX on CHOP whether CHOP plays an additional dynamic role in responsiveness to low glucose exhibits a latency of at influencing the level of C/EBP proteins, the expression of least 24 h. This indicates that it is due to an altered C/EBP isoforms was evaluated during exposure of CHOP- cellular phenotype rather than a direct interaction between expressing cells to the differentiation procedure at high the cAMP and pathway the pathway glucose inducing CHOP. concentration. Western blot analysis demonstrates that CHOP- Ectopic expression of CHOP attenuates adipocytic expressing cells exposed to the differentiation cocktail differentiation 1 of 3T3-L cells failed to induce C/EBPa. levels, though C/EBPP The above results led to an examination of the role CHOP transiently induced at the time of exposure to the adipo- may play in inhibiting the process of adipogenesis. 3T3- genic cocktail, were not sustained (Figure 4A). The normal LI cells were transfected with recombinant retrovirus that transient induction of C/EBP6 was unaffected by the expresses wild-type CHOP and the neo gene (as a dominant ectopic expression of CHOP. These results suggest that, selectable marker). Transfected cells were selected by in the dynamic setting of adipogenesis, CHOP is also culture in G418 and clones were expanded and subjected to implicated in inhibiting differentiation by a novel mechan- 4656 CHOP and adipogenesis z- - K =- me ,. I.W -, (-. >A7 II Pa a.. e.t t I Pa1I~rentalIL 1 Emptz- '1' ('110P~ NI' N,. ( CIlIOP-o - 'I '\ /~~~~- ('HOP LZ- CHOP BR- C1 3 4 6 ( "I' }IOP .l ,'r IS 1. ,i I, / .I ( IO0P 11- V. 2.5 .: EC ,1 (hIOP C'REB n4J 30one Fig. 3. CHOP inhibits adipocytic differentiation. (A) Clones of 3T3-Ll cells derived by transfection with retrovirus encoding wild-type CHOP (WT), CHOP deleted at the leucine zipper dimerization domain CHOP bearing an internal deletion of the DNA-binding basic region (BR-), cells (LZ-), infected with an retrovirus (Empty T) and the parental line (Parental Ll) were compared for their ability to differentiate to adipocytes. empty of 60 mm tissue culture plate stained with Oil-Red-Oil on day 8 of the differentiation process are shown. (B) Western blot analysis of Photographs cells infected with the retroviral vectors. As noted previously, CHOP LZ- and CHOP BR- migrate anomalously (Ron and CHOP-expressing 1992; M.Ubeda et al., in preparation). (C) Six representative clones of CHOP WT and CHOP LZ- 3T3-L1 cells were stained with Oil-Red- Habener, Oil on 8 of differentiation. The variation in intensity of the stain provides an estimate of the clonal variation in differentiation in both types of day cells. Extracts from 3T3-L1 cells undifferentiated sister plates of the CHOP WT expressing clones shown in (C) (lanes 1- (D) parental (parental Ll), cells cultured in media or low were evaluated by Western blot with antibodies to CHOP (upper panel) 6), parental arginine-free (Arginine-) glucose and CREB internal lower (the control, panel). 4657 N.Batchvarova, X.-Z.Wang and D.Ron ...g .} ie X ;i.1 To evaluate further the mechanism of CHOP-induced inhibition of C/EBPa and protein levels, we performed Northern blot analysis of RNA from cells ectopically ;_ 4 ..-, :: -z expressing CHOP. The absence of detectable C/EBPa and - -_ !: 1- the non-sustained induction of protein are due to ...0- C/EBPP P. the attenuated expression of their loom" Iim respective genes (Figure ._.. w *mm - '-------I-- 'NOWAO-- 4C). Thus we conclude that of CHOP presence during the differentiation process attenuates the proper coordinate expression of C/EBP isoforms a and The somewhat P. delayed peak in C/EBPa and mRNA accumulation (day P3 6) compared with the level protein (day 4) reflects the fact that the two experiments were not performed in parallel and demonstrates the inter-experimental variation ;... 1W s:.'..- in rates of differentiation. Irn ... . ... Ectopic expression of C/EBPa induces adipocytic differentiation of CHOP-containing 3T3-11 cells C/EBPa to lesser (and a degree plays a pivotal C/EBPP) role in adipocytic differentiation of 3T3-L1 cells. Inter- ference with the normal induction of C/EBP isoforms by CHOP could explain its inhibitory effect on the differentiation process. If this is the case then ectopic of 'rescue' expression C/EBPa may the CHOP-induced defect in 3T3-L I cells adipogenesis. ectopically expressing CHOP were transfected with a second recombinant retro- # # ~~~~~~~~~~~~~~~~~~~~~~~~~I _~~~~~~~~~~~~~~~~~~~~~~~~~~~~ virus that encodes 40- C/EBPa and contains the hisD gene as .,. off a dominant selectable f.0 marker. Following a 2 week selection itof with histidinol to enrich for the C/EBPa recombinant cells, foci of lipid containing cells were observed (Figure Cells with 5A, upper panels). transfected the empty inhibits the normal differentiation-associated induction Fig. 4. CHOP retroviral vector and likewise of and Western blot of extracts from CHOP selected showed no adipo- C/EBPa (A) analysis P. LZ- and CHOP WT cells at the indicated time of the points differentiation cytic (Figure 5A, upper panels). Appro- differentiation with antisera to and process, probed CIEBPa, 6. P of in the of transfected priate expression C/EBPa pool cells Estimate of the of nuclear extracts (B) C/EBP DNA-binding activity was revealed Western as is the by blotting, ectopically- prepared from CHOP LZ- and CHOP cells (lanes 1-4) WT-expressing CHOP expressed (Figure 5B). (lanes 5-8), before 1 and and after 2-4 and (lanes 5) (lanes 6-8) induction with the adipogenic cocktail. An of a The most of this result autoradiogram gel-shift straightforward interpretation is assay using the APRE site as a C/EBP-binding angiotensinogen gene that CHOP mediates its inhibitory influence on adipo- probe is shown. CHOP Bacterially-expressed wild-type (bCHOP WT, via its effects genesis on C/EBPa. However, an effect of lanes 4 and 8) or CHOP deleted at the leucine zipper (bCHOP LZ-, CHOP on some other of aspect adipocytic differentiation lanes 3 and 7) were added to the nuclear extract the in vitro during that is unrelated to binding reaction. (C) Northern blot of C/EBP isoforms from C/EBP isoform expression is not analysis control (right panel) and CHOP WT (left panel) cells undergoing excluded this 'rescue' by genetic experiment-C/EBPa adipocytic differentiation. The blot was with labeled mouse re-probed may serve as a dose simply 'high suppresser' of such tubulin cDNA to serve as an internal control. The 2 from day sample CHOP We also noted that the activity. degree of adipocytic the control cells is accidentally underloaded. differentiation in the cells expressing ectopic C/EBPa was augmented by treating the cells with adipogenic cocktail ism-blocking the accumulation of C/EBPa and 5 (not shown). This indicates that the ectopic C/EBPa is proteins. also capable of synergizing with endogenous adipogenic Following exposure to the adipogenic cocktail, CHOP- factors in the CHOP-expressing cells. expressing cells do manifest a increase in the significant C/EBP DNA-binding activity. This is demonstrated by the ability of nuclear extracts to shift the mobility of a 'classic' Discussion C/EBP-binding site probe (Figure 4B, lanes 5 and 6). Had The role its levels been obligatory played by C/EBPa and , in adipocytic high enough, CHOP would have been differentiation is capable of inhibiting all such C/EBP probably due to their activity as transcrip- DNA-binding activity, tion factors as adding excess bacterially-expressed CHOP to the gel- regulating the expression of downstream shift assay inhibits all adipogenic genes. DNA-binding activity (lanes 4 and CHOP-C/EBP heterodimers have a very 8). Bacterially-expressed CHOP that is lacking the leucine different DNA-binding activity than dimers of C/EBPs zipper dimerization domain is without effect, indicating (Ron and Habener, 1992; M.Ubeda et al., in preparation) that heterodimerization is the basis for the in vitro inhibi- and for a subset of 'classic' C/EBP target genes CHOP tion of DNA-binding activity in this assay (lanes 3 and acts as a C/EBP inhibitor. Inhibition of adipogenesis 7). We conclude that inhibition of DNA-binding activity by CHOP is therefore a predictable consequence of its by CHOP is incomplete and suggest that blocking expres- biochemical properties. the However, mechanism of sion of C/EBP's may be the predominant mode of CHOP's inhibition revealed in our study is more intricate than effect on adipogenesis. expected. Instead of predominantly stochiometric inhibi- 4658 CHOP and adipogenesis levels of C/EBP DNA-binding activity are still quite low. Such a scenario could explain how a limited amount of early enough in the differentiation CHOP, when present process, antagonizes what would otherwise be a large C H(1-P accumulation of C/EBPa and Furthermore, this model P3. would explain how ectopic expression of C/EBPa might, EKmpt% 1' by simple titration, overcome the inhibitory effect of CHOP. In this model, DNA binding by CHOP is not required for its inhibitory effect on adipogensis, the protein simply needs to form a non-productive heterodimer with relevant C/EBP isoforms early in the differentiation process. C'/EBR1) 1.iP aw- p ," to What then is the role of the CHOP basic region in > JX n w aS I attenuating adipogenesis? In vitro, the basic region is for ability of the protein to inhibit binding dispensable the \50, X44111 their target sequence (M.Ubeda et al., of C/EBPa and to in preparation). In vivo however, the basic region may heterodimers through non- stabilize such non-productive interactions. The observation that the specific chromatin BR- is at lower levels than the CHOP protein expressed wild-type protein might also provide part of the explanation et al., 1994). It remains possible (Figure 3B; Zinszner .Z Z = . however, that part of the effect of CHOP on adipogenesis sequence-specific interaction with is mediated through I.. ,; DNA-binding by CHOP target genes. Sequence-specific a partner and is com- requires dimerization with C/EBP pletely dependent on the integrity of the basic region (M.Ubeda et al., in preparation). Additional indirect sup- of port for the existence of a target gene-mediated effects CHOP is provided by its oncogenic variant, TLS-CHOP. ( ':'I.B In our assay the inhibitory effect of CHOP on adipogenesis is reproducibly more pronounced than that of TLS-CHOP (data not shown), yet both proteins dimerize equally well with C/EBPs (Crozat et al., 1993; Zinszner et al., 1994). C( O-lP m It is conceivable therefore that the two proteins, by virtue of differences in their transcriptional activation domains, Fig. 5. Ectopic expression of C/EBPa promotes adipocytic exert different influences on such target genes. differentiation of CHOP-expressing 3T3-LI cells. (A) CHOP- The finding that ectopic expression of C/EBPa 'rescues' expressing cells were transfected with recombinant retroviruses the adipogenic phenotype of CHOP-expressing cells is encoding C/EBPa or the empty virus (Empty '). Two weeks later the consistent with a model whereby attenuation of differentia- were stained with Oil-Red-Oil and photographed. (B) Western plates blot of extracts from sister plates of those shown in (A) were probed tion is mediated by interference with a C/EBP signaling with antiserum to C/EBPa (upper panel) or CHOP (lower panel). The specifically one mediated by C/EBPa. It is pathway, asterisk in the upper panel indicates a non-specific reactivity of the that both ectopic expression of CHOP and noteworthy antiserum. C/EBP CHOP induction by low glucose levels are associated with marked decrease in C/EBPa level. This is associated with tion of C/EBP activity by hetrodimerization low levels of C/EBPa mRNA and raises the possibility DNA-binding of with CHOP we find a striking that CHOP is a direct negative regulator C/EBPa gene ectopically-expressed mode of attenuation in the C/EBPa (and to a lesser degree C/EBP,) A parallel, phenotypically-dominant expression. however not ruled out. CHOP could action for CHOP, is gene expression. Recent studies that induction of C/EBP isoforms inhibit adipogenesis by activating anti-adipogenic target suggest of C/EBPa a differentiation is autocatalytic: omitting with attenuated expression being during adipocytic genes, rather than its cause. In a inducer of the C/EBP8 marker of failed differentiation dexamethasone, potent early could rescue the from the cocktail leads to diminished the latter case C/EBPa phenotype by isoform, adipogenic with CHOP and the of and a isoforms and results in dimerizing altering target gene accumulation C/EBPP of the CHOP hetero- or activation attenuated differentiation (Yeh et al., 1995). The specificity properties markedly the CHOP effect sites and is could also overcome contains dimer. C/EBPa C/EBPa promoter C/EBP-binding as a et isoforms 1991; by activating 'suppressor' adipogenic target genes activated by C/EBP (Christy al., for CHOP action are consistent of homodimer. Both models et exogenous Legraverend al., 1993); ectopic expression for CHOP to retain its in cells with the established requirement induces the 3T3-L1 C/EBPa endogenous gene dimerization domain and This that CHOP block leucine DNA-binding and suggests may zipper (Lin Lane, 1994). in basic region order to inhibit adipogenesis. and the activation of the C/EBPa insinuating genes by the between the of dimerization in What be significance parallels itself as a partner early might non-productive and culture in of CHOP at a time when ambient inhibition adipogenesis by by the C/EBP autoregulatory cascade, 4659 N.Batchvarova, X.-Z.Wang and D.Ron low Our results indicate glucose? that failure to attain an between the of an altered form of presence CHOP, TLS- when in adipocytic cultured low is not CHOP and an tissue phenotype glucose myxoid liposarcoma, adipose tumor due to substrate of cells that simply deprivation are (Aman et al., 1992) may indicate that this or similar otherwise committed to the normally differentiation pathways play a role in normal metabolism of pro- adipose cess. these untoward culture Rather, conditions affect tissue and derangement therein may participate in the the intrinsic differentiation process-there is development of markedly neoplasia. accumulation of and delayed sustained C/EBPa poorly accumulation of in low media. C/EBP, glucose The Materials and methods same alterations are observed in cells CHOP expressing low ectopically. seems to exert a Furthermore, glucose Cell culture and induction of differentiation more effect on profound differentiation when adipocytic 3T3-L1 pre-adipocytes were maintained in Dulbecco's modified Eagle's in the medium (DMEM) with 10% calf serum. cells of present early differentiation And it is Only early passages were process. during utilized for transfection and differentiation experiments. Differentiation this window of time that (around day 2) differentiating induction was performed as described previously (Student et al., 1980). cells cultured in low activate CHOP. This glucose temporal Briefly, 2-day post confluent cultures were placed in DMEM at the correlation that CHOP be a role suggests may playing in indicated concentration of glucose (14 or 2 mM) supplemented with some of the effects of low 10% fetal bovine serum (FBS, Intergen), mediating inhibitory glucose 10 of insulin gg/ml (Novo Nordisk), 0.25 dexamethasone on gM (Sigma) and 0.5 mM differentiation. 3-isobutyl-l- adipocytic methyl-xanthine (Sigma). The start point of differentiation is referred to As noted the differentiation itself influ- above, process as day 0. Forty-eight hours later (day 2) 3-isobutyl-l-methyl-xanthine ences the of the CHOP to low responsiveness gene glucose and dexamethasone were removed from the culture medium and the In levels. the absence of low does not IBMX, glucose cells were placed in fresh medium containing only insulin and 10% FBS. Media was replaced induce CHOP to a measurable In its every 2 days thereafter. The accumulation of degree (Figure 2). cytoplasmic triglyceride droplets was visualized by staining with Oil- CHOP is induced low but presence by glucose as then, Red-Oil (Aldrich Chemical Company, Inc.) as previously described the hormonal cocktail is cells withdrawn, down-regulate (Preece, 1989). CHOP expression (Figure 1, experiment 2). However, when cells differentiation at a undergoing high glucose Retroviral expression vectors and isolation of stable cell concentration are switched to a low lines glucose level, profound Replication defective ecotrophic retroviruses CHOP and encoding its induction of the CHOP is observed gene (Figure 1, derivatives were constructed introducing by the respective cDNAs into We note that both the IBMX effect on experiment 4). day the pSRa proviral plasmid and co-transfected with a plasmid encoding and the second in CHOP to low peak responsiveness helper functions into COS1 cells to the of leading production viral in committed cells are associated with glucose (on day 6) particles in the culture et The supematant (Muller al., 1991). full-length CHOP cDNA was excised from the CHOP-10 levels of et Yeh et 1995 pBS plasmid (Ron and high (Cao al., 1991; al., C/EBPP Habener, 1992) and ligated into To create pSRa. the leucine zipper and The CHOP is to this Figure 1). promoter responsive deleted form of CHOP (CHOP a BamHI-NheI LZ-), fragment of CHOP factor et the transcription (Sylvester al., 1994), suggesting was ligated into pSRa. This encodes a in which the protein 34 C- that a role in the of possibility plays inducibility terminal amino acids of CHOP are vector-encoded replaced by C/EBPP sequence. CHOP low It is to out that To create the basic mutant form of CHOP by a region (CHOP BR-), the glucose. important point corresponding cDNA was excised from the CHOP-Cb-NLSpcDNA late in CHOP occurs in even smaller, peak expression (Barone et al., 1994) and into The retroviral ligated pSRa. vector committed cells at concentra- differentiating high glucose encoding C/EBPca was constructed by introducing the rat cDNA into tion and Ron and (Figure 1 iB, experiment Habener, the proviral plasmid LXSHD. This virus the expresses gene of interest Carlson et Whether this is due to a 1992; al., 1993). from the retroviral LTR and the hisD a dominant selectable gene, marker gene, is driven by an internal SV40 similar mechanism as that associated with low early promoter (Hartman and glucose is Mulligan, 1988). Viral particles were in cells produced COS1 as not known. of the culture However, frequent replacement described above. Retroviral transfection of 3T3-LI cells was performed media can attenuate this later that peak, suggesting deple- as described previously (Zinszner et al., that 1994), except transfected tion of another essential nutrient a role in CHOP may play cells were selected with 450 tg/ml G418 and 0.2 (Gibco BRL) mM induction. L-histidinol (Sigma) in histidine-free DMEM-10% calf serum. One of the well-characterized effects of low is glucose Western blot analysis to induce that a role as in the proteins play chaperones Whole cell extracts from 3T3-LI cells were prepared in Laemmli buffer, reticulum endoplasmic et This (ER) (Poussegur al., 1977). electrophoresed on a 12% SDS-polyacrylamide gel (or 10% for the effect has been ascribed to interference with normal C/EBPa transferred to a immunoblot), nitrocellulose filter (Micron of secreted glycosylation proteins rather than depletion of Separations and then reacted with the Inc.) indicated antisera. The immunoreactive stores protein species were visualized by the enhanced As chemi- energy (Lee, 1987). previously pointed out, luminescence detection system (ECL, DuPont-NEN). To equilibrate of the stressful many events that induce CHOP also induce sample loading, parallel-run protein gel was stained with Coomassie accumulation of defective proteins in the ER (Carlson blue. The rabbit antisera to C/EBPa and and CREB have been C/EBPP et The al., 1993). trigger for CHOP induction by low described et previously (Ron al., 1992). The polyclonal antisera against glucose be stress in the ER C/EBP8 were from may induced by such malfolded purchased Santa Cruz Biotechnology, Inc. The monoclonal to antibody CHOP, 37-9C8 recognizes an epitope in the N- proteins. Adipocytic differentiation is associated with a terminus of the protein and crude hybridoma supernatant was used at a induction significant of cell secretory activity leading, for dilution of 1:10. to exuberant example, production of matrix glycoproteins et (Kuri-Harcuch al., 1994). The increase in secretory Northern blot analysis is to be activity likely associated with Total RNA was isolated from an increased 3T3-LI cells at various time points during the differentiation process by the for stress in the phenol/guandinium isothiocyanate potential ER. We speculate that one method of extraction (RNAzol B, Biotex Laboratories, Inc.). The RNA important function of CHOP may be to modulate the were samples electrophoresed through formaldehyde-containing agarose normal differentiation so as not to incur a critical process blotted to membranes gels, and nylon (Amersham) hybridized to level of such stress. The association and cDNA as C/EBPa described strong 0 probes et epidemiological previously (Ron al., 1992). 4660 CHOP and adipogenesis To control for RNA loading and transfer efficiency the blot was Sylvester,S.L., ap Rhys,C.M.J., Luethy,J.D. and Holbrook,N.J. (1994) J. with the cDNA for of Dr hybridized ox-tubulin (a gift Nicholas Cowan, Biol. Chem., 269, 20119-20125. New York University, New York). Williams,P.M., Danesch,U., Ringold,G. and Heller,R. (1992) Chang,D.J., Mol. Endocrinol., 6, 1135-1141. DNA-binding assay Yeh,W.-C., Cao,Z., Classon,M. and McKnight,S. (1995) Genes Dev., 9, EMSA experiments were performed exactly as previously described 168-181. using the angiotensinogen APRE sequence as a probe (Ron and Habener, Zinszner,H., Albalat,R. and Ron,D. (1994) Genes Dev., 8, 2513-2526. 1992). Where indicated, 100 ng of purified bacterially-expressed gluta- thione S-transferase-CHOP fusion protein was added to the DNA- Received on June 13, 1995 reaction to the addition of the radiolabeled binding prior probe. Acknowledgements We are indebted to Niki Holbrook and Lennart Philipson for useful discussions and to Ueli Schibler for emphasizing the importance of stochiometric considerations in the action of inhibitors. This work was supported by a PHS research award (DK47119) and by generous support from the Arthritis Foundation. D.R. is a Pew Scholar in Biomedical Sciences. References Ailhaud,G., Grimaldi,P. and Negrel,R. (1992) Annu. Rev. Nutr., 12, 207-233. Aman,P., Ron,D., Mandahl,N., Fioretos,T., Heim,S., Arhenden,K., and Willen,H., Rydholm,A. Mitelman,F. (1992) Genes Chrom. Cancer, 5, 271-277. Barone,M.V., Crozat,A.Y., Tabaee,A., Philipson,L. and Ron,D. (1994) Genes Dev., 8, 453-464. Birkenmeier,E.H., Gwynn,B., Howard,S., Jerry,J., Gordon,J.I., Land- and schultz,W.H. McKnight,S.L. (1989) Genes Dev., 3, 1146-1156. and Cao,Z., Umek,R.M. McKnight,S.L. (1991) Genes Dev., 5, 1538- and Carlson,S.G., Fawcett,T.W., Bartlett,J.D., Bernier,M. Holbrook,N.J. (1993) Mol. Cell. Biol., 13, 4736-4744. Christy,R.J., Yang,V.W., Ntambi,J.M., Geiman,D.E., Landschulz,W.H., Friedman,A.D., Nakabeppu,Y., Kelly,T.J. and Lane,M.D. (1989) Genes Dev., 3, 1323-1335. and Proc. Christy,R.J., Kaestner,K.H., Geiman,D.E. Lane,M.D. (1991) Natl Acad. Sci. USA, 88, 2593-2597. and Crozat,AY., Aman,P., Mandahl,N. Ron,D. (1993) Nature, 363, 640-644. Fornace,A.J., Neibert,D.W., Hollander,M.C., Luethy,J.D., Papathan- asiou,M., Fragoli,J. and Holbrook,N.J. (1989) Mol. Cell. Biol., 9, 4196-4203. and Genes 1654- Freytag,S.O., Paielli,D.L. Gilbert,J.D. (1994) Dev., 8, and 113-116. Green,H. Kehinde,O. (1974) Cell, 1, Hartman,S. and Mulligan,R. (1988) Proc. Natl Acad. Sci. USA, 85, 8047-8051. and Herrera,R., Ro,H.S., Robinson,G.S., Xanthopoulos,K.G. Spiegel- Mol. Cell. 5331-5339. man,B.M. (1989) Biol., 9, and Proc. Natl Acad. Sci. Kaestner,K.H., Christy,R.J. Lane,M.D. (1990) 251-255. USA, 87, and Kuri-Harcuch,W., ArgUello,C. Marsch-Moreno,M. (1994) Differen- 173-178. tiation, 28, Trends Biochem. 20-23. Lee,A.S. (1987) Sci., 12, and Legraverend,C., Antonson,P., Flodby,P. Xanthopoulos,K. (1993) Nucleic Acids 1735-1742. Res., 21, and Genes 533-544. Lin,F.-T. Lane,M.D. (1992) Dev., 6, and Proc. Natl Acad. Sci. 8757-8761. Lin,F.-T. Lane,M. (1994) USA, 91, and Cancer 5-10. Luethy,J.D. Holbrook,N.J. (1992) Res., 52, Muller,A.J., Young,J.C., Pendergast,A.-M., Pondel,M., Landau,N.R., 1785-1792. and Mol. Cell. Littman,D.R. Witte,O.N. (1991) Biol., 11, 941-947. and Pastan,I. Poussegur,J., Shiu,R.P.C. (1977) Cell, 11, Little Brown and Preece Manual Technicians. (1989) for Histological Co., Boston, pp. 3814-3817. and Cancer Price,B. Calderwood,S. (1992) Res., 52, and Nature Genet., Rabbitts,T.H., Forster,A., Larson,R. Nathan,P. (1993) 175-180. 4, Genes and Habener,J.F. Dev., 6, 439453. Ron,D. (1992) and J. Clin. Habener,J.F. (1992) Ron,D., Brasier,A.R., McGehee,R.E.,Jr 223-233. Invest., 89, and Enerb6ck,S. Samuelsson,L., Stromberg,K., Vikman,K., Bjursell,G. EMBO 3787-3793. (1991) J., 10, and J. Biol. 4745-4750. Chem., 255, Student,A., Hsu,R. Lane,M. (1980) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png The EMBO Journal Springer Journals

Inhibition of adipogenesis by the stress‐induced protein CHOP (Gadd153).

Batchvarova, N.; Wang, X. Z.; Ron, D.
The EMBO Journal , Volume 14 (19) – Oct 1, 1995
Download PDF
8 pages

Loading next page...
 
/lp/springer-journals/inhibition-of-adipogenesis-by-the-stress-induced-protein-chop-gadd153-0xUxwiGPVU

References (6)

Publisher
Springer Journals
Copyright
Copyright © European Molecular Biology Organization 1995
ISSN
0261-4189
eISSN
1460-2075
DOI
10.1002/j.1460-2075.1995.tb00147.x
Publisher site
See Article on Publisher Site

Abstract

The EMBO Journal vol.14 no.19 pp.4654-4661, Inhibition of adipogenesis by the stress-induced protein CHOP (Gaddl53) in lead to a decrease in the of the N.Batchvarova, X.-Z.Wang and D.Ron' vitro, expression protein (Ron et al., 1992; Williams et al., 1992). Interfering with of Biomolecular of Medicine Skirball Institute Medicine, Departments the of C/EBPa antisense-RNA inhibits expression by New York and Cell Biology and the Kaplan Cancer Center, University et Lin and adipogenesis (Samuelsson al., 1991; Lane, New NY USA Medical Center, York, 10016, of C/EBPa in 1992) and ectopic expression pre-adipocytic 'Corresponding author fibroblasts converts them to et adipocytes (Freytag al., Lin Detailed of the 1994; and Lane, 1994). analysis cells is on Adipocytic conversion of 3T3-L1 dependent of C/EBP isoforms expression patterns during adipocytic induction of transcription factors from the C/EBP has led to a model the differentiation whereby early family that activate promoters of adipogenic genes. We the C/EBP6 and induces the expression of C/EBPa nuclear that find that expression of CHOP, a protein et Yeh et This model is gene (Cao al., 1991; al., 1995). dimerizes avidly with C/EBP isoforms a and and the existence of functional supported by C/EBP-binding directs the resulting heterodimer away from classic et sites in the promoter of the C/EBPa gene (Christy al., this differentia- C/EBP-binding sites, markedly inhibits the observation that of an 1991), and by over-expression tion process. Surprisingly, the presence of CHOP early form of blocks the inactive, dominant-negative C/EBPP in the differentiation process inhibits C/EBPa and differentia- expression of C/EBPa and inhibits adipocytic gene expression. Ectopic expression of C/EBPa et tion (Yeh al., 1995). the effect of CHOP on differentia- bypasses inhibitory with CHOP is a nuclear protein that dimerizes avidly further evidence that CHOP action is tion, providing C/EBP isoforms a and (Ron and Habener, 1992). mediated by inhibition of C/EBPa gene expression the CHOP Dormant under normal growth conditions, rather than the encoded merely inhibiting protein's is induced to level gene, also known as GADD153, high A similar of attenuated DNA-binding activity. pattern metabolic during the cellular stress caused by toxins, of and J is also observed in cells expression C/EBPa et inhibitors and nutrient deprivation (Fomace al., 1989; induced to differentiate in media with low glucose Price and Luethy and Holbrook, 1992; Calderwood, 1992; concentration. This stressed culture condition is Because of its Ron and Habener, 1992). unusually-struc- associated with induction of CHOP and endogenous bZIP CHOP does not homodimerize and tured domain, marked attenuation of the differentiation Our process. heterodimers bind to a subset CHOP-C/EBP only special data suggest that CHOP functions as an inducible X.-Z. of C/EBP-binding sites (M.Ubeda, H.Zinszner, in to inhibitor of adipocytic differentiation response and in For Wang, I.Wu, J.F.Habener D.Ron, preparation). with the metabolic stress. It does so by interfering some that are many C/EBP target genes, including adipose- accumulation of adipogenic C/EBP isoforms. serves as a specific, CHOP transcriptional inhibitor (Ron Keywords: adipose/C/EBP/CHOP/GADD153/glucose in and Habener, 1992). CHOP plays a special role the of cellular in the in vivo. control growth adipocytic lineage for this comes from the association Evidence very strong Introduction between the of an form of the presence oncogenic protein an in and myxoid liposarcoma, adipose tissue tumor man to mature Differentiation of pre-adipocytic fibroblasts et Rabbitts et Barone (Crozat al., 1993; al., 1993; et al., constitutes a well-established model for the adipocytes 1994; Zinszner et al., 1994). Despite being an inhibitor of tissue in vitro and development adipose (Green Kehinde, of CHOP is induced differentia- C/EBPs, during adipocytic 1974; Ailhaud et al., 1992). This differentiation process tion and Carlson et Recent (Ron Habener, 1992; al., 1993). is associated with the coordinate expression of a group of evidence that this is linked to low suggests glucose related transcription factors, the C/EBP proteins (Christy concentrations and cellular stress under culture conditions et Herrera et al., 1989; al., 1989; Cao et al., 1991). These in vitro (Carlson et al., 1993). We undertook this study to nuclear proteins contain a highly similar DNA-binding further explore the relationship between cellular stress, and dimerization bZIP domain, and are capable of binding CHOP and adipocytic differentiation. and activating the promoters of adipose-specific genes (Christy et al., 1989; Herrera et al., 1989; Kaestner et al., 1990). Results The C/EBPax isoform appears to play a particularly important role in the development and maintenance of Low glucose concentration induces CHOP and adipocytic differentiation. Expression of this isoform cor- inhibits adipocytic differentiation relates temporally with the acquisition of an adipocytic Low glucose concentration in the media has been found phenotype (Birkenmeier et al., 1989) and metabolic pro- to induce the CHOP gene in HeLa and cells 3T3-LI cesses that antagonize adipogenesis, such as the acute- (Carlson et al., 1993). In an attempt to study the process phase response in vivo and tumor-necrosis factor treatment of adipocytic differentiation under conditions in which the 44 Oxford University Press CHOP and adipogenesis A B C D)av: 2 3 4 6 8 o6 1 Lipid (onilent: - . (G t __: FXp C/EBHPI~ I15.4 -. - C H()P (!1i:HI>ti C/1FBl'f1 E =2 I1) I _ _ 14mInM - 3.8 L__Ll_ (Il ()HP ,_,_ ___Z E w C(TB!lo', C. EB1'iR 4.4 -a1mM - r c ti()ll C(; VBl't ,BI ('HO!'>1 rn,\1~~~ -~~~~~ ._ CVEBPII, 10.1 /BION Da':-l Ii1--- 4-I 6 7 C-HOP Fig. 1. Low glucose concentration in the culture media inhibits adipogenesis and induces CHOP Adipocytic differentiation of confluent cultures of 3T3-Ll cells was induced by 2 day exposure to an adipogenic cocktail (Serum, Insulin, IBMX and Dex). Cells cultured at low glucose (2 mM) were compared with those at high glucose concentration (14 mM). (A) Graphic display of experimental design: the glucose concentration is plotted against time. The stippled bar indicates treatment with the adipogenic cocktail. (B) Western blot analysis of whole cell lysates harvested at the indicated time points and probed with antisera specific to the C/EBP isoforms and CHOP. (C) Photograph of plates at day 8 of the differentiation process stained with the lipophilic dye Oil-Red-Oil. Lipid content was estimated by measuring the intensity of the red channel in the scanned photomicrograph. The experimental design, Western blot and stained plates are aligned in the horizontal plane of the figure. CHOP is gene activated, 3T3-Ll cells were subjected to glucose media, the cells committed to differentiation were the standard differentiation procedure while cultured in deprived of glucose at the time of maximal lipid accumula- media that contains a low concentration of glucose (2 mM, tion (days 3-8). Such cells differentiate almost normally see IA Figure for experimental design). When compared and accumulate a significant amount of lipid (Figure 1, with cells undergoing differentiation in media with high experiment 4). Conversely, cells deprived of glucose during glucose concentration (14 mM), the low glucose cells the early phase of differentiation, at a time when lipid accu- to This is were found differentiate much more slowly. mulation is absent, and then switched to high glucose media in reflected a 4-fold reduced accumulation of lipid, accumulate lipid very slowly (Figure 1, experiment 3). We exerts effect on measured by the intensity of staining of the cells with the conclude that low glucose its major adipog- in failure the time of maximal lipophilic dye Oil-Red-Oil (Figure IC) and their enesis, not as a substrate, during lipid rather as an inhibitor of com- to attain a cuboidal morphology, typical of adipocytes (not accumulation (days 3-8), but A 2 mM was chosen in the shown). concentration of glucose mitment to differentiation, early process (days 0-3). a modest decrease in Further evidence that low levels influence the because it is associated with only glucose no in cell internal of the differentiation comes rate of cell growth and change viability (data dynamics process of isoforms. The from of the C/EBP not shown). analysis expression in onset and attenuated as an source of in cul- induction of C/EBPa is delayed Glucose serves important energy cells when its are substrates for the in magnitude in the low glucose compared tured cells and glycolytic products concentration I The of acids. We therefore determined whether with the high glucose (Figure B). synthesis fatty which under our standard culture inhibition of differentiation in low induction of C/EBP,B, the observed adipocyte is sustained off concentration reflects substrate conditions throughout day 8, rapidly tapers glucose deprivation. By in to low concentration cells that had the initial of differ- cells exposed glucose (Figure 1B). switching undergone stages of CHOP reveals that to the in concentration to low Analysis protein exposure entiation high glucose (days 0-3) 4655 N.Batchvarova, X.-Z.Wang and D.Ron the . differentiation in i'~H t procedure high glucose concentration. A .( F with clones transfected Compared with the retro- empty the cells virus, wild-type CHOP-expressing exhibited markedly differentiation delayed (Figure 3A). Deletion of %:. ji. .... ..... :_ _ the basic or I DNA-binding region leucine dimeriza- zipper i, 1. R 6"mummm _0 _ as _ O .. . ~ ... ~ - tion domain abolished the anti-adipogenic effect of CHOP All CHOP derivatives (Figure 3A). were expressed at detectable as levels, though noted previously (Zinszner et the basic al., 1994), mutant accumulates to lower region levels than the other forms Over 50 clones (Figure 3B). 2. IBMX is Fig. required for the early induction of CHOP (day 2) by of cells CHOP-expressing and controls were evaluated in low glucose. 3T3-L1 pre-adipocytes were maintained in the indicated the course of four independent transfection experiments concentration of glucose from -1. Dexamethasone day (<0.25 ,uM, and the inhibitory effect was observed each time. Figure Dex), 3-isobutyl-l-methyl-xanthine both or (0.5 mM, IBMX), neither 3C a were added provides of the clonal variation in to the cultures from day 0-2. Whole cell obtained sample lysates adipogenic on days 1, 2 and 3 were analyzed Western blots differentiation by probed with of wild-type CHOP-expressing cells and, antibodies to CHOP (upper and CREB panel) (lower panel). for leucine minus comparison, zipper CHOP-expressing cells. adipogenic cocktail induces To (on days 0-2) a state of determine whether the inhibitory effect of CHOP increased responsiveness to the effect of 2 mM on glucose. adipogenesis is observable at physiologically-attainable The transient peak in CHOP level of expression noted on day 2 protein, CHOP expression in the transfected in cells differentiating in low is clones glucose highly reproducible was with the level compared of the protein in (see below). A second in CHOP peak expression, and one parental 3T3-L1 cells culture in low glucose or arginine which corresponds to the observed deficient previously induction media (another inducer of CHOP). CHOP levels of CHOP and in the six during adipogenesis (Ron wild-type clones (photographed in Habener, 1992; Figure 3C) Carlson et al., around 6. were no 1993), begins day The greater than those attained in the intensity parental cells of this later peak correlates with the positively to exposed to low glucose or arginine depleted degree media (Figure which cells have undergone commitment to differentiation 3D). In this small there is sample no correlation between (Figure 1, compare 1 the experiments and 3) and is level of CHOP and the strongly degree to which adipogenesis accentuated low by glucose concentration at these later is inhibited. as most However, clones expressed significant time points (Figure 1, compare experiments 1 and 3 with levels of the protein and most exhibited marked attenuation 2 and 4). Regardless of where in the differentiation of process adipocytic differentiation, it is likely that beyond a it occurs, the induction of CHOP is associated with an certain threshold level of CHOP other elements that vary attenuation in the level of and C/EBPa and correlates stochastically in the 3T3-LI population account for the with an inhibition in the process of differentiation. clonal variation in adipocyitc differentiation. We conclude The transient peak in CHOP observed in low glucose on that a physiologically-attainable level of CHOP is capable day 2 of the differentiation protocol suggests a of inhibiting adipogenesis in 3T3-LI synergism cells. between low and the glucose adipogenic cocktail. The effects of individual of the cocktail components on CHOP CHOP interferes with the normal differentiation- inducibility by low glucose were therefore In studied. the associated induction of C/EBPa and /3 absence of any of the 2 mM components cocktail, CHOP inhibits C/EBPa and , glucose DNA-binding activity to a did not induce CHOP to a detectable degree (Figure subclass of C/EBP target genes 2A). (Ron and Habener, 1992; in However, the presence of added 3-isobutyl-1-methyl- M.Ubeda et al., in preparation). Therefore, the observed xanthine (IBMX, a 3 ',5' nucleotide phosphodiesterase effect of CHOP on adipogenesis could be due to simple low inhibitor) glucose led to the brisk induction of CHOP stochiometric interaction with these C/EBP isoforms and (Figure 2B), comparable in level with that observed with inhibition of their binding to adipogenic target genes. the full differentiation cocktail (Figure 2D). Dexametha- However, analysis of cells in differentiating low glucose sone, the other essential ingredient in the differentiation concentrations suggested that inhibition of differentiation cocktail, was without effect (Figure 2C) and IMBX did and induction of CHOP are also associated with decreased not induce CHOP in the presence of normal expression glucose of C/EBPa and f8 (Figure iB). To determine concentration (Figure 2E). The effect of IBMX on CHOP whether CHOP plays an additional dynamic role in responsiveness to low glucose exhibits a latency of at influencing the level of C/EBP proteins, the expression of least 24 h. This indicates that it is due to an altered C/EBP isoforms was evaluated during exposure of CHOP- cellular phenotype rather than a direct interaction between expressing cells to the differentiation procedure at high the cAMP and pathway the pathway glucose inducing CHOP. concentration. Western blot analysis demonstrates that CHOP- Ectopic expression of CHOP attenuates adipocytic expressing cells exposed to the differentiation cocktail differentiation 1 of 3T3-L cells failed to induce C/EBPa. levels, though C/EBPP The above results led to an examination of the role CHOP transiently induced at the time of exposure to the adipo- may play in inhibiting the process of adipogenesis. 3T3- genic cocktail, were not sustained (Figure 4A). The normal LI cells were transfected with recombinant retrovirus that transient induction of C/EBP6 was unaffected by the expresses wild-type CHOP and the neo gene (as a dominant ectopic expression of CHOP. These results suggest that, selectable marker). Transfected cells were selected by in the dynamic setting of adipogenesis, CHOP is also culture in G418 and clones were expanded and subjected to implicated in inhibiting differentiation by a novel mechan- 4656 CHOP and adipogenesis z- - K =- me ,. I.W -, (-. >A7 II Pa a.. e.t t I Pa1I~rentalIL 1 Emptz- '1' ('110P~ NI' N,. ( CIlIOP-o - 'I '\ /~~~~- ('HOP LZ- CHOP BR- C1 3 4 6 ( "I' }IOP .l ,'r IS 1. ,i I, / .I ( IO0P 11- V. 2.5 .: EC ,1 (hIOP C'REB n4J 30one Fig. 3. CHOP inhibits adipocytic differentiation. (A) Clones of 3T3-Ll cells derived by transfection with retrovirus encoding wild-type CHOP (WT), CHOP deleted at the leucine zipper dimerization domain CHOP bearing an internal deletion of the DNA-binding basic region (BR-), cells (LZ-), infected with an retrovirus (Empty T) and the parental line (Parental Ll) were compared for their ability to differentiate to adipocytes. empty of 60 mm tissue culture plate stained with Oil-Red-Oil on day 8 of the differentiation process are shown. (B) Western blot analysis of Photographs cells infected with the retroviral vectors. As noted previously, CHOP LZ- and CHOP BR- migrate anomalously (Ron and CHOP-expressing 1992; M.Ubeda et al., in preparation). (C) Six representative clones of CHOP WT and CHOP LZ- 3T3-L1 cells were stained with Oil-Red- Habener, Oil on 8 of differentiation. The variation in intensity of the stain provides an estimate of the clonal variation in differentiation in both types of day cells. Extracts from 3T3-L1 cells undifferentiated sister plates of the CHOP WT expressing clones shown in (C) (lanes 1- (D) parental (parental Ll), cells cultured in media or low were evaluated by Western blot with antibodies to CHOP (upper panel) 6), parental arginine-free (Arginine-) glucose and CREB internal lower (the control, panel). 4657 N.Batchvarova, X.-Z.Wang and D.Ron ...g .} ie X ;i.1 To evaluate further the mechanism of CHOP-induced inhibition of C/EBPa and protein levels, we performed Northern blot analysis of RNA from cells ectopically ;_ 4 ..-, :: -z expressing CHOP. The absence of detectable C/EBPa and - -_ !: 1- the non-sustained induction of protein are due to ...0- C/EBPP P. the attenuated expression of their loom" Iim respective genes (Figure ._.. w *mm - '-------I-- 'NOWAO-- 4C). Thus we conclude that of CHOP presence during the differentiation process attenuates the proper coordinate expression of C/EBP isoforms a and The somewhat P. delayed peak in C/EBPa and mRNA accumulation (day P3 6) compared with the level protein (day 4) reflects the fact that the two experiments were not performed in parallel and demonstrates the inter-experimental variation ;... 1W s:.'..- in rates of differentiation. Irn ... . ... Ectopic expression of C/EBPa induces adipocytic differentiation of CHOP-containing 3T3-11 cells C/EBPa to lesser (and a degree plays a pivotal C/EBPP) role in adipocytic differentiation of 3T3-L1 cells. Inter- ference with the normal induction of C/EBP isoforms by CHOP could explain its inhibitory effect on the differentiation process. If this is the case then ectopic of 'rescue' expression C/EBPa may the CHOP-induced defect in 3T3-L I cells adipogenesis. ectopically expressing CHOP were transfected with a second recombinant retro- # # ~~~~~~~~~~~~~~~~~~~~~~~~~I _~~~~~~~~~~~~~~~~~~~~~~~~~~~~ virus that encodes 40- C/EBPa and contains the hisD gene as .,. off a dominant selectable f.0 marker. Following a 2 week selection itof with histidinol to enrich for the C/EBPa recombinant cells, foci of lipid containing cells were observed (Figure Cells with 5A, upper panels). transfected the empty inhibits the normal differentiation-associated induction Fig. 4. CHOP retroviral vector and likewise of and Western blot of extracts from CHOP selected showed no adipo- C/EBPa (A) analysis P. LZ- and CHOP WT cells at the indicated time of the points differentiation cytic (Figure 5A, upper panels). Appro- differentiation with antisera to and process, probed CIEBPa, 6. P of in the of transfected priate expression C/EBPa pool cells Estimate of the of nuclear extracts (B) C/EBP DNA-binding activity was revealed Western as is the by blotting, ectopically- prepared from CHOP LZ- and CHOP cells (lanes 1-4) WT-expressing CHOP expressed (Figure 5B). (lanes 5-8), before 1 and and after 2-4 and (lanes 5) (lanes 6-8) induction with the adipogenic cocktail. An of a The most of this result autoradiogram gel-shift straightforward interpretation is assay using the APRE site as a C/EBP-binding angiotensinogen gene that CHOP mediates its inhibitory influence on adipo- probe is shown. CHOP Bacterially-expressed wild-type (bCHOP WT, via its effects genesis on C/EBPa. However, an effect of lanes 4 and 8) or CHOP deleted at the leucine zipper (bCHOP LZ-, CHOP on some other of aspect adipocytic differentiation lanes 3 and 7) were added to the nuclear extract the in vitro during that is unrelated to binding reaction. (C) Northern blot of C/EBP isoforms from C/EBP isoform expression is not analysis control (right panel) and CHOP WT (left panel) cells undergoing excluded this 'rescue' by genetic experiment-C/EBPa adipocytic differentiation. The blot was with labeled mouse re-probed may serve as a dose simply 'high suppresser' of such tubulin cDNA to serve as an internal control. The 2 from day sample CHOP We also noted that the activity. degree of adipocytic the control cells is accidentally underloaded. differentiation in the cells expressing ectopic C/EBPa was augmented by treating the cells with adipogenic cocktail ism-blocking the accumulation of C/EBPa and 5 (not shown). This indicates that the ectopic C/EBPa is proteins. also capable of synergizing with endogenous adipogenic Following exposure to the adipogenic cocktail, CHOP- factors in the CHOP-expressing cells. expressing cells do manifest a increase in the significant C/EBP DNA-binding activity. This is demonstrated by the ability of nuclear extracts to shift the mobility of a 'classic' Discussion C/EBP-binding site probe (Figure 4B, lanes 5 and 6). Had The role its levels been obligatory played by C/EBPa and , in adipocytic high enough, CHOP would have been differentiation is capable of inhibiting all such C/EBP probably due to their activity as transcrip- DNA-binding activity, tion factors as adding excess bacterially-expressed CHOP to the gel- regulating the expression of downstream shift assay inhibits all adipogenic genes. DNA-binding activity (lanes 4 and CHOP-C/EBP heterodimers have a very 8). Bacterially-expressed CHOP that is lacking the leucine different DNA-binding activity than dimers of C/EBPs zipper dimerization domain is without effect, indicating (Ron and Habener, 1992; M.Ubeda et al., in preparation) that heterodimerization is the basis for the in vitro inhibi- and for a subset of 'classic' C/EBP target genes CHOP tion of DNA-binding activity in this assay (lanes 3 and acts as a C/EBP inhibitor. Inhibition of adipogenesis 7). We conclude that inhibition of DNA-binding activity by CHOP is therefore a predictable consequence of its by CHOP is incomplete and suggest that blocking expres- biochemical properties. the However, mechanism of sion of C/EBP's may be the predominant mode of CHOP's inhibition revealed in our study is more intricate than effect on adipogenesis. expected. Instead of predominantly stochiometric inhibi- 4658 CHOP and adipogenesis levels of C/EBP DNA-binding activity are still quite low. Such a scenario could explain how a limited amount of early enough in the differentiation CHOP, when present process, antagonizes what would otherwise be a large C H(1-P accumulation of C/EBPa and Furthermore, this model P3. would explain how ectopic expression of C/EBPa might, EKmpt% 1' by simple titration, overcome the inhibitory effect of CHOP. In this model, DNA binding by CHOP is not required for its inhibitory effect on adipogensis, the protein simply needs to form a non-productive heterodimer with relevant C/EBP isoforms early in the differentiation process. C'/EBR1) 1.iP aw- p ," to What then is the role of the CHOP basic region in > JX n w aS I attenuating adipogenesis? In vitro, the basic region is for ability of the protein to inhibit binding dispensable the \50, X44111 their target sequence (M.Ubeda et al., of C/EBPa and to in preparation). In vivo however, the basic region may heterodimers through non- stabilize such non-productive interactions. The observation that the specific chromatin BR- is at lower levels than the CHOP protein expressed wild-type protein might also provide part of the explanation et al., 1994). It remains possible (Figure 3B; Zinszner .Z Z = . however, that part of the effect of CHOP on adipogenesis sequence-specific interaction with is mediated through I.. ,; DNA-binding by CHOP target genes. Sequence-specific a partner and is com- requires dimerization with C/EBP pletely dependent on the integrity of the basic region (M.Ubeda et al., in preparation). Additional indirect sup- of port for the existence of a target gene-mediated effects CHOP is provided by its oncogenic variant, TLS-CHOP. ( ':'I.B In our assay the inhibitory effect of CHOP on adipogenesis is reproducibly more pronounced than that of TLS-CHOP (data not shown), yet both proteins dimerize equally well with C/EBPs (Crozat et al., 1993; Zinszner et al., 1994). C( O-lP m It is conceivable therefore that the two proteins, by virtue of differences in their transcriptional activation domains, Fig. 5. Ectopic expression of C/EBPa promotes adipocytic exert different influences on such target genes. differentiation of CHOP-expressing 3T3-LI cells. (A) CHOP- The finding that ectopic expression of C/EBPa 'rescues' expressing cells were transfected with recombinant retroviruses the adipogenic phenotype of CHOP-expressing cells is encoding C/EBPa or the empty virus (Empty '). Two weeks later the consistent with a model whereby attenuation of differentia- were stained with Oil-Red-Oil and photographed. (B) Western plates blot of extracts from sister plates of those shown in (A) were probed tion is mediated by interference with a C/EBP signaling with antiserum to C/EBPa (upper panel) or CHOP (lower panel). The specifically one mediated by C/EBPa. It is pathway, asterisk in the upper panel indicates a non-specific reactivity of the that both ectopic expression of CHOP and noteworthy antiserum. C/EBP CHOP induction by low glucose levels are associated with marked decrease in C/EBPa level. This is associated with tion of C/EBP activity by hetrodimerization low levels of C/EBPa mRNA and raises the possibility DNA-binding of with CHOP we find a striking that CHOP is a direct negative regulator C/EBPa gene ectopically-expressed mode of attenuation in the C/EBPa (and to a lesser degree C/EBP,) A parallel, phenotypically-dominant expression. however not ruled out. CHOP could action for CHOP, is gene expression. Recent studies that induction of C/EBP isoforms inhibit adipogenesis by activating anti-adipogenic target suggest of C/EBPa a differentiation is autocatalytic: omitting with attenuated expression being during adipocytic genes, rather than its cause. In a inducer of the C/EBP8 marker of failed differentiation dexamethasone, potent early could rescue the from the cocktail leads to diminished the latter case C/EBPa phenotype by isoform, adipogenic with CHOP and the of and a isoforms and results in dimerizing altering target gene accumulation C/EBPP of the CHOP hetero- or activation attenuated differentiation (Yeh et al., 1995). The specificity properties markedly the CHOP effect sites and is could also overcome contains dimer. C/EBPa C/EBPa promoter C/EBP-binding as a et isoforms 1991; by activating 'suppressor' adipogenic target genes activated by C/EBP (Christy al., for CHOP action are consistent of homodimer. Both models et exogenous Legraverend al., 1993); ectopic expression for CHOP to retain its in cells with the established requirement induces the 3T3-L1 C/EBPa endogenous gene dimerization domain and This that CHOP block leucine DNA-binding and suggests may zipper (Lin Lane, 1994). in basic region order to inhibit adipogenesis. and the activation of the C/EBPa insinuating genes by the between the of dimerization in What be significance parallels itself as a partner early might non-productive and culture in of CHOP at a time when ambient inhibition adipogenesis by by the C/EBP autoregulatory cascade, 4659 N.Batchvarova, X.-Z.Wang and D.Ron low Our results indicate glucose? that failure to attain an between the of an altered form of presence CHOP, TLS- when in adipocytic cultured low is not CHOP and an tissue phenotype glucose myxoid liposarcoma, adipose tumor due to substrate of cells that simply deprivation are (Aman et al., 1992) may indicate that this or similar otherwise committed to the normally differentiation pathways play a role in normal metabolism of pro- adipose cess. these untoward culture Rather, conditions affect tissue and derangement therein may participate in the the intrinsic differentiation process-there is development of markedly neoplasia. accumulation of and delayed sustained C/EBPa poorly accumulation of in low media. C/EBP, glucose The Materials and methods same alterations are observed in cells CHOP expressing low ectopically. seems to exert a Furthermore, glucose Cell culture and induction of differentiation more effect on profound differentiation when adipocytic 3T3-L1 pre-adipocytes were maintained in Dulbecco's modified Eagle's in the medium (DMEM) with 10% calf serum. cells of present early differentiation And it is Only early passages were process. during utilized for transfection and differentiation experiments. Differentiation this window of time that (around day 2) differentiating induction was performed as described previously (Student et al., 1980). cells cultured in low activate CHOP. This glucose temporal Briefly, 2-day post confluent cultures were placed in DMEM at the correlation that CHOP be a role suggests may playing in indicated concentration of glucose (14 or 2 mM) supplemented with some of the effects of low 10% fetal bovine serum (FBS, Intergen), mediating inhibitory glucose 10 of insulin gg/ml (Novo Nordisk), 0.25 dexamethasone on gM (Sigma) and 0.5 mM differentiation. 3-isobutyl-l- adipocytic methyl-xanthine (Sigma). The start point of differentiation is referred to As noted the differentiation itself influ- above, process as day 0. Forty-eight hours later (day 2) 3-isobutyl-l-methyl-xanthine ences the of the CHOP to low responsiveness gene glucose and dexamethasone were removed from the culture medium and the In levels. the absence of low does not IBMX, glucose cells were placed in fresh medium containing only insulin and 10% FBS. Media was replaced induce CHOP to a measurable In its every 2 days thereafter. The accumulation of degree (Figure 2). cytoplasmic triglyceride droplets was visualized by staining with Oil- CHOP is induced low but presence by glucose as then, Red-Oil (Aldrich Chemical Company, Inc.) as previously described the hormonal cocktail is cells withdrawn, down-regulate (Preece, 1989). CHOP expression (Figure 1, experiment 2). However, when cells differentiation at a undergoing high glucose Retroviral expression vectors and isolation of stable cell concentration are switched to a low lines glucose level, profound Replication defective ecotrophic retroviruses CHOP and encoding its induction of the CHOP is observed gene (Figure 1, derivatives were constructed introducing by the respective cDNAs into We note that both the IBMX effect on experiment 4). day the pSRa proviral plasmid and co-transfected with a plasmid encoding and the second in CHOP to low peak responsiveness helper functions into COS1 cells to the of leading production viral in committed cells are associated with glucose (on day 6) particles in the culture et The supematant (Muller al., 1991). full-length CHOP cDNA was excised from the CHOP-10 levels of et Yeh et 1995 pBS plasmid (Ron and high (Cao al., 1991; al., C/EBPP Habener, 1992) and ligated into To create pSRa. the leucine zipper and The CHOP is to this Figure 1). promoter responsive deleted form of CHOP (CHOP a BamHI-NheI LZ-), fragment of CHOP factor et the transcription (Sylvester al., 1994), suggesting was ligated into pSRa. This encodes a in which the protein 34 C- that a role in the of possibility plays inducibility terminal amino acids of CHOP are vector-encoded replaced by C/EBPP sequence. CHOP low It is to out that To create the basic mutant form of CHOP by a region (CHOP BR-), the glucose. important point corresponding cDNA was excised from the CHOP-Cb-NLSpcDNA late in CHOP occurs in even smaller, peak expression (Barone et al., 1994) and into The retroviral ligated pSRa. vector committed cells at concentra- differentiating high glucose encoding C/EBPca was constructed by introducing the rat cDNA into tion and Ron and (Figure 1 iB, experiment Habener, the proviral plasmid LXSHD. This virus the expresses gene of interest Carlson et Whether this is due to a 1992; al., 1993). from the retroviral LTR and the hisD a dominant selectable gene, marker gene, is driven by an internal SV40 similar mechanism as that associated with low early promoter (Hartman and glucose is Mulligan, 1988). Viral particles were in cells produced COS1 as not known. of the culture However, frequent replacement described above. Retroviral transfection of 3T3-LI cells was performed media can attenuate this later that peak, suggesting deple- as described previously (Zinszner et al., that 1994), except transfected tion of another essential nutrient a role in CHOP may play cells were selected with 450 tg/ml G418 and 0.2 (Gibco BRL) mM induction. L-histidinol (Sigma) in histidine-free DMEM-10% calf serum. One of the well-characterized effects of low is glucose Western blot analysis to induce that a role as in the proteins play chaperones Whole cell extracts from 3T3-LI cells were prepared in Laemmli buffer, reticulum endoplasmic et This (ER) (Poussegur al., 1977). electrophoresed on a 12% SDS-polyacrylamide gel (or 10% for the effect has been ascribed to interference with normal C/EBPa transferred to a immunoblot), nitrocellulose filter (Micron of secreted glycosylation proteins rather than depletion of Separations and then reacted with the Inc.) indicated antisera. The immunoreactive stores protein species were visualized by the enhanced As chemi- energy (Lee, 1987). previously pointed out, luminescence detection system (ECL, DuPont-NEN). To equilibrate of the stressful many events that induce CHOP also induce sample loading, parallel-run protein gel was stained with Coomassie accumulation of defective proteins in the ER (Carlson blue. The rabbit antisera to C/EBPa and and CREB have been C/EBPP et The al., 1993). trigger for CHOP induction by low described et previously (Ron al., 1992). The polyclonal antisera against glucose be stress in the ER C/EBP8 were from may induced by such malfolded purchased Santa Cruz Biotechnology, Inc. The monoclonal to antibody CHOP, 37-9C8 recognizes an epitope in the N- proteins. Adipocytic differentiation is associated with a terminus of the protein and crude hybridoma supernatant was used at a induction significant of cell secretory activity leading, for dilution of 1:10. to exuberant example, production of matrix glycoproteins et (Kuri-Harcuch al., 1994). The increase in secretory Northern blot analysis is to be activity likely associated with Total RNA was isolated from an increased 3T3-LI cells at various time points during the differentiation process by the for stress in the phenol/guandinium isothiocyanate potential ER. We speculate that one method of extraction (RNAzol B, Biotex Laboratories, Inc.). The RNA important function of CHOP may be to modulate the were samples electrophoresed through formaldehyde-containing agarose normal differentiation so as not to incur a critical process blotted to membranes gels, and nylon (Amersham) hybridized to level of such stress. The association and cDNA as C/EBPa described strong 0 probes et epidemiological previously (Ron al., 1992). 4660 CHOP and adipogenesis To control for RNA loading and transfer efficiency the blot was Sylvester,S.L., ap Rhys,C.M.J., Luethy,J.D. and Holbrook,N.J. (1994) J. with the cDNA for of Dr hybridized ox-tubulin (a gift Nicholas Cowan, Biol. Chem., 269, 20119-20125. New York University, New York). Williams,P.M., Danesch,U., Ringold,G. and Heller,R. (1992) Chang,D.J., Mol. Endocrinol., 6, 1135-1141. DNA-binding assay Yeh,W.-C., Cao,Z., Classon,M. and McKnight,S. (1995) Genes Dev., 9, EMSA experiments were performed exactly as previously described 168-181. using the angiotensinogen APRE sequence as a probe (Ron and Habener, Zinszner,H., Albalat,R. and Ron,D. (1994) Genes Dev., 8, 2513-2526. 1992). Where indicated, 100 ng of purified bacterially-expressed gluta- thione S-transferase-CHOP fusion protein was added to the DNA- Received on June 13, 1995 reaction to the addition of the radiolabeled binding prior probe. Acknowledgements We are indebted to Niki Holbrook and Lennart Philipson for useful discussions and to Ueli Schibler for emphasizing the importance of stochiometric considerations in the action of inhibitors. This work was supported by a PHS research award (DK47119) and by generous support from the Arthritis Foundation. D.R. is a Pew Scholar in Biomedical Sciences. References Ailhaud,G., Grimaldi,P. and Negrel,R. (1992) Annu. Rev. Nutr., 12, 207-233. Aman,P., Ron,D., Mandahl,N., Fioretos,T., Heim,S., Arhenden,K., and Willen,H., Rydholm,A. Mitelman,F. (1992) Genes Chrom. Cancer, 5, 271-277. Barone,M.V., Crozat,A.Y., Tabaee,A., Philipson,L. and Ron,D. (1994) Genes Dev., 8, 453-464. Birkenmeier,E.H., Gwynn,B., Howard,S., Jerry,J., Gordon,J.I., Land- and schultz,W.H. McKnight,S.L. (1989) Genes Dev., 3, 1146-1156. and Cao,Z., Umek,R.M. McKnight,S.L. (1991) Genes Dev., 5, 1538- and Carlson,S.G., Fawcett,T.W., Bartlett,J.D., Bernier,M. Holbrook,N.J. (1993) Mol. Cell. Biol., 13, 4736-4744. Christy,R.J., Yang,V.W., Ntambi,J.M., Geiman,D.E., Landschulz,W.H., Friedman,A.D., Nakabeppu,Y., Kelly,T.J. and Lane,M.D. (1989) Genes Dev., 3, 1323-1335. and Proc. Christy,R.J., Kaestner,K.H., Geiman,D.E. Lane,M.D. (1991) Natl Acad. Sci. USA, 88, 2593-2597. and Crozat,AY., Aman,P., Mandahl,N. Ron,D. (1993) Nature, 363, 640-644. Fornace,A.J., Neibert,D.W., Hollander,M.C., Luethy,J.D., Papathan- asiou,M., Fragoli,J. and Holbrook,N.J. (1989) Mol. Cell. Biol., 9, 4196-4203. and Genes 1654- Freytag,S.O., Paielli,D.L. Gilbert,J.D. (1994) Dev., 8, and 113-116. Green,H. Kehinde,O. (1974) Cell, 1, Hartman,S. and Mulligan,R. (1988) Proc. Natl Acad. Sci. USA, 85, 8047-8051. and Herrera,R., Ro,H.S., Robinson,G.S., Xanthopoulos,K.G. Spiegel- Mol. Cell. 5331-5339. man,B.M. (1989) Biol., 9, and Proc. Natl Acad. Sci. Kaestner,K.H., Christy,R.J. Lane,M.D. (1990) 251-255. USA, 87, and Kuri-Harcuch,W., ArgUello,C. Marsch-Moreno,M. (1994) Differen- 173-178. tiation, 28, Trends Biochem. 20-23. Lee,A.S. (1987) Sci., 12, and Legraverend,C., Antonson,P., Flodby,P. Xanthopoulos,K. (1993) Nucleic Acids 1735-1742. Res., 21, and Genes 533-544. Lin,F.-T. Lane,M.D. (1992) Dev., 6, and Proc. Natl Acad. Sci. 8757-8761. Lin,F.-T. Lane,M. (1994) USA, 91, and Cancer 5-10. Luethy,J.D. Holbrook,N.J. (1992) Res., 52, Muller,A.J., Young,J.C., Pendergast,A.-M., Pondel,M., Landau,N.R., 1785-1792. and Mol. Cell. Littman,D.R. Witte,O.N. (1991) Biol., 11, 941-947. and Pastan,I. Poussegur,J., Shiu,R.P.C. (1977) Cell, 11, Little Brown and Preece Manual Technicians. (1989) for Histological Co., Boston, pp. 3814-3817. and Cancer Price,B. Calderwood,S. (1992) Res., 52, and Nature Genet., Rabbitts,T.H., Forster,A., Larson,R. Nathan,P. (1993) 175-180. 4, Genes and Habener,J.F. Dev., 6, 439453. Ron,D. (1992) and J. Clin. Habener,J.F. (1992) Ron,D., Brasier,A.R., McGehee,R.E.,Jr 223-233. Invest., 89, and Enerb6ck,S. Samuelsson,L., Stromberg,K., Vikman,K., Bjursell,G. EMBO 3787-3793. (1991) J., 10, and J. Biol. 4745-4750. Chem., 255, Student,A., Hsu,R. Lane,M. (1980)

Journal

The EMBO Journal – Springer Journals

Published: Oct 1, 1995

There are no references for this article.