Published online before print February 13, 2008
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
that modulate neutrophil activation
* Davis Heart and Lung Research Institute and Division of Hematology/Oncology,
Division of Cardiovascular Medicine, Department of Internal Medicine, and
Division of Pulmonary and Critical Care Medicine, The Ohio State University College of Medicine, Columbus, Ohio, USA
1Correspondence: Davis Heart and Lung Research Institute and Division of Hematology/Oncology, The Ohio State University College of Medicine, B311 Starling-Loving Hall, 320 West Tenth Avenue, Columbus, OH 43210-1240, USA. E-mail: belinda.avalos{at}osumc.edu
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
2 glycoprotein (LRG), a member of the leucine-rich repeat family of proteins, was recently shown to be up-regulated during neutrophil differentiation. Its precise role in granulopoiesis, however, remains unknown. In this paper, we show that the transcription factors PU.1 and C/EBP
that regulate the expression of multiple myeloid-specific genes also bind to the LRG promoter. We also demonstrate that LRG localizes to the same cytoplasmic compartment as myeloperoxidase and that G-CSF treatment of the 32Dcl3 myeloid cell line induces nuclear translocation of LRG. Stable transfection of LRG into 32Dcl3 cells resulted in accelerated, G-CSF-mediated neutrophil differentiation and induction of CD11b expression. In contrast, constitutive expression of LRG in 32Dwt18 cells, expressing a chimeric erythropoietin (Epo)/G-CSFR consisting of the EpoR extracellular domain fused to the G-CSFR transmembrane and cytoplasmic domains, failed to induce accelerated neutrophil differentiation and CD11b expression in response to Epo stimulation. LRG-mediated accelerated differentiation and CD11b expression were found to correlate with an increased level of phospho-Stat3 but not with PU.1 or p27kip1 levels. Hence, similar to other genes involved in neutrophil differentiation, the expression of LRG also appears to be regulated by PU.1 and C/EBP. Collectively, these findings suggest a role for LRG in modulating neutrophil differentiation and expression of CD11b via nonredundant G-CSFR signals.
Key Words: granulopoiesis • transcription factors
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
2 glycoprotein (LRG) was purified from human serum more than 25 years ago [1
], establishing a novel family of proteins characterized by the presence of leucine-rich repeats in their amino acid sequences. Although several family members have been shown to be involved in signal transduction, protein–protein interaction, or cell adhesion and development [2
], a function for LRG has remained elusive. Consensus binding sites for several myeloid-specific transcription factors are present within the putative promoter region of the gene for human LRG, and its expression was previously shown by our laboratory to be up-regulated during neutrophilic differentiation [3
]. Notably, the gene for human LRG localizes to the same region on chromosome 19p13.3, to which the genes for multiple primary neutrophil granule enzymes also localize. A specific role for LRG in granulopoiesis has not been elucidated. G-CSF has been shown to be the major regulator of granulopoiesis [4 ] and supports not only the survival, proliferation, and neutrophilic differentiation of myeloid progenitor cells but also activates certain functions of mature, terminally differentiated neutrophils [5 6 7 ]. The biological activities of G-CSF are mediated by specific receptors on the surface of responsive cells. G-CSF binds to the extracellular portion of the G-CSFR, resulting in activation of a complex signaling cascade that includes the Jak and Stat kinases. Expression of the G-CSFR, like many other myeloid-specific genes, including genes for several neutrophil granule proteins, has been shown to be regulated by PU.1 and the C/EBP family of transcription factors.
Studies in mice with knockout or knock-in mutations in the G-CSFR gene suggest that the G-CSFR generates unique, nonredundant signals required for polymorphonuclear neutrophil (PMN) production and marrow egress to maintain homeostatic levels of circulating PMN [7 8 9 10 ]. G-CSFR knockout mice have chronic neutropenia with a uniform decrease in myeloid cells in the bone marrow. PMN from these mice also exhibit selective defects in activation [11 , 12 ]. These observations suggest that the G-CSFR is a critical regulator of PMN differentiation and activation.
Previous work by our laboratory to further characterize the mechanisms by which G-CSF promotes neutrophil differentiation using representational difference analysis resulted in the isolation of cDNA and genomic clones for human and murine LRG (mLRG). To further characterize the physiological role of LRG in granulopoeisis, we stably transfected the mLRG cDNA into 32Dcl3 cells endogenously expressing the wild-type murine G-CSFR and also into 32Dwt18 cells stably transfected with a chimeric erythropoietin (Epo)/G-CSFR and examined the effects of constitutive expression of LRG on their proliferation, differentiation, and activation. Previous investigators have used the 32Dwt18 cell line in place of 32Dcl3 cells to study myeloid maturation, as 32Dwt18 cells transferred to Epo-containing media do not undergo the early massive cell death observed when 32Dcl3 cells are transferred to G-CSF-containing media [13 ]. We report here significant differences in these cell lines and provide further evidence that unique signaling pathways are activated by the G-CSFR, which involves LRG. Our data also suggest a role for the extracellular domain of the G-CSFR in signal transmission.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
mLRG cDNA was cloned into the pcDNA 3.1D/V5-His-TOPO vector (Invitrogen, Carlsbad, CA, USA) in-frame with the V5 epitope tag and transfected into 32Dcl3 cells by electroporation (300V, 960 µF, single pulse). mLRG cDNA in-frame with the V5 epitope tag was also cloned into the pcDNA6/HisB vector (Invitrogen) and transfected into 32Dwt18 cells by electroporation as described above. Clones expressing mLRG were selected by growth in media containing G418 (300 µg/ml, for 32Dcl3 cell transfection) or Blasticidin (10 µg/ml, for 32Dwt18 cell transfection) at 48 h after transfection. Expression of mLRG mRNA in positive clones was confirmed by RT-PCR. To confirm expression of mLRG protein, Western blotting was performed on whole cell lysates (WCL) from positive clones with the anti-V5 antibody (Invitrogen).
Differentiation assay
Purified recombinant human G-CSF was generously provided by Amgen Inc. (Thousand Oaks, CA, USA) and purified human Epo (Procrit, epoitin alfa) by Ortho Biotech Inc. (Bridgewater, NJ, USA). 32Dcl3 and 32Dwt18 cells were removed from IL-3-containing media, washed twice in PBS to remove residual IL-3, and then transferred to G-CSF (10 ng/ml, for 32Dcl3 cells) or Epo (0.1–1 u/ml, for 32Dwt18 cells) containing media to induce neutrophilic granulocyte differentiation. At varying time-points, aliquots were removed for analysis of differentiation, which was monitored by Wright-Giemsa staining and flow cytometric analysis.
Flow cytometric analysis
Cells (1x106) were washed twice with cold HBSS solution supplemented with 1% BSA and 0.1% sodium azide, resuspended in 500 µl cold HBSS/BSA/Azide solution, and incubated with FITC-conjugated rat anti-mouse CD11b or FITC-conjugated rat anti-mouse CD13 antibody (PharMingen, San Diego, CA, USA) for 1 h at 4°C. After incubation, cells were washed twice with cold HBSS/BSA/sodium azide solution and fixed in 500 µl cold 1% paraformaldehyde solution. Cells were then analyzed by flow cytometry on a FACSCalibur (Becton Dickinson Immunocytometry Systems, San Jose, CA, USA) using CellQuest software.
Immunofluorescence staining and confocal microscopy
A total of 5 x 104 32Dcl3 cells per clone was spun onto noncharged slides and fixed in 4% formaldehyde. The fixed cells were washed in PBS and then permeabilized with 1% Triton X-100 in TBS for 30 min at room temperature. To minimize nonspecific antibody binding, the slides were incubated with 1% BSA in TBS for 1 h at room temperature. To detect expression of the V5-tagged mLRG protein, a murine anti-V5 antibody (Invitrogen) was used as the primary antibody in conjunction with the tetramethylrhodamine isothiocyanate (TRITC)-conjugated donkey anti-mouse F(ab)'2 secondary antibody (Jackson ImmunoResearch Laboratories, Inc., West Grove, PA, USA). To detect mouse myeloperoxidase (MPO), goat anti-human MPO antibody (cross-reactive with mouse, Santa Cruz Biotechnology Inc., Santa Cruz, CA, USA) and donkey anti-goat F(ab)'2 antibody conjugated with Cy5 (Jackson ImmunoResearch Laboratories, Inc.) were used. Hoescht (Molecular Probes, Eugene, OR, USA) was used for nuclear staining. ProLong anti-fade mounting media (Molecular Probes) were placed on each slide after staining. Slides were examined under confocal microscopy using a Zeiss LSM510 multiphoton confocal microscope. The excitation/emission spectra for the fluorophores used were as follows: Hoescht 360/415 nm; TRITC 550/570 nm; Cy5 633/660 nm.
Subcellular fractionation
Subcellular fractionation of 32Dwt18 transfectants was performed as described [14
]. Briefly, cells were washed with PBS, resuspended in cold Krebs-Ringer glucose buffer (130 mM NaCl, 5 mM KCl, 1.27 mM MgSO4, 0.95 mM CaCl2, 5 mM glucose, 10 mM NaH2PO4/Na2HPO4, pH 7.4, 5 mM PMSF), and incubated on ice for 5 min. Cells were then spun down, resuspended at 108cells/ml in disruption buffer [100 mM KCl, 3 mM NaCl, 3.5 mM MgCl2, 10 mM Piperazine N,N'-bis2(ethan-sulfonic acid), pH 7.2, 1 mM ATPNa2, 0.5 mM PMSF] with one complete mini-protease inhibitor cocktail tablet (Roche Diagnostic Corp., Indianapolis, IN, USA), and disrupted by nitrogen cavitation (pressurized under nitrogen for 5 min at 380 pounds per square inch). The cavitate was collected drop-wise into disruption buffer containing a final concentration of 0.5 mM EGTA. Nuclei and intact cells were pelleted by centrifugation at 400 g for 15 min. The postnuclear supernatant was applied onto a three-layer Percoll gradient solution (densities 1.050, 1.065, and 1.090 g/ml) and was centrifuged at 37,000 g for 30 min at 4°C. After centrifugation, six fractions of 5 ml each were collected. Percoll was removed by ultracentrifugation at 100,000 g for 90 min. Biological materials above the Percoll pellet were collected and were next subjected to Western blot analysis.
Western blot analysis
WCL from unstimulated and differentiated cells were prepared as described [15
]. Protein concentrations were determined using a BCA kit (Pierce, Rockford, IL, USA). For each sample, 50 µg protein was separated on 10% SDS-PAGE and the proteins transferred to nitrocellulose membranes, which were probed with the respective antibodies, and proteins were visualized by ECL (Amersham, Piscataway, NJ, USA). Antibodies to phospho-Stat3 (pStat3) and p27kip1 were purchased from Cell Signaling Technology Inc. (Beverly, MA, USA). Antibodies to Stat3, PU.1, and Raf-1 were purchased from Santa Cruz Biotechnology Inc. Antibody to V5 tag was purchased from Invitrogen. A rabbit polyclonal antibody to the p85 subunit of PI-3K was purchased from Upstate Biotechnology Inc. (Lake Placid, NY, USA).
Chromatin immunoprecipitation assay (ChIP)
ChIP assays were performed following the protocol for the acetyl-histone H4 ChIP assay kit (Upstate Biotechnology), using buffers supplied in the kit. Briefly, 5 x 106 cells were cross-linked with 1% formaldehyde for 30 min at 37°C. Prior to cross-linking, an aliquot of the cells was removed for analysis of input chromatin DNA. After cross-linking, cells were washed twice in cold PBS, resuspended in lysis buffer, and incubated on ice for 10 min. The samples were then subjected to sonification to shear protein-cross-linked DNA into fragments of 1–3 kb and diluted with immunoprecipitation buffer. After preclearing with a salmon sperm DNA/protein A agarose slurry, samples were then incubated with antibodies to PU.1 (4 µg), C/EBP (8 µg), or C/EBP (4 µg) or with normal rabbit IgG (4 µg) at 4°C overnight. All antibodies were purchased from Santa Cruz Biotechnology Inc. Samples containing no antibody were also included as controls. After sequential washing with low salt, high salt, LiCl, and 10 mM Tris–HCl,1 mM EDTA, pH 8, buffers, the protein–DNA complex was eluted with freshly made elution buffer (1% SDS, 0.1 M NaHCO3). Cross-linking was then reversed by addition of 5 M NaCl at 65°C for 4 h. After proteinease K treatment, DNA was isolated by phenol/chloroform extraction, followed by ethanol precipitation. Equal volumes of DNA products were used as PCR templates. Primers (forward: AATCCCCACCTCACCCTTAACTAG; reverse: CTATGTCTGTCCTCAAGACCCCAG) were designed to specifically amplify the sequences within the mLRG promoter corresponding to consensus binding sites for C/EBP, C/EBP, and PU.1 (nt 794–990, 197 nt product).
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
bind to the LRG promoter to the amplified sequence within the putative LRG promoter could be detected (Fig. 1
, lanes 5 and 9), which increased significantly after 3 days of culture in Epo-containing media to induce differentiation (Fig. 1
, lanes 6 and 10). In contrast, no significant changes in binding of C/EBP to the same region in the LRG promoter were observed during the same time period following transfer to Epo-containing media. As negative controls, reactions were included in which no antibody (Fig. 1
, lanes 1 and 2) or normal rabbit IgG (Fig. 1
, lanes 3 and 4) was added. As an additional negative control, PCR reactions were also performed in the absence of input DNA (Fig. 1
, lane 11). As a positive control, isolated DNA (Input, Fig. 1
, lane 12) was subjected to PCR amplification prior to immunoprecipitation using the same primers.
 View larger version (31K): [in a new window] |
Figure 1. ChIP indicated functional binding sites for PU.1 and C/EBP within the promoter region of the mLRG gene. 32Dwt18 cells were cross-linked with 1% formaldehyde, washed, and resuspended in lysis buffer. Samples were sonicated to shear the protein–DNA complexes and immunoprecipitated with antibodies for PU.1, C/EBP, and C/EBP. The complexes were washed and eluted, and the cross-linkage was reversed. Isolated DNA was then subjected to PCR using primer sets specific for the consensus-binding sites of PU.1, C/EBP, and C/EBP with the putative promoter region of the mLRG gene. Normal rabbit IgG (Rab IgG) was used as a negative control. No antibody control (No Ab) and input DNA control (Input) were also included. Diff, Differentiation.
|
 View larger version (20K): [in a new window] |
Figure 2. LRG exhibited the same cellular localization as that of MPO. 32Dcl3 cells stably transfected with V5 epitope-tagged mLRG or vector alone were cytospun onto slides and fixed with 4% formaldehyde. LRG and MPO were detected by mouse anti-V5 and goat anti-human MPO primary mAb, which were then detected using TRITC-conjugated donkey anti-mouse IgG and Cy5-conjugated donkey anti-goat IgG secondary antibodies, respectively. Nuclei were stained with Hoescht. (A) Cellular localization of LRG in LRG-transfected cells. (B) Cellular localization of MPO in LRG transfectants. (C) Nuclear staining with Hoescht in LRG transfectants. (D) Merged figure of A–C. (E) Cells transfected with vector alone only stained positively for MPO. (F) LRG transfectants stained with secondary antibodies only.
|
 View larger version (26K): [in a new window] |
Figure 3. Analysis of subcellular fractions from 32Dwt18 transfectants. 32Dwt18 cells stably transfected with V5-tagged mLRG (undifferentiated, A; differentiated, B) or with empty vector only (Vec; C) were resuspended in disruption buffer and disrupted by nitrogen cavitation. The nuclei and intact cells were pelleted by centrifugation, and the resultant postnuclear supernatants were loaded onto Percoll gradients. After centrifugation, six continuous fractions were collected. After ultracentrifugation to remove Percoll, the biological materials were collected and were then subjected to SDS-PAGE analysis. Samples were immunoblotted with an antibody to the V5 epitope (upper panels) or murine MPO (lower panels).
|
50 kDa on SDS-PAGE. Three single clones with the highest LRG expression levels were used for the differentiation assay. A mixed pool of three single clones transfected with empty vector only was also used as control. Following initial transfer from IL-3- to G-CSF-containing media, massive cell death was repeatedly observed in five independent experiments. Cell viability decreased by more than 85% within 24 h. Surviving cells were observed to slowly recover and consequently, undergo terminal differentiation into mature neutrophils. Notably, LRG transfectants displayed higher viability compared with cells transfected with vector alone, despite the massive cell death (data not shown), suggesting that constitutive expression of LRG increased the survival of cells after initial transfer to G-CSF-containing media.
Constitutive expression of LRG also significantly accelerated the differentiation of 32Dcl3 cells into mature neutrophils. Morphological analysis of Wright-Giemsa-stained cells demonstrated that by Day 7 of culture with G-CSF, more than 50% of LRG transfectants displayed segmented nuclei consistent with the nuclear morphology of mature neutrophils (Fig. 4A and 4B ). In contrast, only 5–10% of cells transfected with vector alone exhibited segmented nuclei (Fig. 4A and 4B) . To confirm neutrophil maturation, we examined the surface expression level of CD11b (membrane-activated complex-1), a marker for neutrophil differentiation, by flow cytometry. Compared with untransfected, parental 32Dcl3 cells, there was a clear-cut log shift to the right in LRG-transfected cells (Fig. 4C) , and accelerated expression of CD11b was detected in these cells. Taken together, these results indicate that constitutive expression of LRG accelerates neutrophil differentiation in 32Dcl3 cells.
 View larger version (40K): [in a new window] |
Figure 4. Accelerated neutrophilic differentiation in LRG-transfected 32Dcl3 cells in response to G-CSF treatment. (A) Cells stably transfected with vector alone or mLRG cDNA were washed out of IL-3 and transferred to G-CSF-containing media. At indicated time-points, aliquots of the cell were cytospun onto slides. Neutrophilic differentiation was monitored by Wright-Giemsa staining. (B) Bar graphs indicate the fraction of the total cell population of cells transfected with vector alone and the LRG clone at each stage of differentiation at the indicated time-points. (C) Aliquots of cell were stained with FITC-conjugated CD11b antibody and subjected to FACS analysis.
|
Initially, we used a high concentration of Epo (1 u/ml) as previously used to study the CCAAT displacement protein (per cut) in differentiating 32Dwt18 cells [13 ]. Massive cell death was not observed after transfer of the cells from IL-3 to Epo-containing media. However, at this concentration, we did not observe accelerated neutrophil differentiation in LRG-transfected clones. As a result of concerns that any accelerated differentiation might be masked by strong Epo signaling at the high concentration, we reduced the Epo concentration to as low as 0.1 u/ml. At the lower concentration of Epo, we observed decreased viability of the cells similar to that observed with 32Dcl3 cells following their transfer to G-CSF-containing media. Notably, the LRG-transfected clones displayed a higher viability than cells transfected with empty vector alone (data not shown). However, even at the lower concentration of Epo, there was no evidence of accelerated differentiation in LRG transfectants by morphology (Fig. 5A and 5B ) or by CD11b expression (Fig. 5C) . These results suggest different signaling pathways are activated in response to G-CSF and Epo in 32Dcl3 and 32Dwt18 cells, respectively, during neutrophilic differentiation.
 View larger version (39K): [in a new window] |
Figure 5. Lack of accelerated neutrophilic differentiation in LRG-transfected 32Dwt18 cells in response to Epo treatment. (A) Cells stably transfected with vector alone or mLRG cDNA were washed out of IL-3 and transferred to Epo-containing media. At the indicated time-points, aliquots of cells were cytospun onto slides and neutrophilic differentiation monitored by Wright-Giemsa staining. (B) Bar graphs indicate the fraction of the total cell population transfected with vector alone or LRG at various stages of neutrophilc differentiation at each time-point. (C) Aliquots of cells were stained with FITC-conjugated CD11b antibody and analyzed for CD11b expression by FACS analysis.
|
We first investigated activation of Stat3, which has been reported to be required for G-CSF-induced neutrophilic differentiation [17 ]. In 32Dcl3 cells, we detected a rather weak signal for pStat3 in cells transfected with vector alone and cultured for 24 h in G-CSF. In contrast, we detected a strong signal for pStat3, which was apparent as early as 8 h in LRG transfectants. Notably, the expression level of Stat3 remained the same throughout this time (Fig. 6A ), suggesting that Stat3 may be involved in LRG-accelerated neutrophilic differentiation of 32Dcl3 cells. In 32Dwt18 cells, we detected similar patterns of Stat3 and pStat3 in vector-transfected cells and LRG transfectants cultured in Epo-containing media (Fig. 6B) . This may explain why we did not observe accelerated, neutrophilic differentiation of these cells following stable transfection with LRG. We also examined the 32Dwt18 cells in G-CSF-containing media. To our surprise, we did not detect any pStat3 signals, even as late as 24 h following G-CSF treatment (Fig. 6C) . In addition, the expression level of Stat3 significantly decreased at 24 h in G-CSF in LRG-expressing cells and cells transfected with vector alone (Fig. 6C) . This suggests that the endogenous, full-length G-CSFR may not be functional in 32Dwt18 cells.
 View larger version (23K): [in a new window] |
Figure 6. Stat3 is activated at an early stage in LRG-transfected 32Dcl3 cells but not in LRG-transfected 32D wt18 cells. Cells were washed out of IL-3-containing media and transferred to media containing G-CSF or Epo as indicated. WCL were made from cells harvested at the indicated time-points and used for immunoblotting with Stat3 and pStat3. (A) Vector or LRG-transfected 32Dcl3 cells in G-CSF-containing media. (B) Vector or LRG-transfected 32Dwt18 cells in Epo-containing media. (C) Vector or LRG-transfected 32Dwt18 cells in G-CSF-containing media.
|
 View larger version (26K): [in a new window] |
Figure 7. p27kip1 and PU.1 display similar patterns of expression in LRG-transfected 32Dcl3 and 32Dwt18 cells, which were washed out of IL-3-containing media and transferred to media containing G-CSF or Epo as indicated. WCL were made from cells harvested at the indicated time-points and used for immunoblotting with p27kip1antibody (top panels). The membrane was stripped and reblotted with PU.1 (middle panels) and actin antibodies (bottom panels). (A) Vector or LRG-transfected 32Dcl3 cells in G-CSF-containing media. (B) Vector or LRG-transfected 32Dwt18 cells in Epo-containing media.
|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
to their respective LRG-binding sites dramatically increases within the initial 3 days of Epo-induced neutrophil maturation, and the binding pattern for C/EBP did not vary significantly. PU.1 has been shown to activate transcription of primary neutrophil granule enzymes such as neutrophil elastase [18
]. C/EBP is the predominant C/EBP family member in immature myeloid cells and appears to play an important role in regulation of early granulopoiesis [19
20
21
]. C/EBP is expressed in late-stage granulocytes, and C/EBP null mice have increased numbers of immature myeloid cells in their bone marrow and hyposegmented dysfunctional neutrophils [20
21
22
]. Our results indicate that PU.1 and C/EBP bind to the putative LRG promoter region that we identified previously, and binding of these transcription factors appears to modulate the expression of LRG during neutrophil maturation. Up-regulation of LRG expression seems to be an event that occurs at a later stage during granulocytic differentiation and maturation. The gene for human LRG localizes to the same chromosome region to which genes for multiple, primary neutrophil granule enzymes also localize, suggesting that LRG may be another primary granule protein. Lack of commercially available antibodies to LRG has hindered the determination of its cellular localization. We have stably transfected a V5-tagged LRG cDNA into 32Dcl3 or 32Dwt18 cells, so that specific antibody to the V5 epitope tag could be used. Immunofluorescence staining and confocal microscopy results displayed similar diffuse granular distribution patterns within the cytoplasmic compartment for LRG and MPO, suggesting that LRG localizes to primary neutrophil granules. Subcellular fractionation experiments further confirmed the colocalization of LRG and MPO within the primary neutrophil granules. Notably, LRG in differentiated cells tended to move to fractions of lighter density, indicating its progression in post-translational modification along the secretion pathways, which is possibly related to its functions in neutrophil maturation. Heparin-binding protein (HBP), another primary neutrophil granule protein that belongs to the family of serine proteases, was also recently shown to localize to secretory vesicles in addition to primary granules in mature human neutrophils but not in HL-60 cells [23 ]. It is plausible that LRG, similar to HBP, has dual localization. In undifferentiated 32D cells, LRG was produced and stored in primary granules. When its expression is up-regulated during differentiation and maturation, however, LRG could also be targeted to other compartments, such as the secretory vesicles. This dual localization may imply that LRG has multiple functions in granulopoiesis.
The most import issue in LRG characterization is to determine its physiological roles. Our previous work showed that the expression of LRG is induced during neutrophilic granulocyte differentiation in response to G-CSF, leading us to the hypothesis that LRG may be involved in granulopoiesis. Constitutive expression of LRG significantly accelerated neutrophilic differentiation in 32Dcl3 cells, demonstrated by nuclear morphologic changes and cell-surface expression levels of CD11b. Proliferation of LRG transfectants was decreased compared with cells transfected with empty vector alone, consistent with an expected decrease in proliferation as cells undergo differentiation, and also consistent with our observation that LRG transfectants undergo accelerated, neutrophilic differentiation (data not shown).
The mechanisms of this accelerated differentiation are still under investigation. By immunoblot analysis, LRG-transfected 32Dcl3 cells displayed a stronger pStat3 signal at an earlier time-point in response to G-CSF treatment. Jak kinases and Stat proteins have been shown to be activated in hematopoietic cells by the majority of hematopoietic cytokines [24 ]. Overexpression of Jak3 in 32Dcl3 cells was recently reported to accelerate neutrophil differentiation in response to G-CSF [25 ]. This effect was mediated in part through Stats and cell-cycle regulatory proteins. It is possible that accelerated differentiation induced by overexpression of LRG is mediated by similar signaling pathways involving Stat3. Whether LRG facilitates the phosphorylation and action of Stat3, or LRG directly interacts with other genes involved in granulocytic differentiation after it translocates to the nucleus upon G-CSF stimulation remains to be determined. On the other hand, our results suggested that PU.1 or the cyclin-dependent kinase inhibitor p27kip1 might not be involved in the accelerated differentiation of 32Dcl3 cells, as we did not detect any difference in the expression patterns of p27kip1 or PU.1 between 32Dcl3 and 32Dwt18 cells (in which constitutive expression of LRG does not accelerate neutrophilic differentiation). This is in contrast with previous reports about Stat3-mediated activation of p27kip1 [26 ] and PU.1 [27 ]. One explanation for this is that Stat3 may be necessary but not sufficient for the activation of p27kip1 and PU.1.
Previous studies suggest that LRG interacts with TGFR-β and modulates the activity of TGF-β [28 ]. Notably, TGF-β has been shown to modulate hematopoietic cell proliferation with inhibitory and stimulatory activities. TGF-β was reported to increase GM-CSF-driven granulocyte differentiation, which was postulated to occur via increased surface expression of the GM-CSFR induced by TGF-β [29 ]. Recently, Saito et al. [30 ] reported isolation of LRG from high endothelial venule cells and proposed a role for LRG in adhesive interactions between lymphocytes and the endothelium. They also reported that LRG, immobilized on plastic wells, could bind to TGF-β in vitro. Therefore, LRG may aid in the control of granulopoiesis by helping to modulate the surface expression of different receptor types, including TGFR-β, GM-CSFR, and possibly the G-CSFR. Another recent publication provided evidence that specific isoforms of LRG were detected in the proliferative endometrium of women undergoing assisted reproduction, suggesting that LRG has a role in implantation and/or decidualization [31 ]. As G-CSFR is highly expressed in endometrium, the interaction between LRG and G-CSFR may also play an important part in the preparation of the uterus for embryonic growth factor activity.
In these experiments, we used two model systems: 32Dcl3 and 32Dwt18 cells. It would be expected that the signaling pathways mediated by the same cytoplasmic domain of the G-CSFR should lead to the same cellular processes in the two cell model systems. However, the effects of constitutive expression of LRG in the two systems are quite different. Our data indicate that the full-length G-CSFR plays a critical role in neutrophil activation, in addition to neutrophilic differentiation. In vivo studies by Link’s group [32 ] demonstrated that in homozygous mutant mice expressing a G-CSFR/EpoR, in which the extracellular and transmembrane domains of G-CSFR are fused to the cytoplasmic domain of the EpoR, morphologically mature neutrophils were produced, but their mobilization was greatly impaired. Surface adhesion molecules are important factors involved in these biological activities. Interestingly, we showed that CD11b, the major β2 integrin expressed on neutrophils, is only induced by the endogenous, full-length G-CSFR, and its expression is accelerated by the constitutive expression of LRG. Thus, LRG may interfere with the metabolism of adhesion molecules to facilitate functional maturation of neutrophils in response to G-CSF.
In the conventional view, the role of the extracellular domains of cytokine receptors is simply to bind to the respective cytokine and mediate receptor dimerization. It is the cytoplasmic domains that play critical roles in signal transduction, as they bear the docking sites for cytoplasmic tyrosine kinases. A recent study about the 
319 G-CSFR mutation identified in a patient with severe congenital neutropenia has provided additional insights into G-CSFR signaling [33
]. This mutation deletes part of the extracellular portion and the entire transmembrane and cytoplasmic domains. Expression of this mutant in Ba/F3 cells, alone or in combination with the wild-type G-CSFR, reproduces the dominant-negative phenotype observed in the patient. Proper targeting of the receptor to the cell membrane is disrupted, and the response to G-CSF is also abrogated. Similar defects in receptor processing and assembly have been reported with truncation mutants of the EpoR as well, indicating that sequences in the extracellular domain are critical for correct expression and sorting of both mature receptor complexes to the plasma membrane [34
, 35
]. Here, we present further evidence that the extracellular domain of the G-CSFR is indeed indispensable for the unique signaling pathways mediated by the full-length G-CSFR. We thus propose the model shown in Figure 8
. In response to G-CSF, the full-length, wild-type G-CSFR undergoes unique, conformational changes in the cytoplasmic domain, activating the associated tyrosine kinases. This results in the recruitment and activation of Stat proteins, including Stat3, and possibly other unknown molecules. The activated Stats then translocate to the nucleus and activate the transcription of target genes. The cell undergoes terminal differentiation, and CD11b expression is induced. On the other hand, in response to Epo, the chimeric Epo/G-CSFR may undergo different conformational changes within the cytoplasmic domain. This may lead to the activation of different Stat molecules or other factors and in the end, different target genes. LRG may be an important factor involved in the nonredundant signaling pathway mediated by the full-length G-CSFR, and its constitutive expression accelerates neutrophilic differentiation in 32Dcl3 cells. Additional studies to more precisely define these signaling pathways should help to further clarify the functions of LRG in granuolopoiesis.
 View larger version (36K): [in a new window] |
Figure 8. Schematic diagram for LRG function in the unique signaling pathway mediated by the full-length G-CSFR. Refer to text for detailed description. TF, Transcription factor; P, phosphorylated.
|
Received November 14, 2007; revised January 16, 2008; accepted January 24, 2008.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
2-glycoprotein from human serum (author’s transl) Hoppe Seylers Z. Physiol. Chem. 358,639-646[Medline]2-glycoprotein, a novel marker of granulocytic differentiation J. Leukoc. Biol. 72,478-485 and G-CSF receptor signals cooperate to induce the myeloperoxidase and neutrophil elastase genes Leukemia 15,779-786[CrossRef][Medline]-deficient mice Proc. Natl. Acad. Sci. USA 94,569-574-deficient mice Proc. Natl. Acad. Sci. USA 94,13187-13192 Mβ 2 (CD11b/CD18) expression by cytokines is regulated by Stat3-dependent activation of PU.1 J. Biol. Chem. 277,19001-190072-glycoprotein detected in the proliferative endometrium of women undergoing assisted reproduction are associated with spontaneous pregnancy Fertil. Steril. Epub ahead of print.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
