Skip to main page content

• HOME
• Current Issue
• Archive
• Alerts
• About ASM
• Contact us
• Tech Support
• Journals.ASM.org

This Article Abstract Full Text (PDF) Supplemental material Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Reprints and Permissions Copyright Information Books from ASM Press MicrobeWorld Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Jimenez, M. A. Articles by Rosen, E. D. Search for Related Content PubMed PubMed Citation Articles by Jimenez, M. A. Articles by Rosen, E. D.

 Previous Article  |  Next Article 

Molecular and Cellular Biology, January 2007, p. 743-757, Vol. 27, No. 2
0270-7306/07/$08.00+0     doi:10.1128/MCB.01557-06
Copyright © 2007, American Society for Microbiology. All Rights Reserved.

Critical Role for Ebf1 and Ebf2 in the Adipogenic Transcriptional Cascade ^,

Maria A. Jimenez,^1,2 Peter Åkerblad,³ Mikael Sigvardsson,⁴ and Evan D. Rosen^1,2*

Division of Endocrinology, Beth Israel Deaconess Medical Center, Boston, Massachusetts,¹ Broad Institute of Harvard and MIT, Cambridge, Massachusetts,² Astra Zeneca R&D, Mölndal, Sweden,³ Department of Stem Cell Biology, Lund University, Lund, Sweden⁴

Received 21 August 2006/ Accepted 10 October 2006

Top Abstract Introduction Materials and Methods Results Discussion References

 
The Ebf (O/E) family of helix-loop-helix transcription factors plays a significant role in B lymphocyte and neuronal development. The three primary members of this family, Ebf1, 2, and 3, are all expressed in adipocytes, and Ebf1 promotes adipogenesis when overexpressed in NIH 3T3 fibroblasts. Here we report that these three proteins have adipogenic potential in multiple cellular models and that peroxisome proliferator-activated receptor (PPAR) is required for this effect, at least in part due to direct activation of the PPAR1 promoter by Ebf1. Ebf1 also directly binds to and activates the C/EBP promoter, which exerts positive feedback on C/EBP expression. Despite this, C/EBP is dispensable for the adipogenic action of Ebf proteins. Ebf1 itself is induced by C/EBPß and , which bind and activate its promoter. Reduction of Ebf1 and Ebf2 proteins by specific short hairpin RNA blocks differentiation of 3T3-L1 cells, suggesting a critical role for these factors and the absence of functional redundancy between members of this family. Altogether, these data place Ebf1 within the known transcriptional cascade of adipogenesis and suggest critical roles for Ebf1 and Ebf2.

Top Abstract Introduction Materials and Methods Results Discussion References

 
The last decade has seen an enormous surge of interest in the biology of adipocytes, including the developmental processes by which these cells are formed. This has been fuelled by the convergence of a worldwide epidemic of obesity and diabetes (26, 37), with the recent realization that adipose tissue is an active secretory organ regulating a wide array of physiological processes.

Adipogenesis represents a complex series of transcriptional events through which multipotent mesenchymal precursor cells become committed to the adipocyte lineage and ultimately express all of the genes typical of mature fat cells (31). These transcriptional events integrate a variety of extracellular signals that direct fat cell formation in time and space. Most of our knowledge concerning the transcriptional events mediating adipogenesis has come from cultured cell lines such as 3T3-L1 and 3T3-F442A (16, 17) that can be differentiated into fat cells by empirically determined hormonal cocktails. These in vitro systems appear to recapitulate most of the developmental events that take place in vivo and offer the advantage of synchronous differentiation and ease of manipulation. Studies of these cells, as well as cultured mouse embryonic fibroblasts (MEFs) derived from mice with various targeted gene ablations (and more recently, studies from intact animals) have revealed a transcriptional cascade in which the nuclear hormone receptor peroxisome proliferator-activated receptor (PPAR) plays a critical role. In cultured adipocytes and fat pads in vivo, PPAR is both necessary and sufficient for adipogenesis (4, 33, 41). Other transcription factors have also been shown to play an important role in adipogenesis, including the CCAAT/enhancer binding proteins C/EBP, C/EBPß, and C/EBP and the basic helix-loop-helix protein SREBP1c (34). In 3T3-L1 cells exposed to a proadipogenic hormonal cocktail, the levels of C/EBP and C/EBPß rise rapidly (6, 50). Within a few days, the levels of these proteins decrease while C/EBP and PPAR levels increase. C/EBP and PPAR sustain each other's expression through a positive feedback loop (12, 49) and remain elevated throughout the life of the terminally differentiated adipocyte, where they participate in the control of genes involved in lipogenesis, insulin sensitivity, and other pathways.

Although most studies have been devoted to the role of PPAR and C/EBP proteins, it is known that a large number of other transcription factors and cofactors are regulated during adipogenesis. These factors may play a role in the expression of subsets of genes within the terminally differentiated adipocyte, or they may have a more fundamental impact on the process of differentiation per se. Consistent with this idea, there have been recent reports linking other transcriptional regulators to the adipogenesis cascade, such as the Krüppel-like zinc finger transcription factor family members KLF2 (which represses adipogenesis) (3) and the proadipogenic KLF5 (29) and KLF15 (27). Similarly, the zinc finger-containing factor KROX 20 (7) has been shown to participate early in the adipogenic cascade. A number of other transcription factors have also been shown to negatively regulate adipogenesis, for example, GATA-2 and -3 (39, 40), the forkhead transcription factors FoxO1 (28) and FoxA2 (47), the HMG proteins TCF/Lef (21), and SMAD-3 (8). These findings suggest that fat cell development is a more complex process than previously appreciated, requiring the integration of multiple transcriptional regulators to determine differentiation and function of the mature adipocyte.

In this report, we have focused on the adipogenic potential of a family of transcription factors known as Ebf proteins. The prototypical member of this family, Ebf1, was identified by its role in B-cell lymphopoiesis (18) and was independently identified as playing a role in the development of olfactory neurons (42). Ebf proteins have been implicated in other developmental processes as well, including bone formation (22) and neuronal differentiation (9, 14, 44). Like many other developmental regulators, the Ebf family is highly conserved evolutionarily, with orthologs in Caenorhabditis elegans and Drosophila melanogaster in addition to all vertebrates tested (11, 23). Mammals have four Ebf proteins encoded by distinct genes, designated Ebf1 through 4 (43, 45). Ebf proteins bind their cognate DNA sequences as homo- or heterodimers, the formation of which is mediated by an atypical helix-loop-helix motif. DNA binding by Ebf factors requires a zinc-coordinating motif without homology to classic zinc fingers (19).

We have shown that Ebf1, 2, and 3 are expressed in mouse and human adipose tissue as well as in 3T3-L1 adipocytes (1). More importantly, Ebf1 was shown to promote adipogenesis in 3T3-L1 cells, untransformed MEFs, and NIH 3T3 fibroblasts, while a "dominant-negative" Ebf1, consisting of the Drosophila engrailed repressor fused to the Ebf DNA binding domain, inhibited the differentiation of 3T3-L1 cells. Furthermore, transcriptional profiling in NIH 3T3 fibroblasts suggested that Ebf1 and PPAR might induce distinct gene sets early in adipogenesis (2).

In the current set of experiments, we examine the role of other Ebf proteins in adipogenesis. Specifically, we demonstrate that multiple Ebf proteins are adipogenic in gain-of-function experiments in NIH 3T3 fibroblasts and in various MEF lines. We show that C/EBP and PPAR1 are direct Ebf targets and that Ebf proteins likely amplify the actions of C/EBPß and C/EBP by acting immediately downstream of those factors. Finally, we use short hairpin RNA (shRNA)-mediated knockdown to show that Ebf1 and Ebf2 (but not Ebf3) are required for adipogenesis, indicating that these closely related factors have functionally nonredundant roles in this developmental process.

Top Abstract Introduction Materials and Methods Results Discussion References

 
Cell culture. All cells were cultured in Dulbecco's modified Eagle's medium (Invitrogen) with 10% bovine calf serum (Invitrogen) at 5% CO₂. After retroviral infection (see below), cells were allow to grow to confluence in either 100-mm or 60-mm dishes in Dulbecco's modified Eagle's medium with 10% fetal bovine serum. Two days postconfluence, cells were exposed to differentiation medium containing dexamethasone (1 µM), insulin (5 µg/ml), and isobutylmethylxanthine (0.5 mM) (DMI). After 2 days, cells were maintained in medium containing insulin (5 µg/ml) until ready for harvest at 7 days. PPAR^flox/– and PPAR^–/– cells and C/EBP^+/+ and C/EBP^–/– cells were generated as described previously (32, 49). Human preadipocytes were isolated and differentiated as described previously (38).

Oil-red-O staining. After 7 to 10 days of differentiation, cells were washed once in phosphate-buffered saline and fixed with formaldehyde (Formalde-fresh; Fisher) for 15 min. The staining solution was prepared by dissolving 0.5 g oil-red-O in 100 ml of isopropanol; 60 ml of this solution was mixed with 40 ml of distilled water. After 1 h at room temperature, the staining solution was filtered and added to dishes for 4 h. The staining solution was then removed, and cells were washed twice with distilled water.

Generation of retroviral constructs and retroviral infections. Retroviruses were constructed in pMSCV vectors (Clontech) with puromycin- or hygromycin-selectable markers. Ebf1, Ebf2, and Ebf3 cDNA were first subcloned into the pSG5-KF1M2 Flag-tag vector (gift from D. Dowhan), and then the Flag-tagged cDNAs were moved to pMSCV. Additionally, all constructs were subcloned into pCDNA3 (Invitrogen) for in vitro transcription and translation. All three Ebf constructs produced a band of the correct size upon immunoblotting with anti-Flag antibody (see Fig. S1 in the supplemental material) (Sigma). PPAR2 pMSCV retroviral vector was constructed as described previously (32). Viral constructs (10 µg) were transfected using CellPhect transfection kit (Amersham Biosciences) into Phoenix packaging cells along with constructs encoding gag-pol (5 µg) and vesicular stomatitis virus glycoprotein G (5 µg), and supernatants were incubated in the presence of 3 µM trichostatin A (Sigma) and either used immediately or snap frozen and stored at –80°C for later use. Viral supernatants were added to the cells for 4 h in the presence of Polybrene (8 µg/ml). Selection with puromycin (4 µg/ml) or hygromycin (400 µg/ml) was started 24 h after infection and continued until cells infected with a control virus carrying no antibiotic resistance cassette were all killed (usually 4 days for puromycin and 7 days for hygromycin). Experiments were repeated three times.

Isolation of adipocytes, macrophages, and nonmacrophage SVCs from perigonadal adipose tissue. Five-week-old male C57BL/6J mice (n = 10) were obtained from The Jackson Laboratory (Bar Harbor, MN) and were fed a standard diet (rodent diet 8664; Harlan Teklad, Madison, WI). At 19 weeks of age, mice were euthanized by CO₂ inhalation, and epididymal adipose tissue (0.5 g) was collected and placed in Krebs Ringer HEPES buffer containing 10 mg/ml fatty-acid-poor bovine serum albumin (Sigma-Aldrich, St. Louis, MO). The tissue was minced into fine pieces and centrifuged at 1,000 x g for 10 min to remove erythrocytes and other blood cells. Minced tissue was then digested in 0.12 units/ml of low-endotoxin collagenase (Liberase 3; Roche Applied Science, Indianapolis, IN) at 37°C in a shaking water bath (80 Hz) for 45 min. Samples were then filtered through a sterile 300-µm nylon mesh (Spectrum Laboratories Inc., Rancho Dominguez, CA) to remove undigested fragments. The resulting suspension was centrifuged at 500 x g for 10 min to separate stromal vascular cells (SVCs) from adipocytes. Adipocytes were removed and washed with Krebs Ringer HEPES buffer. They were then suspended in Tri reagent phenol guanidine thiocyanate solution (Molecular Research Center, Inc., Cincinnati, OH), and RNA was isolated according to the manufacturer's instructions. The SVC fraction was incubated in erythrocyte lysis buffer (0.154 mM NH₄Cl, 10 mM KHCO₃, 0.1 mM EDTA) for 2 min. Cells were then centrifuged at 500 x g for 5 min and resuspended in 100 µl of fluorescence-activated cell sorter buffer (phosphate-buffered saline containing 5 mM EDTA and 0.2% fatty-acid-poor bovine serum albumin). The cells were incubated in the dark on a bidirectional shaker with FcBlock (20 µg/ml; BD Pharmingen, San Jose, CA) for 30 min at 4°C. They were then incubated for 50 min with allophycocyanin-conjugated primary antibody against F4/80 (5 µg/ml; Caltag Laboratories Inc., Burlingame, CA) and phycoerythrin-conjugated antibody against CD11b (Mac-1; 2 µg/ml). F4/80 and CD11b are cell surface markers of mature macrophages (5, 25) that are not found on preadipocytes (10). Control aliquots of SVCs were incubated with allophycocyanin-labeled (2 µg/ml) and phycoerythrin-labeled (5 µg/ml) isotype control antibodies (Caltag Laboratories Inc., Burlingame, CA). After incubation, cells were washed and suspended in fluorescence-activated cell sorter buffer. F4/80⁺/CD11b⁺ macrophages and F4/80^–/CD11b^– nonmacrophage SVCs were isolated with a MoFlo (DakoCytomation, Fort Collins, CO) fluorescence-activated flow sorter. After sorting, F4/80⁺/CD11b⁺ and F4/80^–/CD11b^– cells were suspended in Trizol (Invitrogen) and RNA was isolated according to the manufacturer's instructions.

RNA extraction and quantitative PCR (Q-PCR). Total RNA was isolated from cultured cells using Trizol reagent according to the manufacturer's instructions. First-strand cDNA was synthesized from 2 µg of total RNA and oligo(dT) primers using Moloney murine leukemia virus reverse transcriptase (RETROscript; Ambion). Real-time PCR was performed using SYBR green PCR Master Mix (Stratagene). Specific mouse primers for each gene were as follows: cyclophilin-F, 5'-GCATACAGGTCCTGGCATCT; cyclophilin-R, 5'-TTCACCTTCCCAAAGACCAC; C/EBP-F, 5'-TGTGCGAGCACGAGACGTC; C/EBP-R, 5'-AACTCGTCGTTGAAGGCGG; C/EBPß-F, 5'-CTATTTCTATGAGAAAAGAGGCGTATGTAT; C/EBPß-R, 5'-ATTCTCCCAAAAAAGTTTATTAAAATGTCT; C/EBP-F, 5'-TGCCCACCCTAGAGCTGTG; C/EBP-R, 5'-CGCTTTGTGGTTGCTGTTGA; Ebf1-F, 5'-AGGTTGGATTCTGCTACGAAACTT; Ebf1-R, 5'-TGATTCCTCTTAAAAAGGCCTGA; Ebf2-F, 5'-AGCACAAAACTACTTATTCCGATGG; Ebf2-R, 5'-GTCCAACATGGCCGCTTG; Ebf3-F, 5'-TCGTGAATATGCACCGTTTTG; Ebf3-R, 5'-ATGAGTACAGAAAAAATGTCTCGAGG; aP2-F 5'-TTCGATGAAATCACCGCAGA; aP2-R, 5'-AGGGCCCCGCCATCT; PPAR2-F, 5'-GCATGGTGCCTTCGCTGA; PPAR2-R, 5'-TGGCATCTCTGTGTCAACCATG; Adipsin-F, 5'-CCTTGCAATACGAGGACAAAGA; Adipsin-R, 5'-CACACCCCAACCAGCCAC; PPAR1-F, 5'-TGAAAGAAGCGGTGAACCACTG; PPAR1-R, 5'-TGGCATCTCTGTGTCAACCATG; CHOP-10-F, 5'-CGGAACCTGAGGAGAGAGTG; CHOP-10-R, 5'-GGACGCAGGGTCAAGAGTAG; Pref1-F, 5'-GAAATAGACGTTCGGGCTTG; Pref1-R, 5'-AGGGGTACAGCTGTTGGTTG; GATA2-F, 5'-TGCATGCAAGAGAAGTCACC; GATA2-R, 5'-ACCACCCTTGATGTCCATGT; GATA3-F, 5'-GTCATCCCTGAGCCACATCT; GATA3-R, 5'-GTAGAAGGGGTCGGAGGAAC. Primers for human genes were as follows: Ebf1-F, 5'-TTTTTCGGGTTATCGCTCAG; Ebf1-R, 5'-AATCTCCTCCCCCTTGAAAA; Ebf2-F, 5'-GGGACAAGGACAAAAGAT; Ebf2-R, 5'-TCAATTCAGGTTGCGTTT; Ebf3-F, 5'-GACCCAAATCACTGCTGGTT; Ebf3-R, 5'-TCGATTTCTGCTGCAGTTTG. All reactions were normalized to cyclophilin levels and performed in triplicate. Statistics were performed using the Student t test.

Protein extraction and Western blot analysis. Whole-cell protein lysates were prepared according to the manufacturer's protocol using a radioimmunoprecipitation assay (RIPA) buffer containing 0.1% sodium dodecyl sulfate (Boston BioProducts) and protease inhibitor cocktail Complete (Roche). Western blot analyses were performed as described previously (20); 30 µg of total protein was loaded on a 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis gel for each sample. Mouse anti-FLAG antibody (3040; Sigma), 1/2,000 dilution, mouse anti-PPAR antibody (SC 7273; Santa Cruz Biotechnology), 1/2,000 dilution, and anti-mouse immunoglobulin G (IgG)-peroxidase conjugate (Sigma), 1/5,000 dilution, were used to detect Flag and PPAR proteins. SuperSignal West Pico stable peroxide solution (Pierce) was used for chemiluminescent detection.

Transfection of adipocytes. 3T3-L1 adipocytes were induced to differentiate as above. Five days after the addition of DMI, cells were transfected using the Amaxa nucleofection system (Amaxa Biosystems). A six-well plate adipocyte was transfected with 3 µg of Ebf1-pCDNA3 or empty vector along with a ß-galactosidase expression vector (2 ng) as a control for transfection efficiency. Cells were harvested for RNA extraction 24 h after transfection. Statistical analyses were performed using the Mann-Whitney test.

Chromatin immunoprecipitation (ChIP) assay. Genomic DNA was extracted, and immunoprecipitation was performed according the manufacturer's protocol (Upstate Biotechnology). Cross-linking was performed as described in the manufacturer's protocol. DNA was sheared using a sonic dismembrator model 100 (Fisher Scientific) at half maximum speed for 10 s. Immunoprecipitation was performed using 2 µg of the following antibodies: C/EBP, C/EBPß, and C/EBP (SC-61, SC-150, and SC-636; Santa Cruz Biotechnology) or anti-Flag M2 antibody (F1814; Sigma). The following primers were used to screen for C/EBP binding sites on the murine Ebf1 promoter: site A-F, 5'-TTGAATAGCCCTTAGCTGC; site A-R, 5'-GACTACAGTCCACTCCCATTG; site B-F, 5'-CCTCACCTGTACAATGG; site B-R, 5'-GCAGAGACTGCAGTCAACG; site C-F, 5'-TGTGGAACCGCACCGCG; site C-R, 5'-ATCCGGGCTTGCGACCTCG. As a negative control, we used primers to a site approximately 2 kb upstream of the transcriptional start site: F, 5'-AGCCATGAAGGGAGAGGAAT; R, 5'-CCGGTTCTGCTTTGCATACT. The following primers were used to screen for Ebf binding sites in the PPAR1 promoter: F, 5'-TCCATGACAGACATGGACA; R, 5'-CCGGAGGAGCCGAGCCG. The upstream site was used as a negative control 2 kb from the actual site: F, 5'-CTTAATTCATATAAATAT; R, 5'-CATGTCCATGTCTGTCATG. To check the Ebf site in the C/EBP promoter, the following primers were used: F, 5'-TCTCTCTCCACTAGCACTATGC; R, 5'-AACTGGCTCGCGCCCGCGCA. For the negative control, a site 2 kb upstream was checked with 5'-GTTTTTAGCCGTCCTCTTG (F) and 5'-GTCTCGTTCACCCTCCACCT (R) primers. A parallel immunoprecipitation with only the secondary antibody (anti-mouse IgG) or the Flag antibody on 3T3-L1 cells infected with the empty retroviruses and differentiated as mentioned above were performed.

Transactivation assays. The C/EPB promoter was constructed by the ligation of a 300-bp PCR fragment containing the proximal promoter (–302/+1) as a fragment into the KpnI site of pGL3 basic (Promega). The 0.8-kb Ebf1 promoter in pGL3 basic was described by Smith et al. (36). The C/EBP, C/EBPß, and C/EBP expression vectors were constructed by cloning into the pCDNA3 vector. NIH 3T3 cells were transfected at 80% confluence using Lipofectamine 2000 (Invitrogen). Transfections (200 ng of each expression construct, 500 ng of each reporter construct) were performed in triplicate and repeated three times. A ß-galactosidase expression vector (2 ng) was cotransfected into cells as a control for transfection efficiency. Luciferase and ß-galactosidase activity were assessed 24 h after transfection using the Galacto-Star luciferase reporter assay (Roche) according to the manufacturer's instructions. The 0.6-kb PPAR2 promoter was the gift of Gökhan Hotamisligil (Harvard School of Public Health), while the 0.4-kb PPAR1 promoter was the gift of Jan Reddy (Northwestern University). Statistics were performed using the Student t test.

shRNA constructs. Target sequences for Ebf1, Ebf2, and Ebf3 were designed by querying the Whitehead small interfering RNA algorithm (http://jura.wi.mit.edu/bioc/siRNAext) as well as the small interfering RNA designer software from Clontech; at least three sequences represented by both algorithms were subcloned into the pSIREN RetroQ vector (Clontech) using the EcoRI and BamHI restriction sites. An effect had to be seen with at least two hairpins to be reported, although only data from the hairpin delivering the most efficient knockdown is shown. The best target sequences are as follows: Ebf1, GAACCCTGAAATGTGCCGA; Ebf2, CCGTTCTGTGCAGCTGGAAGGA; Ebf3, AATCACAACCAGATCCCCACC. Retroviral production and infection of 3T3-L1 cells was as described above. After selection, cells were grown to confluence and induced to differentiate as described above. Oil-red-O staining was performed at day 7. Each experiment was repeated three times.

EMSA. DNA probes were labeled with [-³²P]ATP by incubation with T4 polynucleotide kinase, annealed, and purified on a 6% polyacrylamide Tris-borate-EDTA gel. Recombinant in vitro-translated protein was incubated with labeled probe (20,000 cpm, 3 fmol) for 30 min at room temperature in binding buffer (10 mM HEPES, pH 7.9, 70 mM KCl, dithiothreitol, 1 mM EDTA, 1 mM ZnCl₂, 2.3 mM MgCl₂, 4% glycerol) with 0.75 µg of poly(dI/dC) per reaction mixture (Amersham). DNA competitors were added 10 min before the addition of the DNA probe. The samples were separated on a 6% polyacrylamide Tris-borate-EDTA gel, which was dried and subjected to autoradiography. Recombinant protein was generated by coupled in vitro transcription-translation using a reticulocyte lysate kit (Promega). A total of 0.5 µl of a 25-µl reaction mix was used for electrophoretic mobility shift assays (EMSAs). Oligonucleotides used for EMSA were as follows: mouse mb-1 promoter Ebf binding site forward, 5'-GAGAGAGACTCAAGGGAATTGTGG-3'; mouse mb-1 promoter Ebf binding site reverse, 5'-CCACAATTCCCTTGAGTCTCTCTC-3'; Ebf binding site mutated mouse mb-1 promoter forward, 5'-AGCCACCTCTCAGCCGTTTTGTGG-3'; Ebf binding site mutated mouse mb-1 promoter reverse, 5'-CCACAAAACGGCTGAGAGGTGGCT-3'; mouse PPAR1 Ebf binding site forward, 5'-CCTGTAACTCCAGCTCCAGGGGAGCCCACACCT-3'; mouse PPAR1 Ebf binding site reverse, 5'-AGGTGTGGGCTCCCCTGGAGCTGGAGTTACAGG-3'; mouse C/EBP Ebf binding site forward, 5'-CGTTTGGACACCAGGGGGCGATGCC-3'; mouse C/EBP Ebf binding site reverse, 5'-GGCATCGCCCCCTGGTGTCCAAACG-3'; mutated mouse C/EBP Ebf binding site forward, 5'-CGTTTGGACACCAAGAGGCGATGCC-3'; mutated mouse C/EBP Ebf binding site reverse, 5'-GGCATCGCCTCTTGGTGTCCAAACG-3'; mutated mouse PPAR1 binding site forward, 5'-CCTGTAACTCCAGCTCCAAGAGAGCCCACACCT-3; mutated mouse PPAR1 binding site reverse, 5'-AGGTGTGGGCTCTCTTGGAGCTGGAGTTACAGG-3'.

Diet-induced obesity. Eight- to 10-week-old male C57BL/6 mice were given a high-fat, high-carbohydrate (n = 11) or normal chow (n = 5) diet for 40 days, after which epididymal fat depots were isolated. Tissues were homogenized in Trizol (Invitrogen), and RNA was isolated according to the manufacturer's recommendations. Ebf levels were measured with Q-PCR and normalized to 36B4; results were analyzed with one-way analysis of variance.

Top Abstract Introduction Materials and Methods Results Discussion References

 
Ebf proteins are differentially regulated during adipogenesis. To characterize the timing of Ebf induction during adipogenesis, we performed a detailed time course examination in 3T3-L1 cells. Q-PCR-based expression analysis confirmed that Ebf1, 2, and 3 are expressed during adipogenesis, but to differing degrees, such that by day 7 postdifferentiation, one sees relative levels of Ebf1 > Ebf2 > Ebf3 (Fig. 1A). It is now clear that a number of factors, including CREB (51), Krox20 (7), and lipin 1 (30), are induced very early in the developmental cascade within the first few hours after DMI treatment. We looked at whether any of these Ebf isoforms were seen at these early time points and found that Ebf1 showed a profound increase in expression within 15 min of DMI treatment to levels equivalent to those seen in mature adipocytes. Ebf1 returned to baseline within 3 h but increased again gradually to reach a maximum after 7 days of differentiation. There was a tendency for a small induction in Ebf2 and Ebf3 levels immediately after DMI, but neither of these factors reached the same levels as Ebf1 during early or late time points. In general, Ebf2 increased during adipogenesis and reached a maximum after 7 days, while Ebf3 expression remained relatively low throughout the differentiation process. Levels of Ebf1 and Ebf2 rose before those of C/EBP, PPAR1, and PPAR2 (Fig. 1B), but the early spike in Ebf1 levels paralleled a similar rise in C/EBPß and C/EBP (Fig. 1C). Interestingly, Ebf1 shows a similar expression pattern when human preadipocytes are differentiated in vitro, suggesting a role for this factor in human fat development (Fig. 1D). In human fat, however, Ebf3 rises during differentiation while Ebf2 levels remain low, the inverse of the pattern seen in murine 3T3-L1 cells. We also looked at relative levels of Ebf expression in fat pads from adult mice. In mature adipocytes, Ebf1 levels were highest by far, followed by Ebf2 and then Ebf3 (data not shown), similar to what was observed in mature 3T3-L1 adipocytes. Ebf1 and Ebf3 were more highly expressed in adipocytes than in cells of the stromal-vascular fraction, which includes preadipocytes, while Ebf2 levels were high even in the stromal-vascular cells (Fig. 1E). This is consistent with elevated levels of Ebf2 seen in stromal cells from other sites such as bone marrow (M.S., unpublished results). Ebf isoforms were expressed at low levels in F4/80⁺ macrophages.

View larger version (28K): [in this window] [in a new window]

FIG. 1. Ebf isoforms are induced early and late during 3T3-L1 adipogenesis. (A) Levels of Ebf isoforms measured by Q-PCR in the first few hours (15 and 30 min and 1, 3, 6, and 12 h) after DMI treatment (left) and over the next several days (1 to 7 days, right). Data are means ± standard deviations (SD) (n = 3) and are expressed as inductions relative to Ebf1 at day 0. (B) C/EBP, PPAR2, PPAR1, aP2, and adiponectin expression in the same cells. Data are means ± SD (n = 3) and are expressed as inductions relative to day 0. (C) C/EBPß and expression in the first few hours (15 and 30 min and 1, 3, 6, and 12 h) after DMI treatment. Data are means ± SD (n = 3) and are expressed as inductions relative to day 0. (D) Human preadipocytes were cultivated in vitro and induced to differentiate for 15 days. RNA was collected at 0, 6, and 24 h and 4 and 15 days, and Ebf mRNA levels were measured by Q-PCR. Data are expressed as inductions relative to Ebf1 at day 0. (E) Expression of Ebf isoforms in isolated mouse mature adipocytes, F4/80+ macrophages, and remaining stromal-vascular cells. Data are expressed as means ± SD (n = 10) and are expressed as levels relative to SVF expression for each isoform. ***, P < 0.005 versus SVF.

Ectopic expression of Ebf proteins in NIH 3T3 cells promotes adipogenesis. We next sought to determine which Ebf proteins can promote fat differentiation in nonadipogenic NIH 3T3 fibroblasts. These cells were infected with retroviruses expressing Ebf1, Ebf2, Ebf3, PPAR2, or empty vector. After selection with puromycin, cells were grown to confluence and induced to differentiate with DMI. Seven days after induction, cells were stained for neutral lipid with oil-red-O (Fig. 2A). Adipocytes were easily detected in dishes containing cells expressing any Ebf isoform (representing 10 to 15% of all cells), while vector-infected cells exhibited no staining. As previously shown, cells overexpressing PPAR2 differentiate with very high efficiency (90%). Consistent with the oil-red-O staining, expression of fat-specific genes was elevated in Ebf-expressing cells (Fig. 2B).

View larger version (49K): [in this window] [in a new window]

FIG. 2. Ectopic expression of Ebf isoforms in NIH 3T3 cells promotes adipogenesis. (A) Micrographs of oil-red-O-stained NIH 3T3 cells 7 days after induction of differentiation with DMI. Cells were transduced with retroviruses expressing either Ebf1, Ebf2, Ebf3, PPAR2, or empty vector. (B) Adipsin and aP2 mRNA levels in these cells before (day 0) and after (day 7) DMI treatment as measured by Q-PCR. Data are expressed as means ± standard deviations (n = 3) and are expressed as inductions relative to the vector control at day 0. ***, P < 0.005 versus day 0.

PPAR1 and C/EBP are direct targets of Ebf1. PPAR is the most potent adipogenic transcription factor discovered to date and is required for fat cell differentiation in vitro and in vivo. We sought to determine whether Ebf proteins could induce adipogenesis in the absence of PPAR by using immortalized PPAR^–/– MEFs (32). Retrovirally mediated expression of Ebf1, Ebf2, or Ebf3 (see Fig. S2A in the supplemental material) was unable to induce differentiation in these cells, as measured by lipid accumulation (Fig. 3A) or by marker expression (see Fig. S2B in the supplemental material). In contrast, control fibroblasts containing a single intact PPAR allele (PPAR^flox/–) were differentiated by all three Ebf proteins. As expected, retroviral introduction of PPAR2 allowed robust differentiation of both cell types.

View larger version (94K): [in this window] [in a new window]

FIG.3. PPAR1 and C/EBP are direct targets of Ebf1. (A) Immortalized PPAR–/– and PPARflox/– MEFs were infected with retroviruses expressing either Ebf1, Ebf2, Ebf3, PPAR2, or empty vector. Cells were induced to differentiate with DMI and then stained with oil-red-O on day 7. (B) NIH 3T3 cells were cotransfected with the 0.4-kb PPAR1 (black bars) or empty pGL2 vector (white bars) along with an expression vector (pCDNA3) expressing the Ebf1, Ebf2, or Ebf3 cDNA. Cells were harvested 24 h after transfection. ***, P < 0.005. Data are expressed as inductions relative to the vector control (n = 9). (C) NIH 3T3 cells were cotransfected with the 0.3-kb C/EBP promoter (black bars) or empty pGL3 vector (white bars) along with an expression vector (pCDNA3) expressing the Ebf1 cDNA. Cells were harvested 24 h after transfection. Results are expressed in relative luciferase units (RLU), with inductions relative to that of the vector control. All results are expressed as means ± standard deviations (SD) (n = 6). *, P < 0.05; ***, P < 0.005 versus control (pCDNA3 vector). (D) Ebf binding sites from the mouse and human mb-1, mouse C/EBP, and mouse PPAR1 promoters (top). EMSA (bottom left) showing the binding of in vitro-translated Ebf1 to the mouse mb-1 promoter binding site (lane 2). Competition was performed using Ebf binding sites from the mouse mb-1 (mb-1 RE, lanes 3, 4), mouse C/EBP (C/EBP RE; lanes 7, 8), and mouse PPAR1 (PPAR1 RE, lanes 9, 10) promoters. A mutated Ebf binding site mb-1 promoter competitor (mb-1 M RE, lanes 5, 6) was used as a control. The indicated cold competitor oligonucleotides were added at 150x or 450x molar excess. F, free probe. EMSA (bottom right) showing competition of Ebf1 binding to the mouse mb-1 promoter (lane 1) using 150x or 450x molar excess of wild-type Ebf binding sites as well as mutated Ebf binding sites from mouse C/EBP (lanes 2 to 5) and PPAR1 (lanes 6 to 9) promoters. (E) Flag-tagged Ebf1 or vector was transduced into 3T3-L1 cells, and cross-linked DNA was prepared at day 0 and during differentiation (15 min, 30 min, 1 h, 3 h, 24 h, 48 h, 4 days, and 7 days post-DMI). ChIP assays were performed using anti-Flag antibody or an IgG control using PCR primers directed at regions of the C/EBP and PPAR1 promoters containing the putative Ebf sites. An upstream region (2 kb) from each promoter was used as a control. An additional control was performed by performing the immunoprecipitation with the anti-Flag antibody on cells transfected with empty vector; this showed no nonspecific activity of the Flag antibody (data not shown). (F) PPARflox/– cells were transduced with an Ebf1-expressing retrovirus, selected, and induced to differentiate with DMI. RNA was harvested at the indicated time points after DMI treatment, and Q-PCR was used to assess C/EBP mRNA levels. Data are expressed as inductions relative to the vector control. Data are expressed as means ± SD (n = 3). *, P < 0.05 versus vector only. Statistics were performed using one-way analysis of variance. (G) 3T3-L1 adipocytes were transfected with an Ebf1 expression construct or empty vector 5 days post-DMI. RNA was harvested 24 h later for Q-PCR-based expression analysis. Data are expressed as inductions relative to the vector control. Results are expressed as means ± SD (n = 6). *, P < 0.05; **, P < 0.01 versus vector.

To assess whether Ebf proteins directly induce PPAR expression, we looked at their effect of Ebf1 on the 0.6-kb PPAR2 promoter and the 0.4-kb PPAR1 promoter in transient-transfection assays. No effect was seen on the PPAR2 promoter construct (data not shown), but Ebf1 induced expression from the PPAR1 promoter 2.3-fold (Fig. 3B). We also looked at a possible effect of Ebf1 on the 0.3-kb C/EBP promoter. Ebf1 increased luciferase expression in this assay by 4.5-fold above baseline (Fig. 3C), demonstrating a potential direct transcriptional link between these two adipogenic factors. Putative Ebf binding sites within the C/EBP and PPAR1 promoters were identified and aligned with the consensus Ebf site as well as a well-characterized Ebf binding site from the mouse immunoglobulin -associated protein (mb-1) promoter (15, 35) (Fig. 3D). EMSA confirmed that these putative sites could effectively compete Ebf1 binding from the mb-1 oligonucleotide (Fig. 3D). Oligonucleotides bearing point mutations in the Ebf binding site of the C/EBP and PPAR1 promoters; however, do not compete for binding to the mb-1 promoter (Fig. 3D). To assess Ebf binding to the native promoters, we also performed ChIP assays in 3T3-L1 cells transduced with Flag-tagged Ebf1 (Fig. 3E). Ebf1 is bound to the PPAR1 and C/EBP promoters even prior to differentiation; this binding increases strongly by 1 h after DMI exposure before gradually decreasing at later time points. This peak at 1 h coincides with the period of maximum endogenous Ebf1 expression in 3T3-L1 differentiation (Fig. 1A), strongly suggesting that the interaction is functionally relevant. Further support for this notion was obtained by showing that retroviral introduction of Ebf1 into PPAR^flox/– fibroblasts causes a large induction in endogenous C/EBP levels relative to vector-infected control cells (Fig. 3F). This difference is seen both before and after induction of differentiation with DMI. We did not see a similar effect with PPAR1, which rose after DMI to an equivalent degree in vector- and Ebf1-infected cells (data not shown). Although the reason for this is unclear, we decided to test whether having a full complement of PPAR might enhance a direct effect of Ebf1 on PPAR1. We therefore transfected mature 3T3-L1 adipocytes with an Ebf1 expression plasmid and found enhanced expression of both endogenous C/EBP and PPAR1 mRNA (Fig. 3G).

Ebf1 induces C/EBP expression via C/EBP. Interestingly, levels of C/EBP and C/EBP were elevated by Ebf1 overexpression in both PPAR^flox/– and PPAR^–/– fibroblasts (Fig. 4A) but not in NIH 3T3 cells (Fig. 4B). NIH 3T3 cells express low levels of C/EBP during forced differentiation, in contrast to most other models of adipogenesis. This fact, in concert with the coincident rise of C/EBP and Ebf1 within the first hour after DMI (Fig. 1A and C), led us to consider whether a feedback loop might exist in which Ebf1 induces C/EBP levels in a C/EBP-dependent manner. C/EBP^–/– MEFs were employed to test whether genetic ablation of C/EBP could prevent Ebf1 from inducing C/EBP. These cells, along with C/EBP^+/+ control MEFs, were transduced with an Ebf1-expressing retrovirus (see Fig. S3A in the supplemental material) and harvested just prior to the induction of differentiation (i.e., on day 0). C/EBP was induced by Ebf1 in C/EBP^+/+ control cells but not in C/EBP^–/– cells (Fig. 4C). The same pattern was seen when Ebf2 or Ebf3 was expressed in these cells (data not shown). C/EBPß was not induced by Ebf1 in any cellular context. Taken together, these data strongly suggest a cascade in which Ebf1 might participate in the very early stages of adipogenesis by promoting C/EBP expression via C/EBP. Surprisingly, the total absence of C/EBP did not prevent fat cell formation in the presence of Ebf1, 2, or 3, as measured by lipid accumulation (Fig. 4D) or by marker expression (see Fig. S3B in the supplemental material). This suggests that Ebf proteins, at least in the setting of ectopic overexpression, can promote adipogenesis without inducing C/EBP through C/EBP.

View larger version (78K): [in this window] [in a new window]

FIG.4. Ebf1 induces C/EBP expression in a C/EBP-dependent manner, but does not require C/EBP to promote adipogenesis. (A) C/EBP isoform expression in PPARflox/– and PPAR–/– cells transduced with an Ebf1-expressing retrovirus or empty vector. RNA was harvested at day 0, prior to the addition of DMI. Data are expressed as inductions relative to vector control. Data are expressed as means ± standard deviations (SD) (n = 3). **, P < 0.01; ***, P < 0.005 versus vector. (B) C/EBP isoform expression in NIH 3T3 cells transduced with an Ebf1-expressing retrovirus or empty vector. RNA was harvested at day 0, prior to the addition of DMI. Data are expressed as inductions relative to vector control. Data are expressed as means ± SD (n = 3). (C) C/EBP isoform expression in C/EBP+/+ and C/EBP–/– cells transduced with an Ebf1-expressing retrovirus or empty vector. RNA was harvested at day 0, prior to the addition of DMI. Data are expressed as inductions relative to vector control. Data are expressed as means ± SD (n = 3). ***, P < 0.005 versus vector. (D) C/EBP+/+ and C/EBP–/– cells transduced with empty retrovirus or virus expressing PPAR2, Ebf1, Ebf2, or Ebf3 were differentiated (DMI) for 7 days and stained with oil-red-O.

C/EBPß and C/EBP promote the induction of Ebf1 during adipogenesis. C/EBPß and are among the earliest actors in the adipogenic transcriptional cascade. We noted that the sharp early rise in Ebf1 was contemporaneous with a spike in these two factors, and the more gradual elevation in Ebf1 and Ebf2 that occurred over the first few days lies between the peak expression of C/EBPß and on the one hand and PPAR and C/EBP on the other (Fig. 1). We thus were interested in the possibility that C/EBPß and/or could be a direct inducer of Ebf expression. To address this, we first tested whether C/EBP proteins could induce expression from an Ebf1 promoter-luciferase construct. We found that C/EBP, ß, and were all able to induce expression of this construct (Fig. 5A). We then identified three potential C/EBP binding elements in this construct using TRANSFAC, designated in Fig. 5A as sites A, B, and C. A combination of deletion mapping and transactivation analysis was employed to assess which of these sites might be required for C/EBP action on the Ebf1 promoter. Removal of site A reduced the ability of C/EBP and to induce the reporter but did not affect the ability of C/EBPß to do so. Subsequent removal of site B did not affect transactivation in response to any of the C/EBPs, but removal of site C (in addition to A and B) prevented all C/EBP action on the basal promoter. These results suggest that all three C/EBP proteins can transactivate the C site but that the A site is transactivated only by C/EBP or .

View larger version (53K): [in this window] [in a new window]

FIG. 5. C/EBPß and promote induction of Ebf1. (A) Deletion analysis of Ebf1 promoter was performed to identify potential C/EBP sites. Three putative binding sites for C/EBPs (black boxes A, B, C) and two previously identified transcriptional start sites (TS1, TS2) are shown. The Ebf1 promoter construct (black bars) or empty vector (pGL3; white bars) was cotransfected into NIH 3T3 cells with an expression vector containing C/EBP, ß or cDNA, or empty vector (pCDNA3). Cells were harvested 24 h after transfection. Relative inductions (Ebf1 promoter/pGL3) are shown in parentheses above each bar. Data are expressed as inductions relative to that of the vector control. Results are expressed as relative luciferase units (RLU) and means ± standard deviations (n = 9). ***, P < 0.005. (B) ChIP assay on 3T3-L1 cells collected at various time points (day 0, 1, 2, 4, and 7) before and after DMI treatment using antibodies against C/EBP, ß, or and control IgG. The three putative C/EBP sites in the Ebf1 promoter from above (A, B, C) were queried along with an upstream site (2 kb) as a control.

To confirm that C/EBP proteins occupy the Ebf1 promoter during adipogenesis, we performed ChIP assays at various time points during 3T3-L1 differentiation (Fig. 5B). C/EBPß binds site A from day 1 through day 4 and is thereafter replaced by C/EBP, whereas C/EBP does not bind site A. In contrast, site B does not seem to be regulated because all C/EBP proteins occupy this site throughout the differentiation process. Similarly, binding of site C by C/EBP and is not regulated. No binding was seen to a site 2 kb upstream of site A.

Ebf1 and Ebf2, but not Ebf3, are necessary for adipogenesis. Three Ebf proteins induced adipogenesis in at least four independently derived cell lines (NIH 3T3, PPAR^flox/– MEFs, C/EBP^+/+ MEFs, and C/EBP^–/– MEFs). These results could imply that each Ebf isoform exerts a nonredundant action on differentiation. Alternatively, it could be the case that only one or two proteins are required but that any isoform can drive the process forward when overexpressed because of sufficient similarity in their DNA binding activity. To distinguish between these scenarios, we used retroviral delivery of shRNAs to knock down individual Ebf proteins. We first confirmed the efficacy and specificity of each hairpin by cotransfecting NIH 3T3 cells with Flag-tagged Ebf1, Ebf2, or Ebf3 along with the retroviral vector containing the shRNA against each isoform. The absence of adequate antibodies against individual Ebf proteins made this approach necessary. Twenty-four hours after transfection, cell lysates were harvested and analyzed by immunoblotting with an anti-Flag antibody. As shown in Fig. S4 in the supplemental material, we were able to identify hairpins that knocked down individual Ebf proteins at the protein level and that exhibited no cross-reactivity with other Ebfs. We next infected 3T3-L1 preadipocytes with the shRNA-expressing retroviruses. Transduced cells were selected and expanded, and expression of the endogenous Ebf genes was determined by quantitative real-time PCR on the day differentiation was induced. As seen in Fig. 6A, high-efficiency and highly selective knockdown was achieved in these cells as well. At 7 days postdifferentiation, cells were either stained with oil-Red-O or RNA was harvested. As expected, 3T3-L1 cells infected with a control shRNA are able to differentiate fully into adipocytes. In contrast, knockdown of either Ebf1 (Fig. 6B) or Ebf2 (Fig. 6C) completely abolished the differentiation process. Cells were cultured for more than 10 days to confirm that differentiation was not simply delayed. Interestingly, knockdown of Ebf3 did not have any effect on the number and appearance of fat cells relative to control cells (Fig. 6D). These results were confirmed by measuring adipose-selective markers like aP2 and adipsin at day 7. We thus demonstrate a requirement for Ebf1 and Ebf2 in adipogenesis and suggest that these two proteins play nonredundant roles in the differentiation process. Ebf3, in contrast, seems to be less important, consistent with the fact that endogenous levels of this isoform are hardly induced by differentiation at all in these cells (Fig. 1A).

View larger version (36K): [in this window] [in a new window]

FIG. 6. Ebf1 and Ebf2 are required for adipogenesis. (A) 3T3-L1 preadipocytes were transduced with constructs containing shRNAs directed against luciferase (control), Ebf1, Ebf2, or Ebf3. Levels of Ebf1, Ebf2, and Ebf3 mRNA were measured by Q-PCR prior to differentiation. Data are expressed as repression relative to the vector control (n = 3). ***, P < 0.001. (B) 3T3-L1 preadipocytes were transduced with retroviruses expressing control shRNA or Ebf1 shRNA, selected, and differentiated with DMI; oil-red-O staining was performed at day 7. Levels of adipsin and aP2 mRNA at day 7 are shown at the right. Data are expressed as inductions relative to vector control. Results are expressed as means ± standard deviations (SD) (n = 3). ***, P < 0.005 versus control. (C) 3T3-L1 cells treated as described for panel B, except that Ebf2 was targeted by shRNA. Oil-red-O staining was performed at day 7. Levels of adipsin and aP2 mRNA at day 7 are shown at the right. Data are expressed as inductions relative to the vector control. Results are expressed as means ± SD (n = 3). ***, P < 0.005 versus control. (D) 3T3-L1 cells infected as described for panel B, except that Ebf3 was targeted. Oil-red-O staining was performed at day 7. Levels of adipsin and aP2 mRNA at day 7 are shown at the right. Data are expressed as inductions relative to the vector control. Results are expressed as means ± SD (n = 3). ***, P < 0.005 versus control.

Ebf expression is dysregulated in obesity. The importance of Ebf proteins for adipocyte differentiation in vitro led us to investigate whether levels of these factors are dysregulated in obesity. As shown in Fig. 7, Ebf1, 2, and 3 mRNAs are highly induced in the white adipose tissue of mice on a high-fat, high-carbohydrate diet relative to lean controls.

View larger version (8K): [in this window] [in a new window]

FIG. 7. Ebf1, 2, and 3 mRNA levels are induced in white adipose tissue in mice on a high-fat, high-carbohydrate (HF/HC) diet. After 40 days of chow (n = 5) or HF/HC (n = 11) diet, mRNA was isolated from epididymal fat pads from male C57BL/6 mice and analyzed by Q-PCR. Results are displayed as means ± standard errors of the means. ***, P < 0.001; **, P < 0.01 versus chow. Data are expressed as inductions relative to Ebf1 in chow-fed animals.

Top Abstract Introduction Materials and Methods Results Discussion References

 
Cell fates are determined by a complex series of transcriptional events that promote terminal differentiation and concomitantly suppress alternate possible lineages. In the case of adipogenesis, extensive work from a number of groups has identified a key role for several transcription factors, most notably PPAR and the C/EBP proteins , ß, and (34). Based on these studies, a model has been proposed wherein C/EBPß and are among the earliest factors to respond to adipogenic stimuli in a committed precursor cell, and these in turn induce expression of PPAR and C/EBP (Fig. 8) (12, 34, 48). These latter factors appear to mutually reinforce each other's expression, and both promote the establishment of the ultimate terminally differentiated phenotype. PPAR and C/EBP are not equals in this process, however, as ectopic expression of PPAR in C/EBP null fibroblasts rescues adipogenesis, while the converse is not true (32, 49).

View larger version (12K): [in this window] [in a new window]

FIG. 8. Model of adipogenesis. A model placing Ebf1 into the context of the known adipogenic transcriptional cascade. See the text for details.

In the last few years, several new transcription factors have been discovered to play a role in adipogenesis. Ebf1 was also recently shown to play a positive role in this process (1), although it has not been clear in which part of the cascade Ebf1 acts. Furthermore, other Ebf family members, including Ebf2 and Ebf3, are expressed in adipose tissue but had not been tested for a potential role in the developmental process.

We have addressed these issues using a variety of gain-of-function and loss-of-function approaches in multiple adipogenic and nonadipogenic cell types. Our results indicate that there is a short but intense burst of Ebf1 expression within hours of adding the proadipogenic cocktail, followed by a return to a baseline and a more gradual sustained rise that appears to be maintained for the life of the mature adipocyte. Ebf2 and Ebf3 do not display this sharp rise in the early hours of differentiation, and although they do rise gradually later in development, they never reach peak levels equivalent to that of Ebf1.

These factors promote adipogenesis in NIH 3T3 fibroblasts, placing the Ebf family in a fairly elite class of factors that have this activity. Given the high degree of structural similarity within the Ebf family, however, it was not clear if the three factors tested were acting in a generic "Ebf-like" way when overexpressed or whether each factor had a unique function. To address this, we employed retroviral delivery of shRNA to specifically "knock down" each Ebf isoform in turn. To our surprise, differentiation did not proceed in the presence of reduced levels of either Ebf1 or Ebf2, strongly suggesting that these two factors possess nonredundant activities. Specific Ebf3 "knockdown" did not result in changes in lipid accumulation or expression of marker genes like aP2. These experiments establish the importance of Ebf1 and Ebf2 and inform us that these proteins play a role in the process that cannot be adequately compensated for by other members of the family.

Where can we place Ebf action in the adipogenic cascade? Temporally, they appear at or after the rise in C/EBPß and , suggesting that they might themselves be induced by these factors. This was made clear for Ebf1 in particular by the demonstration of C/EBP binding sites in the proximal promoter. ChIP analysis suggests that C/EBPs may occupy the site A in a differentiation-dependent manner. In contrast, sites B and C are occupied by C/EBPs even prior to the induction of differentiation, suggesting an additional permissive role for C/EBPs in the induction of Ebf1, in which binding to some sites is required but not sufficient for induction of Ebf1 message levels. Alternatively, the C/EBPs occupying site B and C might be regulated by posttranslational modification that alters their capacity to activate the Ebf1 promoter during adipogenesis.

Furthermore, our data suggest that Ebfs promote adipogenesis, at least in part, by inducing expression of C/EBP and PPAR1. The evidence that these factors are direct targets of Ebf1 include (i) the presence of Ebf motifs in the PPAR1 and C/EBP promoters, both of which bind Ebf1 in vitro, (ii) the ability of these sites to enable reporter gene transactivation by Ebf1 in a transient-transfection assay, (iii) direct binding of Ebf1 to these sites in the native promoter in intact cells in a differentiation-dependent manner, (iv) enhancement of endogenous PPAR1 and C/EBP expression in adipocytes transfected with Ebf1, and (v) a requirement for PPAR in the adipogenic action of Ebf1.

Our experiments place Ebf action between C/EBPß/ and C/EBP/PPAR, although they do not rule out further downstream actions of Ebf proteins that might also be critical for differentiation. In fact, the ability of Ebf proteins to induce adipogenesis in the absence of C/EBP (as shown in C/EBP^–/– MEFs as well as in NIH 3T3 cells, which express very low levels of C/EBP) suggests that other targets are likely to exist. This is reminiscent of the situation in which ectopic addition of PPAR allows differentiation in the absence of C/EBP and supports the notion that C/EBP is more ancillary to the process than PPAR. Adipocytes differentiated in the absence of C/EBP accumulate lipid and most markers of the fat phenotype but are not insulin sensitive (13, 34). This is the result of improper Glut4 vesicle trafficking in C/EBP null adipocytes (46). We do not know yet if this will also be the case in C/EBP null adipocytes differentiated by ectopic overexpression of Ebf proteins.

Our data also identify a positive feedback loop involving C/EBP, Ebf1, and C/EBP. This was initially suggested by the fact that Ebf1 could induce expression of C/EBP in PPAR^flox/– and PPAR^–/– cells but not in NIH 3T3 cells, which do not express C/EBP. The requirement for C/EBP was confirmed by showing that Ebf1 induces C/EBP in C/EBP^+/+ MEFs but not in MEFs lacking C/EBP. Ebf1 likely exerts this effect by direct transactivation of the C/EBP promoter. We do note, however, that the Ebf1-C/EBP-C/EBP-positive feedback loop is not absolutely required for adipogenesis, as Ebf1 is perfectly capable of causing fat cell formation in the absence of C/EBP. We postulate that this feedback loop can act as an accelerant during the early phases of differentiation in normal cells.

These experiments demonstrate a critical role for several different Ebf family members in adipogenesis. Ebf1 and Ebf2 in particular are shown to be both necessary and sufficient for adipogenesis in vitro. The situation with Ebf3 is less clear, in that it promotes differentiation quite well when overexpressed, but its absence has no significant effect on terminal differentiation. Ebf2 and Ebf3 are extraordinarily highly conserved (>85% at the amino acid level), with Ebf2 possessing two short novel sequences not seen in Ebf3, including one in the C-terminal transactivation domain. One possible explanation for our results is that Ebf2 can compensate for the loss of Ebf3, but the opposite is not true because of effects mediated by the Ebf2 unique C-terminal sequence. Further studies involving domain swapping will be required to clarify this point.

Much remains to be learned about Ebf isoform action in adipocyte biology. A key question centers on additional target genes of Ebf proteins during development and in the mature adipocyte. Other developmentally relevant targets are likely, as discussed above, and could include other known adipogenic transcription factors besides PPAR1 and C/EBP. PPAR2 may be such a target. Our studies revealed no Ebf binding sites in the 0.6-kb PPAR2 promoter, but certainly sites could exist outside of that region. Ebf proteins may also affect antiadipogenic pathways. Fig. S5 in the supplemental material shows the effect of Ebf1 to 3 on the mRNA level of several factors known to repress adipogenesis in transduced NIH 3T3 cells prior to differentiation. The ability of Ebf1 and Ebf2 to repress GATA2 is certainly interesting in this regard.

It is noteworthy that a statistical analysis of overrepresented motifs in the 1-kb promoter regions from 30 adipose-selective genes identified variants of the Ebf consensus sequence as the top-rated hit (data not shown), suggesting that these factors may play a broad role in establishing and maintaining the mature adipose phenotype. Furthermore, the fact that Ebf1, 2, and 3 are elevated in adipose tissue from obese mice suggests a possible role for these proteins in the pathological behavior of adipocytes in the setting of overnutrition. It will be interesting to look at Ebf expression and function in other obese and diabetic models. Ebf1 null mice have been generated and have been reported to be smaller than wild-type littermates, but no specific measurements of body fat or other metabolic parameters were commented upon (14, 24). Ebf3 null animals exhibit perinatal lethality, while Ebf2 null mice have reduced survival and significant neuronal defects (9, 44). We are in the process of characterizing these and other models for metabolic and adipose-related phenotypes. Given the clinical success of PPAR-based therapies, we will be interested to see if manipulation of Ebf proteins might provide additional avenues in the treatment of metabolic disease.

 
This work was supported by NIH grant DK63906 to E.D.R., Swiss National Foundation award PA00A-105047 to M.A.J., and funds from Astra Zeneca.

We thank James Kirkland at the Adipocyte Core, Boston Obesity Nutrition Research Center, NIH grant DK 46200, for the human adipocyte RNA samples; Karen Inouye for cDNA from isolated murine adipocytes, macrophages, and stromal vascular cells; and Simonetta Westerlund and Magdalen Rhedin for analysis of EBF expression in mice. We also thank members of the Rosen lab for thoughtful discussion.

 
* Corresponding author. Mailing address: Division of Endocrinology, Beth Israel Deaconess Medical Center, Boston, MA 02215. Phone: (617) 667-3221. Fax: (617) 667-2927. E-mail: erosen{at}bidmc.harvard.edu.

Published ahead of print on 23 October 2006.

Supplemental material for this article may be found at http://mcb.asm.org/.

Top Abstract Introduction Materials and Methods Results Discussion References

 

1
1. Akerblad, P., U. Lind, D. Liberg, K. Bamberg, and M. Sigvardsson. 2002. Early B-cell factor (O/E-1) is a promoter of adipogenesis and involved in control of genes important for terminal adipocyte differentiation. Mol. Cell. Biol. 22:8015-8025.[Abstract/Free Full Text]
2
2. Akerblad, P., R. Mansson, A. Lagergren, S. Westerlund, B. Basta, U. Lind, A. Thelin, R. Gisler, D. Liberg, S. Nelander, K. Bamberg, and M. Sigvardsson. 2005. Gene expression analysis suggests that EBF-1 and PPARgamma2 induce adipogenesis of NIH-3T3 cells with similar efficiency and kinetics. Physiol. Genomics 23:206-216.[Abstract/Free Full Text]
3
3. Banerjee, S. S., M. W. Feinberg, M. Watanabe, S. Gray, R. L. Haspel, D. J. Denkinger, R. Kawahara, H. Hauner, and M. K. Jain. 2003. The Kruppel-like factor KLF2 inhibits peroxisome proliferator-activated receptor-gamma expression and adipogenesis. J. Biol. Chem. 278:2581-2584.[Abstract/Free Full Text]
4
4. Barak, Y., M. C. Nelson, E. S. Ong, Y. Z. Jones, P. Ruiz-Lozano, K. R. Chien, A. Koder, and R. M. Evans. 1999. PPAR gamma is required for placental, cardiac, and adipose tissue development. Mol. Cell 4:585-595.[CrossRef][Medline]
5
5. Buckley, P. J., M. R. Smith, M. F. Braverman, and S. A. Dickson. 1987. Human spleen contains phenotypic subsets of macrophages and dendritic cells that occupy discrete microanatomic locations. Am. J. Pathol. 128:505-520.[Abstract]
6
6. Cao, Z., R. M. Umek, and S. L. McKnight. 1991. Regulated expression of three C/EBP isoforms during adipose conversion of 3T3-L1 cells. Genes Dev. 5:1538-1552.[Abstract/Free Full Text]
7
7. Chen, Z., J. I. Torrens, A. Anand, B. M. Spiegelman, and J. M. Friedman. 2005. Krox20 stimulates adipogenesis via C/EBPbeta-dependent and -independent mechanisms. Cell Metab. 1:93-106.[CrossRef][Medline]
8
8. Choy, L., and R. Derynck. 2003. Transforming growth factor-beta inhibits adipocyte differentiation by Smad3 interacting with CCAAT/enhancer-binding protein (C/EBP) and repressing C/EBP transactivation function. J. Biol. Chem. 278:9609-9619.[Abstract/Free Full Text]
9
9. Corradi, A., L. Croci, V. Broccoli, S. Zecchini, S. Previtali, W. Wurst, S. Amadio, R. Maggi, A. Quattrini, and G. G. Consalez. 2003. Hypogonadotropic hypogonadism and peripheral neuropathy in Ebf2-null mice. Development 130:401-410.[Abstract/Free Full Text]
10
10. Cousin, B., O. Munoz, M. Andre, A. M. Fontanilles, C. Dani, J. L. Cousin, P. Laharrague, L. Casteilla, and L. Penicaud. 1999. A role for preadipocytes as macrophage-like cells. FASEB J. 13:305-312.[Abstract/Free Full Text]
11
11. Dubois, L., and A. Vincent. 2001. The COE-Collier/Olf1/EBF-transcription factors: structural conservation and diversity of developmental functions. Mech. Dev. 108:3-12.[CrossRef][Medline]
12
12. Elberg, G., J. M. Gimble, and S. Y. Tsai. 2000. Modulation of the murine peroxisome proliferator-activated receptor gamma 2 promoter activity by CCAAT/enhancer-binding proteins. J. Biol. Chem. 275:27815-27822.[Abstract/Free Full Text]
13
13. El-Jack, A. K., J. K. Hamm, P. F. Pilch, and S. R. Farmer. 1999. Reconstitution of insulin-sensitive glucose transport in fibroblasts requires expression of both PPARgamma and C/EBPalpha. J. Biol. Chem. 274:7946-7951.[Abstract/Free Full Text]
14
14. Garel, S., K. Yun, R. Grosschedl, and J. L. Rubenstein. 2002. The early topography of thalamocortical projections is shifted in Ebf1 and Dlx1/2 mutant mice. Development 129:5621-5634.
15
15. Gisler, R., S. E. Jacobsen, and M. Sigvardsson. 2000. Cloning of human early B-cell factor and identification of target genes suggest a conserved role in B-cell development in man and mouse. Blood 96:1457-1464.[Abstract/Free Full Text]
16
16. Green, H., and O. Kehinde. 1975. An established preadipose cell line and its differentiation in culture. II. Factors affecting the adipose conversion. Cell 5:19-27.[Medline]
17
17. Green, H., and O. Kehinde. 1976. Spontaneous heritable changes leading to increased adipose conversion in 3T3 cells. Cell 7:105-113.[CrossRef][Medline]
18
18. Hagman, J., C. Belanger, A. Travis, C. W. Turck, and R. Grosschedl. 1993. Cloning and functional characterization of early B-cell factor, a regulator of lymphocyte-specific gene expression. Genes Dev. 7:760-773.[Abstract/Free Full Text]
19
19. Hagman, J., M. J. Gutch, H. Lin, and R. Grosschedl. 1995. EBF contains a novel zinc coordination motif and multiple dimerization and transcriptional activation domains. EMBO J. 14:2907-2916.[Medline]
20
20. Jimenez, M., C. Yvon, L. Lehr, B. Leger, P. Keller, A. Russell, F. Kuhne, P. Flandin, J. P. Giacobino, and P. Muzzin. 2002. Expression of uncoupling protein-3 in subsarcolemmal and intermyofibrillar mitochondria of various mouse muscle types and its modulation by fasting. Eur. J. Biochem. 269:2878-2884.[Medline]
21
21. Kennell, J. A., E. E. O'Leary, B. M. Gummow, G. D. Hammer, and O. A. MacDougald. 2003. T-cell factor 4N (TCF-4N), a novel isoform of mouse TCF-4, synergizes with beta-catenin to coactivate C/EBPalpha and steroidogenic factor 1 transcription factors. Mol. Cell. Biol. 23:5366-5375.[Abstract/Free Full Text]
22
22. Kieslinger, M., S. Folberth, G. Dobreva, T. Dorn, L. Croci, R. Erben, G. G. Consalez, and R. Grosschedl. 2005. EBF2 regulates osteoblast-dependent differentiation of osteoclasts. Dev. Cell 9:757-767.[CrossRef][Medline]
23
23. Liberg, D., M. Sigvardsson, and P. Akerblad. 2002. The EBF/Olf/Collier family of transcription factors: regulators of differentiation in cells originating from all three embryonal germ layers. Mol. Cell. Biol. 22:8389-8397.[Free Full Text]
24
24. Lin, H., and R. Grosschedl. 1995. Failure of B-cell differentiation in mice lacking the transcription factor EBF. Nature 376:263-267.[CrossRef][Medline]
25
25. McKnight, A. J., A. J. Macfarlane, P. Dri, L. Turley, A. C. Willis, and S. Gordon. 1996. Molecular cloning of F4/80, a murine macrophage-restricted cell surface glycoprotein with homology to the G-protein-linked transmembrane 7 hormone receptor family. J. Biol. Chem. 271:486-489.[Abstract/Free Full Text]
26
26. Mokdad, A. H., E. S. Ford, B. A. Bowman, W. H. Dietz, F. Vinicor, V. S. Bales, and J. S. Marks. 2003. Prevalence of obesity, diabetes, and obesity-related health risk factors, 2001. JAMA 289:76-79.[Abstract/Free Full Text]
27
27. Mori, T., H. Sakaue, H. Iguchi, H. Gomi, Y. Okada, Y. Takashima, K. Nakamura, T. Nakamura, T. Yamauchi, N. Kubota, T. Kadowaki, Y. Matsuki, W. Ogawa, R. Hiramatsu, and M. Kasuga. 2005. Role of Kruppel-like factor 15 (KLF15) in transcriptional regulation of adipogenesis. J. Biol. Chem. 280:12867-12875.[Abstract/Free Full Text]
28
28. Nakae, J., T. Kitamura, Y. Kitamura, W. H. Biggs III, K. C. Arden, and D. Accili. 2003. The forkhead transcription factor Foxo1 regulates adipocyte differentiation. Dev. Cell 4:119-129.[CrossRef][Medline]
29
29. Oishi, Y., I. Manabe, K. Tobe, K. Tsushima, T. Shindo, K. Fujiu, G. Nishimura, K. Maemura, T. Yamauchi, N. Kubota, R. Suzuki, T. Kitamura, S. Akira, T. Kadowaki, and R. Nagai. 2005. Kruppel-like transcription factor KLF5 is a key regulator of adipocyte differentiation. Cell Metab. 1:27-39.[CrossRef][Medline]
30
30. Phan, J., M. Peterfy, and K. Reue. 2004. Lipin expression preceding peroxisome proliferator-activated receptor-gamma is critical for adipogenesis in vivo and in vitro. J. Biol. Chem. 279:29558-29564.[Abstract/Free Full Text]
31
31. Pittenger, M. F., A. M. Mackay, S. C. Beck, R. K. Jaiswal, R. Douglas, J. D. Mosca, M. A. Moorman, D. W. Simonetti, S. Craig, and D. R. Marshak. 1999. Multilineage potential of adult human mesenchymal stem cells. Science 284:143-147.[Abstract/Free Full Text]
32
32. Rosen, E. D., C. H. Hsu, X. Wang, S. Sakai, M. W. Freeman, F. J. Gonzalez, and B. M. Spiegelman. 2002. C/EBPalpha induces adipogenesis through PPARgamma: a unified pathway. Genes Dev. 16:22-26.[Abstract/Free Full Text]
33
33. Rosen, E. D., P. Sarraf, A. E. Troy, G. Bradwin, K. Moore, D. S. Milstone, B. M. Spiegelman, and R. M. Mortensen. 1999. PPAR gamma is required for the differentiation of adipose tissue in vivo and in vitro. Mol. Cell 4:611-617.[CrossRef][Medline]
34
34. Rosen, E. D., C. J. Walkey, P. Puigserver, and B. M. Spiegelman. 2000. Transcriptional regulation of adipogenesis. Genes Dev. 14:1293-1307.[Free Full Text]
35
35. Sigvardsson, M., D. R. Clark, D. Fitzsimmons, M. Doyle, P. Akerblad, T. Breslin, S. Bilke, R. Li, C. Yeamans, G. Zhang, and J. Hagman. 2002. Early B-cell factor, E2A, and Pax-5 cooperate to activate the early B cell-specific mb-1 promoter. Mol. Cell. Biol. 22:8539-8551.[Abstract/Free Full Text]
36
36. Smith, E. M., R. Gisler, and M. Sigvardsson. 2002. Cloning and characterization of a promoter flanking the early B cell factor (EBF) gene indicates roles for E-proteins and autoregulation in the control of EBF expression. J. Immunol. 169:261-270.[Abstract/Free Full Text]
37
37. Strauss, R. S., and H. A. Pollack. 2001. Epidemic increase in childhood overweight, 1986-1998. JAMA 286:2845-2848.[Abstract/Free Full Text]
38
38. Tchkonia, T., Y. D. Tchoukalova, N. Giorgadze, T. Pirtskhalava, I. Karagiannides, R. A. Forse, A. Koo, M. Stevenson, D. Chinnappan, A. Cartwright, M. D. Jensen, and J. L. Kirkland. 2005. Abundance of two human preadipocyte subtypes with distinct capacities for replication, adipogenesis, and apoptosis varies among fat depots. Am. J. Physiol. Endocrinol. Metab. 288:E267-E277.[Abstract/Free Full Text]
39
39. Tong, Q., G. Dalgin, H. Xu, C. N. Ting, J. M. Leiden, and G. S. Hotamisligil. 2000. Function of GATA transcription factors in preadipocyte-adipocyte transition. Science 290:134-138.[Abstract/Free Full Text]
40
40. Tong, Q., J. Tsai, G. Tan, G. Dalgin, and G. S. Hotamisligil. 2005. Interaction between GATA and the C/EBP family of transcription factors is critical in GATA-mediated suppression of adipocyte differentiation. Mol. Cell. Biol. 25:706-715.[Abstract/Free Full Text]
41
41. Tontonoz, P., E. Hu, and B. M. Spiegelman. 1994. Stimulation of adipogenesis in fibroblasts by PPAR gamma 2, a lipid-activated transcription factor. Cell 79:1147-1156.[CrossRef][Medline]
42
42. Wang, M. M., and R. R. Reed. 1993. Molecular cloning of the olfactory neuronal transcription factor Olf-1 by genetic selection in yeast. Nature 364:121-126.[CrossRef][Medline]
43
43. Wang, S. S., A. G. Betz, and R. R. Reed. 2002. Cloning of a novel Olf-1/EBF-like gene, O/E-4, by degenerate oligo-based direct selection. Mol. Cell. Neurosci. 20:404-414.[CrossRef][Medline]
44
44. Wang, S. S., J. W. Lewcock, P. Feinstein, P. Mombaerts, and R. R. Reed. 2004. Genetic disruptions of O/E2 and O/E3 genes reveal involvement in olfactory receptor neuron projection. Development 131:1377-1388.[Abstract/Free Full Text]
45
45. Wang, S. S., R. Y. Tsai, and R. R. Reed. 1997. The characterization of the Olf-1/EBF-like HLH transcription factor family: implications in olfactory gene regulation and neuronal development. J. Neurosci. 17:4149-4158.[Abstract/Free Full Text]
46
46. Wertheim, N., Z. Cai, and T. E. McGraw. 2004. The transcription factor CCAAT/enhancer-binding protein alpha is required for the intracellular retention of GLUT4. J. Biol. Chem. 279:41468-41476.[Abstract/Free Full Text]
47
47. Wolfrum, C., D. Q. Shih, S. Kuwajima, A. W. Norris, C. R. Kahn, and M. Stoffel. 2003. Role of Foxa-2 in adipocyte metabolism and differentiation. J. Clin. Investig. 112:345-356.[CrossRef][Medline]
48
48. Wu, Z., N. L. Bucher, and S. R. Farmer. 1996. Induction of peroxisome proliferator-activated receptor gamma during the conversion of 3T3 fibroblasts into adipocytes is mediated by C/EBPbeta, C/EBPdelta, and glucocorticoids. Mol. Cell. Biol. 16:4128-4136.[Abstract]
49
49. Wu, Z., E. D. Rosen, R. Brun, S. Hauser, G. Adelmant, A. E. Troy, C. McKeon, G. J. Darlington, and B. M. Spiegelman. 1999. Cross-regulation of C/EBP alpha and PPAR gamma controls the transcriptional pathway of adipogenesis and insulin sensitivity. Mol. Cell 3:151-158.[CrossRef][Medline]
50
50. Yeh, W. C., Z. Cao, M. Classon, and S. L. McKnight. 1995. Cascade regulation of terminal adipocyte differentiation by three members of the C/EBP family of leucine zipper proteins. Genes Dev. 9:168-181.[Abstract/Free Full Text]
51
51. Zhang, J. W., D. J. Klemm, C. Vinson, and M. D. Lane. 2004. Role of CREB in transcriptional regulation of CCAAT/enhancer-binding protein beta gene during adipogenesis. J. Biol. Chem. 279:4471-4478.[Abstract/Free Full Text]

---

Molecular and Cellular Biology, January 2007, p. 743-757, Vol. 27, No. 2
0270-7306/07/$08.00+0     doi:10.1128/MCB.01557-06
Copyright © 2007, American Society for Microbiology. All Rights Reserved.

This article has been cited by other articles:

• Gerin, I., Bommer, G. T., Lidell, M. E., Cederberg, A., Enerback, S., MacDougald, O. A. (2009). On the Role of FOX Transcription Factors in Adipocyte Differentiation and Insulin-stimulated Glucose Uptake. J. Biol. Chem. 284: 10755-10763 [Abstract] [Full Text]  
• Badri, K. R., Zhou, Y., Dhru, U., Aramgam, S., Schuger, L. (2008). Effects of the SANT Domain of Tension-Induced/Inhibited Proteins (TIPs), Novel Partners of the Histone Acetyltransferase p300, on p300 Activity and TIP-6-Induced Adipogenesis. Mol. Cell. Biol. 28: 6358-6372 [Abstract] [Full Text]  
• Xu, Z., Yu, S., Hsu, C.-H., Eguchi, J., Rosen, E. D. (2008). The orphan nuclear receptor chicken ovalbumin upstream promoter-transcription factor II is a critical regulator of adipogenesis. Proc. Natl. Acad. Sci. USA 105: 2421-2426 [Abstract] [Full Text]  
• Pennini, M. E., Liu, Y., Yang, J., Croniger, C. M., Boom, W. H., Harding, C. V. (2007). CCAAT/Enhancer-Binding Protein beta and {delta} Binding to CIITA Promoters Is Associated with the Inhibition of CIITA Expression in Response to Mycobacterium tuberculosis 19-kDa Lipoprotein. J. Immunol. 179: 6910-6918 [Abstract] [Full Text]  
• Lagergren, A., Mansson, R., Zetterblad, J., Smith, E., Basta, B., Bryder, D., Akerblad, P., Sigvardsson, M. (2007). The Cxcl12, Periostin, and Ccl9 Genes Are Direct Targets for Early B-cell Factor in OP-9 Stroma Cells. J. Biol. Chem. 282: 14454-14462 [Abstract] [Full Text]  

This Article Abstract Full Text (PDF) Supplemental material Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Reprints and Permissions Copyright Information Books from ASM Press MicrobeWorld Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Jimenez, M. A. Articles by Rosen, E. D. Search for Related Content PubMed PubMed Citation Articles by Jimenez, M. A. Articles by Rosen, E. D.