THE JOURNAL OF BIOLOGICAL CHEMISTRY
© 2002 by The American Society for Biochemistry and Molecular Biology, Inc.
Vol. 277, No. 11, Issue of March 15, pp. 8802–8809, 2002
Printed in U.S.A.
Extracellular Release of the Glycosylphosphatidylinositol
(GPI)-linked Leishmania Surface Metalloprotease, gp63,
Is Independent of GPI Phospholipolysis
IMPLICATIONS FOR PARASITE VIRULENCE*
Received for publication, September 19, 2001, and in revised form, December 28, 2001
Published, JBC Papers in Press, January 2, 2002, DOI 10.1074/jbc.M109072200
Bradford S. McGwire‡§¶, William A. O’Connell‡, Kwang-Poo Chang储, and David M. Engman‡
From the ‡Departments of Pathology and Microbiology-Immunology, Northwestern University Medical School, Chicago,
Illinois 60611, §Section of Infectious Diseases, College of Medicine, University of Illinois, Chicago, Illinois 60612, and
储Department of Microbiology and Immunology, University of Health Sciences/Chicago Medical School,
North Chicago, Illinois 60064
The major zinc metalloprotease of Leishmania (gp63),
an important determinant of parasite virulence, is attached to the parasite surface via a glycosylphosphatidylinositol anchor. Here we report the spontaneous release of proteolytically active gp63 from a number of
Leishmania isolates, causing cutaneous and visceral disease. To investigate the mechanism(s) of gp63 release,
we transfected a gp63-deficient variant of Leishmania
amazonensis with constructs expressing gp63 and various mutants thereof. Surprisingly, approximately half of
wild type gp63 was found in the culture supernatant
12 h post-synthesis. Biochemical analysis of the extracellular gp63 revealed two forms of the protein, one that
is released from the cell surface, and another, that apparently is directly secreted. Release of cell surface gp63
was significantly reduced when the proteolytic activity
of the protein was inactivated by site-specific mutagenesis or inhibited by zinc chelation, suggesting that release involves autoproteolysis. The extracellular gp63
does not contain a glycosylphosphatidylinositol moiety
or ethanolamine, indicating that phospholipolysis is not
involved in the release process. Release of gp63 is also
independent of glycosylation. The finding of proteolytically active, extracellular gp63 produced by multiple
Leishmania isolates suggests a potential role of the extracellular enzyme in substrate degradation relevant to
their survival in both the mammalian host and the insect vector.
A variety of microorganisms possess membrane-bound
and/or secreted proteases that aid in their invasion of host
tissues (1). Similarly, invasive tumor cells possess extracellular
matrix metalloproteases that enhance their metastatic potential (2, 3). Structurally related to these is a 63-kDa surfacelocalized zinc-dependent protease (gp63 or leishmanolysin) expressed by protozoan parasites such as Leishmania,
Trypanosoma, and Crithidia (4 – 8). Common features of these
* This work was supported in part by grants from the National
Institutes of Health and the American Heart Association, including a
Physician-Scientist Postdoctoral Fellowship (to B. S. M.) and an Established Investigator Award (to D. M. E.). The costs of publication of this
article were defrayed in part by the payment of page charges. This
article must therefore be hereby marked “advertisement” in accordance
with 18 U.S.C. Section 1734 solely to indicate this fact.
¶ To whom correspondence should be addressed: Northwestern
University, Dept. of Pathology Ward 6-140 303 E. Chicago Ave. Chicago, IL 60611. Tel.: 312-503-1267; Fax: 312-503-1265; E-mail: [email protected].
endoproteinases include a HEXXH-zinc binding/catalytic domain, multiple sites of N-linked glycosylation, and membrane
linkage with a glycosylphosphatidylinositol (GPI)1 anchor, all
of which may contribute to posttranslationally regulate enzyme
expression (9). The finding that gp63 has an HEXXH domain
and can undergo autoactivation, presumably via a cysteine
switch mechanism, has led to its comparison with matrix metalloproteases (10, 11).
Leishmania gp63 is involved in the binding of the parasite to
and intracellular survival within host macrophages (5, 12–14).
Upon inoculation into the mammalian host by their sandfly
vector, infective flagellated promastigotes are exposed to a
variety of antimicrobial factors, some of which may be inactivated by gp63. For example, gp63 renders Leishmania resistant to complement-mediated lysis; cleavage of surface C3b
enhances parasite binding to macrophage complement receptors, thereby facilitating uptake by the obligate host cell of the
organism (15). In addition, gp63 itself is a ligand for binding to
host surface proteins, such as the fibronectin receptor (16 –18).
Upon macrophage binding, the parasite is phagocytized and
thereafter multiplies as an intracellular amastigote form in
host phagolysosomes (19). The survival of Leishmania amazonensis in this inhospitable environment is greatly facilitated
by the presence of proteolytically active gp63 (13, 14).
Membrane association of surface ligands via GPI anchors is
a common feature of the trypanosomatids (20). Release of these
GPI-anchored proteins is inducible experimentally either by
transgenetic expression of GPI-phospholipase C or deletion or
site-specific mutation of their GPI attachment sequences (9, 21,
22). Release of the GPI-anchored variant surface glycoprotein
(VSG) of the bloodstream African trypanosome, a spontaneous
natural event during its life cycle, has also been well characterized. VSG release occurs via proteolysis during the differentiation of the bloodstream form to the insect procyclic form as
well as via activation of a GPI-phospholipase during periods of
cellular stress (e.g. osmotic or pH) (23).
In this paper, we describe significant extracellular release of
gp63 from Leishmania. Release apparently involves both autoproteolysis and secretion as likely mechanisms. The finding of
extracellular, proteolytically active gp63 released by multiple
Leishmania species may reflect a common requirement of extracellular protease for the parasite life cycle and may also
contribute to disease pathogenesis.
1
The abbreviations used are: GPI, glycosylphosphatidylinositol;
HRP, horseradish peroxidase; PBS, phosphate-buffered saline; PLC,
phospholipase C; SA, streptavidin; VSG, variant surface glycoprotein.
8802
This paper is available on line at http://www.jbc.org
Leishmania gp63 Release
8803
EXPERIMENTAL PROCEDURES
Parasites—Two parasite stocks were used for the majority of these
studies, an attenuated, gp63-deficient variant of L. amazonensis, designated 250 clone 1, which had been cultivated in vitro for more than
300 passages (9, 14), and parental L. amazonensis, designated 12 clone
1, which has remained cryopreserved with minimal in vitro growth
since 1980 (LV 78). The latter has been cultured less than 50 passages
in vitro and is virulent both in vitro and in vivo. The gp63 deficiency of
the variants consistently correlates with a decrease of in vitro and in
vivo infectivity. Cloned parasites from each stock used in this study
were maintained in medium 199 (M199, Invitrogen) supplemented with
10% heat-inactivated fetal bovine serum at 25 °C. Recent clinical isolates of Leishmania tropica and L. infantum from endemic sites in
Turkey were maintained as described above. Construction of gp63 expression plasmids and parasite transfection were performed as described previously (9).
Immunoblot Analysis—Parasite protein lysates were analyzed by
SDS-PAGE. Resolved proteins were electrophoretically transferred to
nitrocellulose membranes (Schleicher & Schuell) for immunoblot analysis using polyclonal rabbit serum raised against denatured and deglycosylated gp63 purified by monoclonal affinity chromatography from
L. amazonensis (5). The secondary antibody used was horseradish peroxidase-conjugated goat anti-rabbit Fc and developed with ECL reagent
(Amersham Biosciences, Inc.).
Metabolic Labeling and Immunoprecipitation—For metabolic labeling, stationary phase promastigotes were washed three times in Hanks’
balanced salt solution and labeled for 30 min in Hanks’ containing 100
␮Ci ml⫺1 [35S]methionine/cysteine (specific activity ⫽ 100 mCi mmol⫺1;
PerkinElmer Life Sciences) per 108 cells. After labeling, cells were
washed three times in excess M199 medium and resuspended in M199
containing 10-fold excess methionine/cysteine and 0.1% bovine serum
albumin to a density of 3 ⫻ 107 cells ml⫺1. Cells were chased and
collected by centrifugation. Under these conditions, no change in cell
density was observed. Both cell pellets and conditioned medium after
centrifugation were retained for further analysis. All medium samples
were cleared by three rounds of centrifugation and passage through a
0.2-␮m filter before immunoprecipitation for 1 h at 4 °C with gp63specific antiserum at a dilution of 1:200. For collecting cell-associated
gp63, cell pellets (4 ⫻ 108 cells) were washed three times in cold
phosphate-buffered saline (PBS) and lysed on ice for 30 min after the
addition of 1 ml of PBS containing 0.5% Nonidet P-40. Insoluble material was removed by centrifugation at 14,000 rpm for 30 min in a
microcentrifuge. The samples were thereafter processed as described
for conditioned medium. Immune complexes, in each case, were collected by the addition of 50 ␮l of 33% protein-A-Sepharose (Sigma) for
1 h at 4 °C followed by three washes with 0.2 N NaCl, 0.5% Nonidet
P-40, 0.1% SDS and then processed for SDS-PAGE and gel autoradiography as described previously (24). Cell-associated and -released gp63
were assessed in the same aliquot of culture to evaluate their relative
levels. The conditions for immunoprecipitation of gp63 were optimized
to ensure the maximal recovery of protein (9). [3H]Ethanolamine incorporation was performed as described previously (25), and cells and
media were analyzed as described above for 35S incorporation. GPIphospholipase C (GPI-PLC) digestions were performed essentially as
described (9). Briefly, culture supernatants from stationary phase cells
were immunoprecipitated using gp63-specific antibodies covalently
linked to protein-A-Sepharose using dimethylpimelimidate (Sigma).
Beads with immune complexes were washed as described above and
washed twice more in GPI-PLC digestion buffer (50 mM Tris-HCl, pH
8.0, 5 mM EDTA, 1% Nonidet P-40) and divided into two fractions, one
treated with 200 ng of recombinant Trypanosoma brucei GPI-PLC (a
gift from K. Mensa-Wilmot) and another, used as a negative control.
Purified cell-derived gp63 in fresh culture medium was processed exactly as described above as a positive control. All samples were incubated for 3 h at 37 °C. To test the association of extracellular gp63 with
membrane vesicles, aliquots of spent medium from metabolically labeled cells were subjected to centrifugation for 1 h at 100,000 ⫻ g. gp63
was immunoprecipitated from the supernatant as described above. This
precipitated gp63, an uncentrifuged aliquot of the culture supernatant,
and the solubilized 100,000 ⫻ g pellet were analyzed by SDS-PAGE and
autoradiography. For graphical representation of kinetics and quantitation, blots were analyzed by laser-scanning densitometry.
Surface Biotinylation—Parasites were washed three times with PBS,
adjusted to a density of 108 ml⫺1 to which 1 ml of 0.5 mg ml⫺1 of
EZ-Link LC-sulfo-NHS-biotin (Pierce) was mixed, and incubated for 20
min on ice. The cells were then collected by centrifugation, resuspended
again in labeling solution, and incubated again for 20 min on ice. The
FIG. 1. Extracellular gp63 accumulates in stationary phase
cultures of several Leishmania species. Proteins of stationary
phase (1 ⫻ 108 ml⫺1) promastigotes were metabolically labeled by
incubating parasites in medium containing [35S]methionine/cysteine
followed by cold medium for the indicated times. gp63 and p36 (internal
protein control (27) from cells and culture medium were immunoprecipitated and analyzed by SDS-PAGE and autoradiography. The Leishmania species are indicated. The data shown are representative of five
independent experiments.
labeling was then quenched by washing the cells three times in excess
M199, and the sample was processed for gp63 analysis as described
above. For zinc chelation experiments, parasites were fixed in 0.5%
glutaraldehyde for 30 min on ice, washed three times in PBS, and
surface-biotinylated as described above. Zinc chelation was performed
by adding a stock solution of 1,10-phenanthroline to provide the various
final concentrations indicated. Biotinylated molecules were collected
using immobilized streptavidin (SA) as described above.
Extracellular gp63 Purification—Spent medium from cultured parasites was collected, filter-sterilized, and concentrated ⬃50-fold using a
Centricon 20 concentrator. Bovine serum albumin was removed using a
Hi-Trap Blue column (Amersham Biosciences, Inc.), and fractions were
assessed for the presence of gp63 by Western blotting. gp63-containing
fractions were pooled and subjected to anion exchange chromatography
using a Hi-Trap QFF column (Amersham Biosciences, Inc.) with a
continuous salt gradient of 0 –1 M NaCl. The fractions were analyzed by
Western blotting with gp63-specific polyclonal serum. gp63-containing
fractions were pooled and subjected to dialysis to remove excess salt
before further treatment. Endoglycosidase H treatment of gp63 was
performed at 37 °C at pH 5.5 before SDS-PAGE electrophoresis.
Protease Assay—Spent medium samples (250 ␮l each) from cultures
of stationary phase parasites were assayed for proteolytic activity using
azocasein as the substrate (26). After incubation, each sample was
diluted with 0.4 ml of PBS before the addition of 0.1 ml of ice-cold 100%
trichloroacetic acid to precipitate undigested azocasein. The precipitates were removed by microcentrifugation for 2 min. Caseinolytic
activity was determined by reading the absorbance of supernatants at
366 nm and expressed as ␮g of azocasein digested per h⫺1 ml⫺1 medium. The proteolytic activity of gp63 was assessed by comparison of
medium pre- and post-immunoprecipitation with gp63-specific
antiserum.
RESULTS AND DISCUSSION
gp63 Is Released from Promastigotes of Different Leishmania
Species—We previously found gp63 in the culture medium of L.
amazonensis transfectants expressing a mutant gp63 defective
in GPI anchor addition (9). To determine whether gp63 is
normally released spontaneously by wild type parasites, recently acquired clinical isolates causing cutaneous (L. tropica
EP39 – 41) or visceral (L. infantum EP32) disease were tested.
Our laboratory stock of wild type L. amazonensis was included
for comparison. Parasites were metabolically labeled with
[35S]methionine/cysteine and then chased for 12– 48 h with cold
methionine/cysteine. gp63 and a stable cytoplasmic protein,
p36 (27), were immunoprecipitated from cell and medium fractions and analyzed by SDS-PAGE and autoradiography (Fig.
1). This analysis revealed substantial levels of radiolabeled
gp63 in the supernatants of all promastigote cultures. Inter-
8804
Leishmania gp63 Release
FIG. 2. Release of gp63 is independent of glycosylation. A,
pulse-chase analysis of cell- and medium-associated gp63. A gp63deficient variant of L. amazonensis (gp63DV Cells) was transfected with
constructs encoding wild type gp63 (gp63WT), a gp63 mutant deficient in
glycosylation (gp63DG), or expression vector alone (vector). Proteins of
stationary phase (1 ⫻ 108 ml⫺1) promastigote transfectants were metabolically labeled by incubating parasites in medium containing
[35S]methionine/cysteine followed by cold medium for the indicated
times. Aliquots of culture were removed, and cell and medium fractions
were separated. gp63 was immunoprecipitated from each and analyzed
by SDS-PAGE and autoradiography. B, graphical representation of
gp63 release. The percent of total radiolabeled gp63 that is found in the
medium is presented as the ratio of protein precipitated from the
medium to protein precipitated from cells ⫹ medium. Background endogenous gp63 from the gp63DV cells (vector lanes) was subtracted from
that of the gp63WT transfectants to achieve measurement of the transgene wild type cellular gp63 (gp63WT).
estingly, radiolabeled gp63 was released by the recent clinical
(EP32-14) isolates, peaking after 12 h and decreasing slightly
from 24 to 48 h post-chase, probably because of degradation of
the protein. The gp63 released by different isolates was 63– 65
kDa in size, consistent with that seen previously (24). Coimmunoprecipitation of all samples with an antiserum specific
for a stable cytosolic protein, p36 (27), revealed its presence
only in cells (Fig. 1, Cell, p36) and not in the medium (Fig. 1,
Medium, p36), indicating that gp63 release did not occur as a
result of cell damage or death. These results demonstrate that
gp63 is released from virulent, wild type Leishmania isolates.
The recent isolates tested for gp63 had undergone less than five
in vitro passages. The release kinetics are similar to those of
our laboratory stock of virulent wild type L. amazonensis (14).
These findings together with the absence of significant gp63
release from avirulent L. amazonensis (Fig. 2A, see the next
paragraph), suggest an association of gp63 release with virulence. It is also possible that specific gp63 isoforms are selectively released reflective of their different functional roles.
Kinetics of gp63 Release and Role of Glycosylation—To investigate the mechanism of gp63 release from cultured promastigotes, we employed a gp63-deficient variant of L. amazonensis (gp63DV), which produces ⬃5% of the normal level of gp63
(14, 28). Analysis of gp63 proteins expressed in transfectant
gp63DV cells from specific transgenes facilitates the analysis of
protein release, because these cells otherwise release little
endogenous gp63 (seen in Fig. 2A, vector). Because glycosylation may play a role in protein trafficking and secretion (29 –31)
and mutations in the glycosylation sites of gp63 lead to reduced
surface expression (9), we examined the release of these mutants proteins. gp63DV promastigotes expressing wild type
gp63 (gp63WT) or a mutant lacking all three sites of N-linked
glycosylation (gp63DG) (9), both, released significant amounts
of gp63 into the culture supernatant (Fig. 2A). The kinetics of
extracellular accumulation of gp63 in these cultures was determined by immunoprecipitation of cell and medium fractions of
[35S]methionine/cysteine pulse-labeled stationary phase promastigotes, as described. Release of gp63WT was detected as
early as 4 h post-chase, with the extracellular concentration
peaking ⬃12 h post-labeling (Fig. 2A). gp63DG was released
with approximately the same kinetics (Fig. 2A, gp63DV and
gp63DG). Because gp63DG produced from the transgene in both
cell and medium fractions has increased electrophoretic mobility, it is possible to visualize the endogenous gp63 produced by
the gp63DV cells as well (Fig. 2) (14). Semi-quantitative analysis of cell-associated and released gp63 revealed that ⬃60 –70%
of gp63WT and gp63DG synthesized during the labeling period
was released by 12 h of chase (Fig. 2B). Thus, release of gp63 is
not affected by its extent of glycosylation. Moreover, the extent
of deglycosylation does not affect the kinetics of gp63 release.
In pulse-chase experiments, the cell-associated deglycosylated
mutant was apparent within 1 h post-chase and became detectable extracellularly within 4 h (Fig. 2), similar to that observed
for wild type (Figs. 2 and 3) and GPI-mutant gp63s (not shown).
In the case of VSG (32), it has been proposed that GPI-anchoring may provide a forward targeting signal inasmuch as VSG
reporters lacking GPI-anchoring motifs have slower forward
movement out of the endoplasmic reticulum, although they do
eventually become secreted (32). In light of this, the at least 3 h
lag time between the appearance of cell-associated gp63 and
release of gp63 may be related to the absence of a GPI moiety
in the latter, hastening its exit from the cell.
Multiple Modes of gp63 Release; Surface-derived gp63—At
least some of the released gp63 is derived from that preexisting
on the parasite surface. This was demonstrated by examining
gp63DV cells expressing gp63WT protein that were either surface-biotinylated (Fig. 3A) or labeled with [35S]methionine/cysteine (Fig. 3B). Biotinylated gp63 was increasingly released in
the medium over the 48-h time period, indicating that it was
derived from the parasite surface. Concurrent analysis of the
release of [35S]methionine/cysteine-labeled gp63 over the same
time period showed a peak extracellular concentration after
12 h of chase, with a minimal decline in protein levels over
successive time points (Fig. 3B). This suggests the existence of
at least two pools of extracellular gp63, released with different
kinetics.
Multiple Modes of gp63 Release; Secreted gp63—There are
several possible explanations for increasing levels of biotinylated, extracellular gp63 in the setting of steady to decreasing
amounts of radiolabeled extracellular gp63 (Figs. 2 and 3). The
simplest is that some of the extracellular gp63 is directly secreted by the parasite without first being localized at the cell
surface. This interpretation is suggested by results of double
labeling experiments. Wild type L. amazonensis promastigotes
were first labeled with [35S]methionine/cysteine, then surfacebiotinylated, and finally chased in medium for increasing time
(Fig. 4). gp63 from cell and medium fractions was isolated by
immunoprecipitation or by chromatography on immobilized
streptavidin for analysis, as described under “Experimental
Procedures” (Fig. 4, Cell and Medium panels). The cellular
fractions contained a constant amount of radiolabeled gp63
Leishmania gp63 Release
8805
FIG. 3. Kinetics of surface and extracellular gp63 accumulation. L. amazonensis promastigotes were either surface-biotinylated (A) or
labeled with [35S]methionine/cysteine (B) and incubated in medium for the indicated times. gp63 was immunoprecipitated from cell and medium
fractions and analyzed by SDS-PAGE and blotting with HRP-SA to detect biotinylated gp63 (A) or autoradiography (B). units ⫽ total pixel density
by laser densitometry.
FIG. 4. Extracellular gp63 accumulation in wild type L. amazonensis. Stationary phase (1 ⫻ 108 ml⫺1) L. amazonensis promastigotes were cultured for 30 min with [35S]methionine/cysteine, chased for
60 min in fresh medium, and surface-biotinylated. The cells were then
incubated in fresh medium for varying times, and cell- and mediumassociated gp63 were purified by immunoprecipitation with gp63-specific antiserum and protein-A-Sepharose. Immunoprecipitates were analyzed by SDS-PAGE followed either by either gel autoradiography (to
detect radiolabeled gp63) or by blotting with SA-HRP (to detect biotinylated gp63). Alternatively, biotinylated surface proteins were purified from the medium by SA chromatography and detected by SDSPAGE and gel autoradiography (gp63 is the major species). Finally,
medium that had been cleared of biotinylated protein by SA chromatography was subjected to gp63 immunoprecipitation followed by SAHRP blotting (to confirm the absence of residual biotinylated gp63) and
autoradiography (to detect nonbiotinylated gp63 in the medium).
(Cell, gp63, Autorad), whereas surface-labeled gp63 steadily
declined with time (Cell, gp63, SA-HRP). Analysis of the medium fraction revealed increasing accumulation of surface-released gp63 (Medium, gp63, SA-HRP) as seen in previous single-labeling experiments. Some of the biotinylated gp63 was
radioactive (Medium, SA, Autorad), indicating that it acquired
a surface location after biosynthesis. We also analyzed the
medium after removal of biotinylated material to determine
whether some of the released gp63 may be biosynthetically
radiolabeled but not biotinylated (Fig. 4, Medium Post-SA
Clear). The gp63 contained in this ”SA-cleared“ medium did not
contain any biotinylated gp63 (Medium Post-SA Clear, gp63,
SA-HRP) but did contain increasing concentrations of radiolabeled, nonbiotinylated gp63 (Medium Post-SA Clear, gp63,
Autorad).
Because the vast majority of 35S-labeled gp63 destined for
the surface-release pathway acquires its surface location during the first 60 min of chase (data not shown), it is unlikely that
the radiolabeled, nonbiotinylated gp63 seen in the bottom autoradiogram trafficked through the surface after the biotinylation step. Rather, this material was likely derived from another
intracellular pool that never became localized on the surface
before its release. Another less likely possibility could be that
released gp63 was endocytosed by the parasites. It has recently
been shown that cell surface gp63 appears to be internalized,
probably during surface membrane recycling (33). Thus, the
decreasing surface-biotinylated gp63 may be reduced partially
by this mechanism. Because the total biotinylated cellular gp63
fraction appears to decrease over time this would indicate that
such internalized material is either degraded or released. It is
also possible that a fraction of internalized surface gp63 may be
secreted, accounting for some biotinylated gp63 accumulating
extracellularly. Could secretion of gp63 occur simply as a result
of oversaturation of the GPI-anchoring pathway during protein
synthesis and maturation? Wild type Leishmania release significant amounts of gp63 (Fig. 1). In addition, wild type L.
amazonensis overexpressing transgenetically derived gp63
have increased levels of cell-associated (34) and released gp63
(not shown). If the GPI pathway were constitutively oversaturated, accounting for the natural endogenous release from the
wild type cells, then one would expect only further release of
gp63 in these transfectants and stable amounts of cell-associ-
8806
Leishmania gp63 Release
FIG. 5. Released gp63 lacks a GPI anchor. A, cell- and mediumassociated gp63 was immunoprecipitated (IP) with gp63-specific antiserum digested (⫹) or not digested (⫺) with GPI-PLC and analyzed by
immunoblot analysis with gp63- or cross-reacting determinant (CRD)specific antibodies. B, stationary phase, gp63-deficient variant promastigotes expressing wild type gp63 (gp63WT) or a gp63 point mutant
defective in GPI addition (gp63N577L) was metabolically labeled with
[35S]methionine/cysteine. Media derived from these cells were immunoprecipitated with gp63-specific polyclonal antiserum (gp63Poly) or
monoclonal antibody 6H12, which is specific for a form of gp63 containing a GPI anchor.2
FIG. 6. Released gp63 lacks ethanolamine. Stationary phase wild
type L. amazonensis promastigotes or gp63-deficient variant promastigotes expressing either wild type gp63 or a gp63 point mutant deficient in GPI addition (N577L) were metabolically labeled with either
[35S]methionine/cysteine or [3H]ethanolamine and chased for 18 h with
fresh medium. Cell- and medium-associated gp63 was immunoprecipitated with gp63-specific antiserum and analyzed by SDS-PAGE and
autoradiography.
ated gp63. We postulate that there is a constitutive distribution
of gp63 between that destined for cell membrane location and
that which is released extracellularly. The released protein
likely derives from both proteolytic cleavage of surface gp63
and direct secretion.
Extracellular gp63 Is Devoid of a GPI Anchor and Is Proteolytically Active—Multiple mechanisms could account for the
release of gp63 from the cell surface. We first tested the possibility that extracellular gp63 was released from cell membrane
sloughing or breakup of undetected dead cells in the culture.
Cell culture supernatants were collected from metabolically
labeled cells and subjected to a 100,000 ⫻ g centrifugation for
1 h. gp63 was immunoprecipitated from equal volumes of centrifuged and uncentrifuged spent medium and from the pellet
material and analyzed by SDS-PAGE and autoradiography.
Equal amounts of radiolabeled gp63 were present in the supernatant fractions of both samples, whereas no gp63 was detected
in the pellet, indicating that released gp63 is not associated
with membrane vesicles (not shown). Another potential mechanism could involve GPI-PLC cleavage of the GPI anchor of
gp63, which exposes the inositol 1,2-cyclic monophosphate
epitope known as the cross-reacting determinant (35). Such
GPI-PLC activity has been reported in related parasites Trypanosoma cruzi (36) and T. brucei (37, 38). Immunoprecipitated
extracellular gp63 digested with GPI-PLC failed to react with
FIG. 7. Released gp63 is proteolytically active. Media from stationary phase gp63-deficient promastigotes harboring pX vector, pXgp63WT, or pX-gp63WT and cleared of gp63 by immunoprecipitation (IP)
were tested for caseinolytic activity as described under “Experimental
Procedures.”
FIG. 8. Release of surface gp63 requires zinc. A, wild type L.
amazonensis promastigotes were surface-biotinylated and then chased
in medium for 12 h in the absence (0) or presence of 10 and 25 mM
1,10-phenanthroline. gp63 was then immunoprecipitated from the medium and analyzed by SDS-PAGE and immunoblotting with SA-HRP
(left panel). In a separate experiment, the kinetics of surface release
were examined by analyzing gp63 release after 1 and 12 h of chase in
the presence of 0 or 25 mM 1,10-phenanthroline (OP) as indicated
(middle and right panels). B, stationary phase, gp63-deficient L. amazonensis promastigotes expressing either wild type gp63 or a gp63 point
mutant deficient in protease activity addition (E265D) were lightly
fixed with glutaraldehyde, surface-biotinylated, washed, and incubated
in medium for 24 h. gp63 was immunoprecipitated from the medium
and analyzed by immunoblotting with SA-HRP (left panel). Living
promastigotes were analyzed in a similar manner 12 and 18 h after
surface biotinylation, with cell-associated gp63 purified also as a control
(right panels).
antibodies specific to the cross-reacting determinant (CRD)
epitopes in immunoblotting experiments (Fig. 5A, Medium,
compare ⫺ and (⫹), whereas purified cellular gp63, which
contains an intact GPI-anchor, reacted as expected (Fig. 5A,
Cell). This indicates that released gp63 lacks a GPI anchor.
Monoclonal antibody 6H12, initially used to purify gp63 from L.
amazonensis (5), is specific for gp63 containing the GPI anchor.2 We used this reagent to indirectly test extracellular gp63
for the presence of a GPI anchor. Control immunoprecipitations
using polyclonal serum indicate the presence of extracellular
gp63 from metabolically labeled Leishmania (Fig. 5B, gp63Poly)
from either transfectants (gp63DV Cells) expressing the wild
2
M. Chaudhuri and K.-P. Chang, unpublished data.
Leishmania gp63 Release
type gp63 (gp63WT) or the gp63 mutant in the GPI addition site
(gp63N577L). The monoclonal antibody failed to immunoprecipitate extracellular gp63 from the same supernatants (Fig. 5B,
gp63GPI).
Together these data support the idea that extracellular gp63
lacks a GPI anchor. Alternatively, released gp63 could possess
a GPI anchor either resistant to GPI-PLC digestion, as seen in
T. brucei (39), and/or not reactive with the antibody. To test if
released gp63 had received a GPI anchor during its biosynthesis, we examined the incorporation of ethanolamine, which
covalently links the GPI moiety with the carboxyl group to the
C-terminal amino acid of the protein. Wild type L. amazonensis
(Fig. 6, WT Cells) and deficient variant transfected with the
wild type gp63 transgene (gp63DV Cells ⫹ gp63WT) or the GPI
site-specific mutant (gp63DV Cells ⫹ gp63N577L) were each metabolically labeled separately with [35S]methionine/cysteine or
[3H]ethanolamine, chased for 18 h, and processed for analysis
of cellular and released gp63 as indicated under “Experimental
Procedures.” The gp63N577L mutant was completely released
into the culture supernatant as was previously shown (9). Ethanolamine was incorporated in the cellular gp63 fractions of all
cell types, indicating the presence of the ethanolamine bridge
linking the protein to the GPI anchor (Fig. 6, [3H]Etn, Cell).
gp63 immunoprecipitated from the supernatant was ⬃3–5-fold
less in the deficient variants expressing the wild type gp63
than the GPI addition mutant or untransfected wild type L.
amazonensis (Fig. 6, [35S]Met/Cys, Medium). Ethanolamine
incorporation could not be detected in the released gp63 (Fig. 6,
[3H]Etn, Medium) even after prolonged exposure of the film up
to 1 month (not shown). These data further show that released
gp63 contains neither a GPI anchor nor an ethanolamine molecule, ruling out the possibility that it was released by
GPI-phospholipolysis.
Finally, cell culture supernatants were found to contain proteolytic activity. Culture supernatants from transfectants containing either vector alone (Fig. 7, Vector) or those expressing
wild type gp63 (gp63, Pre-IP) were assessed for caseinolytic
activity. The latter had approximately 10-fold higher levels of
proteolytic activity. This activity is gp63-specific, since the
caseinolytic activity in the medium from transfectants could be
removed by immunoprecipitation with anti-gp63 antiserum
(Fig. 7, compare gp63, Pre- and Post-IP). Taken together, these
data show that extracellular gp63 is soluble, proteolytically
active, and devoid of a GPI anchor. The finding that gp63 is
released naturally from Leishmania species is surprising, since
the molecule is an abundant membrane protein linked to the
surface via a GPI anchor. Release of GPI-linked surface molecules has been reported in the trypanosomatid protozoa. T.
brucei VSG can be shed by two mechanisms. First, it can be
proteolytically cleaved from the surface of bloodstream parasites upon their differentiation into the procyclic forms (40).
Interestingly, a zinc-dependent metalloprotease has been implicated in this process, and a likely candidate is a gp63 homologue (32). VSG can also be released by GPI-PLC-mediated
cleavage of its GPI anchor, the stimulus for which may involve
multiple triggers such as osmotic stress, trypsin digestion, or
changes in pH (23).
Proteolytic Cell Surface Release of gp63—gp63 release from
the cell surface may result from proteolytic cleavage, possibly
autoproteolysis. Surface-biotinylated parasites were assessed
for their ability to release gp63 in the presence of increasing
amounts of the zinc chelator, 1,10-phenanthroline, a well
known inhibitor of gp63 proteolytic activity (13). Because this
compound is toxic to cells during the incubation time frame
used, we chose to use lightly glutaraldehyde-fixed cells in
which gp63 was still proteolytically active on the surface (13,
8807
FIG. 9. Size differences among cell-associated and extracellular forms of gp63. gp63 purified from the medium (M) of cultured
parasites or from the cell surface (C) was analyzed by SDS-PAGE before
(untreated, UT) and after endoglycosidase H (Endo H) treatment. a, the
63-kDa gp63 form present in both cellular and medium fractions before
endoglycosidase H treatment; b, the 61-kDa gp63 form present in both
cellular and medium preparations after endoglycosidase H treatment;
c, 59-kDa gp63 form present in the medium-derived gp63 both before
and after endoglycosidase H treatment.
26). We performed Western blot analysis of immunoprecipitated gp63 from the supernatants of surface-biotinylated, fixed
wild type Leishmania (Fig. 8). The amount of gp63 released
was diminished to ⬃50% in the presence of 10 mM chelator and
is barely detectable in the presence of 25 mM chelator (Fig. 8A).
Further increases of the concentration of chelator resulted in
precipitation of the inhibitor and cells. The kinetics of gp63
release from parasites over a 12-h period was tested in the
presence of 25 mM chelator. Under these conditions, the release
of biotinylated gp63 was reduced by ⬃7-fold. These results
further show that the release is a zinc-dependent process (Fig.
8A, right panels).
To more directly test the autoproteolytic hypothesis, transfectants expressing surface-localized wild type gp63 (gp63WT)
and those expressing a site-specific mutant in the active site of
the enzyme (gp63E265D) (9) were compared for their capacity to
release surface gp63. Using both fixed and live parasites, quantitatively less surface-biotinylated gp63 was released from
transfectants expressing the proteolytically inactive mutant
(Fig. 8B). This further indicates that the release of gp63 from
the cell surface is mediated by gp63 autoproteolysis but does
not fully exclude the possibility that another protease or an
unrelated process dependent upon divalent cations may be
involved. In a similar fashion, the GPI-linked matrix metalloprotease, MMP-17, has been shown to be released extracellularly from transfected mammalian cells by matrix metalloproteinase activity (41). Whether this is an autoproteolytic event is
not known.
We purified extracellular gp63 using anion exchange chromatography and compared its electrophoretic mobility with
that of cell-associated gp63 (Fig. 9). Purified extracellular gp63
(UT, M) is composed of two predominant forms of ⬃63 (a) and
59 kDa (c), whereas cellular gp63 (UT, C) is composed of one
predominant form of 63 kDa. Endoglycosidase H treatment of
cellular gp63 reduces the size of the protein to 61 kDa (b)
consistent with the loss of N-linked glycans. Endoglycosidase
treatment of extracellular gp63 produces two bands of ⬃61 (b)
and 59 kDa (c). The 61-kDa protein probably represents fulllength, deglycosylated gp63, whereas the smaller, 59-kDa protein has likely undergone proteolysis and is devoid of N-glycans. The substrate cleavage site specificity of gp63 is not
known. Peptide bond cleavage has been noted to occur Nterminal of hydrophobic residues at the P1⬘ position with basic
residues at P2⬘ and P3⬘ (42) or at serine or threonine residues
(43). The gp63 sequence between the third glycosylation site
and the GPI addition site contains one serine adjacent to a
threonine residue (at positions 540 and 541 in the L. amazonen-
8808
Leishmania gp63 Release
FIG. 10. Hypothetical model of release/secretion of gp63. The two mechanisms of gp63 release identified in this
report are shown. GPI-anchored gp63
may be released from the cell surface by
autoproteolytic cleavage N-terminal to
the GPI addition site. Alternatively, unmodified gp63 may be directly secreted
via vesicular trafficking through the
flagellar pocket.
sis sequence) and downstream serine (at position 543) and
threonine (position 553) residues, which represent potential
autocleavage sites.
The collective results shown in this paper allow us to
develop a representation of the proposed pathways of gp63
release (Fig. 10). Both follow the classic secretory pathway
with routing of gp63 from the endoplasmic reticulum to the
Golgi network and then into the flagellar pocket. Vesicles
containing both membrane-bound and free gp63 fuse with
flagellar pocket membrane, with free gp63 thereafter released extracellularly. Membrane-bound gp63 is located
throughout the external cell membrane, and a proportion of
this is released directly via zinc/proteolysis. We have
documented the significant release of active gp63, a metalloprotease with broad substrate specificity, from the human
pathogen Leishmania. The membrane-bound form of gp63
has been shown to be important for parasite virulence, being
involved in the evasion of complement-mediated lysis, macrophage attachment, and intracellular survival of parasites. A
myriad of functions can be hypothesized for an extracellularly released metalloprotease with broad substrate specificity. gp63 digests gelatin and fibrinogen (9, 44) and, thus, can
be envisioned to play a role in the degradation of extracellular matrix proteins by parasites after their inoculation by the
insect vector. This function may serve to allow greater local
mobility of invading parasites, enhancing their capacity to
bind to and parasitize host macrophages. Extracellular gp63
could allow parasites to evade a variety of anti-microbial
factors in the extracellular environment, such as complement
(18, 45). It will be of interest to determine the capacity of
amastigotes to release gp63, possibly influencing macrophage
function or escape of the organism. Furthermore, within the
sandfly vector, release of extracellular gp63 from promastigotes could enhance the hydrolysis of protein substrates and
play a nutritional and/or protective role (6). We are currently
testing the potential role(s) of gp63 in these processes.
Acknowledgments—We thank Drs. K. Mensa-Wilmot (University of
Georgia, Athens, GA) and J. Bangs (University of Wisconsin, Madison,
WI) for reagents and helpful suggestions. We also thank Drs. S.
Ozensoy, Z. Alkan and A. Ozcel (Ege University, Izmir, Turkey) for the
clinical isolates.
REFERENCES
1. McKerrow, J. H., Sun, E., Rosenthal, P. J., and Bouvier, J. (1993) Annu. Rev.
Microbiol. 47, 821– 853
2. Johansson, N., Ahonen, M., and Kahari, V. M. (2000) Cell. Mol. Life Sci. 57,
5–15
3. Yu, Q., and Stamenkovic, I. (2000) Genes Dev. 14, 163–176
4. Bouvier, J., Schneider, P., and Etges, R. (1995) Methods Enzymol. 248,
614 – 633
5. Chang, C. S., and Chang, K. P. (1986) Proc. Natl. Acad. Sci. U. S. A. 83,
100 –104
6. Inverso, J. A., Medina-Acosta, E., O’Connor, J., Russell, D. G., and Cross, G. A.
(1993) Mol. Biochem. Parasitol. 57, 47–54
7. El-Sayed, N. M., and Donelson, J. E. (1997) J. Biol. Chem. 272, 26742–26748
8. Grandgenett, P. M., Coughlin, B. C., Kirchhoff, L. V., and Donelson, J. E.
(2000) Mol. Biochem. Parasitol. 110, 409 – 415
9. McGwire, B. S., and Chang, K. P. (1996) J. Biol. Chem. 271, 7903–7909
10. Button, L. L., Wilson, G., Astell, C. R., and McMaster, W. R. (1993) Gene 134,
75– 81
11. Macdonald, M. H., Morrison, C. J., and McMaster, W. R. (1995) Biochim.
Biophys. Acta 1253, 199 –207
12. Russell, D. G., and Wilhelm, H. (1986) J. Immunol. 136, 2613–2620
13. Chaudhuri, G., Chaudhuri, M., Pan, A., and Chang, K. P. (1989) J. Biol. Chem.
264, 7483–7489
14. McGwire, B., and Chang, K. P. (1994) Mol. Biochem. Parasitol. 66, 345–347
15. Brittingham, A., Morrison, C. J., McMaster, W. R., McGwire, B. S., Chang,
K. P., and Mosser, D. M. (1995) J. Immunol. 155, 3102–3111
16. Rizvi, F. S., Ouaissi, M. A., Marty, B., Santoro, F., and Capron, A. (1988) Eur.
J. Immunol. 18, 473– 476
17. Soteriadou, K. P., Remoundos, M. S., Katsikas, M. C., Tzinia, A. K., Tsikaris,
V., Sakarellos, C., and Tzartos, S. J. (1992) J. Biol. Chem. 267,
13980 –13985
18. Brittingham, A., Chen, G., McGwire, B. S., Chang, K. P., and Mosser, D. M.
(1999) Infect. Immun. 67, 4477– 4484
19. Chang, K. P., and Dwyer, D. M. (1976) Science 193, 678 – 680
20. Ferguson, M. A., Brimacombe, J. S., Cottaz, S., Field, R. A., Guther, L. S.,
Homans, S. W., McConville, M. J., Mehlert, A., Milne, K. G., and Ralton,
J. E. (1994) Parasitology 108, (suppl.) 45–54
21. Mensa-Wilmot, K., LeBowitz, J. H., Chang, K. P., al-Qahtani, A., McGwire,
B. S., Tucker, S., and Morris, J. C. (1994) J. Cell Biol. 124, 935–947
22. Garg, N., Tarleton, R. L., and Mensa-Wilmot, K. (1997) J. Biol. Chem. 272,
12482–12491
23. Rolin, S., Hanocq-Quertier, J., Paturiaux-Hanocq, F., Nolan, D., Salmon, D.,
Webb, H., Carrington, M., Voorheis, P., and Pays, E. (1996) J. Biol. Chem.
271, 10844 –10852
24. Chang, C. S., Inserra, T. J., Kink, J. A., Fong, D., and Chang, K. P. (1986) Mol.
Biochem. Parasitol. 18, 197–210
25. Mensa-Wilmot, K., LeBowitz, J. H., Chang, K. P., al-Qahtani, A., McGwire,
B. S., Tucker, S., and Morris, J. C. (1994) Braz. J. Med. Biol. Res. 27,
177–184
26. Chaudhuri, G., and Chang, K. P. (1988) Mol. Biochem. Parasitol. 27, 43–52
27. Liu, X., and Chang, K. P. (1994) Mol. Biochem. Parasitol. 66, 201–210
28. Liu, X., and Chang, K. P. (1992) Proc. Natl. Acad. Sci. U. S. A. 89, 4991– 4995
29. Morrison, C. J., Easton, R. L., Morris, H. R., McMaster, W. R., Piret, J. M., and
Dell, A. (2000) Biotechnol. Bioeng. 68, 407– 421
30. Zheng, X., Lu, D., and Sadler, J. E. (1999) J. Biol. Chem. 274, 1596 –1605
31. Allen, S., Naim, H. Y., and Bulleid, N. J. (1995) J. Biol. Chem. 270, 4797– 4804
32. Bangs, J. D., Ransom, D. M., McDowell, M. A., and Brouch, E. M. (1997) EMBO
J. 16, 4285– 4294
33. Weise, F., Stierhof, Y. D., Kuhn, C., Wiese, M., and Overath, P. (2000) J. Cell
Sci. 113, 4587– 4603
34. Chen, D. Q., Kolli, B. K., Yadava, N., Lu, H. G., Gilman-Sachs, A., Peterson,
D. A., and Chang, K. P. (2000) Infect. Immun. 68, 80 – 86
35. Zamze, S. E., Ferguson, M. A., Collins, R., Dwek, R. A., and Rademacher, T. W.
Leishmania gp63 Release
(1988) Eur. J. Biochem. 176, 527–534
36. Redpath, M. B., Carnall, N., Webb, H., Courel, M., Amorim, A., Guther, M. L.,
Cardoso de Almeida, M. L., and Carrington, M. (1998) Mol. Biochem. Parasitol. 94, 113–121
37. Mensa-Wilmot, K., Hereld, D., and Englund, P. T. (1990) Mol. Cell. Biol. 10,
720 –726
38. Webb, H., Carnall, N., and Carrington, M. (1994) Braz. J. Med. Biol. Res. 27,
349 –356
39. Field, M. C., Menon, A. K., and Cross, G. A. (1991) J. Biol. Chem. 266,
8392– 8400
40. Ziegelbauer, K., Stahl, B., Karas, M., Stierhof, Y. D., and Overath, P. (1993)
8809
Biochemistry 32, 3737–3742
41. Itoh, Y., Kajita, M., Kinoh, H., Mori, H., Okada, A., and Seiki, M. (1999) J. Biol.
Chem. 274, 34260 –34266
42. Bouvier, J., Schneider, P., Etges, R., and Bordier, C. (1990) Biochemistry 29,
10113–10119
43. Ip, H. S., Orn, A., Russell, D. G., and Cross, G. A. (1990) Mol. Biochem.
Parasitol. 40, 163–172
44. Nunes, A. C., and Ramalho-Pinto, F. J. (1996) Braz. J. Med. Biol. Res. 29,
1633–1640
45. Joshi, P. B., Sacks, D. L., Modi, G., and McMaster, W. R. (1998) Mol. Microbiol.
27, 519 –530