| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Department of Internal Medicine, University Hospital Groningen, The Netherlands
Correspondence: Ymke van der Geld, Department of Clinical Immunology, University Hospital Groningen, Hanzeplein 1, 9713GZ Groningen, The Netherlands. E-mail: Y.M.van.der.geld{at}med.rug.nl
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONGENE LOCALIZATION AND GENE...PROCESSING, STORAGE, AND...FUNCTIONS OF PR3CONCLUSIONREFERENCES |
|---|
Key Words: Wegener’s granulomatosis • myeloid leukemia • serine protease
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONGENE LOCALIZATION AND GENE...PROCESSING, STORAGE, AND...FUNCTIONS OF PR3CONCLUSIONREFERENCES |
|---|
Friedrich Wegener first described WG in 1936 as a distinct entity within the group of primary vasculitides. WG is characterized by granulomatous inflammation in the upper and lower respiratory tract, systemic vasculitis affecting small blood vessels, and pauci-immune necrotizing crescentic glomerulonephritis [13 ]. The research on WG took a big leap forward when van der Woude and colleagues in 1985 established the close association between active WG and anti-neutrophil cytoplasmic antibodies (ANCA) [14 ]. The discovery of these autoantibodies suggested an autoimmune pathogenesis for WG. Five years later the target antigen for ANCA in WG was identified as PR3 [10 , 15 , 16 ]. PR3-ANCA has been established as a specific marker for WG, and detection of these autoantibodies is helpful in the diagnosis and follow-up of WG patients [17 , 18 ]. Furthermore, in most cases a strong correlation has been observed between PR3-ANCA and disease activity of WG as titers of PR3-ANCA rise before a relapse of WG in many patients [17 , 19 20 21 22 ]. Also, PR3-reactive T cells have been reported in patients with WG [23 24 25 26 ].
It is interesting that PR3 is overexpressed in a variety of acute and chronic myeloid leukemia cells [27 ]. In 1996 Ballieux et al. [24 ] identified several PR3-derived peptides restricted to HLA-A2. Cytotoxic T lymphocytes (CTL) specific for one of these HLA-A2-restricted peptides derived from PR3 preferentially lysed chronic myeloid leukemia cells [28 ] and also inhibited granulocyte-macrophage colony-forming unit activity from HLA-A2+ patients with chronic myeloid leukemia [29 ]. The extent of lysis and colony inhibition was proportional to PR3 overexpression in leukemia cells. Furthermore, an HLA-A2-PR3-peptide tetramer could be used to identify and select PR3-specific CTLs from healthy donors, which could be used in leukemia-specific adoptive immunotherapy [30 , 31 ].
So, PR3 has been implicated in two diseases either as autoantigen or as target for immunotherapy. Most reviews on PR3-ANCA have focused on the autoantibodies in ANCA-associated vasculitis. This review will focus on PR3 and the characteristics of PR3 that might implicate this particular antigen, and not other neutrophil-derived serine proteinases, in the pathogenesis of WG, and as target for immunotherapy in myeloid leukemias. We will discuss three main subjects: namely, the genetic localization and gene regulation of PR3, the processing, storage, and expression of the PR3 protein, and the physiological functions of PR3.
|
GENE LOCALIZATION AND GENE REGULATION |
|---|
|
TOPABSTRACTINTRODUCTIONGENE LOCALIZATION AND GENE...PROCESSING, STORAGE, AND...FUNCTIONS OF PR3CONCLUSIONREFERENCES |
|---|
|
|
1-antitrypsin. So it is possible that these variants of
PR3 differ in their potential to damage connective tissues, which could
lead to more damage. However, this has to be further investigated.
Gene regulation
The mRNA of PR3 is expressed in early cells of the myeloid and
monocytic lineage [8
, 38
, 41
].
Transcription of PR3, together with HLE, CatG, and myeloperoxidase
(MPO), is restricted to the promyelocyte and promonocyte stage of
myeloid differentiation, and transcription is down-regulated upon
granulocyte and monocyte maturation [12
, 34
,
42
]. It has been shown that mRNA expression of PR3 is
tightly down-regulated once the promyelocytic cell lines, HL60
[8
, 12
, 43
44
45
] and NB4
[43
], or the promonocytic cell lines, U937
[8
, 34
] and THP-1 [45
], are
treated with dimethyl sulfoxide (DMSO) [12
,
44
] or retinoic acid [12
, 43
]
to induce granulocyte differentiation. Also, treatment with
1,25-dihydroxyvitamin D3
[1,25-(OH)2D3] [8
,
12
], PMA [8
, 12
,
34
], 12-O-tetradecanoylphorbol 13-acetate
(TPA) [44
], or bile acids [45
], all of
which induce monocyte differentiation, caused down-regulation of PR3
mRNA expression. This tight control of the PR3 mRNA expression was
shown to be effective at the transcriptional as well as at the
posttranscriptional level [43
, 44
].
Therefore, the PR3 gene belongs to a group of genes that are directly
and negatively regulated during myeloid differentiation
[44
], as has previously been shown for the MPO gene
[46
].
Recently, the transcriptional control of the PR3 gene has been
analyzed. The 5’ flanking region of the PR3 gene contains a TATA-box
[34
, 38
, 47
], a PU.1
regulatory element, a CG-element, and a potential site for the CAAT
binding protein (C/EBP) and c-Myb transcription factors. The CAAT box
promoter element, which is present in the promoter regions of HLE
[36
] and CatG [35
], is not present in the
PR3 promoter. Furthermore, the 5’ region of the PR3 gene includes a
NF-
B element on the antisense strand, five ß-globulin-specific
elements, and two ets-core elements. Three of the five ß-globulin
elements are preceded by a thymine, forming the retinoic acid
regulatory elements [34
, 47
]. The 5’
flanking region of PR3 is outlined in Figure 1A
.
Within the base pairs -200 to 1 of the PR3 promoter, two elements were
identified that are sufficient for maximal expression of PR3 mRNA.
These two elements are a PU.1 consensus located at position -101 and a
CG-element located at position -190 [47
]. The PU.1
consensus binds a myeloid nuclear protein, the transcription factor
PU.1, which is expressed in myeloid and B cells but not T cells
[48
]. The CG-element binds a nuclear protein of 
40
kDa that is found in myeloid as well as non-myeloid cells such as HeLa
cells. The CG-element is complementary to the binding site of the
transcription factor Sp1, but in the PR3 promoter region no binding of
Sp1 could be observed. The activity of these two elements is dependent
on the presence of the -91 to 0 promoter region, which contains the
potential sites for C/EBP and c-Myb and the TATA box
[47
].
In the CatG promoter the C/EBP and c-Myb sites are not present, whereas the PU.1 and CG-elements are present in both the HLE [36 ] and the CatG [35 ] promoter regions. This suggests that the latter sites are important control elements for tissue- and developmental stage-specific expression of the neutrophil and monocyte specific serine proteases.
Because PR3 [47 ], HLE [36 , 37 ], CatG [35 ], and MPO [49 ] are similarly regulated during myeloid differentiation most of the regulatory elements present in the promoter region are alike. Nonetheless, there are some minor differences (shown in Table 2 ), two of which were already mentioned above. A glucocorticoid-response element and an ocotomer-binding factor response element present in the CatG and HLE genes are not present in the PR3 gene [34 , 38 ]. The significance of these differences in regulatory elements between the PR3 gene and the genes for HLE, CatG, and MPO is not known. These differences could be important for differential regulation of the PR3 gene, which might have implication for possible aberrant expression of PR3 in patients with WG and myeloid leukemia.
|
|
PROCESSING, STORAGE, AND EXPRESSION OF PR3 |
|---|
|
TOPABSTRACTINTRODUCTIONGENE LOCALIZATION AND GENE...PROCESSING, STORAGE, AND...FUNCTIONS OF PR3CONCLUSIONREFERENCES |
|---|
Processing and granular sorting of PR3 is very similar to that of HLE [57 ], CatG [57 , 58 ], and AZU [59 ] [reviewed in refs. 60 61 62 ]. The processing of PR3 is depicted in Figure 2 . During translation PR3 is synthesized as a prepro-enzyme, shown in Figure 1C . The prepro-enzyme of PR3 is processed in four consecutive steps into a mature form consisting of 222 amino acids. First, the signal peptide is removed upon translocation of prepro-PR3 into the endoplasmic reticulum (ER), yielding a pro-form of PR3 with an amino-terminal and carboxy-terminal propeptide [63 , 64 ]. The amino-terminal propeptide of two amino acids, Ala-Glu for PR3, is a typical feature of myeloid serine proteases [65 , 66 ]. Next, PR3 is glycosylated on both its potential sites for N-linked glycosylation. High-mannose oligosaccharide side chains are added in the ER and converted into complex saccharide chains with terminal sialic acid residues in the Golgi complex [15 , 63 , 64 , 67 ]. As a third processing step the propeptide of PR3 is removed in the post-Golgi organelle [63 , 64 ]. The enzyme responsible for the removal of this propeptide is a cysteine proteinase distinct from dipeptidyl peptidase I (DPPI), which cleaves the propeptide of CatG [68 ], HLE [57 ], and AZU [59 ]. After removal of the two amino acid propeptide the substrate-binding pocket becomes accessible and PR3 becomes potentially enzymatically active [63 , 64 , 69 ]. As a fourth step, a seven-amino-acid carboxy-terminal extension is removed during processing of PR3 [63 , 64 ]. Cleavage takes place next to Arg222, suggesting that a trypsin-like proteinase is involved in the carboxy-terminal processing of PR3. During processing of PR3 four disulfide bridges are formed keeping its three-dimensional structure intact (Fig. 3 ). In the acidic environment of the azurophilic granules the mature form of PR3 is maintained in a conformationally inactive form. Upon translocation into neutral pH PR3 undergoes a final conformational change into an active form. This pH-dependent phenomenon has thus far only been described for PR3 [70 ].
|
|
It is interesting that during synthesis minor portions of the pro-form of PR3 [63 , 64 , 75 ] and other neutrophil- and monocyte-derived serine proteases [57 , 58 , 68 ] escape granular targeting and are secreted (Fig. 2) . This phenomenon has been regarded as an imperfection of the granular targeting process. Recently a function of the secreted pro-form of PR3 was described as being a putative negative feedback regulator of granulopoiesis [75 ] (see below). The overwhelming majority of PR3 is stored in the azurophilic granules as a mature form without amino- and carboxy-terminal propeptides. The secreted form of PR3, however, still contains the amino-terminal propeptide. This pro-form of PR3 is probably targeted for secretion before the post-Golgi organelle and may thus also still have the carboxy-terminal extension.
Only recently it was shown that PR3 is not only stored in the azurophilic granules but is also present in secretory vesicles and in the specific granules, whereas MPO and HLE were only present in azurophilic granules [76 ]. Whether this PR3 in the secretory and specific granules is comparable to the secreted pro-form of PR3 or the mature form of PR3 stored in the azurophilic granules still has to be established.
PR3 migrates on a sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) gel as a triplet of approximately 29–32 kDa. These isoforms probably differ in their carbohydrate structures, as has been shown for HLE and CatG [77 ]. In HLE and CatG certain isoforms were destined for granular targeting and others for secretion. Glycosylation isoforms of PR3 may underlie differences in targeting of PR3. So certain isoforms of PR3 may be destined for granular targeting and others for secretion or expression on the plasma membrane. This might explain the differential expression of PR3 in neutrophils and monocytes.
Differences in processing and storage between PR3, HLE, CatG, and MPO are summarized in Table 3 . In summary, PR3, in contrast to HLE, CatG, and MPO is also present in the secretory vesicles and specific granules. Upon activation these secretory vesicles are readily degranulated, and PR3 might be translocated to the plasma membrane. PR3-ANCA in WG patients can bind to these membrane-bound and secreted forms of PR3. The existence of this readily mobilized pool of PR3 might thus play a relevant role in the pathophysiology of WG.
|
On resting neutrophils of healthy controls a clear bimodal distribution of PR3 surface expression has been observed [82 , 83 ]. Three distinct phenotypes of donors could be defined corresponding to a low, i.e. less than 20% of neutrophils with membrane PR3 (mPR3), intermediate, and high percentage of mPR3-expressing neutrophils. These distinct phenotypes of mPR3 expressing neutrophils varied among individuals, and a particular phenotype was extremely stable for each individual in time as well as after neutrophil activation. This peculiar type of distribution seems to be restricted to PR3 and might be genetically controlled [82 , 83 ]. The occurrence of a high percentage of neutrophils expressing mPR3 is increased in patients with ANCA-associated vasculitis and rheumatoid arthritis when compared to the healthy population [79 , 83 ].
Upon priming of neutrophils by tumor necrosis factor (TNF-) as
may occur during a viral or bacterial infection and subsequent to
activation by fMLP or IL-8 most of the released PR3 is expressed on the
plasma membrane, whereas HLE and MPO are partly expressed on the
membrane and mainly released into the extracellular medium
[76
, 80
].
The way PR3 is associated with the plasma membrane and the functional
significance of membrane-associated PR3 are still under investigation.
The interaction between PR3 and the plasma membrane does not seem to
result from charge interactions [76
], whereas binding of
HLE or CatG is largely charge dependent [84
]. Membrane
PR3 interaction does not depend on membrane sialic acid residues, nor
is PR3 anchored to the plasma membrane via a GPI-link
[76
]. Membrane-associated PR3 appears as a 29-kDa band
on SDS-PAGE gels [76
] and can still bind to
1-antitrypsin (1-AT or 1
protease inhibitor, 1-PI) [85
]. This
rules out the possibility that PR3, complexed to 1-AT,
is bound to the serine proteinase inhibitor (Serpin) enzyme complex
(SEC)-receptor, as earlier proposed [86
]. The
association of PR3 with the plasma membrane seems to be covalent
[76
]. However, association of PR3 to the membrane also
has been described through lipid interactions, in contrast to that of
HLE and MPO [87
]. This lipid-associated PR3 is still
enzymatically active, as has also been shown for HLE and CatG
[84
]. This excludes the possibility that the secreted
pro-form of PR3, which is enzymatically inactive, is bound to the
plasma membrane. Only recently a specific membrane molecule on
endothelial cells has been described that interacts with PR3
[88
]. The same molecule is possibly responsible for the
membrane association of PR3 on neutrophils. Such a receptor on
neutrophils has been reported for HLE [89
,
90
].
Differences in expression between PR3, HLE, CatG, and MPO are summarized in Table 3 . In summary, PR3, in contrast to HLE and MPO, is expressed on the plasma membrane of resting neutrophils. It has even been shown that the occurrence of a high percentage of neutrophils expressing mPR3 is a risk factor for vasculitis. This suggests that PR3 expression on the plasma membrane of neutrophils is clinically relevant and might favor the occurrence of chronic inflammation.
PR3 expression in non-myeloid cells
Whether endothelial cells express PR3 is still a matter of debate.
Mayet et al described that endothelial cells and a number of other
non-myeloid cells express PR3 after stimulation with proinflammatory
cytokines [91
, 92
]. This PR3 was suggested
to be synthesized de novo [91
,
93
], but the production of PR3 mRNA was transient
[94
]. Although Sibelius et al. [94
] have
confirmed PR3 expression on endothelial cells, other groups could not
confirm the presence of this protein nor mRNA expression of PR3 in
cytokine-stimulated endothelial cells [88
,
95
, 96
]. If endothelial cells produce PR3
PR3-ANCA can bind directly to PR3 expressed on endothelial cells,
recruit neutrophils and monocytes via complement activation, and cause
damage to the vessel wall. This would give an explanation for the
vasculitis and glomerulonephritis seen in WG.
It is interesting that PR3 and MPO themselves can bind to endothelial cells in vitro [97 , 98 ]. Only recently a membrane molecule on endothelial cells has been identified that specifically binds PR3 [88 ]. In addition, endothelial cells could internalize PR3 and MPO, whereas only MPO could also be internalized by epithelial cells [99 ]. These recent data emphasize the fact that PR3 can also have a specific interaction with endothelial cells.
The expression of PR3 by non-myeloid cells is difficult to explain considering that one of the transcription factors, PU.1, that is necessary for the transcription of PR3, is expressed in myeloid cells and B cells only. One explanation could be that in endothelial cells another nuclear protein resembling PU.1 binds to the PR3 promoter region inducing transient expression of the PR3 gene. Another explanation could be that the transcription of PR3 is leaky. This is unlikely because PR3 is an active protease and if expressed in an environment not equipped to store this protease, PR3 would easily degrade other proteins.
|
FUNCTIONS OF PR3 |
|---|
|
TOPABSTRACTINTRODUCTIONGENE LOCALIZATION AND GENE...PROCESSING, STORAGE, AND...FUNCTIONS OF PR3CONCLUSIONREFERENCES |
|---|
Regulation of myeloid differentiation by PR3
Several data support the possible role for PR3 in growth and
differentiation of granulocytes and monocytes (shown in a hypothetical
scheme in Figure 4
). In hematopoietic progenitor cells PR3 expression is up-regulated
by granulocyte colony-stimulating factor (G-CSF) as the PR3 gene was
shown to be a G-CSF-responsive gene [101
]. This
G-CSF-induced PR3 expression is dependent on the PU.1, C/EBP, and c-Myb
binding sites in the PR3 promoter region. The transcription factor PU.1
played a major role in inducing this expression of PR3. Constitutive
overexpression of the active form of PR3 induced factor-independent
growth of hematopoietic progenitor cells expressing the G-CSF receptor
[101
]. So, in the early steps of myeloid differentiation
up-regulation of PR3 causes growth of progenitor cells (Fig. 4
, part
1). Later during myeloid differentiation, after the promyelocytic
stage, PR3 expression is down-regulated and growth arrest and further
differentiation along the granulocytic lineage is induced
[12
] (Fig. 4
, part 4). During synthesis of PR3 at the
promyelocytic stage small amounts of the pro-form of PR3 escape
granular targeting and are secreted [63
,
64
]. This secreted pro-form can reduce the fraction of
granulocyte and monocyte colony-forming units (CFU-GM) in S-phase of
hematopoietic progenitor cells, which implies a function for PR3 as a
putative feedback regulator of myeloid differentiation (Fig. 4
, part
2). It is interesting that this feedback regulation by PR3 is
reversible and abrogated by G-CSF and GM-CSF [75
].
Furthermore, during the promyelocytic stage, PR3 can exert certain
functions that can regulate cell differentiation at this stage
[102
103
104
]. At the promyelocytic stage it was shown that
PR3 can truncate the nuclear factor NF-B subunit
[102
], degrade the 28-kDa heat shock protein, which is a
potential component of signal transduction pathways
[103
], and truncate the Sp1 transcription factor
[104
] (Fig. 4
, part 3). Decrease of the degradation of
these proteins upon down-regulation of PR3 might provide a mechanism
through which differentiating promyeloid cells can acquire a new
repertoire of gene expression. Of the above-mentioned functions of PR3
in myeloid differentiation only the truncation of the NF-B subunit
could also be caused by CatG or HLE [102
].
|
|
Another physiological role ascribed to PR3 and HLE is migration of neutrophils through the basement membranes. PR3 has an elastase-like enzymatic activity and can degrade a variety of extracellular matrix proteins and basement membrane proteins like elastin [9 , 81 , 109 ], fibrinogen and proteoglycans [110 ], fibronectin, laminin, vitronectin, and collagen type IV [2 ]. PR3 can also degrade casein and hemoglobin [81 ].
Enzymatically active PR3 has been implicated in the modulation of
inflammatory mediators. PR3, and not HLE or CatG, is involved in the
cleavage and inactivation of C1 inhibitor [111
].
Furthermore, PR3 as well as HLE can cleave membrane-bound TNF-
[112
, 113
], IL-1ß [113
],
and the IL-2 receptor [114
]. Whether these cleaved
products of TNF- and IL-1ß are biologically active has still to be
determined. PR3, HLE, and CatG are able to degrade and inactivate IL-6
[115
] as well as activate latent transforming growth
factor ß (TGF-ß). However, PR3 is most potent in this activation of
TGF-ß [116
]. Only cleavage of IL-8 to a biologically
more active form has thus far been described for PR3 only
[117
].
PR3, HLE, and CatG have been implicated in modulating the activity of platelets and endothelial cells [118 , 119 ], possibly by cleavage and subsequent inactivation of proteinase-activated receptors (PAR). PR3 and HLE are able to potentiate the CatG-induced platelet activation [118 ]. PR3 and HLE can cleave PAR on platelets and endothelial cells downstream of the thrombin cleavage site, thus disabling the activation of these receptors by thrombin [120 ], indicating that PR3 and HLE can inhibit the biological actions of thrombin. Finally, PR3 can inhibit the activation of neutrophil NADPH-oxidase. It is interesting that this inhibition was independent of the enzymatic activity of PR3 [121 ].
Of the above-mentioned physiological functions of PR3 cleavage of IL-8 and C1 inhibitor and inhibition of NADPH-oxidase have thus far been described for PR3 only (Table 4) . Granulocyte activation as seen in WG and subsequent degranulation may locally increase PR3 levels, resulting in cleavage of C1 inhibitor and IL-8. Degradation of C1 inhibitor could possibly lead to further activation of the complement pathway at sites of neutrophil infiltration causing more neutrophil migration, activation, and tissue damage. Cleavage of IL-8 to a biologically more active form can amplify the chemotactic activity of IL-8 and further enhance neutrophil recruitment to sites of inflammation, which could possibly lead to more tissue damage. Modulation of cytokines by neutrophil- and monocyte-derived serine proteases might be important for the cytokine balance at local foci of inflammation where the physiological inhibitors of proteases are not abundantly present.
PR3 and endothelium
Systemic vasculitis affecting small blood vessels and pauci-immune
necrotizing crescentic glomerulonephritis are the major features of WG
[13
]. Several in vitro studies have provided
evidence that PR3 may directly contribute to this vascular injury. PR3
as well as MPO can bind to human umbilical vein endothelial cells
(HUVEC) in vitro. Anti-PR3 or anti-MPO antibodies
[97
, 98
, 122
] can still
recognize PR3 and MPO bound to HUVEC, respectively. Binding of PR3 to
HUVEC could be inhibited by 1-AT [97
].
PR3, HLE, and CatG can induce detachment and cytolysis of HUVEC, which
was dependent on their enzymatic activity [123
].
Furthermore, PR3 and HLE but not CatG could enhance IL-8 production of
HUVEC in vitro. The induction of IL-8 production by PR3 was
independent of its enzymatic activity, whereas that by HLE was
dependent on the enzymatic activity [124
]. Recently, it
was shown that PR3 can also induce MCP-1 production of HUVEC in
vitro [125
]. Finally, PR3 and HLE can induce
apoptosis of human endothelial cells [126
,
127
]. Only recently it was shown that an enzymatically
inactive carboxy-terminal domain of PR3 was responsible for apoptosis
induction [128
]. The above-mentioned activities of PR3
on endothelial cells might be receptor mediated as, only recently, a
membrane molecule on endothelial cells that interacts with PR3 has been
identified [88
]. This supports data showing that PR3,
both enzymatically active and inactive, binds to and enters human
endothelial cells and induces apoptosis [99
].
So, PR3, once released, could activate endothelial cells and contribute to the perpetuation of the inflammatory process in systemic vasculitis by enhancing IL-8 and MCP-1 production with subsequent recruitment of neutrophils, monocytes, and T cells. Furthermore, released PR3 can bind to endothelial cells, be internalized, and induce apoptosis of endothelial cells, which may result in vascular damage.
Inhibitors of PR3
PR3, HLE, and CatG are strong proteolytic enzymes, and the
activity of these proteases is precisely regulated by endogenous
protein inhibitors called serpins (serine protease inhibitors;
Table 5 ). The main physiological inhibitor for PR3 as well as HLE is
1-antitrypsin (1-AT or
1-protease inhibitor) [2
,
129
], whereas the physiological inhibitor of CatG is
1-antichymotrypsin [2
]. PR3 and HLE were
also inhibited by elafin [130
], a human skin-derived
inhibitor, and to a lesser extent by 2-macroglobulin
[2
, 129
]. In contrast to HLE and CatG, PR3
is not inhibited by secretory leukoprotease inhibitor (SLPI) and is
only weakly inhibited by eglin c [2
]. HLE is inhibited
by glycosaminoglycans independent of the substrate, whereas PR3
activity to large substrates like elastin, but not to small substrates,
is partially inhibited by glycosaminoglycans [131
]. Only
recently it was shown that DNase partially inhibited the enzymatic
activity of HLE and CatG but had no effect on the enzymatic activity of
PR3 [132
].
|
1-AT has been found in plasma of WG
patients and healthy controls, although free PR3 could also be detected
[133
, 134
]. Under normal conditions the
activity of PR3 and HLE is tightly regulated by a large excess of
circulating anti-proteases like 1-AT. The
1-AT is encoded by a highly polymorphic gene, with more
than 75 alleles, defining severely (Z allel), medium, and non-deficient
(M allel) 1-AT phenotypes [reviewed in ref. 135 ].
Certain variants of 1-AT may bind and inactivate PR3 to
a lesser extent, thus influencing the balance between protease and
anti-protease at the site of inflammation. Several groups have
described that PR3-ANCA-positive vasculitis is more common in patients
with the deficient ZZ phenotype [136
137
138
139
140
]. Abnormal
1-AT phenotypes are associated with an increased risk of
WG as well as with increased morbidity [141
,
142
]. These severe 1-AT deficiencies are
not specific for patients with antibodies to PR3 because they have also
been described as a risk factor for the development of emphysema
[135
], various hepatic disorders, and several autoimmune
diseases, among which are uveitis and systemic lupus erythematosus
[reviewed in refs. 143 144
].
Apart from the genetically determined imbalance between PR3 or HLE and
1-AT an acquired imbalance between PR3 or HLE and
1-AT seems a factor that influences activity of PR3 or
HLE. Under pathological conditions 1-AT can be degraded
and inactivated by a number of proteins released from activated
neutrophils like HLE [145
, 146
],
metalloproteinases [147
], and reactive oxygen species
[148
], thus increasing the local activity of released
PR3 and HLE. Furthermore, oxidation of 1-AT by oxygen
radicals reduces the association of 1-AT with HLE
[149
]. In patients with WG PR3-ANCA can interfere with
the inhibition of PR3 by 1-AT [150
,
151
]. Inactivation or low serum levels of
1-AT may cause PR3 and HLE to escape from their
physiological inhibitor, resulting in tissue destruction
[152
]. The genetically determined or acquired deficiency
of 1-AT influences both the activity of PR3 and HLE, and
thus may be a risk factor for patients with WG. However, this
1-AT deficiency seems not to account for the specific
development of autoreactivity to PR3 because it does not distinguish
between PR3 and HLE.
One important issue is that PR3, in contrast to HLE, is not inhibited by SLPI (Table 5) . SLPI is present in large amounts in respiratory epithelial lining fluid and may play a significant role in protecting the normal respiratory epithelium, especially in the upper airways, where levels of SLPI are highest. Because PR3 is not inactivated by SLPI PR3 may play a significant role in protease-mediated airway damage, and may thus be a risk factor for patients with WG are for chronic obstructive pulmonary diseases.
In sputum of patients with cystic fibrosis higher levels of PR3 than
HLE were found compared to patients with chronic bronchitis without
cystic fibrosis. PR3 in sputum appeared as a free enzymatically active
enzyme or complexed to inhibitors like 1-AT. The
clinical relevance of high amounts of PR3 in the sputum is illustrated
by its correlation with the severity of cystic fibrosis, whereas the
latter did not correlate with HLE levels [153
]. PR3 as
well as HLE can trigger airway gland secretion [153
] and
can induce emphysema in hamsters upon intratracheal administration
[9
]. Pulmonary active WG patients did not have an
elevated level of free PR3 in their bronchoalveolar lavage fluid and,
in contrast, mainly PR3/1-AT complexes were found
[154
]. Therefore it is likely that 1-AT
is the primary physiologically relevant inhibitor protecting the lower
respiratory tract against PR3 and HLE.
Consequently, damage to lung tissue can result from, on the one hand,
prolonged exposure to serine proteases like PR3 and HLE as in chronic
inflammation and, on the other hand, from an insufficiency of
appropriate inhibitors as in the case of genetically determined or
acquired deficiency of 1-AT. So, the
protease/anti-protease imbalance is a general contributor to the
diverse conditions that result in chronic pulmonary inflammation. The
local ratio of PR3 and HLE and antiproteases may determine the final
proteolytic damage by neutrophils and monocytes.
|
CONCLUSION |
|---|
|
TOPABSTRACTINTRODUCTIONGENE LOCALIZATION AND GENE...PROCESSING, STORAGE, AND...FUNCTIONS OF PR3CONCLUSIONREFERENCES |
|---|
At the level of gene localization and gene regulation PR3, HLE, and CatG are not very polymorphic proteins. Only minor variability is seen in the PR3 sequence (Table 1) . Whether this variability can account for differences at the protein level is still unknown. However, the polymorphism of PR3 cannot account for alloresponsiveness of donor lymphocytes against recipient leukemic cells. Furthermore, most regulatory elements in the promoter region of PR3 are common for the neutrophil-derived serine proteases. Nonetheless, three regulatory elements present in the promoter regions of HLE or CatG are not present in the promoter region of PR3 (Table 2) . Whether these differences can account for a possible aberrant expression of PR3 in patients with WG or myeloid leukemia still has to be determined.
The processing of PR3 is very similar to that of other neutrophil-derived serine proteases, but major differences are seen in the expression and storage of PR3. PR3 is, apart from its presence in azurophilic granules, also present in the secretory vesicles and specific granules of neutrophils and monocytes, and PR3 is expressed on the plasma membrane of resting neutrophils (Table 3) . It is intriguing that this membrane expression is bimodal and that the presence of a high percentage of neutrophils with PR3 membrane expression seems to be a risk factor for vasculitis.
The physiological functions of PR3 that depend on its enzymatic activity are very similar for PR3, HLE, and CatG because all three proteins are serine proteases. Modulation of inflammatory mediators has been described for PR3, HLE and CatG, but cleavage of IL-8 and C1 inhibitor and inhibition of NADPH-oxidase has thus far been described for PR3 only (Table 4) . One major difference in function between neutrophil-derived serine proteases is the unique role PR3 may play as a feedback regulator in myeloid differentiation. Furthermore, PR3, both enzymatically active and inactive, binds to and enters human endothelial cells and induces apoptosis. These activities of PR3 on endothelial cells may be receptor mediated.
A final difference between the serine proteases relates to their inhibition by natural inhibitors. Most of the inhibitors of HLE or CatG can also inhibit the enzymatic activity of PR3 (Table 5) with the exception of SLPI, which does inhibit HLE and CatG but does not inhibit PR3.
What is the significance of the above-summarized differences between PR3 and the other serine protease homologs for our understanding of the role of PR3 as autoantigen in WG or as target for immunotherapy in myeloid leukemias?
For our understanding of the role of PR3 in the pathogenesis of WG four of the above-mentioned differences may be of relevance. First, the role of PR3 as a feedback regulator in myeloid differentiation might be relevant as a defect in the down-regulation of PR3 at the promyelocytic stage (Fig. 4 , part 4) may cause enhanced myeloid differentiation, which may lead to enhanced inflammation. Second, PR3, both enzymatically active and inactive, binds to and enters human endothelial cells that could be relevant for the inflammatory potential of PR3 in WG. PR3 can bind to endothelial cells, activate these cells and induce apoptosis of endothelial cells, with damage to the vessel wall. The latter two factors define PR3 as a protein with high inflammatory potential. It is possible that autoimmunity to PR3 is induced in an inflammatory milieu where PR3 is more immunogenic. Also, an aberrant clearance of PR3 and improper inhibition of PR3 can enhance the inflammatory potential of PR3. Third, upon activation of neutrophils PR3 is readily secreted from the secretory vesicles and can cause tissue damage when not properly inhibited. Furthermore, the occurrence of a high percentage of neutrophils expressing PR3 on their membrane shown to be a risk factor for vasculitis, might favor the occurrence of chronic inflammation. This bimodal expression of PR3 might be genetically controlled. Whether this bimodal expression of PR3 can be accounted for by a bi-allelic restriction fragment length polymorphism described in the PR3 locus has still to be determined. The interaction of this membrane-associated PR3 with PR3-ANCA might be important for the pathophysiology of WG. Finally, aberrant expression of PR3, due to polymorphisms in the amino acid sequence or variants in the promoter region might be important for the induction of autoantibodies to PR3.
For our understanding of the role of PR3 in the development of leukemia or as target for immunotherapy in myeloid leukemias the role of PR3 as a feedback regulator in myeloid differentiation may be of relevance. A defect in the down-regulation of PR3 during the promyelocytic stage of differentiation (Fig. 4 , part 4) could lead to increased secretion of PR3 and thereby contribute to the suppression of normal myeloid differentiation via the secreted pro-form of PR3 (Fig.4 , part 2). The factor-independent growth of hematopoietic progenitors cells induced by PR3 could represent one of the early stages in the development of leukemia (Fig. 4 , part 1). So, overexpression of PR3 could lead to uncontrolled growth of myeloid cells as seen in leukemia. In contrast, in myeloid leukemias PR3 is often overexpressed. Overexpressed PR3 can be processed and presented in MHC class I molecules that might induce CTLs specific for PR3 peptides. These CTLs specific for peptides derived from PR3 have been shown to preferentially lyse myeloid leukemia cells and inhibit granulocyte-macrophage colony forming unit activity, which could be used in leukemia-specific adoptive immunotherapy.
In conclusion, PR3 is a very intriguing protein with a large array of physiological functions. More research on these physiological functions of PR3 can give more insight into the role PR3 may play in the pathophysiology of WG and myeloid leukemia.
Received September 6, 2000; revised September 30, 2000; accepted October 3, 2000.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONGENE LOCALIZATION AND GENE...PROCESSING, STORAGE, AND...FUNCTIONS OF PR3CONCLUSIONREFERENCES |
|---|
