| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Published online before print February 3, 2004
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
,
,1
* Immunogenetics Laboratory,
Center for Animal Functional Genomics, Department of Animal Science, and
Molecular Reproduction Laboratory, Departments of Animal Science and of Physiology, Michigan State University, East Lansing
1Correspondence: Department of Animal Science, 1205E Anthony Hall, Michigan State University, East Lansing, MI 48824. E-mail: burtonj{at}msu.edu
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: CD62L • bone marrow • bovine
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Glucocorticoids are well known for their potent anti-inflammatory actions in mammals [11
, 12
], an important component of which is their ability to modulate expression of a variety of adhesion molecules on circulating leukocytes and vascular endothelia cells [13
, 14
]. In particular, exogenous and endogenous glucocorticoids down-regulate surface expression of L-selectin on circulating neutrophils in human and other model systems, with bovine neutrophils showing heightened sensitivity to the steroid [15
16
17
18
19
20
21
22
23
24
]. We have recently demonstrated that circulating neutrophils from cattle experiencing high blood cortisol concentrations during normal parturition have down-regulated L-selectin mRNA that correlates with down-regulated L-selectin surface expression and neutrophilia [25
]. This suggested to us that glucocorticoid may have genome-level effects on bovine L-selectin expression that influence blood neutrophil homeostasis. We have also demonstrated that these neutrophils express cytosolic mRNA and protein for the 
-isoform (GR) of the glucocorticoid receptor [25
, 26
], which is the isoform known to bind glucocorticoids and influence transcription of hormone-responsive genes [27
]. Available literature argues against a genome-level effect of glucocorticoids on L-selectin expression, as human blood neutrophils treated with dexamethasone (DEX) in vitro for 30–210 min do not down-regulate surface L-selectin [28
, 29
]. In addition, studies in rabbits indicated that the target of steroid-induced L-selectin down-regulation may not be existing blood neutrophils but rather the maturation pool of neutrophils in bone marrow that then get released into blood [22
, 30
]. However, it remains unresolved which neutrophil pool in cattle down-regulates L-selectin expression in response to glucocorticoids and the mechanisms involved.
As blood neutrophil L-selectin mRNA-abundance changes during bovine parturition are relatively acute (occurring within 3–6 h of the sharp peak in blood cortisol), correlate with neutrophilia, and persist for 
36 h [25
], we hypothesized in the present study that existing blood neutrophils are primary targets for glucocorticoid-induced L-selectin down-regulation in cattle. We also hypothesized that the effect of the steroid on bovine L-selectin is mediated directly at a gene-expression level in blood neutrophils, via hormone-activated GR. To test these hypotheses, we simultaneously monitored changes in levels (abundance) of L-selectin mRNA and protein in bovine-blood and bone marrow neutrophils exposed to DEX in vivo and in normal blood neutrophils treated for 4 h with the glucocorticoid ± a GR antagonist in vitro. DEX was chosen as the glucocorticoid for all experiments, as it binds with high specificity and affinity to GR [31
]. Our results show that L-selectin expression in mature neutrophils of bovine bone marrow and blood is sensitive to DEX regulation in vivo; acute bone marrow release of immature neutrophils does not explain the pronounced neutrophilia or down-regulated L-selectin expression profiles of blood neutrophils in DEX-treated cattle; and DEX regulation of L-selectin mRNA abundance in bovine-blood neutrophils is mediated via glucocorticoid receptors and likely occurs at a gene-expression level.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
antagonist was mifepristone (RU486; Sigma Chemical Co.). Total leukocyte counts in whole blood and isolated cell preparations were made using the Z1 Coulter particle counter system (Beckman Coulter Particle Characterization, Miami, FL). Leukocyte differentials were done using immunostaining and flow cytometry (FACSCalibur flow cytometer and Cell Quest software, Becton Dickinson, San Jose, CA). Fluorescence-activated cell sorter lysis solution (Biosciences, San Jose, CA) was used (according to the manufacturer’s recommendations) to fix cells before immunostaining. The monoclonal antibodies (mAb) used were anti-G1 (neutrophils; clone MM20A), anti-CD14 (monocytes; clone MM61A), and anti-CD2 (T lymphocytes; clone BAQ95A), all purchased from VMRD (Pullman, WA) and diluted to working concentrations of 14 µl/ml 0.01% BSA-containing phosphate-buffered saline (PBS). Neutrophil purity checks were performed by G1 immunostaining and flow cytometry [25 ], using density dot-plots of 90° light-scattering on the y-axes and G1 fluorescence intensity on the x-axes. The mAb used for all analyses of L-selectin protein was clone BAQ92A (VMRD). The secondary antibody used for all flow cytometric applications was phycoerythrin (PE)-conjugated goat anti-mouse immunoglobulin G (IgG)1 (1:800 final dilution; Code #M32004, Caltag Laboratories, Burlingame, CA). The sheath fluid used for flow cytometric data acquisition was Haema-Line 2 from ABX Diagnostic Inc. (Irvine, CA).
For immunoblot assays, the cell-disruption solution consisted of 0.34 M sucrose, 10 mM Tris-Cl, pH 6.8, 5 mM EGTA, 100 µM Na3VO4, 10 mM NaF, 2 mM phenylmethylsulfonyl fluoride, 10 mM benzamidine, and 7 µg/ml each leupeptin and pepstatin (Sigma Chemical Co.). The BAQ92A antibody was diluted 1:3000. The anti-ß-actin antibody (1:3000) was from ABCAM (Cambridge, UK). The detection antibody (1:10,000) was goat anti-mouse IgG1 horseradish peroxidase-conjugate (Bethyl, Montgomery, TX). Immunoblots were developed using the SuperSignal West Pico chemiluminescent substrate system (Pierce, Rockford, IL).
TRIzol reagent (Invitrogen Life Technologies) was used for cell lysis (at 1.0 ml per 1x107 cells) and subsequent RNA isolations (per the manufacturer’s instructions). The L-selectin and ß-actin cDNA probes used for Northern blot and slot blot analyses were described in detail elsewhere [25 ]. The bovine-specific primer pairs used for quantitative real-time reverse transcriptase-polymerase chain reaction Q-RT-PCR) analyses of L-selectin mRNA abundance were designed in-house (Primer Express Software, Perkin Elmer Applied Biosystems, Foster City, CA) and synthesized at a commercial facility (Qiagen-Operon, Inc., Alameda, CA). For L-selectin, these were forward 5'-ACGGGAAAAAAGGATTACTATGGA-3', reverse 5'-GCCTATAGTTGCATATGTATCAAATTTTCA-3' (product length=144 bp; Tm=74°C); for bovine ß-actin (the normalizing gene), these were forward 5'-CGCCATGGATGATGATATTGC-3', reverse 5'-AAGCGGCCTTGCACAT-3' (product length=66 bp; Tm=84°C). The Q-RT-PCR assay was performed using Applied Biosystems 7000 DNA sequence detection equipment and software and the SYBR Green PCR master mix system (both from Perkin Elmer Applied Biosystems).
Experiment 1: DEX regulation of L-selectin expression in two main neutrophil pools
Animals and glucocorticoid treatments
Twenty-eight castrated male calves (steers) of the Holstein breed, averaging 108.4 ± 1.8 days of age and 106.7 ± 4.5 kg body weight, were used to simultaneously investigate the in vivo effects of DEX on surface L-selectin expression and L-selectin mRNA abundance in blood and bone marrow neutrophils. The animals were housed at Michigan State University’s Dairy Teaching and Research Facility (East Lansing) and cared for according to the facility’s standard operating procedures, and the Institutional Animal Care and Use Committee approved their use for this study.
Sixteen of the steers were randomly allocated to four of five DEX treatment groups, with four DEX-treated animals per group. The glucocorticoid was administered intramuscularly at 0.10 mg/kg body weight per dose. Steers in 3- and 6-h treatment groups received a single dose of DEX at time 0 h and were processed for tissue collections 3 and 6 h later, respectively. Steers in the 12-h treatment group received two doses of DEX at 0 and 6 h and were processed 12 h after the 0-h dose. Steers in the 24-h group received three doses of DEX at 0, 6, and 12 h and were processed 24 h after the 0-h dose. The additional doses of DEX were given to the 12- and 24-h groups to ensure continuous exposure of blood and bone marrow neutrophils to high levels of the glucocorticoid. The 12 remaining steers not treated with DEX were processed as a fifth 0-h (control) group. This large number of control steers was used to help account for animal variation in basal expression of neutrophil surface L-selectin and L-selectin mRNA, which is not present following DEX administration in cattle [25 , 32 ].
Blood and bone marrow collections
Steers were fitted with indwelling jugular catheters [33
] 2–4 days before sample collections. Catheters were filled with 3.5% (w/v) sodium citrate to maintain patency and sealed until animals were processed. Two animals were processed per collection day and usually included one DEX-treated steer (morning collection) and one control steer (afternoon collection). First, 120 ml blood per steer was collected through catheters at the designated times using 20 cc lure-lock syringes (Becton Dickinson; Fisher Scientific, Pittsburgh, PA) precoated with ACD immediately before use. These samples were aliquoted into a series of four 50-ml graduated tubes that were preloaded with 4 ml ACD and mixed gently before placing on ice and transporting (7-min drive) to the laboratory for immediate processing.
Immediately following blood collections, steers were killed in their housing stalls by administering a lethal dose (97.5 mg/kg body weight) of sodium pentobarbital (Fatal-Plus Solution, Vortech Pharmaceuticals, Dearborn, MI) through the catheters. As soon as death was confirmed (20 s), animals were removed into a connected surgery room, where an autopsy saw (Model BD040, Mopec, Detroit, MI) was used to split the sternums longitudinally. Bone marrow was removed within 5 min from all exposed sternebrae [34
] using surgical scoops. Samples for each animal were cut into 0.1-g pieces and placed into a series of eight 50-ml graduated tubes preloaded with 10 ml ACD until the volume of sample per tube was 15 ml. Tubes were placed on ice and transported to the laboratory for immediate processing.
Leukocyte differential counts in whole blood
Small aliquots (1 ml) of blood from each steer were reserved for total (number of leukocytes/µl whole blood) and differential leukocyte counting (as percent of total leukocytes). Percentages of neutrophils, monocytes, and CD2+ T cells in 10,000 total leukocytes acquired were obtained following immunostaining and flow cytometric analysis [35
]. Cell types were summed for each sample and subtracted from 100% to determine the proportion of unstained leukocytes present, which were called "other." Differential leukocyte counts (number of cells/µl starting blood) were then calculated by multiplying the percentage data by the total leukocyte count for each sample.
Percoll isolation of blood neutrophils
Blood neutrophil isolations were conducted using Percoll (1.084 g/ml) density-gradient centrifugation [25
]. Aliquots (500 µl) of cell suspensions from each animal were used for electronic counting, flow cytometric determination of neutrophil purity, and preparation of slides for microscopy. Remaining cells were pelleted before lysis and storage (–80°C) in TRIzol reagent for subsequent RNA isolations.
Percoll separation of bone marrow cells into three myeloid-maturation fractions
Myeloid-maturation fractions of bone marrow cells were separated on three-layer, discontinuous Percoll gradients (1.03, 1.06, and 1.08 g/ml Percoll), using a method modified from refs. [36
, 37
]. First, 15 ml 0.9% sodium chloride containing 2% dextrose was added to each fresh bone marrow sample. The tubes were gently mixed and placed on ice for 20 min. Bone fragments were removed by passing the samples through four layers of cheesecloth. Remaining cell suspensions were centrifuged (600 g for 10 min at 4°C), supernatants were aspirated and discarded, and the cell pellets were suspended in 5 ml hypotonic lysis buffer for 90 s. Isotonicity was reconstituted with hypertonic restoring solution (2.5 ml/tube). Cells were pelleted (600 g for 10 min at 4°C), supernatants were aspirated and discarded, and the cell pellets were suspended in 40 ml PBS. The cell suspensions were then divided into four 10-ml aliquots, which were layered on top of the Percoll gradients, and the tubes were centrifuged at 1000 g for 20 min at 4°C. Cells floating on the middle (F1) and bottom (F2) Percoll layers were collected by gentle aspiration and pooled into individual tubes per animal. Pellet cells (F3) were pooled into a separate tube per animal. Tubes were centrifuged (600 g for 5 min at 4°C), and supernatants were aspirated and discarded. Cell pellets were washed using 40 ml PBS and suspended in 4 ml PBS for electronic cell counting, neutrophil purity checks, measurements of surface L-selectin expression, and microscopic slide preparation. Remaining cells were centrifuged once more, and the pellets were lysed and stored (–80°C) in TRIzol reagent for subsequent RNA isolation.
Purity of neutrophils isolated from blood and bone marrow
Percoll-isolated blood and bone marrow cells were G1-immunostained for determination of neutrophil purity by flow cytometry [25
]. Resulting %G1+ cell data were used as covariates during statistical analysis of L-selectin mRNA abundance data (see below) to account for the proportion of neutrophils contributing RNA to each sample assayed.
Characterization of neutrophil-maturation stages in isolated blood and bone marrow cells
Microscopy (Nikon Optiphot, Garden City, NJ, microscope) was used to determine the relative proportions (%) of various neutrophil types in 0- and 12-h samples of Percoll-isolated blood neutrophil and F1, F2, ad F3 bone marrow smears [38
]. Three hundred cells were counted in each of two fields per slide under oil immersion (1000x magnification). Circulating band and segmented neutrophil counts (number of cells/ml blood) were computed as percent cell type x total leukocyte count [35
].
Flow cytometric analysis of surface L-selectin expression
Starting samples for L-selectin immunostaining [25
] were 100 µl whole blood (for blood neutrophil analyses) and 1 x 106 cells isolated by Percoll density gradients (for bone marrow cell analyses). L-selectin surface expression was recorded as mean fluorescence intensity (MFI) in PE fluorescence histograms of neutrophils gated out from other leukocyte populations based on their characteristically high forward- and 90° light-scattering properties in density dot-plots [16
, 25
, 36
] of 10,000 cells per sample.
RNA isolation
Total RNA extracted from cells stored at –80°C in TRIzol reagent was checked for concentration and purity using a DU-650 spectrophotometer (Beckman Coulter, Scaumburg, IL) and the 260 and 280 nm readings.
Northern and slot-blot analyses of L-selectin mRNA abundance
Details of the Northern blot (8 µg RNA per lane) and slot-blot assays (5 µg RNA per slot, done in duplicate per sample) are reported elsewhere [25
]. Blots were hybridized first with 32P (Perkin Elmer Life Science, Boston, MA)-labeled L-selectin cDNA and then with labeled ß-actin cDNA. Autoradiographs of the slot-blots were subjected to scanning densitometry (BioRad Laboratories, Hercules, CA), and L-selectin mRNA abundance was recorded as a ratio to ß-actin mRNA abundance on each spot [25
, 39
].
Experiment 2: direct effects of DEX on L-selectin expression in blood neutrophils
Four additional Holstein steers of similar age and weight to the animals used in the in vivo study of Experiment 1, but not treated with DEX, were used as neutrophil donors for in vitro experiments. Percoll-isolated blood neutrophils were treated in vitro with DEX (0 or 10–7 M) [40
], alone or in the presence of RU486 (10–4 M) [41
]. Briefly, sets of cell cultures (25 ml 5x106 neutrophils/ml in sterile 50-ml conical tubes) treated with DEX or DEX ± RU486 were positioned vertically in an orbital shaker (Forma Scientific Model 420, Thermo Electron Corp., Marietta, OH) and incubated at 39°C (normal body temperature for cattle) and 90 rpm over 8 h. Duplicate aliquots (100 µl) of the DEX-treated cells were removed from the cultures at 0, 2, 4, and 8 h for flow cytometric analysis of surface L-selectin expression. Except for the small aliquot of DEX-treated cells needed for the 8-h flow cytometric analysis, remaining cells from all treatments were lysed in TRIzol reagent at 4 h for subsequent RNA isolation and Q-RT-PCR or Northern blot analysis of L-selectin mRNA abundance. Additional cultures of DEX-treated (0 or 10–7 M) neutrophils were incubated in 10 mm culture plates (1x108 cells in 10 ml per plate) at 39°C in moist 5% CO2 air for 0, 2, 4, and 8 h before preparation of their cytosolic fractions for immunoblot analysis.
Immunoblot analysis of cytosolic L-selectin
Cultured neutrophils were prepared as described [42
] for immunoblot analysis. After cell sonication in disruption solution, lysates were centrifuged at 1000 g for 10 min at 4°C, and then at 100,000 g for 30 min at 4°C, supernatants collected and boiled in Laemmli sample buffer (BioRad Laboratories) for 20 min, centrifuged (12,000 g for 15 min at 4°C), and placed on ice for loading onto 11% sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis gels and subsequent transfer of separated proteins to nitrocellulose membranes (BioRad Laboratories). Membranes were blocked with SuperBlockTM blocking buffer in Tris-buffered saline [(TBS); Pierce] for 1 h at room temperature. After washing [TBS/Tween 20 (TBST) buffer; 10 mM Tris-Cl, pH 7.5, 150 mM NaCl, and 0.1% Tween-20], membranes were probed with BAQ92A for 1 h at room temperature and washed again, and detection antibody was added at room temperature for 1 h. Blots were developed (Furtura Model E film processor), photographed (BioRad gel documentation system), stripped (62.5 mM Tris-HCl, pH 6.8, 100 mM ß-mercaptoethanol, and 2% SDS; 50°C for 30 min), and reprobed with anti-ß-actin antibody.
Analysis of L-selectin mRNA abundance
Altered expression of L-selectin mRNA in the cultured neutrophils was assessed by Northern blot analysis or Q-RT-PCR. The Q-RT-PCR [43
] was performed in duplicate per sample using 2.5 ng-starting cDNA. ß-Actin-normalized L-selectin gene-expression changes were computed using a method that monitors relative changes in mRNA abundance based on differences in the PCR-amplified target reaching a fixed threshold cycle (CT) number for a set treatment (calibrator) versus all other treatments [44
]. For our analyses, the CT for untreated neutrophils (i.e., no DEX) at 4 h in culture was the calibrator used to determine relative changes in ß-actin-normalized L-selectin gene expression in DEX-treated cells.
Experiment 3: in vivo effects of DEX on cytosolic L-selectin
Four blood donor steers were used for a final experiment to determine if DEX administration in vivo influenced abundance of cytosolic L-selectin protein in isolated blood neutrophils. Blood (120 ml) was collected into ACD before and 9 h after DEX (0.10 mg/kg body weight) was administered (at 0 and 6 h) into each animal. Cytosolic fractions were prepared from the cells immediately following their isolation and were subjected to immunoblot analysis.
Statistical methods
The main hypothesis of this study was that existing blood neutrophils are the primary targets of glucocorticoid-induced L-selectin down-regulation in vivo. Therefore, our first goal in analysis of data from Experiment 1 was to determine if DEX treatment affected L-selectin surface expression and mRNA abundance in blood and bone marrow neutrophils. The response variables were checked for normality before statistical analysis using the CAPABILITY procedure of SAS [45
]. L-selectin surface expression (MFI) data were distributed normally and thus analyzed without transformation. However, ß-actin-normalized L-selectin slot-blot data required logarithmic transformation to stabilize variances before statistical analysis.
Data sets were analyzed statistically using the MIXED procedure of SAS [46
] and a model that included fixed effects of DEX administration (none vs. treated), treatment group (0, 3, 6, 12, 24 h), and the interaction between them and random steer within treatment group and error effects. The %G1+ neutrophils (mRNA abundance data only), steer age (in days), and steer weight (in kg) were also included in the model as covariates. An unstructured error covariance structure was assumed for animals within the same treatment group. Statistical significance of the tested effects was considered when P 
0.05, with treatment group being the main effect of interest. For data presentation, least squares means (LSM) for each treatment group were computed with Dunnett adjustments [46
, 47
] to test if the 3-, 6-, 12-, and (or) 24-h DEX groups were significantly different from the 0-h (control) group (P=0.05). Where applicable, LSM were back-transformed to the original scale of measurement. A similar analysis was performed to test the effect of treatment group on %G1+ neutrophils in blood and bone marrow F1, F2, and F3 cells. Hypothesis testing continued using correlation analysis (CORR procedure of SAS [46
]) to determine possible relationships between the various phenotypes measured in blood neutrophils and bone marrow F1, F2, and F3 cells. Significance of such relationships was declared when Pearson product-moment correlations had P < 0.05.
The second hypothesis of this study was that the glucocorticoid effect on L-selectin expression in neutrophils occurs at a gene-expression level, is direct, and is mediated via hormone-activated GR. Experiments 2 (in vitro) and 3 (in vivo) were designed to address this hypothesis. L-selectin surface expression determined by flow cytometry, cytosolic L-selectin abundance determined by immunoblot analysis, and L-selectin mRNA abundance changes determined by Northern blot and Q-RT-PCR were presented as original flow cytometric plots, gel figures, or bar graphs (mean density units±SEM) for the indicated treatment scenarios.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
1/3 G1+ cells (33.0±1.85%); and F3, primarily G1+ neutrophils (78.93±2.56%; Fig. 3
). Cells isolated from whole blood contained 
96.38 ± 0.51% G1+ neutrophils (Fig. 3)
.
|
|
|
|
|
|
|
Experiment 2
DEX directly affects L-selectin gene expression in blood neutrophils
Analyses of L-selectin mRNA abundance in cultured blood neutrophils showed that DEX directly reduced L-selectin mRNA abundance (Fig. 8a
) and that RU486 inhibited this effect of DEX (Fig. 8b)
. However, in vitro effects of DEX on cell-surface L-selectin were variable and apparently dependent on levels present at 0 h (Fig. 9
). Immunoblot analysis of cytosolic fractions from the cultured neutrophils revealed massive loss of intact L-selectin after 2 h in culture, even in untreated cells (not shown). Thus, changes in blood neutrophil L-selectin mRNA abundance appeared to become uncoupled from protein expression when the cells were removed from the circulation and cultured in vitro.
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Regarding the first controversy, our earlier work showed that the onset of DEX-induced neutrophilia with concurrent down-regulation of surface L-selectin on circulating bovine neutrophils was rapid (within 3 h) and pronounced enough as to suggest that existing blood neutrophils must have been the main population of affected cells [16 ]. In the same study, the band neutrophil count of whole blood increased slightly during peak neutrophilia but was at least 200-fold less than counts of segmented neutrophils and thus could not explain the complete down-regulation of surface L-selectin observed on all circulating neutrophils [16 ]. In rabbits, however, bromodeoxyuridine labeling and cell-transfer studies strongly implicated the maturation pool of neutrophils in bone marrow as the key cells responsive to DEX-induced L-selectin down-regulation [22 ]. Based on these results, the authors suggested that acute release of affected myeloid-lineage cells from bone marrow into blood might be responsible for changes in surface L-selectin on circulating neutrophils in DEX-treated rabbits [22 ]. However, in a previous study, these investigators also showed that bone marrow release of neutrophils accounted for only a small percentage (10%) of the neutrophilia following DEX administration [30 ]. Instead, mobilization of blood neutrophils from the marginating pool (61%) and extended half-life of existing blood neutrophils (29%) accounted for most of the increase in circulating neutrophil counts of DEX-treated rabbits [30 ]. In previous cattle studies, glucocorticoid did not mobilize bone marrow reserves of mature neutrophils following endotoxin challenge or adrenocorticotropic hormone administration [50 , 51 ]. Instead, neutrophilia in these animals derived primarily from the marginating pool of neutrophils. Given these combined results [16 , 30 , 50 , 51 ], we found it difficult to reconcile that release of L-selectinlow bone marrow neutrophils could so rapidly dilute out the existing pool of L-selectinhigh blood neutrophils following DEX administration as to decrease L-selectin expression profiles in blood to virtually zero within 24 h of hormone injection. We hypothesized in the current study that existing blood neutrophils must be primarily contributors to acutely down-regulated L-selectin expression profiles in circulating cells from animals treated with DEX. If true, this would mean that glucocorticoids have a direct, inhibitory effect on L-selectin expression in blood neutrophils.
Three key findings from our experiments led us to accept our hypothesis and conclude that existing blood neutrophils are the primary targets of L-selectin down-regulation during acute glucocorticoid challenge. First, DEX did not influence expression of surface L-selectin in F1 or F2 bone marrow cells or cause loss of F1 or F2 cells from bone marrow or increases in these cell types in the circulation. This was true even in the 12- and 24-h DEX-treatment groups, which had the most pronounced neutrophilia. Other investigators have shown that the peripheral blood of calves rarely contains band or immature myeloid-lineage cells, even during endotoxin stress and anemia [52 ]. Together with the fact that F1 and F2 bone marrow fractions in our study were composed primarily of myeloblasts, promyelocytes, myelocytes, and metamyelocytes, we conclude that the bovine L-selectin expression system in early-immature and intermediate-immature myeloid-lineage cells is insensitive to glucocorticoids and thus does not influence the L-selectin profiles of blood neutrophils following DEX administration.
Second, although band and segmented neutrophils in the F3 fraction of bone marrow ultimately (24 h) responded to DEX with down-regulation of surface L-selectin, excessive loss of these cells from bone marrow into blood did not occur as neutrophilia progressed in the 3-, 6-, and 12-h DEX-treatment groups. This is in contrast to effects of acute endotoxin challenge in cattle, which causes near total loss of the bone marrow reserve of mature neutrophils within 8 h [34 ]. Thus, DEX effects on the L-selectin system in the current study must have occurred primarily in mature neutrophils of blood, which had higher basal expression (0 h) of surface L-selectin and more pronounced down-regulation of the molecule between 3 and 24 h post-hormone administration than F3 cells. This marked L-selectin down-regulation on circulating cells may have been supported by interactions of the blood neutrophils with vascular endothelial cells, which is well known to cause L-selectin shedding during vascular margination [1 , 2 , 3 ]. In fact, demargination likely accounted for a majority of the early rise in circulating neutrophils with modestly reduced surface L-selectin expression we observed in the 3- and 6-h groups [32 , 50 , 51 ]. However, in the 12- and 24-h groups, where neutrophilia peaked, continued massive loss of surface L-selectin on circulating cells must have been a result of effects of DEX on blood neutrophils themselves, possibly related to increased longevity of existing neutrophils [43 , 53 54 55 ] and (or) direct inhibition of L-selectin gene expression (see below).
Proposed direct, genome level effects of glucocorticoids on L-selectin expression in blood neutrophils have not been widely championed in the literature, as in vitro treatment of human cells for 1–3.5 h with DEX does not cause down-regulation of surface L-selectin [28 , 29 ]. In fact, our own results using bovine-blood neutrophils treated with DEX in vitro partially supported this conclusion, as the glucocorticoid exhibited variable effects on surface L-selectin between 2 and 8 h in culture. However, such short-term culture experiments are really designed to measure acute "shedding" of surface L-selectin in response to DEX, not abundance changes of the molecule inside the cell. This is an important distinction, as detection of changes in surface L-selectin depends on the neutrophils having normal expression and turnover of the molecule at the plasma membrane. In vivo, surface L-selectin on circulating neutrophils is rapidly turned over, as it is constantly cleaved by a membrane-associated cysteine metalloproteinase (called sheddase) when the cells tether and role on L-selectin in the vasculature under blood flow [56 57 58 ]. Normal L-selectin turnover at the plasma membrane must be halted when neutrophils are placed into static conditions outside of blood vessels, as our untreated blood neutrophils lost minimal amounts of surface L-selectin 2–4 h following Percoll isolation and even increased plasma membrane expression by 8 h in culture. The lack of clear down-regulation of surface L-selectin in the cultured neutrophils treated with DEX suggested that glucocorticoid does not activate sheddase, as has been shown for human neutrophils [29 ], and may have instead been a result of the culture conditions used. This point is best substantiated by our in vivo experiment, as a time-lag of several hours occurred between peak neutrophilia and nadir expression of surface L-selectin. If DEX had acted by activating sheddase, down-regulation of surface L-selectin would have been immediate [29 ].
Conversely, normal blood neutrophils constitutively synthesize L-selectin in vivo [2
]. Although the protein is not localized to storage granules in human or bovine neutrophils [59
, 60
], a small reservoir of immunoreactive L-selectin is present near the inner leaflet of the plasma membrane in the cytosol of bovine neutrophils, and this can be mobilized for increased surface expression in vitro [60
]. It is possible that this reservoir of cytosolic L-selectin may facilitate limited recovery of surface L-selectin following sheddase-induced shedding of the molecule during normal vascular margination in vivo and following Percoll isolation of the cells for in vitro work. Given this, we elected to use immunoblot analysis to determine if DEX affected cytosolic levels of L-selectin protein. Cloning, sequencing [61
], and expression of recombinant bovine L-selectin in Escherichia coli [62
] have revealed that the mRNA encoding this molecule is 3 kb in size, that the mature, nonglycosylated protein is 37.4 kDa, and that nine putative N-linked glycosylation sites exist in the extracellular domain of the protein. Thus, depending on degree of glycosylation, numerous sizes of L-selectin are predicted to exist in the cytosol of bovine neutrophils. This was clearly shown on our immunoblot, where at least three intact L-selectin proteins of 37 kDa and 60–67 kDa were present in the cytosol of blood neutrophils from animals not treated with DEX. However, 9 h after DEX was administered into these animals, the 37-kDa form of L-selectin was dramatically reduced. Reductions in some 60–67 kDa L-selectin species were also noted but were more variable. Thus, we propose that DEX had an acute influence on newly synthesized (37 kDa) L-selectin protein but less effect on variably glycosylated forms of L-selectin (60–67 kDa) that may have been present in the cells before the hormone was administered. Unfortunately, rapid (within 2 h) degradation of cytosolic L-selectin made it impossible to demonstrate consistent effects of DEX on 37 kDa cytosolic L-selectin in cultured neutrophils. This suggested one of two possibilities: that removal of neutrophils from the circulation changed L-selectin turnover in and on the cells, precluding detection of DEX effects in vitro, or that DEX effects on neutrophil L-selectin expression were not direct. Others have reported no in vitro effects of DEX on neutrophil-surface L-selectin expression [28
, 29
], claiming that DEX effects in vivo must thus be indirect. However, massive loss of cytosolic L-selectin and low L-selectin turnover at the plasma membrane in cultured cells of the current study suggests instead that in vitro approaches to studying regulation of this protein in blood neutrophils may be inappropriate and lead to results that should be interpreted with caution. Thus, we favor the first possibility put forward above, which along with our hypothesis, was substantiated by a third key finding of our study, that DEX had clear effects on L-selectin mRNA abundance in vivo and in vitro. This is discussed next.
If production of new 37 kDa L-selectin proteins is inhibited by glucocorticoid, as our immunoblot data suggest, two possible mechanisms could have contributed to L-selectin down-regulation in the DEX-treated animals. The first is that DEX may have decreased new protein synthesis and (or) increased degradation of already-synthesized L-selectin proteins. Although our experiments were not designed to study protein synthesis, we found no evidence on our immunoblots that DEX caused additional degradation of L-selectin protein in blood neutrophils beyond what was present in the cells before DEX was administered. Conversely, our experiments have provided strong evidence for a second mechanism of action, whereby DEX reduced the abundance of L-selectin mRNA available for translation. For example, Northern and slot-blot analyses of RNA from blood neutrophils (and from F3 bone marrow cells) revealed pronounced DEX-induced decreases in L-selectin mRNA abundance when the steroid was administered in vivo. We have also shown that this occurs in blood neutrophils of cows with high blood concentrations of endogenous glucocorticoids (cortisol) associated with the stress of parturition [25
]. That this phenomenon occurs directly in existing blood neutrophils was demonstrated through our current in vitro experiments, where 4-h DEX treatment of purified blood neutrophils caused inhibition of L-selectin mRNA expression. Our in vitro experiments also showed that the L-selectin mRNA abundance change was mediated through GR activation, as RU486 blocked this effect of DEX. We acknowledge that the abundance of ß-actin mRNA in some neutrophil samples selected for Northern blots was less than in others and believe this reflected animal variation and (or) unequal sample loading, as glucocorticoid did not affect ß-actin on slot-blots or the immunoblot presented here or in previous studies [25
, 39
, 43
].
Thus, although we cannot rule out inhibited protein synthesis and (or) enhanced degradation as possible mechanism(s) by which DEX regulates cytosolic L-selectin, we can conclude that direct receptor-mediated inhibition of mRNA abundance is one mechanism used by the glucocorticoid to regulate L-selectin expression in bovine-blood neutrophils. As L-selectin mRNA abundance was also decreased in F3 bone marrow cells of the 24-h DEX-treatment group in vivo, we further propose that glucocorticoid treatment may ultimately inhibit critical adhesion properties of this reservoir of neutrophils by interfering with L-selectin gene expression. This may be an important aspect of long-term glucocorticoid treatment on host-inflammatory responses and disease susceptibility. The actual molecular mechanisms underlying DEX regulation of L-selectin mRNA abundance in bovine-blood neutrophils and F3 bone marrow cells await future studies that monitor the rate of L-selectin gene transcription and degree of mRNA stability in hormone-treated cells.
In conclusion, we have shown that DEX directly affects the L-selectin gene-expression system in bovine-blood neutrophils by activating the cells’ glucocorticoid receptors and decreasing L-selectin mRNA abundance. This may have been responsible for the rapid decrease in intracellular 37-kDa L-selectin followed by progressive loss of surface L-selectin between 9 and 24 h post-DEX administration in vivo. Direct effects of glucocorticoids on marginating and circulating neutrophils could explain why there is acute neutrophilia with continuous loss of surface L-selectin on blood neutrophils in animals experiencing stress or treated with DEX. Results of our study also suggested that immature, myeloid-lineage cells developing in bone marrow do not contribute to the L-selectin expression profiles of blood neutrophils over 24 h in DEX-treated cattle. However, band and segmented neutrophils of bovine bone marrow eventually down-regulated L-selectin gene and surface expression in response to DEX, and this may be important to the maintenance of an anti-inflammatory state in animals treated with glucocorticoid or stressed for longer than 24 h. Finally, removal of neutrophils from the circulation for culture must uncouple their L-selectin expression system, as intracellular L-selectin proteins were rapidly degraded in vitro, where DEX had variable effects on surface L-selectin despite its consistent down-regulating effects on L-selectin mRNA.
|
ACKNOWLEDGEMENTS |
|---|
Received October 25, 2003; revised December 30, 2003; accepted January 2, 2004.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
) that were driven by increases in circulating G1+ neutrophils (
). Circulating numbers of CD14+ monocytes (•) and CD2+ T cells (
) did not change in response to DEX administration, but there was a decrease in some other leukocyte subset(s) (x) that may have been 
T cells [
