Apoptosis is a form of programmed cell death involving a chain of biochemical events characterized by distinct morphological changes in the cell including cell shrinkage, pyknosis, plasma membrane blebbing and finally karyorrhexis and formation of apoptotic bodies (Kerr et al., 1972). The apoptotic bodies consist of the cytoplasma and tightly packed organelles enclosed within an intact plasma membrane, and are cleared from the system through phagocytosis by macrophages, paranchymal cells or neoplastic cells. During this entire process, there is no release of cellular contents into the surrounding tissue. Moreover, the phagocytes release anti-inflammatory factors upon interaction with the apoptotic cells, resulting in effectively no inflammatory reaction (Fadok et al., 1998).
Apoptosis is mainly induced through two pathways – the extrinsic or the death receptor pathway and the intrinsic or the mitochondrial pathway. A third route involves cytotoxic T-cell-mediated apoptosis initiated by perforin/granzyme (Fig. 1). Effectors in the different pathways of apoptosis are proteases of the caspase (cysteine-dependent aspartate-specific proteases) family. Caspases are synthesized as inactive zymogens in the cells and are activated either by autocatalytic processing initiated by adaptor-protein-mediated aggregation or by proteolysis by other active caspases (Thornberry & Lazebnik, 1998). Once the caspase cascade is initiated, the cell is irreversibly committed to apoptotic death. The activation of caspases is therefore securely regulated in the cell. The inhibitor of apoptosis proteins (IAPs) can directly inhibit the activity of subsets of caspases and as such play an important role in the regulation of both the intrinsic and the extrinsic pathways (Deveraux et al., 1997). The IAPs are characterized by the baculovirus IAP repeat (BIR) domain and the RING domain. The BIR domain is responsible for the direct binding of IAPs to caspases and confers an antiapoptotic ability to the IAPs (Salvesen & Duckett, 2002; Vaux & Silke, 2005). The IAPs have been shown to occur in high molecular complexes that enhance the stability and activity of the individual IAP proteins in the cells (Dohi et al., 2004; Rajalingam et al., 2006).
Mitochondria play a central role in apoptosis regulation. Besides being essential for the intrinsic pathway, mitochondria are also involved in apoptosis induction by the extrinsic pathway in the so-called ‘Type II cells,’ which require signal amplification by the mitochondrial pathway.
If the mitochondrial membrane becomes permeabilized, apoptogenic factors – cytochrome c, Smac and AIF – are released into the cytosol, where they can initiate the caspase activation cascade. Mitochondrial membrane permeabilization is under the control of members of the Bcl-2 family of proteins, characterized by the presence of at least one of the four conserved Bcl-2 homology domains, BH1-4 (Adams & Cory, 1998). A subset of this family, the BH3-only proteins, acts as a damage sensor in the cells, becoming activated in response to specific apoptotic stimuli (Huang & Strasser, 2000). Activated BH3-only proteins can activate the proapoptotic Bax and/or Bak, eventually leading to mitochondrial permeabilization (Wei et al., 2001). The entire process is tightly regulated by another subset of the Bcl-2 family, the antiapoptotic proteins Bcl-2, Mcl-1, Bcl-XL, Bcl-w and A1 (Cory & Adams, 2002).
It is believed that mitochondrial permeabilization and apoptosis induction are determined by the levels of the anti- and proapoptotic members of the Bcl-2 family. Overexpression of the proapoptotic BH3-only proteins can induce apoptosis in cells. On the other hand, overexpression of the antiapoptotic members such as Bcl-2 can strongly block the induction of apoptosis.
Apoptosis plays an important role in embryogenesis, development of the nervous and immune systems and tissue homeostasis. In addition, it represents a vital defence mechanism against damaged or infected cells. Cytotoxic T lymphocytes can kill tumour cells or virus-infected cells by inducing apoptosis (Russell & Ley, 2002). Apoptosis is induced in infected cells and the resulting apoptotic bodies are phagocytosed, thus preventing the spread of infection to the neighbouring cells. For example, it has been shown that the programmed cell death of Mycobacterium avium-infected cells is an important defence mechanism, which acts by sequestering Mycobacteria and aids in their killing by activation of newly recruited macrophages (Fratazzi et al., 1997).
In an excellent work carried out by Schaible et al. (2003), it was shown that Mycobacteria-infected antigen-presenting cells (APCs) undergo apoptosis and the resulting apoptotic bodies carry mycobacterial antigens to the uninfected bystander APCs. Apoptosis of the infected cells thus represents an effective mechanism for the activation of CD8 T cells with antigens that might normally remain hidden inside the phagosome.
Resistance to apoptosis in the infected cells is therefore important for pathogens, especially those residing inside the cell, to evade the host immune response.
Chlamydia and apoptosis
Several parasitic pathogens, including viruses and bacteria, modulate host cell apoptosis to escape from the host cell immune response and prolong their stay in the host. Bacteria such as Salmonella, Shigella and Yersinia spp., defend themselves from host macrophages by inducing apoptosis (Hilbi et al., 1997). Others like Helicobacter, Neisseria, Staphylococcus and Listeria spp. induce cell stress by bacterial toxins or effector proteins, resulting in apoptosis in the host cell (Guzman et al., 1996; Weinrauch & Zychlinsky, 1999; Galmiche et al., 2000; Muller et al., 2000).
In contrast to these, obligate intracellular bacteria such as Rickettsia and Chlamydia confer resistance to the host cell against a variety of apoptotic stimuli, including the action of cytotoxic T cells (Clifton et al., 1998; Fan et al., 1998). The bacteria are thus able to maintain a long-term relationship inside the host cell. In case of Chlamydia, this could help the bacteria to complete their replication cycle inside the cell, at the end of which numerous elementary bodies are produced and released to infect other cells. Chlamydia can sometimes develop long-term chronic infections that could contribute to diseases such as atherosclerosis. Protection of the host cell against apoptosis would acquire greater significance for the bacteria under such conditions. It is not surprising, therefore, that persistently infected cells have also been shown to be potently resistant to various forms of apoptosis (Dean & Powers, 2001).
Although apoptosis inhibition by Chlamydia has been widely reported, there have also been reports of induction of apoptosis by Chlamydia (Gibellini et al., 1998; Ojcius et al., 1998; Byrne & Ojcius, 2004). However, subsequent studies have confirmed that Chlamydia has a predominantly antiapoptotic effect and that the ‘Chlamydia-induced cell death’ is different from apoptosis, although some of the morphological features may be similar (Greene et al., 2004; Ying et al., 2006; Paschen et al., 2008). In the background of the block in the apoptotic pathways during infection, this cytotoxicity might be the result of cell stress and damage mediated by the host innate immune reaction or by the bacteria themselves. The presence of a bacterial inclusion in the cell would inevitably cause damage and stress in the cell, and it is not surprising that some cells die under the circumstances – the cell is after all still mortal, and although it cannot be forced to commit suicide, it can still be murdered!
Chlamydia has evolved various strategies to survive intracellularly, whereby it not only prevents the detection of the host cell by the immune system but also protects it from destruction by the immune system in case of recognition. The bacteria can downregulate expression of major histocompatibility complex class I and II molecules, thus avoiding the recognition of the host cell by CD+ T cells (Zhong et al., 1999, 2000). It has also been reported that Chlamydia-infected macrophages can induce apoptosis in the neighbouring T cells in an in vitro system by secretion of tumour necrosis factor (TNF)-α (Jendro et al., 2004). T cells have been shown to play an important role in recognizing Chlamydia-infected cells for clearance, and therefore killing of T cells might help the bacteria in establishing a chronic infection. This might also explain the diminished CD8+ T-cell recall response in Chlamydia-infected cells compared with other infections (Loomis & Starnbach, 2006).
Chlamydia further protects the host cell against a cytotoxic immune response, by its remarkable ability of making the cell strongly resistant to apoptosis. Apoptosis resistance has been attributed to all major chlamydial species viz. Chlamydophila pneumoniae, Chlamydia trachomatis, Chlamydophila psittaci, Chlamydophila caviae, as well as Chlamydia muridarum (Fan et al., 1998; Coutinho-Silva et al., 2001; Rajalingam et al., 2001; Zhong et al., 2006). Of these, C. trachomatis and C. pneumoniae have been widely studied for their strong antiapoptotic effect and relevance to human diseases.
The chlamydial infection protects cells against different forms of apoptosis described above (extrinsic, intrinsic and granzyme B), mediated by various stimuli including staurosporine, TNF-α, etoposide, granzyme B/perforin, and UV light (Fan et al., 1998; Fischer et al., 2001; Rajalingam et al., 2001). Initial work characterizing the apoptosis resistance showed that only the inclusion-carrying cells were protected, in contrast to any uninfected neighbouring cells, suggesting that the effect was exerted by the intracellular bacteria and not through a factor secreted into the medium (Rajalingam et al., 2001). Further, chlamydial protein synthesis was essential for the effect because treating the cells with the prokaryotic transcription inhibitor rifampin or the translation inhibitor chloramphenicol abrogated the apoptosis inhibition (Fan et al., 1998). It was also reported that because Chlamydia could block apoptosis even in the presence of the eukaryotic translation inhibitor cycloheximide, host cell synthesis was not required for apoptosis inhibition. However, recent reports have shown that a block in the upregulation of certain antiapoptotic proteins in the host during infection sensitizes the cell to apoptosis (Paland et al., 2006; Rajalingam et al., 2006, 2008) (discussed in detail later). In light of these studies, the efficiency of cycloheximide to block eukaryotic protein synthesis completely is questioned.
Analysis of the apoptotic pathway in the infected cell showed that there was a block in the release of cytochrome c from the mitochondria into the cytosol. Cytochrome c is required for activation of the procaspase-9 by the apoptosome formation. Caspase-9 eventually cleaves and activates caspase-3. As expected, in the infected cells, there was no processing/activation of caspase-9 and caspase-3 (Fan et al., 1998). This implied that there was a block in the mitochondrial permeabilization, which in turn is controlled by Bax and Bak. Later studies showed that the chlamydial-infected cells failed to achieve activation of these regulators of mitochondrial permeabilization (Fischer et al., 2004b; Xiao et al., 2004). It was further shown that the infected cells are not resistant to apoptosis in Type I cells, which do not require the mitochondrial pathway for activation of the effector caspases (Fischer et al., 2004a). From these data, it became evident that Chlamydia confers a block at the mitochondrial level. This would also explain the resistance to apoptosis initiated through different pathways that collude at the mitochondria.
The following question now remained: how do Chlamydia achieve this block in the pathway upstream of the mitochondria?
It has been reported that during chlamydial infection, there is a broad-scale degradation of the different BH3-only proteins in the cell. The mRNA levels of these proteins were not affected, indicating that the downregulation was at the protein level. The chlamydial protease-like activity factor was shown to be responsible for targeting those active proteins with an exposed BH3 domain, for degradation (Fischer et al., 2004b; Dong et al., 2005; Ying et al., 2005; Paschen et al., 2008), implying that Chlamydia destroy most, if not all, active BH3-only proteins. In the absence of the BH3-only proteins, death signals cannot be transmitted to the mitochondria, and this could account for the block in apoptosis upstream of the mitochondria.
Recently, our group observed that there was no noticeable degradation of the BH3-only proteins during C. trachomatis infection of HeLa cells (Rajalingam et al., 2008). However, the infected cells were still able to strongly resist apoptosis induced by various stimuli including TNF-α, staurosporine and granzyme B. The cells were found to be sensitized to apoptosis on being treated with mitogen-activated protein kinase (MAPK) inhibitors or silencing of some antiapoptotic genes in the cell (discussed in detail below). These contradictory results could have been because of differences in the experimental conditions (different cell types, Chlamydia serovars, etc.) However, it becomes clear that the degradation of the BH3-only proteins cannot be the only mechanism used by Chlamydia to achieve the strong and widespread inhibition to apoptosis induced by various stimuli.
In an elegant experiment analysing the extent of apoptosis inhibition in the infected cells, it was observed that the cytosolic extract from chlamydial-infected cells resisted the activation of caspase-3 even on treatment with cytochrome c (Fischer et al., 2001). This indicated that Chlamydia-infected cells block apoptosis downstream of the mitochondria as well, which would prevent the activation of caspases in spite of cytochrome c release. Evidently, Chlamydia interfere with the host apoptotic machinery at different levels – not only at the mitochondria but also downstream of it.
Genetic studies in the infected cells showed that Chlamydia indeed widely interfere with the host protein synthesis. Infection with C. trachomatis led to the upregulation of some genes including certain antiapoptotic genes (Hess et al., 2001; Xia et al., 2003). Prominent among the antiapoptotic proteins transcriptionally upregulated by Chlamydia was cIAP-2, a member of the IAP family. The significance of this antiapoptotic protein in conferring apoptosis resistance to the infected cells became clear, when it was shown that its silencing with RNAi technology sensitized the infected cells to apoptosis (Rajalingam et al., 2006). It was also observed that although their levels were not increased upon infection, the silencing of XIAP and cIAP-1 also sensitized the host cells to apoptosis. This gives credence to the growing opinion that the IAPs work together in a concerted mechanism in a complex to block the activation of caspase-3. Chlamydial infection stabilized the IAP–IAP complexes. Silencing of any of the components of the complex – XIAP, cIAP-1 or cIAP-2 – caused disruption of the complex, which in turn affected the stability of the remaining components. In this scenario, the IAP-mediated block on caspase-3 was abrogated, and the infected cells could undergo apoptosis.
During apoptosis induction in an uninfected cell, the inhibition of caspase-3 processing and activation by IAPs is countered by the release of Smac from the mitochondria (Rajalingam et al., 2007). In the presence of Chlamydia, there is no mitochondrial outer membrane permeabilization and therefore Smac is not released into the cytosol. Indeed, the incomplete processing of caspase-3 observed in the infected cells upon apoptosis induction is similar to that seen after apoptosis induction in cells with Smac knockdown (Rajalingam et al., 2006, 2007).
To identify the host factors required by C. trachomatis to confer this strong block in mitochondrial permeabilization, an RNAi screen was carried out by our group (unpublished data). Prominent among the targets whose ablation led to the sensitization of infected cells to apoptosis by TNF-α was the antiapoptotic Bcl-2 member Mcl-1. Chlamydial infection leads to the activation of the Raf/MEK/ERK and the PI3/AKT survival pathways in the host cells by an as yet unknown mechanism (Su et al., 2004). It was observed that the activation of the Raf/MEK/ERK pathway was required for the upregulation of Mcl-1 during infection. Further, the activation of the PI3K/AKT pathway helped to stabilize and accumulate Mcl-1 in the cell to high amounts, which could effectively block the release of Smac from the mitochondria. Inhibition of the MAPK pathways with chemical inhibitors could stop the increase in Mcl-1 levels, overcoming the block in the release of Smac, and sensitized the cell to apoptosis during the early stages of infection. The role of MAPK pathways in Chlamydia mediated resistance turned out to be even more crucial, when it was observed that these pathways were also required for the infection-induced upregulation of cIAP-2 protein. Interestingly, Mcl-1 depletion failed to sensitize the cells in the late stages of infection (around 48 h in case of C. trachomatis), suggesting that different mechanisms were at play at different stages of infection (Rajalingam et al., 2008).
The activation of the PI3K pathway in the infected cells has also been shown to sequester the BH3-only protein BAD away from the mitochondria. It was seen that activation of the PI3K pathway led to the phosphorylation of BAD, after which it was recruited at the surface of the chlamydial inclusion (Verbeke et al., 2006). This would further increase the resistance to those apoptosis inducers (e.g. staursporine) that rely on BAD for activating the mitochondrial pathway.
From these observations, it appears that during early stages of infection, chlamydial infection leads to a MAPK-dependent increase of the antiapoptotic proteins Mcl-1 and cIAP-2. Mcl-1 upregulation inhibits mitochondrial outer membrane permeabilization, thus blocking the release of Smac and cytochrome c into the cytosol. cIAP-2 upregulation leads to the stabilization of the IAP–IAP complex, which would directly block the activation of caspase-3. The role of Mcl-1 in sustaining bacterial infection has also been shown for Helicobacter pylori infections, indicating that the activation of MAPK pathways for Mcl-1 upregulation could be a more common strategy for resisting host apoptotic pathways during bacterial infection (Mimuro et al., 2007).