| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Departments of
* Surgery and
Pathology, Immunology and Laboratory Medicine, University of Florida College of Medicine, Gainesville, Florida 32610, USA
1Correspondence: Department of Surgery, University of Florida College of Medicine, Room 6116, Shands Hospital, 1600 S.W. Archer Rd., PO Box 100286, Gainesville, FL 32610-0286, USA. E-mail: moldawer{at}surgery.ufl.edu
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONAPOPTOSIS DURING INFLAMMATIONCONCLUSIONSREFERENCES |
|---|
or Fas ligand.
Apoptosis proceeds via auto-activation of cytosolic and/or
mitochondrial caspases, which can be influenced by the pro- and
anti-apoptotic members of the Bcl-2 family. In experimental animals,
not only can treatment with inhibitors of apoptosis prevent lymphoid
cell apoptosis; it may also improve outcome. Although clinical trials
with anti-apoptotic agents remain distant due in large part to
technical difficulties associated with their administration and tissue
targeting, inhibition of lymphocyte apoptosis represents an attractive
therapeutic target for the septic patient.—Oberholzer, C., Oberholzer,
A., Clare-Salzler, M., Moldawer, L. L. Apoptosis in sepsis: a new
target for therapeutic exploration.
Key Words: Bcl-2 • caspases • lymphocytes • neutrophils
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONAPOPTOSIS DURING INFLAMMATIONCONCLUSIONSREFERENCES |
|---|
.
Despite recent progress in antibiotics and critical care therapy,
sepsis is still associated with a high mortality rate. Septic shock and
sequential multiple organ failure/dysfunction syndrome (MOF/MODS)
correlate with poor outcome (2)
, and septic shock is the
most common cause of death in intensive care units (3)
.
Experimental data have provided a solid rationale for innovative
therapies and clinical trials, yet the outcome from this syndrome has
not improved significantly in the past 40 years (4)
.
The immunological cascade resulting in the sepsis response can be
initiated by tissue injury, ischemia-reperfusion injury, gram-positive
organisms, and fungi as well as gram-negative organisms and their
constituent endotoxin. The sepsis response may begin with an infectious
nidus, which may either invade the bloodstream, leading to
dissemination and positive blood cultures, or proliferate locally and
release various microbial products into the bloodstream. In multiple
trauma or hemorrhagic shock, the direct tissue or secondary
ischemia-reperfusion injury may also lead to an increased appearance of
microorganisms and exotoxins from the gut. The host response to these
microbial products or to the trauma and ischemia-reperfusion injury
itself leads to the rapid activation of the innate immune response and
the release of a variety of humoral mediators, including
glucocorticoids, catecholamines, and proximal proinflammatory cytokines
like tumor necrosis factor (TNF-), interleukin-1 (IL-1), and
IL-6 (5)
.
A vigorous induction of the innate immune system can and often does
have catastrophic effects on the patient with sepsis syndrome.
Exaggerated production of proinflammatory cytokines and the induction
of more distal mediators such as nitric oxide, platelet activation
factor, and prostaglandins have been implicated in the endothelial
changes and induction of a procoagulant state that leads to
hypotension, inadequate organ perfusion, and necrotic cell death
associated with MODS. This proinflammatory state has been defined as
being a systemic inflammatory response syndrome (SIRS)
(6)
.
However, most patients survive this initial SIRS event and the
proinflammatory state ultimately resolves. The proinflammatory
cytokines and humoral mediators responsible for the induction of the
innate immune response and SIRS also contribute to the development of
acquired or specific immune defects. The patient frequently enters an
immunological state characterized by T cell hyporesponsiveness, anergy,
and a defect in antigen presentation that has been recently termed a
compensatory anti-inflammatory response syndrome (CARS)
(7)
.
The role that apoptosis plays in sepsis syndromes and in the development of CARS and MODS has not been adequately explored, but there is rapidly developing evidence to suggest that increased apoptotic processes may play a determining role in the outcome to sepsis syndromes. In particular, increased apoptosis, particularly in lymphoid tissues and potentially in some parenchymal tissues from solid organs, may contribute to the sepsis-associated MODS and can be a potential therapeutic target for intervention. Direct apoptotic organ injury and the immune suppression secondary to apoptotic losses in T cell, B cell, and NK cell populations may contribute significantly to the risk of secondary opportunistic infections. Therapeutic efforts at modulating the apoptotic response, particularly by interfering with cell signaling pathways that lead to caspase-mediated apoptosis, represent an attractive therapeutic target for the septic patient. In the current review, we address the scientific rationale for modifying apoptotic pathways in sepsis syndrome and propose several potential therapeutic targets.
|
APOPTOSIS DURING INFLAMMATION |
|---|
|
TOPABSTRACTINTRODUCTIONAPOPTOSIS DURING INFLAMMATIONCONCLUSIONSREFERENCES |
|---|
). Physical or chemical injury, such as the deprivation of oxygen that
occurs in the heart muscle during myocardial infarction or in sepsis
when the microvasculature is disrupted by fibrin deposition and
intravascular coagulopathies, often leads to cell disintegration or
necrosis. The necrotic tissue is phagocytosed and degraded by
phagocytic cells, often leading to inflammation.
|
An alternative form of cell death is known as programmed cell death or
apoptosis. Apoptosis is a normal cellular process that is crucial for
tissue remodeling and/or development. For example, most thymocytes
undergo an apoptotic cell death when they fail positive selection or
are negatively selected as a result of recognizing self-antigens
(8)
.
During apoptosis, cells undergo a non-necrotic cellular suicide that,
in contrast to necrosis, generally does not produce inflammation and
injury in the tissue (9)
. Cells undergoing apoptosis
typically show DNA fragmentation, condensation of chromatin, membrane
blebbing, cell shrinkage, and finally, disassembly into
membrane-enclosed vesicles (10)
. A hallmark of this type
of cell death is the fragmentation of nuclear DNA into multiples of 200
base pairs through the activation of endogenous nucleases that cleave
the DNA between nucleosomes.
Cytokines like TNF- and Fas ligand (FasL), glucocorticoids and
granzymes can induce apoptosis in several cell populations, whereas
other cytokines such as IL-1, IL-6, and G-CSF often inhibit apoptosis.
Increased levels of several of these proinflammatory cytokines
(TNF-, IL-1ß, IL-6, G-CSF) have been reported in sepsis syndromes
and have been implicated as principal effectors of the host immune
responses during endotoxemia (11)
and severe trauma
(12)
. In addition, increased FasL expression has been
reported in rodent models of endotoxemia and in generalized peritonitis
(13)
; elevated soluble Fas ligand has also been detected
in critically ill patients [head injury and acute respiratory distress
syndrome (ARDS)] in cerebrospinal fluid as well as in bronchial fluid
(14
, 15)
.
TUMOR NECROSIS FACTOR-INDUCED APOPTOSIS AND SIGNAL TRANSDUCTION
There is growing appreciation that alterations in lymphoid cell
and neutrophil apoptosis are normal components of the innate immune
response. As such, it is not surprising that several of the humoral
factors that initiate and regulate the innate immune response also
regulate apoptotic processes in lymphoid tissues and in neutrophils.
TNF-, FasL, and glucocorticoids appear to be the principal humoral
factors that may induce apoptosis in these cell populations.
TNF- is produced by macrophages and T cells in response to infection
(16)
. TNF- is involved in both inflammation and
cellular apoptosis, and is synthesized as a 26 kDa membrane-associated
precursor that is cleaved to the soluble 17 kDa form by
TNF--converting enzyme, a novel matrix metalloproteinase
(17)
. The biological responses to TNF- are mediated by
ligand binding via two structurally distinct receptors: p55 (TNF-R I,
CD120a) and p75 (TNF-R II, CD120b) (18)
. Expression of
these two receptors appears to be differentially regulated and shows
some tissue specificity. Both receptors are transmembrane glycoproteins
(Fig. 2
). The two TNF receptors, however, differ significantly in their binding
affinities (19)
as well as their intracellular signaling
pathways (20)
. The intracellular signaling domains of the
TNF-R1 (p55) receptor actually shares greater homology with the
intracellular signaling domains of Fas/CD95 receptor than it does with
TNF-R2 (p75), particularly with regard to a highly conserved
intracellular domain called the ‘death domain’ (DD). This sequence
plays a pivotal role in TNF-’s ability to trigger apoptosis in the
cell.
|
Binding of homotrimeric TNF- causes trimerization of its receptors.
Trimerization of the TNF-R1 (p55) (18)
initiates the
physical association of the receptors’ death domains. Subsequently, an
adapter protein termed TRADD (TNF receptor 1-associated death domain)
binds through its own death domain to the clustered receptor death
domains of the TNF-R1. TRADD signaling lies at the bifurcation of the
apoptotic and proinflammatory transduction pathways of TNF- (Fig. 2)
(21)
. TRADD functions as a platform adapter that recruits
several signaling molecules to the activated receptor like TRAF2 (TNF
receptor-associated factor 2) (22)
and RIP (receptor
interacting protein) (23)
stimulating the pathway, leading
to either nuclear factor (NF)-
B-inducing kinase (NIK) to activate
NF-B or c-Jun N-terminal kinase to activate activator protein 1
(AP-1) (24)
. On the other hand, binding of Fas-associated
death domain (FADD) mediates activation of apoptosis (25)
.
FADD couples the TNF-R1-TRADD complex to induce the recruitment and
activation of procaspase-8 molecules (26)
. Activation of
caspase-8, also known as FADD-like IL-ß-converting enzyme by the FADD
complex is a principal mechanism initiating physiological and
pathophysiological apoptosis through caspase-3 (Fig. 2)
(27)
. Moreover, FADD represents the convergence point for
signaling between the p55 TNF receptor and Fas receptor
(27)
. Besides FADD, TNF-R1 can engage an adapter called
RIP-associated ICH-1/ZED-3-homologous protein with a death domain
(RAIDD) or caspase and RIP adapter with death domain (28)
.
RAIDD binds through a death domain to the death domain of RIP and
through a caspase recruitment domain (CARD) motif to a similar sequence
in the death effector, caspase-2, thereby inducing apoptosis. The TNF
type II receptor participates in the proinflammatory signal of TNF-
via TRAF2, which mediates activation of NF-B (29)
.
FAS LIGAND-INDUCED APOPTOSIS AND SIGNAL TRANSDUCTION
Fas ligand (FasL) is a 40 kDa type II integral membrane protein
belonging to the TNF/nerve growth family (30)
.
Membrane-bound FasL can be proteolytically cleaved by
metalloproteinases generating a soluble FasL (31)
. The
specific receptor for FasL is Fas (CD95, Apo-1), a 45 kDa type 1
transmembrane protein and a member of the TNF receptor family
(32)
. Fas is ubiquitously expressed in various tissues
with an abundant expression in the thymus, liver, heart, and kidney and
is controlled mainly by tight regulation of FasL expression
(33)
. On the other hand, FasL is predominantly expressed
in activated lymphocytes and natural killer (NK) cells and
constitutively expressed in tissues such as testis and eye. The Fas and
FasL pathway plays an important role in physiological processes that
require apoptosis. These processes include the peripheral deletion of
activated mature T cells at the end of an immune response, or the
killing of virus-infected or cancer cells by cytotoxic T cells and by
natural killer cells, and the killing of inflammatory cells at ‘immune
privileged’ sites (34)
.
Fas ligand binding to Fas leads to oligomerization and clustering of
the Fas DD (Fig. 2)
. The DD are actually a protein–protein interaction
domain that binds an adapter protein FADD (35)
. The
resulting complex is termed the death-inducing signaling complex. FADD
also contains a death effector domain that acts as a CARD-activating
procaspase-8 (36)
. Caspase-8 then activates downstream
effector caspases. The net results of these enzymes is induction of
apoptosis, which is associated with the exposure of phosphatidyl
serines on the exterior cell surface promoting rapid phagocytosis of
dying cells. Ligation of FasL to its receptor leads to apoptosis by at
least two different pathways depending on the cell type
(37)
. T cell lines that overexpress Bcl-2 are not
susceptible to CD95-induced apoptosis (38)
. In contrast,
Bcl-2 overexpression has no influence on Fas-mediated apoptosis in the
SKW6 cell line (37)
. In the former, it is believed that
ligation of FasL to the Fas receptor leads to mitochondrial release of
cytochrome c and the cleavage of caspase-3. In the latter
cell line, CD95 leads to caspase-8 activation, which then bypasses the
mitochondria and directly results in activation of the downstream
mediators, such as caspase-3 (37)
(Fig. 2)
.
Giordano et al. have shown that Fas expression can be induced and
up-regulated by inflammatory cytokines such as IL-1 (39)
,
which are elevated during sepsis and trauma (11
, 40)
. This
raises the intriguing possibility that inflammation per se may promote
Fas-dependent injury. In addition, soluble FasL has been detected in
cerebrospinal fluid in patients after severe head injury
(14)
and soluble FasL concentrations correlated with the
severity of brain injury. Furthermore, increased FasL has been detected
in bronchial fluid from patients suffering ARDS (15)
.
Moreover, Ottonello and colleagues have shown that soluble FasL is a
potent chemoattractant for human neutrophils without evoking their
secretory responses (41)
and thus further supports the
Fas-FasL system-mediated cell injury independent of its
apoptosis-inducing properties.
CYTOTOXIC T CELLS (CD8) CAN INDUCE TARGET CELLS TO UNDERGO
PROGRAMMED CELL DEATH IN SEPSIS SYNDROMES
Cytotoxic T lymphocytes and NK cells are known to induce lethal
damage on their target cells by inducing granule exocytosis of
granzymes or via the FasL system. However, the role that cytotoxic T
cells play in the host response to sepsis is unclear. Faist and
colleagues have reported that after a burn injury, peripheral blood
CD8+ numbers did not change, but their secretion of IL-4 was markedly
increased (42)
. There is strong evidence that cytotoxic T
cells kill their targets largely by programming them to undergo
apoptosis. When cytotoxic T cells are mixed with target cells and
rapidly brought into contact by centrifugation, they can program
antigen-specific target cells to die within 5 min, although death may
take hours to become fully evident. An early feature of T cell killing
is degradation of target cell DNA, while later effects include the loss
of membrane integrity, which may also be induced by other cytotoxic
mechanisms. The short period required by cytotoxic T cells to program
their targets to die reflects the release of preformed effector
molecules by the T cell, which activate an endogenous apoptotic pathway
within the target cell. The elimination of infected cells without
destruction of healthy tissue requires the cytotoxic mechanisms of CD8+
T cells.
The principal mechanism through which cytotoxic T cells act is by the
calcium-dependent release of specialized lytic granules upon
recognition of antigen on the surface of a target cell. These granules
are modified lysosomes that contain at least two distinct classes of
cytotoxins, proteins that are expressed selectively in cytotoxic T
cells and stored in the lytic granules in active form. CD8+ T cells
carry out their killing function by releasing two types of preformed
cytotoxins: granzymes, which induce apoptosis in any target cell, and
the pore-forming protein perforin, which creates holes in the target
cell membrane through which granzymes can enter. Recently, it has been
shown that granzyme B can cleave the ubiquitous cellular enzymes,
caspase-3 and caspase-8, which play a key role in the programmed cell
death of all cells (43
, 44)
. Ligation of the T cell
receptor similarly induces de novo synthesis of perforin and
granzymes in CD8+ T cells, so that their supply of lytic granules is
replenished. This makes it possible for a single CD8+ T cell to kill
many targets in succession.
The membrane-bound molecule, FasL, expressed by CD8+ and TH1 CD4+ T cells is also capable of inducing apoptosis by binding to Fas on target cells. Distribution of Fas on the surface of target cells is generally ubiquitous, so the rate-limiting step is often the up-regulation of Fas ligand on the surface of the effector cells or the recruitment of Fas ligand expressing cells to the tissues. These properties allow the cytotoxic T cell to attack and destroy virtually any cell that is infected with a cytosolic pathogen.
GLUCOCORTICOIDS INDUCE APOPTOSIS THROUGH CYTOSOLIC CASPASES
Glucocorticoid hormones have been documented to be increased in
sepsis syndromes and have been implicated as regulators of T cell
growth, differentiation, and death. In general, T lymphocytes,
particularly thymocytes, are especially sensitive to
glucocorticoid-mediated apoptosis. The intracellular signaling pathways
involved in glucocorticoid-induced apoptosis appear to be distinct from
the proximal pathways invoked by either TNF- or FasL, although both
appear to be caspase-3 dependent. Thymocyte apoptosis induced by
glucocorticoids can be prevented by the broad-acting caspase inhibitor
ZVAD-fmk as well as by a more specific caspase-3 inhibitor, DEVD-CHO.
Unlike apoptosis-inducing pathways invoked by TNF- or FasL, which
require caspase-8, glucocorticoid-mediated activation of caspase-3
appears to be dependent on activation of caspase-9 (Fig. 3
). Evidence that caspase-9-mediated activation of caspase-3 is
independent of caspase-8 comes from studies of embryonic fibroblasts
from caspase-9 knockout mice. These cells are resistant to
glucocorticoid-induced apoptosis, but are surprisingly sensitive to
apoptosis induced by either UV radiation or FasL (45)
.
|
There is controversy, however, regarding how caspase-9 is activated in
glucocorticoid-induced apoptosis. Resistance to apoptosis in caspase-9
null cells is associated with a retention in mitochondrial membrane
potential, despite translocation of cytochrome c to the
cytosol. These findings suggest that in glucocorticoid-induced
apoptosis, caspase-9 activation of caspase-3 occurs downstream of
cytochrome c (46)
.
The evidence to date suggests that proximal activation of caspase-9
occurs through acid-sphingomyelinase- and ceramide-dependent pathways.
Glucocorticoid binding to its receptor rapidly induces diacylglycerol
generation through the actions of a protein kinase C- and phospholipase
C-dependent event. Diacylglycerol is the primary activator of acid
sphingomyelinase and ceramide generation, which appears to be dependent
for activation of caspase-9 (47
, 48)
.
THE ROLE OF CASPASES IN APOPTOSIS
Caspases are constitutively expressed as proenzymes that contain
three domains. Activation involves proteolytic processing between
domains, followed by association of the large and small subunits to
form a heterodimer. Caspases inactivate proteins that protect living
cells from apoptosis, and they contribute to cell death by direct
disassembly of cell structures (49)
. It is not known which
caspase is responsible for cleavage of the diverse proteins such as
PARP (poly (ADP-ribose) polymerase), lamin B, actin, and others under
physiological conditions. Some caspases show overlapping specificities
for some substrates (caspase-3 and -7 both cleave PARP), whereas
caspase-6 is the only caspase known to cleave lamins (50)
.
Caspases themselves are substrates for other caspases and activate each
other in positive feedback loops. Another example of the function of
caspases is the cleavage of Bcl-2 proteins, negative regulators of
apoptosis, whereby the protein is not only inactivated, but a fragment
is also released that may directly promote apoptosis (51)
.
Moreover, caspase activity itself is tightly regulated. By
phosphorylation of procaspase-9 through the kinases Akt and p21-Ras,
procaspase-9 proteolytic processing through cytochrome c is
defective (52)
.
Depending on the signal, different caspases are activated,
leading to apoptosis. When ligand binding occurs (FasL, TNF-) to
their death receptors, caspase-8 is activated; on the other hand,
caspase-9 is involved in apoptosis induced by cytotoxic agents
(dexamethasone, gamma irradiation) (34
, 45)
. Yet binding
of Apaf-1 (apoptosis protease-activating factor 1), cytochrome
c and dATP must be present for activation of procaspase 9,
(Fig. 4
) (53)
. Caspase-9 appears to be an essential requirement
for thymocyte apoptosis in stressful conditions (45)
.
Another pathway inducing apoptosis has recently been shown to be
mediated through caspase-12, which is located in the endoplasmic
reticulum (ER) and leads to apoptosis by ER stress (i.e., accumulation
of excess proteins in the ER). This pathway is independent of the other
known caspase-dependent pathways (54)
.
|
Some caspases are active before they are proteolytically processed, but
do not induce apoptosis because of the presence of endogenous caspase
inhibitors, such as members of the inhibitor of apoptosis protein (IAP)
family (Fig. 4)
. These proteins rapidly inactivate active caspases.
When a sufficient concentration of activated caspases accumulates and
the IAPs cannot neutralize them, apoptosis proceeds (55)
.
In the different pathways inducing apoptosis, caspase-3 appears to play
a central role as most pathways result in the activation of caspase-3.
The precursor form of caspase-3 is localized in both the cytosol and
the intermembrane space of the mitochondria (56)
.
THE ROLE OF THE Bcl-2 FAMILY IN APOPTOSIS
Analyses of the proteins that induce or block apoptosis have
been extensively explored in the worm, Caenorhabditis
elegans. In this simpler organism, cellular death is a precisely
defined sequence during development, and the consequences of genetic
manipulations upon death or survival of these cells has been accurately
defined. Several ‘ced’ genes (C. elegans death) have been
identified. Remarkably, homologues of these genes have been found in
mammalian cells and often serve similar functions. One example of an
important effector of apoptosis in the worm is a protein called Ced-3.
Its mammalian homologue is an aspartate-directed cysteine protease that
belongs to a family of proteases, one prototype of which is an enzyme,
called IL-1-converting enzyme (ICE, caspase-1), that converts the
precursor form of the cytokines IL-1ß and IL-18 to their active
forms. Different Ced 3/ICE-like enzymes have been identified. They are
normally present in the cytoplasm in an inactive form and are
catalytically activated on various stimuli. The targets of these
proteases are matrix proteins and nucleoproteins whose degradation
results in nuclear DNA fragmentation and apoptosis (57)
.
Activation of these pro-apoptotic proteases is blocked by a family of
proteins, the prototype of which is Bcl-2 (B cell lymphoma-2). Bcl-2 is
widely expressed in immature tissues prenatally, but becomes highly
restricted with maturation. In the adult, Bcl-2 expression is in
immature cell populations, in hormonally responsive epithelia that
undergo cycles of hyperplasia and involution, and in neurons of the
peripheral nervous system (58)
. Changes in Bcl-2
expression match cell differentiation more closely than patterns of
death. Another member of the Bcl family is Bcl-x, which is present in a
long (Bcl-xL) and a short form
(Bcl-xS). The former functions as an
anti-apoptotic, like Bcl-2, whereas the latter, as well as a related
protein called Bax, promotes cell death. The Bcl-2 family of proteins
regulate apoptosis especially in lymphocytes (Fig. 4)
. Fluctuations in
the levels of expression of Bcl-2 or Bcl-xL
during lymphocyte maturation and activation appear to correlate
inversely with the cell’s susceptibility to apoptosis. Forced
overexpression of Bcl-2 or Bcl-xL results in
enhanced survival of immature lymphocytes and prolonged antibody
responses (59)
. On the other hand, disruption of Bcl-2 or
Bcl-xL leads to reduced survival of mature or
immature lymphocytes (60)
. The family members of Bcl-2 can
dimerize with one another, antagonizing or enhancing the function of
each other. In this manner, the ratio of inhibitors to activators in a
cell may determine the propensity of the cell to undergo apoptosis
(61)
. Another mechanism to regulate dimerization of Bcl-2
family members is by phosphorylation. Bad, a pro-apoptotic member of
the Bcl-2 family, is phosphorylated and loses its ability to bind
Bcl-xL. This enables the dissociated
Bcl-xL to execute its anti-apoptotic function.
One of the factors inducing phosphorylation of Bad is IL-3
(62)
. Further pro-apoptotic members such as Bak and Bax,
trigger the release of caspases from death antagonists via
heterodimerization and by inducing the release of mitochondrial
apoptogenic factors [apoptosis-inducing factors (AIF) and
procaspase-3] into the cytoplasm (Fig. 4)
(63)
. In
unstimulated cells, Bax is located in the cytosol and is in peripheral
association with intracellular membranes, including mitochondria, but
inserts into mitochondrial membranes after initiation of a death signal
(64)
. Bax can heterodimerize with multiple anti-apoptotic
members and also induce apoptosis by directly inducing cytochrome
c release (65)
. On the other hand,
heterodimerization with Bcl-2 leads to inhibition of its apoptotic
function (66)
. Bcl-2 and Bcl-xL are
localized to the outer mitochondrial membranes and endoplasmic
reticulum as well as nuclear membranes. The Bcl-2 family members may
also act as ion channels, which may play a role in the cell death
pathway (67)
. Bcl-xL binds to one
portion of Apaf-1, whereas procaspase-9 binds to its
NH2-terminal CARD. Bcl-xL
may inhibit the association of Apaf-1 with procaspase-9 and thereby
prevent its activation (68)
. Conversely, Bik, a
pro-apoptotic protein, may free Apaf-1 from the death inhibitor
(69)
. When Apaf-1 is freed from
Bcl-xL, it forms a complex with cytochrome
c and dATP and activates procaspase-9, which then leads to
activation of caspase-3 initiating apoptosis (Figs. 3
and 4)
.
BAR (bifunctional apoptosis regulator) is an apoptosis modulator
cross-linking two apoptosis pathways, namely, the cytosolic and the
mitochondrial pathway. It contains a DED-like domain capable of
suppressing apoptosis signaling through Fas (cytosolic), as well as a
domain that mediates interactions with the Bcl-2 family members and
suppresses Bax-induced apoptosis in yeast and mammalian cells
(mitochondrial) (70)
.
THE ROLE OF MITOCHONDRIA IN APOPTOSIS
At least three general mechanisms are known by which the
mitochondria are involved in activation-induced cell death, including
1) the release of proteins that trigger activation of the
caspase family of proteases, 2) disruption of electron
transport, oxidative phosphorylation and adenosine triphosphate (ATP)
production, and 3) alteration of cellular
reduction-oxidation (redox) potential (71)
. During
apoptosis, Bid, a member of the pro-apoptotic Bcl-2 family, is
activated as a result of caspase-8 cleavage (Fig. 3)
. This leads to
cytochrome c release (72)
, which can be
inhibited by the presence of Bcl-2 on these organelles
(73)
. Once cytochrome c is released, the cell
either dies by a rapid apoptotic mechanism involving Apaf-1-mediated
caspase activation or through a slower necrotic process. The latter is
a result of a collapse of electron transport that occurs when
cytochrome c is depleted from the mitochondria, resulting in
a variety of sequelae including generation of oxygen free radicals and
decreased production of ATP (Fig. 4)
. The consequence of this release
may depend on the cell type. If endogenous caspase inhibitors (IAP) are
present in sufficient quantities, the release of cytochrome
c may fail to induce the caspase-dependent apoptosis and the
slow loss of the electron transport chain may lead to cellular
necrosis. Conversely, if cytochrome c is available in
excess, caspases can be activated and still enough cytochrome
c may be docked by its high-affinity binding sites to
maintain electron transport. ATP production can therefore be continued
while caspases are activated and induce apoptosis through cleavage of
cytosolic and nuclear substrates (71)
.
Cytochrome c, Apaf-1, and dATP form a complex that activates
caspase-9 (Figs. 3
and 4)
(74
, 75)
. Apaf-1 has been shown
to also interact with other caspases (caspases 4 and 8), upstream of
caspase 9 (75)
. The relevance of these interactions is
unknown. Another caspase-activating protein released from mitochondria
is AIF. In vitro, AIF has been shown to process purified
procaspase-3. Its activity is blocked by ZVAD-fmk, a broad caspase
inhibitor, raising the possibility that AIF is in reality another
caspase (76)
.
LYMPHOCYTE APOPTOSIS IS EXAGGERATED IN THE CRITICALLY ILL
Apoptosis of mature T lymphocytes occurs through at least two
distinct processes: antigen-driven and lymphokine withdrawal (Fig. 4)
.
Active T cell apoptosis takes place indirectly either by increased
glucocorticoid release or by the antigen-induced expression of death
cytokines, like FasL and TNF- (77
, 78)
. These death
cytokines engage specific receptors that assemble caspase-activating
protein complexes. Fas-deficient T cells exhibit reduced but clearly
evident T cell receptor (TCR) -induced death, and the residual
apoptosis is blocked by inhibiting TNF- (78)
. In
resting T cells, the genes for FasL and TNF are weakly induced by TCR
stimulation, but in IL-2 stimulated T cells these death cytokines are
induced more strongly (79)
. This difference can explain in
part why TCR engagement kills cycling but not resting T cells, and
thereby might explain the observed lymphopenia in septic patients. Both
FasL and TNF- are found in cell surface-anchored forms in T cells
and can be readily cleaved from the membrane by metalloproteinases.
Evidence supports the concept that a single T cell can kill itself
through these autocrine signaling pathways (31
, 80)
.
However, Fas is constitutively present on circulating T cells and
increases during aging (81)
.
The activation of mature T lymphocytes results in the coexpression of Fas and FasL on the cells. A high concentration of the growth factor IL-2 enhances the expression of FasL on antigen-stimulated T cells and the development of sensitivity to Fas-mediated apoptosis. Thus, IL-2 is both a growth factor for T cells and a feedback regulator of T cell responses. Reactivation, especially in the presence of IL-2, leads to engagement of Fas by FasL and triggering of an apoptotic pathway, which can be responsible for activation-induced cell death and the prevention of uncontrolled activation of lymphocytes. In some cell populations, such as CD8+ T lymphocytes, activation-induced cell death is apparently triggered not through Fas, but through TNF receptor signaling. Conversely, apoptosis of CD4+ T cells, is usually a result of the interaction of two coexpressed molecules on activated cells: Fas (CD 95) and FasL.
Not all forms of lymphocyte apoptosis are a consequence of activation.
In fact, many lymphocytes are programmed to die unless protected by
receptor-mediated stimulation or growth. This type of programmed cell
death is due to neglect and does not appear to involve the Fas/TNF
receptor family. The susceptibility of proliferating T cells for
apoptosis positions IL-2 as a key, but generally unrecognized,
regulator of T cell apoptosis (82)
. The fate of cycling T
cells is thus linked to the prevailing state of the immune response.
Without continuous antigen stimulation, the expression of IL-2 and its
receptor falls, and passive or ‘lymphokine withdrawal’ apoptosis
ensues.
Cell death due to passive lymphokine withdrawal may result from
the cytoplasmic activation of caspases regulated partly by mitochondria
and the Bcl-2 protein. Therefore, passive apoptosis decreases the
expanded population of the T cells at the end of an immune response
(83)
. Conversely, if cycling T cells are strongly
stimulated by antigen, active or ‘antigen-induced’ apoptosis occurs,
which can be due to Fas ligand and TNF-. Even though cycling T cells
are programmed to die after strong antigen re-engagement, effector
functions, such as lymphokine production and T cell cytotoxicity, are
still potently expressed (79)
. The net effect is a
fine-tuned feedback response for eliminating T cells if there is too
much or too little antigen and IL-2 (Fig. 5
).
|
O’Sullivan and others have shown that impairment of the adaptive
immune response after traumatic or thermal injury is characterized by
failure of IL-2 production (84
, 85)
, which is produced by
TH1 cells. Major injury leads to a predominance
of TH2 cells; because expression of IL-2 is
decreased during major injury, the passive or ‘lymphokine withdrawal’
apoptosis of T cells ensues. This process may explain why TNF- and
FasL do not appear to be directly involved in injury-induced apoptosis.
It is presumed that this increased apoptotic loss of T lymphocytes
further increases the susceptibility to sepsis that is manifested in
severely injured patients. Furthermore, it has been shown that
increased lymphocyte apoptosis in peripheral blood T cells from burned
patients appears to contribute to decreased lymphocyte immune
responsiveness (86)
.
APOPTOSIS IN LYMPHOID ORGANS IN SEPSIS
A recent clinical study by Hotchkiss and colleagues demonstrated
increased apoptosis in lymphocytes (spleen and lymph nodes) and
gastrointestinal columnar epithelial cells (colon and ileum), as well
as a pronounced lymphopenia in patients who died from sepsis
(87)
. In addition, marked increases in activated caspase-3
and reduced Bcl-2 expression were also seen in these tissues. In
contrast, patients dying from nonseptic causes did not show an increase
in apoptosis in any of these cell populations, nor was there a similar
increase in caspase-3 activity (87)
. This organ-associated
lymphoid cell apoptosis is consistent with clinical reports indicating
an increased frequency of lymphopenia in hospital patients with sepsis
(88)
. The presence of lymphopenia has been documented to
correlate with an adverse outcome (89)
.
Similar findings have been seen in murine models of sepsis
(90)
. Rodents subjected to either a scald burn injury or
generalized peritonitis secondary to a cecal ligation and puncture
demonstrated a very rapid onset of increased apoptosis in lymphoid
organs, usually within 3 h (spleen, thymus, and bone marrow)
(Table 1
). We reported that after a scald burn injury, caspase-3-dependent
apoptosis increased in lymphoid organs, and up to 35% of T cells in
the thymus were undergoing apoptosis (91)
. Similarly,
Ayala and colleagues observed increased apoptosis in both mature and
immature T cell populations from thymus, spleen, and bone marrow of
mice after a cecal ligation and puncture (92)
. Hotchkiss
also observed that the increased apoptosis in lymphoid organs appeared
to play a direct role in the adverse outcome to a cecal ligation and
puncture, since selective Bcl-2 overexpression in T lymphocytes or a
systemic administration of caspase inhibitors led to a decrease in
apoptosis in thymus and spleen, as well as increased survival
(90
, 93)
.
|
Although there is a consensus that apoptosis is increased in lymphoid
cell populations during sepsis syndrome (94)
, the humoral
or endocrine factors that stimulate this increased apoptosis are still
not fully known. For example, Ayala and his group have reported in
rodents that thymocyte apoptosis seen during polymicrobial sepsis is
primarily a direct response to corticosteroids and can be controlled
in vivo by the steroid receptor blocker mifepristone.
Furthermore, thymocyte-induced apoptosis was Fas ligand and TNF-
independent (95
, 96)
. We came to similar conclusions
regarding the increased apoptosis in thymus and spleen after a scald
burn injury, since treatment with mifepristone blocked the increased
apoptosis in these tissues (97)
. The increased apoptosis
was still seen in endotoxin-resistant mice (C3H/HeJ) and in mice
with deletions or mutations in TNF-
(tnf-/-) and FasL
(gld).
The role played by TNF- and FasL in sepsis-induced lymphoid cell
apoptosis remains controversial and may depend on the sepsis model
studied. TNF- appears to play a role in thymocyte-induced apoptosis
in animals injected with high-dose lipopolysaccharides
(98)
. This may also be cell type specific, since a decease
in both intraepithelial gut lymphocyte apoptosis and in mortality was
seen in FasL-deficient mice (gld), suggesting FasL to be an
important contributor to survival and the loss of lymphoid tissues from
the gut (99)
. An additional study also suggested FasL as a
mediator for splenocyte apoptosis, contributing to the depression of
splenocyte immune responses seen during polymicrobial sepsis
(100)
.
APOPTOSIS IN PARENCHYMAL CELLS
Although there is general agreement that apoptosis is
increased in lymphoid cells during sepsis, data supporting a
significant parenchymal cell apoptosis in solid organs during sepsis
are much less convincing. In a murine endotoxemia model, endogenously
produced TNF- induced massive hepatocyte apoptosis and death;
however, hepatocyte apoptosis required the simultaneous presence of
transcriptional inhibition by either D-galactosamine or actinomycin-D
(101)
. Caspase-3 activity was essential for
TNF--induced hepatic parenchymal cell apoptosis, since pretreatment
of mice with synthetic substrates for caspase-3 (DEVD-CHO) prevented
mortality (102)
. This increased hepatocellular apoptosis
was an important signal for the transmigration of primed neutrophils
sequestered in sinusoids (101
, 103)
. TNF- induced
neutrophil sequestration in hepatic sinusoids during sepsis and
endotoxemia, which was also associated with hepatocyte necrosis
(104)
.
Similarly, in a rodent model of endotoxemia after priming with P.
acnes, blocking Fas ligand and TNF- prevented hepatocyte
apoptosis and death (105)
. Moreover, overexpression of
hepatitis B antigen induced FasL-dependent hepatocyte apoptosis.
Despite the evidence that hepatocyte apoptosis is increased in some
models of experimental endotoxemia, hepatocyte apoptosis appears not to
be a common occurrence in other models of sepsis or endotoxemic shock.
High-dose endotoxin-induced mortality is not associated with
significant hepatocyte apoptosis, although hepatocyte necrosis is
widespread (106)
. In an animal model of generalized
peritonitis (cecal ligation and puncture), the presence of widespread
hepatocellular apoptosis is controversial. Hiramatsu was unable to see
any significant increase in hepatocyte apoptosis in mice after a cecal
ligation and puncture (107)
. We have examined apoptosis
and caspase-3 activities in the liver of mice after a cecal ligation
and puncture, and have found only very modest increases when compared
to endotoxin and transcriptional inhibition or to FasL activation (data
not shown). We recently surveyed mouse solid organs for increased
apoptosis 3–24 h after a scald burn injury and found no evidence of
increased solid organ apoptosis or caspase-3 activity in liver, lungs,
kidney, or skeletal muscle (91)
.
Hotchkiss and colleagues examined parenchymal cell apoptosis in
patients who had immediately died from sepsis and found increases in
apoptosis that were limited to epithelial cells of the gut
(87)
. Although apoptosis was markedly increased in gut
lymphoid tissues of the spleen and thymus, increased apoptosis was not
seen in parenchymal cells from either the heart, lungs, kidney, or
liver.
NEUTROPHIL APOPTOSIS DURING SEPSIS
Neutrophils are inflammatory cells with potent oxidative and
proteolytic potential that are usually the first line of defense
against invading pathogens. Activated neutrophils produce cytotoxic
factors leading to deleterious inflammatory processes, including tissue
injury (108)
.
While lymphoid cells are undergoing accelerated apoptosis, spontaneous
neutrophil apoptosis associated with sepsis or SIRS is delayed
(109
, 110)
. This decreased apoptosis is thought to be
important in enhancing tissue injury in ARDS, SIRS, and burn injury by
promoting a disbalanced tissue load of neutrophils and uncontrolled
release of toxic metabolites injurious to endothelial cells’
mitochondria and collagen (109
, 111)
. Bacterial products
and cytokines released during sepsis can delay neutrophil apoptosis and
delayed neutrophil apoptosis has been associated with severe clinical
sepsis (110
, 112)
. In trauma patients, neutrophil
apoptosis correlated with MODS in these patients (113)
.
Further evidence of decreased neutrophil apoptosis during sepsis was
shown in a study where plasma of patients with SIRS decreased apoptosis
in neutrophils from healthy individuals. Neutralization of GM-CSF and
addition of IL-10 attenuated the delayed apoptosis (114)
.
Delayed neutrophil apoptosis after exposure with GM-CSF appears to be
in part due to phosphorylation of Bad, which results in the
dissociation of Bad from other members of the Bcl-2 family, leading to
prolonged cell survival. Anti-inflammatory cytokines such as IL-10 can
reverse the anti-apoptotic effect of SIRS on neutrophil life span
(110
, 114)
, whereas proinflammatory cytokines like IL-1
and IL-6 prolong survival of neutrophils by inhibiting apoptosis
(115)
. IL-10 has been reported to end pulmonary
inflammation in vivo by promoting neutrophil apoptosis
(116)
. Glucocorticoids inhibit apoptosis of human
neutrophils, which might be another reason for the decreased apoptosis
in these cells (117)
.
THERAPEUTIC OPTIONS
Although anti-apoptotic therapies for septic patients are
presently unavailable, further exploration of this clinical approach
appears warranted. The pro- and anti-apoptotic pathways regulating T
cell and B cell death are rapidly being delineated, not only in
lymphoproliferative diseases, but also in acute inflammation
(83)
. Therapeutic targets critical for inducing lymphocyte
apoptosis include, among others, caspase-8, caspase-3, and PARP.
Alternatively, stimulating the overexpression of anti-apoptotic
proteins like Bcl-2, Bcl-xL, or IAP or inhibiting
the expression of pro-apoptotic proteins like Bid and Bax could be
appropriate therapeutic targets (Table 2
and Fig. 4
).
|
Several formidable challenges must be overcome in order to develop
anti-apoptotic therapies in the septic patient. These include
successfully targeting both the appropriate signaling pathway and the
specific cell population. There have been several attempts at blocking
the circulating mediators or humoral factors during sepsis that induce
apoptosis, like TNF-, FasL, granzymes, or glucocorticoids. Although
this is the most technically feasible approach, it has not proved
successful in treating the increased lymphocyte apoptosis that
characterizes sepsis syndromes. This is due partly to the observation
that there is considerable redundancy in actions among these mediators.
For example, TNF-, FasL, and glucocorticoids are all pro-apoptotic,
and their expression or release is increased acutely in sepsis
syndromes (13
, 96)
. In fact, we have shown that early
after a scald burn injury or generalized peritonitis in the mouse,
increased TNF- and FasL expression has been documented in the thymus
and spleen where increased T cell apoptosis is occurring. In these same
experimental models of sepsis and burn injury, however, blocking
TNF- and FasL had only minimal effects on lymphoid apoptosis in
spleen, thymus, and gut-associated lymphoid tissue. Inhibiting
glucocorticoid receptor binding was more effective (96)
.
Glucocorticoid blockade, unfortunately, is not a feasible approach for
treating the septic patient because of its obligatory role in
carbohydrate and energy homeostasis in stressful conditions.
A more appropriate schema is to target directly the specific intracellular pathways and effectors leading to sepsis-induced cell death, such as the caspases or PARP, whose activation may be the common product of mitochondrial or cytosolic apoptotic pathways. These more distal targets may provide greater specificity than the humoral mediators that induce them, since they represent common intracellular signaling pathways for diverse extracellular mediators.
One experimental approach to prevent apoptosis in animals has been to
forcibly overexpress endogenous proteins that interfere with apoptotic
processes in lymphoid tissues. Overexpression of the anti-apoptotic
Bcl-2 protein in T and B lymphocytes actually improved survival after a
cecal ligation and puncture, and prevented apoptosis in lymphoid organs
(118)
. Moreover, transgenic mice overexpressing Bcl-2 in
gut epithelial cells were resistant to ischemia-reperfusion injury in
the gut (119)
. In Fas-mediated acute liver injury,
apoptosis of hepatocytes can be decreased in vivo in
different ways. Bcl-2-overexpressing hepatocytes of transgenic mice
show decreased liver apoptosis (120)
. Also, Bid-deficient
mice have a higher survival rate and a decreased hepatocellular liver
injury after injection of ant-Fas antibody compared to wild-type mice
(72)
. Deletion of either pro-apoptotic protein suggests a
beneficial role. Furthermore, caspase inhibitors such as ZVAD have been
shown to efficiently block Fas-mediated liver destruction
(121)
. Alternative approaches include overexpression of
the natural inhibitors of caspases, including the IAP family of
proteins.
The second major challenge will be to target only the
lymphocyte and transiently block the increased apoptosis
only during sepsis. Since apoptosis is a normal
physiological process essential for the death and removal of several
cell populations of lymphocytes, delayed apoptosis has been implicated
in the malignant transformation of several cell populations, including
B lymphocytes (122)
. In addition, delayed apoptotic
removal of neutrophils has been imputed in the pathogenesis of adult
respiratory distress syndrome (111)
. Inhibition of
neutrophil apoptosis through NF-B-dependent pathways in sepsis or
SIRS appears to prolong the life of neutrophils once they have
extravasated the blood compartment into lung parenchymal tissues and to
potentially increase oxidative damage in the lung (123)
.
Thus, targeted blockade of apoptosis in lymphocyte populations
must be specific enough to target primarily those lymphocyte cell
populations undergoing increased apoptosis, and to be sufficiently
transient to prevent the risk of malignant transformation associated
with prolonged blockade of cell death.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONAPOPTOSIS DURING INFLAMMATIONCONCLUSIONSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONAPOPTOSIS DURING INFLAMMATIONCONCLUSIONSREFERENCES |
|---|
