CCAAT Enhancer-binding Protein β Regulates Constitutive Gene Expression during Late Stages of Monocyte to Macrophage Differentiation*

Human monocyte to macrophage differentiation is accompanied by pronounced phenotypical changes and generally proceeds in the absence of proliferation. The molecular events governing this process are poorly understood. Here, we studied the regulation of the macrophage-specific chitotriosidase (CHIT1) gene promoter to gain insights into the mechanisms of transcriptional control during the differentiation of human blood monocytes into macrophages. We used transient transfections to define a cell type-specific minimal promoter that was mainly dependent on a proximal C/EBP motif that bound multiple C/EBP factors in gel shift assays. In depth analysis of occupied promoter elements using in vivo footprinting and chromatin immunoprecipitation analyses demonstrated the differentiation-associated recruitment of C/EBPβ and PU.1 at the proximal promoter in parallel with CHIT1 mRNA induction. Notably, the induction of C/EBPβ promoter binding strongly correlated with increased nuclear levels of Thr-235-phosphorylated C/EBPβ protein during the differentiation process, whereas C/EBPβ mRNA and total protein expression remained relatively stable. Our data suggest an important constitutive gene regulatory function for C/EBPβ in differentiated macrophages but not in human blood monocytes.

Peripheral blood monocytes are able to differentiate into morphologically and functionally divergent effector cells, including macrophages, myeloid dendritic cells, and osteoclasts (1). These cell types together with their bone marrow progenitors constitute the mononuclear phagocyte system, a widely used model to study cellular differentiation, lineage commitment, and mechanisms of cell type-specific gene regulation. Extensive research during the last two decades revealed several transcription factors that are important for lineage commitment and subsequent differentiation of myeloid progenitor cells as well as myeloid-specific gene regulation. These include members of the Ets-, CCAAT-enhancer-binding protein (C/EBP) 2 , and core binding factor (CBF) families and several other transcription factors (2,3). In particular the Ets family transcription factor PU.1 and C/EBP family members C/EBP␣ and C/EBP␤ represent master regulators of the myeloid lineage. PU.1 is known to be essential for normal monocyte/macrophage differentiation and cell type-specific gene expression (4 -6), and both C/EBP family members are implicated in myeloid development and myeloid gene expression and are capable of reprogramming committed B-or T-cell progenitors to macrophages (7)(8)(9). In humans, the differentiation process of progenitor cells into monocyte/macrophage-like cells has been studied mainly using myeloid progenitor cell lines like HL-60, U937, and THP-1, which acquire monocyte/macrophage-like phenotypes upon vitamin D3 or phorbol ester treatment (10). However, the above cell lines, treated or untreated, differ from primary macrophages (11) and are not necessarily well suited to investigate the transition of primary monocytes into macrophages, which generally proceeds in the absence of DNA synthesis (12). It is clear that the acquisition of a mature macrophage phenotype is distinct from monocytic lineage commitment. It requires expression of a new set of genes, which in turn must be regulated by a macrophage-associated yet poorly characterized network of transcription factors.
To gain insight into the molecular mechanisms of monocyte to macrophage differentiation, our laboratory investigates the regulation of genes that are specifically up-regulated during the differentiation of primary human blood monocytes into macrophages in vitro (13)(14)(15)(16). The two mammalian members of the glycosyl hydrolase family 18, chitotriosidase (CHIT1) (17) and human cartilage 39-kDa glycoprotein (CHI3L1) (18) show a unique cell-type restricted and maturation-associated expression in monocyte-derived macrophages and are direct neighbors on chromosome 1q31-q32. Earlier studies by us and others have shown that the expression of both genes strictly correlates with late macrophage differentiation (13,17,19). Plasma levels of both proteins are markedly elevated in some pathological conditions (20 -23).
Because of their binding specificities, both genes have been implicated in innate immunity and may play a role in defense against chitin-containing pathogens (24). Chitotriosidase, the product of the human CHIT1 gene, is an active chitinase and was shown to inhibit hyphal growth of chitin-containing fungi (25). Genetic association studies on children undergoing acute myeloid leukemia therapy indicate that variant alleles in CHIT1 may influence susceptibility to infections with Gram-negative bacteria (26). CHIT1 activity is increased in patients affected by malaria infection or other hematological disorders in which activated macrophages are involved (27)(28)(29). The specific biological role of both chitinase family members, however, is difficult to study since both genes appear to be differentially regulated in primates and rodents (30,31).
Under normal conditions expression of CHIT1 is prominent in human alveolar macrophages, suggesting that the up-regulation of this gene is part of the normal macrophage differentiation program in the human lung (30). High levels of CHIT1 transcription were also detected in lipid-laden macrophages accumulated in various organs during Gaucher disease (32), subsets of macrophages in atherosclerotic plaques (20), and in Kupffer cells from patients with non-alcoholic fatty liver disease (33), suggesting that lipid mediators may play a role in regulating macrophage CHIT1 induction.
The similar expression profile in human monocyte-derived macrophages in conjunction with their close proximity on chromosome 1 suggests common regulatory mechanisms for both chitinase family genes. Understanding the regulation of model genes like CHIT1 and CHI3L1 on the molecular level may significantly enhance our knowledge of macrophage specific gene regulatory networks.
We recently performed a thorough analysis of the CHI3L1 promoter in macrophages revealing a major role for Sp1-family proteins at this promoter (31). The current study investigates the structure and regulation of the proximal CHIT1 promoter as well as the binding and recruitment of transcription factors at the proximal promoter region of CHIT1 during monocyte to macrophage differentiation. We show that the differentiationassociated CHIT1 induction is accompanied by the binding of transcription factor PU.1 and, most strikingly, of C/EBP␤ to the proximal CHIT1 gene promoter. Whereas C/EBP␤ total mRNA or protein levels remain largely unchanged, we also show that C/EBP␤ recruitment to its target gene promoter is preceded by an increase in phosphorylation of nuclear C/EBP␤ during the differentiation process.

EXPERIMENTAL PROCEDURES
Chemicals-All chemical reagents were purchased form Sigma-Aldrich unless otherwise noted. Protease inhibitors and enzymes were obtained from Roche Applied Science unless otherwise noted. Kinase inhibitors LY294002 and GW2580 were obtained from Calbiochem. U0126 was purchased from Promega (Mannheim, Germany). Standard oligonucleotides were synthesized by Operon Biotechnologies GmbH (Cologne, Germany). High performance liquid chromatography-purified oligonucleotides were obtained by Metabion (Planegg-Martinsried, Germany). Sequences are given in supplemental Table 1. DNA sequencing was done by Geneart (Regensburg, Germany).
Cells-Peripheral blood monocytes were separated by leukapheresis of healthy donors followed by density gradient centrif-ugation over Ficoll/Hypaque and subsequent counter current centrifugal elutriation in a J6M-E centrifuge (Beckman Instruments) as previously described (13). Monocytes were Ͼ85% pure as determined by morphology and expression of CD14 antigen. Supernatants of monocyte cultures were routinely collected and analyzed for the presence interleukin-6, which was usually low, indicating that monocytes were not activated before or during culture. To generate macrophages, isolated monocytes were cultured in endotoxin-free RPMI 1640 medium (Biochrom KG, Berlin, Germany) supplemented with vitamins, antibiotics, pyruvate, nonessential amino acids (all from Invitrogen), 5 ϫ 10 Ϫ8 M ␤-mercaptoethanol, and 2% human pooled AB-group serum on Teflon foils or Petri dishes for up to 11 days. For non-adherent cultures, monocytes were cultured in roller bottles in the presence of 100 ng/ml human CSF-1 (Chiron). Kinase inhibitors GW2580 (10 M), LY294002 (100 M), and U0126 (10 M) were added after 2 or 3 days of culture as indicated. Viability of cells was analyzed by trypan blue exclusion. The human monocytic cell line THP-1 was grown in RPMI 1640 medium supplemented with 10% fetal calf serum (Invitrogen). The human cervical carcinoma cell line HeLa and the human hepatocellular liver carcinoma cell line HepG2 were maintained in Dulbecco's modified Eagle's medium (Invitrogen) plus 10% fetal calf serum.
RNA Preparation-Total RNA was isolated from different cell types by the guanidine thiocyanate/acid phenol method (34) or using the RNeasy mini kit (Qiagen, Hilden, Germany).
RNA Ligase-mediated RACE-PCR-Ten g of total RNA from monocyte-derived macrophages (d7) were used for cDNA synthesis with the FirstChoice TM RLM-RACE kit (Ambion Cambridgeshire, UK). Primers CHIT1-OUT and CHIT1-IN (sequences are given in supplemental Table 1) were used to amplify full-length 5Ј-cDNA fragments. PCR products were cloned into pCR2.1-TOPO (TOPO cloning kit, Invitrogen), and inserts from several individual plasmid-containing bacterial colonies were sequenced.
Real-time Reverse Transcription PCR-RNA (2.5 g) was treated with DNA-free TM DNase (Ambion). One g of DNasetreated RNA was reverse-transcribed using Superscript II MMLV-RT (Promega). Real-time PCR was performed on a LightCycler system (Roche Applied Science) using the Quanti-Tect SYBR Green PCR kit (Qiagen) according to the manufacturer's instructions. Primers are given in supplemental Table 1. Melting curves were analyzed to control for specificity of the PCR reactions. Expression data for CHIT1, CHI3L1, CEBPA, CEBPB, and CEBPD genes were normalized for expression of the ACTB or HPRT gene as indicated. The relative units were calculated from a standard curve plotting three different concentrations of log dilutions against the PCR cycle number at which the measured fluorescence intensity reached a fixed value. For each sample, data of two donors (two independent runs each) were averaged.
Plasmid Construction and Purification-Cloning of the proximal promoter region of the CHIT1 gene was originally performed using the Promoter Finder DNA Walking kit (Clontech, Palo Alto, CA) and primers CHIT1-R1 and CHIT1-R2 following the manufacturer's recommendations. The ϳ1550-bp product obtained from the SspI library was cloned into pCR2.1-TOPO (TOPO Cloning kit, Invitrogen), sequenced, and subcloned (as HindIII/XhoI fragment) into pGL3 basic to obtain the CHIT1-1503 construct. Deletions were done by restriction digestion (CHIT1-419, NheI/XbaI; CHIT1-357, PvuI/SmaI) and subsequent religation or by PCR using CHIT1-1503 construct and primer CHIT1-247, CHIT1-149, or CHIT1-76 with primer GL2 and subsequent cloning into pGL3 basic. Mutation of putative binding sites was done by PCR-mediated mutagenesis using primers CHIT1-mutCEBP or CHIT1-mutGC with primer GL2. PCR fragments with correctly introduced alterations were subcloned back into pGL3 basic. Primer sequences are given in supplemental Table  1. Expression plasmids for human C/EBP-family proteins were a gift from Dr. Alan Friedman. For transient transfections, plasmids were isolated and purified using the EndoFree plasmid kit from Qiagen.
Transient DNA Transfections-THP-1 and HeLa cells were transfected as described earlier (35). HepG2 were transfected in 12-well tissue culture plates using Lipofectamine 2000 reagent (Invitrogen) and 1.5 g of total DNA (1 g of reporter plasmid, 0.5 or 0.1 g of individual expression plasmids, 0.4 g of empty expression plasmid, if required, and 0.05 g of Renilla control vector) according to the manufacturer's instructions. Triplicate transfections were harvested after 24 h. Cell lysates were assayed for firefly and Renilla luciferase activities using the Dual Luciferase Reporter Assay system (Promega) on a Lumat LB9501 (Berthold, Bad Wildbach, Germany). Firefly luciferase activity of individual transfections was normalized against Renilla luciferase activity.
Electrophoretic Mobility Shift Assay-Nuclear extracts from in vitro differentiated human macrophages or THP-1 cells were prepared using Nonidet P-40 lysis as described (36). Doublestranded oligonucleotides corresponding to the proximal C/EBP site (pCEBP_S and AS) were annealed and end-labeled with ␣-[ 32 P]dCTP using Klenow DNA polymerase. Sequences of competitors and mutated oligonucleotides are listed in supplemental Table 1. Antisera used in supershift analyses were from Santa Cruz Biotechnology Inc. (Heidelberg, Germany). Gels were fixed, dried, and imaged using storage phosphor screens and a 9200 Typhoon scanner (GE Healthcare).
In Vivo Genomic Footprinting with DMS-In vivo DMS footprinting was performed as described previously (37) using 1.5 g of genomic DNA from DMS-treated cells. In vitro DMS treatment of naked DNA was performed essentially as described by Maxam and Gilbert (38). Ligation-mediated PCR was carried out using a LP21-25 linker and Bio-CHIT1foot1F, CHIT1foot2F, and CHIT1foot3F. Differences in DMS accessibility were visualized on a denaturing polyacrylamide gel after a labeling reaction with a Cy5-end-labeled linker LP25 primer (Cy5-LP25). Gels were scanned on a 9200 Typhoon scanner (GE Healthcare). Footprinting was performed with cells from two different donors.
Chromatin Immunoprecipitation-Preparation of crosslinked chromatin was performed as described previously (39) with some modifications. Briefly, 10 ϫ 10 6 cells were treated in 1/10th volume of fixation buffer (500 mM HEPES/KOH, pH 7.9, 0.1 M NaCl, 1 mM EDTA, pH 8.0, 0.5 mM EGTA, pH 8.0, 11% formaldehyde) for 10 min at room temperature and quenched by 0.125 M glycine. Cells were resuspended in 250 l of L1A (10 mM HEPES/KOH, pH 7.9, 85 mM KCl, 1 mM EDTA, pH 8.0) and lysed in 250 l of L1B (L1A ϩ 1% Nonidet P-40) for 10 min on ice. Cross-linked chromatin was sheared to an average DNA fragment size around 400 -600 bp using a Branson Sonifer 250 (Danbury, CT). After centrifugation, 4 l of supernatant was used as input. After preclearing with Sepharose CL-4B beads (Sigma) for 2 h, chromatin samples from 2 ϫ 10 6 cells were immunoprecipitated overnight with 2.5 g of rabbit polyclonal antibodies anti-C/EBP␣, -␤, and -␦ and anti-PU.1 (sc-61, sc-150, sc-636, and sc-352, respectively, Santa Cruz Biotechnology Inc.) or anti-rabbit IgG (Upstate/Millipore GmbH, Schwalbach, Germany). Immunocomplexes were recovered by 2 h of incubation with protein A-Sepharose beads (GE Healthcare) at 4°C. After reverse cross-linking, DNA was purified using the QIAquick PCR purification kit (Qiagen). Enrichment of specific DNA fragments in the immunoprecipitated material was determined by quantitative PCR on a LightCycler system (Roche Applied Science) using the QuantiTect SYBR Green PCR kit (Qiagen) according to the manufacturer's instructions. Primers are listed in supplemental Table 1. For each individual primer pair, a standard curve was generated using serial log dilutions of input DNA, and the melting curves for each PCR product were analyzed to control for specificity of the PCR reactions.

RESULTS
Expression of CHIT1 during Monocytic Differentiation-CHIT1 and CHI3L1 genes are direct neighbors on chromosome 1q32.1. Earlier studies suggested that both CHIT1 and CHI3L1 are strongly up-regulated during late stages of macrophage maturation (13,17,19). Real-time PCR analysis was performed to compare the expression profile of both genes. As shown in Fig. 1A, both genes are induced more than 10,000-fold during monocytic differentiation. The overall kinetics of mRNA induction appeared very similar; the observed increase in CHIT1 expression showed a slight delay (approximately 1 day) as compared with CHI3L1. The close physical proximity in combination with the similar expression profile suggests common regulatory mechanisms for both genes.
THP-1 cells, a well studied myeloid cell line, acquire features of differentiated macrophages when treated with phorbol esters. Although expression levels were comparably low, CHIT1 was also strongly induced in PMA-treated THP-1 cells, suggesting that THP-1 cells may share regulatory mechanisms of CHIT1 induction with primary cells (Fig. 1B).
To further investigate the regulation of CHIT1, we studied its expression in monocytes non-adherently cultured in the presence of CSF-1. As shown in Fig. 1C, CSF-1 was sufficient to up-regulate CHIT1 under non-adherent conditions, although the increase was slightly reduced as compared with adherent macrophages. No viable cells were recovered from non-adherent monocyte cultures in the absence of CSF-1. The above data suggested that autocrine (adherence-induced) (40) or paracrine CSF-1 could play a role in CHIT1 regulation.
To study the potential role of CSF-1, we cultured adherent monocytes in the presence of GW2580, a potent inhibitor of CSF-1 receptor-mediated signaling. The inhibitor had little effect on CHIT1 induction when added immediately or after 1 day (data not shown) or 2 days of culture (Fig. 1D), suggesting that autocrine CSF-1 plays a minor role in adherently differentiating macrophages. We next tested the effect of inhibitors for kinases MEK (U01026) and phosphatidylinositol 3-kinase (LY294002) on CHIT1 regulation. Because immediate administration of these inhibitors induced massive cell death, they were added on day 2 at the onset of CHIT1 induction, and CHIT1 expression was analyzed on day 4 where the majority of cells was viable in all culture conditions as determined by trypan blue exclusion. Interestingly, all macrophages detached from the culture substrate after 24 h of treatment with phosphatidylinositol 3-kinase inhibitor, which is in line with previous studies implicating phosphatidylinositol 3-kinase signaling with macrophage adhesion (41,42), whereas macrophages were adherent in all other treatments. As shown in Fig. 1D, only the phosphatidylinositol 3-kinase inhibitor LY294002 had a strong impact on CHIT1 and CHI3L1 expression. In treated macrophages we did not observe repression of other genes including Toll-like receptor 4 (TLR4) or the macrophage marker carboxypeptidase M (CPM) (15) (data not shown), suggesting that the reduced induction was specific for CHIT1 and CHI3L1.
Structure of the Proximal CHIT1 Promoter-To gain further insights into the molecular mechanisms of CHIT1 gene regulation, we defined the proximal promoter of the human CHIT1 gene. We first determined transcription start sites in human macrophages using RLM-RACE PCR. Several cloned RACE PCR products were sequenced defining three transcriptional start sites located 25-30 bp downstream of an atypical TATAlike box (supplemental Fig. 1). Within the human proximal CHIT1 promoter region, several putative binding sites for ubiquitous transcription factors (Sp1/KLF, AP-1) and tissue-restricted factors (e.g. C/EBP, ETS/PU.1) were identified by in silico analysis using the MatInspector data base (Genomatix, Munich, Germany). Alignment of human, chimp, mouse, and rat sequences indicated a moderate degree of sequence conservation and little conservation of putative binding sites at the proximal promoter between primates and rodents (supplemental Fig. 1), suggesting that the regulation of orthologous enzymes may have changed during evolution.
Reporter Gene Analysis of the 5Ј-Flanking CHIT1 Sequence Region-We next cloned a 1.5-kilobase fragment of the 5Ј-proximal promoter region of the human CHIT1 gene as well as several deletions into a luciferase reporter plasmid. Because primary human monocytes/macrophages are not suited for transient transfections, we performed reporter assays in the myeloid cell line THP-1 and the non-myeloid cell line HeLa (cervical carcinoma) to identify important cis elements of the human CHIT1 promoter. As shown in Fig. 2A, significant reporter activity was detected only in THP-1 cells. Maximal reporter activity in THP-1 cells was observed with as little as 76 bp upstream of the transcriptional start site, suggesting that this region may be important for tissue-specific expression of CHIT1. This region contained putative GC-rich and C/EBP elements that were subsequently mutated in the Ϫ76 construct. As shown in Fig. 2B, mutation of the C/EBP site caused an almost complete loss (Ͼ90%) of reporter activity, whereas mutation within the GC-box reduced the reporter activity to 20% that of the wild type construct. Similar results were obtained in transfected and PMA-treated THP-1 cells (data not shown). A double-stranded oligonucleotide corresponding to the proximal C/EBP site was labeled and used in electrophoretic mobility shift assay with nuclear extracts from THP-1 cells (data not shown) and human macrophages. As shown in Fig. 2C, super-shift analyses indicated the prominent binding of C/EBP␤ to the proximal CHIT1 motif in macrophages, whereas binding of C/EBP␣, -␤, -␦, and -⑀ were detected in THP-1 cells (data not shown). To test, whether C/EBP factors were able to transactivate the CHIT1 promoter in reporter gene assays, HepG2 cells were used for co-transfection assays with the Ϫ247 CHIT1 reporter construct and expression plasmids for C/EBP␣, -␤, and -␦. As shown in Fig.  2D, all three factors significantly induced the CHIT1-reporter gene dose-dependently, with C/EBP␣ and -␦ being the most potent transactivators.
DNA-Protein Interactions at the Proximal CHIT1 Promoter in Primary Monocytes and Macrophages-A marked up-regulation of CHIT1 mRNA expression is detected in THP-1 cells after PMA treatment (Fig. 1C); however, untreated THP-1 cells do not express significant endogenous levels of mRNA. The observed high activity of the CHIT1 promoter in transient transfections of untreated THP-1 cells may indicate different levels of accessibility of endogenous and reporter genes for endogenous transcription factors. We next asked whether the induction of CHIT1 during monocytic differentiation is accompanied by an assembly of transcription factor complexes on the CHIT1 promoter. Transcription factor occupancy at the proximal CHIT1 promoter was analyzed using in vivo footprinting. As shown in Fig. 3, DMS in vivo footprinting of the proximal CHIT1 promoter revealed a number of differences in transcription factor occupancy between primary monocytes and monocyte-derived macrophages. In macrophages several sites corresponding to C/EBP␤ consensus motifs, GC-rich sequences and PU.1 motifs were clearly protected, indicating the high affinity binding of transcription factors at these sites. In contrast, the pattern of footprints only slightly differed from naked DNA in monocytes, suggesting that nuclear proteins are specifically recruited to the above sites during the differentiation process.
To determine which C/EBP family member might bind the C/EBP consensus sites in vivo, we performed chromatin immunoprecipitation assays with cross-linked chromatin from freshly isolated human blood monocytes and in vitro differentiated macrophages (at days 3, 5, and 8) using antibodies against C/EBP␣, -␤, and -␦ as well as PU.1 (Fig. 4). Specifically precipitated DNA fragments were quantified using real-time PCR and primers proximal to the promoter region as indicated in Fig. 4. We detected a strong increase of promoter-bound C/EBP␤ during monocyte to macrophage maturation. The increased binding was specific for C/EBP␤; C/EBP␣ and C/EBP␦ were not significantly precipitated in mature macrophages, which is in line with the sole binding of C/EBP␤ to the proximal site in electrophoretic mobility shift assay (Fig. 2). We also detected a significant increase in PU.1 binding during the differentiation of macrophages. The recruitment of both PU.1 and C/EBP␤ paralleled the induction of mRNA expression in macrophages (Fig. 1), suggesting that both factors may be involved in the cell  type-specific induction of the CHIT1 gene. We also tried chromatin immunoprecipitation assays for Sp1 and Sp3 as possible binding candidates for the GC-rich elements; however, no specific enrichment was obtained for these factors at the CHIT1 promoter (data not shown). A similar recruitment of both PU.1 and C/EBP␤ was also detected during PMA-induced differentiation of THP-1 cells (supplemental Fig. 2A).
Regulation of C/EBP␤ during Monocyte Differentiation-The above data strongly suggested a role for C/EBP␤ in CHIT1 regulation. We first investigated, whether regulation of CEBPB (the gene encoding C/EBP␤) might account for its differential binding at the proximal promoter, and studied its mRNA expression, in comparison with the two other family members CEBPA and CEBPD using real-time PCR. Whereas CEBPD transcription was markedly down-regulated (20-fold) after monocyte adherence and CEBPA was up-regulated 2-fold, mRNA expression of CEBPB remained unchanged during the time course studied (Fig. 5A). In contrast, CEBPB was up-regulated in THP-1 cells after PMA treatment (supplemental Fig.  2B), suggesting that regulation of CEBPB mRNA may play a role in THP-1 differentiation but not in primary cells. To test whether CEBPB is regulated on the protein level, Western blot analysis was performed using whole cell extracts from monocytes and in vitro differentiated macrophages. Fig. 5B shows a representative immunoblot using a polyclonal C/EBP␤ antibody. Alternative start codon usage generates three C/EBP␤ isoforms represented by 48-kDa (LAP*) and 45-kDa (LAP) isoforms and a short isoform of 17 kDa (LIP). An additional, specific band (35 kDa) likely represents a proteolytic cleavage product of LAP isoforms. Although both LAP isoforms slightly increased after overnight culture of monocytes, C/EBP␤ protein levels were relatively stable during the remaining time course, suggesting that changes in C/EBP␤ protein levels may not account for the increase in differentiation associated recruitment of C/EBP␤ to the CHIT1 promoter. Earlier studies indicated that C/EBP␤ activity may also be controlled by its phosphorylation status (43). Therefore, we investigated the presence of Thr-235-phosphorylated C/EBP␤ protein in nuclear extracts during the differentiation time course. As shown in Fig. 5C, immunoblots using polyclonal antibodies against phosphorylated C/EBP␤ revealed a marked increase of nuclear protein levels with monocytic differentiation. Nuclear, phosphorylated C/EBP␤ was undetectable until day 2, but its levels strongly increased upon further culture, suggesting that C/EBP␤ phosphorylation may account for the differential C/EBP␤ binding to the CHIT1 promoter. A similar increase in C/EBP␤ phosphorylation was also detected in PMA-treated THP-1 cells (supplemental Fig. 2C).
We next tested whether the kinase inhibitors that interfered with CHIT1 expression (Fig. 1), would have an impact on C/EBP␤ expression or phosphorylation. As shown in Fig. 5D, treatment with the indicated inhibitors did not decrease CEBPB mRNA expression. However, we observed a reduction in nuclear Thr-235 phosphorylation of C/EBP␤ when macrophages were treated with phosphatidylinositol 3-kinase inhibitor LY294002 (Fig. 5E), suggesting that phosphatidylinositol 3-kinase may directly or indirectly contribute to the differentiation induced phosphorylation of C/EBP␤. No loss of phosphorylation was detected in GW2580-or U0126-treated cells (Fig.  5E), suggesting that CSF-1 or MEK activity are not required for Thr-235 phosphorylation of C/EBP␤ in human macrophages.
We additionally investigated the subcellular localization of total and phosphorylated C/EBP␤ using an independent approach. Fig. 6 shows representative immunofluorescent staining of monocytes and macrophages using the above antibodies. In monocytes, C/EBP␤ was detected in the cytoplasm and nuclei of monocytes using the anti-C/EBP␤ (Fig. 6A), but no staining was obtained with the phosphospecific antibody (Fig. 6B). In contrast, both antibodies stained C/EBP␤ preferentially in nuclei of mature macrophages (Fig. 6, A and B). The immunofluorescence data suggest that C/EBP␤ is already present in monocyte nuclei; however, it is not significantly phos-phorylated at Thr-235. Nuclear staining of monocytes with the total C/EBP␤ antibody generally appeared weaker, suggesting that nuclear levels of C/EBP␤ may further increase during the differentiation process. Most notably, phosphorylation at Thr-235 is only evident in macrophages but not in monocytes.

DISCUSSION
The differentiation of human peripheral blood monocytes into macrophages is accompanied by marked changes in cell morphology, gene regulation, and function. The current study demonstrates that the transcription factor C/EBP␤, although expressed at comparable levels during the monocytic differentiation process, is increasingly phosphorylated at Thr-235, suggesting that human macrophages (in contrast to peripheral monocytes) contain high constitutive levels of functional C/EBP␤. We also show that C/EBP␤, among other factors like PU.1, is abundantly recruited to the CHIT1 promoter in a differentiation-associated manner and necessary for its activity.
C/EBP␤ is a basic leucine zipper transcription factor and member of the C/EBP family. It is transcribed from an intronless gene, and a single transcript is translated into three different isoforms, full-length LAP* (AA 1-345), LAP (AA 24 -345), and the short LIP (amino acids 199 -345). Both LAP isoforms constitute transcriptional activators, whereas the short LIP polypeptide acts as a dominant negative inhibitor of C/EBP function (7,44). C/EBP␤ is known to play important roles, e.g. in the control of cellular differentiation, metabolism, and inflammation, particularly in adipocytes, hepatocytes, and macrophages (7,44,45). Several studies using cell lines or C/EBP␤-deficient macrophages established that C/EBP␤ contributes to the inducible expression of several inflammatory genes including interleukin (IL) 1␤, IL-12, IL-6 or PTGS2 in mouse macrophages (46, 47). C/EBP␤ or -␣ overexpression were found to induce the reprogramming of B-cells or pre-Tcells into macrophages (8,9), suggesting that these C/EBP factors play important roles in linage commitment and the establishment of a macrophage-specific transcription factor network. The activity of C/EBP␤ appears to be regulated on various levels, including transcriptional, translational, or posttranslational (7,44). The contributions of various regulatory mechanisms to C/EBP␤ function in macrophages, however, are unclear. In the monocyte/macrophage lineage, it has been shown that the transcription of CEBPB increases during the differentiation of progenitor cells (48). We observed a similar induction of CEBPB during PMA-induced differentiation of THP-1 cells. However, the presented data suggest that the transition of monocytes into mature macrophages is not associated with major changes in CEBPB mRNA, which remained stable during the differentiation process. Total protein levels slightly increased after adherence (day 1); in particular, the longer LAP isoforms were induced. Our immunohistochemical staining indicates that total nuclear levels of C/EBP␤ also slightly increase during differentiation of monocytes into macrophages. The clearest difference between monocytes and macrophages was observed on the level of C/EBP␤ phosphorylation. Protein phosphorylation has been previously identified as a major regulator of C/EBP␤ function. Phosphorylation at specific serine or threonine residues appears to relieve repression of the transactivation domain and control nuclear translocation. Phosphorylation of C/EBP␤ at Thr-235, for example, was recently shown to determine the association of C/EBP␤ with the active Mediator complex (49), suggesting that this event determines the transcriptional properties of C/EBP␤. A recent study showed that C/EBP␤ is constitutively phosphorylated at multiple sites and functional in primary mouse macrophages (50); however, this study did not observe constitutive phosphorylation of Thr-188 (corresponding to Thr-235 in human C/EBP␤). Our studies on the specific regulation of the CHIT1 gene promoter and general C/EBP␤ properties during monocyte to macrophage differentiation in vitro show that in human macrophages (but not in monocytes) C/EBP␤ is constitutively phosphorylated at Thr-235 and bound to the human CHIT1 promoter. Binding of C/EBP␤ is accompanied by CHIT1 transcription; however, expression of the above activation markers (interleukin-1␤ or -6 or PTGS2) is further down-regulated (data not shown), suggesting that additional factors may specify the function of C/EBP␤ on individual promoters. It is unclear so far which particular signals or kinases are responsible for C/EBP␤ phosphorylation and activation during human monocytic in vitro differentiation. The Thr-235 residue was previously shown to be a target of the RAS-dependent mitogen-activated protein kinase cascade (43), which is triggered, for example, by CSF-1. However, our inhibitor studies indicate that neither autocrine CSF-1 nor mitogen-activated protein kinase signaling is involved in the regulation of CHIT1 or the constitutive phosphorylation of C/EBP␤ at Thr-235 in human monocyte-derived macrophages. The observed effect of a phosphatidylinositol 3-kinase inhibitor suggests that this kinase is either directly or indirectly involved in CHIT1 expression and C/EBP␤ phosphorylation.
We also noted a differentiation-associated recruitment of PU.1 to the proximal CHIT1 promoter. Because PU.1 is also present and detectable in monocytes, 3 it is likely that at least the proximal promoter is remodeled during differentiation, revealing transcription factor binding sites that were inaccessible in monocytes. The factors that may induce chromatin remodeling at the CHIT1 promoter are unknown. Previous studies in chicken myeloid cells showed that the long isoform of C/EBP␤ (LAP*) is able to recruit the SWI/SNF complex to myeloid target genes (51). Thus, C/EBP␤ might be responsible for the activation of the CHIT1 gene. However, it is also possible that other factors are involved, e.g. those binding the GC-rich sequences of the CHIT1 promoter that were occupied in macrophages but not in monocytes. Further studies are needed to identify those additional factors that participate in the regulation of CHIT1. For the adjacent CHI3L1 gene, we previously determined a minor role for C/EBP factors in regulating the CHI3L1 promoter. However, it is possible that reporter assays in THP-1 cells do not accurately mirror the necessity of cis-acting elements in macrophages in vivo. Chromatin immunoprecipitation assays indicate that C/EBP␤ is also recruited to the CHI3L1 promoter in a differentiation-associated manner, suggesting that it may also play a role in regulating CHI3L1 and possibly other macrophage-specific genes.
In summary, our study provides a detailed characterization of the human CHIT1 gene promoter. We show that transcription factor C/EBP␤ is an important regulator of the CHIT1 gene and in general that the activity of C/EBP␤ is controlled during monocytic differentiation likely through the phosphorylation of Thr-235 in macrophages but not in monocytes.