|Home | About | Journals | Submit | Contact Us | Français|
L-selectin is an adhesion molecule expressed by neutrophils that broadly directs their infiltration in to sites of inflammation, and is also present at relatively high levels in the serum of normal individuals. It is well established that L-selectin is efficiently shed from the surface of neutrophils upon their activation, a process that regulates its density and binding activity. Neutrophil programmed cell death is critical for the resolution of inflammation, and L-selectin down-regulation is induced during this process as well. The mechanisms underpinning this latter process are much less understood, and were investigated here. Using ADAM17 radiation chimeric mice, we demonstrate for the first time that during early events of death receptor-mediated neutrophil apoptosis, L-selectin down-regulation occurs primarily by ADAM17-mediated shedding. This was observed as well upon using shRNA to knock-down ADAM17 expression in Jurkat cells, a well-studied cell line in terms of the molecular processes involved in the induction of apoptosis. These findings directly reveal that ADAM17 activity occurs during programmed cell death. Hence, the cleavage of particular ADAM17 substrates may be an additional component of the anti-inflammatory program initiated by apoptotic neutrophils. Of interest was that during later stages of induced leukocyte apoptosis, soluble L-selectin production occurred independent of ADAM17, as well as membrane events such as blebbing and microparticle production. This process may provide an explanation for the lack of diminished serum L-selectin levels in ADAM17-null mice, and suggests a mechanism for the homeostatic maintenance of soluble L-selectin levels in the blood of healthy individuals.
Ectodomain shedding is a regulated proteolytic process that directs the cleavage of cell surface proteins, typically at a juxta-membrane site, resulting in the release of a soluble extracellular domain fragment (1–3). The functional implications of ectodomain shedding are diverse, as it can promote an immediate and prolonged formation of soluble agonists and antagonists, as well as regulate the density of receptors and adhesion molecules. The majority of a growing list of shed proteins are cleaved by ADAM173, a disintegrin and metalloprotease (4). Leukocytes express ADAM17 and a number of their products undergo ectodomain shedding (5). By directly examining leukocytes deficient in functional ADAM17, it has been reported that this sheddase cleaves various inflammation regulating factors, including the pleiotropic cytokine TNFα, its two receptors TNFRI (CD120a) and TNFRII (CD120b), and L-selectin (CD62L) (6–11).
ADAM17’s enzymatic activity is inducible and its function has been primarily studied in the context of cell activation (12–14). The leukocyte-expressed adhesion molecule L-selectin is perhaps one of the best characterized shed molecules in terms of structure/function analyses and its targeting by ADAM17 (8, 14–18), and thus is an excellent model protein to study ADAM17 activity in leukocytes. L-selectin is constitutively expressed at high levels by resting neutrophils, and essentially all molecules are cleaved within minutes of neutrophil activation by a variety of physiological stimuli (19).
In addition to their pro-inflammatory role during activation, neutrophils mediate an anti-inflammatory program upon programmed cell death (20, 21), which includes an efficient down-regulation of cell surface L-selectin expression (22, 23). In our study, we directly exam ADAM17’s role of in this process by using leukocytes that lack expression of the sheddase. Our findings show that during early events of induced neutrophil apoptosis, L-selectin down-regulation occurs primarily by ADAM17-mediated ectodomain shedding, indicating ADAM17 activity during the anti-inflammatory program. We also report that at later time points of induced neutrophil apoptosis, soluble L-selectin production can occur in a manner independent of ADAM17. This process may be important for regulating the high levels of soluble L-selectin in the blood of healthy individuals, a potential anti-adhesive mechanism.
The anti-human L-selectin mAb DREG-200 has been previously described (24). The anti-L-selectin mAb LAM1-116 (detects both murine and human L-selectin) conjugated to PE, biotin, or allophycocyanin (APC), and recombinant human Fas ligand were purchased from Ancell (Bayport, MN). PE-conjugated anti-mouse LFA-1 (CD11a/CD18) mAb, PE- and APC-conjugated anti-mouse L-selectin (Ly-22) mAb, and Alexa Fluor-488 anti-mouse Ly-6G (Gr-1) were purchased from eBioscience (San Diego, CA). The anti-human LFA-1 (CD11a/CD18) mAb R3.1 has been previously described (25). The anti-Fas IgM mAb CH-11 was purchased from Upstate (Lake Placid, NY). The anti-ADAM10 and ADAM17 mAbs (clones 163003 and 111623, respectively) and TNFα were purchased from R&D Systems (Minneapolis, MN). Biotin, FITC-, PE-, and APC-conjugated F(ab')2 goat anti-mouse IgG secondary antibodies as well as APC-conjugated streptavidin were purchased from Jackson Immunoresearch (West Grove, PA). DMEM, RPMI, PBS, HBSS, Hepes, and molecular grade water were purchased from Mediatech (Hevdon, VA). FBS was purchased from Atlas Biologicals (Fort Collins, CO). Penicillin-Streptomycin solution was purchased from Cellgro (Manassas, VA). PMA was purchased from Sigma-Aldrich (St. Louis, MO). TAPI-1 was purchased from Peptides International (Louisville, KY). Leupeptin, 6-aminocaproic acid, APMSF-HCl, aprotinin, pepstatin A, and chymostatin were all purchased from Sigma. For cell apoptosis detection, a FLICA Poly Caspase kit (FAM-VAD-FMK) was purchased from ImmunoChemistry Technologies (Bloomington, MN) and FITC and PE-conjugated annexin-V-was purchased from BD Pharmingen (San Diego, CA).
All experimental procedures were performed in accordance with protocols approved by the Animal Care and Use Committee of the University of Minnesota. ADAM17 deficiency in mice is lethal between embryonic day 17.5 or soon after birth (7), and therefore radiation chimeric mice were generated by reconstitution with embryonic day 15–16, liver hematopoietic cells that either lacked or expressed functional ADAM17, as previously described (8, 9). These mice are referred to as either ADAM17 chimeric or wild-type chimeric mice, respectively. ADAM8 knock-out mice were generously provided by Dr. Andy J.P. Docherty (Celltech R&D, Slough, United Kingdom) via Dr. Carl P. Blobel (Hospital for Special Surgery, Weill Medical College of Cornell University) (26). Age-matched mice were used for all experiments.
All reagents used for cell isolation and incubations were sterile and tested for endotoxin. Cell viabilities were assessed by exclusion of the vital dye trypan blue. Mouse bone marrow neutrophils were isolated as previously described (8, 9). For neutrophil treatments, cells were suspended in PBS at the indicated concentrations and initially equilibrated to culture conditions at 37°C in 5% CO2. Some cell samples were incubated with the broad-spectrum zinc metalloprotease inhibitor TAPI (50 µg/ml) for 30 min, and then were treated with TNFα (50 ng/ml) with or without cycloheximide (35 µM) for the indicated time points. Jurkat cells (an acute T cell leukemia cell line) were maintained in DMEM/F12 supplemented with 8% heat-inactivated FBS and antibiotics at 37°C in 5% CO2. For Jurkat cell treatments, cells were suspended in RPMI at the indicated concentrations and initially equilibrated to culture conditions at 37°C in 5% CO2. Some cell samples were incubated with TAPI (50 µg/ml) for 30 min, and then treated with PMA (25 ng/ml), recombinant human FasL, or anti-Fas antibody CH-11 (both of these reagents were used at a final concentration of 0.5 µg/ml) for the indicated time points. In additional experiments, Jurkat cells were incubated with TAPI and/or various broad-spectrum inhibitors of serine, cysteine, or aspartate proteases (leupeptin, 100 µM; 6-aminocaproic acid, 10mM; APMSF-HCl, 50 µM; aprotinin, 20 µg/ml; pepstatin A, 3 µM; chymostatin, 20 µM) for 30 min, and then treated with anti-Fas antibody CH-11 for the indicated time points. The working concentrations of these inhibitors were based on those frequently used in the literature (http://sciencegateway.org/resources/protease.htm).
For expression of shRNA in Jurkat cells, the human ADAM17 sequence 5’-GCTTGATTCTTTGCTCTCA-3’, which was determined to knock-down ADAM17 expression, or the ADAM17 sequence 5’-CAGCAGCTGGAGTCCTGTG-3’, which did not knock-down ADAM17 expression and was used as a control to assess off-target effects, were followed by a 9 nucleotide non-complementary spacer (TTCAAGAGA). With their respective reverse complement sequence, the double-stranded DNA constructs were inserted into the pLL3.7 lentiviral vector (Addgene Inc., Cambridge, MA), per the manufacture’s instructions, after vector digestion with XhoI and HpaI (New England Biolabs, Ipswich, MA). Lentivirus generation and transduction were performed as previously described (14, 17).
For cell staining, Fc receptor and nonspecific antibody binding sites were blocked and cells were stained with particular mAbs, as previously described (8, 27). Briefly, mouse bone marrow leukocytes were triple-stained with anti-Gr-1, anti-mouse L-selectin or anti-LFA-1, and annexin-V. For cell staining that involved annexin-V (for labeling of externalized phosphatidylinositol) or FLICA (labeling of active caspases), non-fixed cells were used, and following their staining with antibodies, the cells were then treated with either reagent (per the respective manufacture’s instructions) and immediately examined by flow cytometry, as previously described (23). ADAM17 and ADAM10 are typically expressed at low levels and so their staining on Jurkat cells was amplified by treating cells with an anti-ADAM17 or anti-ADAM10 mAb, biotin-conjugated F(ab’)2 goat anti-mouse IgG, and then APC-conjugated streptavidin. Isotype-matched negative control mAbs were used to evaluate levels of nonspecific staining. All stained cells were analyzed on a FACSCanto instrument (Becton Dickinson, San Jose, CA).
Soluble murine or human L-selectin in tissue culture media supernatants were quantified by ELISA, as previously described (8, 17). In some cases, media supernatants were filtered (0.22 µm pore size) then centrifuged at 100,000 × g prior to their analysis by ELISA.
Human neutrophils upon the induction of apoptosis down-regulate their surface L-selectin in a manner involving metalloproteases (23). L-selectin is shed by ADAM17 following neutrophil activation (8), yet recent studies have revealed that other sheddases can also cleave L-selectin (28–30). To directly establish the role of ADAM17 in L-selectin shedding during death receptor-induced apoptosis, we examined murine neutrophils lacking functional ADAM17. Because important differences can occur between human and murine neutrophils (31), we initially established the effects of induced apoptosis by murine neutrophils on their expression of L-selectin. Our primary means of inducing apoptosis by isolated human neutrophils has been through the Fas receptor (23), due to its well established signaling cascades and its direct manner of caspase activation (32). However, our Fas-binding reagents for human leukocytes (see below) demonstrated little activity with murine neutrophils (data not shown). Instead, TNFα in the presence of cycloheximide was used, which is a reproducible method for inducing neutrophil apoptosis (33–36). Murine TNFα was initially tittered down to a concentration that induced nominal L-selectin shedding via neutrophil activation during the timeframe of our assays. We observed that our source of TNFα at 50 ng/ml had little effect on phosphatidylserine exposure and L-selectin shedding by murine neutrophils, whereas when combined with cycloheximide, both events were markedly enhanced (Fig. 1). L-selectin surface expression was down-regulated in a relatively rapid manner and typically occurred before appreciable reactivity by annexin-V with the neutrophils, as determined by flow cytometry. In Figures 1 and and2,2, the two events were evaluated at different time points. The down-regulation in L-selectin surface expression and production of its soluble extracellular fragment were consistently observed in our assays by 90 min after the induction of apoptosis (Fig. 1), and at this time point, soluble L-selectin production was essentially abolished by broad-spectrum zinc metalloprotease inhibitors, such as TAPI (Fig. 1). Such inhibitors, however, did not prevent the cells from undergoing apoptosis, as previously described for human neutrophils (23).
Targeting the Adam17 gene in all cells of mice is lethal and most die soon after birth (7). However, radiation chimeric mice reconstituted with hematopoietic cells lacking functional ADAM17 (ADAM17 chimeras) are viable (8, 9). An assessment of neutrophils from wild-type chimeras (reconstituted with wild-type hematopoietic cells) revealed that they also shed their L-selectin upon the induction of apoptosis by TNFα/cycloheximide, as determined by flow cytometry and ELISA (Fig. 2A). In contrast to these cells, the shedding of L-selectin by neutrophils from ADAM17 chimeras was quite minimal at early time points of induced apoptosis (Fig. 2A). Neutrophils from both chimeras demonstrated similar levels of phophatidylserine exposure after their treatment with TNFα/cycloheximide (Fig. 2A), indicating that a lack of functional ADAM17 did not prevent the apoptotic process.
ADAM8 has also been implicated in L-selectin shedding by activated neutrophils (29), and thus we examined its contribution in the shedding of L-selectin by apoptotic neutrophils as well. This was also done in a direct manner by using neutrophils obtained from ADAM8 knock-out mice. In contrast to ADAM17-deficient neutrophils, however, we observed that ADAM8-deficient neutrophils efficiently down-regulated their surface L-selectin upon their treatment with TNFα/cycloheximide (Fig. 2B).
We also directly examined the role of ADAM17 in L-selectin shedding by human leukocytes. This was done by using shRNA specific for ADAM17 to knock-down its expression. Targeting ADAM17 in human neutrophils is problematic considering that they are short-lived and that their manipulation by transfection or transduction consistently induced some level of activation, which may confound our results. Instead, Jurkat cells, a human T cell line, were used. The advantages of these cells are that they express high levels of endogenous L-selectin, unlike most human myeloid cell lines, and that Jurkat cells are one of the best studied cell types with regard to Fas-induced apoptosis (37). We found that the engagement of their Fas receptor led to efficient L-selectin shedding, which was performed by using either soluble recombinant human FasL (rhFasL) or the anti-Fas IgM mAb CH-11 (supplemental Fig. 1A–C). As with neutrophils, the induction of L-selectin shedding by Jurkat cells occurred as an early apoptotic event prior to phosphatidylserine exposure (supplemental Fig. 1A). The treatment of Jurkat cells with rhFasL consistently induced L-selectin shedding and phosphatidylserine exposure more rapidly than CH-11 treatment (data not shown), which may have to do with different manners of Fas engagement.
Jurkat cells were transduced with a bicistronic lentiviral vector that expressed GFP and ADAM17 shRNA in a proportional manner. Greater than 50% of the transduced Jurkat cells expressed GFP, and the GFP-positive cells expressed ≈50% less cell surface ADAM17 than the GFP-negative cells (Fig. 3A), whereas surface ADAM10 expression, the ADAM family member most similar to ADAM17 (38), was equivalent for both populations (Fig. 3A). Control ADAM17 shRNA was observed to have no effect on the surface levels of ADAM17 expression (Fig. 3A). Following the treatment of Jurkat cells transduced with active ADAM17 shRNA with rhFasL, L-selectin surface expression down-regulated to a greater extent by GFP-negative cells than by GFP-positive cells (Fig. 3B). When the transduced cells were treated with TAPI prior to the induction of apoptosis, L-selectin shedding by the GFP-negative cells was blocked, demonstrating that their down-regulation of L-selectin occurred by ectodomain shedding (Fig. 3B). Taken together, the above findings reveal ADAM17 activity upon the induction of leukocyte apoptosis, and that it is the primary sheddase of L-selectin during this process.
When monitoring L-selectin shedding by Jurkat cells undergoing induced apoptosis in the presence of TAPI over an extended time period, we noted that though initially blocked, considerable production of soluble L-selectin did occur at later time points. As shown in Figure 4A, soluble L-selectin production was assayed for up to 4 hr by Jurkat cells that were either untreated, treated with CH-11, or treated with CH-11 and TAPI. Two hours after the induction of apoptosis, high levels of soluble L-selectin were detected, which was essentially abolished by the presence of TAPI. At the 3 and 4 hr time points, however, apoptotic Jurkat cells in the presence of TAPI produced significantly higher levels of soluble L-selectin than time-matched, untreated cells (Fig. 4A). At these later time points of apoptosis, cell surface L-selectin also underwent a corresponding down-regulation in expression in the presence of TAPI (Fig. 4B). This late apoptotic process did not result in a general down-regulation in the expression of cell surface molecules, as the surface levels of LFA-1 on the same cells were only modestly decreased (Fig. 4B). We also monitored ADAM17 surface expression on Jurkat cells at various time points of apoptosis. Jurkat cells treated with CH-11 for 2 hours demonstrated surface ADAM17 expression levels equivalent to time-matched, untreated cells, whereas cells treated for with CH-11 for 4 hours demonstrated a marginal decrease in ADAM17 staining when compared to time-matched, untreated cells. Hence, soluble L-selectin production occurred at later stages of apoptosis in the presence of a broad-spectrum metalloprotease inhibitor and during the down-regulation of ADAM17 surface expression.
An obvious consideration for assays involving TAPI is its stability during extended cell incubation periods. We examined this in a cell-based assay by collecting media supernatants from Jurkat cells treated with CH-11 for 4 hr in the presence or absence of TAPI, and using the media supernatants to suspend untreated Jurkat cells, which were then overtly activated with PMA to induce L-selectin shedding. We found that activated Jurkat cells incubated in media supernatant from the CH-11/TAPI-treated cells did not down-regulate their expression of surface L-selectin, whereas Jurkat cells incubated in media supernatant from the CH-11-treated cells did (Fig. 4C), indicating activity by the metalloprotease inhibitor. We have also added TAPI to Jurkat cells in culture prior to and at regular time points after the induction of apoptosis, and again noted a significant down-regulation in L-selectin expression at the later time points of apoptosis (data not shown).
In addition to a broad-spectrum metalloprotease inhibitor, we examined the effects of other protease inhibitors on the late apoptosis release of soluble L-selectin, in particular inhibitors of serine proteases, which can also cleave L-selectin and have been implicated in ectodomain shedding (39–41). As shown in supplemental Figure 2, Jurkat cells treated with TAPI to block the initial shedding of L-selectin during the induction of apoptosis along with various other broad-spectrum protease inhibitors produced similar levels of soluble L-selectin as cells treated with TAPI alone.
Membrane blebbing and microparticle production are additional mechanisms by which certain surface proteins can be down-regulated during leukocyte apoptosis. Interestingly, CD43 and CD16, which undergo ectodomain shedding upon neutrophil activation (42), can be down-regulated in expression during neutrophil apoptosis by these processes (43, 44). To determine whether L-selectin released from the surface of Jurkat cells treated with CH-11 and TAPI was associated with membrane particles, these were removed from the media supernatant by filtration and ultracentrifugation. For instance, media supernatant from Jurkat cells treated with CH-11 and TAPI for 4 hr was divided into two equal volumes and one of the aliquots was filtered (0.22 µm pore size) then centrifuged at 100,000 × g for 1 hr. Soluble L-selectin levels in the two aliquots were then compared by ELISA. If soluble L-selectin produced by apoptotic Jurkat cells in the presence of TAPI was primarily associated with membrane particles, we would expect that its levels would be significantly decreased after the described treatments; however, this was not the case (Fig. 4D).
Soluble L-selectin occurs at high levels in the blood of healthy humans and mice (45, 46). Interestingly, the levels of soluble L-selectin in the blood of ADAM17 and wild-type chimeric mice are not significantly different (8), suggesting another mechanism for its production. As with Jurkat cells, we observed that murine neutrophils also down-regulated their surface L-selectin at later time points of induced apoptosis in the presence of TAPI. For instance, by 3 hr of induced apoptosis, neutrophils from unmanipulated mice treated with TNFα/cycloheximide in the presence of TAPI demonstrated an appreciable down-regulation in L-selectin expression, as determined by flow cytometry (Fig. 5A). Again, surface levels of LFA-1 on the same cells decreased only marginally at this time point (Fig. 5A). In contrast to early time points following the induction of apoptosis, neutrophils isolated from ADAM17 chimeras also demonstrated a down-regulation in surface L-selectin expression as apoptosis progressed (Fig. 5B), resulting in a corresponding increase in soluble L-selectin levels (Fig. 5C). Altogether, the above findings reveal that soluble L-selectin production can occur by distinct mechanisms invoked at different stages of induced apoptosis.
Emerging evidence indicates that ectodomain shedding by leukocytes is a key post-translational mechanism for regulating inflammation. In part, this process modulates the activity of various cytokines, cytokine receptors, and adhesion molecules (5). ADAM17 induction occurs upon cell activation, and its biological functions have been mainly examined in this context. The fate of neutrophils following their infiltration into sites of inflammation is programmed cell death, which is an effector activity that plays a vital role in resolving acute inflammation (20, 21). In this study, we provide the first direct evidence that ADAM17 cleaves L-selectin upon the induction of neutrophil apoptosis, which appears to be an early apoptotic event. We also report that as the apoptotic process progresses, L-selectin can be released from the surface of leukocytes by a mechanism distinct from ADAM17-mediated shedding and the membrane events of blebbing and microparticle production.
We have previously reported that L-selectin undergoes efficient shedding upon the induction of human neutrophil apoptosis by a mechanism that involves metalloproteases (23). L-selectin shedding has been reported to occur in an ADAM17-dependent and independent manner (8, 28–30), and thus we directly examined its role in L-selectin shedding upon death receptor-induced leukocyte apoptosis using two distinct genetic approaches. Fas signaling of apoptosis in Jurkat cells has been extensively studied (37), and these cells, which express high levels of surface L-selectin, were employed to knock-down ADAM17 expression using shRNA. This approach decreased the expression of cell surface ADAM17, but not the expression of its most similar family member ADAM10, and also reduced L-selectin shedding upon Fas engagement when compared with control cells. We also examined the shedding of L-selectin by neutrophils lacking functional ADAM17 following the induction of apoptosis. In contrast to wild-type neutrophils, ADAM17-deficient neutrophils demonstrated a near complete abrogation of L-selectin shedding at early time points of induced apoptosis. ADAM8 has been implicated in L-selectin shedding by neutrophils as well (29), and we also directly examined its role by using neutrophils from ADAM8 knock-out mice. We observed that these cells efficiently down-regulated their surface L-selectin expression soon after the induction of apoptosis. Hence, our findings indicate that in addition to neutrophil activation, ADAM17 is a primary sheddase of L-selectin upon death receptor-induced apoptosis.
Activation-induced L-selectin shedding regulates the receptor’s cell surface density and neutrophil adhesiveness (18, 47). The purpose of L-selectin shedding during neutrophil apoptosis is less clear at this time, though receptor down-regulation by apoptotic neutrophils likely contributes to a general diminution in their responsiveness to inflammatory cues (22). Neutrophils that have recently infiltrated sites of inflammation already express low levels of L-selectin due to their activated state (8, 18, 48), and thus it would seem that nominal levels of cell surface L-selectin would be available for further ADAM17-mediated cleavage upon the induction of neutrophil apoptosis. However, activated neutrophils express L-selectin mRNA, actively synthesize the adhesion molecule, and they can even re-express appreciable levels of surface L-selectin following their activation, but prior to undergoing apoptosis (23) (49, 50). Consequently, ADAM17-mediated shedding during neutrophil apoptosis at sites of inflammation may produce physiologically relevant levels of localized soluble L-selectin, among additional substrates. Other settings in which ADAM17-dependent shedding may occur upon leukocyte apoptosis is during the clearance of senescent neutrophils from the blood and the turnover of lymphocytes during negative selection, which may be a mechanism for the down-regulation of various receptors and/or the production of assorted soluble agonists and antagonists.
ADAM17’s enzymatic activity is robustly induced upon cell activation with various stimuli, which occurs in part through an intrinsic process (51, 52). We speculate that ADAM17’s enzymatic activity is enhanced upon the induction of leukocyte apoptosis as well, though at this time we cannot rule out that changes in L-selectin’s conformation and/or cell surface distribution may also contribute to its shedding during this cellular process. We have recently reported that redox modifications of cysteinyl sulfhydryl groups in the ectodomain of mature ADAM17 up-regulates its enzymatic activity (14). It will be interesting to determine whether this inducer mechanism may regulate ADAM17’s activity upon death receptor-induced neutrophil apoptosis as well. In addition, increased conversion of ADAM17’s pro-form to mature-form has also been reported during leukocyte apoptosis (53).
Our data also reveals that during the late stages of neutrophil apoptosis, L-selectin expression can be down-regulated by a mechanism distinct from ADAM17, which resulted in the production of soluble L-selectin. CD16 and CD43 can undergo ectodomain shedding as well upon neutrophil activation, and have been reported to be released from the surface of neutrophils by membrane blebbing and microparticle production during apoptosis (43, 44). An approach used to discern soluble adhesion molecules that are cleaved from those that are associated with membrane particles is by sedimentation of the latter using ultracentrifugation (44, 54). We found that a combination of filtration (≥ 0.22 µm particle size) as well as ultracentrifugation (100,000 × g) did not significantly reduce the levels of soluble L-selectin in the tissue culture media supernatant from cells undergoing prolonged apoptosis in the presence of the broad spectrum zinc metalloprotease inhibitor TAPI, suggesting that it may have been produced by a proteolytic process.
ADAM10 has been reported to cleave L-selectin in the absence of its primary sheddase ADAM17 (30). However, it would seem unlikely that ADAM10 is playing a major role in soluble L-selectin production by apoptotic leukocytes considering that it has been shown to down-regulate in expression during this process (23), and that soluble L-selectin production occurred in the presence of TAPI, which is known to block ADAM10 activity (55–57). Serine proteases have been implicated as sheddases as well and can also cleave L-selectin (39–41). However, various broad-spectrum serine protease inhibitors, as well as inhibitors of aspartic and cysteine proteases, also did not block soluble L-selectin production by late apoptotic leukocytes. Thus, the precise proteolytic process by which soluble L-selectin production occurs during later stages of leukocyte apoptosis remains to be determined. In addition, it will be interesting to determine whether this putative protease targets other cell surface determinants. For instance, surface expression of PSGL-1, a ligand of L-selectin (58), can also be down-regulated during neutrophil apoptosis, which appears to involve a non-ADAM17 proteolytic process as well (54).
The current study is the first to establish mechanisms of soluble L-selectin production upon death receptor-induced, neutrophil apoptosis. We have previously reported that ADAM17 deficient neutrophils can generate soluble L-selectin upon spontaneous apoptosis, which appears to involve other metalloproteases (8). The biological significance of ADAM17-independent processes that mediate the release of surface L-selectin from apoptotic neutrophils is unclear at this time. Soluble L-selectin is maintained at high levels in the blood of healthy individuals and mice (45, 46). We have reported that the levels of soluble L-selectin in the blood of ADAM17 and wild-type chimeric mice are not significantly different (8), suggesting that ADAM17 may not be the principle means of maintaining blood levels of soluble L-selectin. However, Venturi et al. have reported that soluble L-selectin in the blood is primarily derived by proteolytic cleavage, as gene-targeted mice expressing non-cleavable L-selectin have considerably lower levels of serum L-selectin (18). In conjunction with our findings, it is tempting to speculate that soluble L-selectin levels in the blood of normal individuals may be maintained by a proteolytic mechanism other than ADAM17, perhaps occurring during leukocyte apoptosis.
In conclusion, ADAM17 appears to be active during the various phases of a neutrophil’s short life span; such as during their time in circulation [e.g. maintains L-selectin at an appropriate cell surface density (8)], upon their activation (8), and during death receptor-mediated apoptosis. The induction of ectodomain shedding during leukocyte apoptosis likely has a much broader purpose than just cleaving L-selectin. Indeed, various leukocyte determinants that undergo ectodomain shedding have an important role in regulating the resolution of inflammation, including TNFα, TNFRI, TNFRII, FasL, TRAIL, and IL-6R (53, 59–63), and thus this process may be another component of the anti-inflammatory program initiated by neutrophils (20) (21).
We greatly appreciate the gifts of ADAM17(TACE)+/ΔZn mice from Dr. Roy Black (Amgen Inc.) and ADAM8 knock-out mice from Dr. Andy Docherty (Celltech R&D) and Dr. Carl Blobel (Weill Medical College of Cornell University). We thank Adam Nettles for his technical assistance.
1The study described was supported by Grant Number R01HL61613 from the National Institutes of Health.
3Abbreviations used in this paper: ADAM17, a disintegrin and metalloprotease; APC, allophycocyanin; rhFasL, recombinant human Fas ligand; CHX, cycloheximide
Publisher's Disclaimer: This is an author-produced version of a manuscript accepted for publication in The Journal of Immunology (The JI). The American Association of Immunologists, Inc. (AAI), publisher of The JI, holds the copyright to this manuscript. This version of the manuscript has not yet been copyedited or subjected to editorial proofreading by The JI; hence, it may differ from the final version published in The JI (online and in print). AAI (The JI) is not liable for errors or omissions in this author-produced version of the manuscript or in any version derived from it by the U.S. National Institutes of Health or any other third party. The final, citable version of record can be found at www.jimmunol.org.