Role of complement and complement regulators in the removal of apoptotic cells

Abstract

Apoptosis, followed by rapid phagocytic clearance, is the primary mechanism by which organisms dispose of unwanted cells. The intracellular and extracellular composition of an apoptotic cell changes to decrease immunogenicity and enhance its uptake. By changing their extracellular composition, apoptotic cells acquire the capacity to bind complement initiation molecules such as C1q and MBL. Binding of these molecules can lead to complement activation. Membrane bound complement inhibitors are down-regulated during apoptosis, which would leave the cell less protected against complement activation; however, recent data show that fluid-phase complement inhibitors may compensate for this loss of regulation. Importantly, binding of complement is a process that mainly takes place during the late stages of apoptosis. Most cells will be cleared before that stage under steady state conditions, but during overwhelming apoptosis or impaired phagocytosis, apoptotic cells may remain in tissues for a longer time and acquire complement proteins. Based on the data from deficiencies of early complement components and the development of systemic lupus erythematosus with accumulation of dead cells, it is clear that, under certain conditions, apoptotic cells persist, becoming necrotic and overloading the scavenging capacities of the complement system. Although the complement system is also involved in inducing apoptosis in target cells, this review will focus on the role of complement in the clearance of apoptotic cells.

Introduction

Tissue homeostasis is achieved by replacing old cells with new when needed, which requires the efficient and non-inflammatory removal of the old cells. Most cells are removed from the body via a process that involves apoptosis and efficient phagocytosis of the dead cells (Henson and Hume, 2006). Host-derived dead cells are a valuable source of self-antigens that can be used to instruct the adaptive immune system to distinguish between self and non-self. However, autoimmunity could arise if these self-antigens were presented in a pro-inflammatory context. In many clinical conditions apoptosis and clearance of apoptotic cells has been suggested to play an important role. Understanding how the complement system is involved in this process will be valuable for intervention in autoimmune diseases, inflammatory conditions, neurological pathologies and malignancies.

The body has devised redundant mechanisms that ensure that the process of cell removal and replacement is kept in an anti-inflammatory state (Gardai et al., 2006). This is mediated by careful induction of cell death via apoptosis: an active process of cell death that involves activation of several intracellular pathways that make sure the intracellular content, including the nucleus, is condensed, fragmented and stored in ‘safe’ vesicles. Several changes also take place at the cell surface that makes the cell recognizable to the surrounding cells and phagocytes as being dead and ready to be removed. One of the best known changes is the loss of lipid bilayer asymmetry (or membrane flip-flop), exposing high levels of phosphatidyl-serine, which by itself already constitutes a pro-phagocytic ‘eat me’ signal (Fadok et al., 1998). Conversely, several membrane proteins normally present on the cell surface are either internalized or shed (Elward and Gasque, 2003, Elward et al., 2005, Gardai et al., 2005). Down-regulation of several molecules, such as CD47, CD200 and sialic acid (Hoek et al., 2000, Oldenborg et al., 2000, Wright et al., 2000), may already initiate recognition because they serve as ‘don’t-eat me’ signals. Other molecules that are normally only found in the intracellular compartment may acquire access to the cell surface and enhance recognition via receptors on neighboring cells or on phagocytes (Gardai et al., 2005, Gardai et al., 2006). In addition to changes from within, the cell also acquires proteins from the extracellular environment that bind to the apoptotic cell surface and may serve as opsonins (Krysko et al., 2006). An array of non-complement molecules have been reported to do so, including β2 glycoprotein, milk fat globule protein (MFG-E8), protein S, growth arrest specific gene number 6 (Gas-6), thrombospondin, IgM, pentraxin 3 (PTX3), serum amyloid P component (SAP), C-reactive protein (CRP), and several components of the complement system such as C1q, MBL, ficolins and C3 also bind. It is currently unknown whether this process is equally important for cells that are present in a protein rich environment, such as circulating blood cells, as compared to cells within tissues. Numerous in vitro studies have shown that binding of these plasma proteins enhanced uptake by phagocytes; however, it is interesting to note that other studies have reported inhibition of uptake for cells coated with PTX3 (van Rossum et al., 2004) and C4b-binding protein (C4BP)–protein S complex (Kask et al., 2004). The relative binding affinities of these different proteins and the ligands involved are less well characterized. However, differences in experimental protocol, cell types studied and precisely how investigators define early and late apoptosis have led to big differences between studies regarding the time point at which dying cells acquire complement components as opsonin. Although some findings were reported to show binding during early apoptosis, it is now becoming increasingly clear that binding of complement initiation molecules and complement activation are events that mainly take place on late apoptotic cells (Gaipl et al., 2001, Trouw et al., 2007). This is in contrast to some of the earlier apoptotic events such as membrane flip-flop, down-regulation of ‘don’t-eat me’ signals and binding of other plasma proteins, which initiate phagocytosis long before complement comes into play.

The initial stages of apoptosis can be divided into membrane flip-flop (as characterized by increased surface annexin-V binding) and exposure of nucleosomes from the nucleus displayed at the cell surface (Radic et al., 2004), followed by a gradual increase in membrane permeability (frequently measured as loss of propidium iodide exclusion). It is important to note that these are only changes to the cell that were easy to determine experimentally and do not per se represent essential stages in the process of cell death.

Binding of complement initiation molecules does not generally take place on early apoptotic cells despite the fact that the membrane has undergone flip-flop. In fact, flip-flop of membranes also takes place, in a reversible fashion, for example, on activated B cells (Dillon et al., 2001) and subsets of monocytes (Appelt et al., 2005). Although C1q can, under some circumstances, detect early changes on apoptotic cells (possibly involving nucleic acid exposure at the cell surface), several key complement inhibitors will act to control the activation of the complement cascade (Elward et al., 2005). Several lines of evidence suggest that the cell surface needs to be sufficiently remodeled for the initial complement molecules to be able to bind (Trouw et al., 2007). Purified C1q has been observed to bind directly to apoptotic cells, but it is likely that adaptor molecules, such as IgM, will also be involved in binding from whole serum. This again underscores the importance of determining which proteins and ligands are involved in this process in vivo. Due to the redundant nature of the apoptotic process, this may not be easy, but a more complete picture is now emerging for which complement pathways are activated by dying cells and which proteins associate with dying cells, making possible a clarification of the overall picture by systematic investigation of each protein. In addition, it is important to carefully consider the concentrations of reagents used for these studies. Initial studies have focused on classical pathway activation using relatively low concentrations of human serum that exclude, or underestimate, the contribution of the alternative pathway of complement.

As mentioned above, apoptosis is an important mechanism in tissue homeostasis, but it is important to realize that it also occurs in places with active inflammation. The environment in which apoptosis take place will have impact on complement activation, regulation and on the pro- or anti-inflammatory context of phagocytosis. Also the anatomical location of apoptosis and the contribution of complement on the phagocytosis of such cells will be important; an apoptotic cell present in blood will be interacting differently with complement compared to a cell present in the brain. These two aspects have not been addressed in great detail and would be relevant in the context of the potential use of complement inhibition for inflammatory conditions.