Published online before print October 30, 2007
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
,
,1,2,,2,,,,
,,,3
* Institute for Molecular Bioscience,
Cooperative Research Centre for Chronic Inflammatory Diseases (CRC-CID),
ARC Special Research Centre for Functional and Applied Genomics, and
School of Molecular and Microbial Sciences, University of Queensland, Brisbane, Queensland, Australia
1 Correspondence: Institute for Molecular Bioscience, Bldg. 80, University of Queensland, Brisbane, Queensland 4072, Australia. E-mail: d.ovchinnikov{at}imb.uq.edu.au
 TOP ABSTRACT REFERENCES |
|---|
Key Words: myeloid lineage • macrophages • Langerhans cells • Gal4-UAS system
Cells of the mononuclear phagocyte system (MPS), including blood monocytes, embryonic and tissue macrophages, as well as dendritic cells, play pivotal roles in both innate and acquired immunity [1 ]. While a number of transgenic mice have been developed to enable gene inactivation or repression in myeloid cells [2 , 3 ], no tools for gene overexpression have been created. The binary Gal4-UAS system has been widely used in model organisms from Drosophila to zebrafish, and successfully trialed in mice [4 5 6 ]. We used a novel approach for the generation of transgenics for tissue-specific and easily traceable overexpression of genes, by generating a transgene that consists of both Gal4-expressing and Gal4-reporting modules. We chose the promoter of the mouse Csf1r (c-fms) gene to direct Gal4 expression, since we have previously successfully used it to drive transgene expression in the myeloid lineage [7 ]. Here, we describe the generation of a transgenic mouse line in which mononuclear phagocytes and other cells of the myeloid lineage are labeled with high levels of ECFP fluorescence, and the use of this binary system to produce targeted overexpression of an epitope-tagged gene of interest selectively in myeloid cells.
The promoter of the Csf1r gene, in addition to its activity in the myeloid lineage, was found to be expressed in some placental structures, primarily in the giant trophoblast cells at the interface between maternal and embryonic tissues, and the parietal endoderm [7 , 8 ]. To prevent undesirable expression of the transgene in trophoblast cells, we deleted the trophoblast-specific Csf1r promoter (Ptro, Fig. 1A ), located directly upstream of the macrophage-specific promoter (Pmac). The deleted bases -454 to -298 (relative to the Csf1r ATG) of the 7.2 kb Csf1r fragment encompassed the previously identified trophoblast-specific transcription start sites [7 ] and a TATA-box, located upstream in close proximity of those sites. This deletion did not compromise the expression of either EGFP or the luciferase reporters, driven by this promoter, in stable transfectants of RAW264.7 macrophage-like cell line (data not shown). The modified promoter was then cloned upstream of the Gal4VP16 transcriptional activator [9 ] and SV40 early polyadenylation site, and the functionality of the generated "Gal4 driver" construct was confirmed by cotransfection with the "Gal4-responder" plasmid (expressing EGFP in a Gal4-dependent manner) into the RAW264.7 macrophage-like cell line. High levels of EGFP were observed in cells containing both plasmids, while the Gal4 responder plasmid alone did not produce any detectable fluorescent reporter (data not shown). ECFP was chosen as a reporter for in vivo use since it has spectral characteristics most distinct from the commonly used EGFP and EYFP fluorescent reporters and allows for visualization in tissues with minimal background, thus broadening the range of future combinatorial applications of the transgene. Both EGFP and ECFP Gal4-responder plasmids were generated by cloning the corresponding fluorescent protein’s coding region from pExFP-N1 (BD Biosciences, San Diego, CA, USA) into pGene/V5-His B plasmid (Invitrogen, Carlsbad, CA, USA).
 View larger version (41K): [in a new window] |
Figure 1. Csf1r-Gal4VP16/UAS-ECFP (MacBlue) transgene and its expression in the embryo. (A) Schematic diagram depicting Csf1r-Gal4VP16 construct, expressing Gal4 protein binding as a dimer to each of the 6 upstream activating sequences (UAS) to drive ECFP reporter transcription. (B) Brightfield and ECFP fluorescence images of an 8 dpc transgenic embryo (still in its yolk sac), showing expression in forming blood islands. All - allantoins, s - somite. (C) ECFP+ yolk sac macrophages at 10.5 dpc. v-yolk sac blood vessel. (D) ECFP+ macrophages in a 9 dpc transgenic embryo, where they appear more concentrated in the anterior third of the embryo. (E) 10.5 dpc Csf1r-Gal4VP16 transgenic embryo. By this stage, ECFP+ macrophages are evenly distributed throughout the embryo, while the highest concentration is apparent in the forming liver. (F) A hindlimb (anterior is up) of a 12.5 dpc transgenic embryo, showing ECFP+ phagocytes concentrating in the anterior and interdigital necrotic zones, while being excluded from the forming digital cartilaginous condensations. ECFP fluorescence images were captured using a Leica MZ16FA dissecting scope (Leica, Wetzler, Germany) equipped with a Leica DFC480 CCD camera using Leica IM400 (v. 4.0), and adjusted to recapitulate the observer’s perception. Original magnifications: x40.8 (B), x110 (C), x35.3 (D), x15.3 (E), x35.4 (F).
|
Blood smears from one of the founders carrying both Gal4 and ECFP cointegrated transgenes contained 
5% ECFP+ leukocytes, and this line was chosen for full characterization. Transmission rate of the cointegrated transgenes (which always cosegregated, n > 500) in crosses with wild type (WT) animals was statistically indistinguishable from expected, indicating that these transgenes had no effect on embryonic viability. Studies were carried out on animals of a mixed CD1/C57BL6 background.
ECFP expression in transgenic embryos was first detectable at 8 dpc (Fig. 1B) in the blood islands of the proximal yolk sac, where the first appearance of committed macrophage progenitors of embryonic origin has been reported [10 ]. From 9 dpc, a network of macrophages could be seen throughout the yolk sac, some associated with forming blood vessels (Fig. 1C) with the first appearance of macrophage-like cells in the embryo proper (Fig. 1D) . At 10.5 dpc, in addition to diffusely distributed macrophages, a higher concentration of ECFP+ cells was observed in the liver (Fig. 1E) , which is infiltrated by hematopoietic progenitors and stem cells at this time [11 ]. From this time on, the density of fluorescent macrophages increased in the embryo, while no expression was observed in the placenta or giant trophoblast cells, indicating success in deleting the trophoblast-specific promoter (see supplementary data). ECFP+ macrophages were concentrated in areas of increased apoptotic cell death, such as necrotic zones of the developing limb, consistent with their proposed scavenger function [12 ] (Fig. 1F) . This expression pattern closely mimics that of known embryonic macrophage-specific genes: Csf1r and Csf1r-EGFP transgene [7 ], lysozyme, PU.1, and macrophage mannose receptor [13 ].
The removal of the trophoblast promoter did not mitigate expression in tissue macrophages. Screening of multiple tissues revealed similar populations of stellate macrophages to the MacGreen mouse [7 ] (Fig. 2 and data not shown), visualized more readily because of the low background fluorescence observable through ECFP filter set. Figure 2A shows expression of ECFP in macrophages of the spleen (colocalization with F4/80 on sections is also shown in Fig. 2F ). Within the white pulp, in keeping with expression of EGFP in MacGreen mice [7 , 14 ], ECFP was present in F4/80-negative interdigitating dendritic cells (Fig. 2F) . The high levels of ECFP allowed visualization of intricate architecture of Langerhans cells of the skin (Fig. 2B , also see the supplemental data), considered to be precursors of myeloid antigen-presenting dendritic cells [1 ]. Reconstruction of the three-dimensional structure of these cells (performed in an unfixed skin in isotonic solution) revealed that most of the dendritic projections are directed toward and terminate at the plane of the epithelial surface (Fig. 2B and supplemental data).
 View larger version (40K): [in a new window] |
Figure 2. Csf1r-Gal4VP16/UAS-ECFP expression and functionality. (A) Bright-field and ECFP fluorescent image of the spleen from a 6-wk-old Csf1r-Gal4VP16/UAS-ECFP transgenic. The highest concentration of ECFP+ cells is observed in the marginal zones around the white pulp nuclei. (B) 3D reconstruction of two Langerhans cells (from the ear skin) using confocal microscope imaging. The cells’ processes are projected away from the viewer, toward the plane coinciding with the epithelial surface, where most of them terminate (C). FACS analysis of ECFP-expressing populations in peripheral blood. All cells expressing high levels of ECFP are F4/80+. (D) Western blot analysis showing expression of the Schlafen4-V5 protein (of expected size of 70 kDa) in bone marrow of double transgenics (carrying both Csf1r-Gal4/UAS-ECFP and UAS-Schlafen4-V5 transgenes), but not in single UAS-Schlafen4-V5 transgenic animal. Ribosomal S6 protein is used as a loading control. (E) Confocal imaging of BMMs from Gal4 and Gal4+UAS-Schlafen4-V5 transgenic animals. Note that Schlafen4-V5 protein is excluded from the nucleus. Red, phalloidin staining for actin; blue, ECFP; green, Schlafen4-V5 protein. In Schlafen4-V5-expressing cells, cytoplasmic ECFP fluorescence is masked by high levels of tagged protein expression. (F) Cryosections of spleens of single (Csf1r-Gal4/UAS-ECFP) and double (Csf1r-Gal4/UAS-ECFP;UAS-Schlafen4-V5) transgenic mice, showing specificity of the V5-tagged Schlafen4 protein expression. Notice that most F4/80+ macrophages with characteristic stellar morphology (arrows) express Schlafen4-V5. Like the ECFP, Schlafen4-V5 was also detectable in F4/80– interdigitating dendritic cells in lymphoid areas (arrowheads). No V5-tagged protein could be detected in spleens of UAS-Schlafen4-V5 single transgenic animals (not shown). RP, red pulp; WP, white pulp. Original magnifications: x9.8 (A), x60 (B), x100 (E), x20 (F).
|
10 times lower levels than monocytes (Fig. 2C)
. In peripheral blood, all ECFP+ cells (5% of total leukocytes) were CD11b-positive, and all were negative for lymphocyte markers, e.g., B220 (Fig. 2C
, and data not shown). All ECFPhi leukocytes expressed F4/80, a definitive monocyte/macrophage marker [1
] (Fig. 2C)
. In bone marrow, around 7% of cells appeared ECFP+‘. Antibody staining and FACS analysis was performed as described [7
]. Antibodies used were anti-F4/80-R-PE (Serotec, Oxford, UK), anti-Ly-6G-R-PE (BD PharMingen, San Diego, CA, USA), and rat anti-CD45R/B220 (BD PharMingen) combined with Phycoerythrin (PE) conjugated anti-rat F(ab’)2 (Cell Sciences, Canton, MA, USA). Analysis was performed on a BD LSRII flow cytometer (BD Biosciences), and data were processed and presented using FlowJo software (www.FlowJo.com). To test the ability of the MacBlue transgenic mouse line to drive overexpression of a Gal4-dependent transgene, we crossed it to a transgenic mouse designed to express V5-tagged mouse Schlafen4 protein (a macrophage-specific protein of unknown function [15 ]) from a Gal4-inducible promoter. Expression of a V5-tagged protein of expected size was detected (using mouse anti-V5 (Serotec) and horse anti-mouse IgG-HRP antibodies (Cell Signaling Technology, Danvers, MA, USA) in bone marrow from animals carrying both the Gal4 and UAS-Schlafen4-V5 transgenes (Fig. 2D) . Expression of Schlafen4-V5 was detected in the majority of ECFP+ BMMs from MacBlue; UAS-Schlafen4-V5 animals (Fig. 2E) . The observed granular cytoplasmic subcellular localization of the protein recapitulated the pattern seen in transfected RAW264.7 cells (data not shown). Expression of Schlafen4-V5 could also be readily detected in ECFP+ tissue macrophages—for instance spleen macrophages—as corroborated by costaining for F4/80, and in some F4/80– dendritic cells of the spleen’s white pulp (Fig. 2F) . Schlafen4-V5 protein was visualized using rabbit anti-V5 (Sigma, St. Louis, MO, USA) and goat anti-rabbit-Alexa488 (Molecular Probes, Eugene, OR, USA) antibodies as in [16 ]. Confocal mages were taken with a Zeiss LSM 510 META (Zeiss, Oberkochen, Germany) confocal microscope, using a blue diode laser, an Argon/2 laser and a HeNe1 laser at 405 nm, 488 nm, and 543 nm, respectively. RAW264.7 and BMM culture was performed as described [7 ].
In summary, we have generated and validated a new transgenic tool that allows both specific labeling of cells of the mononuclear phagocyte system (during embryonic development and adulthood) with high levels of ECFP expression, and overexpression of Gal4-dependent transgenes in those cells. Although a number of protein markers specific for cells of the MPS have been well established [17 ], so far, no mouse tools allowing gene manipulation in the MPS and not other cells of the myeloid lineage, primarily neutrophilic granulocytes, have been generated [2 , 3 ]. We have recently performed comparative microarray profiling of macrophages and neutrophilic granulocytes, which failed to identify any macrophage-restricted transcripts, and by inference-promoters [18 ]. That study also revealed that the Csf1r-EGFP (MacGreen) transgene was expressed in all of the granulocytes. Most likely, explanations for the greatly reduced expression of the MacBlue transgene in granulocytes are that the trophoblast promoter region contributes to expression in those cells, or that Gal4-VP16 is not translated or is less active in those cells. Regardless of the reasons, the exclusion from granulocytes, as well as trophoblasts, augments the value of the largely mononuclear phagocyte-restricted transgenic mouse tool we have generated.
MacBlue (Csf1r-Gal4VP16/UAS-ECFP) mice are available to researchers on request from D.A.O. (d.ovchinnikov{at}imb.uq.edu.au, IMB, Brisbane, Australia) or D.A.H. (david.hume{at}bbsrc.ac.uk, The Roslin Institute, Edinburgh, UK).
3 Current address: The Roslin Institute, University of Edinburgh, Roslin EH25 9PS, UK.
Received August 31, 2007; revised October 8, 2007; accepted October 8, 2007.
|
TOPABSTRACTREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
