(Journal of Leukocyte Biology. 2001;70:18-29.)
© 2001
by Society for Leukocyte Biology
Unlocking the secrets of cytotoxic granule proteins
Mark J. Smyth*,
Janice M. Kelly*,
Vivien R. Sutton*,
Joanne E. Davis*,
Kylie A. Browne*,
Thomas J. Sayers
and
Joseph A. Trapani*
* Cancer Immunology Division, Trescowthick Laboratories, Peter MacCallum Cancer Institute, Melbourne, Australia; and
Laboratory of Experimental Immunology, National Cancer Institute, FDR-DC, NIH, Frederick, Maryland
Correspondence: Mark J. Smyth, Cancer Immunology Division, Trescowthick Laboratories, Peter MacCallum Cancer Institute, Locked Bag 1, A’Beckett St, 8006, Melbourne, Australia. E-mail: m.smyth{at}pmci.unimelb.edu.au
 |
ABSTRACT
|
|---|
Cytotoxic lymphocytes largely comprise CD8+ cytotoxic T
cells and natural killer cells and form the major defense of higher
organisms against virus-infected and transformed cells. A key function
of cytotoxic lymphocytes is to detect and eliminate potentially harmful
cells by inducing them to undergo apoptosis. This is achieved through
two principal pathways, both of which require direct but transient
contact between the killer cell and its target. The first, involving
ligation of TNF receptor-like molecules such as Fas/CD95 by their
cognate ligands, results in mobilization of conventional, programmed
cell-death pathways centered on activation of pro-apoptotic caspases.
This review concentrates on the second pathway, in which the toxic
contents of secretory vesicles of the cytotoxic lymphocyte are secreted
toward the target cell, and some toxins penetrate into the target cell
cytoplasm and nucleus. In addition to invoking a powerful stimulus to
caspase activation, this "granule-exocytosis mechanism" provides a
variety of additional strategies for overcoming inhibitors of the
caspase cascade that may be elaborated by viruses. The key molecular
players in this process are the pore-forming protein perforin and a
family of granule-bound serine proteases or granzymes. The molecular
functions of perforin and granzymes are under intense investigation in
many laboratories including our own, and recent advances will be
discussed. In addition, this review discusses the evidence pointing to
the importance of perforin and granzyme function in pathophysiological
situations as diverse as infection with intracellular pathogens, graft
versus host disease, susceptibility to transplantable and spontaneous
malignancies, lymphoid homeostasis, and the tendency to auto-immune
diseases.
Key Words: apoptosis • lymphocyte • granzyme • perforin • granulysin
|
THE GRANULE EXOCYTOSIS MODEL
|
|---|
Cytotoxic T lymphocytes (CTL) and natural killer (NK) cells play a
key role in immune responses toward virus-infected and malignant cells
[1
]. The granule exocytosis model describes one pathway
by which these effector cells act upon target cells (reviewed in
[2
3
4
5
6
7
]). Recognition and binding between a cytotoxic
lymphocyte and its target lead to vectorial exocytosis of specific
cytoplasmic granules toward the target cell [5
6
7
].
Delivery of this "lethal hit" is known to occur via proteins
contained within the granules [8
9
10
], many of which
interact with or enter the target cell. This brings about cell death
typified by cytosolic and nuclear-apoptotic changes. These cytotoxic
granules are quite complex organelles that not only contain
pro-apoptotic proteins restricted to CTL and NK cells
[11
, 12
] but also others ubiquitously
expressed within lysosomes [11
]. Thus, cytotoxic
lymphocytes package secretory and lysosomal proteins within a
specialized, cytotoxic organelle, which serves as a secretory lysosome
(Table 1
)
|
DENSE CORE GRANULE COMPONENTS— A SUMMARY
|
|---|
The dense core of a cytotoxic granule contains the
proteins responsible for target cell destruction [11
].
These include perforin (pfp; otherwise termed pore-forming protein or
cytolysin), calreticulin, and many serine proteases or granzymes
(granule-associated enzymes; grz) [11
, 13
,
14
], which are complexed with chondroitin sulfate-rich
proteoglycans. Other than calreticulin, the expression of these
proteins is largely restricted to activated CTL and NK cells
[13
, 14
].
Pfp
Pfp is a pore-forming protein with patchy homology to the
C9 complement component [15
] and is synthesized as a
70-kDa inactive precursor, which is cleaved at the C-terminus to yield
a 60-kDa active form. This processing occurs in an acidic compartment
(pH 5.5), and agents such as concanamycin A that disrupt granule
acidification prevent pfp processing [16
,
17
]. Cleavage of the pro-piece occurs at the boundary of
a synaptotagmin-like C2 domain at the pfp carboxyl terminal; this
domain is then capable of binding the plasma membrane and initiating
pore formation in a calcium-dependent manner [16
]. The
clipped pro-piece is short (12–20 residues) but contains a bulky
glycan attached at 528Asn. Its removal permits pfp monomers
to undergo a conformational change in the presence of calcium ions,
which allows binding of phosphorylcholine groups on membrane lipids and
coalescence with other pfp molecules to form poly(pfp) pores
[16
, 18
19
20
]. Pfp insertion into the target
cell membrane is a stimulus that amplifies the endocytic uptake of
other granule constituents and their delivery into the target cell
cytosol. Thus, now it seems that the formation of large transmembrane
pores is not necessary for uptake of other granule constituents
[21
]. However, the specific molecular basis for this
process is not well understood.
Granulysin
Granulysin is a member of the saposin family of lipid-binding
proteins [22
23
24
]. Although functionally related to
other anti-bacterial peptides, defensins and magainins, granulysin is
structurally distinct. It is active against a broad range of microbes,
including Gram-positive and Gram-negative bacteria, fungi, and
parasites. Two major protein products of 15 and 9 kDa are encoded by
the granulysin gene. Recombinant 9-kDa granulysin disrupts artificial
liposomes and cell membranes, damages mitochondria, and activates
caspase-9 to induce apoptosis in nucleated cells [25
].
Identification of this molecule indicates a broader and perhaps more
significant role for killer cells in innate and acquired anti-microbial
defenses. Thus far, granulysin has been shown to kill extracellular
Mycobacterium tuberculosus directly by reducing its membrane
integrity and to decrease the viability of intracellular M.
tuberculosus by a pfp-dependent mechanism [26
].
Grz
Grz are serine proteases belonging to the chymotrypsin superfamily
[27
, 28
]. The major features that define
this family of serine proteases have been reviewed previously
[29
, 30
]. The crystal structure of grzB has
been solved now, and the overall similarity between grz and
chymotrypsin, verified [31
]. Quite a number of grz and
their genes have been characterized in mouse, rat, and human CTL and NK
cells (Table 2
) These can be divided into three subfamilies on the basis of their
gene structure (indicating their evolutionary relationships),
proteolytic specificity, and biological functions. GrzH is the product
of a gene fusion unique to humans [32
] and is the only
known human grz with chymase activity [33
]. As yet,
human counterparts of mouse grzC to -G have not been isolated. All grz
are produced as preproproteins, the "pro" segment usually
comprising an acidic, inactivating dipeptide [13
,
14
, 34
, 35
]. The leader
sequence enables the nascent grz to be processed through the
endoplasmic reticulum (ER) and Golgi apparatus during synthesis so that
they may be ultimately targeted to the secretory pathway
[13
, 14
]. The glycosylation of grz is quite
heterogeneous, but generally, their packaging into cytotoxic granules
is dependent on the mannose 6-phosphate pathway [36
].
Below, we will discuss how grz induce apoptotic death, however much
experimental evidence indicates additional, potential functions for
grz.
Calreticulin
Calreticulin is a calcium-storage protein that carries a
COOH-terminal KDEL sequence for retention in the lumen of the ER.
However, in CTLs, calreticulin also colocalizes with pfp and is
released with it on degranulation [37
]. Pfp interacts
with the P-domain of calreticulin, which has high,
Ca2+-binding affinity and chaperone function,
but the molecules dissociate upon exposure to higher, extracellular
Ca2+ concentrations. Therefore, calreticulin
may act as a Ca2+-regulated chaperone for pfp,
potentially serving to protect the killer cell during biogenesis of
granules and regulating pfp function after release [38
].
Chondroitin sulfate proteoglycans
Chondroitin sulfate proteoglycans are protease-resistant, possess
chondroitin sulfate A side-chains, and are specifically exocytosed upon
contact with sensitive targets [39
, 40
].
These negatively charged macromolecules may regulate the packaging and
delivery of pfp and positively charged (basic) grz [41
].
Recently, it has been proposed that polyvalent uptake of grz and other
molecules by target cells may occur as a result of their attachment to
serglycin-rich moieties [42
].
Chemokines
Chemokines exocytosed by CTL are believed to be important in the
noncytolytic inhibition of the human immunodeficiency virus (HIV)-1
replication. One study has shown that grzA, macrophage inflammatory
protein (MIP)-1
, and RANTES (regulated on activation, normal T
expressed and secreted) are localized in the cytolytic granules of
HIV-1-specific CD8+ CTL [43
]. These
mediators are co-secreted after T-cell receptor (TCR) triggering,
facilitating lysis of virion-producing cells and the inhibition of free
virus. In addition, RANTES, MIP-1, and MIP-1ß are secreted by CTL
as a macromolecular complex containing sulfated proteoglycans,
providing a potentially rapid response to pathogens, using preformed
and prepackaged chemokines.
|
GENERIC, LYSOSOMAL COMPONENTS
|
|---|
The multivesicular (peripheral) domains of cytolytic granules
contain lysosomal hydrolases, including acid phosphatase,
-glucosidase, arylsulphatase, ß-glucoronidase, cathepsins B and D,
cathepsin A-like protective protein (CAPP; which possesses serine
carboxypeptidase and deamidase activities), and lysosomal membrane
proteins, Lamp-1, Lamp-2, and CD63. In addition to their protein
content, the lytic granules have other properties in common with
lysosomes. This region has an acidic pH, comparable with that of
endosomes and lysosomes. The multivesicular domains of the granules are
rich in the 270-kD mannose-6-phosphate receptor, normally absent from
mature lysosomes but present in earlier endocytic compartments. Thus,
the granules represent an unusual, dual-function organelle, where a
regulated secretory compartment, the dense core, is contained within a
prelysosomal compartment, the multivesicular domain. How the biogenesis
of the "secretory lysosome" differs from that of a conventional
secretory granule is unclear, however a combination of lysosomal and
other sorting signals appears to be required. Similar organelles are
also found in other hemopoietic subsets of cells. Therefore,
hemopoietic cells may possess specialized mechanisms that allow the
correct sorting of secreted products to the lysosome, and these signals
may differ from those in conventional secretory (e.g., neurosecretory)
cells. Studies on Chediak-Higashi syndrome (CHS) patients support the
idea that granules are specialized, secretory lysosomes, because their
hemopoietic cells are unable to secrete their granule contents, and
their conventional, secretory cells are able to do so
[36
]. CTL from CHS patients cannot secrete the giant
granules in which their cytotoxic proteins are stored, thereby
suggesting that the defect lies in protein sorting or membrane fusion.
Dipeptidyl peptidase I (DPPI; cathepsin C) is a lysosomal, cysteine
protease and major posttranslational-processing enzyme responsible for
generating activated myeloid and lymphoid granule serine proteases
[44
]. DPPI was first shown definitively to process and
activate human grzB [45
], and subsequently, the
generation of DPPI-/- mice indicated that DPPI plays an
essential role in the in vivo processing and activation of
grzA and -B [46
].
|
FUNCTIONS FOR CYTOTOXIC GRANULE PROTEINS
|
|---|
Target cell death
Pfp
The crucial role of pfp became evident when several laboratories
created pfp-deficient, gene knock-out mice (pfp-/-)
[8
, 9
, 47
]. These mice have
normal T- and NK-cell development [8
], however their
cytolytic lymphocytes are compromised, and the mice are highly
susceptible to certain intracellular pathogens [8
,
9
, 48
]. In vitro, pfp-deficient
CTL and NK cells are defective in their killing of allo-reactive or
xeno-reactive tumor cell lines and NK-sensitive target cells,
respectively [49
, 50
]. Despite its clear
biological importance, little is known about the molecular function of
pfp, and only recently have dogmas concerning its mechanism of action
been questioned. The inability thus far to assign pfp functions to
discrete parts of the molecule represents a major gap in our
understanding of effector lymphocyte biology. Pfp shares functional,
antigenic, and ultrastructural similarities with complement proteins
C6–C9, as described some time ago, and functions such as membrane
insertion and polymerization have been tentatively explained on the
basis of these similarities [51
52
53
54
]. As already
discussed, definitive evidence has emerged that the C-terminal domain
of pfp is the site of calcium ion-binding and initiates lipid insertion
[16
]. Based on the structure of synaptotagmin and
related molecules, it is postulated that after cleavage at the carboxy
terminal, multiple aspartate residues of the C2 domain become
approximated in three dimensions to bind a calcium ion
electrostatically. The refolded pfp molecule becomes highly reactive
with lipids as a result of exposure of amphipathic domains elsewhere in
the molecule and is able to attach to and be inserted in the plasma
membrane.
A pfp receptor?
Although pfp can form transmembrane channels in synthetic, lipid
membranes with no protein content, it has long been suspected that, for
example, on the basis of markedly differing susceptibility of various
cell types to pfp, factors other than lipid composition can regulate
the activity of pfp. The recent studies of Berthou et al.
[55
] have provided some evidence of this proposition.
These investigators have proposed that the lysolipid PAF (known as
platelet activating factor because it promotes platelet aggregation) is
co-released with pfp when NK cells degranulate and may potentiate pfp
lysis by forming a molecular bridge between pfp and the PAF receptor
(PAF-R) Thus, ternary complexes containing PAF, PAF-R, and pfp may
achieve membrane disruption more efficiently by enhancing the binding
of pfp to phosphorylcholine. Furthermore, as PAF-R expression is
inducible with interferons (IFNs), local, inflammatory mediators may
modulate the sensitivity of cells to pfp at a focus of infection. If
verified, the studies by Berthou et al. [55
]
may also provide an opportunity to manipulate pfp function for
experimental and even therapeutic purposes.
What does pfp actually do?
As with complement, purified pfp can cause cell lysis by forming
discrete pores 12–20 nm in diameter but, by itself, cannot account for
the morphological changes of apoptosis, such as chromatin condensation
and DNA fragmentation [56
]. The nuclear changes occur
before cell membrane damage [57
,
58
] and are reproducible in vitro with
concentrations of pfp that cause minimal, membrane-permeability changes
[21
]. Thus, the proposition that large membrane pores
are necessary for grz access to the target cell cytosol has come into
question recently. Despite the structural similarities of the pores
formed by purified pfp and complement, grzB did not trigger apoptosis
of target cells when delivered by complement [21
].
However, listeriolysin (LLO), a virulence factor of Listeria
monocytogenes (LM) that causes lysis of endosomes, was able to deliver
grzB potently in the absence of measurable plasma-membrane damage
[21
]. The pro-apoptotic activity of LLP was inhibited
when the pH of endo-lysosomes was raised to neutral with ammonium
chloride or bafilomycin [21
]. Direct support for an
endosomolytic function of pfp came from the observation that brefeldin
A (BFA) inhibited pfp-induced release of grzB from endosomes, blocked
its translocation to the nucleus, and inhibited cell death. Consistent
with BFA having no effect on receptor-mediated uptake via endocytosis,
BFA had no effect in kinetic or absolute terms on grzB uptake into the
cell in the absence of pfp [21
].
Because the concentration of pfp delivered to the target cell surface
by a CTL has not been determined, a key question is whether
endosomolysis is relevant physiologically at pfp concentrations causing
appreciable cell-membrane damage. We have found that when target cells
are incubated with concentrations of pfp, which caused 100%
51Cr release, freely diffusible, fluorescent proteins of
9–13 kDa remain excluded for over 1 h, whereas 32-kDa grzB and
65-kDa grzA continued to be delivered to the cytoplasm and nucleus
within a few minutes [21
]. This indicates the delivery
of apoptotic mediators by pfp is indeed highly selective, and thus, grz
entry into cells cannot be ascribed to passive diffusion simply through
pfp pores, even if target cells are placed under severe, osmotic
stress.
Grz
Previous reviews have described the gene structure, chromosomal
loci, and protease specificity of grz family members [29
,
30
, 59
, 60
]. The major, recent
advance in terms of grz structure has been the description of a crystal
structure for grzB in complex with a macromolecular inhibitor. The
primary specificity for Asp residues occurs through a side-on
interaction with a "buried" side chain of Arg226 of
granzyme B. A further nine amino acids make contact with the substrate
and dictate the extended substrate-specificity profile
[31
]. There has also been considerable progress in
further elucidating the role of grz in apoptosis [60
].
Early studies unequivocally illustrated that grzA and B could
collaborate with pfp to kill target cells [57
,
61
, 62
]. The most potent granzyme in this
context is grzB, and grzB-deficient mice have impaired ability to
induce rapid DNA fragmentation in the target cell [63
].
Two other grz were purified, which induce target cell DNA fragmentation
with much slower kinetics, and these were the tryptases grzA and
tryptase-2 [57
]. However, grzA-deficient mice have no
generalized defect of target cell DNA fragmentation, indicating that
unlike grzB, the absence of grzA can be compensated by other grz
[64
]. We and others [65
66
67
68
] have shown
that grzB, abundant in cytolytic granules, is the protease largely
responsible for eliciting the nuclear changes of apoptosis. We and
several other groups [67
68
69
] have shown that grz can
enter cells independently of pfp but remain sequestered in endosomes
and so do not damage cells unless pfp is also present. Thus, the
provision of a membrane pore by pfp is not necessary for grz to enter
the target cell cytoplasm. Cell-surface binding of
125I-grzB is saturable and can be competed by unlabeled grz
[69
], suggesting uptake through a specific receptor.
Recently, the 270-kDa, cation-independent, mannose 6-phosphate receptor
was demonstrated to be a receptor through which grzB can enter the
endosomal compartment [70
]. Most surprisingly, it was
claimed that expression of the same receptor on H2k-expressing
fibroblasts was required for their rejection by allogeneic T cells,
suggesting a primary role for grzB in the allogenic-effector response
and predicting a further, possible mechanism for immune escape by
tumors [70
].
GrzB function in cell death: a crucial role for Bid-cleavage rather
than direct caspase activation
Sublytic pfp can induce the redistribution of grzB or dimeric grzA
(65 kDa) from endosomes into the cytosol and can amplify greatly
cellular uptake of grz [67
]. Apoptotic changes are
apparent within only 2 min, and migration of grzB out of endosomes and
its appearance in the nucleus are precise predictors of apoptotic death
[65
]. In addition to inducing programmed, cell-death
pathways operating through caspases, grzB can directly cleave
cytoplasmic substrates such as the actin-binding protein, filamin
[71
], and nuclear poly(ADPribose) polymerase (PARP) and
nuclear matrix antigen at sites different than those preferred by
caspases [72
]. Grz entry into the nucleus occurs before
apoptotic, nuclear-membrane disruption [58
] and is
dependent on an unknown cytosolic carrier protein but does not require
expenditure of energy [67
].
Through its unique ability to cleave after aspartate residues, grzB can
cleave many pro-caspases in vitro [73
74
75
].
However, in intact cells, there is a requirement for mitochondrial
perturbation [76
, 77
], without which direct
caspase activation occurs only very slowly [78
]
(Fig. 1
) We showed recently that a vital part of the apoptotic signal
imparted by grzB to mitochondria is through direct cleavage of the
pro-apoptotic, BH3-only, Bcl-2 family member Bid, which is specifically
cleaved at a site 16 amino acids down-stream of that used by caspases
[78
]. Surprisingly, grzB can also induce death through a
caspase-independent mechanism that involves damage to non-nuclear
structures and is probably mediated by direct, grz-mediated disruption
of mitochondria [66
, 79
, 80
].
These caspase-independent pathways may safeguard against viruses that
delay programmed cell death by expressing serpins such as the caspase-8
inhibitor of cowpoxvirus, cytokine response modifier A (crmA)
[81
].
View larger version (21K):
[in this window]
[in a new window]
|
Figure 1. A simplified scheme for cell death mediated by grzB. Following the
entry of grzB into the target cell by endocytosis and its liberation
into the cytosol by pfp, the principal role of grzB (thick arrow) is to
cleave the pro-apoptotic, Bcl-2-like molecule Bid, specifically at the
grzB cleavage site at amino acid 75, which is strongly conserved
between mouse and humans. Direct processing of caspases by grzB is
inefficient in the absence of mitochondrial perturbation (dotted line).
Truncated Bid disrupts mitochondria, resulting in the release of
pro-apoptotic mediators, the precise nature of which is unknown at
present (dashed arrow). These mediators may include Diablo/Smac and
posssibly cytochrome c and induce efficient caspase processing and
generic caspase-mediated cell death. Insertion of truncated Bid and
release of apoptotic mediators from mitochondria can be negatively
regulated by overexpression of Bcl-2. Cell death in response to grzB
can also occur without efficient caspase activation through a
non-nuclear mechanism, which is poorly understood but may involve
cleavage of cytoskeletal elements (dashed arrow; for a deeper
explanation, see text and ref. [76
]).
|
|
GrzA function in cell death
GrzA is a specific tryptase, which becomes concentrated in the
nucleus of targeted cells and synergistically enhances DNA
fragmentation induced by grzB and pfp [58
]. Little is
known about grzA-induced cell death. Using recombinant, mutant,
inactive grzA, two grzA-binding proteins, PHAP (putative HLA-associated
protein) II and heat shock protein (hsp)27, were isolated
[82
], however neither of these has been demonstrated to
be important yet in apoptosis mediated by grz. GrzA-induced membrane
perturbation, nuclear condensation, and DNA damage are unimpaired by
caspase blockade, however entry of grA to the nucleus is totally
inhibited by Bcl-2 overexpression [83
]. Recently, grzA
was shown to induce single-strand DNA breaks rather than
oligonucleosomal fragmentation [84
]. GrzA enhances DNA
accessibility to exogenous endonucleases and degrades histone H1
in vitro into 
16-kDa fragments. Histone digestion may
provide a mechanism for unfolding compacted chromatin to facilitate
endogenous DNase access to DNA [85
]. CTL and NK cells of
mice deficient in grzA and grzB (grzAB-/-) induced
51Cr-release in target cells at levels and with kinetics
similar to those of normal mice [86
]. This contrasts
with their inability to induce apoptotic nuclear damage in target
cells, suggesting that grzA and -B are critical for CTL/NK
granule-mediated nucleolysis, with grzB being the main contributor, and
target cell death is pfp-dependent and -independent of both proteases.
|
GRANZYME SERPINS
|
|---|
Serprins contol many physiological processes by the balance of
serine-protease activities and their regulated blockade. Serpins, such
as crmA, form irreversible complexes with their cognate proteases by
acting as pseudo-substrates. The inhibitory loop of these serpins
contains sequences recognized specifically by the protease. Primary
sequence recognition is encoded by the P1 residue, but neighboring
residues upstream (P2, P3, etc.) or downstream (denoted P1', P2', etc.)
of P1 can also influence recognition and cleavage. The two ends of the
inhibitory loop are flexible, hinge-like structures, which enable the
loop to become mobile after cleavage by the cognate protease,
commencing a conformational change that locks the two molecules into a
complex often strong enough to withstand boiling in sodium dodecyl
sulfate (SDS)-containing buffer. Evidence has emerged recently that
cytotoxic lymphocytes synthesize their own serpins, which act within
the cytosol, to safeguard against missorting or mispackaging of
pro-apoptotic granzymes. The proteinase inhibitor (PI) residue of the
CTL/NK serpin PI-9 is Glu, allowing it to inhibit grzB
specifically [87
]. The choice of Glu at the P1 residue
seems puzzling at first because grzB prefers cleaving at Asp in most
instances [88
]. Bird and colleagues [89
]
showed that mutation of the P1 residue to Asp resulted in poor complex
formation with grzB and, furthermore, that the mutated molecule
acquired a crmA-like ability to inhibit caspases, which wild-type PI-9
does not possess. Therefore, PI-9, which is absent from cytotoxic
granules but present in high concentrations in the cell cytosol, can
block toxic grzB molecules, which leak out of granules, without
inhibiting physiological death of the CTL occurring through the Fas
pathway [89
]. Many new intracellular serpins have been
described recently [90
] in CTL/NK cells, and it is
likely that effector cells are armed with appropriate, protective
serpins specific for all of their grz.
Previous observations that the highly conserved, poxvirus-encoded
serpins inhibit cytotoxic activities of alloreactive CTL via granule
and/or Fas-mediated pathways were taken to indicate their involvement
in immune evasion by poxviruses. The striking similarities between crmA
and PI-9 suggest that viral products such as crmA may have arisen from
the capture and mutation of homeostatic elements such as PI-9 or a
related serpin expressed in infected cells. The data in support of the
paramount importance of CTL and its effector molecule perforin in the
recovery from primary ectromelia virus infection question the role of
serpins in the evasion of CTL killing [91
92
93
]. Further
analysis of poxvirus interference with target cell lysis by
alloreactive CTL revealed that suppression affects the Fas-mediated
and, to a lesser extent, the granule exocytosis pathway primarily and
that serpin-2 is the main contributor to suppression for both killing
pathways [93
].
|
WHAT TYPES OF IMMUNE RESPONSE ARE DEPENDENT ON CYTOTOXIC GRANULE
PROTEINS?
|
|---|
Control of virus infections
Analysis of pfp-deficient mice has identified pfp as the
preeminent effector molecule in T-cell-mediated control of many virus
infections. Lymphocytic choriomeningitis virus (LCMV)-specific CTL are
responsible for virus eradication and the onset of pathology associated
with the disease, depending on the timing, route of entry, and strain
of the virus [8
]. CD8+ T cells are activated
by LCMV in pfp-/- mice but fail to clear the virus
effectively. Rather, a large, proliferative expansion and persistence
of antigen-reactive T and B cells and antigen-presenting cells (APC)
occur in pfp-deficient mice challenged with LCMV. This presentation
bears a striking resemblance to that of pediatric patients with the
autosomal-recessive immunodeficiency, familial hemophagocytic
lymphohistiocytosis (FHL), about half of whom have been shown to lack
NK cell activity as a result of inherited, structural mutations in the
pfp gene [94
]. It is possible that human FHL is
triggered by a virus, which is normally cleared in a pfp-dependent way.
Previous studies have established that NK/CD8+ T
cytotoxicity was not essential to resolve most cytopathic poxvirus
infections and that secretion of IFN-
by CD4+ and
CD8+ T cells was crucial in immunity against poxviruses.
However, a lack of pfp renders the relatively resistant C57BL/6 mice
highly susceptible to the natural mouse pathogen ectromelia, a
cytopathic orthopoxvirus [91
]. Pfp-deficient mice showed
increased mortality, elevated virus titers, increased cytopathic damage
in their liver and spleen, and increased circulating liver transaminase
levels [91
]. It is interesting that mice deficient in
grzA and -B were virtually as susceptible as pfp-deficient mice to this
virus, and mice deficient in either grz alone were susceptible only
partially [92
]. Death of grzAB-/- mice
occurred despite the expression of functionally active pfp and the
absence of an intrinsic defect in generating splenic, cytolytic T
cells.
The increased sensitivity of grzAB-/- mice to ectromelia
is the most significant phenotype that grz-deficient mice display. It
has been suggested that grz are important effector molecules in this
setting, but it remains unclear why grz are so important in host
protection from ectromelia. In contrast to ectromelia, cowpox virus is
more virulent in the presence of pfp than in its absence. An additional
lack of grzA increases the virulence of cowpox virus. NK cells and
CD8+ CTL also have a protective role against cytopathic
murine cytomegalovirus (MCMV) infection. Spleen NK cells control MCMV
infection in a pfp-dependent manner [95
], however in the
liver, production of IFN- by NK cells was the predominant mechanism
that regulated MCMV DNA synthesis. More recent data extend previous
studies on the critical role of NK/CD8+ T cells in the
early control of MCMV infection by showing that pfp and grzA and -B
contribute to viral elimination in the salivary glands, however none of
these molecules alone was essential for final control of infection
[96
]. Control of ganglionic herpes simplex virus (HSV)
infection depends on CD8+ cells but not on the death of
infected neurons. It has been shown recently that grzA restricts the
interneuronal spread of HSV and influences ganglionic virus load
significantly. Thus, several other virus models are emerging where grz
may play a more subtle role in controlling viral load, and these
mechanisms may not involve cell death of virus-infected cells.
Theiler’s virus, a murine picornavirus, infects the central nervous
systems of C57BL/6 mice and is cleared in a pfp-dependent process,
which requires CD8+ cytotoxic T cells [97
].
Some controversy exists as to the role of pfp in HSV infection. One
study indicates pfp is essential for host protection against ocular HSV
challenge, but not herpetic stromal keratitis (HSK), an inflammatory
disease of the cornea that often results in blindness
[98
]. Another study suggested that pfp-dependent
cytotoxicity is an important effector mechanism in the production of
HSK, but viral clearance from the eyes of pfp-/- mice was
not impaired [99
]. By contrast, these findings show that
pfp is sometimes important in the pathogenesis of viral infection
rather than viral clearance per se. Further, supporting an important
role for pfp in the pathogenesis of infection is the exacerbation of
Coxsackievirus B3-induced myocarditis by pfp [100
] and
pfp-mediated immunopathology in IFN--/- mice infected
with LCMV [101
].
|
CONTROL OF BACTERIAL INFECTION
|
|---|
Granule protein-mediated control of bacterial infections has not
been widely studied, however many intracellular bacteria are controlled
in part in a pfp-dependent way. Kagi et al.
[8
] first demonstrated that the absence of pfp-mediated
cytotoxicity resulted in the delayed clearance of LM from the spleen
but not the liver after primary infection. Clearly, pfp-dependent and
-independent mechanisms of CD8+ T-cell-mediated clearance
of Listeria were evident, however protection against a secondary
infection was impaired drastically in pfp-/- mice.
Subsequently, immunity for several Listeria antigens was shown to be
mediated by pfp-expressing CD8+ T cells. Analysis of
epitope-specific CD8+ T-cell expansion by major
histocompatibility complex (MHC) class I tetramer staining and
enzyme-linked immunospot (ELISPOT) revealed no deficiency in the
primary or secondary response to LM infection in pfp-/-
mice [102
]. These data show that reduced resistance to
LM observed with pfp-/- mice is a consequence of a
deficiency in effector function and not a result of suboptimal,
CD8+ T-cell priming. Recently, the same group has
demonstrated that pfp-independent immunity in the spleen requires
CD8+ T cell-derived tumor necrosis factor (TNF)
[103
]. The second microorganism of great interest has
been M. tuberculosus. CTL have been shown to be protective
against M. tuberculosus infections in the mouse, and these
effectors are cytolytic toward M. tuberculosus-infected
cells and release IFN- in response to mycobacterial antigen. CTL
have also been shown to kill intracellular pathogens by a
granule-dependent mechanism involving perforin and granulysin, as
described above [26
].
Pfp plays a role in host protection against several other
microorganisms. For example, Encephalitozoon cuniculi is a
protozoan parasite shown recently to cause opportunistic infection in
immunocompromised individuals. Protective immunity in the normal host
is CD8+ T-cell-dependent, and pfp-/- mice are
particularly sensitive to parasite challenge [104
].
CD8+ T cells have also been shown to be required for acute
resistance to infection with the protozoan parasite, Trypanosoma
cruzi, the causative agent of Chagas’ disease. Although
pfp-dependent cytolytic mechanisms can clearly affect acute resistance
to T. cruzi infection, this contribution may be strain- and
challenge-dose-dependent [105
]. Pfp also plays a limited
role in host resistance to Toxoplasma gondii, particularly
during the chronic stage of infection [106
]. The
clearance of several other microorganisms tested including
Chlamydia pneumoniae and Chlamydia trachomatis
was not pfp-dependent.
|
CONTROL OF TUMOR GROWTH
|
|---|
Tumor cells of different tissue origins have now been
characterized for their sensitivity to pfp-mediated cytotoxicity, and
in general, cytolytic lymphocytes kill the majority of these tumors in
a pfp-dependent manner [107
, 108
]. Some
tumor cells, such as acute myeloid leukemias, may be able to
protect themselves from pfp-mediated cell death through impaired
binding of pfp [109
]. Pfp-/- mice were
challenged with syngeneic lymphoid tumors that were MHC class
I-deficient and were shown to be controlled by NK cells in a
pfp-dependent manner [110
]. Subsequently, we supported
this study by demonstrating that pfp accounts for all the effector
function of the NK cells mediating rejection of MHC class I-deficient
lymphoid tumors in the peritoneal cavity [111
]. van den
Broek et al. [112
] also challenged naive
pfp-/- mice with syngeneic tumor cell lines of various
tissue origin. Most of the tumors were rejected ten- to 100-fold more
efficiently by wild-type mice, and the difference between wild-type and
pfp-/- mice was more marked following priming. Others
have demonstrated the relative importance of pfp in graft versus
leukemia effects following transplantation [113
114
115
].
Adoptive transfer or biological immunotherapies that stimulate
NK/NKT and CTL responses have also been demonstrated to mediate
their anti-tumor activities via pfp [116
117
118
]. These
studies in experimental settings underline the importance of tumor
rejection via the pfp pathway.
Despite the abundance of information that supports a key role for pfp
in host immunity against experimental tumors, the role of CTL and NK
cells, and in particular cytotoxicity, in tumor immune surveillance has
remained a controversial question. Pfp-/- mice have
provided an ideal model in which to revisit this issue. Approximating
models of spontaneous tumor formation, van den Broek et al.
[112
] demonstrated that pfp-/- mice were
more susceptible to sarcoma induction than wild-type mice after
receiving the chemical carcinogen, methylcholanthrene (MCA) or
oncogenic Moloney sarcoma virus. Subsequently, we have explored
MCA-induction of sarcoma in pfp-/- and other
gene-targeted mice. Clearly, there is a role for NKT and NK cells in
host protection, and pfp is not the only protective effector mechanism
used by the immune system [119
]. We also examined
whether tumorigenesis would be accelerated in tumor-prone,
p53-deficient mice that also lacked pfp expression. From this study, we
were able to show that pfp-deficient mice have a high incidence of
malignancy in several distinct lymphoid cell lineages (T, B, NKT),
indicating a specific requirement for pfp in protection against
lymphomagenesis [120
]. These highly malignant neoplasms
were strongly rejected by pfp-expressing mice, indicating that the
problem in pfp-deficient mice was poor tumor destruction rather than
inefficient tumor cell recognition or antigen processing. The
susceptibility to lymphoma was accelerated by a simultaneous lack of
expression of the p53 gene. This was the first study to demonstrate
that lymphocyte-mediated cytotoxicity plays an important role in
promoting host resistance to spontaneous tumor formation. Although the
pfp-/- mice in our previous aging studies have not been
exposed to LCMV nor do they develop histiocytic infiltrates early in
life, it is possible that other microorganisms might provide antigenic
stimulation also leading to lymphoproliferation, extended cell
survival, and an increased pool of cells vulnerable to oncogenesis in
pfp-/- mice. Alternatively, some recent experimental
evidence suggests that pfp may normally control B cell hyperplasia
[121
], suggesting that pfp also plays an
immunoregulatory role (see below). In any event, the increased lymphoma
incidence associated with pfp deficiency might have its genesis in an
increased number of premalignant cells and the absence of the cytotoxic
mechanism that normally eliminates them.
|
HOMEOSTASIS OF THE LYMPHOID COMPARTMENT
|
|---|
To prevent uncontrolled expansion, the massive proliferation of T
cells during an acute immune response has to be followed by their
controlled deletion. Mounting evidence now supports an immunoregulatory
role for the granule exocytosis pathway of cell death. The earliest
indication of this role was the exacerbated lymphoid expansion of mice
doubly deficient for pfp and the apoptosis-inducing Fas ligand
[122
]. These mice spontaneously develop infiltrates of
highly activated CD8+ T cells in their kidneys, liver,
pancreas, and uterus and die between 5 and 12 weeks of age
[123
]. Related studies, which examined the effects of
combining pfp deficiency and Fas mutation, suggested that pfp-mediated
cytotoxicity plays a specific role in regulating systemic autoimmunity
[124
]. These conclusions have been recently supported by
studies in a model of graft versus host disease [121
].
In the setting of B- and T-cell activation, pfp plays an important
immunoregulatory role in the prevention of humoral autoimmunity through
the elimination of autoreactive B cells and Ag-specific T cells.
Moreover, an ineffective, initial CTL response can evolve into a
persistent, antibody-mediated response and, with it, the potential for
sustained, humoral autoimmunity.
A role for pfp in CD8+ T-cell homeostasis has also been
investigated in the context of staphylococcal enterotoxin activation
[123
, 125
]. Previous in vitro
studies have shown that these CTL effectively lyse MHC class
II-expressing cells presenting the proper superantigen. Injection of
staphylococcal enterotoxin B (SEB) into pfp-/- mice
results in dramatically increased, selective expansion and prolonged
persistence of CD8+, but not CD4+, SEB-reactive
T cells [123
]. Repeated injections of staphylococcal
enterotoxin A to pfp-/- mice resulted in significantly
less B-cell depletion compared with control mice [125
].
This suggests that superantigen-activated CD8+ T cells lyse
MHC class II+ APC in a pfp-dependent manner in
vivo. In a model of LCMV infection, secondary immunization of TCR
transgenic (for LCMV) pfp-/- mice with the LCMV-specific
peptide led to an increased proliferation of transgenic
CD8+ T cells, which was not explained by failure to deplete
professional APC [126
]. These results are supported by
an additional study [127
] and point to a novel mechanism
of T-cell homeostasis in which the acquisition of pfp-dependent,
cytotoxic activity regulates the expansion and persistence of
CD8+ effector T cells in vivo.
T-cell memory depends on factors that regulate expansion and death of
CD8+ effector T cells after antigenic stimulation. This
differentiation of effector T cells into memory T cells is critical for
an effective and controlled immune reaction [127
]. There
is now considerable interest in determining what effector molecules
might control this differentiation step. Mice deficient in pfp and
IFN- exhibited increased expansion, altered immunodominance, and
decreased death of antigen-specific CD8+ T cells after
infection with an attenuated strain of LM, which was cleared from these
mice [128
]. Expansion of CD8+ T cells was
controlled by pfp, whereas IFN- was responsible for immunodominance
and the death phase. Another study with Epstein-Barr virus (EBVP)
infection supports this hypothesis, because that primary response is
closely regulated, and the majority of cells are programmed to die via
a cytokine-rescuable pathway, leaving only small populations of memory
T cells surviving [129
]. Thus, it is likely that pfp and
IFN- regulate distinct elements of CD8+ T-cell
homeostasis, independent of their role as effector molecules.
|
OTHER POTENTIAL grz FUNCTIONS?
|
|---|
Because of their many protease specificities, grz have been
suggested to participate in lymphocyte functions such as antigen
processing, extravasation, and migration of mature T cells. GrzA has
been demonstrated to cleave several extracellular matrix proteins
in vitro [130
], and Sayers et al.
[131
] found that grzB inhibited the growth of adherent
tumor cell lines by preventing their adhesion to extracellular matrix
proteins. Perhaps a most striking finding is that the majority of
autoantigens targeted across the spectrum of human systemic autoimmune
diseases are cleaved efficiently by grzB in vitro and during
cytotoxic lymphocyte, granule-induced death, generating unique
fragments not observed during other forms of apoptosis
[132
]. The grzB cleavage sites in autoantigens contain
amino acids in the P2 and P3 positions that are preferred by grzB but
are not tolerated by caspase-8. In contrast, nonautoantigens are not
cleaved by grzB or are cleaved to generate identical fragments arising
in other forms of apoptosis. Systemic autoimmune diseases are a
genetically complex, heterogeneous group of disorders in which the
immune system targets a diverse but highly specific group of
intracellular autoantigens. These targeted molecules are not unified by
common structure, function, or distribution but become clustered and
concentrated in surface blebs when cells undergo apoptosis.
Collectively, these results focus attention on the role of the
cytotoxic lymphocyte, granule-induced death pathway in the initiation
and propagation of systemic autoimmunity. Future work in this area
promises to be very revealing.
GrzA has also been demonstrated to induce cellular responses mediated
by -thrombin [133
], however whether grzA plays any
physiological role in these processes remains to be determined.
Biological functions for all other grz, including grzM and grzH, have
not been defined. The presence of active grz in serum and synovial
fluid of rheumatoid joints has been described [134
], but
their pathophysiologic role remains unclear as yet. It still remains
likley that the role of grz in functions other than cytolysis will be
confirmed best in vivo in homozygous grz-deficient mice. The
derivation of DPPI-/- mice and other grz-/-
mice will be useful tools in this regard.
|
CONCLUDING REMARKS
|
|---|
The last decade of research into cytotoxic granule proteins allows
us to take away several important messages and pose interesting
questions for future investigation. Firstly, it is clear that the
function of most granule proteins revolves around pfp. Mutations of pfp
in mice and man verify the central role pfp plays in host-immune
reaction and regulation. The functions of grz and granulysin, the
best-defined effector molecules in granules, are pfp-dependent. Despite
clearly defined, apoptotic pathways triggered by grzB, experiments in
grz-deficient mice make us question whether grz play a generalized role
in lymphocyte-mediated cell death. Combined with the healthy cytolytic
activity of grzAB-/- CTL/NK cells, these studies
illustrate the potential relevance of grz to the clearance of specific
pathogens. Perhaps the ability of grz to target the nucleus and trigger
DNA fragmentation is specific for viral infections in cytotoxic cells.
In addition, we need to determine how CTL/NK cells can kill target
cells in a pfp-dependent manner that is independent of grzA and grzB.
Obviously, the key to regulating CTL/NK cell-granule function is to
understand exactly how pfp facilitates granule-protein function. There
is a great need to relate the structure of pfp to its molecular
function to enable the rational design of agents that can control pfp
action. Other proteins housed within the cytotoxic granules, which are
undefined yet, will be important in the normal processing and control
of pfp activity and may provide important clues to regulating pfp
activity. This decade also promises many surprises as the secrets of
cytotoxic granules are unraveled.
|
ACKNOWLEDGEMENTS
|
|---|
M. J. S. and J. A. T. are Principal Research
Fellows of the National Health and Medical Research Council (NHMRC) of
Australia. J. M. K., V. R. S., and K. A. B. are supported by project grants from the NHMRC. J. E. D. is supported by a post-graduate scholarship from the Anti-Cancer
Council of Victoria. T. J. S. is supported by the Intramural
Research Support Program, SAIC, National Cancer Institute-Frederick,
Frederick, MD. The content of this publication does not necessarily
reflect the views or policies of the Department of Health and Human
Services, nor does mention of trade names, commercial products, or
organizations imply endorsement by the U.S. Government. This project
has been funded in whole or in part with federal funds from the
National Cancer Institute, under contract number N01-CO-56000.
Received February 5, 2001;
revised February 12, 2001;
accepted February 15, 2001.
|
REFERENCES
|
|---|
-
Trinchieri, G., Perussia, B. (1984) Human natural killer cells. Biologic and pathologic aspects Lab. Investig. 50,489-513[Medline]
-
Henkart, P. A. (1985) Mechanism of lymphocyte-mediated cytotoxicity Annu. Rev. Immunol. 3,31-58[Medline]
-
Podack, E. R. (1985) The molecular mechanism of lymphocyte-mediated tumor cell lysis Immunol. Today 1,21-27
-
Doherty, P. C. (1993) Cell-mediated cytotoxicity Cell 75,607-612[Medline]
-
Kupfer, A. (1991) T-cell effector functions: mechanisms for delivery of cytotoxicity and help Annu. Rev. Cell Biol. 7,479-504
-
Peters, P. J., Geuze, H. J., Ban der Donk, H. A., Borst, J. (1990) A new model for lethal hit delivery by cytotoxic T lymphocytes Immunol. Today 11,28-32[Medline]
-
Young, J. D-E., Cohn, Z. A. (1986) Cell-mediated killing: a common mechanism? Cell 46,641-642[Medline]
-
Kagi, D., Ledermann, B., Burki, K., Seiler, P., Odermatt, B., Olsen, K. J., Podack, E. R., Zinkernagel, R. M., Hengartner, H. (1994) Cytotoxicity mediated by T cells and natural killer cells is greatly impaired in perforin-deficient mice Nature 369,31-37[Medline]
-
Lowin, B., Beermann, F., Schmidt, A., Tschopp, J. (1994) A null mutation in the perforin gene impairs cytolytic T lymphocyte- and natural killer cell-mediated cytoxicity Proc. Natl. Acad. Sci. USA 91,11571-11575[Abstract/Free Full Text]
-
Talento, A., Nguyen, M., Law, S., Wu, J. K., Poe, M., Blake, J. T., Patel, M., Wu, T-J., Manyak, C. L., Silberklang, M., Mark, G., Springer, M., Sigal, N. H., Weissman, I. L., Bleackley, R. C., Podack, E. R., Tykocinski, M. L., Koo, G. C. (1992) Transfection of mouse cytotoxic T lymphocytes with an antisense granzyme A vector reduces lytic activity J. Immunol. 149,4009-4015[Abstract]
-
Burkhardt, J. K., Hester, S., Lapham, C. K., Argon, Y. (1990) The lytic granules of natural killer cells are dual-function organelles combining secretory and pre-lysosomal compartments J. Cell Biol. 111,2327-2340[Abstract/Free Full Text]
-
Peters, P. J., Geuze, H. J., Van der Donk, H. A., Slot, J. W., Griffith, J. M., Stam, N. J., Clevers, H. C., Borst, J. (1989) Molecules relevant for T cell-target cell interaction are present in cytolytic granules of human T lymphocytes Eur. J. Immunol. 19,1469-1475[Medline]
-
Jenne, D. E., Tschopp, J. (1988) Granzymes, a family of serine proteases released from granules of cytolytic T lymphocytes upon T cell receptor stimulation Immunol. Rev. 103,53-71[Medline]
-
Bleackley, R. C., Lobe, C. G., Duggan, B., Ehrman, N., Fregeau, C., Meier, M., Letellier, M., Havele, C., Shaw, J., Paetkau, V. (1988) The isolation and characterization of a family of serine protease genes expressed in activated cytotoxic T lymphocytes Immunol. Rev. 103,5-19[Medline]
-
Muller-Eberhard, H. J. (1986) The membrane attack complex of complement Annu. Rev. Immunol. 4,503-528[Medline]
-
Uellner, R., Zvelebil, M. J., Hopkins, J., Jones, J., MacDougall, L. K., Morgan, B. P., Podack, E., Waterfield, M. D., Griffiths, G. M. (1997) Perforin is activated by a proteolytic cleavage during biosynthesis which reveals a phospholipid-binding C2 domain EMBO J 16,7287-7296[Medline]
-
Kataoka, T., Takaku, K., Magae, J., Shinohara, N., Takayama, H., Kondo, S., Nagai, K. (1994) Acidification is essential for maintaining the structure and function of lytic granules of CTL. Effect of concanamycin A, an inhibitor of vacuolar type H(+)-ATPase, on CTL-mediated cytotoxicity J. Immunol. 153,3938-3947[Abstract]
-
Tschopp, J., Schafer, S., Masson, D., Peitsch, M. C., Heusser, C. (1989) Phosphorylcholine acts as a calcium dependent receptor molecule for lymphocyte perforin Nature 337,272-274[Medline]
-
Young, J. D., Hengartner, H., Podack, E. R., Cohn, Z. A. (1986) Purification and characterization of a cytolytic pore-forming protein from granules of cloned lymphocytes with natural killer activity Cell 44,849-859[Medline]
-
Liu, C-C., Walsh, C. M., Young, J. D-E. (1995) Perforin: structure and function Immunol. Today 16,194-201[Medline]
-
Browne, K. A., Blink, E., Sutton, V. R., Froelich, C. J., Jans, D. A., Trapani, J. A. (1999) Cytosolic delivery of granzyme B by bacterial toxins: evidence that endosomal disruption, in addition to transmembrane pore formation, is an important function of perforin Mol. Cell. Biol. 19,8604-8615[Abstract/Free Full Text]
-
Pena, S. V., Hanson, D. A., Carr, B. A., Goralski, T. J., Krensky, A. M. (1997) Processing, subcellular localization, and function of 519 (granulysin), a human late T cell activation molecule with homology to small, lytic, granule proteins J. Immunol. 158,2680-2688[Abstract]
-
Hanson, D. A., Kaspar, A. A., Poulain, F. R., Krensky, A. M. (1999) Biosynthesis of granulysin, a novel cytolytic molecule Mol. Immunol. 36,413-422[Medline]
-
Krensky, A. M. (2000) Granulysin: a novel antimicrobial peptide of cytolytic T lymphocytes and natural killer cells Biochem. Pharmacol. 59,317-320[Medline]
-
Gamen, S., Hanson, D. A., Kaspar, A., Naval, J., Krensky, A. M., Anel, A. (1998) Granulysin-induced apoptosis. I. Involvement of at least two distinct pathways J. Immunol. 161,1758-1764[Abstract/Free Full Text]
-
Stenger, S., Hanson, D. A., Teitelbaum, R., Dewan, P., Niazi, K. R., Froelich, C. J., Ganz, T., Thoma Uszynski, S., Melian, A., Bogdan, C., Porcelli, S. A., Bloom, B. R., Krensky, A. M., Modlin, R. L. (1998) An antimicrobial activity of cytolytic T cells mediated by granulysin Science 282,121-125[Abstract/Free Full Text]
-
Henkart, P. A., Berrebi, G. A., Takayama, H., Munger, W. E., Sitkovsky, M. B. (1987) Biochemical and functional properties of serine esterases in acidic cytoplasmic granules of cytotoxic T lymphocytes J. Immunol. 139,2398-2405[Abstract]
-
Jenne, D. E., Masson, D., Zimmer, M., Haefliger, J-A., Li, W-H., Tschopp, J. (1989) Isolation and complete structure of the lymphocyte serine protease granzyme G, a novel member of the granzyme multigene family in murine cytolytic T lymphocytes. Evolutionary origin of lymphocyte proteases Biochemistry 28,7953-7961[Medline]
-
Smyth, M. J., O’Connor, M. D., Trapani, J. A. (1996) Granzymes: a variety of serine protease specificities encoded by genetically distinct subfamilies J. Leukoc. Biol. 60,555-562[Abstract]
-
Trapani, J. A. (1997) Dual mechanisms of apoptosis induction by cytolytic lymphocytes Int. Rev. Cytol. 182,111-192
-
Waugh, S. M., Harris, J. L., Fletterick, R., Craik, C. S. (2000) The structure of the pro-apoptotic protease granzyme B reveals the molecular determinants of its specificity Nat. Struct. Biol. 7,762-765[Medline]
-
Haddad, P., Jenne, D., Tschopp, J., Clement, M-V., Mathieu-Mahul, D., Sasportes, M. (1990) Structure and evolutionary origin of the human granzyme H gene Int. Immunol. 3,57-66[Abstract/Free Full Text]
-
Edwards, K. M., Kam, C-M., Powers, J., Trapani, J. A. (1999) The human cytotoxic T cell granule protease granzyme H has chymotrypsin-like (chymase) activity and is taken up into cytoplasmic vesicles reminiscent of granzyme B-containing endosomes J. Biol. Chem. 274,30468-30473[Abstract/Free Full Text]
-
Caputo, A., Garner, R. S., Winkler, U., Hudig, D., Bleackley, R. C. (1993) Activation of recombinant murine cytotoxic cell proteinase-1 requires deletion of an amino-terminal dipeptide J. Biol. Chem. 268,17672-17675[Abstract/Free Full Text]
-
McGuire, M. J., Lipsky, P. E., Thiele, D. L. (1993) Generation of active myeloid and lymphoid granule serine proteases requires processing by the granule thiol protease dipeptidyl peptidase I J. Biol. Chem. 268,2458-2467[Abstract/Free Full Text]
-
Griffiths, G. M., Isaaz, S. (1993) Granzymes A and B are targeted to the lytic granules of lymphocytes by the mannose-6-phosphate receptor J. Cell Biol. 120,885-896[Abstract/Free Full Text]
-
Dupuis, M., Schaerer, E., Krause, K. H., Tschopp, J. (1993) The calcium-binding protein calreticulin is a major constituent of lytic granules in cytolytic T lymphocytes J. Exp. Med. 177,1-7[Abstract/Free Full Text]
-
Andrin, C., Pinkoski, M. J., Burns, K., Atkinson, E. A., Krahenbuhl, O., Hudig, D., Fraser, S. A., Winkler, U., Tschopp, J., Opas, M., Bleackley, R. C., Michalak, M. (1998) Interaction between a Ca2+-binding protein calreticulin and perforin, a com
TOP
ABSTRACT
THE GRANULE EXOCYTOSIS MODEL
