Cell death has been recognised as an important part of life for the past 130 years, and many terms have been used to refer to the various ways that tissues or cells die, but until the 1960s very little work was actually done into defining these processes and determining exactly what happens. In 1972 Kerr et al. described the characteristic morphological patterns that could be seen in a form of cell death that they named ‘apoptosis' and described as the “extensive deletion of cells with little tissue disruption”, a process complementary to mitosis in tissue homeostasis. It was discovered that specific cells always die by apoptosis, which occurs at precise times during tissue development and could be caused at other times by certain triggers, allowing the conclusion to be made that there must be some sort of cellular mechanism that causes the cells to die, so the term ‘programmed cell death' (PCD) was coined. An enormous amount of research has been done since this time and apoptosis is now very well understood, although there are still various gaps in what is known.
The terms ‘programmed cell death' and ‘apoptosis' have frequently been used interchangeably, this is incorrect though: Apoptosis is a form of PCD (known as type 1 PCD) characterised by its morphology and, more recently, the biochemical pathways which cause it. PCD refers to when a cell dies in a controlled manner to aid the survival or development of the organism, this usually occurs by apoptosis, and as such, apoptosis has easily had the greatest amount of time and money spent on it, but PCD can also occur by several other methods. These methods include, but may not be limited to, autophagy, (type 2 PCD) which involves the lysosomal degradation of cellular contents, cornification, in upper epidermis formation, pyroptosis, a caspase-1 dependant form of cell death, and anoikis, a form of apoptosis involved in the evasion of metastatic cancers through the loss of contact with the ECM. There has been a considerable amount of debate in recent years as to whether some of these processes should be considered as programmed cell death due to certain necrotic-like features and also whether autophagy is a form of PCD or a last resort for survival. All of these processes are metazoan (although some do occur in other organisms, such as autophagy in yeast), but PCD has also been seen it plants. Forms of botanical PCD have been compared to apoptosis - due to the presence of apoptotic like bodies and caspases homologues, but phagocytosis is prevented by the cell wall. They have also been compared to autophagy - due to vacuolar degradation, and oncosis (a necrotic process involving ion pump failure, cellular swelling, and lysis) because of cellular swelling, but again much less is known about plant PCD than apoptosis, although the level of knowledge is growing.
Necrosis is another term that is not always used correctly, it is not a process of death but refers to the signs that an area of tissue has died, such as blackening, inflammation, and swelling. Some forms of PCD can be described as necrotic-like because they release inflammatory substances into the surrounding tissue, unlike apoptosis which involves the careful packaging of cellular contents into apoptotic bodies for phagocytosis by nearby cells.
intrinsicApoptosis can be triggered in two ways: it can be caused by an internal signal (known as intrinsic or mitochondrial apoptosis), the most obvious example of which would be when DNA is irreparably damaged, or it can be caused by and external signal (extrinsic or death receptor apoptosis), such as through the release of cytokines from a nearby cell. Both of these methods involve many proteins and very complex interconnected pathways, which will be briefly explained but not covered in a great deal of depth. Whichever pathway is used, they both result in the cleavage and activation of a family of proteins known as caspases (cysteine aspartate proteases). Caspases begin the process of killing the cell through the cleavage of many proteins, including other caspases, resulting in a massive feedback loop, which ends with the inevitable death of the cell. This process is known as a caspase cascade.
Intrinsic apoptosis caused by DNA damage is triggered when the MRN complex detects the damage and recruits ATM, which results in the recruitment of ATR and activation of both. ATM/ATR can then activate CHK2, which activates p53, or they can activate p53 themselves. p53 is an important transcription factor central to many pathways involved in the cell cycle, cell death, survival, senescence, and others. It interacts with over 100 proteins and the human genome is believed to have hundreds of p53 binding sites, and as such is a very important protein in the control of the cell and in cancer research due to the massive implications if it becomes mutated. p53 is also able to bind to breaks in DNA, suggesting a role in damage detection. After being activated, p53 up regulates the transcription of a variety of apoptotic genes including caspases and both pro and anti-apoptotic Bcl-2 family proteins, which compete between keeping the cell alive and killing it, their activity and expression can also be affected by many other factors. Bax and Bak, two pro-apoptotic Bcl-2 family proteins form a pore in the mitochondrial outer membrane, causing it to become more permeable, a process known as MOMP (mitochondrial outer membrane permeablisation), allowing various substances normally only present in the mitochondria to be released to the cytosol including cytochrome c and Smac (also known as DIABLO). Cytochrome c binds to apaf-1 (apoptotic peptidase activating factor 1) and procaspase-9, forming an apoptosome causing caspase-9 to become active. Caspase-9 can then activate caspase-extrinsic3 and caspase-7, resulting in a cascade and apoptosis.
Extrinsic apoptosis is caused by an extracellular signal such as tumour necrosis factor α (TNFα), FASL, or TRAIL, each of these ligands and their receptors form slightly different apoptotic complexes known as DISCs (death inducing signalling complex) involving a protein known as FADD (FAS-associated death domain), which activates procaspase-8. Extracellular TNFα binding to its receptor results in the formation of two complexes: TRADD, TRAF2, RIP1, and TNF-Receptor 1 form complex I while TRADD, RIP1, FADD, TRAF2, and procaspase-8 form complex II. Under normal conditions complex I would indirectly inhibit complex II through Flip but under apoptotic conditions, it is unable to, resulting in the activation of caspase-8. FASL binds to FAS allowing FADD and procaspase-8 to bind and become activated. When TRAIL binds its receptor, it induces the binding of FADD, procaspase-8 and procaspase-10, resulting in activation and apoptosis. Caspase-8 goes on to activate caspase-3, caspase-4, caspase-7, and caspase-9, initiating the caspase cascade.
Once the caspase cascade has begun, the cell undergoes the morphologic changes associated with apoptosis first described in detail by Kerr et al. The cytoplasm begins to condense, followed by the aggregation of chromatin close to the nuclear envelope, which then breaks down. This is followed by the membrane beginning to bleb (it was suggested by Majno and Joris in 1995 that this process should be referred to as budding rather than ‘blebbing' due to the ischemic connotations of ‘blebbing'). While this is occurring, cytochrome c is released from the mitochondria (due to MOMP, which still occurs in extrinsic apoptosis but is not part of its activation) and oxidises phosphatidyl serine (PS), which is normally only found on the cytoplasmic side of the plasma membrane. As translocases are no longer functioning (due to cleavage), scramblases move oxidised PS to the extracellular side of the membrane, where it can interact with opsonins such as MFG-E8 and Gas6. The PS and opsonins act as ‘eat me' signals, causing nearby cells (both professional and non-professional phagocytes) to phagocytose the apoptotic bodies which have now formed and will each contain some of the partially degraded cellular contents.
The caspases were mostly discovered in the 1990s as mammalian homologues to the apoptotic C. elegans CED genes. They were renamed Caspases in 1996, with the most recent caspase - caspase-18 - being discovered in 2008 through genomic methods. Caspase is an abbreviation for cysteine aspartate protease; the main catalytic residue is a cysteine, which cuts after aspartate residues, although the Drosophila caspase-9 homologue DRONC is able to cleave after glutamate just as well. Only 13 of the 18 known mammalian caspases are found in humans and caspase-12 acts as a pseudogene in most people. The caspases have traditionally been split into three groups: inflammatory caspases (1, 4, 5, 11, 12, 13, 14), which are involved in cytokine maturation, apoptotic initiator or apical caspases (2, 8, 9, 10) and apoptotic effector or executioner caspases (3, 6, 7). Caspase-15 to caspase-18 are not included in these groups as there has not been a significant amount of research done into their functions yet (a PubMed search for caspase-15 yields 4 papers, all from the same lab). Despite not playing a major role in apoptosis, most of the inflammatory caspases have been linked to other forms of cell death, such as caspase-1, caspase-4, and caspase-5 in pyroptosis. So far, the only one of the first 14 caspases that has not been linked to cell death is caspase-14, which is involved in keratinocyte differentiation, it is also the only one not expressed in a wide variety of cells.
Caspases are synthesised in a zymogenic or pro-enzyme form containing three domains, an N-terminal pro domain, followed by a large domain, followed by a C-terminal small domain. The N-terminal domain is much larger in initiator caspases (219 residues in human caspase-10) than in effector caspases (23 residues in human caspases-6 and 7), and as such has extra functions. It can contain either a caspase recruitment domain (CARD - found in 1, 2, 4, 5, and 9) or a death effector domain (DED - found in 8 and 10). CARDs are involved in activation and interactions with other caspases and regulators, DEDs have been linked to activation. Caspases are activated through cleavage after aspartate residues, as this is also the method by which caspases cut their targets, it allows them to autoproteolyse/autoactivate, and to activate other caspases. Initiators are activated through autoproteolysis; they form a complex with other proteins containing two procaspases, which activate each other through cleavage between the large and small domains/subunits, followed by cleavage between the prodomain and the large domain/subunit. This creates an active tetrameric caspase, which can be seen as a ‘homodimer of heterodimers', with the active sites sitting in the large domains, near the interfaces with the small domains.
The active initiators can then activate other initiators and effectors, beginning the cascade. Effectors do not need to form a dimer or a complex to be activated; they are also present in higher concentrations than initiators, which allows a fairly weak apoptotic signal to be amplified, still resulting in apoptosis. It has been shown that caspase-3 is the main effector caspase and is necessary and sufficient for apoptosis to take place; caspase-3 knockouts have a wide range of deformities and will usually die before birth, while caspase-6 and 7 knockouts have much less severe problems.
Although caspase cascade initiation can be seen as a final ‘go' signal for apoptosis, apoptosis is not directly caused by caspases. They are fairly specific in cleaving, although they do cleave a wide range of targets. These targets include cytoskeletal proteins, nuclear lamins, proteins involved in DNA replication and repair, protein kinases, proteins involved in signal transduction and gene expression, proteins involved in the cell cycle, proteins that have a direct role in apoptosis (caspases, Bcl-2 family, DNases, proteases, and inhibitors etc.), and proteins which have been linked to diseases (sometimes explaining the pathogenicity of the disease, should the protein be cleaved incorrectly). It is believed that the majority of proteins that are degraded during apoptosis are actually cleaved by non-caspase proteases that are activated (directly or indirectly) by caspases, and that most protein cleavages/degradations that occur are not required for correct apoptotic function. As an extension of this, most of the morphological changes seen are hypothesised to be controlled by a select few proteins. However, it is largely unknown specifically which proteins control these processes. One example is that the membrane blebbing/budding has been shown as a direct result of ROCK I cleavage and hyperactivation.
As well as having their apoptotic roles, there has been an increasing amount of evidence over the past 10 years for unrelated, non-apoptotic roles for the apoptotic caspases. It has been shown that they have roles in the cell cycle, proliferation, and differentiation of many cells, but the exact nature of these roles is unclear. Schwerk et al. postulated that differentiation can occur through a form of ‘incomplete apoptosis', this may be due to only partial caspase activation, through inhibition, or through anti-apoptotic factors. Caspase-3 has been linked to lens and erythrocyte differentiation and also to thrombocyte formation. Caspase-8 has been linked to lymphocyte development as its mutation can result in impaired B, T, and NK cell function, but using traditional caspase-8 inhibitors does not inhibit their proliferation or differentiation, suggesting that this particular function is not mediated by the apoptotic active site, probably occurring through the action of another active site elsewhere on the protein. In recent years, evidence for many more functions of caspases has been seen, and it is now appears that caspases have a diverse range of functions. There are three questions surrounding the non-apoptotic functions that need answering: what non-apoptotic roles exist, what are their mechanisms, and how is it possible for the caspases to be activated without initiating a cascade, killing the cell? This dissertation will examine some examples of such roles and attempt to answer these questions.
Caspase-9 activation in myoblast differentiation
A non-apoptotic role for caspase-9 in muscle Differentiation
T Murray, J McMahon, B Howley, A Stanley, T Ritter, A Mohr, R Zwacka, H Fearnhead - J Cell Sci. 121 3786-3793 (2008)
Apoptosis and the differentiation of myoblasts have several cellular processes in common, such as the disassembly and reorganisation of actin filaments; it is therefore not much of a leap to suggest that they may have certain similarities at the molecular level and that their mechanisms may ‘overlap'. Various caspases have been shown to have roles in differentiation, but it has not yet been completely revealed how they are activated or how they perform their functions without initiating a cascade and causing apoptosis. Fernando et al. showed in 2002 that caspase-3 is present in its active form in differentiating myoblasts and that it cleaves and activates a protein essential to myotube formation. They also showed that it does not cause apoptosis to occur on a large scale. In this study, Murray et al. aimed to find out how caspase-3 is activated in myotube formation.
Myoblasts are mononucleate muscle progenitor cells that differentiate to form multinucleate myotubes; part of this process is the fusion of adjacent myoblasts. At the same time, the level of Myosin heavy chain (MyHC) expression increases significantly, it is not expressed in undifferentiated myoblasts, but is expressed in differentiated myotubes, so Murray et al. used these two changes to monitor myoblast differentiation.
Methods, Results & Conclusions
shRNABecause at least one active caspase was present in the myoblasts, it was possible that some of the cells were apoptotic and could skew the results. To ensure that this did not happen, DAPI staining was used to analyse nuclear morphology of the C2C12 myoblasts used, it was shown that at least 99.8% were not undergoing apoptosis. As caspase-3 is an executioner caspase, the authors began by assuming that it is likely to be activated by an initiator caspase, as it would be in apoptotic cells. Western blots were performed on cell extracts at various stages to determine whether any initiator caspases were present in their active form. The authors tested for caspase-9active caspases-1, caspase-2, caspase- 8, and caspase-9. Only caspase-9 ( ?) was shown to be present in its cleaved form during myoblast differentiation, suggesting that a form of the intrinsic apoptotic pathway may be used to activate caspase-3.
bcl To assess whether caspase-9 was part of the differentiation process, the level of expression was reduced using short hairpin RNA (shRNA) delivered through an adenoviral vector. Under these conditions, less than two days after transduction, the level of active caspase-9 and was seen to be around 50% lower than in normally differentiating myoblasts (image not shown). This resulted in the level of differentiation also being reduced significantly, as measured by the decreased amount of cell fusion events ( ?). The rate of expression of MyHC was not seen to decrease when caspase-9 expression was reduced, leading to the authors suggesting that this was independent of cell fusion events and caspase expression.
To test whether caspase-9 was activated by the intrinsic pathway, as it would be in apoptosis, the expression of Bcl-xL, an anti-apoptotic Bcl-2 family protein was increased, and a western blot was performed to detect any cytochrome c or Smac in the cytosol as a sign that MOMP had occurred. Bcl-xL over expression clearly reduced the number of fusion events ( ?), but no cytochrome c or Smac were detected in normally differentiating cells, suggesting that although the pathway that is used to activate caspase-9 has its similarities to the intrinsic pathway, there are also significant differences.
In this paper the authors showed that caspase-3 is activated by caspase-9 in differentiating myoblasts and that the level of differentiation can be reduced through the expression if Bcl-xL, suggesting that caspase-9 is activated in the same way that it is activated in intrinsic apoptosis, through mitochondrial outer membrane permeablisation, but MOMP was not seen to have occurred. It is therefore unclear how caspase-9 is activated, although it may be activated by a mechanism related to the mitochondrial pathway found in intrinsic apoptosis that lacks certain apoptotic features. Indeed, the absence of these features may be what prevents apoptosis from occurring. Regardless of how this occurs, it is clear from the point of view of this dissertation that non-apoptotic roles have been seen for active caspase-3 and caspase-9 with a method of activation related to their activation in intrinsic apoptosis, although it is unclear how apoptosis is prevented in this situation.
Caspase-3 in glia differentiation
Regional differences in the temporal expression of non-apoptotic caspase-3-positive Bergmann glial cells in the developing rat cerebellum
Bergman glia are specialised glial cells found in the vermis, the area between the two hemispheres of the cerebrum. The vermis contains a large amount of Purkinje cells, which are supported by Bergman glia. Radial glia in the anterior vermis differentiate into Bergman glia in the first two weeks after birth, while glia in the posterior vermis begin to differentiate towards the end of the first month after birth ( ?). The authors of this study had previously shown that differentiating neurons contained active caspase-3, and later extended this to Bergman glia. They showed that the presence of caspase-3 in these cells does not cause apoptosis, but not how it is prevented, and that inhibition of caspase-3 halted differentiation, providing direct evidence that caspase-3 has a role in the differentiation of glia. In this study, they analysed the changes in expression of caspase-3 in Bergman glia from different lobules of the vermis over the course of differentiation.
This study involved the immunostaining of paraformaldehyde (polymethanal) fixed brain slices taken from 9, 15, 21, and 30 day-old rats (n≥6 for each stage of development). The cells were viewed under a confocal microscope to determine if the Bergman glia were expressing caspase-3 and statistical analyses were performed to show how caspase-3 expression changes over the course of the first month in the early developing anterior lobe of the vermis and the intermediate to late developing posterior lobe. ? shows an image of differentiating Bergman glia and Purkinje cells typical of the images used in this experiment.
Results & Conclusions
This study found that although caspase-3 expression is seen in low levels in both the anterior and posterior vermis nine days after birth, it is present in higher levels in the early developing anterior lobules ( ?). There is a large increase in both the anterior and posterior expression by day 15, with higher posterior expression. This posterior expression stays high up to day 21, while the anterior expression drops, this aligns with the fact that differentiation in the anterior vermis will have finished by this point. By day 30, the anterior expression has dropped slightly more and the posterior expression has dropped significantly, as differentiation is well under way. Although these results do correlate with the processes of differentiation, and do show a level of correlation between caspase-3 expression and differentiation, they do not give an enormous amount of insight into this correlation. More useful results could have been obtained if more rats were used for each day, and samples were taken more frequently over a two month period. Regardless, there is a level of correlation between when caspase-3 is expressed, and when the cells differentiate, this strengthens the evidence gathered by Strahlendorf's lab that caspase-3 has a role in the differentiation of Bergman glia, although they have not attempted to show what its role in differentiation is, or how apoptosis is prevented.
The cell cycle contains checkpoints between G1 and S phase, between G2 and M phase, and between the various stages of mitosis. These checkpoints are used to ensure that the cell, and more specifically its DNA, is not damaged and to allow any repairs to take place if it is damaged, ultimately preventing tumour growth. Intrinsic apoptosis is induced if the cell is damaged beyond repair. Various mechanisms involved in the cell cycle have been found and are well understood; however, in this paper Hashimoto et al. suggest a role for several caspases, particularly caspase-7, throughout the cell cycle and during mitosis that had apparently been completely missed in previous studies.
Methods & Results
mitosisWhile studying caspase-3 in apoptotic cells Hashimoto et al. noticed that a small amount of the non-apoptotic control cells also seemed to contain active caspase-3, upon further inspection they noticed that all of the cells that contained caspase-3 were undergoing mitosis ( ?). Western blots were performed on cell lysates to confirm the presence of caspase-3 in mitotic cells, and to check for the presence of any other caspases, they also checked if any common caspase targets were being cleaved. The western blots showed that caspase-3, caspase-7, caspase-8, and caspase-9 were all present in their cleaved forms in mitotic cells ( ?) and that three targets of caspase-3 and caspase-7 were being cleaved (PARP, lamin B1, and PKCδ - image not shown).
siRNATo determine how much of an effect the caspases had on the progression of the cell cycle, cells were cultured 3+7with Z-Asp-CH2-DCB, a broad-spectrum caspase inhibitor, the cells were visualised and it was seen that the nuclei did not have normal mitotic morphology (image not shown). Each of the caspases detected in the western blots and caspase-4 for reasons unclear then 8+9individually had their expression knocked down using small interfering RNA (siRNA). The amount of cells on the second and third days after transfection relative to a specific amount selected on the first day was then calculated. Cells treated with siRNA for caspase-3, caspase-4, caspase-8, or caspase-9 showed little to no significant decrease in rates of proliferation. On the other hand, treating cells with siRNA for caspase-7 halted mitosis altogether ( ?). This major mitotic role for caspase-7 was confirmed by knocking down its expression in one cell culture with shRNA, and restoring its expression in a caspase-7-/- cell culture using plasmids. Although the shRNA did not prevent proliferation altogether, it did reduce it and a large amount of the cells died during mitosis (data not shown). The cells with restored caspase-7 showed increased levels of proliferation and decreased levels of death during mitosis, although not as large changes as the shRNA knockdowns (data not shown).
The data presented by Hashimoto et al. clearly demonstrate that several caspases can be found in their active forms throughout the cell cycle, but in higher concentrations during mitosis, and that caspase-7 is required for the normal progression of cell division. This paper particularly focuses on caspase-7 because of all the active caspases present during mitosis, caspase-7 appears to have the most vital role, as demonstrated in the caspase-7 knockdowns, which were unable to proceed through normal mitosis, and the restored caspases-7-/- cells, whose proliferative ability was improved. The authors suggested that the cells transfected with shRNA and the caspase-7-/- cells did not completely lose their ability to divide because other caspases were able to compensate for the loss to a certain extent, something that has been noticed with caspases in apoptosis.
Interestingly, knock down of the initiators caspase-8 and caspase-9 did not have a significant effect on proliferation, so it can be assumed that it also did not have an effect on caspase-7 expression, suggesting that caspase-7 is activated by either a caspase-8 or caspase-9 independent form of an already known pathway, or by an entirely different pathway. It is also unclear what the roles of the caspases are during mitosis, although three cleaved caspase targets were detected, this does not constitute a mechanism. This paper did not attempt to explain how the caspase cascade and apoptosis are prevented, but they did briefly mention that several recent papers have suggested that IAPs (Inhibitors of apoptotic activity) may function as caspase inhibitors and as E3 ubiquitin ligases. What they did do though was to clearly show a novel role for caspases in a process whose mechanisms had previously been thought of as largely solved. This opens the possibility that other cellular processes that are currently believed to be more or less completely understood may have mechanisms that are not yet known about, possibly involving apoptotic caspases in non-apoptotic roles.
NHE1 is a sodium potassium antiporter that is involved in the regulation of cellular pH. Its expression had previously been shown to be reduced under oxidative stress, specifically in the presence of hydrogen peroxide. This reduced expression contributed to an overall increased cellular sensitivity to apoptotic stimuli. The authors of this paper examined the role of caspases in the down regulation of NHE1 gene transcription under oxidising conditions, although they do not state why they thought caspases may have such a role, it is apparent from their findings that they do.
Kumar et al. began by confirming that non-lethal levels hydrogen peroxide reduce the level of expression of NHE1, this was shown using various methods based around using rat L6 muscle cells transfected with the mouse NHE1 promoter attached to the firefly luciferase gene (termed L6 1.1kb cells). Luciferase activity was measured both with and without H2O2 and the levels of expression were compared. This method was then repeated using z-VAD-fmk, a broad-spectrum caspase inhibitor, to show that caspase inhibition nullified the NHE1 down regulating effect of H2O2. To find out which caspases were involved in the prevention of NHE1 expression, the same experiment was performed again using caspase inhibitors specific to caspase-3 (z-DEVD-fmk), caspase-6 (z-VEID-fmk), caspase-8 (z-IETD-fmk), and caspase-9 (z-LEHD-fmk). siRNA was used to knock down expression, confirming the involvement of specific caspases and western blots were performed to confirm which caspases were present in their active forms, and show that the siRNA knocked down their expression. A recent paper showed that induction of apoptosis under oxidising conditions involved redox cycling of iron, to examine whether iron played any role in this situation, DFO - an iron chelating agent, and FeCl3 were used in similar experiments.
Results & Conclusions
The first luciferase reporter gene assay showed that treating the cells with increasing concentrations of H2O2 proportionally reduced NHE1 expression ( ?), confirming the previously seen effect. The addition of a broad-spectrum caspase inhibitor partially reversed this effect, restoring NHE1 expression ( ?), although only one concentration was used; using several concentrations of the inhibitor would have allowed a comparison between the kinetics of this effect and caspase inhibition, which would have confirmed beyond any doubt that caspases were being inhibited. When specific caspases inhibitors were used, inhibiting caspase-8 or caspase-9 had no effect, but inhibition of caspase-3 partially restored normal expression, while caspase-6 inhibition restored it completely ( ?). The involvement of caspase-3 and caspase-6 was confirmed by the western blots and by the siRNA, which had the same effect as inhibitors (data not shown). Addition of DFO increased the level of expression of NHE1, and addition of FeCl3 increased caspase activity, indicating that caspases are activated by iron in low, non-lethal oxidising conditions as well as lethal ones.
The NHE1 promoter region contains an AP-2 binding site. AP-2 proteins are transcription factors that are known targets for cleavage by caspases, so this is likely to be the mechanism by which caspase-3 and caspase-6 down regulate NHE1 expression under oxidising conditions. This also suggests that caspases would also down regulate any other genes with AP-2 binding sites in the promoter region in the presence of H2O2, although it is unknown how H2O2 activates caspases. The lack of an effect of caspase-8 and caspase-9 inhibition on the expression of NHE1 implies that the activation of caspase-3 and caspase-6 in this response is independent of normal apoptotic pathways. The confirmation that the activation of caspases is mediated by iron suggests that they are activated in a similar way to how they are activated in lethal oxidising conditions, which is currently unknown.
This paper shows that caspases are activated and are able to regulate gene expression in non-apoptotic conditions, a role very different to the now fairly frequently seen, although not well characterised roles in differentiation and proliferation. It also shows that they are activated by an alternative, currently unknown pathway. The paper does not suggest how their activity is controlled, or how a cascade is prevented.
This paper is interesting because, like the previous paper, it analyses a role for caspase-8 that is not linked to proliferation and differentiation, but also because it appears that a second active site is used, meaning the process is completely unrelated at a molecular level. In this role, caspase-8 is functional in its zymogenic form. In a previous study, procaspase-8 was shown to play a role in the activation of calpains and rac in the assembly of lamellipodia, and accordingly, caspase-8 knockouts had impaired cell motility. The authors of this study had previously shown that caspase-8 was able to form an apoptosis-regulating complex with integrins, and that epidermal growth factor was able to cause the Src-mediated phosphorylation of caspase-8 on tyrosine 380, a modification that inhibited its apoptotic activity. This study aimed to investigate whether the phosphorylation of procaspase-8 is linked to its role in cell motility and whether the phosphorylation by Src is a part of the formation of a motility-regulating complex.
Three assays were used in this study, in the first, caspase-8-/- NB7 cells were transfected with GFP tagged caspase-8, GFP tagged mutant Y380F caspase-8, or left as they were. Cells expressing the transfected caspase-8 were selected and this expression was confirmed through western blotting. The cells were attached to coverslips with fibronectin and shortly thereafter, ‘wounded' with a pipette tip. After being allowed to heal for a short period of time, some the cells were fixed and stained using antibodies so that localisation of caspase-8 could be visualised, others were viewed live.
In the second, the same cells were allowed to form monolayers with confluent densities on dishes coated with fibronectin. The monolayers were given a 1500µm wound, which was allowed to heal. After 20 hours, the extent of wound closure was measured and averages were calculated. In the third assay GST pull downs of cell extracts from human embryonic kidney cells were performed using GST bound to the SH2 domain of Src to see if a complex formed between it and procaspase-8, cell extracts containing both wild type and Y380F mutant caspase-8 were used. The methods used were clearly explained and enough repeats (n=30) were used in the migration assay to eliminate the possibility that these results may have occurred by random chance.
Results & Conclusions
The results of this study showed that wild type procaspase-8 is recruited to the membrane of non-migrating cells spontaneously to allow them to connect to the medium ( ?A). This localisation is dependent on the phosphorylation of tyrosine 380 ( ?B), a post-translational modification that, as well as activating this function, inhibits the apoptotic function of the protein. In migrating cells procaspase-8 is recruited to the leading edge of the cell to assist in lamellipodia formation ( ?C). It was also shown through a C360A mutation that cysteine 360, the main active site residue, is not required for recruitment of procaspases-8 to the membrane (image not shown). Strangely, images of wounded Y380F mutant cells were not included in the paper, however the method indicates that such images were taken and the discussion suggests that they showed no mutant procaspase-8 recruitment to the periphery. Dr Stupack was unclear on the matter when asked for further information.
The study also showed that the inability of Y380F mutant procaspse-8 to be phosphorylated hindered the motility of the cells, presumably because the mutation prevents the caspase-8 from performing its function. The mutation did not render the cells completely immobile, but did significantly reduce their motility to something similar to that of caspases-8 knockouts ( ?).
Finally, it was shown that Src does form a complex with procaspase-8 confirming a direct interaction ( ?). This complex only forms readily at low levels though, as shown by the faint band, so the authors stated that motility related functions of procaspase-8 probably do not just rely on the interaction with Src, and that further interactions need to be found. Senft et al. performed a related study at the same time, and found that procaspase-8 phosphorylated at tyrosine 380 regulates phosphatidylinositol-3 kinase activity, which controls cell motility, although they were unable to produce exact mechanisms.
This paper clearly demonstrates that zymogenic procaspase-8 has a role entirely unrelated to the apoptotic role of the cleaved form. Presumably, due to the large polar nature of a phosphate group, cleavage and autoactivation through the extrinsic pathway, and therefore caspases cascade initiation and apoptosis will be prevented by phosphorylation of the protein. This phosphorylation also activates its alternative role in cell motility, which is yet to be completely described.
The papers analyzed in this dissertation show that the apoptotic caspases, which were previously thought to only be involved in the induction of a caspase cascade and apoptosis, are also required for the correct execution of a wide variety of cellular processes. Most of the apoptotic caspases have been linked to the differentiation or proliferation of at least one form of tissue, but recently caspases have been linked to many other cellular processes, including gene regulation and cell motility as previously discussed. Evidence of non-apoptotic functions has so far been found for all of the apoptotic caspases apart from caspase-10.
In three of the papers discussed, the caspase(s) involved appear to be activated by significantly different mechanisms to their mechanism of activation in apoptosis. Murray et al. showed that caspase-3 is activated in myoblast differentiation by caspase-9, as it normally would be in an apoptotic cell. They also showed that Bcl-xl, which inhibits MOMP and the release of cytochrome c, decreases caspase-9 expression, preventing differentiation. However, MOMP was not seen to occur, suggesting that although caspase-9 may be activated in a way similar to its activation in apoptosis, there are definite differences. Hashimoto et al. showed that while initiator caspases are present in their cleaved forms during the cell cycle, down regulating their expression does not prevent caspase-7 activation or mitosis. Suggesting that caspase-7 is activated by an entirely different, currently unknown mechanism. Similarly, Kumar et al. showed that inhibition of initiator caspases did not prevent caspase-3 or caspase-6 mediated gene regulation.
As well as those differences between non-apoptotic activation and apoptotic activation, Murray et al., Hashimoto et al., and Kumar et al. all showed that the amount of active caspase during their particular non-apoptotic process was lower than it would normally in apoptosis. The differences between the methods of activation and the level of active caspase could be possible reasons why apoptosis is not induced.
The first four papers discussed looked at alternative functions of cleaved caspases. Barbero et al. looked at a function of caspase-8 in its un-cleaved proenzyme or zymogen form. Because caspase-8 does not need cleaving to become active in this situation, normal apoptotic methods of activation do not apply. Instead, they showed that it is phosphorylated and that the phosphorylation is required for its function in cell motility and probably prevents normal apoptotic cleavage. This discovery of a role for caspase-8 in its proenzyme form leads to questions that should be answered in future studies: Do any other caspases perform functions in their proenzyme forms? If so, are their methods of activation related in any way to the method of activation in this situation?