Vol. 276, Issue 6, R1587-R1594, June 1999
Vestigial respiratory burst activity in wound macrophages
Christopher C.
Nessel,
William L.
Henry Jr.,
Balduino
Mastrofrancesco,
Jonathan S.
Reichner, and
Jorge E.
Albina
Department of Surgery, Rhode Island Hospital and Brown
University, Providence, Rhode Island 02903
 |
ABSTRACT |
Macrophages from experimental wounds in
rats were tested for their capacity to generate reactive oxygen
intermediates. Measurements of superoxide and
H2O2
release, O
2-dependent lucigenin
chemiluminescence, oxygen consumption, hexose monophosphate shunt flux,
and NADPH oxidase activity in cell lysates indicated, at best, the
presence of a vestigial respiratory burst response in these cells. The
inability of wound cells to release
O2 was not rekindled by priming
with endotoxin or interferon-
in vivo or in vitro. NADPH oxidase
activity in a cell-free system demonstrated that wound macrophage
membranes, but not their cytosols, were capable of sustaining maximal
rates of O2 production when mixed
with their corresponding counterparts from human neutrophils. Immune
detection experiments showed wound macrophages to be particularly
deficient in the cytosolic component of the NADPH oxidase
p47-phox. Addition of
recombinant p47-phox to the human
neutrophil-cell membrane/wound macrophage cytosol cell-free oxidase
assay, however, failed to support
O2 production. Present findings
indicate an unexpected deficit of wound macrophages in their capacity
to generate reactive oxygen intermediates.
rodent; monocytes; inflammation
| |
INTRODUCTION |
THE REACTIVE OXYGEN intermediates (ROI) produced by
phagocytes through the NADPH oxidase enzyme complex are thought to
mediate some of the anti-infectious and antitumor activities of
macrophages. The capacity of these effector cells to produce ROI, in
turn, appears to be modulated by tissue environmental factors. In this regard, in vitro maturation of blood monocytes, their passage into
noninflamed compartments like the peritoneal cavity, or their differentiation into Kupffer cells has been reported to
result in a decreased ability to produce ROI (12, 19). A reduction in
ROI production by monocytes extravasated into normal tissues could be
interpreted as protective and beneficial, in light of the potential for
bystander damage by the unchecked release of such ROI.
In contrast with events in normal tissues, the participation of
macrophages in inflammatory responses has been indicated to correlate
with an increased ability of these cells to generate oxygen-derived
radicals (10). It was with this premise in mind, namely, that
macrophages participating in inflammatory responses should have an
increased capacity to release ROI, that experiments to determine the
production of these intermediates by macrophages obtained from
experimental wounds were performed. Data to be shown unexpectedly
demonstrate that these cells have a markedly reduced capacity to
produce and release ROI in vitro following stimulation with classic
activators of the respiratory burst oxidase and point to cytosolic
components of the NADPH oxidase system in the macrophages as mediators
of this deficit.
| |
METHODS |
Wound model, cells, and cell culture.
The experimental wound used was the subcutaneously implanted polyvinyl
alcohol sponge (PVA) model previously described (4). Briefly, 10 circular sterile PVA sponges (M-Pact, Eudora, KS) measuring ~1 cm in
diameter were implanted subcutaneously in the dorsum of anesthetized
male Fischer rats (VAF-Plus; Charles River Laboratories, Wilmington, MA) through a midline skin incision. The sponge wounds were retrieved 3-10 days after wounding (4). For macrophage harvest, sponges were
removed from the animals after CO2
euthanasia and minced into Hanks' balanced salt solution (HBSS, GIBCO;
Life Technologies, Grand Island, NY) containing 1% FCS
(Hyclone, Logan, UT). The cells contained in the sponges were isolated
by repeated rapid compression using a Stomacher (Tecmar, Cincinnati,
OH) for 30 s. Erythrocytes were lysed by brief exposure to distilled
water, and the remaining cells plated in culture medium consisting of L-arginine-free RPMI 1640 (Life
Technologies) with 1% BSA (Sigma, St. Louis, MO) or 1% FCS, 5 × 105 M 2-mercaptoethanol, 10 mM MOPS, and antibiotics at a cell density of 1.5 × 106 cells/ml. After 2-h
incubation, nonadherent cells were washed out and fresh medium was
added to the culture. Cells obtained in this manner are >90%
macrophages, according to Wright, nonspecific esterase,
and KU-1 antibody staining (6). Unless specifically indicated, studies
were performed using macrophages obtained from 10-day-old wounds.
Peritoneal macrophages were obtained from wounded and nonwounded rats
by peritoneal lavage (1, 4). Cells were allowed to adhere prior to
assay for superoxide release, phagocytosis, or glucose metabolism. When
indicated, cells were incubated overnight in culture
medium at a density of 2 × 106 cells/ml. Culture medium was
supplemented with murine recombinant interferon-
(rIFN-, 10 U/ml; Genzyme, Cambridge, MA) and/or 1 µg/ml
lipopolysaccharide (LPS, Escherichia
coli serotype 055:B5; DIFCO, Detroit, MI) when so indicated.
Additional wounded animals were injected with 2 µg/kg LPS
intraperitoneally 12 h prior to the harvest of peritoneal cells and
10-day-old wounds. Peritoneal and wound cells were isolated and used in
a O2 release assay as indicated below.
For chemiluminescence, O2
consumption, and
H2O2
release measurements, which require nonadherent cells, wound
macrophages were selected by adherence, detached using ice-cold
Ca2+- and
Mg2+-free HBSS, and resuspended as
indicated. Rat resident peritoneal macrophages cannot be detached from
tissue culture treated or microbiologic quality plastic ware without
substantial loss of viability. Unfractionated peritoneal lavage cells
were used, therefore, for the aforementioned determinations.
Differential count of these cells was as follows: 54.9%
monocyte/macrophages, 24% lymphocytes, 11.8% mast cells, 0.5%
polymorphonuclear leukocytes (PMNs), and 8.8% others
(basophils, eosinophils, and unidentified). Preliminary experiments
demonstrated no significant difference in
O2 production normalized for cell
number between unfractionated and adhered peritoneal lavage cells in a
ferricytochrome c
reduction assay after phorbol 12-myristate 13-acetate
(PMA) stimulation (unfractionated peritoneal lavage cells = 6.1 ± 1.0 nmol · 106
cells1 · h1 vs. adherent
peritoneal lavage cells = 7.2 ± 0.6 nmol · 106
cells1 · h1;
n = 8 per treatment,
P > 0.05, unpaired
t-test).
Human PMNs were isolated from the blood of normal volunteers using
Ficoll-Hypaque-dextran centrifugation. Rat PMNs were obtained from
peritoneal lavage 6 h after the intraperitoneal injection of 10 ml 1%
oyster glycogen (Sigma). Rat monocytes were separated from blood
obtained by cardiac puncture using Lympholyte M (Cedarlane, Ontario, Canada).
Superoxide release. Superoxide release
was measured using the ferricytochrome
c reduction assay described by Pick
and Mizel (21). Briefly, adherent resident peritoneal or wound cells (3 × 105/well) were overlaid
with 100 µl of a 160 µM solution of ferricytochrome c (Sigma) in phenol red-free HBSS.
Control wells contained cells and superoxide dismutase (SOD, 300 U/ml;
Sigma). Cells were stimulated with PMA (LC Services, Woburn, MA) or
opsonized zymosan. The cells were then incubated at 37°C for 60 min, and the absorbance at 550 nm against a reference 630-nm filter was
measured in a spectrophotometric plate reader (model EL340; Biotek
Instruments, Winooski, VT). Superoxide production was calculated from
the difference in absorbance at 550 nm (corrected for absorbance at 630 nm) and the extinction coefficient for the absorbance of reduced
ferricytochrome c using the equation
E550 nm 630 nm = 2.1 × 104
M1 · cm1.
Results thus obtained were normalized by cell number. Maximal O2 release from resident
peritoneal macrophages was found to occur with 200 nM PMA or 100 µg/ml opsonized zymosan. No detectable
O2 release from wound macrophages was found when agonist concentrations were increased up to 10-fold.
H2O2
release assay.
Hydrogen peroxide release was assayed fluorometrically using
horseradish peroxidase and 4-OH-phenylacetic acid as described by
Segura-Aguilar (23). For this determination, 1.5 × 106 peritoneal lavage cells or
wound cells were placed in 3 ml of 50 mM potassium
phosphate (pH 8.0), 1 mM EDTA, 1 mM sodium azide, 1 mM
4-OH-phenylacetic acid, and 8 U/ml horseradish peroxidase (Sigma). When
so shown, SOD was added to the cuvettes at 300 U/ml. Developing fluorescence was monitored using a luminescence spectrometer (model LS-50B; Perkin-Elmer, Newton Center, MA). Results are shown as
relative fluorescence intensity.
Lucigenin-dependent chemiluminescence.
For this determination, cells were dispensed at
107 cells/ml in 0.5 ml in phenol
red-free HBSS containing 1% FCS and 200 µM lucigenin (Sigma) into
the chamber of a photometer (Chem-Glow II; SLM Aminco, Urbana, IL).
Cells were stimulated or not with PMA (200 nM). The luminescent signal
was captured using an analog-to-digital converter (model MP100; Biopac
Systems, Goleta, CA) and proprietary data acquisition software. Signal calibration was obtained using a standardized radioactive source.
Oxygen consumption. Cells were
dispensed at 6 × 106
cells/ml in HBSS + 1% FCS into the chamber of an oxygen uptake system
(model 203B; Instech Laboratories, Plymouth Meeting, PA). Continuous recording of the O2 content of the
media was afforded using an Instech
O2 electrode and the Biopac model
MP 100 data acquisition system. The respiratory burst response was
triggered using PMA (200 nM).
Glucose metabolism. Glucose metabolism
was measured essentially as described previously (3). Cells (1 × 106 in 500 µl) were dispensed
into 17 × 100-mm polystyrene tubes, and nonadherent cells were
removed by washing 2 h later. Cells were then incubated in the presence
of [1-14C]glucose or
[6-14C]glucose
(DuPont-NEN, Cambridge, MA) at 0.5 µCi/ml and 1 mM glucose. An
aliquot of media was removed from the cultures immediately after the
addition of the respective radiolabeled substances and used for the
determination of the initial glucose-specific radioactivity. The tubes
were then closed with rubber stoppers fitted with center wells
containing filter paper and 200 µl Oxosol
14C (National
Diagnostics, Somerville, NJ). After incubation for 15-120 min at 37°C in a shaking water bath, 1 ml of 0.7 N
trichloroacetic acid was added to the tubes to evolve dissolved
CO2, and the tubes were replaced
in the water bath for an additional hour to allow the trapping of
14CO2
by the Oxosol 14C. The center
wells were then removed, and radioactivity in trapped 14CO2
was counted in a scintillation counter (LKB Pharmaceutical, Piscataway,
NJ). The deproteinized culture supernatants were analyzed by HPLC (3),
and the radioactivity contained in glucose or lactate was separated
using fraction collection. Glycolysis was calculated from the
appearance of 14C from
[1-14C]glucose in
lactate and the initial glucose-specific radioactivity. Hexose
monophosphate shunt (HMPS) activity was calculated from the production
of
14CO2
from [1-14C]glucose
and the initial glucose-specific radioactivity without correction for
14CO2
from [6-14C]glucose.
Glucose oxidation was calculated from the production of
14CO2
from [6-14C]glucose
and the initial glucose-specific radioactivity. Preliminary work
indicated the rates of glycolytic flux, HMPS activity, and glucose
oxidation in unstimulated cells were linear for at least 4 h in culture.
NADPH oxidase activity in cell-free
lysates. Peritoneal lavage cells or wound-derived
macrophages were dispensed at 1 × 106 in 1 ml of 17 mM potassium
phosphate buffer with 1.2 mM magnesium chloride, 5 mM potassium
chloride, 123 mM sodium chloride, 160 mM ferricytochrome
c, 6 mM sodium azide,
and 5 mM glucose in a 1-cm light-path cuvette and placed in a
spectrophotometer (model DMS90; Varian Techtron, Mulgrave, Australia).
Under constant stirring, absorbance at 550 nm was continuously
monitored and recorded. Following stabilization at 37°C, cells were
stimulated with PMA (200 nM), and the change in absorbance recorded for
5 min. At that time, cells were lysed with a mixture of deoxycholate
(0.0625%, wt/vol) and Tween 20 (0.0625% vol/vol) as indicated in
(27). Changes in absorbance after detergent addition were recorded for an additional 45 s, and NADPH (0.6 mM) was added to the cuvette. After
a 3-min recording, SOD (300 U/ml) was incorporated into the mixture.
Cell-free NADPH oxidase reconstitution
assay. Human blood PMNs or wound macrophages were
suspended at 108
cells/ml in relaxation buffer (0.1 M KCl, 3.5 mM NaCl, 3.5 mM MgCl2, 1.25 mM EGTA, and 10 mM HEPES buffer, pH 7.3) containing protease
inhibitors. Plasma membranes and cytosols in postnuclear supernatants
were separated after cell disruption by sonication by
ultracentrifugation at 134,000 g for
30 min (Air Fuge; Beckman, Palo Alto, CA) and stored frozen. NADPH
oxidase activity was measured by mixing
106 cell equivalents of cell
membranes and cytosols in 75 mM
KPO4 (pH 7.0), 0.2 mM
ferricytochrome c, 4 mM
MgCl2, 10 µM FAD, 1 mM EGTA, 10 µM GTPS, 200 µM NADPH, and 100 µM SDS. Control reactions contained SOD (300 U/ml). Superoxide production was calculated from the
reduction of ferricytochrome c as
described above.
Purified recombinant p47-phox from a
baculovirus expression vector was a kind gift of Dr. Thomas L. Leto
(13). A
p47-phox-glutathione-S-transferase (p47-phox-GST) fusion
protein was a kind gift of Dr. Bernard M. Babior (20). Both recombinant
p47-phox forms have been shown to
reconstitute activity in
p47-phox-deficient cell-free NADPH oxidase reconstitution assays (8, 13) and were added to the cell-free
system at 0.1-1 µg/ml.
Immunodetection of p47-phox and
p67-phox. Postnuclear supernatants of cell lysates (4 × 106 cell equivalents/lane
for rat PMNs and resident peritoneal and wound macrophages, and 5 × 105/lane for human
PMNs) were size fractionated in 10% SDS-polyacrylamide gel and transferred to a nitrocellulose membrane. Membranes were blocked in 5% BSA in PBS (pH 7.5)-0.05% Tween 20, incubated for 1 h
with a 1/1,000 dilution of goat antisera against human
p47-phox or
p67-phox (generously provided by Dr.
Thomas L. Leto; Ref. 13), and then with a rabbit anti-goat horseradish
peroxidase-conjugated antibody at 1/16,000 dilution.
p47-phox and
p67-phox were detected in a
chemiluminescence reaction using ECL reagent (Amersham, Arlington Heights, IL).
Phagocytosis. Sheep red blood cells
(sRBC) were incubated with a subagglutinating dose of monoclonal
anti-sRBC, subclass IgG2a (Accurate Scientific, Westbury, NY), for 1 h
at 37°C and radiolabeled with 100 µCi
Na51CrO4
for 1 h. Radiolabeled sRBC were added to macrophage monolayers for 1 h,
and phagocytosis was determined from radioactivity remaining in the
wells after the lysis of extracellular sRBC with distilled H2O and plate washing.
Endotoxin assay. Endotoxin in serum
and wound fluid was detected following manufacturer's instructions
using a chromogenic assay (QCL1000; Whittaker M.A. Bioproducts,
Walkersville, MD).
Data presentation and analysis.
Results shown are means ± SD from a representative of
5-10 independent experiments. Depending on the number of cells
required for each assay, 3-10 animals were used per experiment as
donors of peritoneal cells or wound macrophages. For each cell type,
cells from all animals in each experiment were pooled prior to assay.
Usual yield of peritoneal cells was of 1.2-1.5 × 107 cells per animal. Usual yield
of day 10 wound macrophages was of
5-7 × 106 cells per
animal. Unless otherwise indicated, statistical analysis was performed
by ANOVA-Newman-Keuls or Mann-Whitney U test.
| |
RESULTS |
Wound-derived macrophages do not produce detectable
O2 or
H2O2.
Table 1 illustrates the production of
O2 by freshly harvested resident
peritoneal and by wound macrophages obtained 10 days following injury.
Resident peritoneal macrophages, used here as a benchmark against which
to compare and contrast wound macrophages and as a positive control for
the production of ROI in each experiment, released
O2 measured as SOD-inhibitable
ferricytochrome c reduction (21) after
stimulation with PMA or zymosan (Table 1). Cytochrome
c reduction was fully inhibited by
pretreating the cells with diphenyleneiodonium (10 µM) or by adding
SOD (300 U/ml) to the assay. Rat blood monocytes produced 4.23 ± 0.35 nmol O2/h when stimulated with PMA.
Identically stimulated wound macrophages, in turn, produced no
detectable O2. Additional
experiments used wound macrophages harvested 3 or 5 days after injury.
These cells, like those harvested 10 days after wounding,
failed to release detectable superoxide when stimulated with PMA or
with zymosan.
The lack of detectable O2 release
from wound macrophages did not result from a reduced viability of the cell preparation. Data in Table 1 indicate that, even after overnight culture, wound macrophages were markedly more phagocytic than resident
peritoneal lavage cells for opsonized sRBC. In addition, trypan blue
exclusion testing of freshly harvested and overnight cultured wound
macrophages indicated the preservation of membrane integrity in >95%
of the cells.
The procedures required to isolate macrophages from the wounds did not
result in loss of the cellular capacity to release ROI. This is so
because peritoneal lavage cells processed in the manner described in
the METHODS for wound cells (mixed
with sterile sponges, compressed in the Stomacher, and adhered for 2 h)
produced as much O2 after PMA
stimulation as did cells that were just adhered (6.2 ± 0.2 nmol
O2/106
freshly adhered cells vs. 5.9 ± 0.1 nmol
O2/106
cells processed through the Stomacher,
P > 0.05, Mann-Whitney U test).
The lack of detectable O2
production by wound macrophages in the ferricytochrome
c reduction assay did not result from
their release of substances with cytochrome oxidase activity (i.e.,
H2O2,
NO or ONOO). Data in Fig.
1 show, in this regard, that
PMA-stimulated resident peritoneal macrophages rapidly and
abundantly released products that induced lucigenin chemiluminescence.
Wound macrophages, in contrast, exhibited a markedly delayed and
blunted chemiluminescence response (Fig. 1). Lucigenin
chemiluminescence required O2 production, since it was fully abrogated by SOD (300 U/ml).
View larger version (16K):
[in this window]
[in a new window]
|
Fig. 1.
O2-dependent lucigenin chemiluminescence.
Peritoneal lavage cells or wound macrophages were dispensed at
107 cells/ml into 0.5 ml HBSS
containing 200 µM lucigenin and placed in the chamber of a Chem-Glo
II photometer. Cells were stimulated with phorbol 12-myristate
13-acetate (PMA, 200 nM), and the chemiluminescent signal was recorded
and expressed in volts (V). The square wave was generated using a
radioactive standard and indicates similar instrument calibration for
both runs. Chemiluminescence was completely abrogated by superoxide
dismutase (SOD, 300 U/ml) (not shown).
|
|
In parallel with findings regarding
O2 release, results shown in
Fig. 2 demonstrate the abundant produc-tion of
H2O2
by PMA-stimulated resident peritoneal macrophages and its negligible
release from stimulated 10-day wound macrophages. Indeed, although the
addition of SOD to the assay increased the amount of detectable
H2O2
liberated from resident peritoneal macrophages, its presence was
without effect in wound cells.
View larger version (17K):
[in this window]
[in a new window]
|
Fig. 2.
H2O2
release by peritoneal lavage (open symbols) or wound macrophages (solid
symbols). Cells were isolated as described in
METHODS.
H2O2
release after PMA stimulation (200 nM) was measured over time using
4-OH-phenylacetic acid and horseradish peroxidase. When indicated, SOD
(300 U/ml) was present in the cuvettes prior to the addition of PMA.
 , Wound macrophages without SOD;  , wound macrophages with SOD;
 , peritoneal lavage cells without SOD;  , peritoneal lavage cells
with SOD.
|
|
Wound-derived macrophages lack a respiratory burst
following PMA stimulation. To determine whether the
lack of detectable ROI release from wound macrophages actually resulted
from an inoperative respiratory burst system and not from the efficient
removal of its products by intracellular antioxidant mechanisms, the
rate of oxygen consumption by these cells and by peritoneal lavage cells was established prior to and after PMA stimulation. Figure 3 presents data from one such experiment
and shows a robust increase in O2
consumption by the peritoneal cells following stimulation with PMA and
a complete lack of respiratory burst activity in similarly stimulated
wound macrophages harvested 10 days after injury.
View larger version (13K):
[in this window]
[in a new window]
|
Fig. 3.
O2 consumption by peritoneal
lavage or wound macrophages. Cells isolated as described in
METHODS were placed in the chamber of
an Instech model 203B O2 uptake
system. Oxygen content in culture media was monitored continuously
prior to and after addition of PMA (200 nM).
|
|
Interestingly, the rate of spontaneous
O2 consumption by wound
macrophages was higher than that of resident peritoneal cells. Because
spontaneous O2 consumption in both
cells was abrogated by KCN (5 mM) (not shown), it appears to reflect
mitochondrial respiration. In agreement with this conclusion,
wound-derived macrophages oxidized more glucose through the
tricarboxylic acid cycle than their peritoneal counterparts (wound
macrophages = 1.3 ± 0.3 pmol glucose/min per
106 cells vs. resident peritoneal
macrophages = 0.2 ± 0.1 pmol glucose/min per
106 cells;
P < 0.05, unpaired
t-test). Ruling out that the increased O2 consumption of resident
peritoneal macrophages following PMA stimulation was related to the
production of NO, neither
NG-monomethyl-L-arginine
(0.5 mM) nor N-iminoethyl-L-ornithine (0.1 mM), two inhibitors of NO synthase with different
mechanisms of action, affected basal or stimulated
O2 consumption by the cells (not shown).
Impact of PMA on glucose metabolism through glycolysis
and the HMPS. NADPH utilized during the generation of
ROI by the NADPH oxidase is furnished primarily by the HMPS. Flux of
glucose metabolites though the shunt is greatly increased during the
respiratory burst response. Resident peritoneal and wound macrophages
had similar rates of HMPS activity prior to PMA stimulation (wound
macrophages = 0.16 ± 0.01 nmol glucose/min per
106 cells vs. resident peritoneal
macrophages = 0.14 ± 0.01 nmol glucose/min per
106 cells;
P > 0.05, unpaired
t-test). Peritoneal macrophages
rapidly and substantially increased their HMPS flux when treated with PMA (Fig. 4,
inset). In a manner reminiscent of
findings in chemiluminescence experiments, wound macrophages increased
their HMPS activity only to a minor extent after phorbol ester
stimulation.
View larger version (20K):
[in this window]
[in a new window]
|
Fig. 4.
Effects of PMA stimulation on glucose metabolism through glycolysis and
the hexose monophosphate shunt (HMPS) by resident peritoneal
(A) and wound macrophages
(B). Cells were cultured with
[1-14C]glucose and
stimulated or not with PMA (200 nM). Cumulative glucose flux through
the indicated pathways after PMA treatment was determined every 15 min
for 120 min and is shown in the main graph. Columns indicate the
consumption of glucose through the pathways of interest by 120 min by
cells not treated with PMA. Insets:
glucose metabolism through the HMPS by each cell type with and without
PMA stimulation. Axes for the insets
are the same as those for the main graph.
|
|
The contrasting magnitude of the HMPS response to PMA in resident
peritoneal and wound macrophages can be better seen if it is shown in
the context of their capacity to metabolize glucose through that
pathway and through glycolysis. Area plots in Fig. 4 show
[1-14C]glucose
metabolism through these two pathways by either cell type for up to 120 min in culture following PMA stimulation. For comparison, fluxes are
also shown for cells cultured without PMA for 120 min (bars in
Fig. 4). As can be seen in Fig. 4, the fraction of [1-14C]glucose that
was processed through the HMPS by PMA-treated resident peritoneal
macrophages accounted for at least 50% of all
[1-14C]glucose that
was metabolized by these cells for the duration of the experiment.
Peritoneal macrophages not treated with PMA processed ~10% of
available
[1-14C]glucose through
the HMPS.
Wound macrophages were more glycolytic than their peritoneal
counterparts, markedly increased their glycolytic flux after PMA, thus
showing their capacity to bind and respond to phorbol esters, and
metabolized at most 5% of total
[1-14C]glucose
processed through the HMPS, whether stimulated with PMA or not.
Wound-derived macrophages do not show evidence of
SOD-inhibitable NADPH oxidase activity following cell
lysis. Experiments were also performed
that measured NADPH oxidase activity in cell-free lysates. Results
shown in Fig. 5 illustrate the activity of
NADPH oxidase in peritoneal lavage cells and in wound macrophages when measured following cell lysis. Results in Fig. 5 for
resident peritoneal cells demonstrate, sequentially, the production of O2 that followed stimulation with
PMA, its suppression after cell lysis, and its recovery upon addition
of NADPH (0.6 mM) to the cell lysate. Ferricytochrome
c reduction by peritoneal cell lysates
was fully suppressed by SOD.
View larger version (13K):
[in this window]
[in a new window]
|
Fig. 5.
Evidence that wound macrophage lysates do not demonstrate NADPH oxidase
activity. Cells were harvested and processed as described in
METHODS, and the reduction of
ferricytochrome c was recorded at 550 nm. Arrows, times of addition for PMA (200 nM), deoxycholate + Tween
(DOC), NADPH (0.6 mM), and SOD (300 U/ml). OD, change in optical
density.
|
|
Intact wound cells, in turn, elicited no ferricytochrome
c reduction after PMA stimulation, but
initiated rapid cytochrome reduction upon lysis and NADPH addition.
Ferricytochrome c reduction by wound
macrophage lysates, however, did not reflect
O2 production, because it was not
inhibitable by SOD. Data obtained in these experiments indicate the
vestigial respiratory burst found in wound macrophages does not result
from a reduced, limiting NADPH content in these cells.
In vivo or in vitro priming of peritoneal or wound
macrophages fails to alter
O2 release.
It has been reported that culture with LPS or IFN- as priming agents
increases the responsiveness of macrophages to subsequent stimulation
in terms of respiratory burst activity (17). Alternatively, overnight
culture of murine macrophages with LPS and IFN- was shown to
suppress NADPH oxidase activity, an effect putatively assigned to the
induction of inducible NO synthase in the cells and the
subsequent inhibitory effects of NO on the respiratory burst enzyme
system (11). Neither priming nor inhibition of PMA-induced
O2 production was found when rat peritoneal macrophages were cultured overnight in the presence of
IFN- (1-100 U/ml), LPS (1 µg/ml), or their combinations,
prior to stimulation (Table 2). Wound
macrophages did not release detectable O2 after overnight culture,
regardless of culture conditions.
Almost identical results were obtained when peritoneal and wound
macrophages were isolated from animals that had received intraperitoneal LPS (2 µg/kg) or saline solution 12 h prior to cell
harvesting. That LPS injection resulted in significant LPS content in
plasma and wound fluid by the time of cell harvest was demonstrated by
the finding of 2.7 ± 0.7 endotoxin units/ml in plasma and 2.7 ± 0.9 endotoxin units/ml in wound fluid obtained from the harvested
sponges (n = 5 for each
determination). Endotoxin activity in plasma and sponge fluid of
control animals injected with saline was below the limit of detection
of the assay (0.05 endotoxin units/ml). Superoxide release from
PMA-stimulated peritoneal macrophages from wounded, LPS-treated animals
was 3.37 ± 0.12 nmol · 106
cells1 · h1
and not different from that released from peritoneal cells from wounded, saline-treated animals (4.33 ± 0.18 nmol · 106
cells1 · h1;
P > 0.05, unpaired
t-test). Wound macrophages from these
animals did not release detectable
O2.
Cell-free reconstitution experiments point to a
cytosolic defect as the mechanism for the reduced NADPH oxidase
activity of wound macrophages. To attempt to clarify
the mechanism for the vestigial respiratory burst activity of wound
macrophages, experiments were performed using human PMN membranes or
cytosols in combination with the corresponding components of wound
macrophages in a cell-free NADPH oxidase assay (25). Results from these
experiments, shown in Fig. 6, indicated
that wound macrophage cell membranes and human PMN cell membranes were
capable of sustaining similar rates of SOD-inhibitable ferricytochrome
c reduction when combined with human
PMN cytosols. In contrast, macrophage-derived cytosols were unable to
attain the same rate of O2
production when combined with human PMN membranes as did PMN cytosols.
These findings suggested that the mechanism for the vestigial
respiratory burst of wound macrophages resides in their cytosolic, and
not in their membrane-bound, NADPH oxidase components.
View larger version (13K):
[in this window]
[in a new window]
|
Fig. 6.
Wound macrophage cell membranes (B),
but not their cytosols (A), allow
maximal O2 production when mixed
with their human polymorphonuclear leukocyte (PMN) counterparts in a
cell-free NADPH oxidase reconstitution system. Human PMN or wound
macrophage cell membranes or cytosols from
106 cells/well were mixed as
described in METHODS and assayed for
O2 production in a ferricytochrome
c reduction assay.
|
|
Wound macrophages contain less p47-phox than human or
rat PMNs and rat resident peritoneal macrophages: addition of
recombinant p47-phox to cell-free human PMN membrane-wound macrophage
cytosol reconstitution system does not increase NADPH oxidase
activity. Attempting to identify the
cytosolic component responsible for the reduced NADPH oxidase activity
of wound macrophages, the p47-phox and
p67-phox components of the oxidase
system were measured in cell lysates, using human and rat blood PMNs
and rat resident peritoneal macrophage lysates for comparison. As can
be seen in Fig. 7, macrophage cell lysates
contained less p47-phox and
p67-phox than PMN lysates. Most
strikingly, wound macrophage lysates contained virtually no
immunoreactive p47-phox antigen by
Western blot analysis.
View larger version (40K):
[in this window]
[in a new window]
|
Fig. 7.
Western blot analysis of p47-phox and
p67-phox in cell lysates. Postnuclear
supernatants of rat glycogen-elicited peritoneal PMNs
(lane 1), resident peritoneal
macrophages (lane 2), wound
macrophages (lane 3), and human
blood PMNs (lane 4) were size
fractionated and immunoreactive
p47-phox and
p67-phox detected as described in the
METHODS. Lanes
1-3 were loaded with 4 × 106 cell equivalents and
lane 4 with 5 × 105 cell
equivalents.
|
|
On the basis of these findings then, NADPH oxidase activity was
measured in the cell-free NADPH oxidase assay using human PMN cell
membranes and wound macrophage cytosols supplemented with recombinant
p47-phox added either as a purified
product obtained from a baculovirus expression system or as a
p47-phox-GST fusion protein. In
neither case was NADPH oxidase activity increased by the addition of
the recombinant protein (results not shown).
| |
DISCUSSION |
Work in this laboratory has focused on the phenotypic characterization
of macrophages participating in the reparative processes that follow
tissue injury. Experiments reported here addressed the capacity of
wound macrophages to produce and release ROI and demonstrate these
cells to be remarkably silent in this regard. This
particular functional deficit of wound macrophages contrasts with their
demonstrated ability to carry out phagocytosis, not only of opsonized
latex beads as reported here but also of apoptotic wound neutrophils
(16), and their capacity to release cytokines (2).
Evidence for a vestigial, markedly blunted, respiratory burst capacity
in wound macrophages was provided by results from chemiluminescence assays and by a subtle increase in HMPS activity in wound macrophages following PMA stimulation. Tests of wound macrophages for
O2 or
H2O2
release, measurements of O2
consumption, and assays for NADPH activity in cell lysates yield
uniformly negative results. The relative inability of the wound
macrophage to generate ROI appears to be acquired early in the process
of repair, since its immediate precursor, the blood monocyte, is
capable of producing O2, whereas
macrophages isolated 3 or 5 days after wounding fail to release ROI
after appropriate stimulation.
Additional findings in these experiments included marked metabolic
differences between wound and peritoneal macrophages. It was found, in
this connection, that wound macrophages have a higher basal rate of oxygen consumption and glucose oxidation than peritoneal macrophages. In this regard, it has been proposed that the
oxidative/glycolytic partition of glucose metabolism in macrophages is
substantially influenced by the oxygen content of their natural
microenvironment. Alveolar macrophages, normally exposed to a high
oxygen environment, were shown to be more oxidative, and
correspondingly less glycolytic, than peritoneal macrophages which
reside in conditions of relative oxygen deprivation (7, 22).
Interestingly, wound macrophages are thought to exist in a relatively
hypoxic milieu. Niinikoski et al. (18), using a rat wound model similar
to that employed in the currently reported experiments, demonstrated a
persistently reduced oxygen tension in the wound over time, with a
PO2 of
~10 mmHg, 10 days after wounding. Silver, in turn, indicated that the
oxygen tension in the core of rabbit ear chamber wounds ranges between
0 and 2 mmHg (24). The finding of higher oxidative capacity in wound
macrophages than in peritoneal macrophages appears to contradict the
concept that the oxidative capacity of macrophages is determined by the
prevailing oxygen content of their site of residence.
Returning to the markedly reduced capacity of wound macrophages to
generate ROI, efforts to explain current observations from a molecular
standpoint failed to provide fully satisfactory results. Findings in
cell-free NADPH oxidase reconstitution experiments, however, were
interesting in demonstrating that wound macrophage cell membranes are
equivalent to those obtained from human PMNs in supporting
O2 superoxide production in a cell-free NADPH oxidase assay system. Results from these experiments also pointed toward a cytosolic component(s) of the macrophage oxidase
system as responsible for the cells' comparatively reduced ability to
release O2.
Western blot analysis of the cytosolic NADPH oxidase components
p47-phox and
p67-phox in cell lysates obtained from
rat or human PMNs and from rat resident peritoneal or wound macrophages showed macrophages to contain lesser amounts of both proteins than
PMNs. This reduced content was more marked in the case of p47-phox and especially striking in
wound macrophages. These cells contained basically no detectable
p47-phox. Supplementation of the
cell-free NADPH oxidase assay with recombinant
p47-phox, however, failed to rekindle
oxidase activity. This observation suggests other unidentified
deficiencies in oxidase components in wound macrophages (i.e., the
Ras-related GTPase Rac; Ref. 26) or defects in the assembly of those
components in these cells as explanations for their reduced capacity to
release ROI.
Present findings may be interpreted in the light of the proposed
functions of the ROI produced by phagocytes through the NADPH oxidase.
These functions, both beneficial and potentially detrimental to the
host, include the elimination of invading microorganisms and the damage
of adjacent tissues (9, 15). In regard to the first
function, abundant evidence indicates the participation of ROI in the
antibacterial actions of phagocytes (9). The sterile environment of a
noninfected, normally healing wound may not provide the signals found
in infected spaces for the acquisition of the capacity to generate ROI.
Alternatively, the presence of molecules with macrophage-deactivating
capacity (i.e., cytokines like interleukin-4, interleukin-10, or
transforming growth factor-
; prostanoids like
PGE2 or others) within the wound
may result in the decreased capacity of wound cells to produce ROI that
is here described. In this regard, then, it was of interest to observe that endotoxin arrival into the wound did not promote respiratory burst
capacity in wound macrophages.
In connection to the potential for bystander injury
mediated by ROI released within the confined limits of a local
inflammatory response, the lack of substantial tissue injury during
acute inflammation, as opposed to chronic, persistent inflammation has
been recently highlighted (5). The inability of wound macrophages,
which constitute the predominant cell type in this model in wounds 3 days old or more, may contribute an explanation for the aforementioned lack of tissue injury during acute inflammation.
Data reported here expand the characterization of the wound macrophage
phenotype. Previous work from this laboratory indicated that wound
macrophages are not cytotoxic for NO-sensitive tumor cells, even when
stimulated to produce abundant NO, an observation suggestive of a
defect in the production or delivery of cytotoxic molecules by these
cells (14). Also reported was the ability of wound macrophages to
exhibit activation-associated traits when cultured under anoxic
conditions (2). Contributing to the definition of the wound macrophage
phenotype, results shown here illustrate the complete, or almost
complete, inability of wound macrophages to release ROI.
| |
ACKNOWLEDGEMENTS |
This work was supported by National Institute of General Medical
Science Grants GM-42859 (J. E. Albina) and GM-51493 (J. S. Reichner),
by the Anita Allard Memorial Fund, and by allocations to the Department
of Surgery by Rhode Island Hospital.
| |
FOOTNOTES |
The costs of publication of this
article were defrayed in part by the
payment of page charges. The article
must therefore be hereby marked
"advertisement"
in accordance with 18 U.S.C. §1734 solely to indicate this fact.
Address for reprint requests and other correspondence: J. E. Albina,
Division of Surgical Research, Dept. of Surgery, Rhode Island Hospital,
593 Eddy St., Providence, RI 02903 (E-mail:
Jorge_Albina{at}Brown.edu).
Received 9 November 1998; accepted in final form 18 February 1999.
| |
REFERENCES |
1.
Albina, J. E.,
J. A. Abate,
and
W. L. Henry, Jr.
Nitric oxide production is required for murine resident peritoneal macrophages to suppress mitogen-stimulated T cell proliferation.
J. Immunol.
147:
144-148,
1991[Abstract].
2.
Albina, J. E.,
W. L. Henry, Jr.,
B. Mastrofrancesco,
B.-A. Martin,
and
J. S. Reichner.
Macrophage activation by culture in an anoxic environment.
J. Immunol.
155:
4391-4396,
1995[Abstract].
3.
Albina, J. E.,
and
B. Mastrofrancesco.
Modulation of glucose metabolism in macrophages by products of nitric oxide synthase.
Am. J. Physiol.
264 (Cell Physiol. 33):
C1594-C1599,
1993[Abstract/Free Full Text].
4.
Albina, J. E.,
C. D. Mills,
A. Barbul,
C. E. Thirkill,
W. L. Henry, Jr.,
B. Mastrofrancesco,
and
M. D. Caldwell.
Arginine metabolism in wounds.
Am. J. Physiol.
254 (Endocrinol. Metab. 17):
E459-E467,
1988[Abstract/Free Full Text].
5.
Ali, H.,
B. Haribabu,
R. M. Richardson,
and
R. Snyderman.
Mechanisms of inflammation and leukocyte activation.
Med. Clin. North Am.
81:
1-28,
1997[Medline].
6.
Bodenheimer, H. C.,
R. A. Faris,
C. Charland,
and
D. C. Hixson.
Characterization of a new monoclonal antibody to rat macrophages and Kupffer cells.
Hepatology
8:
1667-1672,
1988[Medline].
7.
Butterick, C. J.,
D. A. Williams,
L. A. Boxer,
R. A. Jersild, Jr.,
N. Mantich,
C. Higgins,
and
R. L. Baehner.
Changes in energy metabolism, structure and function in alveolar macrophages under anaerobic conditions.
Br. J. Haematol.
48:
523-532,
1981[Medline].
8.
El Benna, J.,
J.-W. Park,
J. M. Ruedi,
and
B. M. Babior.
Cell-free activation of the respiratory burst oxidase by protein kinase C.
Blood Cells Mol. Dis.
21:
201-206,
1995[Medline].
9.
Fantone, J. C.,
and
P. A. Ward.
Role of oxygen-derived free radicals and metabolites in leukocyte-dependent inflammatory reactions.
Am. J. Pathol.
107:
397-418,
1982.
10.
Johnston, R. B., Jr.,
C. A. Godzik,
and
Z. A. Cohn.
Increased superoxide anion production by immunologically activated and chemically elicited macrophages.
J. Exp. Med.
148:
115-127,
1978[Abstract/Free Full Text].
11.
Jun, C.-D.,
J.-Y. Lee,
B.-S. Lee,
B.-M. Choi,
J.-Y. Um,
H.-J. Kwak,
K.-Y. Ji,
H.-M. Kim,
and
H.-T. Chung.
Generation of nitric oxide inhibits formation of superoxide in macrophages during activation.
Biochem. Mol. Biol. Int.
34:
1-8,
1994[Medline].
12.
Lepay, D. A.,
C. F. Nathan,
R. M. Steinman,
H. W. Murray,
and
Z. A. Cohn.
Murine Kupffer cells. Mononuclear phagocytes deficient in the generation of reactive oxygen intermediates.
J. Exp. Med.
161:
1079-1096,
1985[Abstract/Free Full Text].
13.
Leto, T. L.,
M. C. Garrett,
H. Fujii,
and
H. Nunoi.
Characterization of neutrophil NADPH oxidase factors p47-phox and p67-phox from recombinant baculoviruses.
J. Biol. Chem.
266:
19812-19818,
1991[Abstract/Free Full Text].
14.
Mateo, R. B.,
J. S. Reichner,
and
J. E. Albina.
NO is not sufficient to explain maximal cytotoxicity of tumoricidal macrophages against an NO-sensitive cell line.
J. Leukoc. Biol.
60:
245-252,
1996[Abstract].
15.
McCord, J. M.
Superoxide radical: controversies, contradictions, and paradoxes.
Proc. Soc. Exp. Biol. Med.
209:
112-117,
1995[Abstract].
16.
Meszaros, A. J.,
J. S. Reichner,
and
J. E. Albina.
Macrophage phagocytosis of wound neutrophils.
J. Leukoc. Biol.
65:
35-42,
1999[Abstract].
17.
Morel, F.,
J. Doussiere,
and
P. V. Vignais.
The superoxide-generating oxidase of phagocytic cells: physiological, molecular and pathological aspects.
Eur. J. Biochem.
201:
523-546,
1991[Medline].
18.
Niinikoski, J.,
C. Heughan,
and
T. K. Hunt.
Oxygen and carbon dioxide tensions in experimental wounds.
Surg. Gynecol. Obstet.
133:
1003-1007,
1971[Medline].
19.
Pabst, M. J.,
H. B. Hedegaard,
and
R. B. Johnston, Jr.
Cultured human monocytes require exposure to bacterial products to maintain an optimal oxygen radical response.
J. Immunol.
128:
123-128,
1982[Abstract].
20.
Park, J.-W.,
J. El Benna,
K. E. Scott,
B. L. Christensen,
S. J. Chanock,
and
B. M. Babior.
Isolation of a complex of respiratory burst oxidase components from resting neutrophil cytosol.
Biochemistry
33:
2907-2911,
1994[Medline].
21.
Pick, E.,
and
D. Mizel.
Rapid microassays for the measurement of superoxide and hydrogen-peroxide production by macrophages in culture using an automatic enzyme immunoassay reader.
J. Immunol. Methods
46:
211-226,
1981[Medline].
22.
Romeo, D.,
G. Zabucchi,
T. Marzi,
and
F. Rossi.
Kinetic and enzymatic features of metabolic stimulation of alveolar and peritoneal macrophages challenged with bacteria.
Exp. Cell Res.
78:
423-432,
1973[Medline].
23.
Segura-Aguilar, J.
A new direct method for determining superoxide dismutase activity by measuring hydrogen peroxide formation.
Chem. Biol. Interact.
86:
69-78,
1993[Medline].
24.
Silver, I. A.
The measurement of oxygen tension in healing tissue.
Prog. Respir. Res.
3:
124-135,
1969.
25.
Smith, R. M.,
J. T. Curnutte,
L. A. Mayo,
and
B. M. Babior.
Use of an affinity label to probe the function of the NADPH binding component of the respiratory burst oxidase of human neutrophils.
J. Biol. Chem.
264:
12243-12248,
1989[Abstract/Free Full Text].
26.
Tardif, M.,
M.-J. Rabiet,
T. Christophe,
M.-D. Milcent,
and
F. Boulay.
Isolation and characterization of a variant HL60 cell line defective in the activation of the NADPH oxidase by phorbol myristate acetate.
J. Immunol.
161:
6885-6895,
1998[Abstract/Free Full Text].
27.
Tsunawaki, S.,
and
C. F. Nathan.
Release of arachidonate and reduction of oxygen. Independent metabolic bursts of the mouse peritoneal macrophage.
J. Biol. Chem.
261:
11563-11570,
1986[Abstract/Free Full Text].
Am J Physiol Regul Integr Compar Physiol 276(6):R1587-R1594
0002-9513/99 $5.00
Copyright © 1999 the American Physiological Society
TOP
ABSTRACT
INTRODUCTION
