Changes in gene expression of granulocytes during in vivo granulocyte colony-stimulating factor/dexamethasone mobilization for transfusion purposes

Granulocyte concentrates constitute a promising adjuvant tool in the treatment of neutropenic and immunocompromised patients experiencing life-threatening infections in which the exclusive use of modern antimicrobial drugs and additional growth factors is ineffective.^1-3  Donor stimulation with a combination of granulocyte colony-stimulating factor (G-CSF) and dexamethasone has become the standard procedure to increase the number of neutrophils in the circulation of the donors and, thus, to collect a sufficient amount of cells for the preparation of granulocyte concentrates.⁴ 

In addition to its important function in granulopoiesis, G-CSF has been reported to modulate several granulocyte functions in vitro. For instance, G-CSF increases the chemotactic ability, enhances cell adhesion to vascular endothelium, promotes phagocytosis, and primes the NADPH-oxidase activity (for a review, see Eyles et al⁵ ). However, glucocorticosteroids such as dexamethasone have been suggested to suppress some granulocyte functions, including mobility, adhesion, and microbial killing.^6-8 

Despite all the aforementioned studies, we and others have found that donor granulocytes, when mobilized for transfusion purposes, show virtually normal functional characteristics in vitro in the presence of some minor phenotypic changes, whereas their life span is consistently prolonged.^9,10 

Both G-CSF and dexamethasone are well-established prosurvival factors for neutrophilic granulocytes,^11-13  and this effect involves various survival signaling pathways. G-CSF has been shown to increase in neutrophils the mRNA expression of A1/Bfl-1, an antiapoptotic member of the Bcl-2 family of proteins.¹⁴  The prosurvival effect of dexamethasone in neutrophils has been connected to the stabilization of Mcl-1, another antiapoptotic Bcl-2 family member, during neutrophil culture.¹⁵  Furthermore, G-CSF inhibits the mitochondria-dependent activation of caspase-3 in neutrophils via control of the calpain-dependent degradation of the X-linked inhibitor of apoptosis (XIAP).^16,17  In these studies, the antiapoptotic effect of G-CSF was shown to depend on de novo protein synthesis.

Calpains form a family of calcium-dependent cysteine proteases of which calpain-1 (μ-calpain, calpain I), calpain-2 (m-calpain, calpain II), and the natural inhibitor of calpains, calpastatin, are ubiquitously expressed. Several studies have implicated calpain activity in spontaneous neutrophil apoptosis,^18,19  and there is a growing number of indications that calpains play an important role in the early phase of programmed cell death.^17,20 

Although some of the major functional characteristics relevant for host defense seem well preserved in the donor granulocytes, the extent to which other relevant properties of these cells are altered by the in vivo preactivation is as yet unclear. To understand how G-CSF and dexamethasone may induce the granulocytes to obtain their increased survival capacity, we performed a comparative study with the use of Agilent microarrays (Agilent Technologies Netherlands, Amstelveen, The Netherlands) coupled with real-time reverse transcription–polymerase chain reaction (RT-PCR), flow cytometry, and immunoblotting. In the present study, we demonstrate that mobilization of granulocytes with G-CSF and dexamethasone has a considerable impact on gene expression of these cells, with more than 1000 genes being strongly affected. These changes could be partially mimicked by in vitro culture of neutrophils in the presence of G-CSF and dexamethasone. However, more than 75% of changes in gene expression were unique for in vivo mobilization. The affected genes encoded proteins involved in cellular transcriptional activity and protein synthesis, immune response and inflammation, as well as cell survival and apoptosis. We show that among the genes involved in control of cell apoptosis, treatment with G-CSF and dexamethasone induced an increased expression of calpastatin, the endogenous inhibitor of calpains. This increased expression also could be accomplished by culturing neutrophils with G-CSF/dexamethasone in vitro. Furthermore, calpains are demonstrated to contribute importantly to neutrophil apoptosis, whereas the increase in calpastatin mRNA and protein levels directly corresponds to the prolonged lifespan of neutrophils when treated with G-CSF/dexamethasone in vivo or in vitro.

Granulocytes from 3 different healthy donors (2 males and 1 female) were studied before and after treatment with G-CSF and dexamethasone. Donors received G-CSF (5 μg/kg subcutaneously) and dexamethasone (8 mg orally). The study was approved by the ethical medical committee at Sanquin Research and Landsteiner Laboratory and was conducted with informed consent given in accordance with the Declaration of Helsinki. Blood samples were taken just before the donor treatment (control sample, fresh) and 16 to 20 hours after G-CSF and dexamethasone administration (in vivo treatment).

Neutrophils from the control sample were isolated and directly prepared for RNA isolation (see the section “RNA isolation, amplification, labeling, and hybridization”) or cultured overnight in the absence (control sample, apoptosis) or presence of G-CSF (Amgen Europe, Breda, The Netherlands) and dexamethasone (Sigma-Aldrich, St Louis, MO; in vitro treatment). Thus, 4 pools of RNA were obtained for comparison (Figure S1, available on the Blood website; see the Supplemental Materials link at the top of the online article).

Heparinized venous blood was collected from the donors, and the granulocytes were isolated as described.^21,22  In short, the granulocytes and mononuclear cells were separated over isotonic Percoll with a specific density of 1.076 g/mL. Erythrocytes in the pellet were lysed in ice-cold medium containing 155 mmol/L NH₄Cl, 10 mol/L KHCO₃, and 0.1 mmol/L ethylene diamine tetraacetic acid (EDTA), pH 7.4. Granulocytes were washed and resuspended in Hepes-buffered saline solution (HBSS, containing 132 mmol/L NaCl, 6.0 mmol/L KCl, 1.0 mmol/L CaCl₂, 1.0 mmol/L MgSO₄, 1.2 mmol/L potassium phosphate, 20 mmol/L Hepes, 5.5 mmol/L glucose, and 0.5% (wt/vol) human serum albumin, pH 7.4). The purity of granulocytes isolated with this method was more than 95%. Overnight culture was performed in HBSS with or without addition of 100 ng/mL G-CSF and 1 μmol/L dexamethasone.

Total cellular RNA was extracted from a minimum of 20 × 10⁶ cells with TRIzol reagent (Invitrogen, Breda, The Netherlands) according to the protocol provided by the manufacturer, with the following minor modifications. An additional phenol-chloroform extraction was performed and the isopropanol precipitation at −20°C was facilitated by the addition of 20 μg/mL glycogen (Roche Applied Science, Almere, The Netherlands). Purity and integrity of the RNA samples were confirmed on the Agilent 2100 bioanalyzer (Agilent Technologies Netherlands) using the RNA 6000 Nano LabChip kit. Finally, mRNA was amplified with the MessageAmp II Kit (Applied Biosystems/Ambion, Foster City, CA). Labeling, hybridization, and data extraction were performed at ServiceXS (Leiden, The Netherlands), as has been described elsewhere.²³ 

The microarray slides were scanned with the Agilent dual-laser DNA microarray scanner. Default settings of Agilent Feature Extraction preprocessing protocols were used to obtain normalized expression values from the raw scans. Exact protocol and parameter settings are described in the Agilent Feature Extraction Software User Manual 8.5 (http://chem.agilent.com/scripts/LiteraturePDF.asp?iWHID = 37 629). The default Agilent normalization procedure, called Linear & Lowess, was applied. Rosetta Resolver (Rosetta Biosoftware, Seattle, WA) was used for analysis of the data. Genes were defined as differentially transcribed if the average expression level changed at least 3-fold compared with unstimulated cells (0 hours, control sample) in all 3 donors. The microarray data have been deposited in National Center for Biotechnology Information's Gene Expression Omnibus²⁴  and are accessible through GEO Series accession number GSE12841 (http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE12841).

PCR amplification was performed on a LightCycler instrument (Roche Applied Science), and analyzed with LightCycler Software version 3.5 (Roche Molecular Biochemicals, Mannheim, Germany). The reaction was performed with Lightcycler FastStart DNA Master^PLUS SYBR Green I (Roche Diagnostics, Indianapolis, IN). The annealing temperature used for all primers was 65°C. The reaction mixture consisted of 4 μL of cDNA, 1 μL of relevant primer combination, and 4 μL of SYBR Green I mix in a total volume of 20 μL. All amplified cDNA was compared with the standard within the same run, and in every run the same standard was used, although there was very little variation in the standard between runs.

For amplification, the following LightCycler protocol was used. The chemical cleft of the Taq polymerase was removed by preincubation for 10 minutes at 95°C; the template was amplified for 40 cycles, with annealing of the primers at 65°C. The fluorescence was measured at the end of each cycle at 72°C. At the end of 40 cycles, a melting curve was generated to determine the unique features of the DNA amplified. The specific size of the product was determined on a 1% (wt/vol) agarose gel. Subsequently, the obtained band was purified using the GFX PCR DNA and Gel Band purification kit (Amersham Biosciences, a division of GE Healthcare, Little Chalfont, United Kingdom) according to the manufacturer's instructions to remove excess dNTPs and primers. The product was sequenced by Big-dye Terminator Sequencing and ABI Prism software (Applied Biosystems, Foster City, CA). The sequence was verified with BLAST (http://www.ncbi.nlm.nih.gov/BLAST/) to determine specificity. All products obtained were unique and had no overlap with other isoforms.

As a source of cDNA for standard curves to which all samples were normalized, neutrophils were isolated from an apheresis buffy coat obtained from the Sanquin Blood Bank North-West Region (Amsterdam, The Netherlands). Serial 10-fold dilutions from the cDNA obtained were made to which each sample was quantified with the method described in Technical Note No. LC 13/2001 (Roche Applied Science), as has been described elsewhere.²⁵ 

Cell surface expression of various receptors on granulocytes was assayed in total leukocyte samples by flow cytometry (ie, fluorescence-activated cell-sorting [FACS]), with saturating concentrations of commercially available monoclonal antibodies (MoAbs), either directly labeled with fluorescein isothiocyanate (FITC), phycoerythrin (PE), or indirectly labeled with allophycocyanin (APC). CD52-FITC, CD55-FITC, CD59-FITC, and CD69-PE were from Sanquin Reagents (Amsterdam, The Netherlands); EMR3-FITC and CD177-unconjugated were from AbD Serotec (Oxford, United Kingdom); and goat F(ab′)₂ anti–mouse-IgG-APC was from Southern Biotech (Birmingham, AL).

Samples were analyzed on an LSRII flow cytometer equipped with FACSDiva software (BD, Franklin Lakes, NJ). Cells were gated based on their forward and side scatter, and 10 000 gated events were collected per sample (100% positive staining for CD16 and negative for CD36 [monocytes] or CD56 [NK cells] confirmed purity of the analyzed population).

To detect apoptosis, cells were labeled for 10 minutes on ice with FITC-labeled annexin V (Bender Med Systems, Vienna, Austria), diluted 1:500 in HBSS, and supplemented with 2.5 mM CaCl₂. Annexin V labeling was followed by a single wash step with the same medium, whereupon the cells were resuspended in HBSS 2.5 mM CaCl₂ containing 1 μg/mL propidium iodide (PI; Sigma-Aldrich). After an additional 5 minutes on ice, the samples were analyzed on a FACScan flow cytometer (BD). Surviving cells were defined as the cells in the lower left quadrant that stained negative for both annexin V and PI. A total of 10 000 events were collected for each sample, and data were analyzed with the use of CellQuest Pro software (BD).

Total cell lysates were prepared by treating the cells with a lysis buffer (250 mmol/L sucrose, 70 mmol/L KCl, 0.5% Triton X-100 (vol/vol), 0.5% β-octylglucoside (vol/vol), 2 mmol/L NaVO₄, 1 mmol/L NaF, 1 mM EDTA, supplemented with a complete protease inhibitor cocktail mix (PIM; Roche Diagnostic, Almere, The Netherlands) and 2 mmol/L diisopropylfluorophosphate (DFP; Fluka Chemica, Steinheim, Switzerland) in phosphate-buffered saline [PBS]) for 30 minutes on ice. Afterward, samples were mixed with 4× Laemli sample buffer (LSB; 50 mmol/L Tris-HCl, pH 6.8, 10% glycerol [vol/vol], 5 mmol/L DTT [DL-dithiothreitol, Sigma] 1% β-mercaptoethanol, 1% sodium dodecylsulfate [SDS; m/v], 10 μg/mL bromophenol blue) and boiled for 15 minutes at 95°C. All samples were stored at −20°C before subjection to SDS–polyacrylamide gel electrophoresis (SDS-PAGE). Samples were run on 10%, 1.5-mm polyacrylamide gels in a protean-3 mini system (Bio-Rad Laboratories, Veenendaal, The Netherlands). The equivalent of 1.5 × 10⁶ cells was loaded in each lane.

After electrophoresis, proteins were transferred to polyvinyl difluoride membranes (PVDF, Bio-Rad), which were subsequently blocked for 60 minutes with blocking buffer (5% nonfat dry milk [m/v, Elk; Campina, Zaltbommel, The Netherlands] in Tris-buffered saline, 0.1% Tween-20 [vol/vol]). The membranes were immune-labeled with specific antibodies in blocking buffer containing 2 mmol/L NaN₃ overnight at 4°C. After washing, the membranes were labeled with fluorescently labeled secondary antibodies (either IRDye 800CW or IRDye 700 [Li-COR Bioscience, Lincoln, NE]), and the protein bands were visualized with the Odyssey Infrared Imaging System (Li-COR), and analyzed with the accompanying software (version 2.1).

Significantly regulated genes were selected with Rosetta Resolver (Rosetta Biosoftware, Seattle, WA). Genes with a fold change greater than or equal to 3, together with a P value cutoff of .01 (by 1-way analysis of variance [ANOVA] test with the Benjamini-Hochberg false-discovery rate correction), were considered significantly different across the different cell populations. Genes differentially expressed in mobilized granulocytes were categorized by reported or putative functions by the use of the OntoExpress program (http://vortex.cs.wayne.edu/projects.htm#Onto-Express). Graphs were drawn and statistical analysis was performed with GraphPad Prism version 5.00 for Windows (GraphPad Software, San Diego, CA). Data were evaluated by paired, one-tailed Student t test where indicated. The results are presented as the mean plus or minus SEM, as indicated.

Treatment of healthy donors with a combination of a single dose of G-CSF and dexamethasone has been described to result in an increase in the number of circulating neutrophils within 2 to 4 hours after administration, which peaks within 12 to 16 hours (Figure 1A) and returns back to normal levels within 48 to 72 hours.²⁶  Granulocytes for transfusion usually are collected during the peak phase of mobilization (ie, after overnight treatment).

We used a microarray approach to analyze gene expression patterns in mobilized cells in comparison with the cells obtained from the same healthy control donors before G-CSF/dexamethasone administration. The administration of G-CSF/dexamethasone results in both mobilization of immature neutrophils from the bone marrow, as indicated by staining with EMR3, a myeloid-specific member of the epidermal growth factor-7-transmembrane (EGF-TM7) family of adhesion class TM7 receptors, which has been recently described as a marker for mature granulocytes (Figure 1A),²⁷  as well as direct effects on the already circulating cells. To estimate the contribution of the latter, we also performed in vitro G-CSF/dexamethasone incubations and analyses on neutrophils taken before the in vivo treatment (see also the scheme for experimental design in Figure S1). This step also allows the evaluation of changes in gene expression associated with the G-CSF/dexamethasone–induced delay in apoptosis.

By the use of Agilent Whole Human Genome microarrays, we screened approximately 32 000 gene transcripts. This analysis revealed that approximately 1000 genes were differentially expressed in the cells after the administration in vivo of G-CSF/dexamethasone compared with those isolated before mobilization (Figure 1B and Document S1). Even more genes were differentially expressed after treatment in vitro of neutrophils with G-CSF/dexamethasone (Figure 1B and Document S1).

Approximately 150 genes were similarly regulated in both experimental systems (in vivo vs in vitro, Table 1), but still more than 800 genes demonstrated a unique expression pattern for the in vivo stimulation.

We used LightCycler RT-PCR to verify the changes in gene expression as identified by the microarray analysis. We selected 6 representative genes, encoding well-characterized cell surface receptors, from the microarray dataset for subsequent confirmation by RT-PCR and flow cytometry (Figure 2A). For these genes, there was a strong correlation (r = 0.94) between mRNA levels determined by either microarray or RT-PCR analysis. Mobilization of granulocytes resulted in increased expression levels of complement regulatory factors (CD55 and CD59) and a strong increase in the level of CD177 mRNA, as well as the simultaneous down-regulation of CLEC2A (CD69), CD52, and EMR3 mRNA levels.

Because all transcripts selected for the validation of the microarray data by quantitative RT-PCR represented surface receptors expressed on the plasma membrane, we determined the protein expression levels by flow cytometry (Figure 2B). Both complement regulatory factors CD55 (also known as decay-accelerating factor [DAF]) and CD59 were found to be present on fresh neutrophils. Their expression increased significantly after treatment with G-CSF/dexamethasone in vivo. Granulocyte mobilization also resulted in the increased expression levels of CD177 (also known as the human neutrophil antigen [HNA]-2a or NB1). Increased expression of CD177 on neutrophils has previously been reported in response to G-CSF treatment for hematopoietic stem cell mobilization.²⁸  Control neutrophils clearly expressed the GPI-anchored proteins CD52 and, to a lesser extent, CD69, whereas the surface expression was either diminished (CD52) or had become completely null (CD69) on donor neutrophils mobilized in vivo, confirming the microarray data.

The expression of EMR3 also was found to be lower on the surface of in vivo mobilized cells. This may correspond with the lower maturation state of the granulocytes “purged” from the bone marrow by the combined treatment with G-CSF and dexamethasone, as also indicated by the concomitant neutrophilic left shift (data not shown).

Significant numbers of genes with different expression patterns between the control granulocytes and those isolated after G-CSF/dexamethasone administration fell into 6 main functional gene ontology (GO) categories: (1) transcription and regulation of transcription, (2) signal transduction, (3) immune responses, (4) cell survival and apoptosis, (5) cell adhesion and motility, and (6) cell metabolism (Table 2). Similar analysis has been applied to the genes presented in Table 1 to compare these categories of gene patterns (Table S1).