| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
*
Department of Pathology, Division of Cytometry, Cancer Research Facility, University of New Mexico School of Medicine, Albuquerque, NM 87131; and
National Flow Cytometry Resource, Los Alamos National Laboratory, Los Alamos, NM 87545
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
L-selectin shedding is one important aspect of the normal physiologic regulation of L-selectin adhesive function. Cell surface expression of this adhesion molecule is characteristically down-modulated in response to cell activation (3, 15, 16). This proteolytic release from neutrophils inhibits subsequent L-selectin-dependent interactions with other neutrophils and endothelial cells at inflammatory sites (17, 18). Lymphocyte L-selectin is shed in response to activation by PMA (15), bacterial superantigens (19), or, like neutrophil L-selectin, by the treatment of cells with Abs to L-selectin (20, 21). L-selectin loss results in profound changes in T cell recirculation pathways (22), and studies with L-selectin-deficient mice have revealed a dramatic (70–90%) reduction in the number of lymphocytes in peripheral lymph nodes (22, 23). The released, soluble L-selectin retains binding capacity and may function as an adhesive buffer by preventing leukocyte adhesion at sites of subacute inflammation (24). Increased levels of plasma L-selectin are found in several disease states, including AIDS (25).
L-selectin shedding is the result of a proteolytic cleavage close to
its transmembrane domain, conducted by a constitutively active membrane
metalloprotease (26, 27), recently shown to be identical
with TNF-
converting enzyme (28). Several groups,
including ours, have shown that hydroxamic acid-based inhibitors of
matrix metalloproteases, such as
(N-{D,L-[2-(hydroxyaminocarbonyl)-methyl]-4-methylpentanoyl}-L-3-(tert-butyl)-alanyl-L-alanine,
2-aminoethyl amide
(TAPI-2),3 inhibit the
L-selectin sheddase and have used these compounds to study the
physiological consequences of L-selectin retention (18, 29, 30, 31). While it is clear that the susceptibility to this
protease is determined by the tertiary structure of L-selectin, the
mechanism by which these conformational changes are modulated remains
unclear (27, 32). Interestingly, calmodulin has recently
been found to be associated with the cytoplasmic domain of L-selectin,
and calmodulin inhibitors were shown to induce L-selectin shedding
through a protease-dependent mechanism (33).
We report here on the regulation of L-selectin shedding by sulfhydryl reagents. In an effort to begin to understand the mechanisms of shedding, we have studied in detail the effect of phenylarsine oxide (PAO), which we found induces activation-independent L-selectin release from neutrophils, lymphocytes, and eosinophils. PAO is an organic trivalent arsenical that cross-links vicinal thiols in the Cys-x-y-Cys sequence by forming stable dithioarsine rings (34, 35). The dithiols 2,3-dimercaptopropanol (DMP), also known as British anti-lewisite, and its membrane-impermeable sulfonic acid analogue 2,3-dimercaptopropanesulfonic acid (DMPS), known to remove PAO from its protein target(s) (36), effectively block PAO-induced L-selectin shedding. PAO affects many cell functions, including receptor internalization (37), glucose uptake (38), neutrophil NADPH oxidase (39), platelet activation (40), protein tyrosine phosphatase activity (41), and IL-1 converting enzyme-related apoptosis (42). Although most of these effects are imparted at low concentrations of PAO (<<10 µM), they may require PAO to enter the cell. Here we present evidence suggesting that PAO induces L-selectin shedding by interacting with a cell surface target and that the entrance of PAO into the cell is not required. Moreover, we propose that a likely target of PAO in this process is a membrane-resident protein disulfide isomerase (PDI) (43, 44, 45), a redox-sensitive enzyme that catalyzes oxidation-reduction reactions through an internal, vicinal dithiol-dependent, disulfide-sulfhydryl interchange.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Human venous blood was collected from healthy volunteers into sterile syringes containing heparin (10 U/ml of blood; Elkins-Simms, Cherry Hill, NJ). The blood was separated on Mono-Poly resolving medium (ICN Biochemicals, Aurora, OH) by centrifugation at 500 x g for 22 min at 12°C. The granulocyte and mononuclear (for lymphocytes) layers were collected separately and washed in HHB buffer (110 mM NaCl, 10 mM KCl, 10 mM glucose, 1 mM MgCl2, and 30 mM HEPES, pH 7.40), then pelleted at 400 x g for 10 min. The cells were resuspended in HHB buffer containing 0.1% human serum albumin (Armour, Kankakee, IL) and 1.5 mM CaCl2, at 107 cells/ml. The buffer was depleted of endotoxin by affinity chromatography over columns containing polymyxin B-Sepharose (Detoxi-gel, Pierce, Rockford, IL) and autoclaving for 1 h. All plastic ware was autoclaved for at least 45 min. Eosinophils were identified by labeling the granulocyte population with VLA-4 mAb (IgG1, PE-anti human CD49d, PharMingen, San Diego, CA) at 0.50 µg/ml, then gating on the FL2-positive population with a FACScan cytometer (Becton Dickinson, Lincoln Park, NJ). This method of identifying eosinophils was verified by flow cytometry cell sorting (Elite, Coulter, Miami, FL) of very late Ag-4 (CD49d) and L-selectin-positive granulocytes and subsequent immunohistochemical analysis of the sorted population (Ref. 46 and E. B. Lynam and L. A. Sklar, unpublished observations).
Reagents
Neutrophils were activated with fMLF (Sigma, St. Louis, MO) for 10 min at 37°C at a final concentration of 100 nM. PAO, DMP, DMPS, N-ethylmaleimide, N-acetyl-L-cysteine, glutathione, iodoacetate, nitro blue tetrazolium, iodoacetamide, mersalyl acid (o-((3-hydroxomercurio-2-methoxypropyl)carbomyl)phenoxyacetic acid), thimerosal (mercury-((O-carboxyphenyl)thio)ethyl sodium salt), 5,5'-dithio-bis-(2-nitrobenozoic acid) (DTNB), PMSF, and p-aminophenylmercuric acetate were all obtained from Sigma. Bromobimanes (monobromobimane, dibromobimane, and monobromotrimethylammoniobimane) were purchased from Molecular Probes (Eugene, OR), and 4,4'-diisothiocyanatostilbene-2,2'-disulfonic acid disodium salt were purchased from Fluka (Buchs, Switzerland). Diamide (azodicarboxylic acid bis-dimethylamide), azodicarbonamide, As2O3, CdCl2, and Sb2O3 were obtained from Aldrich (Milwaukee, WI). Stock solutions of DMPS were prepared in sterile water. PAO solutions were prepared in DMSO (Sigma) and gently heated until PAO went into solution. PAO-induced shedding was accomplished by incubating cells with PAO (at 100 nM unless otherwise specified) for 10 min at 37°C. DMP and DMPS were used at a final concentration of 50 µM unless otherwise stated. Neutrophils were incubated with these reagents for 10 min at 37°C. TAPI-2 (provided by Dr. Roy A. Black, Immunex, Seattle, WA) was prepared in DMSO and used at a final concentration of 100 µM; it was administered to the cells for 10 min at 4°C before stimulation with fMLF or addition of PAO.
PAO reversal assays
PAO reversal assays were performed by first preincubating cells with 100 nM PAO for 10 min at 4°C, then either DMP or DMPS was added. Cells were incubated for another 10 min at 4°C, followed by an additional 10-min incubation at 37°C.
Analysis of surface Ag expression
Direct immunofluorescence labeling of control and treated cells
was performed in a final volume of 200 µl at
106 cells/ml by incubating cells with mAb for
1 h at 4°C. Leu 8-FITC (IgG2a; Becton Dickinson Monoclonal
Antibodies, Lincoln Park, NJ), a fluorescent mAb that recognizes
L-selectin, was used at a final concentration of 0.625 µg/ml.
Likewise, Leu 15-PE (IgG2a; Becton Dickinson Monoclonal Antibodies), a
fluorescent mAb that recognizes the -subunit (CD11b) of Mac-1, was
used at 1.25 µg/ml. The relative expression of the receptors was
quantitated using a FACScan Flow Cytometer (Becton Dickinson).
Immunophenotyping assay
Control cells and PAO-treated cells (100 nM for 10 min at 37°C) were labeled for surface expression of several epitopes. Direct immunofluorescence labeling of cells was performed for detection of L-selectin and ß2 integrin with Abs Leu 8-FITC and Leu 15-PE. Indirect immunofluorescence was used to detect the remaining epitopes, including PDI. Cells (1 x 106) in 200 ml of HHB were incubated for 40 min at 4°C with appropriate Abs. The Abs were against CD14, CD16 (both at 10 µg/ml; Dako, Carpinteria, CA), CD43 (8 µg/ml; IgG2a; Camfolio (Becton Dickinson), San Jose, CA), CD54 (8 µg/ml; BioSource, Camarillo, CA), PSGL-1 (PL1; IgG1; 10 µg/ml; a gift from Dr. Rodger McEver, University of Oklahoma, Oklahoma City, OK). After incubation the cells were washed by centrifugation for 10 min at 400 x g at 4°C. The second Ab, goat anti-mouse IgG-FITC polyclonal Ab (Becton Dickinson Antibodies) at a concentration of 6.25 µg/ml, was added, and cells were incubated for an additional 20 min at 4°C. After a final wash, the specific labeling for each Ab was analyzed by flow cytometry. Expression of PDI on the cell surface was determined similarly. Anti-PDI mAbs (clone RL90 (IgG2a) and clone RL77 (IgG2b), both 1.5 mg/ml) were obtained from Affinity BioReagents (Golden, CO). Both were used at the final dilution of 5 µl/100 µl (105) cells. Matched isotype control Abs (Coulter, Hialeah, FL) were used to measure any nonspecific staining. The results are reported as the relative mean channel fluorescence.
PAO time-course experiments
For these experiments, isolated neutrophils, eosinophils, or lymphocytes were warmed to 37°C, and a zero point sample was withdrawn and placed on ice. PAO was then added (1 µM for neutrophils and eosinophils, and 5 µM for lymphocytes). Cell samples were withdrawn at 1-min intervals and placed on ice. Thereafter, the cells were labeled for 40 min with Leu 8-FITC for the lymphocyte preparation or with Leu 8-FITC plus anti-VLA-4-PE for the granulocyte population. This permits the simultaneous identification and quantitation of L-selectin on neutrophils and eosinophils as described above.
Soluble L-selectin ELISA
Fifty-microliter aliquots of neutrophils suspended in HEPES buffer at 106 cells/ml were treated with 100 nM PAO, 100 nM PAO followed by 50 µM DMPS, 100 µM TAPI-2 followed by 100 nM PAO, 100 nM fMLF, or 100 µM TAPI-2 followed by 100 nM fMLF, with necessary incubations as outlined above. Also included was an untreated control kept at 4°C and a DMSO-treated sample subjected to a 10-min 37°C incubation. Aliquots were then centrifuged, and the supernatants were removed and prepared according to the test protocol of the Bender MedSystems (Boehringer Ingelheim Group, Vienna, Austria) sL-selectin ELISA kit. Neutrophils were resuspended in buffer and labeled for L-selectin expression with Leu 8-FITC (as above), then analyzed by FACScan.
Neutrophil aggregation measurements
Methods for aggregation measurements have been described previously (6). Briefly, control and treated cells, in a volume of 500 µl at 4 x 106 cells/ml, were labeled with the nucleic acid stain LDS-751 (Exciton, Dayton, OH) at 0.04 µg/ml for 7 min at 37°C. Samples were equilibrated for 2 min at 37°C under conditions of shear mixing using a small bar magnet (7 x 2 mm; VWR Scientific, Media, PA) above a heated stir device at 500 rpm. Samples were then activated with 0.1 µM fMLF, and data were acquired at specific intervals after stimulation. We report the percentage of cells that formed aggregates.
Dual population aggregation of neutrophils with ICAM-1 transfectants
This method has been previously described (6). Briefly, a transfected murine melanoma cell line expressing ICAM-1, Uill/E3, was aggregated with neutrophils to test neutrophil integrin activity. For dual population aggregation experiments, Ui11/E3 cells were labeled using a membrane-linked stain, PKH2-GL (Sigma). Labeled Ui11/E3 were suspended in HHB buffer containing 1.5 mM CaCl2 and 0.1% human serum albumin, and 250 µl of labeled Uill/E3 cells at 3 x 106 cells/ml were combined with 250 µl of LDS-75 (40 ng/ml; Exciton)-labeled neutrophils at 3 x 106 cells/ml. The singlet and aggregate events were quantitated using FACScan research software. An analysis gate was placed around each specific cluster of events. We report here the percentage of neutrophils that were involved in two-color heterotypic aggregates.
Induction of L-selectin shedding by anti-PDI Abs
Neutrophils were preincubated in the absence or the presence of anti-PDI mAbs (clone RL90 (IgG2a) and clone RL77 (IgG2b), Affinity BioReagents) or matched isotype control mAbs (Coulter). Five microliters of Ab was added to 4 x 105 cells in a final volume of 200 µl, then the sample was incubated for 30 min on ice. Following a 10-min incubation at 37°C, the cells were washed with ice-cold HEPES buffer and assayed for L-selectin expression with Leu 8-FITC as described above.
Induction of L-selectin shedding with bacitracin
Bacitracin (100 mM; Sigma) or purified bacitracin A (a gift from Leo Kesner, Biology Department, State University of New York Health Sciences Center, Brooklyn, NY) stock was prepared in HHB and used to treat neutrophils at a final concentration of 3 mM. To rule out LPS contamination, some neutrophil samples were pretreated for 30 min at 4°C with 20 µg/ml of MY4 (Coulter), a mAb that blocks the LPS receptor CD14 (8). After the 30-min incubation at 37°C, the cells were placed on ice and assayed for L-selectin expression with Leu 8-FITC.
Interaction of PDI with PAO affinity resin
ThioBond (Invitrogen, San Diego, CA), an agarose-based support covalently modified with PAO, was washed twice with PBS (pH 7.2; Life Technologies, Grand Island, NY). A 500-µl aliquot placed in a 1.5-ml microfuge tube was activated with 1 ml of 20 mM 2-ME (Sigma). The tube was rocked at room temperature for 60 min. The resin was allowed to settle by gravity, and the supernatant was decanted. The resin was washed three times with PBS. PDI (20 µg; Calbiochem, San Diego, CA) was solubilized in 500 µl of PBS and added to the activated ThioBond. The sample was rocked for 90 min at room temperature. The resin was gravity settled, and the PDI solution was decanted and saved. Five subsequent washes were performed. To elute the bound protein, 500 µl of 0.5 M ß-ME was added, and the sample was rocked at room temperature for 30 min. The eluate was then collected. Twenty-microliter fractions from the flow-through volume, each wash, elution, and a PDI control were assayed by SDS-PAGE. A control was generated by incubating 500 µl of ThioBond in 1 ml of a 10 mM DTT (Sigma) solution. DTT irreversibly inactivates the resin. PDI (20 µg) was then added to the resin, followed by the washing and elution procedure described above.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
We have previously reported on the activation of integrin function
by sulfhydryl reactive agents (47, 48). During the course
of these investigations we observed that thiol-reactive agents also
regulated L-selectin shedding. In general, the oxidizing and
thiol-blocking reagents promote shedding (Table I
). These include membrane-permeable
thiol-reactive iodoacetate, monobromobimane, dibromobimane,
4-aminophenylmercuric acetate, and N-ethylmaleimide.
The membrane-impermeable reagents include DTNB, mersalyl acid,
thimerosal (sodium ethylmercurithiosalacylate),
4,4'-diisothiocyanatostilbene-2,2'-disulfonic acid, and quaternary BBr.
Their monothiol reactivity, rather than membrane impermeability, is
likely to account for their uniformly low effectiveness. Nitro blue
tetrazolium is a superoxide scavenger and, like hydrogen peroxide, a
potent oxidant. Azodicarbonamide is a structural analogue of diamide
that is well known for its ability to cross-link thiols
(49), while PMSF is reactive with thiol nucleophils. In
addition, diagnostic inhibitors of enzymes with active site dithiol
groups, such as arsenite and Cd2+ (50, 51), also induce L-selectin shedding (not shown). Trivalent
arsenite (As2O3) and
antimony (Sb2O3) are the
most potent (inducing full shedding in 10 min at about 50 µM), while
the divalent cadmium (CdCl2) requires higher
concentrations (1 mM). In contrast, high concentrations (5 mM) of the
dithiol reducing agents were shown not to induce L-selectin shedding,
but, rather, to block shedding when neutrophils were activated
with formyl peptide (Fig. 1). Monothiol
reducing reagents do not affect the rate of fMLF-induced shedding. We
have shown previously that activation of cell adhesion occurs normally
(47).
|
|
a). A 10-min incubation with
100 nM PAO reduces L-selectin levels to that comparable with a 5-min
treatment with 100 nM fMLF. Activation-dependent down-regulation of
L-selectin is tightly coordinated with an increase in the surface
expression of the Mac-1 (CD11b/CD18) ß2
integrin (6). Marginal up-regulation of Mac-1 in response to PAO (Fig. 2b), much less than that induced by activation with fMLF,
indicates that PAO induces L-selectin shedding without activating the
neutrophils. Similarly, PSGL-1 (a ligand for P- and L-selectin), CD14
(LPS receptor), CD16 (Fc
receptor), CD43 (sialophorin; a major
sialoglycoprotein shown to interact with ICAM-1), and CD54 (ICAM-1)
were not affected by 100 nM PAO (Fig. 2b).
|
c), with 50% shedding at 20 min with 1 µM
PAO. The requirement for an increased concentration of PAO in whole
blood compared with the purified neutrophil populations probably
reflects an increased number of PAO-reactive sites present in whole
blood. Reversibility of PAO binding
The dithiol, heavy metal chelating compound DMP was developed as
an antidote for the arsenical war gas and has been extensively used for
treatment of arsenical or mercury poisoning (52, 53). DMP
(Fig. 3) is able to reverse the binding
of PAO to its target (36). It is thought that DMP competes
for PAO on the PAO-protein complex by reducing the vicinal sulfhydryls,
stripping PAO from its target protein(s), and forming a stable, soluble
chelate (52) (Fig. 3).
|
a). DMP at this
concentration was not able to inhibit fMLF-induced shedding of
L-selectin, indicating a selective reversal of PAO effects by DMP
rather than an inhibition of fMLF signaling or the proteolytic cleavage
of L-selectin.
|
) is a membrane-impermeable analogue of DMP that has
previously been used to define a site of PAO action with respect to the
surface of the plasma membrane (54). Lower concentrations
(50 µM) of DMPS, like the membrane-permeable DMP, did not interfere
with the fMLF-induced shedding. DMPS was, however, still able to
inhibit the PAO-induced shedding of L-selectin from neutrophils (Fig. 4a). This suggests that the mechanism by which PAO causes
L-selectin to be shed is an extracellular process and is not due to
intracellular signaling. Several other thiol reagents noted in Table I,
reported to be membrane impermeable, induced L-selectin shedding. In
contrast, most other previously reported PAO-induced events in
neutrophils appear to be related to intracellular actions (39, 55, 56). Neither DMP nor DMPS had adverse effects on Mac-1 (Fig. 4b). These compounds did not up-regulate the integrin on
their own and did not interfere with fMLF-induced up-regulation. As a
result, we conclude that it is unlikely that either DMP or DMPS
interacts with the pathway that regulates L-selectin shedding. We
envision that at low concentrations DMP and DMPS prevent the
PAO-induced shedding of L-selectin by abstracting PAO from its distinct
binding site(s) on the cell surface. Rate of L-selectin cleavage in neutrophils, eosinophils, and lymphocytes
To determine whether PAO had a similar effect on L-selectin levels
in other leukocytes, we examined L-selectin expression over time in
lymphocytes and eosinophils along with neutrophils (Fig. 5). Eosinophils initially expressed lower
levels of L-selectin and shed this molecule in response to PAO more
slowly than neutrophils within the same granulocyte population.
Similarly, lymphocytes also showed a lower basal level of L-selectin
expression and, even at increased concentrations of PAO, a considerably
slower rate of PAO-induced L-selectin shedding compared with
neutrophils. PAO (5 µM), however, induced complete L-selectin release
from lymphocytes in 30 min. When the incubation was conducted in the
absence of PAO, none of these three leukocyte types experienced
significant spontaneous L-selectin shedding (Fig. 5). These results
show that although expressing different initial levels of L-selectin
and responding with different rates of shedding, PAO induces shedding
in neutrophils, eosinophils, and lymphocytes.
|
To verify the functional integrity of neutrophils treated with
PAO, we examined the ability of neutrophils to aggregate with one
another. We have previously shown that homotypic aggregation, which
occurs when neutrophils are exposed to fMLF or leukotriene B4 under
shear stress (6, 7), involves two sequential steps that
are analogous to leukocyte-endothelial cell adhesion. The first step is
a low affinity interaction between neutrophil L-selectin and its mucin
counterstructure PSGL-1 on the opposing neutrophil (7).
The second step is a high affinity adhesion between a
ß2 integrin (CD18) and its neutrophil ligand,
most likely ICAM-3 (18, 57). Accordingly, the aggregation
of PAO-treated neutrophils was inhibited (Fig. 6a) due to the loss of
L-selectin. Although DMPS alone at low micromolar concentrations did
not have an adverse effect on aggregation, DMPS was able to rescue the
ability of the cells to aggregate, presumably by removing the PAO from
the critical target protein. This allowed L-selectin to remain on the
surface and to initiate the aggregation process upon stimulation
with fMLF.
|
To verify that the absence of aggregation in PAO-treated
neutrophils was due to the loss of L-selectin and was not the result of
PAO interfering with intracellular signaling or integrin activation, a
murine melanocyte cell line transfected with ICAM-1 was used to assess
ß2 integrin function (Fig. 6b). The
adherence of neutrophils with the ICAM-1-transfected cells is dependent
solely on the integrin step (58). Although treatment with
100 nM PAO inhibited homotypic neutrophil aggregation (Fig. 6a), it did not inhibit the adhesion of neutrophils to
target cells (Fig. 6b). This demonstrated that at this
concentration PAO treatment does not interfere with the signaling
pathways that lead to an increase in the adhesive competence of
neutrophil integrins.
Inhibition of PAO-induced shedding of L-selectin by TAPI-2
TAPI-2, a hydroxamate-based inhibitor of matrix metalloproteases,
has previously been shown to inhibit the activation-induced shedding of
L-selectin from neutrophils, eosinophils, and lymphocytes (18, 29, 30, 31). Here we show that TAPI-2 also inhibits PAO-induced
shedding in neutrophils (Fig. 7). Thus,
TAPI-2 appears to be able to inhibit the activation-independent release
of L-selectin as well as the activation-dependent release.
|
To confirm that the L-selectin analysis by flow cytometry
represented shedding from the neutrophil surface, an sL-selectin ELISA
was performed on cell supernatants. Neutrophils were treated with 100
nM PAO, 100 nM PAO followed by 50 µM DMPS, 100 µM TAPI-2 followed
by 100 nM PAO, 100 nM fMLF, or 100 µM TAPI-2 followed by 100 nM fMLF.
Control samples and samples containing TAPI-2 displayed negligible
levels of sL-selectin (<<0.20 ng/ml). The PAO/DMPS samples registered
slightly higher reading at 
0.4 ng/ml, while the PAO and fMLF samples
displayed sL-selectin levels in the 1.6–2.0 ng/ml range. This
indicates that PAO, like fMLF treatment, results in the release of the
L-selectin molecule into the medium. As further conformation, the cells
from which the supernatants were taken were subsequently labeled for
L-selectin surface expression. The results mimicked those shown in
Figs. 2a, 4a, and 7, in which control, DMPS, and
TAPI-2 samples maintained near normal levels of cell surface L-selectin
expression, while PAO- and fMLF-treated cells displayed minimal levels
of L-selectin.
A hypothesis was developed in which extracellular PAO regulates the
susceptibility of the L-selectin molecule to a constitutively active,
TAPI-2-inhibitable, protease. We postulated that PDI
(43, 44, 45), known for its ability to rearrange disulfide
bonds within a variety of substrate proteins, could promote an
interchange between its thiols and the disulfide bonds of the 24
cysteine residues of L-selectin. To explore this, we first determined
that PDI is indeed expressed on the neutrophil cell surface (Fig. 8a). Additionally, two
anti-PDI monoclonals, both previously reported to inhibit PDI
activity (59, 60), were found to induce L-selectin
shedding (Fig. 8b). We further verified the involvement of
PDI using another known inhibitor of PDI activity, bacitracin (Fig. 8c). This antibiotic inhibits PDI, but not thioredoxin, the
other enzyme also present at the cell surface and known to catalyze
oxido-reduction reactions (60). LPS receptor CD14-blocking
Abs (My4) were used to verify that LPS contamination of bacitracin was
not involved in the induction of L-selectin shedding from these
bacitracin-treated neutrophils. Moreover, bacitracin-induced shedding
was not due to neutrophil activation, which would result in the
characteristic quantitative up-regulation of the cell surface
ß2 integrins (data not shown). These results
were confirmed using bacitracin further purified by Dr. Kesner (State
University of New York). Lymphocytes also express PDI on the cell
surface (61) (as confirmed by us) and respond to
anti-PDI Abs by shedding L-selectin, albeit much more slowly than
neutrophils. Lymphocytes only respond to prolonged (>30-min)
bacitracin treatment (data not shown). Finally, we have obtained the
first direct evidence for the interaction of PAO and PDI. ThioBond
resin (Invitrogen), an agarose-based support covalently attached to
PAO, specifically binds purified PDI (Fig. 8d). ThioBond did
not bind BSA, and resin inactivated by treatment with DTT was not able
to bind PDI.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The predominant mechanism for regulating L-selectin-mediated
adhesion is its proteolytic shedding from cell surfaces during an
immunological or inflammatory response (15). Several
laboratories have concluded that the L-selectin sheddase is a
constitutively active protease, most likely identical with TNF-
converting enzyme (28), and that it is the conformational
status of L-selectin molecule that determines the susceptibility of the
L-selectin molecule to the proteolytic cleavage (26, 27, 32). It has been postulated that ligand binding or cellular
activation induces the protease-susceptible conformation in the
membrane-proximal region of L-selectin. Cellular activation has also
been reported to induce a transient conformational change in L-selectin
molecules that correlates with an increase in L-selectin avidity for
its ligand PPME (polyphosphomannan monoester core) (62).
Thus, ligand and cell activation-induced changes in L-selectin
conformation may be a mechanism that ensures that the rapid increase in
L-selectin receptor avidity is efficiently down-modulated through its
subsequent proteolytic release (62).
PAO and sulfhydryl regulation
We have followed a lead that suggests regulation of shedding
through extracellular sulfhydryl chemistry (Table I and Fig. 1).
Moreover, purified neutrophils and eosinophils respond rapidly to low
micromolar doses of PAO without compromising the signaling and adhesive
functions of other adhesion molecules such as Mac-1 (Figs. 2, 4, and 6). Lymphocytes also respond to PAO by shedding their L-selectin,
although they require somewhat higher concentrations and longer
incubation times (Fig. 5). Homotypic aggregation, here used as a model
for the L-selectin-dependent adhesive processes and known to play a
physiologic role in inflammatory amplification, is inhibited by PAO
(Fig. 6). Other L-selectin-dependent interactions, such as the
recruitment of leukocytes to the inflammatory sites and lymphocyte
recirculation through the lymph nodes, are expected to be profoundly
affected by the PAO-induced L-selectin loss.
Because the membrane-impermeable PAO-reversing reagent DMPS blocks
PAO-induced L-selectin release, the critical PAO target protein is
likely to reside on the outside of the plasma membrane (Fig. 4). This
extracellular location of the L-selectin shedding regulatory protein is
substantiated by the analogous, although far less potent, effect of the
membrane-impermeable monothiol-reactive reagents. The higher
specificity of the dithiol-reactive PAO combined with its extracellular
site of action provide an opportunity to cause L-selectin shedding with
membrane-impermeable analogues of PAO. Restricting PAO access to the
cell surface promises to minimize toxicities associated with
intracellular PAO.
Mechanism of PAO action
Although we have not yet formally excluded the idea that PAO
interacts directly with L-selectin, we have obtained evidence for a
regulatory molecule such as PDI (Fig. 8). Inhibition of PDI activity by
DTNB, anti-PDI Abs, and bacitracin lead to L-selectin shedding. If
cell surface PDI could act as a regulatory protein that retains
L-selectin in a noncleavable conformation, then the inhibition of its
oxido-reductive capacity by PAO, through interaction with a
thioredoxin-like active site Cys-Gly-His-Cys, is expected. Similarly,
mono-thiol-reactive N-ethylmaleimide and DTNB, both inducers
of L-selectin shedding, are routinely used to block PDI activity,
albeit at high concentrations (63). Although PAO has not
yet been reported to block PDI active sites, it is known to cross-link
the homologous active site in thioredoxin (64). PDI is a
subunit of the tri-iodothyronine receptor (65), and the
recombinant rat tri-iodothyronine receptor has been shown to bind
specifically to a PAO affinity column (66). Additionally,
we have shown that purified PDI specifically binds to a PAO affinity
column, providing further evidence for an interaction between PDI
and PAO.
One remaining speculation is that L-selectin is a substrate for PDI.
PDI is known to catalyze disulfide bond interchange in a spectrum of
substrate proteins, and this isomerase function depends on the
integrity of the vicinal-dithiol active sites. PDI is also a chaperone
whose function does not depend on its isomerase activity
(43, 44, 45). The chaperone activity is thought to be due to
PDI binding to proteins that have a tendency to aggregate in the
denatured state by promoting the correct folding of the protein. PDI is
a critical component of protein complexes such as the -subunit of
prolyl-hydroxylase, N-glycosyl transferase, and the
triglyceride transfer protein complex, where it is required to maintain
triglyceride transfer protein in catalytically active form and to
prevent its aggregation (67).
Despite its Lys-Asp-Glu-Leu endoplasmic reticulum retention signal, PDI has been detected on the surface of many cell types, including hepatocytes, platelets, and lymphocytes (60, 61, 68, 69), and is implicated in many cell surface processes. A PDI homologue, cognin, plays a role the adhesion-dependent aggregation of retinal cells (70). PDI modulates the conformation of thrombospondin where the isomerization of disulfide bonds is likely to have a profound effect on its ligand binding and adhesive capacity (71). Couet et al. have demonstrated that cell surface PDI is involved in the shedding of human thyrotropin receptor ectodomain (60). By analogy, our results implicate PDI in the release of L-selectin. One crucial difference is that it is the inhibition of neutrophil PDI that permits L-selectin shedding. Moreover, L-selectin shedding can be induced by treatment with the same PDI-blocking reagents (DTNB, bacitracin, and anti-PDI Abs) that impede the release of the thyrotropin receptor. It remains to be determined whether this PDI-mediated mechanism operates independently or is a part of the calmodulin-controlled, L-selectin shedding pathway (33).
Neutrophils contain appreciable immunoreactive PDI in their specific granules, and degranulation with PMA treatment releases PDI into the medium (72). We now show that PDI is present on the resting neutrophil cell surface. Neither the oxidative status of the PDI Cyc-x-y-Cys active sites, the impact of cell activation, nor the role of the released PDI in shedding of L-selectin is known.
Because L-selectin initiates the interaction of leukocytes with activated endothelium, L-selectin appears to play a pivotal role in inflammatory disease. These include acute respiratory distress syndrome, ischemia-reperfusion injury that follows myocardial infarction and stroke, the pathogenesis of multiorgan failure that follows sepsis, and the diseases of eosinophilic inflammation, such as asthma and dermatitis. Cell surface PDI-blocking agents, such as impermeable PAO analogues that interfere with both leukocyte adhesion to the endothelium and leukocyte-leukocyte interactions, could play an important role in limiting or perhaps preventing damage in acute as well as chronic inflammatory diseases. Furthermore, HIV-induced CD4+ lymphocyte depletion may be due to L-selectin signaling (14). Abrogation of this signaling by shedding L-selectin might mitigate the decrease in CD4+ cell count that is typical of late stages of AIDS pathogenesis.
In summary, we present our model, illustrated in Fig. 9. This model proposes a novel mechanism
by which the cellular chaperone and oxido reductase, PDI, regulates the
susceptibility of leukocyte L-selectin to shedding. Cell surface PDI
constitutively acts upon L-selectin to maintain disulfide bonds in the
reduced, noncleavable state. Blockade of PDI function permits reversion
of L-selectin to the oxidized, cleavable conformation. In the presence
of the sheddase inhibitor, TAPI-2, L-selectin is retained on the cell
surface. We speculate that following physiological activation,
L-selectin conformation is modulated by oxidation of its critical
sulfhydryls, rendering it sensitive to proteolysis. Our preliminary
data suggest that reactive oxygen and/or nitrogen species, released by
activated cells, might modulate PDI activity under physiologic
conditions. This PDI-mediated mechanism, analogous to that which
mediates chloroplast translational activation (73), offers
a simple, extracellular switch for the regulation of L-selectin
shedding.
|
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Snezna Rogelj, Jones Annex Room 315, Department of Biology, New Mexico Institute of Mining and Technology, Socorro, NM 87801.
3 Abbreviations used in this paper: TAPI-2, (N-{D,L-[2-(hydroxyaminocarbonyl)-methyl]-4-methylpentanoyl}-L-3-(tert-butyl)-alanyl-L-alanine, 2-aminoethyl amide; PAO, phenylarsine oxide; PDI, protein disulfide isomerase; DMP, 2,3-dimercaptopropranol; DMPS, 2,3-dimercaptopropanesulfonic acid; DTNB, 5,5'-dithio-bis-(2-nitrobenozoic acid); PSGL-l, P-selectin glycoprotein ligand-1; SL-selectin; soluble L-selectin.
Received for publication October 29, 1999. Accepted for publication February 1, 2000.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
receptor. Exp. Cell Res. 211:150.[Medline]
This article has been cited by other articles:
|
|
 |
|
  Y. Wang, A. H. Herrera, Y. Li, K. K. Belani, and B. Walcheck Regulation of Mature ADAM17 by Redox Agents for L-Selectin Shedding J. Immunol., February 15, 2009; 182(4): 2449 - 2457. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S.-B. Jeon, H. J. Yoon, S.-H. Park, I.-H. Kim, and E. J. Park Sulfatide, A Major Lipid Component of Myelin Sheath, Activates Inflammatory Responses As an Endogenous Stimulator in Brain-Resident Immune Cells J. Immunol., December 1, 2008; 181(11): 8077 - 8087. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Stefansson, E. J. Su, S. Ishigami, J. M. Cale, Y. Gao, N. Gorlatova, and D. A. Lawrence The Contributions of Integrin Affinity and Integrin-Cytoskeletal Engagement in Endothelial and Smooth Muscle Cell Adhesion to Vitronectin J. Biol. Chem., May 25, 2007; 282(21): 15679 - 15689. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Lee, J. B. Schultz, P. A. Knauf, and M. R. King Mechanical Shedding of L-selectin from the Neutrophil Surface during Rolling on Sialyl Lewis x under Flow J. Biol. Chem., February 16, 2007; 282(7): 4812 - 4820. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Hiroi, K. Okada, S. Imaoka, M. Osada, and Y. Funae Bisphenol A Binds to Protein Disulfide Isomerase and Inhibits Its Enzymatic and Hormone-Binding Activities Endocrinology, June 1, 2006; 147(6): 2773 - 2780. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Janiszewski, L. R. Lopes, A. O. Carmo, M. A. Pedro, R. P. Brandes, C. X. C. Santos, and F. R. M. Laurindo Regulation of NAD(P)H Oxidase by Associated Protein Disulfide Isomerase in Vascular Smooth Muscle Cells J. Biol. Chem., December 9, 2005; 280(49): 40813 - 40819. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Levitan, T. Trebitsh, V. Kiss, Y. Pereg, I. Dangoor, and A. Danon Dual targeting of the protein disulfide isomerase RB60 to the chloroplast and the endoplasmic reticulum PNAS, April 26, 2005; 102(17): 6225 - 6230. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Rabie, A. Strehl, A. Ludwig, and B. Nieswandt Evidence for a Role of ADAM17 (TACE) in the Regulation of Platelet Glycoprotein V J. Biol. Chem., April 15, 2005; 280(15): 14462 - 14468. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Chigaev, G. J. Zwartz, T. Buranda, B. S. Edwards, E. R. Prossnitz, and L. A. Sklar Conformational Regulation of {alpha}4{beta}1-Integrin Affinity by Reducing Agents: "INSIDE-OUT" SIGNALING IS INDEPENDENT OF AND ADDITIVE TO REDUCTION-REGULATED INTEGRIN ACTIVATION J. Biol. Chem., July 30, 2004; 279(31): 32435 - 32443. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Benkerrou, C. Delarche, L. Brahimi, M. Fay, E. Vilmer, J. Elion, M.-A. Gougerot-Pocidalo, and C. Elbim Hydroxyurea corrects the dysregulated L-selectin expression and increased H2O2 production of polymorphonuclear neutrophils from patients with sickle cell anemia Blood, April 1, 2002; 99(7): 2297 - 2303. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. P. Fors, K. Goodarzi, and U. H. von Andrian L-Selectin Shedding Is Independent of Its Subsurface Structures and Topographic Distribution J. Immunol., October 1, 2001; 167(7): 3642 - 3651. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. J. Strausbaugh and S. D. Rosen A Potential Role for Annexin 1 as a Physiologic Mediator of Glucocorticoid-Induced L-Selectin Shedding from Myeloid Cells J. Immunol., May 15, 2001; 166(10): 6294 - 6300. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Trebitsh, E. Meiri, O. Ostersetzer, Z. Adam, and A. Danon The Protein Disulfide Isomerase-like RB60 Is Partitioned between Stroma and Thylakoids in Chlamydomonas reinhardtii Chloroplasts J. Biol. Chem., February 9, 2001; 276(7): 4564 - 4569. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L.-c. Zhao, M. Shey, M. Farnsworth, and M. O. Dailey Regulation of Membrane Metalloproteolytic Cleavage of L-selectin (CD62L) by the Epidermal Growth Factor Domain J. Biol. Chem., August 10, 2001; 276(33): 30631 - 30640. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
