,1
* Division of Pulmonary and Critical Care Medicine, University of Colorado Health Sciences Center, Denver;
Department of Pediatrics, National Jewish Medical and Research Center, Denver, Colorado; and
Department of Rheumatology, University of Washington, Seattle
1Correspondence: Department of Pediatrics, National Jewish Medical and Research Center, 1400 Jackson Street, Denver, CO 80206-2761. E-mail: hensonp{at}njc.org
ABSTRACT
The process of apoptosis includes critically important changes on the cell surface that lead to its recognition and removal. The recognition also generates a number of other local tissue responses including suppression of inflammation and immunity. It is surprising that the ligands generated on the apoptotic cell, which mediates these effects, have received relatively little attention. Some of these candidate molecules and possible mechanisms for their surface expression are addressed herein, with particular emphasis on phosphatidylserine and calreticulin. However, exposure of such ligands is exclusive to apoptosis and may, in fact, occur on viable cells. To partially explain the lack of response to such potential stimuli, the presence on viable cells of "don’t eat me" signals, in this case, CD47 is suggested to prevent such unwarranted actions. Loss or inactivation of the don’t eat me CD47 effects accompanies apoptosis and now allow the cells to be recognized and cleared.
Key Words: apoptosis • efferocytosis • calreticulin • phosphatidylserine • CD47 • collectins
Apoptosis is a mechanism for cell deletion. As such, it involves inactivation of the cell’s replicative apparatus with a presumed protection against abnormal division and also critically, changes to the cell surface, which lead to its recognition and eventual removal. Clearance of cells by phagocytic processes was first noted by Metchnikov at the turn of the 20th century [1 ]. In the early 1980s, a link between apoptosis (the term coined by Wyllie et al. [2 ] to describe a form of programmed cell death) and this cell removal was demonstrated in mammals and nematodes [3 , 4 ]. Since then, a huge effort has been made to understand the pathways leading to apoptosis in most cell types and increasingly, the mechanisms of recognition and uptake of such cells by phagocytes. It is surprising, however, that much less attention has been paid to the surface changes on the apoptotic cells that mediate this recognition. This, then, will be the focus of this short review.
However, before we discuss the recognition ligands on apoptotic cells, a few general points about apoptotic cell clearance and its consequences need to be addressed. The consequence of inefficient apoptotic cell clearance is secondary necrosis or postapoptotic cytolysis with release of cell constituents into the tissues. These are often proinflammatory and proimmunogenic, and indeed, disrupted cell removal has been associated with supply of autoantigens in various autoimmune conditions. In vivo, apoptotic cell removal appears to be highly efficient, to such a degree that we have suggested that detection of apoptotic cells in a tissue should raise the possibility of disrupted clearance mechanisms or a really large overload of the system with apoptotic cells (see, for example, ref. [5 ]). Although most of the attention in the mammalian literature has addressed uptake of apoptotic cells by macrophages (or immature dendritic cells), it is also apparent that many different cell types can mediate this clearance, including fibroblasts, epithelial, endothelial and smooth muscle cells, and stromal cells among others. At this point, the degree to which the various cell types carry out apoptotic cell removal in vivo is not clear, but its overall speed and efficiency argue that the involvement of so-called nonprofessional phagocytes may have been under-represented. Recognition of apoptotic cells and their uptake is observed in all metazoan, and increasingly, the pathways of signaling for uptake by the phagocytosing cells are seen to be extremely conserved (Fig. 1 ). For example, a number of the intracellular signal molecules can be readily exchanged between C. elegans and mammalian systems [6 7 8 ]. In addition, there is increasing evidence to show that the uptake mechanisms, receptors, and signal pathways are not only highly conserved but are significantly different [9 , 10 ] from the phagocytosis associated with immunoglobulin (Ig) Fc receptor recognition (the so-called zipper mechanism described by Griffin and co-workers [11 ]). For these reasons, we have coined the term "efferocytosis" for apoptotic cell uptake (from effero; to carry to the grave; to bury [9 , 12 ]).
 View larger version (19K): [in a new window] |
Figure 1. Pathways of intracellular signaling associated with efferocytosis in Caenorhabditis elegans or mammals. PSRS, Phosphatidylserine (PS) recognition structure.
|
 View larger version (23K): [in a new window] |
Figure 2. The tether and tickle hypothesis for efferocytosis [10
].
|
Accordingly, the universality of the recognition and response to apoptotic cells (whatever the cell type that is becoming apoptotic) has led us to ask if there are general surface changes occurring during apoptosis that drive these responses. The first such ligand described was PS. Normally present only on the inner leaflet of the plasma membrane bilayer, this phospholipid was shown to be exposed on apoptotic cells [20 , 21 ] and to participate in recognition of these cells by a wide variety of responding cells for efferocytosis or the generation of anti-inflammatory mediators. The current thoughts to explain the exposure of PS on the cell surface are illustrated in Figure 3 [22 23 24 ]. Normally, in resting, viable cells, phospholipid asymmetry is maintained with low levels of flip-flop (flip refers to inward and flop to outward movement) across the bilayer, and errant PS is flipped back to the inner leaflet by the activity of aminophospholipid translocase(s). During cell activation, phospholipid flip-flop (sometimes called scrambling, as it is not specific for a given phospholipid head group) is increased, and aminophospholipid translocase is maintained, and exposure of PS in the outer leaflet is transient [24 25 26 ]. By contrast, during apoptosis, there is a coordinate increase in phospholipid flip-flop with permanent inactivation of the aminophospholipid translocase, leading to persistent PS exposure. This has led to a commonly used assay for apoptosis, namely the binding of annexin V (which recognizes PS), but has to be controlled for cell permeability to avoid the annexin gaining access to the intracellular PS. As implied by the comments on activation-induced PS exposure, some caution should also be applied to this assay if the cells in question are likely to be in an activated state.
 View larger version (30K): [in a new window] |
Figure 3. Mechanisms for exposure of PS on the surface of activated or apoptotic cells. PC, Phosphorylcholine; PE, phosphatidylethanolamine; SM, sphingomyelin.
|
 View larger version (26K): [in a new window] |
Figure 4. PS exposure on apoptotic cells occurs in patches and is associated with "rafts." An apoptotic neutrophil is depicted with surface PS detected with fluorescent factor Va and rafts with cholera toxin B.
|
v integrins, as well as Gas 6 [31
]) and probably, protein S [32
], which appears to interact with the receptor tyrosine kinase Mer. Other PS-binding bridge molecules with less clear roles include ß2-glycoprotein 1 [33
]. Conversely, a widely distributed and conserved molecule, described originally as a specific PS receptor (PSR) [34
], now appears not to serve this role and seems to occupy a predominantly nuclear location (see refs. [35
36
37
]). This gene appears critical in vertebrate development, but its potential role in apoptotic cell uptake is still unclear. The study from Bose et al. [35
] showed no evidence of defective uptake in their PSR–/– mice, but other knockouts in mice or other species were not so definitive [38
39
40
41
], and some reconciliation is still needed to finally sort out the discrepancies. The question remains as to whether there is a specific PSR still out there. At this point in time, the above-mentioned bridge molecules do not easily explain all the observed responses to PS, including the stereospecificity [17
, 20
, 42
] and the anti-inflammatory and anti-immunogenic effects, leaving us to postulate the presence of a still-unidentified receptor. An IgM monoclonal antibody Mab217 [34
] stimulates identical responses in a wide variety of cells as seen with PS-exposing apoptotic cells or PS-containing liposomes and may react with this candidate structure (which at the moment, we are calling PSRS). It did not show any diminished binding to the Bose et al. [35
] knockout cells, and in our hands, its binding does not correlate with the level of expression of the "PSR"—further supporting the contention that this latter does not fit the parameters of a receptor, which is involved in PS recognition in apoptotic cell removal. OTHER APOPTOTIC LIGANDS
Over the course of the past 20 years, a number of other molecules or structures have been tentatively identified on, or released from, apoptotic cells. These include carbohydrate ligands, e.g., aminosugars [43 ] or mannose [44 , 45 ], intercellular adhesion molecule-3 [46 , 47 ], lysophospholipids such as lyso-PC [48 49 50 ], and unknown structures, which may include PS contributing to activation of complement [51 ]. The case for oxidized phospholipids, especially oxidized PS, has been made by a number of investigators (see refs. [27 , 50 ]). These may interact with candidate PSRs or binding molecules or with the various members of the scavenger receptor family, which have been implicated in apoptotic cell uptake (see ref. [52 ]). The oxidized versions of PS may be intrinsically resistant to aminophospholipid translocase-induced flipping, thereby remaining on the cell surface [23 ]. Nonspecific surface changes have also been invoked, including alterations in surface charge [43 ], possibly relating to changes in glycosyl groups.
COLLECTINS
The collectin family of innate immune system, pattern recognition molecules may be considered to include mannose-binding lectin (MBL), C1q, surfactant protein D (SP-D), which is found in the lung and at other mucosal surfaces, and SP-A, which is lung-specific. Ficolins are a related group of proteins of similar structure and probably function. Each of the collectin family noted above has been shown to bind to the surface of apoptotic cells and to mediate in vitro and for SP-D, C1q, and MBL in vivo as well, the clearance of apoptotic cells [53
54
55
56
57
]. The structures on the apoptotic cell surface to which they bind have not been characterized but do involve the pattern-recognizing globular heads of these molecules. As the collectin family can bind various carbohydrates, proteins, or lipids, the candidates could be extensive. Nevertheless, the binding allows the collectins to act as opsonins, presenting their conserved, collagenous tail regions to the responding phagocyte to drive uptake of the apoptotic cell. Over the years, many candidate C1q receptors have been described, and the importance of this family in host defense is in keeping with significant redundancy at this level. However, a cC1q receptor, which now appears to bind all four of these collectin family members through their collagenous (hence, c) domains, was originally described by Ghebrehiwet and colleagues [58
, 59
]. This was later shown to be identical to cell-surface calreticulin (CRT) [60
]. Later studies [54
, 55
, 61
] showed that collectin binding to apoptotic cells induced their uptake through interaction of the CRT (which does not traverse the membrane and does not have a candidate signaling motif) with CD91, which is also known as low-density lipoprotein receptor-related protein 1 (LRP1) or 2 macroglobulin receptor and is a highly effective internalization receptor. This effect is illustrated in Figure 5A
.
 View larger version (16K): [in a new window] |
Figure 5. Proposed cis or trans roles for CRT and its interaction with LRP in the process of efferocytosis. 2MR, 2 Macroglobulin receptor; ACAMPs, apoptotic cell-associated molecular pattern molecules.
|
CRT has usually been considered a soluble protein of the endoplasmic reticulum lumen, where it acts as a calcium store and chaperone [62 ]. The issue of how it gains access to the surface is still not understood, although possible transport with membrane proteins, which it serves to chaperone, such as major histocompatibility complex class I, has been suggested [63 ]. The mechanisms for expression, especially during the onset of apoptosis, are an important area, which needs investigation. Similarly, as noted above, the molecule does not contain a transmembrane domain nor glycosylphosphatidylinositol linkages. Its binding partners on the surface also need to be determined. Data presented in ref. [12 ] suggest that surface sites are saturable and that more binds to apoptotic than viable cells, but the structures to which it attaches were not determined. Some may bind to LRP itself, although Murphy-Ullrich and co-workers [64 ] have suggested that a different CRT stimulus, thrombospondin, can drive surface CRT to then secondarily interact with the LRP.
The collectin stimulation of LRP can be seen as a cis action of CRT already on the phagocyte surface, perhaps even already bound to the LRP. However, it has become increasingly apparent that most cell types express CRT on their surface. This raised the possibility that CRT could also act in a trans mode to directly activate LRP on the phagocytic cell (Fig. 5B) , i.e., without intervening bridge molecules such as the collectins [12 ]. During apoptosis, surface CRT is increased on a wide variety of cells (including fibroblasts). In fact, cellular stress is often associated also with new CRT synthesis as well as increased surface expression [65 , 66 ]. Cells deficient in CRT undergo apoptosis normally but are inefficiently removed by phagocytes. However, preincubation with soluble CRT can restore the levels on the surface to those approximating the normal expression on viable or apoptotic cells, respectively. Under these circumstances, apoptotic, reconstituted cells are ingested normally (Fig. 6 ). Furthermore, as predicted, the uptake appears to require the presence and function of LRP on the phagocyte [12 ]. From these studies, we are suggesting that CRT represents another important and widely distributed ligand for recognition of apoptotic cells. However, there are a number of issues raised by this proposal that need to be considered.
 View larger version (18K): [in a new window] |
Figure 6. Surface CRT promotes uptake of apoptotic cells. Apoptotic CRT-deficient mouse embryonic fibroblasts (MEF) are poorly ingested by macrophages in comparison with wild-type cells. However, restoration of uptake was seen if the CRT–/– cells were preincubated with CRT, washed, and then presented to the phagocytes. Hsp90, Heat shock protein 90.
|
A third issue applies to both of these candidate ligands (and possibly others as well). CRT is present on viable cells, and PS can be exposed on stimulated, viable cells. Why then are such cells not removed by near neighbors or wandering phagocytes? We have previously implicated the patching of ligands, including the patched binding of collectins, to the apoptotic cell surface [54 ], as perhaps now being able to stimulate low-affinity, high-avidity receptors on the phagocyte. Although this indeed may play a role, and we are increasingly intrigued by the role of local ligand and receptor distribution on target and ingesting cell, this explanation seems unable to completely account for the observations, especially as PS is often exposed in highly localized regions on stimulated, viable cells [26 ]. Accordingly, we have recently explored the likely role of what John Savill has called "don’t eat me" signals on cells to regulate their recognition and uptake.
CD47 AS A DON’T EAT ME SIGNAL
A study by Brown et al. [67 ] showed that CD31 on viable neutrophils could act homotypically on phagocytes to mediate their disassociation from the phagocyte surface and thereby prevent their ingestion. Defective CD31 signaling in apoptotic cells canceled this effect and allowed engulfment. This led to John Savill developing the concept of don’t eat me signals.
CD47, also known as integrin-associated protein, is widely distributed on cells and can act as a ligand for the heavily glycosylated, immunoreceptor tyrosine-based inhibitory motif-containing, inhibitory protein signal regulatory protein (SIRP) on the phagocyte. Ligation of the SIRP extracellular domain by CD47 on viable cells leads to its tyrosine phosphorylation by unknown kinases [68
] and subsequent activation of inhibitory tyrosine phosphatases such as Src homology-containing tyrosine phosphatase-1 (SHP-1). CD47 stimulation of SIRP is known to block IgG or complement-induced phagocytosis [12
]. In keeping with a putative don’t eat me effect, erythrocytes from CD47-deficient mice are recognized and rapidly eliminated from the blood of wild-type recipients in vivo and phagocytosed by wild-type macrophages in vitro. For these reasons, we hypothesized a role for the CD47-SIRP system in preventing uptake of CRT or PS-expressing, viable cells and for its likely inactivation during apoptosis so that the apoptotic cells can be removed (Fig. 7
). On many cell types, CD47 expression is reduced during apoptosis, and on all, it is redistributed into patches, which are distinct from those containing PS and CRT [12
]. Whether this down-regulation or redistribution is sufficient or whether additional, functional alterations also occur, apoptotic cells (or CD47–/– viable cells) are no longer able to stimulate SIRP and the downstream SHP-1, whereas normal, viable cells or soluble CD47 constructs are highly effective at doing this. Although CD47 appears to be a critical regulator, we suggest that it is highly unlikely to be the only inhibitory signal preventing uptake of viable cells, which may have to be blocked to permit uptake of apoptotic cells. As noted, when CD31 (platelet-endothelial cell adhesion molecule) is expressed on target and phagocyte, this system is a candidate for such an effect, especially as the CD31 signaling may also act via SHPs. The concept in general has even led us to speculate that removal of cells may be more of a default process, which has to be actively kept at bay, rather like the concept of apoptosis itself being a default response of cells, which are not actively exposed to growth/maintenance signals.
 View larger version (16K): [in a new window] |
Figure 7. Proposed mechanisms by which CD47 serves as a don’t eat me signal for viable cells, which is inactivated when the cells become apoptotic.
|
ACKNOWLEDGEMENTS
This work was supported by GM61031, HL68864, HL34303, AI58228, and Pfizer Fellowship Award PN0503104.
Received October 3, 2005; revised December 15, 2005; accepted December 17, 2005.
REFERENCES
vß5 integrin and triggers DOCK180-dependent Rac1 activation for the phagocytosis of apoptotic cells Exp. Cell Res. 292,403-416[CrossRef][Medline] or calreticulin/CD91, lung collectins act as dual function surveillance molecules to suppress or enhance inflammation Cell 115,13-23[CrossRef][Medline]This article has been cited by other articles:
 |
|
 |
|
  A. Kukreja, S. Radfar, B.-H. Sun, K. Insogna, and M. V. Dhodapkar Dominant role of CD47-thrombospondin-1 interactions in myeloma-induced fusion of human dendritic cells: implications for bone disease Blood, October 15, 2009; 114(16): 3413 - 3421. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y.-J. Park, G. Liu, Y. Tsuruta, E. Lorne, and E. Abraham Participation of the urokinase receptor in neutrophil efferocytosis Blood, July 23, 2009; 114(4): 860 - 870. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. S. GREENBERGER Radioprotection In Vivo, March 1, 2009; 23(2): 323 - 336. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Agudo, M. C. Perez-Marin, P. Sobrado-Calvo, U. Lonngren, M. Salinas-Navarro, I. Canovas, F. M. Nadal-Nicolas, J. Miralles-Imperial, F. Hallbook, and M. Vidal-Sanz Immediate Upregulation of Proteins Belonging to Different Branches of the Apoptotic Cascade in the Retina after Optic Nerve Transection and Optic Nerve Crush Invest. Ophthalmol. Vis. Sci., January 1, 2009; 50(1): 424 - 431. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. C. Frasch, K. Z. Berry, R. Fernandez-Boyanapalli, H.-S. Jin, C. Leslie, P. M. Henson, R. C. Murphy, and D. L. Bratton NADPH Oxidase-dependent Generation of Lysophosphatidylserine Enhances Clearance of Activated and Dying Neutrophils via G2A J. Biol. Chem., November 28, 2008; 283(48): 33736 - 33749. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S.-Y. Park, S.-Y. Kim, M.-Y. Jung, D.-J. Bae, and I.-S. Kim Epidermal Growth Factor-Like Domain Repeat of Stabilin-2 Recognizes Phosphatidylserine during Cell Corpse Clearance Mol. Cell. Biol., September 1, 2008; 28(17): 5288 - 5298. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Q. Xiao, C. G. Freire-de-Lima, W. P. Schiemann, D. L. Bratton, R. W. Vandivier, and P. M. Henson Transcriptional and Translational Regulation of TGF-{beta} Production in Response to Apoptotic Cells J. Immunol., September 1, 2008; 181(5): 3575 - 3585. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. M. Henson and R. M. Tuder Apoptosis in the lung: induction, clearance and detection Am J Physiol Lung Cell Mol Physiol, April 1, 2008; 294(4): L601 - L611. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Monks, C. Smith-Steinhart, E. R. Kruk, V. A. Fadok, and P. M. Henson Epithelial Cells Remove Apoptotic Epithelial Cells During Post-Lactation Involution of the Mouse Mammary Gland Biol Reprod, April 1, 2008; 78(4): 586 - 594. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. N. Maiti, K. Balasubramanian, J. A. Ramoth, and A. J. Schroit {beta}-2-Glycoprotein 1-dependent Macrophage Uptake of Apoptotic Cells: BINDING TO LIPOPROTEIN RECEPTOR-RELATED PROTEIN RECEPTOR FAMILY MEMBERS J. Biol. Chem., February 15, 2008; 283(7): 3761 - 3766. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Baumgartner, M. Geiger, S. Zieseniss, J. Malleier, J. A. Huntington, K. Hochrainer, E. Bielek, M. Stoeckelhuber, K. Lauber, D. Scherfeld, et al. Phosphatidylethanolamine critically supports internalization of cell-penetrating protein C inhibitor J. Cell Biol., November 19, 2007; 179(4): 793 - 804. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. A. Trouw, A. A. Bengtsson, K. A. Gelderman, B. Dahlback, G. Sturfelt, and A. M. Blom C4b-binding Protein and Factor H Compensate for the Loss of Membrane-bound Complement Inhibitors to Protect Apoptotic Cells against Excessive Complement Attack J. Biol. Chem., September 28, 2007; 282(39): 28540 - 28548. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V. Venegas and Z. Zhou Two Alternative Mechanisms That Regulate the Presentation of Apoptotic Cell Engulfment Signal in Caenorhabditis elegans Mol. Biol. Cell, August 1, 2007; 18(8): 3180 - 3192. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. M. Kinchen and K. S. Ravichandran Journey to the grave: signaling events regulating removal of apoptotic cells J. Cell Sci., July 1, 2007; 120(13): 2143 - 2149. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. C. James, A. H. Keeble, Z. Khan, D. A. Rhodes, and J. Trowsdale Structural basis for PRYSPRY-mediated tripartite motif (TRIM) protein function PNAS, April 10, 2007; 104(15): 6200 - 6205. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Y. Tyurina, L. V. Basova, N. V. Konduru, V. A. Tyurin, A. I. Potapovich, P. Cai, H. Bayir, D. Stoyanovsky, B. R. Pitt, A. A. Shvedova, et al. Nitrosative Stress Inhibits the Aminophospholipid Translocase Resulting in Phosphatidylserine Externalization and Macrophage Engulfment: IMPLICATIONS FOR THE RESOLUTION OF INFLAMMATION J. Biol. Chem., March 16, 2007; 282(11): 8498 - 8509. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. S. Isenberg, M. J. Romeo, M. Abu-Asab, M. Tsokos, A. Oldenborg, L. Pappan, D. A. Wink, W. A. Frazier, and D. D. Roberts Increasing Survival of Ischemic Tissue by Targeting CD47 Circ. Res., March 16, 2007; 100(5): 712 - 720. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Wehner, C. N. Morrell, T. Reynolds, E. R. Rodriguez, and W. M. Baldwin III Antibody and Complement in Transplant Vasculopathy Circ. Res., February 2, 2007; 100(2): 191 - 203. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. G. Freire-de-Lima, Y. Q. Xiao, S. J. Gardai, D. L. Bratton, W. P. Schiemann, and P. M. Henson Apoptotic Cells, through Transforming Growth Factor-beta, Coordinately Induce Anti-inflammatory and Suppress Pro-inflammatory Eicosanoid and NO Synthesis in Murine Macrophages J. Biol. Chem., December 15, 2006; 281(50): 38376 - 38384. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. M. Henson, R. W. Vandivier, and I. S. Douglas Cell Death, Remodeling, and Repair in Chronic Obstructive Pulmonary Disease? Proceedings of the ATS, November 1, 2006; 3(8): 713 - 717. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Lucas, L. M. Stuart, A. Zhang, K. Hodivala-Dilke, M. Febbraio, R. Silverstein, J. Savill, and A. Lacy-Hulbert Requirements for Apoptotic Cell Contact in Regulation of Macrophage Responses J. Immunol., September 15, 2006; 177(6): 4047 - 4054. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
