Published online before print September 7, 2007
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
* Institute of Human Genetics, University of Regensburg, and
Department of Hematology and Oncology, University Hospital Regensburg, Germany
2Correspondence: Institute of Human Genetics, University of Regensburg, Franz-Josef-Strauss-Allee 11, 93053 Regensburg, Germany. E-mail: thomas.langmann{at}klinik.uni-regensburg.de
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: Tyrobp • retinal degeneration • myeloid promoter
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The membrane adaptor protein Dap12 is expressed on several immune cells, contains an immunotyrosine-based activation motif, and associates with more than 20 different surface receptors to regulate immune responses [10 ]. Although Dap12 has been shown to have a dual role by potentiating or attenuating immune cell functions, a mainly deactivating function has been postulated for myeloid cells [11 ]. Thus, together with triggering receptor expressed on myeloid cells 2 (TREM2), Dap12 blocks LPS/TLR-mediated cellular activation [12 , 13 ] and enhances differentiation of macrophage precursors [14 ]. Furthermore, TREM2 and Dap12 are expressed on brain microglia and cooperatively control phagocytosis of apoptotic neurons [15 , 16 ]. Despite these findings, regulation of Dap12 expression has not been analyzed in resting or activated retinal microglia cells, and nothing is known about cis-acting elements controlling Dap12 transcription in any cell type.
Therefore, our objective in this study was to investigate whether the increased Dap12 expression in the dystrophic retina of Rs1h–/Y mice can be attributed to activated microglia cells. To achieve this goal, ex vivo isolation and culture of retinal microglia from Rs1h–/Y mice and an in vitro model of microglia activation were established. Detailed mRNA expression studies were performed in both cell systems using DNA microarrays and real-time quantitative RT-PCR (qRT-PCR). Moreover, as a first step towards an understanding of the transcriptional regulation of Dap12 in mononuclear phagocytes, luciferase reporter gene assays, EMSA, chromatin immunoprecipitation (ChIP), RNA interference, and re-expression experiments were carried out. These experiments establish PU.1 as a major regulator of Dap12 expression in macrophages.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Isolation and culture of retinal microglia
Retinal tissue from wild-type and Rs1h–/Y mice at Postnatal Day 14 (P14) was isolated from eye bulbs and purified from contaminating vitreous body and retinal pigment epithelium (RPE)/choriocapillaris. Pools of four retinae each were cut into small pieces and incubated for 40 min at 37°C in 1 ml PBS with 1 mg/ml collagenase type I (Sigma Chemical Co., St. Louis, MO, USA), 0.3 mg/ml DNase I (Roche, Indianapolis, IN, USA), and 0.2 mg/ml hyaluronidase (Sigma Chemical Co.). The cell suspension was filtered through a 70-µm cell strainer (Becton Dickinson, San Jose, CA, USA). Cells were washed twice with 10 ml DMEM/10% FCS and finally subjected to Ficoll density gradient centrifugation for 20 min at 2000 rpm (690 g, without brake) in a Heraeus centrifuge for the isolation of mononuclear cells. The interphase was removed carefully and washed with 10 ml DMEM/FCS. The cells were cultured for 11 days in 75 cm2 flasks containing DMEM/10% FCS, supplemented with 50 ng/ml recombinant human M-CSF (R&D Systems, Minneapolis, MN, USA), and phase contrast micrographs were taken with a Nikon Eclipse TE2000-S microscope.
Culture of cell lines
HeLa and RAW264.7 cells were obtained from American Type Culture Collection (Manassas, VA, USA), and BV-2 cells were a gift from Professor Ralph Lucius (Clinic of Neurology, Christian Albrechts University, Kiel, Germany). HeLa and RAW264.7 cells were cultured in DMEM, supplemented with 10% FCS, 100 U/ml penicillin, and 100 µg/ml streptomycin and incubated in 10% CO2 in air at 37°C. The culture of BV-2 cells has been described elsewhere [17
]. PUER cells [18
] were cultivated in IMDM (Invitrogen, Carlsbad, CA, USA), supplemented with penicillin/streptomycin (10,000 U/ml), glutamin (200 mM), β-ME (50 µM), mouse IL-3 (5 ng/ml, Biosource, Camarillo, CA, USA), puromycin (1 µg/ml), and 10% FCS. To activate the PU.1-ER fusion protein, PUER cells were centrifuged and resuspended in medium containing 0.1 µM tamoxifen [4-hydroxytamoxifen (OHT)].
Activation and phagocytosis assay of BV-2 cells
To keep BV-2 cells in a ramified and resting state, cells were incubated in serum-free medium for 24 h. For the gradual transformation into amoeboid cells, 5% and 10% FCS was added to the culture medium, followed by further incubation for 48 h. The phagocytic activity of amoeboid cells cultured in the presence of 10% FCS was assessed as reported earlier [19
]. Briefly, 1 µm blue Latex beads from polystyrene (Sigma Chemical Co.) were added to the wells at a concentration of 1 µl beads/ml, and cells were washed with PBS after overnight incubation. The phagocytosis potential was monitored by counting Latex blue bead-positive cells using microscopy.
RNA isolation and RT
Total RNA was extracted from retinae or cultured microglia cells according to the manufacturer’s instructions using the RNeasy Protect Midi kit or Micro kit, respectively (Qiagen, Valencia, CA, USA). Purity and integrity of the RNA were assessed on the Agilent 2100 bioanalyzer with the RNA 6000 Nano LabChip® reagent set (Agilent Technologies, Santa Clara, CA, USA). The RNA was quantified spectrophotometrically and then stored at –80°C. First-strand cDNA synthesis was performed with the RT system from Promega (Madison, WI, USA), according to the manufacturer’s instructions.
DNA microarray analysis
Generation of ds-cDNA, preparation and labeling of cRNA, hybridization to Affymetrix 430 2.0 mouse genome arrays, washing, and scanning were performed according to the Affymetrix standard protocol. Duplicate microarrays were carried out with pooled RNA from ex vivo microglia isolated at P14 from wild-type and Rs1h–/Y retinae, which were cultured for 11 days in M-CSF medium. Data analysis was carried out as described previously [8
, 20
] and minimal information about a microarray experiment criteria was met [21
]. Functional annotation of coregulated transcripts was performed using BiblioSphere pathway edition (Genomatix Software GmbH, Germany) and the GenMap annotator and pathway profiler (Gladstone Institutes, San Francisco, CA, USA). The microarray dataset of this study is publicly available at the National Center for Biotechnology Information Gene Expression Omnibus (http://www.ncbi.nlm.nih.gov/geo/).
Real-time qRT-PCR
Amplifications of 50 ng cDNA were performed with the iCycler iQTM real-time PCR detection system (Bio-Rad, Munich, Germany) in triplicates in 25 µl reaction mixtures containing 1x TaqMan Universal PCR Master Mix (Applied Biosystems, Foster City, CA, USA), 200 nM primers (Table 1
), and 0.25 µl dual-labeled probe (Roche Probe Library). The reaction parameters were as follows: 2 min, 50°C hold; 30 min, 60°C hold; and 5 min, 95°C hold, followed by 45 cycles of 20 s, 94°C melt, and 1 min, 60°C anneal/extend. Relative quantification with normalization to various reference genes was performed as described earlier [22
].
|
View this table: [in a new window] |
Table 1. Primer Sequences and Probe IDs for Real-Time qRT-PCR Using the Roche Probe Library
|
Clone –860/+118 was used as template for site-directed mutagenesis at the –80 and –26 PU.1 sites. The QuikChange site-directed mutagenesis kit (Stratagene) was applied, according to the manufacturer’s recommendations, with the following mutagenic primers: PU.1(1), 5'-GCGGCCATGTCTATACGTAAATCCTCCCTGCTGC-3', and PU.1(2), 5'-CCACCACCCACCTCACACAAACTCCTTCACTTGGTTGG-3'. Double PU.1 site mutations were generated by applying both primers in a single reaction. Altered nucleotides are depicted in bold and underlined.
RAW264.7, BV-2, and HeLa cells were transfected using 6 µl Fugene6 reagent (Roche) with 3 µg each reporter plasmid in six-well plates. The transfected cell lines were harvested after incubation for 48 h. Cell lysates were assayed for protein concentration using the Bradford assay, and firefly luciferase activity was measured with the Luciferase assay system (Promega) on a FLUOstar OPTIMA (BMG Labtech, Germany). Each experiment was carried out three times, and each measurement was done in triplicate.
EMSA
Nuclear extracts were prepared using the NE-PER nuclear and cytoplasmic extraction reagents (Pierce Biotechnology, Rockford, IL, USA). Double-stranded oligonucleotides were end-labeled with T4 polynucleotide kinase (Fermentas, Ontario, Canada) and [
-32P]ATP. An equivalent of 40,000 cpm double-stranded oligonucleotide probe containing the Dap12 promoter sequence was incubated with 5 µg nuclear extract from BV-2 or RAW264.7 cells as described previously [23
] in a buffer containing 50 mM HEPES/HCl, pH 7.9, 6 mM MgCl2, 50 mM DTT, 100 µg/ml BSA, 0.01% Nonidet P-40 (NP-40), and 1 µg poly(dI-dC) at room temperature for 20 min.
In vitro-translated PU.1 (IVT-PU.1) or a control reaction was generated from a pGEM plasmid containing the open-reading frame of PU.1 in sense or antisense orientation [24
], respectively, and the TNT Quick Coupled Transcription/Translation system (Promega). [35S]Methionine PU.1 was analyzed by SDS-PAGE and autoradiographed to confirm its correct molecular mass of 38 kDa. Supershift analysis was carried out with 2 µl antisera (Santa Cruz Biotechnology, Santa Cruz, CA, USA) against PU.1 (SC352X) and Ets1/2 (SC351X). In competition experiments, nuclear extracts were preincubated with a 100-fold molar excess of wild-type or mutant competitor for 10 min prior to the addition of the radiolabeled probe. The sequences of the oligonucleotides used in EMSA analysis are shown (see 
Fig. 7
). DNA–protein complexes were electrophoretically resolved on a native 8% polyacrylamide gel, dried, and audioradiographed with Kodak BioMax MR films at –80°C.
 View larger version (69K): [in a new window] |
Figure 1. Morphology of ex vivo microglia cells from wild-type (WT) and Rs1h–/Y retinae. Phase contrast micrographs were taken from retinal microglia isolated at P14 and cultured in the presence of 50 ng/ml M-CSF for 11 days. (A and C) Ramified morphology of resting, nonactivated cells (wild-type C57BL/6 littermates). (B and D) Ameboid cell shape typical for activated microglia cells (Rs1h–/Y mice). (A and B) Original magnification, x100; (C and D) original magnification, x200.
|
 View larger version (23K): [in a new window] |
Figure 2. Gene expression profiles of isolated microglia and PU.1 regulatory network. (A) qRT-PCR analysis of selected marker genes in Rs1h–/Y-derived microglia and total retina. The relative expression of each marker gene was analyzed in 50 ng RT cDNA (total RNA equivalent) and calculated for microglia (solid bars) and retina (open bars), respectively. The data shown are the means of two independent cell/tissue preparations from four retinae each. Cd68, Dap12, Clec7a, Casp11, Gfap, and Cralbp transcript abundance was normalized to a set of reference genes (Atp5b, Gusb, Hprt1, and Rp14), which were not significantly different in both RNA sources. Error bars indicate the SD of the mean. (B) Pie chart displaying functional categories represented by the 94 significantly up-regulated genes in Rs1h–/Y versus wild-type microglia. Functional annotation of the up-regulated genes (presented in Supplementary Table 1) was carried out with BiblioSphere Pathway edition (Genomatix Software GmbH) and GenMapp (Gladstone Institutes). Circled numbers indicate the quantity of genes in each segment. (C) PU.1-dependent transcriptional regulatory network. Sixteen genes significantly induced in Rs1h–/Y versus wild-type microglia contain conserved binding motifs for PU.1/Ets. Among these, Csf1r, Msr1, Fcgr2, Fcgr3, and Cd68 are known PU.1 targets [28
29
30
31
] (solid-line arrows). All other genes, including Dap12, have not been shown to be regulated by PU.1 (dashed-line arrows). Gene symbols and the fold change of regulation identified by DNA microarrays are indicated in small boxes.
|
 View larger version (17K): [in a new window] |
Figure 3. Increased Dap12 and PU.1 expression in total retinae and isolated microglia from Rs1h–/Y mice. qRT-PCR analysis of Dap12 and PU.1 in Rs1h–/Y retinae and ex vivo microglia compared with wild-type mice. The data shown are mean values of two independent experiments representing transcript levels from a pool of four retinae or ex vivo microglia preparations each. Error bars indicate the SD of the mean. *, P < 0.05 Rs1h–/Y, versus wild-type using Student’s t-test.
|
 View larger version (120K): [in a new window] |
Figure 4. Induction of Dap12 expression during in vitro activation of BV-2 cells. (A–E) Phase contrast micrographs of BV-2 cells cultured in the absence of FCS (A, ramified morphology), 5% FCS (B, partially activated), or 10% FCS (C and D, fully activated). Activated BV-2 cells incubated with 1 µl Latex blue beads for 1 h (C) and overnight (D). Original magnification, x100. (E) Quantification of phagocytosis by microscopic counting of Latex blue bead-positive cells after overnight incubation. A strong, phagocytic activity was detected in cells cultured in the presence of 10% FCS compared with resting cells. *, P < 0.05, BV-2 10% FCS and BV-2 5% FCS versus BV-2 without (w/o) FCS using Student’s t-test. (F) qRT-PCR analysis of Dap12 in in vitro-activated BV-2 cells and RAW264.7 cells. The data shown are mean values of three independent experiments, and error bars indicate the SD of the mean. **, P < 0.01, BV-2 10% FCS and RAW264.7 versus BV-2 without FCS using Student’s t-test.
|
 View larger version (49K): [in a new window] |
Figure 5. Evolutionary conservation of the proximal Dap12 promoter. A ClustalW alignment is shown using Dap12 promoter sequences from Mus musculus (mouse), Rattus norvegicus (rat), Bos taurus (cow), Homo sapiens (human), Pan troglodytes (chimp), and Macaca mulatta (rhesus monkey). The boxed sequences contain predicted binding sites for PU.1 (–80 and –26) and vitamin D receptor (VDR)/retinoid X receptor (RXR). The core binding sequences for PU.1 are indicated in bold and underlined. The transcription start sites identified using the mouse CAGE library [26
] are shown as asterisks, and the location of the sequences used for EMSA analysis is marked by arrows over the conservation blocks.
|
 View larger version (29K): [in a new window] |
Figure 6. The Dap12 promoter is only active in myeloid cells and requires intact PU.1/Ets binding sites. (A and B, left sides) Overview of luciferase (Luc) reporter plasmids. (A) CMV-Luc-positive control, a promoterless luciferase vector, and the –860/+118 Dap12 promoter luciferase construct were transiently transfected into the nonmyeloid HeLa cell line and into the microglia/macrophage cell lines BV-2 and RAW264.7. Luciferase activity is calculated relative to the empty reporter vector. (B) Deletion and mutagenesis analysis of the Dap12 promoter in RAW264.7 cells. Luciferase activity is calculated relative to the longest wild-type construct –860/+118, which was set to 100%. Values are means from three independent experiments, and error bars indicate the SD of the mean.
|
 View larger version (55K): [in a new window] |
Figure 7. PU.1 interacts with both PU.1/Ets sites in the proximal Dap12 promoter. Labeled, dsPU.1 oligonucleotides PU.1(1) and PU.1(2), containing the sequences shown (lower), were used in EMSA with nuclear extracts from RAW264.7 cells, BV-2 cells, or IVT-PU.1 protein. (A) EMSA with the PU.1(1) oligonucleotide and RAW264.7 nuclear extracts (Lanes 2–6). Competition analysis was carried out using 100-fold molar excess of cold wild-type (Lane 3) or cold mutant (Lane 4) probe. Antibody supershift analyses with specific PU.1 (Lane 5) or Ets1/2 antibodies (Lane 6) are indicated. IVT-PU.1 and a corresponding negative control (asPU.1) were used in Lanes 7–10. (B) EMSA with the PU.1(2) oligonucleotide and RAW264.7 extract (Lanes 2–5), BV-2 extracts (Lanes 6 and 7), and IVT-PU.1 (Lanes 8–11). Competition experiments were performed using 100-fold molar excess of wild-type or mutant oligonucleotides. Supershift analyses were done in the presence of 2 µl of the indicated antisera. Specific PU.1-containing complexes are indicated by arrows, and nonspecific bands are marked (*).
|
RNA interference
Knockdown of PU.1 in RAW264.7 cells was achieved with two independently annealed, ds-small interfering (si)RNAs (Qiagen). The siRNA sequences of murine PU.1 were: PU.1(1), 5'-(GGAUGUUACAGGCGUGCAA)dTdT-3' (sense), and PU1.(2), 5'-(GCA AGAAGAUGACCUACCA)dTdT-3' (sense). BLAST searches confirmed that these sequences were not homologous to any nuclear genes other than PU.1. GAPDH siRNA represented a positive control and a scrambled siRNA molecule a negative control. RAW264.7 cells at 80% confluence in 12-well plates were transfected with 75 ng each of the above siRNAs using the HiPerFect transfection reagent (Qiagen), according to the manufacturer’s protocol. The transfected cells were incubated at 37°C for 48 h before isolation of total RNA and qRT-PCR analysis to confirm the knockdown effect.
Bioinformatic transcription factor binding site and promoter analysis
Promoter sequences from transcripts induced significantly in activated microglia were retrieved with Gene2Promoter (Genomatix Software GmbH) and DataBase of Transcriptional Start Sites (http://dbtss.hgc.jp). Approximately 1 kb DNA of the upstream regulatory regions was analyzed for putative transcription factor binding sites using Matinspector (Genomatix Software GmbH) and Transfac (BIOBASE GmbH, Germany). Only matrices predicted with algorithms and a core similarity of 1.0/a matrix similarity >0.75 were included in the regulatory network analysis. Transcription start sites were determined by the cap analysis of gene expression (CAGE) resource for comprehensive promoter analysis (http://fantom3.gsc.riken.jp) [26
].
Statistical analysis
Data are presented as means, and error bars indicate the SD of the mean. Statistical significance was calculated using the Student’s t-test, and P < 0.05 was significant.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
We adapted a protocol for the isolation of colonic macrophages [27 ] and obtained pure cultured microglia cells without visible contaminations of Müller glial cells, RPE cells, or astrocytes (Fig. 1 ). The cells proliferated in distinct, large colonies when cultured for 3–5 days in medium supplemented with M-CSF. A clear difference in the morphological phenotype was observed between wild-type (Fig. 1A and 1C) and Rs1h–/Y (Fig. 1B and 1D) mice after 11 days in culture. Although wild-type microglia showed long protrusions, typical for resting cells (Fig. 1C) , most microglia from Rs1h–/Y retinae displayed an amoeboid morphology characteristic of activated phagocytes (Fig. 1D) . This morphological difference was lost when cells were cultured more than 2 weeks (data not shown), indicating an adaptation to the culture conditions after prolonged culture.
To estimate the purity of activated ex vivo microglia from Rs1h–/Y retinae, marker transcripts for different retinal cell types were analyzed in isolated cultured cells and total retina by real-time qRT-PCR (Fig. 2A ). The specific myeloid marker Cd68 and the RPE/Müller cell marker Cralbp were highly enriched in microglia or in retina (Fig. 2A , first and last bars). Transcript levels of Dap12, Clec7a, and Casp11 were markedly higher in microglia cells compared with retina when using the same amount of template from both RNA sources. In contrast, Gfap, which has been detected previously in the dystrophic Rs1h–/Y retina, shows lower abundance in isolated microglia, suggesting expression in microglia and also in other retinal cell types. In summary, these data indicate that we have isolated and cultured a pure cell population well suited to detect microglia-specific molecular events in retinal degeneration.
Our next goal was to capture and compare the transcriptional profile of microglia cells from Rs1h–/Y versus wild-type retinae using Affymetrix Mouse Genome 430 2.0 GeneChips. Duplicate microarray analyses from two independent animal pools were performed. Applying high-stringency criteria with at least 2.0-fold difference and a minimum signal intensity of 50, we have identified 143 differentially regulated genes (Supplementary Table 1). It is interesting that the majority of these genes (n=94) was induced significantly in Rs1h–/Y microglia compared with control cells. Molecular annotation of functions and pathways of the 94 up-regulated genes revealed a strong representation of the categories "Surface receptors," "Enzymes and inhibitors," "Chemokines and cytokines," and "Transcription factors," strongly reminiscent of activated macrophages (Fig. 2B)
. In contrast, a significant number of genes could not be grouped into larger, immune-related pathways and have not been connected previously to microglia activation. Also, many genes with no detectable expression difference in the total retina of Rs1h–/Y versus wild-type mice [8
] were now significantly different in isolated and enriched microglia. Likewise, we have detected a significant 3.7-fold mRNA increase of the transcription factor PU.1 in activated microglia (Fig. 2C
, yellow centered box). As nuclear factors are of particular importance for transcriptional responses, regulatory network analysis was performed with all up-regulated gene clusters and the identified transcription factors PU.1, Runx3, Atf3, Msx2, Lmo7, and peroxisome proliferator-activated receptor . Bioinformatic binding-site analysis and literature data mining exclusively revealed a PU.1-dependent network consisting of 16 genes, including five known PU.1 targets (Fig. 2C
, solid-line arrows). There were also 11 novel, putative, PU.1-regulated genes (Fig. 2C
, dashed-line arrows), including Dap12.
To confirm the DNA microarray data, Dap12 and PU.1 transcript levels were determined in total retinae and isolated microglia from Rs1h–/Y and wild-type mice by real-time qRT-PCR. As shown in Figure 3 , Rs1h–/Y retinae showed a threefold higher expression of Dap12 than wild-type tissue, and a more than fivefold difference of Dap12 mRNA levels was observed in Rs1h–/Y compared with wild-type microglia. Although the amount of PU.1 transcripts was not different in total retinae (most likely as a result of an overall low expression level), enriched, activated microglia expressed significantly higher amounts of PU.1 mRNA than cells from wild-type mice (Fig. 3) . These data indicate that Dap12 and PU.1 expression is correlated with the activation state of microglia.
To investigate whether the overexpression of Dap12 is a direct consequence of microglia activation, an independent in vitro model (BV-2 microglia cell line) was established. BV-2 cells are converted into a quiescent state when cultured in the absence of serum for 24 h (Fig. 4A ). Addition of 5% and 10% FCS caused a gradual, morphological change in the majority of cells, such as loss of cytoplasmic protrusions and rounding of the cells (Fig. 4B and 4C) . Experiments measuring the overnight uptake of Latex beads demonstrated a strong phagocytic activity of cells kept in medium with 5% FCS and 10% FCS (Fig. 4D and 4E) compared with resting cells in serum-free conditions. Therefore, these morphological changes and functional characteristics suggest an activated state of BV-2 cells kept in full culture medium. In line with our findings from the ex vivo microglia, an elevated Dap12 expression was detected in BV-2 cells when transformed from the ramified to the amoeboid state by increasing the FCS concentration in the culture medium (Fig. 4F) . It is interesting that the mouse macrophage cell line RAW264.7 showed a BV-2-comparable amount of Dap12 transcripts (Fig. 4F) and high levels of PU.1 mRNA (data not shown). These cells were therefore used for the majority of the following promoter assays.
Myeloid-specific promoter activity of Dap12 depends on proximal PU.1 sites
To further study myeloid Dap12 gene expression and to characterize the potential regulation by PU.1 at the promoter level, we cloned the murine Dap12 upstream regulatory region and determined the transcription start sites with the help of macrophage data from the CAGE library [26
]. Two close transcription initiation sites were found in independent mouse macrophage mRNA pools (Fig. 5
, *). Five 5'-upstream regions from mouse, rat, cow, human, chimpanzee, and rhesus monkey Dap12 revealed a 90-bp promoter region with two highly conserved blocks for PU.1- binding sites (Fig. 5)
. There was no strong sequence homology further upstream of this region, suggesting that the perfectly conserved motif for the myeloid transcription factor PU.1 at position –26 and the less conserved, second PU.1 site at –80 might be functionally important for several species. A VDR/RXR motif with lower similarity score was also identified (Fig. 5)
.
Transient transfections of a pGl4.10 luciferase reporter construct containing 860 bp of the Dap12 promoter and extending until the end of the first exon (–860/+118) revealed a strong promoter activity in BV-2 and RAW264.7 cells, whereas the nonmyeloid cell line HeLa (cervical carcinoma) displayed no significant activity (Fig. 6A ). Analysis of the two deletion constructs, –308/+118 and –104/+118, in RAW264.7 (Fig. 6B) and BV-2 cells (data not shown) indicated the presence of a negative regulatory element between –308 and –104. These findings locate the myeloid-specific core promoter region 104 bp proximal to the most upstream transcription start site.
The proximal Dap12 promoter region lacks a classical TATA box, initiator sequences, or GC-rich motif (Fig. 5) but contains purine-rich sequences characteristic for myeloid-specific promoters [32 ]. We therefore tested whether the two 5'-GGAA-3' binding motifs for PU.1 at –80 and –26 are required for Dap12 promoter activity. Mutant reporter gene constructs changing GGAA to TTTA at both sites were created by site-directed mutagenesis and analyzed by transient transfection. Mutation of the upstream PU.1 site [PU.1(1)] reduced the activity of the full-length promoter by 70% while mutating the PU.1(2) site-diminished promoter activity by over 85% (Fig. 6B) . Modifying both PU.1 sites in the same construct had a similar effect, resulting in an over 80% loss of reporter activity. These results demonstrate that both PU.1 sites are essential for the activity of the Dap12 promoter in myeloid cells.
PU.1 interacts with the Dap12 proximal promoter in vitro and in vivo
To study in vitro binding of PU.1 to the functionally important Ets/PU.1 motifs, EMSA analysis with nuclear extracts from RAW264.7 and BV-2 cells as well as IVT-PU.1 were performed. As shown in Figure 7
, prominent PU.1 binding activity was detected for the –90/–68 PU.1(1) site and the –37/–15 PU.1(2) site (Lanes 2). Complex formation at both sites was abolished completely by competition with an excess of unlabeled oligonucleotides containing the wild-type motif but not a mutated PU.1 site (Fig. 7
, A, Lanes 3 and 4, and B, Lanes 4 and 5). Each DNA–protein complex was confirmed to contain PU.1 by supershift analysis with a polyclonal anti-PU.1 antibody, resulting in a slowly migrating, high molecular complex (Fig. 7
, A, Lane 5, and B, Lanes 3 and 7). Furthermore, IVT-PU.1 strongly bound to the PU.1(1) and PU.1(2) sequences in a dose-dependent manner, whereas the IVT-negative control did not show specific binding (Fig. 7
, A, Lanes 7–10, and B, Lanes 8–11). Complex formation was similar when using RAW264.7 nuclear extract or BV-2 extract (Fig. 7B
, Lanes 6 and 7), which express high levels of PU.1 mRNA and protein. It is notable that the band at the PU.1(1) site was not supershifted completely by PU.1. Addition of an anti-Ets1/2 antibody diminished complex formation significantly, suggesting the presence of Ets1/2 protein in this complex. These results suggest that PU.1 is the only binding protein at site –37/–15 and that the region –90/–68 might contain an additional component, most likely Ets1 or Ets2.
We also analyzed the evolutionarily conserved VDR/RXR binding element located between both PU.1 sites using EMSA (data not shown). However, no specific binding to this Dap12 promoter region could be detected in RAW264.7 or BV-2 cells, indicating that myeloid expression of Dap12 is not mediated via this sequence motif.
By ChIP analysis, we next investigated whether PU.1 interacts with the proximal Dap12 promoter region in vivo. Chromatin was prepared from RAW264.7 cells, which expressed highest levels of Dap12 and PU.1, and immunoprecipitation was achieved with specific antibodies against PU.1 and IgG isotype as a control. After reversal of cross-linking, real-time qPCR analysis was carried out on ChIP–DNA using primers flanking the PU.1 sites in the Dap12 proximal promoter. The Egr1 gene was amplified as a genomic control, as it has no obvious PU.1 binding motif in its regulatory region. As shown in Figure 8A , the Dap12 core promoter region was strongly and specifically immunoprecipitated by anti-PU.1 antibody (Fig. 8 , left bar), whereas the IgG isotype control shows only background binding (right bar). We further studied in vivo binding of PU.1 to the Dap12 regulatory region using ChIP-Chip analysis. Mouse promoter tiling arrays were hybridized with DNA from PU.1 and IgG immunoprecipitations, respectively, and 5 kb upstream and 1 kb downstream of the Dap12 gene were analyzed for significant hybridization signals. Figure 8B displays that four adjacent probes located in the proximal Dap12 promoter region exhibit strong PU.1-specific signals compared with the IgG control (red bars). No additional significant signals were detected further upstream or in the first intron of the Dap12 locus. Taken together, both ChIP assays provide clear evidence for in vivo binding of PU.1 to the proximal Dap12 promoter in myeloid cells.
 View larger version (13K): [in a new window] |
Figure 8. In vivo binding of PU.1 to the proximal Dap12 promoter. Chromatin prepared from RAW264.7 cells was immunoprecipitated with antibodies against PU.1 ( PU.1) or IgG isotype control. (A) qPCR was performed with specific primers against the proximal Dap12 promoter region, and an internal control region from the Egr1 promoter was used as reference to calculate relative enrichment. The data shown are mean values of two independent experiments measured in triplicates, and error bars indicate the SD of the mean. (B) ChIP-Chip analysis of the Dap12 upstream regulatory region. Log2 expression values of tiled probes spanning 1 kb upstream of the first Dap12 exon are depicted. Significant probes are shown in red; nonsignificant probes are shown in blue. Specific in vivo binding of PU.1 to the proximal Dap12 promoter region spanning both PU.1 sites is indicated by four adjacent, significant probes.
|
 View larger version (36K): [in a new window] |
Figure 9. PU.1 knockdown down-regulates endogenous Dap12 expression. RNA interference of PU.1 in RAW264.7 cells was carried out with two independent siRNAs against PU.1 (PU.1_1 and PU.1_2), GAPDH, and a nonsilencing, scrambled siRNA. Transfected cells were harvested after 48 h, and qRT-PCR analysis for GAPDH (positive control), PU.1, and Dap12 was carried out to quantify the knockdown effect. Results are displayed as relative gene expression compared with untransfected cells. The data shown are mean values of three independent experiments, and error bars indicate the SD of the mean. **, P < 0.01; *, P < 0.05, specific siRNA versus scrambled siRNA using Student’s t-test.
|
 View larger version (14K): [in a new window] |
Figure 10. PU.1 re-expression and activation in PU.1–/– cells up-regulate endogenous Dap12 expression. PUER cells were cultured in the presence or absence of 100 nM OHT for the indicated time-points, and Dap12 mRNA levels were determined using qRT-PCR. Results are displayed as relative gene expression levels compared with noninduced PUER cells. The data represent mean values of two independent experiments, and error bars indicate the SD of the mean. **, P <0.01, OHT-induced versus noninduced using Student’s t-test.
|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Our results provide the first report of whole genome transcript profiling in ex vivo microglia from a murine retinal degeneration model. In a pioneering work, Roque and Caldwell [33 ] established isolation and culture of retinal microglia from 8-week Royal College of Surgeons rats and performed a morphological, phenotypic characterization of these cells. In our initial experiments to transfer this method to 14-day-old Rs1h–/Y and wild-type mice, we were not successful in obtaining larger amounts of highly pure cells. We therefore modified and adapted a protocol initially used for the isolation of colonic macrophages [27 ]. As a crucial step, Percoll separation and prolonged culture in M-CSF-supplemented medium were required to achieve a highly enriched microglia population without significant contaminations of other retinal cell types. It is remarkable that ramified and amoeboid cell shapes were maintained over the culture period, and the gene expression profiles of resting and activated cells could be clearly distinguished. Also, a significant overlap to transcript patterns found in total retinae from Rs1h–/Y and wild-type mice [8 ] was established, demonstrating an excellent comparability of the cultured microglia to the in situ situation in the retina. In addition to our data, the power of DNA microarray analysis to study microglia/macrophage activation in the nervous system has been shown recently by Albright and Gonzalez-Scarano [34 ]. In their study, specific expression of chemokines and surface receptors, which partially correspond to our microarray data, could be identified in mixed glial cultures of HIV-induced, degenerating brains. However, as microglia constitute only 60% of the mixed glial cultures [34 ], these data may still be biased by the presence of mRNA molecules of nonmyeloid cell types. An elegant solution to obtain highly pure microglia from neuronal tissues without further selective culture would be the use of transgenic mice with fluorescent markers in the microglia population for cell sorting and isolation. Pursuant to this idea, breeding of Rs1h–/Y mice with MacGreen mice expressing enhanced GFP under a macrophage-selective promoter [35 ] is currently underway in our laboratory. This will greatly facilitate the isolation and further characterization of activated microglia cells in retinal degeneration.
Our DNA microarray and qRT-PCR profiling data have shown significant mRNA induction and a potential functional role of Dap12 in activated microglia cells from degenerating Rs1h–/Y retinae. It is interesting that as an adaptor protein, Dap12 is known to cooperate with several cell-surface receptors to control leukocyte functions. Dap12-interacting receptors include members of the TREM family [11 , 36 ]. Stimulation of one of these receptors, TREM2, causes Dap12 phosphorylation and increased migratory and phagocytic activity of brain microglia. Conversely, knockdown of TREM2 leads to impaired clearance of apoptotic neurons and increased secretion of proinflammatory cytokines [16 ]. Showing altered synaptic functions in the brain of Dap12-deficient mice, Roumier et al. [15 ] also identified predominant expression of Dap12 in amoeboid microglia during hippocampal development. Thus, Dap12 and its partner TREM2 may fulfill a regulatory role in microglia to control phagocytosis of dying neurons without strong concomitant immune activation. In support of this hypothesis, Hamerman et al. [12 ] and Turnbull et al. [13 ] found that Dap12 can attenuate macrophage functions and cytokine production in response to TLR agonists such as LPS. We have also found a dramatic increase of TREM2 mRNA levels in activated microglia by DNA microarray analysis. However, whether Dap12 and TREM2 have an anti-inflammatory and microglia-controlling function in the retina remains to be determined. To address this question will be especially important, as Dap12 might have activating and inhibitory functions, depending on selective associations with other receptors and/or receptor/ligand avidities caused by different triggers.
In activated microglia, we have identified a PU.1-dependent regulatory network consisting of 16 genes including Dap12. In addition to the bioinformatic evidence of PU.1 binding sites in the conserved promoter regions of a number of species, several lines of evidence from our in vitro and in vivo promoter assays suggest that the Dap12 promoter is a direct target for PU.1. Supporting our data, Henkel et al. [37 ] found that retroviral re-expression of PU.1 into PU.1 null monocytic precursor cells leads to a strong and fast induction of Dap12 mRNA levels as identified by substractive hybridization. There was no detectable expression of Dap12 in PU.1-deficient cells, and the specific rescue of PU.1 was sufficient to restore Dap12 expression, indicating direct interaction of PU.1 with the Dap12 promoter. Furthermore, similar cellular phenotypes and antigen expression of mature macrophages in ectopic Dap12- and PU.1-expressing myeloid progenitors indicate that Dap12 may be a target of PU.1 in the control of macrophage maturation and activity [14 ].
In addition to our mRNA data showing increased expression of PU.1 in activated microglia from Rs1h–/Y retinae, little information is available about PU.1 expression in microglia. In line with our observations, high expression levels of PU.1 immunoreactivity were identified in the hemispheres of hypoxia-injured hippocampus of rats [38 ]. It is notable that the authors speculate that the increased PU.1 protein levels in the infarcted cortex may be more likely a result of microglia proliferation than activation. However, gene expression patterns of PU.1 target genes and activation markers would be useful to fully understand the role of PU.1 in hypoxia-induced brain injury. Based on our data and the data from the literature, we propose that PU.1 is critical for the myeloid-specific transcriptional regulation of the murine Dap12 gene. Moreover, as indicated by our gene expression profiles, PU.1 may also play a general role in the regulation of microglia genes during retinal degeneration. Therefore, further characterization of the PU.1 regulatory network in microglia cells is required to better understand the molecular mechanisms of microglia activation in degenerative disorders.
Received July 5, 2007; revised August 9, 2007; accepted August 15, 2007.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
inhibitor, on TLR-4-dependent activation of murine microglia Biochem. Biophys. Res. Commun. 337,110-115[CrossRef][Medline] RIIIA promoter EMBO J. 13,3852-3860[Medline]This article has been cited by other articles:
 |
|
 |
|
  T. Langmann, C. C. L. Lai, K. Weigelt, B. M. Tam, R. Warneke-Wittstock, O. L. Moritz, and B. H. F. Weber CRX controls retinal expression of the X-linked juvenile retinoschisis (RS1) gene Nucleic Acids Res., November 1, 2008; 36(20): 6523 - 6534. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Ebert, T. Schoeberl, Y. Walczak, K. Stoecker, T. Stempfl, C. Moehle, B. H. F. Weber, and T. Langmann Chondroitin sulfate disaccharide stimulates microglia to adopt a novel regulatory phenotype J. Leukoc. Biol., September 1, 2008; 84(3): 736 - 740. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
