|
|
||||||||
Published online before print March 24, 2006
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Department of Pathology, University of Washington, Harborview Medical Center, Seattle
1Correspondence: Department of Pathology, University of Washington, Harborview Medical Center, 325 9th Avenue, Box 359675, Seattle, WA 98104-2499. E-mail: ewraines{at}u.washington.edu
| ABSTRACT |
|---|
Key Words: ADAM protease adhesion cytokine endothelium
| INTRODUCTION |
|---|
Inflammatory cell surface proteins can be regulated by a number of different mechanisms including changes in transcription and translation, internalization and degradation, dimerization or formation of higher order functional complexes, and by post-translational modifications such as phosphorylation. In addition, a wide variety of inflammatory transmembrane cell surface proteins also exist as soluble forms (Table 1 ), suggesting that their cleavage and release from the cell surface may be additional, post-translational regulatory mechanisms. In this review, we summarize recent literature characterizing the cleavage and shedding of a variety of inflammatory cell surface proteins and discuss how this emerging mechanism may play a role in regulating the inflammatory process.
|
| LEUKOCYTE RECRUITMENT DURING INFLAMMATION REQUIRES COORDINATED INTERACTIONS BETWEEN CIRCULATING LEUKOCYTES AND ENDOTHELIAL CELLS |
|---|
|
Chemokines are critical for the rapid, physiological conversion of rolling behavior to firm integrin-dependent arrest [52 53 54 55 56 57 ]. This conversion can occur within seconds following chemokine stimulation of leukocytes and leads to integrin and cytoskeletal activation [56 , 58 59 60]. Chemokines further provide directional cues for specific leukocyte subsets, and the expression of chemokines is finely regulated by environmental stimuli [52 53 54 55 ]. In contrast, the regulatory steps involved in the conversion from firm adhesion to locomotion and endothelial diapedesis are less well-defined. However, transendothelial migration is known to be a rapid event. Once a leukocyte sticks to the endothelium, it only takes a few minutes to reach the subendothelial basement membrane [43 , 52 ].
| AN ATTRACTIVE MECHANISM TO TRANSITION BETWEEN STEPS REQUIRED FOR LEUKOCYTE RECRUITMENT IS CELL SURFACE-ECTODOMAIN SHEDDING |
|---|
Multiple proteases that can act at or near the cell surface, including soluble and transmembrane proteases (Table 1) , have been shown to be ectodomain sheddases. However, the majority of shed proteins identified to date is cleaved by members of three protease families: ADAMs, which are a family of type I transmembrane proteins with a unique domain structure that includes a disintegrin-like adhesive domain and a metalloprotease domain [62 63 64 ]; MMPs, a family of proteases, including transmembrane and secreted members, which have gained recognition for their ability to regulate cell behavior through cleavage of diverse substrates beyond matrix proteins [65 ]; and soluble neutrophil-derived serine proteases, which are released from activated neutrophils and are present in high concentrations at sites of inflammation [66 ]. Inflammatory cells are particularly rich sources of all of these protease families, and endothelial and leukocyte activation is associated with release and/or activation of the zymogen forms of many of the enzymes. As highlighted in Table 1 , ADAMs 17 and 10 appear to play a particularly prominent role in ectodomain shedding of inflammatory proteins at all stages of leukocyte recruitment.
The functional implications of ectodomain shedding are significant and diverse, as many released ectodomains gain or retain biological activity and act as soluble mediators (Fig. 2
). Proteolytic shedding on an individual cell can release transmembrane cytokines expressed as pro-forms such as TNF-
-generating soluble mediators (Fig. 2A)
, reduce the density of a receptor or adhesion molecule and also produce soluble antagonists (Fig. 2B)
, and promote further proteolysis of the initial transmembrane cytoplasmic fragment, which can transcriptionally alter gene expression (Fig. 2C)
. Although all of these effects of ectodomain shedding involve shedding by a cell expressing the enzyme and the substrate (acting in cis), it has also recently been demonstrated that a ligand can be proteolytically released from its membrane tether by a complex on an opposing cell (acting "in trans") composed of the ligands receptor and an ADAM protease [67
] leading to cell-cell repulsion (Fig. 2D)
. Thus, the cell surface is a highly dynamic microenvironment, and protease-mediated ectodomain shedding is a versatile, post-translational regulatory mechanism, which has the potential to rapidly modulate cellular responses at multiple levels.
|
| A HUMAN GENETIC MUTATION SUPPORTS A ROLE FOR ECTODOMAIN SHEDDING IN THE REGULATION OF INFLAMMATION |
|---|
, TNF-R1 [68
]. In the majority of TRAPS patients, TNF-R1 mutations are correlated with impaired cleavage and shedding of the TNF-R1 ectodomain following stimulation and with a corresponding decrease in the serum concentrations of soluble TNF-R1 [68
]. The chronic inflammatory syndrome seen in TRAPS patients can be reversed in the majority of patients with TNF-neutralizing therapy, and therefore, a clear link between the failure to shed TNF-R1 and dysregulation of the inflammatory response has been established [68
69
70
]. Although TRAPS patients highlight the consequence of impaired shedding of a single cytokine receptor, Table 1
provides a list of other inflammatory cell surface proteins whose release is regulated by ectodomain shedding from all of the stages of leukocyte recruitment. | ECTODOMAIN SHEDDING CAN RAPIDLY REGULATE LEUKOCYTE-ENDOTHELIAL INTERACTIONS AT EACH STEP OF LEUKOCYTE RECRUITMENT |
|---|
|
|
|---|
Leukocyte recruitment can be modulated by interference with ectodomain shedding of adhesion molecules on leukocytes and endothelial cells
Tethering to and rolling along the endothelium are the first steps in the cascade of leukocyte recruitment (Fig. 1)
and are mediated by the selectin family of adhesion molecules. L-selectin is constitutively expressed on a wide variety of inflammatory cells including monocytes, neutrophils, eosinophils, B cells, and subsets of T cells and is rapidly shed from the surface of leukocytes following activation by diverse stimuli [71
]. Its binding to carbohydrate counter-ligands on the endothelium in the presence of hydrodynamic shear is essential for leukocyte recruitment and lymphocyte trafficking [72
73
74
75
]. As illustrated in Figure 2B
, shedding of L-selectin reduces receptor density and releases soluble receptors. Even a 50% reduction in leukocyte cell surface levels of L-selectin can result in a 70% decrease in leukocyte migration [76
], and soluble L-selectin can further antagonize leukocyte migration [77
]. Although L-selectin is an adhesion molecule, its ligation or cross-linking can induce leukocyte activation and L-selectin shedding, suggesting that shedding may also serve as a negative-feedback mechanism [78
79
80
].
The primary protease responsible for inducible cleavage of L-selectin appears to be ADAM17 [81 ], and inhibitors of ADAMs and other metalloproteinases have been shown to decrease the rolling velocity of leukocytes in microvessels by 4060% [82 , 83 ]. Inhibition of L-selectin shedding also leads to increased exposure to the endothelium and leukocyte activation [83 ], effects not seen in L-selectin null mice [83 ]. L-selectin is lost from neutrophils in vivo during exudation to sites of inflammation [84 ] and on lymphocytes migrating across high endothelial venules into lymph nodes [85 ]. Metalloproteinase inhibitors, which block L-selectin shedding, arrest lymphocytes on the endothelial lining of high endothelial venules and delay transendothelial migration [85 ]. Similarly, leukocytes in transgenic mice expressing uncleavable forms of L-selectin on an L-selectin null background transmigrate to high endothelial venules more slowly, and they retain their ability to re-enter the peripheral lymph node once activated, as their L-selectin is not shed [86 , 87 ]. These data also illustrate that to reduce rolling velocity, shedding of L-selectin must occur within seconds and that L-selectin shedding can control the time of interaction of a leukocyte with the endothelium [52 ]. Regulation of L-selectin by shedding has been proposed to be a more generalized mechanism of action for diverse anti-inflammatory agents. Nonsteroidal, anti-inflammatory drugs decrease leukocyte recruitment, perhaps in part through their ability to induce metalloproteinase-dependent shedding of L-selectin from neutrophils in vitro and in vivo [88 ]. Together, the data suggest that inhibition of L-selectin shedding increases leukocyte adhesion and transmigration by increasing the duration of leukocyte exposure to the inflamed endothelium, which can also promote leukocyte activation by outside-in signaling.
On the endothelial side, the Ig family of cell-cell adhesion molecules, including ICAM-1 and VCAM-1, involved in rolling/capture and firm adhesion of leukocytes, is up-regulated rapidly on activated endothelial cells [89
, 90
]. Although soluble forms of ICAM-1 and VCAM-1 are elevated in the serum of patients at risk for cardiovascular disease and with a wide variety of inflammatory diseases [91
], more is known about the enzymes involved in VCAM-1 shedding (Table 1)
. Cleavage of VCAM-1 depends on the stimulus and cellular context. ADAM17 regulates its shedding from stimulated endothelial cells [33
, 34
], and elastase and Cat G released by activated neutrophils can also induce its proteolytic release [35
], including during G-CSF mobilization of hematopoietic stem cells [35
]. Although the neutrophil enzymes are required for VCAM-1 shedding during stem cell mobilization, the release of the hematopoietic stem cells is not, suggesting a mechanism independent of neutrophil proteases [92
]. Experiments to evaluate the effect of VCAM-1 shedding on rolling and firm adhesion have not been performed. However, results similar to those with L-selectin would be predicted with a decrease in receptor density and release of soluble receptors (Fig. 2B)
. Thus, shedding of VCAM-1 would decrease its binding to its leukocyte counter-receptor
4ß1 integrin, which could increase rolling velocity and delay firm adhesion. ADAM17 was also recently shown to be capable of inducing shedding of ICAM-1 [32
]. It is thus tempting to speculate that the process of firm adhesion could provide an opportunity for cleavage of VCAM-1 and/or ICAM-1 in trans through soluble or membrane-associated leukocyte proteases (Fig. 2D)
. The rate at which ADAM17 cleaves VCAM-1 and ICAM-1 might appear to be slower than that of L-selectin, as shedding assays for VCAM-1 and ICAM-1 are normally 16 h [32
, 33
] as compared with 1530 min for L-selectin [1
]. However, assay sensitivities are the primary reason for the different incubation times, and shedding rates have not been characterized in detail for any ADAM17 cell surface substrate. Although the in vivo significance of VCAM-1 and ICAM-1 shedding is still poorly understood, the relatively high concentrations of their soluble counterparts in vivo in different inflammatory diseases [91
] suggest that proteolysis could play an important role in their function. In vivo experiments in mice lacking the key enzymes and uncleavable forms of VCAM-1 and ICAM-1 should help clarify the relevance of these shedding events to the regulation of the inflammatory response.
Shedding can regulate signaling in the initial steps of activation required for firm adhesion
TNF-
is produced by inflammatory cells following infection and tissue injury and induces a cascade of endogenous, inflammatory mediators, including adhesion molecules and other cytokines. The importance of TNF-
in the inflammatory response is highlighted by the marked defects in leukocyte recruitment and resistance to endotoxic shock displayed in TNF-
-deficient mice, which lack pro- and soluble TNF-
[93
]. TNF-
is synthesized as a trimeric type II transmembrane-anchored precursor (Fig. 2A)
, and its release from the cell surface as a soluble mediator is regulated by proteolysis [21
, 22
]. In vivo studies have shown that many, but not all, of the inflammatory effects of TNF-
require cleavage and shedding from the cell surface, and targeted deletion of TNF-
from various leukocyte subsets has demonstrated that macrophages and neutrophils are the main source of systemic TNF-
after an inflammatory challenge [94
].
Distinct functions for the transmembrane and soluble forms of TNF-
have been distinguished by studies of mutant mice expressing wild-type or cleavage-resistant forms of TNF-
[94
, 95
]. Knock-in of an uncleavable, membrane-bound, pro-TNF-
allele onto the TNF null background, which completely eliminates inducible cleavage of pro-TNF-
and therefore, generation of soluble TNF-
, rescues only a subset of defects seen in TNF null mice [95
]. Mice expressing uncleavable pro-TNF-
are still resistant to endotoxic shock and show defects in leukocyte trafficking [95
], but the mice rescue TNF-
-null defects in secondary lymphoid organ structure [95
] and display cytotoxic activity, activation of B and T cells [96
, 97
], and endothelial cells [98
].
Effects of TNF-
are mediated by two receptors, TNF-R1 and TNF-R2 [99
]. Experiments using genetically deficient mice have shown that TNF-R1 is the primary receptor, which mediates the majority of inflammatory functions ascribed to soluble TNF-
, and those of membrane-bound pro-TNF-
are mediated predominantly by TNF-R2 [100
101
102
103
]. Like TNF-
, TNF-R1 and TNF-R2 can be shed from the cell surface through regulated proteolysis to generate soluble receptors [104
, 105
]. TNF-R shedding may attenuate TNF-
activity in at least two ways (Fig. 2B)
. First, shedding rapidly down-regulates cell surface expression of TNF-Rs and therefore, decreases the sensitivity of a target cell to a soluble TNF ligand [106
]. Second, as soluble TNF-Rs are able to bind ligands, they may function as competitive antagonists for soluble TNF-
[106
]. Ligation of TNF-R1 by soluble TNF-
induces cleavage and shedding of TNF-R1 and TNF-R2 from human leukocytes in vitro and in vivo [107
, 108
]. Similarly, histamine stimulation in vitro and in vivo leads to rapid shedding of TNF-R1 from endothelial cells and renders endothelial cells refractory to the proinflammatory effects of TNF-
[109
]. As highlighted by TRAPS patients, impaired TNF-R1 shedding is sufficient to induce chronic inflammation [68
].
To date, ADAM17 appears to be the predominant mammalian protease responsible for the inducible cleavage and shedding of TNF-
and TNF-Rs from T cells and myeloid cells [21
, 22
]. However, a variety of in vitro and in vivo studies suggests that additional proteases may possess catalytic activity against pro-TNF-
, TNF-R1, and TNF-R2 (Table 1)
. Increasing evidence suggests that cell type and tissue-specific differences lead to cleavage through ADAM17-independent mechanisms, such as macrophage infiltration and MMP-7-mediated cleavage in intervertebral disc remodeling [23
]. Depending on the activation state of neutrophils, elastase or ADAM17 can cleave TNF-RI, resulting in products with distinct molecular weights [26
] and potentially, different activities. Thus, the TNF/TNF-R system is regulated at multiple levels by members of both metalloproteinase superfamilies of proteases (ADAMs and MMPs) as well as by serine proteases such as neutrophil elastase. Proteolytic shedding of other members of the TNF-R family, such as the lymphoid activation marker CD30 and the immune modulator CD40, which are also shed by ADAM17 [5
, 6
], may similarly contribute to modulation of the inflammatory response (Table 1)
. Although the relative contributions of specific proteases to TNF-
ligand and receptor shedding within the context of specific inflammatory responses remains to be fully understood, defects in the generation of the soluble TNF-
ligand severely impair leukocyte recruitment, and mutations that prevent TNF-RI release lead to chronic inflammation.
Protease-induced function-switching of fractalkine from an adhesion molecule to a chemokine
Fractalkine is another adhesion molecule whose expression is up-regulated on activated endothelial cells and is detected in a variety of inflammatory diseases including atherosclerosis, psoriasis, and glomerulonephritis [110
]. Fractalkine has an unusual structure in which the N terminus encodes a chemokine domain, which is attached to a glycosylated, mucin-like transmembrane stalk [110
111
112
], and the ectodomain can undergo cleavage and shedding from the cell surface [12
13
14
]. The effects of soluble and transmembrane fractalkine are mediated by a single G-protein-coupled chemokine receptor, CX3CR1, which is expressed on inflammatory cells including monocytes, T cells, and natural killer (NK) cells [113
, 114
]. Immobilization of fractalkine on glass substrata and monolayers of fractalkine-transfected cells supports the capture and high-affinity adhesion of leukocytes [111
, 113
, 115
]. Further, using parallel-plate flow-chamber assays, immobilized fractalkine has been shown to synergistically enhance leukocyte adhesion mediated by VCAM-1 [116
]. Therefore, at the cell surface, transmembrane fractalkine is thought to function predominantly as an adhesion molecule, and its shedding would reduce its density (Fig. 2B)
. However, recent studies further suggest that cell surface fractalkine may also play a role in cell-cell signaling. Transmembrane fractalkine, but not its soluble form, has been shown to stimulate interferon-
production by NK cells [117
]. In contrast, following proteolytic release from the cell surface (Fig. 2A)
, soluble fractalkine functions as a potent chemoattractant cytokine for NK cells, T lymphocytes, and dendritic cells [112
, 113
, 118
]. Together, these findings suggest that transmembrane and soluble fractalkine indeed have distinct functions and that cleavage and shedding likely represent an important mechanism of function switching.
Fractalkine cleavage and shedding are mediated by two distinct metalloproteinase activities. ADAM17 is responsible for the rapid shedding of fractalkine following protein kinase C (PKC) activation [12 , 13 ], and ADAM10 is the constitutive fractalkine sheddase [14 ]. A distinct, subcellular storage compartment for fractalkine has also been characterized in endothelial cells and may serve as a source for acute up-regulation of cell surface levels of fractalkine following stimulation [119 ]. Inhibition of ADAM10-mediated fractalkine shedding increased the adhesion and prevented the de-adhesion of monocytic cells to monolayers of cultured cells expressing fractalkine [14 ]. Most simply, cleavage of fractalkine switches its function from a cell surface adhesion molecule to a soluble chemokine with chemotactic as well as autocrine/paracrine signaling capabilities. In addition to function-switching, cleavage of fractalkine may be a mechanism to modify the high-affinity interactions between fractalkine and its receptor, CX3CR1 (Fig. 2B) . Such cleavage may play a role in de-adhesion after the initial capture of CX3CR1-expressing cells by endothelial transmembrane fractalkine expressed at sites of inflammation.
Sequential proteolysis of the adhesion molecule CD44 regulates gene expression
A variety of inflammatory cells express a form of the cell adhesion molecule CD44, which can mediate recruitment to sites of inflammation by binding to endothelial-bound hyaluronan [120
, 121
]. Similar to L-selectin, ligation of CD44 by hyaluronan induces CD44 cleavage and shedding [122
]. Elevated levels of the soluble CD44 are detected in the serum of patients with arthritis and various forms of metastatic cancer [123
124
125
], and the release of soluble CD44 (Fig. 2B)
lowers receptor cell density and generates a receptor antagonist [126
]. An unusual aspect of CD44 shedding biology is that following metalloproteinase-mediated ectodomain cleavage, CD44 can also be cleaved within the transmembrane domain of the resulting cell surface remnant, leading to release of the CD44 cytoplasmic tail domain into the cytoplasm (Fig. 2C)
[27
, 127
, 128
]. In vitro studies have shown that intramembrane cleavage of CD44 requires prior metalloproteinase-mediated cleavage within the ectodomain [128
]. The resulting CD44 cytoplasmic tail fragment translocates to the nucleus, where it can function as a transcriptional activator of several genes, including CD44 itself [128
]. Intramembrane processing of cell surface proteins is thought to be mediated predominantly by aspartyl-class proteases,
-secretases, which possess the ability to cleave type I transmembrane proteins within the hydrophobic environment of the cell membrane [129
].
-Secretase-dependent intramembrane cleavage has also been described for the epithelial cell-cell adhesion molecules N-cadherin and E-cadherin [130
, 131
].
As indicated in Table 1 , multiple enzymes have been shown to be involved in the initial shedding of the CD44 ectodomain using at least three distinct cleavage sites, and MT1-MMP, ADAM10, and ADAM17 were implicated at the cellular level [27 , 28 , 132 ]. It is important that identical cleavage fragments and soluble ectodomains have been shown to exist in vivo, suggesting that they are likely physiologically relevant [132 ]. Together, these studies suggest a potential model, whereby sequential proteolytic cleavage of cell surface CD44 regulates its function at multiple levels. In the context of an inflammatory response, CD44 likely functions as a classic adhesion molecule, mediating the capture and rolling of leukocytes on activated endothelial cells in cooperation with other adhesion molecules (Fig. 2B) such as L-selectin, E-selectin, and fractalkine. Ligation of CD44 by its ligand hyaluronan leads to cross-linking and initiation of an intracellular signaling cascade, which ultimately leads to activation of ectodomain sheddases including MT1-MMP, ADAM10, and/or ADAM17. Although the intracellular mechanisms leading to protease activation are not yet completely understood, it has recently been shown that ligation of CD44 leads to activation of the guanosinetriphosphatase Rac1 followed by the redistribution of CD44 to newly formed membrane ruffles [133 ]. In addition, CD44 has been shown to direct MT1-MMP to such membrane ruffles at the leading edge of migrating cells [134 ]. Once activated, MT1-MMP and perhaps ADAMs then may selectively mediate the cleavage of CD44 within distinct regions of a migrating leukocyte. Based on in vitro studies with metalloproteinase inhibitors, cleavage of the CD44 ectodomain is predicted to regulate cell de-adhesion, which is required for effective cell migration (Fig. 2B) .
Changes in gene expression may initiate specific transcriptional programs, which prepare the leukocyte for downstream effector functions within the extravascular tissue. Thus, sequential processing of CD44 may modulate its activity as an adhesion molecule and also expands its function as a cell-signaling molecule, thereby linking ligands encountered in the extracellular environment to changes ultimately in gene expression. In support of this general mechanism of adhesion molecule regulation, it has recently been shown that
-secretase-mediated cleavage of E-cadherin induces disassembly of cadherin-based adherens junctions, leading to inhibition of cadherin-dependent cell-cell adhesion [130
]. Future studies reconstituting CD44 null mice with wild-type versus uncleavable CD44 mutants deficient in metalloproteinase- or
-secretase-dependent cleavage will help to clarify the emerging roles of sequential proteolysis in CD44 function and leukocyte recruitment.
| ECTODOMAIN SHEDDING AS A POST-TRANSCRIPTIONAL REGULATORY MECHANISM OF THE INFLAMMATORY RESPONSE |
|---|
|
In contrast, if the endothelium has been activated previously by inflammatory stimuli, then additional adhesion molecules including E-selectin, P-selectin, and fractalkine will be expressed on the endothelial surface. Under these conditions, L-selectin adhesion would be supplemented by these additional adhesive interactions such that the leukocyte may remain adherent after L-selectin cleavage. Consistent with this model, it has been shown that metalloproteinase inhibitors no longer decrease leukocyte-rolling velocities in vivo if the endothelium has been activated previously by inflammatory cytokines [82 ]. Through this mechanism, ADAM17-mediated cleavage of L-selectin may ensure that a leukocyte can roll and sample the endothelium at an appropriate velocity, such that a cell is captured if the endothelium expresses additional inflammatory adhesion molecules.
Leukocyte adhesion mediated by endothelial selectins is weak, and a transition to firm integrin-mediated adhesion is thought to be required for effective leukocyte recruitment. Firm adhesion requires activation of leukocyte integrins by endothelial-derived cytokines and subsequent binding of activated integrins to their endothelial ligands including ICAM-1 and VCAM-1 (Fig. 1)
. At this stage, it appears that endothelial cells and leukocytes "communicate" extensively in a juxtacrine cell-cell manner. Cleavage of endothelial transmembrane cytokines such as TNF-
and/or fractalkine may be one mechanism by which cytokines can be transferred between the endothelium and the adherent leukocytes. In contrast, ADAM17-mediated shedding of TNF-Rs from leukocytes and/or endothelium may down-regulate TNF signaling locally at sites of ongoing leukocyte recruitment. Therefore, ectodomain shedding may be an important mechanism by which leukocytes and endothelial cells communicate during the transition to firm adhesion and to prepare the leukocyte for subsequent effector functions.
Following firm adhesion, the leukocyte must migrate across the endothelium to its cell-cell junctions for final transmigration into the extravascular tissue (Fig. 1)
. Effective cell migration requires coordinated formation of new, adhesive interactions at the leading edge of the cell, complemented by de-adhesion at the trailing edge. Intracellular signaling, which induces cycles of integrin activation and deactivation, is one likely mechanism by which adhesive interactions are modified at the leading and trailing edge of a migrating leukocyte, respectively [56
, 58
]. However, proteolysis may be an additional mechanism. Leukocytes could use coactivation of PKC and Rac to induce accumulation of ADAM17 and CD44 at the leading edge to promote their turnover, and extending lamellipodia could induce cell stretching to trigger extracellular calcium influx at the rear of the cell, leading to rapid activation of ADAM10 to facilitate CD44 cleavage and cell detachmenta mechanism proposed to mediate CD44-dependent migration on extracellular matrix [136
]. Further proteolytic processing of the CD44 transmembrane fragment by
-secretase-like enzymes may additionally generate an intracellular cytoplasmic tail fragment leading to transcriptional activation within the migrating leukocyte (Fig. 2C)
. At the same time, leukocyte activation, in close contact to the underlying endothelium, may create a microenvironment in which neutrophil-derived proteases, such as elastase, can cleave endothelial cell surface proteins such as ICAM-1 and VCAM-1, further facilitating cell motility.
In the final step of diapedesis, migrating leukocytes crawl between tightly apposed endothelial cellsa rapid process that can be completed in
90 s. Several mechanisms have been proposed to allow the passage of leukocytes without compromising the integrity of endothelial junctions. Molecules concentrated at the junctions of endothelial cells and involved in the maintenance of their integrity [137
] play a particularly prominent role in regulating signaling and adhesive events in diapedesis [43
]. Homophilic interactions between endothelial and leukocyte-expressed PECAM-1 and CD99 act at sequential steps as the leukocyte crosses the endothelial barrier [138
]. For PECAM-1, it has been shown that targeted recycling occurs in a novel membrane compartment in endothelial cells in the zone where the leukocyte is migrating [139
]. Leukocyte diapedesis has also been demonstrated to induce selective dissolution of endothelial cell-cell contact structures containing PECAM-1 and VE-cadherin [140
]. As PECAM-1 and VE-cadherin have previously been shown to be substrates of metalloproteinases [36
, 38
], it is tempting to speculate that they may also be cleaved and perhaps shed during leukocyte transmigration.
| CONCLUDING REMARKS |
|---|
. L-selectin and TNF-
ligand binding can turn off their signaling and adhesive functions immediately by proteolytic shedding of L-selectin and TNF-Rs. Although antagonist activity of the soluble forms of L-selectin and TNF-Rs can locally prevent subsequent signaling through these pathways, they also allow leukocyte interaction with other signaling and adhesive molecules. Distinct signals can lead to cleavage of a single substrate by different sheddases, as has been shown for CD44, which is cleaved by ADAM10 following stimulation of extracellular calcium influx and by ADAM17 when PKC is activated [27
]. Within a given leukocyte, localized signals may further coordinate different processes at the leading edge versus the rear of the cell through protease-specific cleavage of the same or distinct substrates. Sequential proteolysis, as illustrated by CD44, further provides a mechanism for regulation of the gene program in the leukocyte or endothelial cell. Although our understanding of the specific roles of ectodomain shedding in leukocyte recruitment is minimal at present, impaired shedding of a single cytokine receptor is sufficient to cause a chronic inflammatory disorder as shown for TRAPS patients [68 ] and sends a clear message of the potential importance of shedding. A major challenge will be to characterize the nature and extent to which ectodomain shedding participates in the transitions between and/or coordination of the different steps of leukocyte recruitment. As targeted deletion of two of the enzymes implicated in the cleavage of many inflammatory substrates, ADAM10 and ADAM17, are embryonic or perinatal lethal [1 , 141 ], generation of mice with cell-specific, conditional deletion will be important to examine the role of each of these enzymes in leukocyte recruitment. Our discussion of the substrates involved in particular steps in leukocyte recruitment highlights the need for examination of the contribution of regulated shedding of individual substrates and the kinetics of their shedding. Data with uncleavable mutants have provided important insights for some of these substrates, but additional cleavage-resistant mutants are needed to allow substrate-specific perturbation at different stages of leukocyte recruitment. Only with these important, additional in vivo experiments will we be able to access the full extent to which regulated proteolysis helps to coordinate the complex interactions required for leukocyte emigration into sites of inflammation.
| ACKNOWLEDGEMENTS |
|---|
Received January 19, 2006; revised February 21, 2006; accepted March 2, 2006.
| REFERENCES |
|---|
-converting enzyme J. Immunol. 165,6703-6709
-converting enzyme. Implications for CD40 signaling J. Biol. Chem. 278,32801-32809
-converting enzyme (ADAM17) mediates the cleavage and shedding of fractalkine (CX3CL1) J. Biol. Chem. 276,37993-38001
-converting enzyme mediates the inducible cleavage of fractalkine J. Biol. Chem. 276,44622-44626
subunit J. Immunol. 169,5679-5688
converting enzyme J. Biol. Chem. 275,14608-14614
. Contribution of the stalk sequence to the cleavage pattern of transmembrane proteins Eur. J. Biochem. 267,2624-2631[Medline]
is generated by cleavage that involves the tumor necrosis factor-
-converting enzyme (TACE/ADAM17) J. Biol. Chem. 279,40368-40375
from cells Nature 385,729-733[CrossRef][Medline]
Nature 385,733-736[CrossRef][Medline]
in a model of herniated disc resorption J. Clin. Invest. 105,143-150[Medline]
and IL-1ß from a stimulated human monocytic cell line in the presence of activated neutrophils or purified proteinase 3 Proc. Natl. Acad. Sci. USA 96,6261-6266
and TNF-
and shed by the activity of the disintegrin-like metalloproteinase ADAM10 J. Immunol. 172,6362-6372
}-converting enzyme (TACE/ADAM-17) mediates the ectodomain cleavage of intercellular adhesion molecule-1 (ICAM-1) J. Biol. Chem. 281,3157-3164
-converting enzyme (ADAM 17) J. Biol. Chem. 278,37459-37464
B and cytokine-inducible enhancers FASEB J. 9,899-909[Abstract]
-granule membrane protein, is also synthesized by vascular endothelial cells and is localized in Weibel-Palade bodies J. Clin. Invest. 84,92-99[Medline]