Caspases and Their Substrates

KILLER PROTEASES

Apoptosis is orchestrated by a set of proteases, called caspases, which reside in an inactive form in nearly all of our cells. When activated, some caspases cleave hundreds, or perhaps thousands, of distinct target proteins. Such cleavage leads to all of the features that characterize this type of cell death, and apoptosis depends on the functions of caspases. In this review, we introduce these enzymes and the substrates that they cut to bring about apoptosis.

Caspases are endopeptidases—that is, they cut proteins internally rather than nibble away at the ends of the proteins. Unlike digestive proteases such as trypsin, caspases do not degrade their substrates but, rather, they clip them at discrete sites. In general, they cut specific sequences that end in aspartate residues and cut immediately after this amino acid (this is the “asp” in the term caspase). However, caspases do not cut after every aspartic acid that can be accessed in a protein—they have a preference for certain sequences.

In caspases, the active site of the enzyme includes a cysteine, which therefore classifies caspases as “cysteine proteases” (this is the source of the “c” in the word caspase; hence, a cysteine protease that cuts proteins after aspartic acid residues).

HOW CASPASES CUT PROTEINS

Cysteine proteases, also called thiol proteases, are one of the four major types of proteases characterized by their active sites (the others are serine, aspartyl, and zinc proteases). Although the focus here is on how caspases work, there are similarities among all proteases.

For a caspase to work, the interaction with its protein substrate must be “fast-on–fast-off.” This involves the enzyme briefly holding the target peptide bond at the active site. In a functionally active caspase, a substrate specificity pocket is close to the active cysteine, which is itself near a histidine. This cysteine–histidine dyad is where the action takes place. A crucial residue in the substrate pocket is an arginine, which holds the target aspartate in the substrate in position. This results in the situation shown in Figure 1, and the reaction follows. This step is called acylation; the next step is deacylation, shown in Figure 2. A water molecule is sacrificed, the peptide bond after aspartate is cut, and the caspase is now ready to cut another substrate protein.

Figure 1.

The acylation step of caspase cleavage.

Figure 2.

The deacylation step of caspase cleavage. The mechanism is based on the function of other cysteine proteases and has not been confirmed in caspases.

For a caspase to function, the catalytic dyad, cysteine–histidine, must be brought close to a target protein. This is the function of the substrate specificity pocket in the caspase. That pocket, with its arginine, only permits aspartates (and, to a much weaker extent, glutamate, in some cases) to gain proximity to the dyad. Inactive caspases (“procaspases”), however, do not allow access to the catalytic dyad. To form the pocket, changes to the procaspase must occur to form an active site. How this occurs depends on the type of caspase.

TYPES OF CASPASES

Caspases, like apoptosis, are found only in animals,¹ and, in most animals, there are several different caspases. In general, types of caspases can be distinguished by (1) their functions, (2) the structure of the procaspase, and (3) how they are activated. These are often interrelated, and the distinctions are not absolute.

In vertebrates and probably most other animals, at least two types of caspases are involved in apoptosis (the notable exception is nematodes, as discussed below): executioner caspases (such as in mammals, caspases-3, -6, and -7) and initiator caspases (in mammals, caspases-8 and -9). Initiator caspases are sometimes referred to as “apical” caspases. Another type of caspase, related to initiator caspases, includes inflammatory caspases (in humans, caspases-1, -4, and -5; in rodents, caspases-1 and -11). Additional caspases (in mammals, caspases-2, -10, -12, and -14, among others) are more difficult to place into one of these categories; their functions are less well understood.

All procaspases have three regions, termed the prodomain, large subunit, and small subunit. The latter two are always separated by one or more sites, where they are cleaved by the action of the caspase itself (self-cleavage), as it becomes activated, or by another protease (usually another caspase, but, as we will see in Green 2022a, at least one other protease can do this). Schematic examples of the domain structure of several human caspases are shown in Figure 3.

Figure 3.

Schematics of several human caspases. Caspases-2 and -10 are grouped with the initiator caspases, although their classification is problematic. The representation aligns related sequences and should not be taken to indicate the actual structures of the proteins.

The prodomain and large subunit are often separated by a cleavage site targeted by the caspase itself. The functions of these cleavage sites vary in different caspases, as discussed in Green (2022a). For now, it is sufficient to be aware that there are different types of caspases with different biochemistries and functions. Those that concern us in this review are primarily executioner caspases, which, in vertebrates, are caspases-3, -6, and -7. These caspases become active during apoptosis and cleave substrates that bring about the characteristic features of apoptosis.

CASPASES IN OTHER ANIMALS

Nearly all animals for which we have genomic information appear to have caspases, and, in some invertebrate animals, we know that executioner caspases are important for apoptosis. As mentioned in the Introduction, the pathways of apoptosis in two invertebrates, namely, the nematode Caenorhabditis elegans and the fruitfly Drosophila melanogaster, have been characterized in detail.

In C. elegans, only one caspase, CED3, is important in apoptosis, and CED3 has features of an initiator caspase (e.g., a long prodomain), but it also acts as an executioner caspase, orchestrating apoptosis in the dying cells (Fig. 4).

Figure 4.

The linear organization of CED-3. CARD, caspase recruitment domain. The arrows are auto-cleavage sites.

In flies, several different types of caspases have been identified, and these are shown in Figure 5. Of these, DCP1 and Drice appear to be the most important executioner caspases. Two other caspases, Decay and Damm, might be executioner caspases, based on their sequences, but this is speculation.

Figure 5.

Caspases in Drosophila.

In general, executioner caspases identified in other animals have essentially the same specificities as executioner caspases in vertebrates. In the discussion that follows, we focus on substrates of mammalian executioner caspases, which are the best characterized.

CASPASES SHOW PREFERENCES FOR THEIR SUBSTRATES

Caspase-mediated cleavage events can have catastrophic consequences that ensure not only that the cell dies, but that it does so quickly and “cleanly” (without inducing an inflammatory response). To understand this, we must discuss the sequences that executioner caspases recognize in proteins. And, for this, we need to become familiar with a bit of terminology.

We know that most caspases prefer to cut their substrate proteins just after aspartic acid residues, and this target amino acid after which the cut occurs is called the P1 residue of the substrate. The amino acid that follows the cut is called P1′. The residue just before P1 is P2, preceded by P3, P4, and so on (Fig. 6).

Figure 6.

Numbering scheme for amino acids around a site in a substrate protein that can be cut by caspases.

In nearly all caspase substrates, P1 is obviously aspartate, although very rarely it is glutamate. However, the presence of an aspartate in a substrate is not enough to predict that a caspase will cut at that site. First, the aspartate must be exposed so that the enzyme can access it, and, second, other amino acids contribute to recognition by the caspase. Based on the use of peptide libraries, caspases-3 and -7 share preferences for sequences that contain DXXD/G (or S or A), in which “X” is any amino acid, and “/” indicates the cleavage between P1 and P1′. Other caspases display different preferences, although D (aspartate) is generally preferred at P1. Some of these preferences² are shown in Figure 9.

Figure 9.

Caspase preferences, based on peptides. Sequences shown are optimally cleaved, although other peptide sequences can be cleaved nearly as well.

How do these preferences apply to actual cellular substrates for caspases? We can identify protein substrates for the caspases by adding active caspases to cell extracts and looking for cleavage events. In this way, hundreds of substrates for caspase-3 have been identified, and investigators believe there could be a thousand or more. Some of these are listed in Table 1.

Table 1.

Examples of caspase substrates identified in apoptotic cells

Substrate	Putative or demonstrated functional consequence of cleavage	Cleavage site (D residue number)
Acinus	Involved in chromatin condensation	DELD (1093)
AKT/PKB	Loss of kinase activity. Putative—loss of survival signaling	ECVD (462), TVAD (108), EEMD (119)
ATP11C	Loss of phospholipid flippase activity	QEVD (439), SQTD (445), DAVD (481)
β-Catenin	Reduced α-catenin binding. Putative—loss of cell adhesion	YQDD (145), NDED (164), SYLD (32), ADID (83), TQFD (115), YPVD (751)
c-IAP1	Loss of signaling functions	ENAD (372)
E-Cadherin	Release of intracellular fragment	DTRD (750)
GATA-1	Loss of transcriptional activity, which leads to impaired erythropoiesis	EDLD (125)
Gelsolin	Loss of binding to monomeric actin and triggering of F-actin depolymerization, membrane blebbing	DQTD (403)
iCAD (DFF45)	Release of active CAD endonuclease	DETD (117), DAVD (224)
IL-33	Putative—activation of IL-33, which sensitizes toward a TH2 immune response	DGVD (178)
iPLA2	Increased phospholipid turnover; release of lysophosphatidylcholine (LPC), which attracts monocytic cells	DVTD (183)
Lamin A/C	Breakdown of nuclear envelope	VEID (230)
Lamin B1	Disassembly of nuclear lamina	VEID (231)
Stk4/Mst1	Kinase constitutively active; overexpression of amino-terminal fragment induces apoptotic morphology	DEMD (326)
NDUFS1	Disruption of electron transport (complex I) and transmembrane potential (ΔΨm), leading to production of reactive oxygen species (ROS), loss of ATP production, and mitochondrial damage	DVMD (255)
PAK-2	Kinase constitutively active and activates c-Jun amino-terminal kinase pathway; overexpression of cleaved fragment leads to apoptotic morphology (shrinkage and rounding up)	SHVD (212)
PARP-1	Loss of poly(ADP-ribose) polymerase activity	DEVD (214)
Procaspase-3	Activates protease activity	ESMD (28), IETD (175)
Procaspase-7	Activates protease activity	DSVD (23), IQAD (198)
RIPK-1	Inhibits activation of NF-κB	LQLD (324)
ROCK1	Kinase constitutively active—drives cell contraction and blebbing. Phosphorylates PTEN, which then inhibits Akt/PKB serine/threonine-protein kinase	DETD (1113)
TRAF-1	Inhibits activation of NF-κB	LEVD (163)
Vimentin	Disrupts intermediate filaments	DSVD (85)
Xkr8	Scrambles the lipids of the plasma membrane	DGVD (355)
XIAP	Amino-terminal fragments inhibit caspase-3 and caspase-7 activity. Carboxy-terminal fragment inhibits caspase-9 activity	SESD (242)

IDENTIFYING CASPASE SPECIFICITY WITH COMBINATORIAL PEPTIDE LIBRARIES.

One way to determine cleavage preference is to identify sites in substrates and compare them, as we will see. Another is through the use of combinatorial peptide libraries. This approach requires that we have a method to detect that cleavage of a substrate has occurred. This is most easily done by attaching a fluorescent molecule to the end of a peptide in such a way that it only becomes fluorescent when the bond between it and the peptide is cleaved, resulting in a measurable signal. This is a useful way to measure protease activity in general (Fig. 7).

Figure 7.

Caspase activity detected using a peptide substrate that generates a fluorescent signal after cleavage.

Because for most of the caspases P1 is D (aspartate), this position can be fixed, and then P1′, P1, P2, P3, and P4 (or more) can be varied with every possible amino acid to make a combinatorial library of the set. Each is then tested for its sensitivity to cleavage. An example of the preferences for caspase-3 found in this way is represented in Figure 8.

Figure 8.

Representation of amino acid preferences for caspase-3. Amino acids are shown in single-letter code, and the size of each letter corresponds to its frequency in cleaved peptides. Cleavage occurs between P1 and P1′. (Reprinted from Mahrus et al. 2008. ©2008 with permission from Elsevier.)

When all of the cleavage sites in caspase-3 substrates are compared, we get the pattern shown in Figure 10.

Figure 10.

Representation of caspase-3 cleavage sites in known substrates. (Reprinted from Mahrus et al. 2008. ©2008 with permission from Elsevier.)

Clearly, the combinatorial libraries worked well, giving us a pretty good idea of the preferred sites for caspase-3 in native proteins. However, there could be another problem here: When we make a cell extract, we might allow caspase-3, for example, to access substrates it would not normally see in cells. For example, cells have many subcellular compartments from which caspases could be excluded, and this compartmentalization can be lost when lysing the cell to make an extract. Thus, it is important to know whether caspase-3 has a similar substrate preference during apoptosis in intact cells.

We can address this question by comparing the proteins that are cut during apoptosis. Although we cannot be sure that all of the cuts are due to caspase-3, an exhaustive analysis of substrate cleavage during apoptosis can be performed. The result of such a large-scale approach by mass spectroscopy is informative. This approach identified not only those substrates listed in Table 1, but more than 1000 substrates overall, and revealed a sequence preference (Fig. 11) that matched that predicted by the approaches mentioned above. Caspases with the preference of caspases-3 and -7 might therefore predominate during apoptosis to cut protein substrates. It is not unreasonable to suppose that these are, in fact, caspases-3 and -7.

Figure 11.

Representation of protein cleavage events in cells undergoing apoptosis. (Reprinted from Mahrus et al. 2008. ©2008 with permission from Elsevier.)

KEY SUBSTRATES FOR EXECUTIONER CASPASES IN APOPTOSIS

We know what sort of sequences the executioner caspases like to cut during apoptosis, but how does this lead to the features of this form of cell death? Is the effect of executioner caspase activation really “death by a thousand cuts”? Or do only a few cuts in certain key substrates produce the phenomenon we call apoptosis?

Some insights come from considering how caspase cleavage of a substrate can affect a cell. We can envision four ways: Cleavage could (1) destroy an activity, provided that the cleavage is efficient; (2) trigger an activity by removing an inhibitor or an inhibitory domain in the substrate; (3) convert a protein into a dominant-negative version that inhibits the activity of the intact protein; or (4) have no relevant effect at all. The last is vexing and subtle. Many caspase cleavage events might simply “happen” in a cell that is doomed to die anyway. Therefore, even if an event seems likely to be important, on reflection, its importance could pale in the context of death. For example, the activation of an important transcription factor by an executioner caspase is unlikely to result in the production of a protein if, at the same time, the genome has been dismantled by DNA fragmentation.

The vast majority of proteins cleaved during apoptosis do not have established roles in the events that characterize this form of cell death. Therefore, to gain an understanding of how executioner caspases bring about apoptosis, a different but complementary approach is needed.

A useful way to identify caspase substrates that are important for apoptotic events is to dissect a process that depends on caspase activity and identify the key substrate responsible for that process. Below, we consider the substrates whose cleavage produces specific apoptotic events in different cellular compartments: the nucleus, plasma membrane, and mitochondria.

NUCLEAR EVENTS MEDIATED BY CASPASE CLEAVAGE OF SPECIFIC SUBSTRATES

A striking feature of apoptosis is the fragmentation of chromatin. The DNA is cut into pieces equivalent to one or more nucleosomes (multiples of 180 base-pairs), and this phenomenon depends on executioner caspases being active in the dying cell.

If we treat normal cytosol from living cells with caspase-3 and then add this to isolated nuclei, DNA fragmentation characteristic of apoptosis occurs. This approach allows identification of the enzyme responsible, as well as elucidation of how it is activated by caspase-3. The enzyme is a nuclease that preferentially cuts DNA at accessible sites between nucleosomes: “caspase-activated DNase” (CAD) or, alternatively, “DNA fragmentation factor of 40 kDa” (DFF40). It is present in healthy cells, but it is held in an inactive complex by an inhibitor, called iCAD (also called DFF45). The inhibitor is the caspase substrate, and, when cleaved by caspase-3, it releases the active nuclease to cut the DNA.

It would seem especially dangerous for cells to produce a nuclease capable of fragmenting the genome, but there is an additional safeguard. CAD is completely inactive unless properly folded by a chaperone, and this is a function of iCAD, which then holds the nuclease inactive unless iCAD is cleaved. The basic scheme is shown in Figure 12.

Figure 12.

Cleavage of the iCAD–CAD complex by caspases is responsible for DNA fragmentation during apoptosis. Caspase-3 (right) cleaves iCAD (center), releasing the active CAD, which randomly cuts the chromatin at accessible sites between the nucleosomes (top left). When run on an agarose gel, the DNA forms a ladder (far left) composed of multiples of nucleosome-sized lengths. (Image courtesy of Dr. Yufang Shi, Institutes for Translational Medicine, Soochow University, Suzhou, China; structure [inset], PDB 1GQF [Riedl et al. 2001].)

Cutting of the chromatin by CAD probably facilitates the degradation of the dead cell after it has been engulfed by a phagocytic cell (discussed in Green 2022b). In the absence of functional CAD (achieved by removing CAD or its chaperone, iCAD), DNA fragmentation does not occur in the cell before engulfment, but the cell dies nevertheless. Therefore, although it is unlikely that the cell can survive once caspases have cleaved iCAD and released active CAD, this is not required for cell death.³

So, are CAD and iCAD important? It is difficult to say. Both are present in most of the animals for which we have genomic sequences, with at least one exception. Among the animals, iCAD and CAD are found in hydra, sea anemones, insects, and of course vertebrates (to name a few), but they are not found in nematodes. Apparently, nematodes lost this substrate and its function in apoptosis along the way.

Other nuclear events in apoptosis can be traced to the cleavage of additional caspase substrates. During apoptosis, the chromatin becomes condensed, and this has been linked to the cleavage of the protein acinus by caspase-3 (and probably caspase-7). Precisely how acinus causes chromatin condensation is unclear but might involve phosphorylation of histones or the participation of acinus in DNA fragmentation by CAD.

Disruption of the nuclear envelope also occurs in apoptosis, and this can be traced to the cleavage of lamins by caspase-6. Lamins are important structural elements that preserve nuclear integrity and are involved in the breakdown and regeneration of the nuclear envelope during mitosis. If mutant lamins lacking the cleavage site are introduced, the nuclear envelope remains intact during apoptosis; however, all other aspects of apoptosis seem to proceed normally.

EVENTS AT THE PLASMA MEMBRANE CAUSED BY CASPASES

During apoptosis, the cell undergoes extensive membrane blebbing, extending bulbous outgrowths that often break off as small, membrane-bound bodies. This is an effect, at least in part, of actin polymerization, and pharmacological inhibitors of actin polymerization block blebbing.

Three caspase substrates have been implicated in actin polymerization and blebbing during apoptosis. These are the actin regulator gelsolin and two kinases that function in signaling pathways that control actin organization in cells: p21-activated kinase (PAK) and ROCK1 kinase. In all three, caspase cleavage activates the protein by removing regulatory domains, which results in the dynamic changes in actin organization that cause blebbing. Pharmacological inhibitors of ROCK1 or silencing its expression prevent blebbing and can even stop it after it is under way (Fig. 13). As with the other related events that we have discussed, however, the cell still dies if blebbing is blocked.

Figure 13.

Silencing or inhibition of ROCK1 prevents blebbing during apoptosis. Blebbing seen during apoptosis (A), is eliminated when ROCK1 is silenced with small interfering RNA (siRNA) (B) or blocked with an inhibitor (C). (Courtesy of Michael Olson, Beatson Institute.)

One important consequence of apoptotic cell death is that the cell is eaten by other cells before the integrity of the plasma membrane is lost. The “eat me” signals that ensure that this will occur depend on caspase activation. The most important of these signals is the appearance on the cell surface of the lipid phosphatidylserine, which is recognized by the cells that do the eating (discussed in much more detail in Green 2022b). Phosphatidylserine is normally restricted to the inner leaflet of the plasma membrane, but, during apoptosis, the lipids scramble, and phosphatidylserine is exposed.

The molecule that is responsible for keeping phosphatidylserine (and other phospholipids) localized to the inner leaflet of the plasma membrane is adenosine triphosphatase type 11C (ATP11C), which uses ATP to “flip” phosphatidylserine from the outer to the inner leaflet. If ATP is depleted from a cell, phosphatidylserine can therefore accumulate on the surface. A second protein, CDC50A, is required to chaperone ATP11C to the plasma membrane, and cells without CDC50A constitutively expose phosphatidylserine. ATP11C is a caspase substrate; cleavage by executioner caspases destroys its flippase function, and phosphatidylserine is then exposed on the outer leaflet (Fig. 14). Mutation of the caspase cleavage sites in ATP11C results in cells that do not efficiently expose phosphatidylserine during apoptosis.

Figure 14.

Caspases induce exposure of phosphatidylserine on the outer leaflet of the plasma membrane by two mechanisms. Caspase cleavage (scissors) disrupts the flippase activity of ATP11C (left) and induces the scramblase activity of Xkr8 (right). Both result in loss of phospholipid asymmetry, resulting in exposure of phosphatidylserine on the cell surface.

In addition to this loss of flippase function during apoptosis, the lipids of the plasma membrane also undergo an active scrambling to bring phosphatidylserine rapidly to the outer leaflet. Another caspase substrate is responsible for this event, a protein called Xkr8. Xkr8 associates with a partner protein, which can be either of the two related proteins basigin or neuroplastin. When Xkr8 is cleaved by an executioner caspase, it now forms heterotetramers of two Xkr8 and the two partner molecules. This higher-order complex is responsible for scrambling the lipids of the plasma membrane during apoptosis (Fig. 14).

The exposure of phosphatidylserine on the cell surface can occur without caspase activation, and therefore it is not specific to apoptosis. Because of the nature of the plasma membrane, any disruption in the membrane results in redistribution of phospholipids, including phosphatidylserine. An influx of calcium into the cell can activate another phospholipid scramblase, TMEM16F, resulting in transient exposure of phosphatidylserine. This is important, for example, in platelet activation. Cells without TMEM16F do not externalize phosphatidylserine in response to calcium influx, but they do in response to caspase activation during apoptosis.

MITOCHONDRIAL EFFECTS OF EXECUTIONER CASPASES

When the mitochondrial pathway of apoptosis is engaged (discussed in detail in Green 2022d), activated executioner caspases gain access to the inner membrane of mitochondria (because the outer mitochondrial membrane becomes permeable). Here, several caspase substrates on the inner membrane are found, including the oxidoreductase NDUFS1, which is an integral part of complex I of the electron-transport chain, the major source of energy in the cell.

When NDUFS1 is cut by caspases, several events rapidly follow. Electrons that would normally be used for oxidative phosphorylation are shuttled to oxygen to produce superoxides. These can then be converted to hydroxyl radicals that are damaging to many proteins and membranes. In addition, the proton gradient normally produced by electron transport dissipates, and ATP levels rapidly decrease. This loss of ATP has effects on the plasma membrane, because ion pumps require ATP to operate. Introduction of a noncleavable NDUFS1 mutant into cells prevents these rapid, caspase-dependent events, but it has no effect on other apoptotic changes, such as DNA fragmentation, blebbing, or cell death.

DEATH BY A THOUSAND CUTS?

As we noted, there are many hundreds of caspase substrates that are cleaved when executioner caspases become active. How many of these are actually responsible for cell death? Clearly, if a nuclease such as the CAD becomes active in a cell and fragments the DNA extensively, the cell will not survive. But some specialized cells, such as mammalian red blood cells, persist for more than 100 days without a nucleus (the elimination of the nucleus in red blood cells does not depend on CAD). And cells lacking iCAD or CAD still die by apoptosis. A glance at the substrates in Table 1 and those we discussed above give us an idea why. Indeed, it is remarkable that a few key substrates that have major roles in producing specific changes that we see during apoptosis can be identified. But it is unlikely that only a few key substrates (these or others) are responsible for cell death.⁴

CASPASES AS KILLERS

The short answer to the question of what kills the cell during apoptosis is, of course, “caspases.” Introduction of an executioner caspase into any type of cell causes it to die, and, in animal cells, this occurs by apoptosis.

As we go into the pathways of apoptosis in this book, we will see several examples in which the activation of caspases is not essential for cell death, and other mechanisms come into play. For now, it is important to be aware of studies that reveal just how important caspases really are. For example, in C. elegans, one caspase, CED3, is required for most cell death during development and in germ cells of the adult worm that die in response to stress. Animals with mutations that prevent CED3 activation or function accumulate extra cells that differentiate and persist (Fig. 15). In this animal, cell death depends on completion of the apoptotic pathway and therefore depends on the caspase.

Figure 15.

Extra cells in CED3 mutant nematodes. Normal cell deaths in a wild-type larva (arrows, left) are not seen in a CED3 mutant animal (right). (Reprinted from Ellis and Horvitz 1986. ©1986 with permission from Elsevier.)

This is also true in Drosophila, although again not all cell deaths follow this rule. Drosophila lacking the initiator caspase Dronc accumulate extra cells, which is lethal for the developing fly (Fig. 16).

Figure 16.

Extra cells in Dronc-deficient fly embryos. Cell deaths (stained blue) in wild-type embryos (left) are not seen in embryos lacking the caspase Dronc (right). (Reprinted from Quinn et al. 2000. ©2000 with permission from the American Society for Biochemistry and Molecular Biology.)

Cultured Drosophila cells that lack caspase activity continue to survive and proliferate when exposed to stresses that would normally cause apoptotic cell death, emphasizing the importance of these enzymes in cell demise. Mice lacking an initiator caspase (caspase-9), or an executioner caspase (caspase-3), also accumulate extra cells (Fig. 17). Elimination of caspase-7 as well as caspase-3 exacerbates this effect. However, as we will see in Green (2022e), this effect might not be due to the extra cells “not dying”—it is a more complex effect of the delayed death of cells in the developing neural tube permitting excessive proliferation of developing neurons.

Figure 17.

Extra cells (arrow and asterisk) in the brains of caspase-3-deficient mice (−/−; lower) compared with wild-type mice (wt; upper). (Reprinted with permission from Macmillan Publishers Ltd.: Kuida et al. 1996. ©1996.)

Caspases must therefore be important for the death of some cells. However, a close look at the development of such mutant mice reveals that developmental cell death still occurs. We return to this problem of caspase-independent cell death later (in Green 2022d) as we learn more about the pathways and other forms of cell death.

So far, we have discussed caspases and the way they contribute to apoptosis. But how are the caspases activated to cause this devastation? We consider this problem next.

ADDITIONAL READING

Caspases

Lamkanfi M, Festjens N, Declercq W, Vanden Berghe T, Vandenabeele P. 2007. Caspases in cell survival, proliferation and differentiation. Cell Death Differ 14: 44–55.

A review of the caspases and their functions in life and death.

Yuan J, Shaham S, Ledoux S, Ellis HM, Horvitz HR. 1993. The C. elegans cell death gene ced-3 encodes a protein similar to mammalian interleukin-1 β-converting enzyme. Cell 75: 641–652.

The original landmark paper that identified caspases as being important in apoptosis.

Fernandes-Alnemri T, Litwack G, Alnemri ES. 1994. CPP32, a novel human apoptotic protein with homology to Caenorhabditis elegans cell death protein Ced-3 and mammalian interleukin-1 β-converting enzyme. J Biol Chem 269: 30761–30764.

The first description of mammalian caspase-3, originally called CPP32.

Alnemri ES, Livingston DJ, Nicholson DW, Salvesen G, Thornberry NA, Wong WW, Yuan J. 1996. Human ICE/CED-3 protease nomenclature. Cell 87: 171.

Eckhart L, Ballaun C, Hermann M, VandeBerg JL, Sipos W, Uthman A, Fischer H, Tschachler E. 2008. Identification of novel mammalian caspases reveals an important role of gene loss in shaping the human caspase repertoire. Mol Biol Evol 25: 831–841.

An overview of the evolutionary relationships among mammalian caspases, including caspases-15, -16, -17, and -18.

Kuida K, Zheng TS, Na S, Kuan C, Yang D, Karasuyama H, Rakic P, Flavell RA. 1996. Decreased apoptosis in the brain and premature lethality in CPP32-deficient mice. Nature 384: 368–372.

The first knockout of caspase-3 (called CPP32) and effects on development.

Caspase Specificities

Timmer JC, Salvesen GS. 2007. Caspase substrates. Cell Death Differ 14: 66–72.

Thornberry NA, et al. 1997. A combinatorial approach defines specificities of members of the caspase family and granzyme B. Functional relationships established for key mediators of apoptosis. J Biol Chem 272: 17907–17911.

The identification of caspase preferences through the use of a combinatorial peptide library.

Mahrus S, Trinidad JC, Barkan DT, Sali A, Burlingame AL, Wells JA. 2008. Global sequencing of proteolytic cleavage sites in apoptosis by specific labeling of protein N termini. Cell 134: 866–876.

Another approach to characterization of caspase substrates.

Poreba M, Strozyk A, Salvesen GS, Drag M. 2013. Caspase substrates and inhibitors. Cold Spring Harb Perspect Biol 5: a008680.

Julien O, Wells JA. 2017. Caspases and their substrates. Cell Death Differ 24: 1380–1389.

A useful review of the specificities and functions of caspases, and how they cause cell death.

Functional Caspase Substrates

Luthi AU, Martin SJ. 2007. The CASBAH: A searchable database of caspase substrates. Cell Death Differ 14: 641–650.

An introduction to a searchable database of all known caspase substrates.

Kumar S, van Raam B., Salvesen GS, Cieplak P. 2014. Caspase cleavage sites in the human proteome: CaspDB, a database of predicted substrates. PLoS One 9: e110539.

Another searchable database of caspase substrates.

Sakahira H, Enari M, Nagata S. 1998. Cleavage of CAD inhibitor in CAD activation and DNA degradation during apoptosis. Nature 391: 96–99.

One of the original descriptions of the iCAD–CAD system and its role in DNA fragmentation during apoptosis.

Zhang J, Liu X, Scherer DC, van Kaer L, Wang X, Xu M. 1998. Resistance to DNA fragmentation and chromatin condensation in mice lacking the DNA fragmentation factor 45. Proc Natl Acad Sci 95: 12480–12485.

A clear demonstration of the role of iCAD/DFF45 in DNA fragmentation versus cell death.

Sebbagh M, Renvoize C, Hamelin J, Riche N, Bertoglio J, Breard J. 2001. Caspase-3-mediated cleavage of ROCK I induces MLC phosphorylation and apoptotic membrane blebbing. Nat Cell Biol 3: 346–352.

One of the original descriptions of ROCK as a caspase substrate and its role in blebbing during apoptosis.

Coleman ML, Sahai EA, Yeo M, Bosch M, Dewar A, Olson MF. 2001. Membrane blebbing during apoptosis results from caspase-mediated activation of ROCK I. Nat Cell Biol 3: 339–345.

See above.

Kothakota S, Azuma T, Reinhard C, Klippel A, Tang A, Chu K, McGarry TJ, Kirschner MW, Koths K, Kwiatkowski DJ, et al. 1997. Caspase-3-generated fragment of gelsolin: Effector of morphological change in apoptosis. Science 278: 294–298.

The role of gelsolin in blebbing during apoptosis.

Ricci JE, Munoz-Pinedo C, Fitzgerald P, Bailly-Maitre B, Perkins GA, Yadava N, Scheffler IE, Ellisman MH, Green DR. 2004. Disruption of mitochondrial function during apoptosis is mediated by caspase cleavage of the p75 subunit of complex I of the electron transport chain. Cell 117: 773–786.

The original description of NDUFS1 as a mitochondrial caspase substrate.

Suzuki J, Denning DP, Imanishi E, Horvitz HR, Nagata S. 2013. Xk-related protein 8 and CED-8 promote phosphatidylserine exposure in apoptotic cells. Science. 341: 403–406.

The original description of the caspase-activated phospholipid scramblase Xkr8.

Suzuki J, Imanishi E, Nagata S. 2016. Xkr8 phospholipid scrambling complex in apoptotic phosphatidylserine exposure. Proc Natl Acad Sci 113: 9509–9514.

More insights into the function of Xkr8 during apoptosis.

Segawa K, Kurata S, Yanagihashi Y, Brummelkamp TR, Matsuda F, Nagata S. 2014. Caspase-mediated cleavage of phospholipid flippase for apoptotic phosphatidylserine exposure. Science 344: 1164–1168.

The original description of how executioner caspases inhibit the lipid translocase responsible for keeping phosphatidylserine sequestered to the inner leaflet of the plasma membrane.