| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
,
*
Department of Medicine, University of California at San Diego, La Jolla, CA 92093; and
Department of Pathology and
Center for Molecular Genetics, University of California at San Diego, San Diego, CA 92103
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The parasite exists in two forms, the infectious cyst, which is resistant to many environmental factors, and the disease-causing trophozoite, which colonizes the intestinal lumen but does not invade the mucosa (4, 5). Excystation of ingested cysts occurs in response to luminal host signals, specifically the low pH encountered during passage through the stomach, followed by the elevated pH and proteases of the duodenum (6). Differentiation of trophozoites into cysts is induced by an elevated pH and bile, characteristic of the small intestine (6).
G. lamblia mostly colonizes the duodenum and jejunum, although a significant proportion of patients also have ileal and, occasionally, colonic colonization (7). Despite the often dramatic symptoms of infection, in >95% of patients with nonspecific gastrointestinal complaints and Giardia-positive duodenal biopsies, the duodenal mucosa was considered normal, i.e., without inflammation or grossly altered epithelium, although some degree of villous shortening often can be seen (7). Only a small fraction (3–4%) of patients with Giardia-positive duodenal biopsies have some degree of duodenitis with infiltration of neutrophils and lymphocytes (7). Despite the lack of an inflammatory response, Giardia infection induces a host immune response, as primary infection of mice with Giardia muris confers protection against secondary challenge (8, 9). In G. muris-infected mice, CD4 T cells and B cells are important for clearing the infection (8, 9, 10), while CD8 T cells and NK cells play no role in parasite clearance (8, 11). Similar conclusions appear to apply also to G. lamblia pathogenesis (12), although the host response to G. lamblia has been characterized less extensively. Among potential host effector mechanisms, antigiardial Abs, particularly of the IgA isotype, are thought to play an important role in parasite clearance (13), although Xid mice do not clear a primary Giardia infection despite high levels of antigiardial IgA (10), and IgA-deficient patients have only a slight increase in the incidence of G. lamblia infection (14). This suggests that other host defense mechanisms also play a role in Giardia clearance. For example, defensins, small anti-microbial peptides produced by intestinal epithelial cells, and lactoferrin can kill G. lamblia in vitro (15, 16), although their roles in vivo are not known.
NO is antimicrobial for a wide range of bacterial and parasitic pathogens (17, 18, 19) and has multiple other functions, including a role in neurotransmission and regulation of mucosal barrier integrity and vascular tone in the gut (20). NO is produced enzymatically from arginine through the action of NO synthase (NOS)4, which exists in three isoforms, neuronal NOS (NOS1), inducible NOS (iNOS; NOS2), and endothelial NOS (NOS3). In many cell types, the expression of iNOS is inducible by cytokines and microbial products, and iNOS is the major NOS isoform expressed by intestinal epithelial cells (21, 22). Expression of iNOS in these cells can either be constitutive, as in mouse ileum (23) and isolated normal human duodenocytes (24), or inducible in vivo during colonic inflammation (25) or in vitro by cytokines (21) or in response to infection with invasive bacteria (22, 26). In polarized intestinal epithelial cells, the stable NO end products, nitrite and nitrate, are preferentially detected at the apical side, suggesting that relevant targets for epithelial cell-derived NO and its metabolites may be located on the luminal side of the cells (22).
Because trophozoites colonize in close apposition to epithelial cells in the small intestine (5), epithelial NO production may be a potential host defense against G. lamblia. Therefore, in the present studies, we evaluated the ability of NO to affect giardial viability, growth, and differentiation. We show that NO inhibits growth, encystation, and excystation of G. lamblia, but has no effect on giardial viability. Furthermore, G. lamblia infection of human intestinal epithelial cell cultures inhibits epithelial NO production, suggesting that the parasite has evolved strategies to evade this potential host defense mechanism.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
G. lamblia trophozoites (strain WB, clone C6, ATCC 30957) were cultured at 37°C in trophozoite growth medium (Diamond TYI-S-33 medium containing 10% adult bovine serum, pH 7.1, supplemented with 0.5 mg/ml bovine bile, but without added iron, vitamins, or antibiotics).
Trophozoite growth assays
Trophozoites were chilled, counted, and diluted to 104/ml in trophozoite growth medium. Freshly prepared NO donors or control compounds (e.g., GSNO, GSH, GSSG, spermine NONOate, spermine, sodium nitrite, and sodium nitrate; all obtained from Sigma, St. Louis, MO) were added at the specified concentrations, and cultures were incubated at 37°C. Cultures were harvested by chilling, and live and dead trophozoites were enumerated microscopically in the presence of trypan blue.
Encystation
In vitro encystation was conducted under conditions described previously (27). Trophozoite cultures grown to late log phase were chilled, and trophozoites (3 x 103/ml) were added to freshly prepared preencystation medium (TYI-S-33 medium containing 10% adult bovine serum, pH 7.1, with 500 µg/ml piperacillin and 125 µg/ml amikacin, but without bile). After 72 h in culture, attached trophozoites were refed with encystation medium (TYI-S-33 medium containing 10% adult bovine serum, adjusted to pH 7.8 with 1 M NaOH, supplemented with 5 mM lactic acid, 0.25 mg/ml porcine bile, and antibiotics) and cultured for an additional 18 h, after which the number of encystation-specific secretory vesicles (ESVs) was determined by Nomarski differential interference-contrast microscopy. To obtain in vitro derived cysts for excystation experiments, cultures were kept in encystation medium for a total of 66 h. Subsequently, medium containing the nonattached cells, including the cysts, was removed and centrifuged at 2200 x g for 5 min at room temperature. Trophozoites and incomplete cysts were lysed by incubation in double-distilled water for 20 min at room temperature. Cysts were washed several times with water at 4°C and stored in water at 4°C overnight before excystation experiments to lyse Giardia with incomplete cyst walls.
Excystation
In vitro excystation was conducted by a two-step method (27, 28). In vitro derived cysts were first treated with water for 60 min at 4°C (pre-excystation stage) and then suspended in an acidic excystation stage I solution (pH 4.0) that models conditions in the human stomach, containing reduced glutathione and L-cysteine in HBSS, and incubated for 20 min at 37°C. After centrifugation, the cyst pellet was resuspended in slightly alkaline excystation stage II solution (pH 8.0) that models conditions in the small intestine, containing 1 mg/ml trypsin (type 2, from porcine pancreas) in Tyrode’s salt solution, and incubated for 60 min at 37°C. Cells were collected by centrifugation and resuspended in prewarmed TYI-S-33 growth medium with bovine bile. After 30-min incubation in a water bath at 37°C, motile excysted trophozoites were counted in a hemocytometer.
Immunoblot analysis of giardial cyst wall protein 2 (CWP2) expression
For the analysis of giardial CWP2 expression, cells were harvested after 18 h of encystation and washed three times in cold PBS, with protease inhibitors in the final wash (1 mM PMSF and 1 µM trans-epoxysuccinyl-L-leucylamido(4-guanidino)-butane). Cell pellets were resuspended in cold 6% TCA, and the suspension was incubated on ice for 10 min. Lysates were centrifuged for 3 min at 13,000 rpm in a microfuge at 4°C, and pellets were snap-frozen in liquid nitrogen and stored at -80°C. After thawing, pellets were neutralized with NaOH, resuspended in SDS-PAGE sample buffer, and boiled for 6 min. Protein concentrations were determined by the Bradford method, and 6 µg protein/lane was electrophoretically size-fractionated on a 10% denaturing, reducing polyacrylamide gel. Proteins were transferred to nitrocellulose, and blots were stained with a mAb (8C5) against CWP2 (29) as described previously (30). As a control, blots were reacted with a mAb against the giardial lectin, taglin (31).
Human intestinal epithelial cell lines
HT-29 and HCT-8 human intestinal epithelial cells were obtained from American Type Culture Collection (Manassas, VA). Caco-2 human colon epithelial cells were a gift from S. Tzipori (Tufts University, North Grafton, MA). Cell lines were grown in DMEM supplemented with 10% FCS and 2 mM L-glutamine. In experiments to determine arginine dependence of epithelial NO production, cells were cultured in arginine-free DMEM supplemented with 10% dialyzed FCS (dialysis was performed for 24 h at 4°C using a dialysis membrane with a 1000 m.w. cut-off). To obtain polarized Caco-2 monolayers, 106 cells were seeded onto collagen-coated microporous filter inserts (0.4 µm pore size, 4.7 cm2 growth area; Transwells, Costar, Cambridge, MA) and cultured for 7–10 days. Development of tight junctions was confirmed by determining transepithelial resistance with a Millicell ERS electrical resistance system (Millipore, Bedford, MA).
G. lamblia infection of intestinal epithelial cell cultures
Epithelial cells were seeded into six-well tissue culture plates
and grown for 1–3 days post-confluence, at which time the cultures
contained 2–5 x 106 cells/well. G.
lamblia trophozoites were washed twice with cold PBS and
resuspended in DMEM with 10% FCS and 2 mM glutamine. Trophozoites were
added to each well in a 1-ml volume, and cultures were incubated at
37°C for 18 h or the indicated times for time-course analyses.
Supernatants were collected, centrifuged to remove debris, and stored
at 4°C until analysis for assaying nitrite/nitrate and IL-8 and at
-80°C for assaying amino acid levels. To stimulate epithelial NO and
IL-8 production, a combination of 20 ng/ml human IL-1
, 20 ng/ml
human TNF-, and 50 ng/ml human IFN-
was added to the cultures at
the time of infection. Viable epithelial cells at the end of the
culture period were counted in the presence of trypan blue after
detachment with trypsin/EDTA.
Determination of nitrite/nitrate and IL-8 in culture supernatants
To determine levels of the stable NO end products, nitrite and nitrate, nitrate was first reduced to nitrite by incubating the samples for 60 min at room temperature with 0.05 U/ml nitrate reductase (from maize, Oxford Biomedical, Oxford, MI) and 0.1 mM ß-NADH in a buffer of 25 mM MOPS (pH 7.0) and 0.5 mM EDTA. Nitrite levels were then determined using the Griess reaction. Briefly, 100 µl of supernatant or standard was mixed with 50 µl of 80 mM sulfanilamide in 2 N HCl and 50 µl of 4 mM N-1(-naphthyl)ethylenediamine dihydrochloride in water, and absorbance was determined at 550 nm. The assay was sensitive to 2 µM nitrite and nitrate. IL-8 levels were determined by ELISA as described previously (32). The assay was sensitive to 20 pg/ml IL-8.
Analysis of epithelial iNOS expression by RT-PCR and immunoblots
Total RNA was extracted using an acid guanidinium-phenol-chloroform method (Trizol, Life Technologies/BRL, Gaithersburg, MD). RT-PCR amplification was performed using previously described protocols (33). Primers for human iNOS amplification were 5'-GGT GCT GTA TTT CCT TAC GAG GCG AAG AAG G-3' (sense) and 5'-GGT GCT ACT TGT TAG GAG GTC AAG TAA AGG GC-3' (antisense). Primers for amplification of human ß-actin mRNA were previously described (33). Annealing and extension were conducted at 60°C for iNOS and 72°C for ß-actin. Amplification for 32 cycles yielded a 258-bp fragment for iNOS and a 661-bp fragment for ß-actin. Protein levels of iNOS and actin were determined by immunoblot analysis as described previously (22). Briefly, cell lysates (15 µg total protein/lane) were size-fractionated on a denaturing, nonreducing 8% polyacrylamide minigel and electrophoretically transferred to a nitrocellulose membrane (0.1 µm pore size). Specific proteins were detected using optimal concentrations of rabbit anti-human iNOS (COOH terminus; Santa Cruz Biotechnology, Santa Cruz, CA) or rabbit anti-actin (Sigma) as primary Abs, and peroxidase-conjugated donkey anti-rabbit Ig (Amersham, Arlington Heights, IL) as secondary Ab. Specifically bound peroxidase was detected by enhanced chemiluminescence (ECL system, Amersham) and exposure to x-ray film (XAR-5, Eastman Kodak, Rochester, NY) for 3 min.
HPLC analysis of amino acids
Amino acid levels in culture supernatants were analyzed by precolumn derivatization and HPLC using previously described methods (34). Supernatants were lyophilized and derivitized using a mixture of methanol, triethylamine, water, and phenylisothiocyanate. Samples were relyophilized, dissolved in a sodium phosphate/acetonitrile buffer, and analyzed by HPLC. Amino acid levels in the eluate were determined by absorbance at 254 nm and comparison with known amounts of exogenous standards, using internal standardization (methionine sulfone). The method has an independent amino acid detection limit of 50 nM.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
We first determined whether NO affects viability or growth of
G. lamblia trophozoites in vitro. Trophozoites were cultured
in medium that supports optimal viability and growth (35)
and exposed to compounds (NO donors) that spontaneously release NO in
aqueous solution. Exposure of trophozoites to up to 5 mM of the NO
donor GSNO for a relatively short time period (90 min) had no effect on
their viability (data not shown), suggesting that NO is not cytotoxic
for G. lamblia trophozoites under these conditions. We
therefore tested whether NO donors had long term effects on
trophozoites by assaying growth over several cell divisions. Under
control conditions, the number of trophozoites increased approximately
exponentially over a 42-h period, with a doubling time of about 8
h (Fig. 1
A). Addition of GSNO
inhibited trophozoite growth, but did not increase the number of dead
trophozoites (Fig. 1B). The IC50 for
growth inhibition by GSNO was 
0.7 mM, and growth was almost
completely inhibited at 2 mM. Two control compounds, GSSG and GSH,
which are structurally related to GSNO, but do not release NO, had no
effect on trophozoite growth (Fig. 1B). Furthermore,
incubation of trophozoites with up to 2 mM nitrate, which generally is
the most abundant stable end product of NO, had no effect on growth
(data not shown). Addition of another relatively stable NO end product,
nitrite, was moderately inhibitory for trophozoite growth in a
dose-dependent manner (e.g., addition of 0.1 and 1 mM nitrite inhibited
growth by 25 and 40%, respectively, over a 24-h culture period).
However, nitrite-induced growth inhibition was lower than that induced
by identical concentrations of GSNO, and only a small fraction of NO
derived from GSNO was converted to nitrite under the culture conditions
(e.g., 0.3 mM nitrite was detected by the Griess reaction in
Giardia culture medium containing 2 mM GSNO after incubation
for 42 h without trophozoites). Moreover, GSNO is relatively
stable in aqueous solution, so that only a fraction of GSNO is
converted to NO over any given period of time, whereas added nitrite is
present at its initial concentration throughout the entire culture
period. Thus, nitrite does not appear to be primarily responsible for
mediating the growth inhibitory effects of NO, although it probably
contributes to the inhibition. Taken together, these data indicate that
NO, or possibly one or several of its more stable metabolites, is
cytostatic, but not cytotoxic, for G. lamblia trophozoites
under the conditions of these studies.
|
The number of G. lamblia trophozoites in the small
intestine is determined in part by the balance between growth and
encystation. The latter is initiated when trophozoites are exposed to a
slightly alkaline pH and bile (27, 36). To further explore
the role NO can play in influencing trophozoite numbers in the small
intestine, we determined whether NO affects G. lamblia
encystation. For these experiments, trophozoites were exposed for
18 h to conditions that efficiently induce encystation in vitro
(27, 36), and the number of ESVs was determined
microscopically. Formation of ESVs is a specific and relatively early
event in the G. lamblia encystation pathway
(37). In nonencysting trophozoite populations, <5% of
cells contain ESVs (37), while after 18 h under
encystation conditions >80% of cells contain one or more ESV, with
some cells containing >20 ESVs (Fig. 2A). Addition of GSNO to
encysting G. lamblia cultures completely inhibited the
development of cells with 
8 ESVs/cell and decreased by >65% the
average number of ESVs per positive cell, but had little effect on the
percentage of cells containing at least one ESV (Fig. 2). The
ED50 for the decrease in the number of ESVs per
cell by GNSO was about 0.6 mM, and maximal effects were seen at 2 mM
(Fig. 2C). The number of ESVs per cell also decreased after
addition of a different NO donor, spermine NONOate, to encysting
cultures (data not shown). The control compounds, GSSG and GSH (for
GSNO) and spermine (for spermine NONOate), had little or no effect on
the number of ESVs per cell (Fig. 2 and data not shown).
|
A). GSNO inhibited CWP2
expression in encysting G. lamblia cultures in a
dose-dependent manner, while the control compounds GSSG and GSH had
little or no effect on CWP2 levels (Fig. 3). Expression of a
nonregulated giardial control protein, taglin, was not affected by
addition of 1 mM GSNO, GSSG, or GSH to encysting cultures (data not
shown). Thus, together with the ESV counts, these data show that NO
inhibits G. lamblia encystation.
|
Excystation of G. lamblia is required to establish
infection in the small intestine and cause disease. This developmental
step is controlled by sequential exposure to conditions that ingested
cysts encounter during passage through the stomach into the duodenum,
i.e., acid pH followed by alkaline pH and active proteases
(28). Other host factors, such as specific Abs, can affect
the excystation program and thereby the establishment of infection
(27). We therefore determined whether NO, which can be
produced by epithelial cells in the stomach and small intestine
(23, 24, 39), affects excystation of G.
lamblia. Cysts were initially kept under pre-excystation
conditions (distilled water, designated pre-stage), which lyses and
kills all Giardia with incomplete cyst walls, followed by
incubation under acidic conditions (stage I), and then alkaline
conditions in the presence of trypsin (stage II), after which the
number of motile excysted trophozoites was determined microscopically.
As shown in Table I, GSNO at 5 mM
inhibited excystation of G. lamblia when present during all
three excystation stages, while the control compounds GSSG and GSH had
no effect on excystation. A lower dose of GSNO (2 mM) did not inhibit
excystation, indicating that a 3- to 5-fold higher dose of GSNO was
required for inhibition of excystation compared with inhibition of
growth and encystation (Figs. 1
and 2). This may be related to the
thickness or relative impermeability of the cyst wall compared with the
cell membrane of the vegetative stage, rather than a lower NO
sensitivity of the intracellular signaling events involved in
excystation compared with those in encystation or during vegetative
growth. Another NO donor, spermine NONOate, also inhibited excystation,
while its control compound, spermine, did not (Table I).
|
, incubation with GSNO during excystation stage
II alone was sufficient to inhibit excystation. Furthermore, GSNO had
to be present during stage II to inhibit excystation, while incubation
with GSNO during the pre-excystation stage and/or stage I did not
inhibit excystation. Thus, the presence of an NO donor during stage II
of excystation (modeling the small intestine) was necessary and
sufficient for inhibiting G. lamblia excystation. Infection of human intestinal epithelial cell cultures with G. lamblia inhibits epithelial NO production
The studies described above showed that NO donors can affect growth and differentiation of G. lamblia, suggesting that NO production may be a potential host effector mechanism. Based on this, we hypothesized that G. lamblia may have the ability to modulate this host defense, since a number of pathogens have evolved strategies to evade or take advantage of specific host responses. Intestinal epithelial cells are likely to be the most abundant producers of NO at levels relevant for the parasite, which resides in close contact to these cells in the small intestine (5). Therefore, we determined whether G. lamblia can modulate NO production by intestinal epithelial cells.
Monolayers of the human intestinal epithelial cell line, HT-29, were
infected with varying numbers of G. lamblia trophozoites,
and after 18 h, levels of the stable NO end products nitrite and
nitrate were determined in the culture supernatants. Constitutive NO
production was not detectable in these cultures before or after
G. lamblia infection (Fig. 4A). However, G.
lamblia infection inhibited NO production by >95% in HT-29
cultures stimulated with a combination of cytokines (IL-1, TNF-,
and IFN-) compared with that in cytokine-stimulated cultures that
were not infected (Fig. 4A). Addition of as few as 1
trophozoite/100 epithelial cells caused half-maximal inhibition of
epithelial NO production, and inhibition was maximal when 3
trophozoites/100 epithelial cells were added to the HT-29 cultures.
Furthermore, inoculation of cytokine-stimulated HT-29 cultures with
G. lamblia completely inhibited NO production throughout the
entire culture period (24 h), whereas in uninfected HT-29 cultures NO
production began to increase within 10–12 h after cytokine stimulation
and continued to increase thereafter.
|
B). These findings, together with the observation
that G. lamblia infection did not affect the number of
viable epithelial cells in culture (Fig. 4C), ruled out the
possibility that G. lamblia had cytotoxic effects on the
epithelial cell cultures. Furthermore, we found in additional control
experiments that cultures of G. lamblia alone in the absence
of epithelial cells did not consume exogenously added NO metabolites,
nitrite or nitrate, or interfere with the detection of these products.
G. lamblia cultures also did not produce any NO metabolites
detectable using nitrate reductase and the Griess reaction. Together,
these data suggest that G. lamblia interfered with
epithelial NO production rather than consumed epithelial cell-derived
NO or its metabolites.
To extend the findings obtained in HT-29 cells, cultures of two
additional human intestinal epithelial cell lines, Caco-2 and HCT-8,
were infected with G. lamblia. Cytokine-inducible NO
production by these cells was also significantly inhibited by G.
lamblia infection in an inoculum-dependent manner (644 ± 116
µM NOx in uninfected Caco-2 controls vs
352 ± 86 and 200 ± 67 µM after addition of 3 x
106 and 3 x 107
trophozoites, respectively; 49 ± 1 µM NOx
in uninfected HCT-8 controls vs 13 ± 2 and <2 µM after
addition of 105 and 106
trophozoites, respectively; mean ± SEM; n =
3–4). Caco-2 cells produced 10-fold greater levels of NO after
cytokine stimulation than HT-29 and HCT-8 cells under these conditions,
and a higher trophozoite inoculum was required for inhibition of NO
production by Caco-2 cells. Nonetheless, G. lamblia
infection also inhibited constitutive NO production by Caco-2 cells,
which, in contrast to HT-29 or HCT-8 cells, was high enough to be
readily detectable (82 ± 6 µM NOx in
uninfected controls vs 46 ± 8 and 31 ± 2 µM after
addition of 3 x 106 and 3 x
107 trophozoites, respectively). The latter
finding in conjunction with the lack of G. lamblia effects
on inducible IL-8 secretion by the epithelial cells (Fig. 4B) strongly suggest that G. lamblia infection
did not interfere with the stimulation/signaling of epithelial cells by
cytokines, but, rather, affected the ability of the cells to
produce NO regardless of the mode and level of stimulation. As for
HT-29 cells, cell number and viability of Caco-2 cells were not
affected by the infection, because total protein content (as determined
by Bradford assays of cell lysates) and the ability to reduce the
tetrazolium dye 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium
bromide (a common measure of metabolic function) were not different in
controls and G. lamblia-infected Caco-2 cultures (data not
shown).
Apical G. lamblia infection of polarized intestinal epithelial cell monolayers inhibits both apical and basolateral epithelial NO release
G. lamblia is a strictly lumen-dwelling pathogen with
exclusive access to the apical side of the physiologically polarized
intestinal epithelium. Therefore, to further determine the physiologic
relevance of the G. lamblia-induced inhibition of epithelial
NO production, we used polarized monolayers of Caco-2 cells grown on
collagen-coated microporous filter supports. The polarized epithelial
monolayers were infected with G. lamblia on the apical side,
whereas cytokine stimulation was from the basolateral side, because the
cytokines, IL-1, TNF-, and IFN-, used for these studies are
physiologically produced by cells, such as T cells, NK cells, and
macrophages, that are present in the intestinal lamina propria
underlying the epithelium. As shown in Fig. 5A, apical G.
lamblia infection inhibited total (i.e., apical and basal)
epithelial NO production in response to basolateral cytokine
stimulation by >80%. Constitutive epithelial NO production was also
inhibited by G. lamblia infection, although the difference
did not reach statistical significance. In contrast to NO production,
neither constitutive nor cytokine-induced IL-8 production was
significantly affected by apical G. lamblia infection of
polarized Caco-2 monolayers (Fig. 5B and data not shown).
Furthermore, consistent with a previous report (40),
infection had no effect on the transepithelial resistance of the
polarized monolayers (data not shown).
|
We next investigated giardial requirements for the inhibition of epithelial NO production. Inhibition required viable trophozoites, because addition of heat-killed trophozoites did not affect NO production by cytokine-stimulated HT-29 cultures (110 µM NOx in uninfected controls, <2 µM NOx after addition of 3 x 106/well viable G. lamblia, and 111 µM NOx after addition of 3 x 106/well heat-killed G. lamblia). In contrast, cell-to-cell contact between G. lamblia trophozoites and epithelial cells was not required for inhibition of epithelial NO production. Thus, separation of HT-29 monolayers and trophozoites through a filter membrane (0.2 µm pore size) had no effect on the ability of G. lamblia to inhibit epithelial NO production by cytokine-stimulated HT-29 cells (no filter: 49 µM NOx in uninfected controls, <2 µM NOx after addition of 3 x 106/well trophozoites; with filter interposed: 51 µM NOx in controls, <2 µM NOx after addition of 3 x 106/well trophozoites). These findings indicate that the factor(s) responsible for the G. lamblia-induced inhibition of epithelial NO production is soluble and diffusable, suggesting that viable trophozoites released an inhibitory activity for epithelial NO production and/or consumed a crucial factor or substrate needed for this epithelial function.
G. lamblia infection has no effect on epithelial iNOS expression
NO production in epithelial and other cells is controlled at
several levels, including the availability and transport of arginine
and the levels of NOS, the critical rate-limiting enzyme for the
conversion of arginine to NO. To define the role of the latter in the
G. lamblia-induced inhibition of epithelial NO production,
we focused on the expression of iNOS, because this isoform was
previously shown to be the most abundant NOS isoform in intestinal
epithelial cells (21, 22). Cytokine stimulation of HT-29
cells up-regulated iNOS mRNA and protein expression, as determined by
RT-PCR and immunoblot analysis, respectively, but G. lamblia
infection had no effect on this cellular response (Fig. 6). Thus, G. lamblia
trophozoites did not release an inhibitory activity that interfered
with the potential ability of intestinal epithelial cells to produce NO
(i.e., as reflected in iNOS levels).
|
The finding that G. lamblia did not inhibit epithelial
iNOS expression suggested that infection may reduce the availability or
transport of a factor or substrate needed for NO production. Because
G. lamblia is known to use arginine as an energy source
(41, 42), and arginine is the crucial substrate needed for
NOS to generate NO, we tested the hypothesis that G. lamblia
reduces the availability of arginine in coculture with intestinal
epithelial cells. As a first approach to test the hypothesis, we
determined the levels of arginine in the supernatants from infected and
control cultures by HPLC. As shown in Fig. 7A, G. lamblia
infection of HT-29 cultures reduced arginine levels in the
supernatants, with an inoculum/response curve similar to that for
G. lamblia-induced inhibition of epithelial NO production
(see Fig. 4A). At inocula of 3 trophozoites/100 epithelial
cells, arginine levels were reduced to <2% of those in uninfected
control cultures. This effect became apparent <4 h after infection of
the cultures, indicating rapid and highly efficient arginine uptake by
G. lamblia under the coculture conditions. In contrast,
G. lamblia infection either increased or did not affect the
levels of several other amino acids, including glycine and asparagine
(data not shown). These data indicate that the observed arginine
decrease in G. lamblia-infected intestinal epithelial cell
cultures is specific and not due to a generally increased consumption
of amino acids by G. lamblia, a conclusion consistent with
previous studies on amino acid consumption by G. lamblia
trophozoites cultured in the absence of other cell types and in growth
medium optimal for the parasite (41, 42).
|
2-fold increased arginine levels in the basolateral compartment
compared with the culture medium alone (Fig. 7B), suggesting
that the cells transported arginine from the apical to the basolateral
compartment. Basolateral cytokine stimulation of the monolayers
strongly increased epithelial NO production (Fig. 5A), but
had no effect on apical arginine levels and decreased basolateral
arginine levels only to a small extent (Fig. 7B). The latter
decrease may reflect an increase in arginine uptake required for the
cytokine-induced increase in NO production. When cytokine-stimulated
cultures of polarized Caco-2 monolayers were infected with G.
lamblia on the apical side, arginine levels decreased further in
the apical compartment (to <1% of the level in culture medium) and
were also reduced (by 60–70%) in the basolateral compartment compared
with levels in cytokine-stimulated cultures that were not infected. In
contrast to arginine, glycine was not significantly transported across
control or cytokine-stimulated Caco-2 monolayers, nor were glycine
levels affected by G. lamblia infection (Fig. 7B). Taken together, these findings indicate that G.
lamblia consumed arginine in cultures of both nonpolarized and
polarized intestinal epithelial cells. Moreover, in polarized
epithelial monolayers, G. lamblia appeared to compete with
the epithelial cells for arginine in the apical compartment, as
suggested by the decreased amount of arginine that was transported by
the epithelial cells from the apical to the basolateral
compartment. Arginine addition reverses G. lamblia-induced inhibition of epithelial NO production
To further test the hypothesis that G. lamblia inhibits
epithelial NO production by reducing the availability of arginine, we
next established the concentration/response relationship for the
arginine dependence of epithelial NO production in uninfected cultures.
Confluent HT-29 monolayers were cultured in arginine-free medium,
supplemented with 10% dialyzed FCS, to which titrated amounts of
arginine were added. Cells were stimulated for 18 h with a
combination of cytokines (IL-1, TNF-, and IFN-), and NO
production was determined. Maximal NO production was observed when
300 µM arginine was added to the cultures (97 ± 13 µM
NOx at 300 µM arginine and 75 ± 9 µM
NOx at 10 mM arginine; n = 4),
while NO production was not detectable (<3 µM
NOx) after addition of 
100 µM arginine. In
contrast to NO production, IL-8 secretion was little affected by the
lack of arginine in the cultures (376 ± 27 ng/ml IL-8 without
arginine; 524 ± 64 ng/ml IL-8 with 10 mM arginine;
n = 3). These data show that NO production by
uninfected HT-29 cultures is critically dependent on the availability
of exogenous arginine, and that the concentration/response relationship
for this effect is consistent with the observed decrease in arginine
levels after G. lamblia infection (Fig. 7A) and
the concomitant decrease in NO production (Fig. 4A).
To further define the role of arginine in the G.
lamblia-induced effects on epithelial NO production, we added
increasing concentrations of exogenous arginine to G.
lamblia-infected HT-29 cultures (which were set up in regular,
arginine-containing medium). As shown in Fig. 8A, addition of increasing
concentrations of arginine reversed the G. lamblia-induced
inhibition of epithelial NO production. Thus, after addition of 1 mM or
more arginine to G. lamblia-infected cultures, epithelial NO
production was significantly higher than that in infected cultures not
supplemented with arginine. Moreover, at the two highest concentrations
of arginine added (3 and 10 mM), NO production in G.
lamblia-infected cultures was not significantly different from
that in uninfected control cultures. This effect was specific for NO
production, because arginine addition to G. lamblia-infected
HT-29 cultures did not significantly affect IL-8 secretion (Fig. 8B). Furthermore, the effect was specific for arginine,
because addition of up to 10 mM serine, which is also consumed by
G. lamblia (42), had no effect on the
inhibition of epithelial NO production by G. lamblia (data
not shown). Addition of up to 10 mM arginine or serine to uninfected
control cultures had no effect on epithelial NO or IL-8 production
(Fig. 8 and data not shown). Taken together, these data show that
arginine addition specifically and completely reversed the G.
lamblia-induced inhibition of epithelial NO production.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
NO inhibited both growth and encystation of G. lamblia trophozoites as well as excystation of cysts. Inhibitory effects on growth and excystation would be expected to reduce the number of trophozoites in the intestinal lumen, while inhibition of encystation might have the opposite effect, i.e., increased numbers of trophozoites in the lumen. The net outcome of these opposing effects is difficult to predict. However, growth inhibition may be important for the infected host, because local trophozoite growth is probably crucial for the ability of G. lamblia to establish and maintain infection of the proximal small bowel, whereas inhibition of encystation by NO could reduce the formation and passing of infectious cysts and, thereby, transmission to other potential hosts.
We found NO to be cytostatic, but not cytotoxic, for G. lamblia trophozoites, which is in contrast to a recent report that NO donors appeared to kill trophozoites in vitro (45). In that study the effects of NO donors were assessed in a minimal medium (buffered salt solution), which may have exaggerated the results, because such media do not support prolonged viability and growth of trophozoites even in the absence of NO donors. To avoid such complications, the present studies used complex medium to mimic in vivo conditions that allow long term survival and growth of G. lamblia trophozoites (35). Although it is clear that NO is cytostatic for trophozoites in complex medium, it is not known how NO exerts this effect. We have observed recently that trophozoite cultures exposed to NO donors appear to accumulate cells that are undergoing cell division (data not shown), which suggests that NO may affect the regulation of the giardial cell cycle, although little is currently known about mechanisms that govern cell cycle control in Giardia.
Our finding that G. lamblia and intestinal epithelial cells compete for arginine has several implications for understanding the interaction between pathogen and host. From the host perspective, reduced arginine availability could be considered a virulence mechanism of the pathogen, because it inhibited epithelial NO production, thereby subverting a potential host defense against G. lamblia. Both intestinal epithelial cells and G. lamblia trophozoites have highly efficient arginine transporter systems that have comparable substrate affinities, although the giardial arginine transport system has a 10- to 20-fold higher maximal transport capacity (46, 47, 48, 49), suggesting that Giardia may have an advantage over the host in taking up arginine. In addition, luminal arginine derived from food may encounter G. lamblia before it can be taken up by the epithelium, thus providing a further advantage for the pathogen over the host epithelium, although little is known about arginine levels in the intestinal lumen, extracellular space, or serum during giardial infection. However, a large fraction of arginine in food is in the form of peptides that need to be digested by peptidases before releasing free arginine. Many of the relevant peptidases are located on the apical membrane of intestinal epithelial cells, which may give the host epithelium an advantage over the trophozoites in taking up arginine. Furthermore, serum and, presumably, extracellular fluid contain moderate levels of free arginine (0.1–0.2 mM), and the intestinal epithelium is somewhat permeable for amino acids, which facilitates their diffusion from the serum or extracellular space into the intestinal lumen (50).
Giardial consumption of the substrate arginine required for epithelial NO production is clearly an important mechanism for G. lamblia-induced inhibition of epithelial NO production, as indicated by time course and inoculum/response curves for inhibition of NO production compared with those for arginine levels in the supernatants. However, the data do not exclude the possibility that trophozoites, in addition, may release an inhibitory activity for epithelial NO production. Comparison of concentration/response curves for the arginine dependence of NO production by HT-29 cells using arginine-free medium and regular medium previously conditioned by G. lamblia suggests that the parasite releases such an inhibitory activity, because higher levels of arginine are required for G. lamblia-conditioned medium compared with arginine-free medium to achieve the same levels of epithelial NO production (data not shown). One possibility for a G. lamblia-derived inhibitory activity for epithelial NO production would be the amino acid ornithine, which is released by G. lamblia in exchange for arginine (41, 42) and has been shown to competitively inhibit arginine uptake by epithelial cells (48). Based on this idea, G. lamblia could inhibit epithelial NO production by two separate, mutually reinforcing means: consumption of arginine and release of a competitor such as ornithine that inhibits arginine uptake by the intestinal epithelium.
G. lamblia inhibited NO production in polarized intestinal epithelial cell monolayers despite the fact that trophozoites resided exclusively on the apical side of the monolayers, and arginine depletion was mostly apparent on that side, whereas arginine levels on the basolateral side were only slightly lower than those in culture medium alone. These data suggest that polarized intestinal epithelial cells are largely dependent on apical arginine availability for NO production and do not substantially use arginine present on the basolateral side. Several lines of evidence support this interpretation. NO production in most cells depends completely on extracellular arginine despite the presence of sufficient arginine inside the cells (a phenomenon referred to as the arginine paradox) (51, 52). Efficient transport of extracellular arginine into cells is facilitated by several transporters, including those of the cationic amino acid transporter family of transporters, which are found in close association with some NOS isoforms and are thought to form a functional complex close to the membrane (51). In intestinal epithelial cells, iNOS is largely confined to the apical side of the cell underneath the cell membrane (23, 53), suggesting that functional cationic amino acid transporter/iNOS complexes are also limited to that side of the cell, although such complexes remain to be demonstrated in epithelial cells.
Intestinal epithelial cells are likely to be the most relevant NO-producing cells in the host for affecting giardial growth and differentiation due to the close association between the cells and G. lamblia. Cells immediately underlying the epithelium, e.g., subepithelial fibroblasts or macrophages, conceivably could also contribute a fraction of NO effective against Giardia via diffusion through the epithelial layer. In either case, expression of sufficiently high levels of one or several NOS isoforms by the relevant host cells is a necessary precondition for the release of effective amounts of NO to contribute to the host defense against G. lamblia. Intestinal epithelial cells in the distal small bowel in mice constitutively express iNOS mRNA (23), and isolated human duodenocytes produce NO constitutively (24). It is possible that epithelial iNOS expression may be up-regulated in vivo during the course of G. lamblia infection, although this is not known. Furthermore, inflammation of the proximal small bowel, which is likely to be accompanied by increased epithelial iNOS expression (25), is common in developing countries where coinfections with multiple enteric pathogens are prevalent (54). Under these conditions, increased epithelial iNOS expression caused by one pathogen may help to control infection with another pathogen, such as G. lamblia. This concept is illustrated, for example, by the finding that infection with Trichinella spiralis, which causes mucosal inflammation and increased iNOS mRNA expression in the proximal small bowel (55), interferes with subsequent Giardia infection (56).
Giardia resides strictly in the lumen of the intestinal tract. For effective control and removal of the parasite, the host has to rely on defense mechanisms that are active in the intestinal lumen. One such mechanism could be NO produced by the intestinal epithelium, as suggested by the present studies. Any role of NO in antigiardial defense is likely to be a complement to other luminally active defense mechanisms. For example, Abs are important for controlling Giardia infection because B cell-deficient mice cannot clear infections with Giardia, and secretory Abs of the IgA isotype are generally thought to be crucial for this function (9, 12, 13). However, the exact role of IgA in controlling G. lamblia infection is not clear, since patients with IgA deficiency have only a slight increase in the incidence of G. lamblia infection (14), and Xid mice cannot clear a primary G. muris infection despite high levels of antigiardial IgA Abs in mucosal secretions (10). This suggests that several levels of host defenses exist against Giardia, and that other host defense mechanisms may be able to compensate for the lack of secretory IgA. Furthermore, components of the natural immune system, such as defensins and lactoferrin, may contribute to host defense against Giardia by providing a base level protection against this pathogen (15, 16). Epithelial production of NO could also belong to this category of natural immune defenses. The balance between giardial arginine consumption and epithelial NO production could contribute to the variability in the duration and severity of infections by this ubiquitous parasite.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Lars Eckmann, Department of Medicine 0623D, University of California at San Diego, 9500 Gilman Drive, La Jolla, CA 92093-0623. E-mail address:
3 Current address: Station de Pathologie Aviaire et de Parasitologie, Institut National de la Recherche Agronomique Centre de Recherches de Tours, Nouzilly, France.
4 Abbreviations used in this paper: NOS, NO synthase; iNOS, inducible NOS; GSNO, S-nitrosoglutathione; GSSG, oxidized glutathione; spermine NONOate, 1-N-[3-aminopropyl]-N-[4-(3-aminopropylammonio)butyl]-amino-diazen-1-ium-1,2-diolate; CWP2, cyst wall protein 2; ESV, encystation-specific secretory vesicle.
Received for publication August 9, 1999. Accepted for publication November 22, 1999.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
) or the presence of GSNO at 1 mM (
) or 2
mM (•). After 18 and 42 h, the number of viable trophozoites was
determined. B, Concentration dependence of NO-induced
growth inhibition. Trophozoites (104/ml) in complex growth
medium were grown for 42 h in the absence or the presence of the
indicated concentrations of GSNO (•, 
), GSSG (
), or GSH (
).
Subsequently, the numbers of viable (•, 
).
After 18 h, the number of ESVs was counted microscopically for at
least 50 cells. A, Frequency distribution of the number
of ESVs per cell in controls or after addition of 2 mM of GSNO, GSSG,
or GSH in a single experiment. The average numbers of ESVs per cell in
this experiment were 5.56 (controls), 1.57 (GSNO), 4.12 (GSSG), and
4.62 (GSH). B, Concentration/response curves for the
percentage of cells with at least one ESV per cell. C,
Concentration/response curves for the number of ESVs per cell for all
ESV-positive cells (i.e., those cells that contain at least one ESV).
Results are from a representative experiment. An additional experiment
showed similar findings.
