INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Schistosoma mansoni, the causative agent for schistosomiasis, is an obligate intravascular helminth parasite. Ultrastructural examinations of experimentally infected mice show that during its migration within its definitive mammal hosts, S. mansoni physically interacts with different types of blood vasculature endothelium from several tissues and organs including skin, heart, lung, liver, brain, kidney, intestines, and mesenteries (46, 47). For instance, a few days after percutaneous penetration, the larval-stage schistosomula reach the lungs, where the parasites are in intimate contact with pulmonary capillaries for several days (6, 46). Considering the pivotal function of microvascular endothelial cells (ECs) in inflammation (40), this initial contact may have an important role in the early immunological events following parasite penetration into its vertebrate hosts. Moreover, in the -irradiated vaccine model of murine schistosomiasis, it is believed that pulmonary capillary ECs play an important role in the elimination of challenged parasites by participating in the recruitment of immune cells to the lungs (4, 5, 7). In the same manner, parasite-EC interactions may be important in this mechanism.

In its quiescent state, the capillary endothelium acts as a selective permeability barrier for molecules and circulating cells regulated in large part by intercellular junctions. These are complex structures formed by transmembrane adhesive molecules linked to network of cytoplasmic or cytoskeletal proteins (10). The mechanisms that regulate the opening or the closure of endothelial junctions involve intracellular signals which cause cytoskeletal reorganization involving actin microfilaments (22). In certain pathological conditions, such as inflammation (due to a pathogen, for instance), activation of ECs results in morphologic changes in cell shape (retraction), leading to the opening of tight junctions and/or interendothelial gaps and to an increase of permeability to molecules and cells (13, 15, 19, 40). EC activation by inflammatory agonists also results in the synthesis of an array of cytokines and chemokines and in the expression of adhesion molecules which permit the trafficking of leukocytes to sites of inflammation (3, 35).

The aim of the present report was to investigate whether lung schistosomula could bind to and/or activate resting ECs by modifying the endothelial phenotype in vitro. To this end, the permeability of monolayers of bovine brain capillaries ECs (BBCEC) to a small molecule was measured in response to parasites. Surprisingly, we found that the excretory/secretory (ES) products from schistosomula generate intracellular signals to ECs, leading to a dramatic decrease in transvascular permeability. This endothelial barrier-enhancing effect appeared to be mediated by a cyclic AMP (cAMP)/protein kinase A (PKA) pathway.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Reagents. All reagents were purchased from Sigma (St Quentin-Fallavier, France) unless otherwise indicated. [³H]inulin (3.2 Ci/mmol) and sodium [³²P]orthophosphate (200 mCi/mmol) were purchased from Amersham (Les Ulis, France), and the kinase inhibitors Rp-cAMP, calphostin C, and tyrphostin AG 126 were purchased from Calbiochem (La Jolla, Calif.).

Cell cultures. Mouse microvascular lung endothelial cells (MLE) were grown as described previously (45) in flat-bottomed culture plates (Nunc, Roskilde, Denmark) precoated with 0.2% gelatin in 1:1 (vol/vol) Ham's F12 medium-Dulbecco's modified Eagle's medium (DMEM) supplemented with 5% (vol/vol) heat-inactivated fetal calf serum (FCS), penicillin (100 U/ml), and streptomycin (100 µg/ml) (Gibco, Grand Island, N.Y.). MLE were used between passages 20 and 30. The fibroblast (3T3) and epithelial (HeLa) cell lines were obtained from the American Type Culture Collection (Rockville, Md.). BBCEC were isolated, cloned, and cultured in DMEM supplemented with 10% heat-inactivated calf serum, 10% horse serum (Hyclone, Logan, Utah), 1 ng of basic fibroblast growth factor per ml, 2 mM glutamine, and 50 µg of gentamicin per ml as reported previously (27). Primary cultures of astrocytes were prepared from newborn rat cerebral cortex and cultured as described previously (2). BBCEC were grown to confluent monolayers on inserts in the presence of rat astrocytes (cultured in the lower chamber) (8). Briefly, after two to four passages, BBCEC (10⁵ cells) were seeded on rat tail collagen-coated cell culture inserts (Millicell-CM; pore-size, 0.4 µm; diameter, 30 mm; Millipore, Bedford, Mass.) placed into the astrocyte-containing wells, and cultured for 10 days at 37°C. The medium was renewed to both the luminal and abluminal chambers every 2 days. To eliminate possible interferences between astrocytes and parasites, the inserts were transferred into astrocyte-free wells 2 days prior to stimulation. Both cell types were maintained at 37°C in a humidified atmosphere supplemented with 5% CO₂.

Parasites. In order to have sufficient amounts of parasites, we developed a technique of culture to obtain in vitro lung-transformed schistosomula. Briefly, schistosomula were obtained by the skin penetration procedure (32) from cercariae (Puerto Rican strain) shed from infected Biomphalaria glabrata snails and resuspended in conditioned BBCEC culture medium for 6 days (a period corresponding to the period over which parasites reach the lungs and undergo the adaptive changes necessary for intravascular migration). In vitro lung-transformed schistosomula were successfully tested for the ability to mature in vivo by their surgical transfer to the superior mesenteric vein of naive mice. Glutaraldehyde-fixed parasites were prepared by incubation of live in vitro lung-transformed schistosomula with sterile PBS containing 0.5% (vol/vol) glutaraldehyde for 5 min at 20°C and extensive washing with culture medium.

Collection of ES products from schistosomula. ES products from in vitro lung-transformed schistosomula were obtained by incubating 2.5 × 10³ to 2 × 10⁴ parasites/ml in BBCEC culture medium or in DMEM supplemented or not with 1% FCS at 37°C for 4 h. The culture supernatant was then carefully removed, centrifuged at 2,000 × g for 5 min to eliminate residual parasites, sterilized by filtration, and frozen at 70°C until use. Size fractionation of the schistosomula ES products was performed by successive filtrations on Microsep microconcentrators (molecular size cutoffs of 1, 3, 10, and 100 kDa) as detailed by the manufacturer (Filtron, Northborough, Mass.), thus producing fractions of <1, 1 to 3, 3 to 10, 10 to 100, and >100 kDa.

Cocultures with S. mansoni schistosomula and attachment assay. MLE were cultured to confluence in 24-well gelatin-coated plates (Nunc). Aliquots of either living or glutaraldehyde-fixed schistosomula (200 parasites/10-µl aliquot) were added to each well containing 250 µl of complete culture medium and incubated at 37°C for 4, 12, or 18 h. After incubation, the wells were thoroughly washed four times with phosphate-buffered saline (PBS) containing 1 mM CaCl₂ and 1 mM MgCl₂ to remove unattached parasites, and the remaining adherent parasites were counted by light microscopy. In some cases, cells were fixed for 30 min with glutaraldehyde (0.1% [vol/vol] in PBS) or treated for 30 min with cytochalasin D (0.5 µM), nocodazole (10 µM), or colchicine (1 µM). Cultures were then extensively washed with culture medium prior to attachment assays. Results are expressed as the percentage of the total added schistosomula attached to each well.

Measurement of paracellular permeability. Confluent BBCEC monolayers were incubated with schistosomula (1 parasite/10⁴ cells) for 2, 4, or 8 h; then transendothelial transport measurements were performed as reported previously (9). In experiments in which schistosomula ES products were tested as activators, 150 µl of a 4-h parasite culture supernatant (10⁴ parasites/ml of BBCEC culture medium) was added to the well (final volume, 1.5 ml). After stimulation, inserts were transferred to six-well plates containing 2 ml of prewarmed Ringer-HEPES buffer in each well (abluminal compartment). In the luminal chambers, culture medium was replaced by 2 ml of Ringer-HEPES supplemented with 1.25 µCi of [³H]inulin (5 kDa). At 10, 20, and 30 min thereafter, inserts were placed into the next well. The amounts of labeled inulin crossing the monolayers were measured by scintillation counting of 500-µl aliquots of the medium from the lower compartment. For control, flux across cell-free, collagen-coated inserts was also measured. To calculate the permeability of the monolayer, the clearance principle was used (9). Except for Fig. 2, results are expressed as the endothelial permeability coefficient (P_e, in centimeters per minute).

Determination of intracellular cAMP concentrations. BBCEC were grown to confluence in 35-mm-diameter dishes (approximately 3 × 10⁵ cells), and medium of the monolayers was renewed with 1.5 ml of preheated DMEM supplemented with 1% FCS 30 min before stimulation. Cell monolayers were then exposed to 100 µM isobutylmethylxanthine (IBMX; an inhibitor of cAMP phosphodiesterase) 10 min prior to the addition of potential stimulators. Treatment was accomplished by adding schistosomula ES products (150 µl of a 4 h-culture supernatant; 2.5 × 10³ to 2 × 10⁴ parasites/ml of DMEM-1% FCS) or forskolin (an activator of adenylate cyclase; final concentration, 25 µM) directly to the culture and incubating for various periods of time. After activation, the cells were washed with cold PBS containing 100 µM IBMX and homogenized in ice-cold PBS containing 10% (vol/vol) trichloroacetic acid. After sonication and centrifugation at 2,000 × g for 10 min, the trichloroacetic acid-soluble supernatant was extracted four times with water-saturated diethylether. The concentration of intracellular cAMP was determined in triplicate by the enzyme immunoassay kit provided by Cayman Chemical (Ann Arbor, Mich.).

Treatment of BBCEC with kinase inhibitors and permeability studies. Confluent BBCEC were pretreated for 15 min with the different kinase inhibitors indicated in Fig. 5 or vehicle alone placed in the luminal side of the monolayers before stimulation with forskolin or schistosomula ES products. Ten minutes later, permeability measurements were carried out. For each inhibitor, we initially defined an optimal concentration that does not significantly modify the barrier permeability of unstimulated BBCEC (not shown).

Determination of MLCK phosphorylation. Determination of myosin light-chain kinase (MLCK) phosphorylation in BBCEC was carried out by immunoprecipitating lysates from ³²P-labeled cells with a mouse immunoglobulin G2b anti-MLCK monoclonal antibody (clone K36) exactly as described previously (17, 44). Immunoprecipitates were separated on a 7% polyacrylamide gel under denaturating conditions (21) and transferred to nitrocellulose membranes. The relative intensities of the labeled immunoprecipitated MLCK were quantified by scanning densitometry. The position of EC MLCK was verified by staining the membranes with anti-MLCK antibody.

Statistical analysis. Statistical analysis was performed by using Student's t test. P < 0.05 was considered statistically significant.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

S. mansoni schistosomula attach to cultured ECs. We initially assayed the ability of in vitro lung-transformed S. mansoni schistosomula to attach to MLE monolayers. As shown in Fig. 1A, fixed (nonmotile) schistosomula attached to MLE monolayers more efficiently than live parasites. Over a 4-h period, 13% ± 2% (mean ± standard deviation [SD]) of fixed parasites attached to MLE, reaching a plateau of 28% ± 4% after 18 h of incubation, versus 4% ± 2% and 15% ± 3% for live schistosomula. It is noteworthy that similar adherence values were obtained with in vivo-derived lung schistosomula (not shown). To determine whether schistosomula specifically bind to ECs, parasites were incubated with irrelevant control cells. As shown in Fig. 1B, live schistosomula attached less efficiently to cultured epithelioid (HeLa) cells, fibroblasts (3T3 cells), and cell-free wells (gelatin) than to MLE. To investigate whether parasite binding to MLE was an active or passive process, schistosomula were incubated with glutaraldehyde-fixed MLE. As shown in Fig. 1C, live parasites attached minimally to wells containing glutaraldehyde-fixed cells. In the same manner, schistosomula attachment to MLE was dramatically reduced when cells were pretreated for 30 min with cytochalasin D, an inhibitor of actin polymerization (80% ± 5% inhibition compared to untreated cells). In contrast, the microtubule-disrupting agents colchicine and nocodazole did not significantly impair parasite attachment to MLE (Fig. 1C). Taken together, these results suggest that S. mansoni schistosomula can firmly and specifically attach to ECs in vitro and that this attachment is not a passive cellular process since it requires cytoskeletal activity.

View larger version (13K): [in this window] [in a new window]   FIG. 1.   Attachment of S. mansoni schistosomula to ECs. (A) Live and glutaraldehyde-fixed schistosomula attach to monolayers of MLE. Schistosomula (200/well) were incubated with confluent cultures of MLE for various periods; then cultured wells were washed, and the remaining adherent parasites were counted. *, P < 0.05 compared to live schistosomula. (B) Schistosomula attach more rapidly to MLE than to epithelioid cells (HeLa), fibroblasts (3T3), and cell-free wells (gelatin). *, P < 0.05 compared to HeLa cells. (C) Attachment of schistosomula to MLE requires cell metabolism and actin microfilament function. Confluent MLE were incubated with glutaraldehyde (0.1% in PBS) or with cytochalasin D (0.5 µM), nocodazole (10 µM), and colchicine (2 µM) for 30 min and then exposed with schistosomula for 18 h. *, P < 0.05 compared to unstimulated cells (medium). Values for percent attachment represent arithmetic mean values ± SD (bars) for five separate experiments performed in triplicate.

S. mansoni schistosomula activate ECs in vitro. To further investigate whether interactions of schistosomula with ECs leads to cellular activation, we exploited the barrier to small molecular weight molecules of confluent primary cultures of BBCEC monolayers cocultured with astrocytes. These cells, which have already been demonstrated to be a good model for assessing EC activation in response to different stimuli (8, 9, 13), possess little transcellular vesicular transport and develop tight junctional complexes which limit nonspecific transendothelial and paracellular pathways across the monolayers (8, 33, 37). To validate the endothelial barrier model, the flux of radiolabeled inulin, which penetrates between adjacent ECs paracellularly, was monitored across unstimulated BBCEC monolayers. The obtained values for permeability to inulin (shown in clearance in Fig. 2) confirmed the barrier property of the monolayer (P_e = 0.51 ± 0.07 10³ cm/min). The effect of schistosomula on endothelial permeability was then studied. Interestingly, we found that transmonolayer diffusion of inulin was significantly decreased when cells were incubated with live schistosomula. This effect was time dependent and was maximal (not shown) after 4 to 8 h of coincubation (P_e reduction of 50% ± 6% compared to baseline). This demonstrates that unlike inflammatory agonists, schistosomula activate capillary ECs by reinforcing the endothelial barrier properties.

View larger version (22K): [in this window] [in a new window]   FIG. 2.   Change in transmonolayer permeability of BBCEC to inulin in response to live schistosomula. After 12 days of culture, BBCEC were coincubated with schistosomula (1 parasite/104 cells) for various periods of time. The effects of schistosomula on permeability were studied by using a radiolabeled membrane-impermeant molecule (inulin) as the tracer. Results are expressed as clearance (microliters of tracer diffusing from the luminal to the abluminal compartments) as described elsewhere (9). All values represent means of triplicate inserts ± SD and are representative of five independent experiments. The mean values for percent inhibition are indicated in the text. *, (4 and 8 h), P < 0.05 compared to unstimulated cells (medium).

Endothelial activation does not require parasite-EC contact. To investigate whether interaction of schistosomula with ECs is requisite for cellular activation, schistosomula were physically separated from EC monolayers by a permeable membrane (addition of parasites in the abluminal face), and permeability changes were measured after 4 h of incubation. Compared to unstimulated cells (medium), this caused a decrease in paracellular flux of inulin (expressed as P_e in Fig. 3), as was the case when parasites were added directly to the monolayer (luminal face) (Fig. 3A). In the same manner, adjunction of schistosomula ES products to BBCEC also resulted in a significant decrease in inulin transport through the monolayers (51% ± 4% reduction compared to unstimulated cells). Conversely, incubation of glutaraldehyde-fixed schistosomula with BBCEC (luminal face) did not modify the permeability properties of the monolayer, thus confirming that soluble parasitic factors rather that direct parasite-EC contact are responsible for the generation of intracellular signals leading to permeability modification. To determine an optimal incubation period for EC activation, we performed time course studies with schistosomula ES products. Maximal reduction of BBCEC permeability was achieved with as little as 10 min of exposure of ES products with BBCEC, thereafter reaching a plateau (Fig. 3B). For the rest of the study, BBCEC were activated with schistosomula ES products for 10 min.

View larger version (17K): [in this window] [in a new window]   FIG. 3.   Change in transmonolayer permeability of BBCEC to inulin in response to schistosomula ES products. (A) Schistosomula (1 parasite/104 cells) were applied to the monolayer from the abluminal or luminal side. Schistosomula ES products (150 µl of a 4-h culture supernatant [104 parasites/ml]) or glutaraldehyde-fixed parasites (1 parasite/104 cells) were added to the luminal side (final volume, 1.5 ml). (B) Kinetics of schistosomula ES products on BBCEC permeability. Schistosomula ES products were added to the wells for various periods of time, and the transendothelial passage of inulin was determined. *, P < 0.05 compared to unstimulated cells.

Schistosomula ES products induce an increased concentration of intracellular cAMP in BBCEC. Since cAMP is known to be an important regulator of endothelial barrier properties (11, 23, 28, 39), we inquired whether the cAMP level was increased in BBCEC after exposure to schistosomula ES products. Kinetic experiments indicated a dramatic increase of cAMP in cells stimulated with ES products, with a maximal 2.3-fold increase after 10 min of exposure (Fig. 4A). Similarly, an average threefold increase in the cAMP concentration was observed in cells stimulated for 10 min with forskolin (25 µM), a known adenylate cyclase activator used as a positive control. As shown in Fig. 4B, the cAMP-elevating effect induced by schistosomula ES products to BBCEC was concentration dependent, reaching a plateau at 10⁴ parasites/ml. This experiment indicates that schistosomula ES products contain factors that activate ECs by binding to cell surface receptors that are probably coupled to an adenylate cyclase system.

View larger version (18K): [in this window] [in a new window]   FIG. 4.   Effect of schistosomula ES products on cellular content of cAMP in BBCEC. Cells were incubated in 1.5 ml of DMEM supplemented with 1% FCS and preincubated for 10 min with 100 µM IBMX. (A) Cellular cAMP was assayed after the addition of schistosomula ES products (150 µl of a 4-h culture supernatant [104 parasites/ml of DMEM-1% FCS]) or forskolin (25 µM) (time course, 2.5, 5, 10, and 30 min). The basal intracellular cAMP level is shown (medium). (B) Intracellular concentration of cAMP was determined after 10 min of incubation with various doses of schistosomula ES products (150 µl of a 4-h culture supernatant [2.5 × 103 to 2 × 104 parasites/ml of DMEM-1% FCS]). The experiments shown are representative of three independent experiments and were performed in triplicate. Values represent means ± SD. *, P < 0.05 compared to unstimulated cells.

PKA inhibitor reduces the reinforcement of the barrier function of BBCEC induced by schistosomula ES products. Because elevation of the intracellular cAMP level increases PKA activity in ECs (18), we studied the effects of a specific and potent PKA antagonist (Rp-cAMP) on permeability properties of BBCEC prior to stimulation with schistosomula ES products. Before undertaking this analysis, we defined the optimal concentration of Rp-cAMP to use to counteract the barrier-enhancing effect of the cAMP/PKA agonist forskolin, used as a positive control (not shown). As represented in Fig. 5, the forskolin-induced permeability decrease (expressed as percent reduction of P_e values) of inulin was reversed by Rp-cAMP (maximal inhibition at 50 µM). At this concentration, Rp-cAMP had no significant effect on basal monolayer inulin transport. In the same manner, pretreatment of BBCEC with an identical concentration of Rp-cAMP significantly reduced the barrier-enhancing effect of schistosomula ES products on BBCEC monolayers. Conversely, the use of highly specific protein kinase C (PKC) (calphostin C, 1 µM) and protein tyrosine kinase (PTK) (tyrphostin AG 126, 10 µM) inhibitors did not significantly modify the permeability changes induced by forskolin (not shown) or schistosomula ES products (Fig. 5). Taken together, these results suggest that cAMP represents the crucial second messenger acting on the reinforcement of the barrier function of BBCEC induced by schistosomula ES products and that increased PKA activities in BBCEC probably mediate this effect.

View larger version (22K): [in this window] [in a new window]   FIG. 5.   Effects of kinase inhibitors on endothelial permeability to inulin of BBCEC stimulated with forskolin or schistosomula ES products. Cells were exposed for 15 min to optimal concentrations of Rp-cAMP (50 µM), calphostin C (1 µM), tyrphostin AG 126 (10 µM), or vehicle alone and subsequently treated or not (medium) with forskolin (25 µM) or schistosomula ES products (150 µl of a 4-h culture supernatant [104 parasites/ml]) for 10 min in the continued presence of kinase inhibitors (final volume, 1.5 ml). Results are expressed as percent reduction of Pe values compared to unstimulated cells. *, P < 0.05 compared to cells treated with vehicle alone and then stimulated with forskolin or schistosomula ES products.

Schistosomula ES products increase the level of phosphorylated MLCK in BBCEC. Recently, it was shown that the phosphorylation of MLCK is a key event in the regulation of the endothelial barrier property (17, 44). Since, among other pathways, activation of PKA leads to MLCK phosphorylation, we investigated whether in our model of endothelial culture, schistosomula ES products could time dependently modify the level of phosphorylated MLCK in BBCEC (Fig. 6A). Although endogenous MLCK phosphorylation was observed in resting BBCEC, we found that the extent of MLCK phosphorylation increases in BBCEC stimulated with schistosomula ES products after 5 min of exposure to reach a plateau at 10 min (threefold compared to unstimulated cells). This effect was partially inhibited (65%) in the presence of Rp-cAMP (Fig. 6B), indicating that PKA mediate, at least in part, the phosphorylation of MLCK in BBCEC stimulated with schistosomula ES products.

View larger version (31K): [in this window] [in a new window]   FIG. 6.   Effects of schistosomula ES products on phosphorylation level of the endothelial MLCK. (A) BBCEC (8 × 105 cells/lane) were labeled with 32P, treated for various periods with schistosomula ES products (150 µl of a 4-h culture supernatant [104 parasites/ml of DMEM without FCS]; final volume, 1.5 ml) or not (DMEM) and then subjected to immunoprecipitation under denaturing conditions with the anti-MLCK antibody. (B) BBCEC were exposed for 15 min to optimal concentration of Rp-cAMP (50 µM) or vehicle alone and subsequently treated or not with schistosomula ES products for 10 min in the continued presence of Rp-cAMP. The autoradiograph was analyzed by quantitative laser densitometry. Nonimmune serum, used as a negative control, showed no reactivity. The position of EC MLCK was verified by Western blotting with the anti-MLCK antibody (not shown). The position of the molecular weight marker is indicated.

A low-molecular-weight molecule(s) from the schistosomula ES products is responsible for the decrease permeability of BBCEC. To gain insight into the activating factors present in the schistosomula-released products, we size fractionated the ES products by successive filtrations with various molecular size cutoff concentrators and tested their ability to activate monolayers of BBCEC. As shown in Fig. 7, fractions below 1 kDa, and to a lesser extent between 1 and 3 kDa, led to a decrease in transvascular permeability to inulin, whereas the other fractions were inactive. This finding suggests that a low-molecular-weight substance(s) from the ES products is responsible for the activation of BBCEC.

View larger version (17K): [in this window] [in a new window]   FIG. 7.   Effects of size-fractionated schistosomula ES products on endothelial permeability to inulin of BBCEC. Cells were stimulated with total or size-fractionated (<1-, 1- to 3-, 3- to 10-, 10- to 100-, and >100-kDa) schistosomula ES products (150 µl of a 4-h culture supernatant [104 parasites/ml of DMEM-1% FCS]; final volume, 1.5 ml) for 10 min, and the transendothelial passage of inulin was determined. Results are expressed as percent reduction of Pe values compared to unstimulated cells. *, P < 0.05 compared to unstimulated cells.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

Generally, microorganisms that localize in the vasculature activate and initiate an inflammatory response to the endothelium (14, 25, 36). This leads to the synthesis of cytokines and chemokines and to the expression of cell surface adhesion molecules which permit the diapedesis of immune cells. In parallel, an increase of vascular permeability that may even lead to the disruption of the endothelial barrier and to tissue disorders (for instance, edema formation) is observed. In this study, we show that an intravascular pathogen, i.e., the helminth parasite S. mansoni, may exert an opposite effect on the endothelium by enhancing its paracellular barrier property.

In this study, we were mainly interested in the larval stage of S. mansoni since schistosomula is the first stage to closely interact with the host vasculature endothelium, particularly with the lung capillaries. This initial contact may influence the host immune response but also affect the permeability of the endothelial barrier. In the lungs, schistosomula initiate an inflammatory response, probably by releasing soluble antigens that go through the endothelium and that activate pulmonary antigen-presenting cells to produce inflammatory cytokines (5). This inflammatory state may modify the endothelial barrier properties and favor cell or molecule extravasation. We wondered whether in the vascular compartment, parasite-EC interactions could modulate the endothelial permeability.

To know whether parasites could physically interact with and firmly attach to ECs in vitro, we developed an attachment assay consisting of MLE monolayers and schistosomula. This approach revealed that in vitro (as well as in vivo [not shown])-derived lung schistosomula attached to MLE and to BBCEC (not shown) in a time-dependent manner. Attachment of schistosomula to ECs appears to be an active cellular process since glutaraldehyde-fixed ECs failed to support parasite binding. It is likely that schistosomula stimulate actin polymerization of ECs since cytochalasin D (a microfilament-disrupting agent) abrogated parasite binding to ECs. This observation suggests that at least one endothelial cytoskeleton component, the microfilament, is necessary for parasite binding to ECs. Moreover, we noticed that parasite attachment to ECs was dramatically increased when EC monolayers were pretreated with tumor necrosis factor alpha (TNF-) (not shown), a cytokine known to induce the expression of adhesion molecules on ECs (3). Although we did not address the issue of which molecules might be involved in this binding, inhibition experiments with monoclonal antibodies show that selectin-Lewis^x interactions do not appear to be implicated in these phenomena (not shown), contrary to previous reports suggesting that this might be the case (20, 41).

Whatever the molecules involved in schistosomula-EC binding (if any), we investigated the possibility that ECs may be activated in response to schistosomula. To this end, we used brain capillary ECs grown on porous filters. These cells exhibit a high electrical resistance and a low permeability (8) and have been reported to be a good model for investigation of the endothelial response as a marker of cell activation (8, 12, 13). For instance, inflammatory agonists (such as thrombin, histamine, TNF-, interleukin-1, and interleukin-1) or endotoxin (lipopolysaccharide) increase permeability by generating intracellular signals that cause cytoskeletal reorganization and opening of tight junctions and/or interendothelial gaps (10, 13, 19, 22, 31). Conversely, substances known to elevate the cAMP level inside the cell (such as phosphodiesterase inhibitors) exert a paracellular flux-reducing effect (11, 23, 28, 39). In this study, we investigated the paracellular permeability changes of monolayers of confluent BBCEC to a small molecule (inulin) in response to schistosomula. Interestingly, we found that incubation of in vitro (as well as in vivo [not shown])-derived lung schistosomula with BBCEC resulted in a dramatic decrease of permeability to inulin, a characteristic marker for the generation of an anti-inflammatory phenotype (11). Since noncontact coculture of BBCEC and schistosomula, unlike fixed parasites, also resulted in markedly decreased transmonolayer permeability, it appears that live, intact S. mansoni schistosomula secrete soluble factors that activate BBCEC. In the same manner, ES products were capable of activating ECs in a manner qualitatively similar to that of live parasites. Although ongoing studies are needed, we found that the soluble factor(s) present in the ES products from schistosomula is below 1 kDa in size, suggesting that the most likely candidates are eicosanoids. Indeed, schistosomula have been shown to synthesize a wide array of eicosanoid members, some of which are also found in mammals (1, 34). Data obtained with molecularly purified eicosanoids indicate that some of the parasite eicosanoids may exert an inflammatory reaction in mammalian cells whereas others, such as members of the prostaglandin family, have anti-inflammatory properties. Indeed, some prostaglandins have been shown to increase the endothelial barrier function both in vitro and in vivo via a cAMP-dependent mechanism (24, 26, 30, 43). In our model, it is likely that some similar parasite-derived factor(s) is responsible for the observed effects on ECs.

We then explored the signaling pathway that is induced by the parasites to BBCEC and that leads to the increased barrier function. Consistent with previous reports demonstrating that cAMP analogs or ligands that elevate intracellular cAMP levels reinforce the EC barrier function (11, 16, 23, 24, 26, 28, 39, 43), we found that schistosomula signal ECs by elevating the intracellular cAMP concentration. Moreover, the use of specific kinase inhibitors suggested that the increased impermeability of BBCEC appears to be positively regulated by PKA but not, as already reported, by other kinases such as PKC and PTK (29, 37). Recently, Garcia et al. (17) demonstrated that in a macrovascular endothelial culture model (bovine pulmonary artery ECs), the PKA-mediated phosphorylation of the endothelial MLCK is critical in the reinforcement of the endothelial barrier function. In our model of EC culture, we found that schistosomula ES products increase, in a PKA-dependent manner, the level of phospho-MLCK. Although we have not shown that in BBCEC this event also reduces the kinase activity of MLCK and the subsequent myosin light-chain phosphorylation level (as in bovine pulmonary artery ECs), it is likely that the final products of activating pathway induced by schistosomula ES products also act on cytoskeletal proteins, particularly on the polymerization of F-actin (17, 39).

Taken together, our data indicate that during the parasite's prolonged stay in the lungs, schistosomula could induce an anti-inflammatory phenotype to the endothelium. This finding may have several important consequences on the host response. For instance, the reinforcement of the endothelial barrier properties by schistosomula may prevent the endothelial barrier dysfunction that generally accompanies inflammatory processes (which develop in S. mansoni-infected lungs). This effect may decrease the transendothelial migration of immune cells as well as the extravasation of fluid and macromolecules such as parasitic antigens or host-derived immune mediators and thus control the development of the inflammation in this organ. Moreover, the anti-inflammatory effects of schistosomula to ECs may have important immunological consequences during schistosomiasis. Consistent with the data presented in this report, we have recently shown that schistosomula do not directly signal ECs to express inflammatory markers, such as adhesion molecules (41) or inflammatory cytokines (unpublished data). Conversely, we found that schistosomula are stage specifically able, via a cAMP/PKA pathway, to inhibit the expression of certain adhesion molecules (VCAM-1 and E-selectin) on lung microvascular ECs stimulated with the pro-inflammatory cytokine TNF-. Taken together, these findings suggest that the activating effects exerted by schistosomula to ECs probably play a role in the evasion by the parasite of the inflammatory response.