|Home | About | Journals | Submit | Contact Us | Français|
Sialostatin L (SialoL) is a secreted cysteine protease inhibitor identified in the salivary glands of the Lyme disease vector Ixodes scapularis. Here, we reveal the mechanisms of SialoL immunomodulatory actions on the vertebrate host. LPS-induced maturation of dendritic cells from C57BL/6 mice was significantly reduced in the presence of SialoL. Although OVA degradation was not affected by the presence of SialoL in dendritic cell cultures, cathepsin S activity was partially inhibited, leading to an accumulation of 10 KDa invariant chain intermediate (Ii-p10) in these cells. As a consequence, in vitro antigen-specific CD4+ T cell proliferation was inhibited in a time-dependent manner by SialoL and further studies engaging cathepsin S−/− or cathepsin L−/− dendritic cells confirmed that the immunomodulatory actions SialoL are mediated by inhibition of cathepsin S. Moreover, mice treated with SialoL displayed decreased early T cell expansion and recall response upon antigenic stimulation. Finally, SialoL administration during the immunization phase of experimental autoimmune encephalomyelitis in mice significantly prevented disease symptoms, which was associated with impaired IFN-γ and IL-17 production and specific T cell proliferation. These results illuminate the dual mechanism by which a human disease vector protein modulates vertebrate host immunity and reveals its potential in prevention of an autoimmune disease.
Hard ticks feed on their hosts for an extended period of time, sometimes exceeding 10 days, over which a full range of inflammatory and immune reactions take place at the feeding site as well as systemically on the host body. Thus, active modulation of the host immune response by tick saliva is required for the completion of its long lasting blood feeding, which facilitates transmission of pathogens (1). Indeed, apart from the Lyme disease agent Borrelia burgdorferi, ticks efficiently transmit protozoa and viruses that cause a variety of diseases in humans (2). According to the Centers for Diseases Control and Prevention, Lyme disease is the most common tick-borne disease in North America and Europe and one of the fastest-growing infectious diseases in the U.S. (3).
A key effect of tick saliva on vertebrate immunity is the marked suppression of T cell proliferation, accompanied by downregulation of Th1 cytokines such as IFN-γ and IL-2 (4-6). Reduced proliferative activity of Con A on T lymphocytes was found in different hosts upon infestation with Rhipicephalus sanguineus, Ixodes ricinus and Dermacentor andersoni (7-9). In vitro mitogen-driven and antigen-specific proliferation has also been demonstrated to be inhibited by tick saliva from different species (10-13). Although the immunomodulatory properties of Ixodes salivary cocktail were revealed more than two decades ago (13), the constituents that account for these activities have not been fully characterized. Certain molecular components common to ticks, such as PGE2 have been shown to modulate lymphocyte proliferation (13, 14). Moreover, some proteins have been associated with such suppressive activity, including a 36-kDa protein from D. andersoni (15), an immunosuppressor from I. ricinus (16), an IL-2 binding protein (17) and Salp15 from I. scapularis (18, 19). We have recently characterized a secreted cysteine protease inhibitor from I. scapularis salivary glands that selectively target a limited subset of human cathepsins (20, 21). This inhibitor displays high affinity for cathepsin L (Ki 10−10 M), prompting us to name it sialostatin L (SialoL). When the sialostatin gene was silenced using the iRNA technique in ticks, the vertebrate host (in that case, rabbits) recognized ticks more quickly, leading to tick feeding impairment (21). Taking into consideration both the pivotal role of the sialostatin's enzymatic targets in antigen processing/presentation (22, 23), and the faster immune recognition of ticks in the absence of sialostatin secretion from their salivary glands (21), we have undertaken an investigation on the mechanism of action of this protein. More specifically, we demonstrate that SialoL inhibits microbial-induced maturation of dendritic cells (DCs) as well as antigen-specific T cell proliferation. Furthermore, we show that cathepsin S inhibition accounts for the observed SialoL-mediated effects on immunity and that in vivo treatment of mice with SialoL impairs early CD4+ T cell expansion upon antigenic stimulation and in vitro recall responses. Finally, using a murine model for multiple sclerosis, we show that in vivo administration of SialoL delays disease onset and prevents its symptoms. Collectively, these data shed light on the immunomodulatory mechanism of SialoL and its preventive potential against an autoimmune disease. Beyond the basic knowledge on the mechanisms that ticks have developed to successfully obtain a blood meal, the current work shows SialoL to be an attractive candidate in the development of novel drug formulations for the treatment of immunity related pathological conditions such as autoimmune diseases.
Unless otherwise indicated, protocols followed standard procedures (24), and experiments were performed at room temperature (25° ± 1° C). All water used was of 18 MΩ quality, produced by a MilliQ apparatus (Millipore). If not otherwise stated, all reagents were purchased from Sigma-Aldrich Co., and all cells were cultured at 37°C under an atmosphere of 5% CO2. All experimental protocols involving animals were approved by the Institutional Animal Care and Use Committee (NIAID).
The SialoL gene was overexpressed in Escherichia coli, and the corresponding active protein was purified in 0.8 mM stock solution as previously described (20, 21). Any potential LPS contamination in the stock solution was removed by Arvys Proteins Inc. by detergent extraction; endotoxin presence by the end of the procedure was estimated as lower than 4 × 10−5 endotoxin units per μg of protein (roughly, less than 3 × 10−14g of endotoxin per μg of protein) with a sensitive fluorescent-based endotoxin assay (PyroGene recombinant factor C endotoxin detection system; Lonza Biologics Inc.).
Female mice (6 to 10 wk old) were used. C57BL/6 and BALB/c were purchased from Charles River Laboratories. B6.PL-Thy1a/CyJ mice were purchased from The Jackson Laboratory. OT-II mice (which express transgenic TCR specific for OVA peptide 323−339) were purchased from Taconic Farms. Cathepsin S−/− (C57BL/6 background) and L−/− (C57BL/6/S129 background) mice, and their wild-type (WT) littermates were bred at the Department of Medicine, Brigham and Women's Hospital and Harvard Medical School (Boston, MA). Animals were maintained at an American Association of Laboratory Animal Care-accredited facility at the NIH.
Bone marrow-derived DCs from cathepsin S−/− and L−/− mice and their respective WT littermates were generated as previously described (14). After 6 days of culture with GM-CSF, nonadherent cells were collected and enriched for CD11c+ cells as described elsewhere (25). CD4+ T cells were purified from spleens of OT-II mice using a Dynal® mouse CD4 negative isolation kit (Invitrogen).
DCs from C57BL/6 mice were cultured at 105 cells/well in flat-bottom 96-well cluster plates (Costar; Corning Inc.) and preincubated for 2 h with medium in the presence or absence of SialoL. Subsequently, they were stimulated with 50 ng/ml of ultrapure LPS (InvivoGen). Cell-free supernatants were collected at 6 h for TNF-α and 48 h for IL-12p70 and IL-10 and levels of these cytokines were measured using OptEIA™ ELISA sets according to manufacturer's instructions (BD Biosciences).
Four million bone marrow-derived DCs were preincubated with medium or 3 μM SialoL. After 2 h, medium or LPS (50 ng/ml) were added and incubated overnight. Cells were washed twice and stained with fluorochrome-labeled Abs against CD11c, CD80, CD86, and MHC class II (I-Ab; BD Pharmingen). A total of 100,000 live cell events as gated on forward and side scatter characteristics were acquired. Data were collected using a FACSCalibur (BD Immunocytometry Systems) with CellQuest (BD Biosciences) and analyzed with FlowJo software (Tree Star).
DCs were prepared as described above and incubated with medium or 3 μM SialoL for 24 h. Then, they were washed 3 times in cold PBS and cell pellets were lysed in lysis buffer (50 mM Tris-HCl pH 7.4, 0.5% NP-40 and 5 mM and 5 mM MgCl2) on ice for 2 h. After centrifugation at 15,000 g for 10 min, volumes of each sample containing the same amount of protein were dissolved into nonreduced LDS Sample Buffer (Invitrogen) and either boiled or nonboiled samples were separated on 12% SDS-PAGE. Separated proteins were were transferred onto nitrocellulose filters which were then probed with anti-Ii (CD74 – BD Pharmingen) and anti-cathepsin S (Santa Cruz Biotechnology, Santa Cruz, CA). Horseradish peroxidase- or alkaline phosphatase-conjugated secondary antibodies were used for signal detection. Filters were developed with Western Blue Stabilized Substrate for Alkaline Phosphatase (Promega Corp., Madison, WI) or SuperSignal West Pico Chemiluminescent Substrate for horseradish peroxidase (Thermo Fisher Scientific, Atlanta, GA).
DCs were prepared as described above and preincubated with medium or 3 μM SialoL for 3 h. DQ™ OVA (1 μg/mL) was then added to the cultures and further incubated for 2 h. A control group consisted of cells incubated with medium, but not with DQ™ OVA. Cells were repeatedly washed with PBS/SBF 1% at 4° C and OVA degradation inside the cells was analyzed by flow cytometry (FL-1 channel).
DCs (2.5 × 104/well) were preincubated at different time points with medium or 3 μM SialoL, as indicated. Then, CD4+ T cells (105/well) were added and cultures were incubated for 72 h in the presence or absence of 1 μg/mL OVA (Imject® OVA; Pierce).
For mixed lymphocyte reaction, total splenocytes from BALB/c mice (2.5 × 105/well) were preincubated for 3 h with medium in the presence or absence of SialoL. Subsequently, splenocytes from OT-II mice (2.5 × 105/well) were added and incubated for 72 h.
Proliferation was assessed by adding 10% (v/v) Alamar Blue® (TREK Diagnostic Systems Inc.) in the last 48 h of incubation. Absorbance was measured at 570 nm and 600 nm as previously described (26).
CD4+ T cells from OT-II mice (Thy1.2) were purified as described above and labeled with 2 μM CFSE (Invitrogen) according to manufacturer's instructions. B6.PL-Thyla/CyJ mice (Thy1.1) were i.v. injected with 2.5 × 106 CD4+ T cells and, 1 day later, immunized with 10 μg OVA emulsified in complete Freund's adjuvant (CFA). A volume of 100 μl emulsion/site was s.c. injected into two sites on the flanks of mice near the tail. Six h prior to immunization, concomitant with the immunization, and 6 h after the immunization, mice received s.c. injections of BSA or SialoL (10 μg/ injection) into the same sites of the flanks. Additional control groups consisting of mice receiving only CFA in PBS (as a negative control for the immunization) and PBS as a treatment were analyzed. Lymph nodes (LN) were collected after 4 days and cells were stained with anti-CD4 allophycocyanin, anti-Thy1.2 phycoerythrin (BD Biosciences). A total of 1,000,000 live cell events as gated on forward and side scatter characteristics were acquired and analyzed for CFSE dilution by flow cytometry as described above.
In another set of experiments, naïve C57BL/6 mice were immunized by the same procedure, and their LN cells were collected 30 days post treatment, suspended at 2.5 × 105/well, and restimulated in vitro with medium, OVA (0.1, 1, and 10 μg/mL) or Con A (0.5 μg/ml) for 72 h. Alamar blue was added in the last 24 h of incubation, and proliferation was assessed as described above.
Induction of EAE was performed according to a previously published protocol (27). Female C57BL/6 mice were immunized with 200 μg myelin oligodendrocyte glycoprotein (MOG) 35−55 peptide (MOG p35−55) (Anaspec) emulsified in incomplete Freund's adjuvant (IFA) together with 5 mg/ml Mycobacterium tuberculosis H37RA. A volume of 100 μl emulsion/site was s.c. injected into two sites on the flanks of mice near the tail. At days 0 and 2 following the initial injections, animals received additional i.p. injections of 200 ng pertussis toxin. Six h prior to immunization, concomitant with immunization, and 6 h after immunization, mice received s.c. injections of PBS, SialoL, or SialoL2 (10 μg/ injection) on the flanks into the same sites as the immunization. Mice were scored daily for clinical assessment of disease based on the following criteria: 0, normal; 1, limp tail or hind limb weakness; 2, limp tail and hind limb weakness; 3, one hind limb paralysis; 4, both hind limb paralysis; 5, both hind limb and forelimb paralysis; 6, moribund or dead. Food and water was made accessible to immobile animals, and moribund animals with a score of 6 were euthanized.
The left and right inguinal LN from naïve or MOG-immunized mice treated with BSA (negative control) or SialoL were removed 10 days after immunization. At this period, mice do not show any clinical symptom of EAE. The cells were dispersed through a 40-μM cell strainer (BD Falcon™), and red blood cells were hypotonically lysed. The cells were washed twice and cultured at 2 × 105/well for 72 h in 96-well microplates (BD Falcon™) with either medium or MOG (1 and 5 μg/mL). Proliferation was assessed by adding Alamar Blue® as described above, and levels of IFN-γ (BD Biosciences) and IL-17 (R&D Systems) were determined in cell-free supernatants.
Each experiment was performed at least three times and data are shown as mean ± SEM. Statistical differences were analyzed by Student t-test and analysis of variance (ANOVA). A P value of 0.05 or less was considered statistically significant.
Given that DCs are a major resident cell type in the skin and SialoL is present in I. scapularis salivary secretion (20), we first tested whether the DC response to microbial stimulation is affected by the presence of SialoL. LPS-induced production of IL-12 by DCs was inhibited in a concentration-dependent manner by SialoL, reaching ~60% inhibition at 3 μM (Fig. 1A; P < 0.05). TNF-α production was marginally inhibited in the same conditions, reaching ~25% inhibition at 3 μM SialoL (Fig. 1B; P < 0.05). None of the mentioned effects appear to be associated with alterations in IL-10 production, as no differences were observed in its production in the absence or presence of SialoL (Fig. 1C).
We next tested whether the expression of costimulatory molecules was affected by SialoL. Incubation with SialoL alone did not significantly change expression of CD80 or CD86 in DCs when compared to control cells (incubated with medium only). However, preincubation of DCs with SialoL reduced the expression of these molecules by 40% of the level induced by LPS (Fig. 2). LPS-induced CD40 expression was not affected by preincubation with SialoL (data not shown).
Next, we asked whether the effects of SialoL were mediated by its interaction with intracellular cathepsin. DCs cultured with medium or SialoL were extensively washed, lysed and blotted to detect cathepsin S. Figure 3A shows a reduced intensity of cathepsin S band from SialoL-incubated DCs when compared to medium-incubated DCs. This result was confirmed by densitometry (Fig. 3B) and it was consistently found in 4 different experiments (38.8 ± 6.7% reduction - P < 0.05), suggesting formation of SDS-resistant intracellular complex where SialoL-bound cathepsin S displays a distinct migration pattern compared to free enzyme. In addition, it seems that under denaturing conditions, the complex is preserved in non-boiled samples preventing antibody recognition of bound cathepsin S. To exclude that SialoL affected the expression levels of cathepsin S, the samples were boiled to dissociate the complex and, in fact, boiled samples presented bands of similar intensities, further confirming our hypothesis (Fig. 3A and B). Similar results were found upon cell incubation with LHVS, a well-characterized cathepsin S inhibitor (data not shown) (28).
We next investigated the effects of cathepsin S inhibition on antigen degradation and invariant chain (Ii) cleavage, two essential steps for antigen presentation (23). Regarding the antigen degradation, DCs were preincubated with medium or SialoL and pulsed with DQ-OVA, which emits fluorescence upon proteolysis. Figure 3C shows that DQ-OVA fluorescence is equally detected in DCs in the presence or absence of SialoL, suggesting that SialoL does not affect this step, which is known to have participation of additional cathepsins beyond cathepsin S (23). To evaluate Ii degradation, DCs were incubated with medium or SialoL, and after 24 h, the Ii intermediates were detected in cell lysates using anti-Ii (CD74) antibody. Figure 3D shows that the p31 and p41 isoforms of Ii (Ii-p31 and Ii-p41) are equally detected in cell lysates from both groups while Ii-p10, whose cleavage is dependent of cathepsin S (29, 30), accumulates in the cells incubated with SialoL, but not in control DCs. Additionally, SDS-stable MHC class II/Ii-p10 complexes can be dissociated upon boiling (29) and, in fact, boiled cell lysates from SialoL incubated DCs presented a stronger Ii-p10 band (Fig. 3D).
To determine whether the inhibition of DC maturation and Ii degradation correlate with a functional defect in antigen presentation, we next investigated the modulation of antigen-dependent T cell proliferation by SialoL. A time-dependent inhibition of OVA-specific CD4+ T cell proliferation was observed in the presence of SialoL, reaching ~65% upon 3 h preincubation with the inhibitor (Figure 4A; P < 0.05). In contrast, the mixed lymphocyte reaction was not affected by preincubation with SialoL at any of the concentrations tested ranging from 0.1 to 3 μM, excluding SialoL unspecific toxicity (Fig. 4B). Collectively, these data suggest that SialoL affects DC maturation and function, rather than a direct effect on T cells.
To further demonstrate the cathepsin S-dependent inhibitory activity of SialoL on T cell proliferation, SialoL-exposed or unexposed DCs from cathepsin L−/− and S−/− mice were used as APCs in the OVA-induced proliferation assay. Cathepsin L−/− DCs presented a similar proliferative stimulatory ability when compared with WT DCs (Fig. 5A). Upon preincubation with SialoL, both WT and cathepsin L−/− DC cultures produced a reduction in T cell proliferation (Fig. 5A; P < 0.05). In contrast, T cell proliferation in cathepsin S−/− DC cultures was strongly diminished when compared with WT DCs in the absence of SialoL (Fig. 5B; P < 0.05), while preincubation of these cultures with SialoL inhibited proliferation induced by WT and cathepsin S−/− DCs (Fig. 5B; P < 0.05). Because cathepsin S−/− DCs are not expected to stimulate OVA-specific T cell clones (29), the remaining proliferative activity presented by cathepsin S−/− DC cultures suggests a minor contamination with APCs from the donor (OT-II mice) in the CD4+ T cell preparation, as they were purified by negative selection. Consistent with this concept, we observed a complete lack of T cell proliferation in cathepsin S−/− DC cultures incubated with SialoL (Fig. 5B). Altogether, these results suggest that SialoL inhibits antigen-specific T cell proliferation by a cathepsin S-dependent mechanism.
We next evaluated whether the inhibitory effects of SialoL on T cell proliferation can be also achieved in vivo. To do so, recipient B6.PL-Thy 1a/CyJ mice (Thy 1.1) received CFSE-labeled CD4+ T cells from OT-II mice (Thy 1.2) and were subsequently immunized with OVA in the presence of BSA (control protein) or SialoL (see Material and Methods for immunization protocol). CFSE dilution was analyzed 4 days after immunization. As expected, transferred cells represented only a small fraction of total cells in a non-immunized recipient (Fig. 6A) and did not proliferate (Fig. 6B). In OVA-immunized mice, transferred CD4+ T cells represented around 3% of total cells (Fig. 6C) and virtually all of them proliferated, presenting up to 8−9 divisions (Fig. 6D). Nevertheless, in SialoL-treated mice immunized with OVA, transferred CD4+ T cells represented less than 1% of total cells (Fig. 6E), and nearly half of them did not proliferate (no CFSE dilution - Fig. 6F).
In another set of experiments, naïve C57BL/6 mice were treated with SialoL or BSA and subsequently immunized with OVA. After 30 days, their lymph nodes (LN) cells were isolated, restimulated in vitro, and their proliferative response was compared. Proliferation of LN cells from SialoL-treated animals was reduced in comparison with BSA-treated animals and this inhibition was consistent in all OVA concentrations used upon in vitro restimulation (Fig. 6G; P < 0.05). However, the polyclonal proliferation induced by Con A was not affected in any group (Fig. 6G), suggesting that SialoL treatment affects antigen-specific proliferation but not mitogen-driven proliferation. Of note, when a much higher amount of OVA (10 times more) was employed during immunization, the inhibitory effect on the proliferative response was barely or not observed (data not shown), indicating that an excessive amount of antigen is capable of surpassing this inhibitory activity.
We next determined whether the observed in vivo effects of SialoL could be extended to an autoimmune disease setting. We used the EAE model, because this disease is caused by autoreactive T cells primed by DCs after immunization of mice with antigens derived from a myelin sheath protein (31). Therefore, the disease onset depends on the functionality of DCs. During MOG immunization, mice were treated with PBS, SialoL or SialoL2; the latter being a second cystatin from I. scapularis with at least 500 times lower affinity for cathepsin S (21). Strikingly, SialoL prevented disease symptoms at the onset of EAE between days 12 and 15 after immunization (Fig. 7A). In contrast, mice receiving SialoL2 displayed no differences in clinical score compared with PBS-injected group, serving as an additional negative control group (Fig. 7A). After the peak of EAE onset, in the late milder phase of the disease (day 16 to the end of the monitoring period), there was no difference between the experimental groups. In addition, since SialoL was previously shown to inhibit neutrophil migration (20), we have tested whether this anti-inflammatory activity affects the disease when T cells are already primed by the antigen. Treatment of animals with SialoL at days 12, 14 and 16 after MOG immunization had no effect on the EAE development (data not shown). we have tested this inflammatory activity on MOG-immunized mice
Next, the inguinal LNs that drain the MOG injection site were dissected at day 10 following MOG immunization (before the onset of the clinical signs of EAE). We then evaluated the proliferative response and cytokine production of LN cells upon in vitro restimulation with MOG. In animals treated with BSA at the time of MOG immunization, LN cells were strongly responsive to MOG restimulation, proliferating and producing IFN-γ and IL-17 (Fig. 7B-D). On the other hand, LN cells from mice treated with SialoL at the time of MOG immunization presented a weak proliferative response and barely detectable cytokine production upon in vitro MOG restimulation (Fig. 7B-D). This decreased response was not associated with increased IL-10 or TGF-β production (data not shown).
We have previously shown that SialoL is a constituent of the I. scapularis salivary secretion (20), and therefore is co-administered with the pathogens transmitted while these ticks are attached to the skin. In this environment, DCs are a major resident cell type and probably among the first cells activated by pathogen products such as TLR ligands from these microorganisms. Thus, we determined whether the DC response to LPS, a typical TLR ligand, is affected in the presence of SialoL by measuring cytokines and costimulatory molecules involved in the maturation process of these cells. In vitro LPS-induced production of IL-12 (Fig. 1A) and, less extensively TNF-α (Fig. 1B), were affected by SialoL presence in the culture. This observed down-modulation is apparently not associated with changes in IL-10 production, as levels of this cytokine were similar in the presence or absence of SialoL (Fig. 1C). We have previously demonstrated that I. scapularis saliva inhibits LPS-induced TNF-α and IL-12 production and increases IL-10 production by DCs and that PGE2 is the major low molecular weight mediator responsible for this activity in saliva (14). The results presented here suggest that inhibition of TNF-α and IL-12, but not the increase of IL-10 production, may be also attributed to salivary SialoL, which is not present in the low molecular weight fraction characterized in our previous work. Additionally, CD80 and CD86 expression in DCs, which are increased upon incubation with LPS, are reduced in the presence of SialoL (Fig. 2). Nevertheless, the participation of cysteine proteases in DC maturation has never been observed, and it is currently under investigation in our laboratory.
DCs are also potent APCs that initiate adaptive immune responses through activation of T cells. Antigen processing and presentation by APCs is a complex process with two concomitant pathways, both dependent on cathepsins (23, 32). The antigen degradation, which produces antigenic peptides to bind MHC class II molecules is a process only partially dependent on cathepsin S, and other cathepsins are able to replace it in the absence of the former (23). In contrast, the final proteolysis of the invariant chain associated with MHC class II molecules is exclusively dependent of cathepsin S (33, 34). Western blot analysis consistently showed a partial decrease of cathepsin S intensity in cell lysates of DCs incubated with SialoL, suggesting intracellular complex formation between SialoL and cathepsin S, a finding that is consistent with the picomolar affinity between the enzyme and inhibitor (21). This binding was reversible upon boiling (Fig. 3A-B), indicating that binding of tight inhibitor to cathepsin S either masks or distorts the epitope recognized by the antibody used to reveal this enzyme in Western blot assays, and experiments performed in the same conditions with the well-characterized cathepsin S inhibitor LHVS (28, 30) confirmed these findings (data not shown). To investigate whether cathepsin S inhibition by SialoL interfered with the two essential steps for antigen presentation, we firstly analyzed DQ-OVA cleavage inside DCs. Figure 3C shows that DQ-OVA fluorescence is equally detected in DCs in the presence or absence of SialoL, suggesting that the presence of SialoL inside these cells is not associated with a general lysosomal poisoning. Then, Ii-p10 intermediate, whose cleavage is catalyzed by cathepsin S, was detected in cell lysates. Strikingly, Ii-p10 was detected in cells incubated with SialoL (Fig. 3D) and this accumulation is indicative of a defective peptide loading into MHC class II molecules.(29) The inhibitory role of SialoL in OVA-specific proliferation (Fig. 4A) confirms the biological significance of our previous findings. This result is consistent with previous studies carried out with nonspecific and specific inhibitors of cysteine proteases (34, 35). However, the strict and narrow target specificity of SialoL (20) restricts its action to antigen-specific proliferation systems, as no inhibitory activity on mixed lymphocyte reaction has been observed (Fig. 4B). In addition, the defective T cell proliferation when using cathepsin S−/− DCs, but not cathepsin L−/− DCs (Fig. 5), is in agreement with a previous work (23). Thus, our results confirm a dominant role for the inhibition of cathepsin S by SialoL over that of cathepsin L, as the molecular mechanism that accounts for its effects on antigen presentation. Together with the above-described activity on DC maturation (Figure 1 and and2),2), these data suggest that SialoL does not directly affect T cell function, but rather interferes with both TLR-dependent and antigen-dependent functions of DCs.
The immunomodulatory action of SialoL was also observed in vivo. More specifically, SialoL treatment reduced CD4+ T cell proliferation in vivo relative to BSA treatment upon OVA immunization (Fig. 6), suggesting that APCs exposed to the inhibitor displayed reduced functionality. Inhibition of early T cell expansion by SialoL affected subsequent recall responses, as ex vivo recall response to OVA was impaired in mice receiving SialoL, while Con A-induced proliferation was not affected (Fig. 6G). This suggests that the inhibitory action of SialoL in vivo is specific for antigen-dependent proliferation and not for mitogen-driven polyclonal proliferation. As a proof of principle, we tested SialoL activity in an animal model for human multiple sclerosis, i.e., EAE. The disease is mediated by autoreactive T cells that recognize antigens of the myelin sheath of the central nervous system, including myelin basic protein and MOG (36). Induction of EAE requires T cell responses initiated by DC priming of naïve CD4+ T cells (31) and can also occur by passive transfer of myelin antigen–specific Th17 or Th1, but not Th2 cells.(37-39) Markedly, treatment of mice with SialoL during the immunization phase of EAE concomitant with MOG injection resulted in delayed EAE development and milder disease symptoms in comparison with PBS-treated group (Fig. 7A). SialoL2, a second cystatin described in I. scapularis saliva that displays similar inhibitory activity for cathepsin L but it is more than 500 fold weaker cathepsin S inhibitor in comparison to SialoL (21), presented no effect on the symptoms development (Fig. 7A). In addition, draining LN cells from SialoL-treated mice were markedly less responsive to in vitro MOG restimulation than LN cells from control mice (Fig. 7B-D). Based on the in vitro and in vivo results, it is reasonable to conclude that SialoL negatively affected DC maturation and function, thus interfering with MOG-specific T cell stimulation and causing the observed effect on disease onset. This scenario is further corroborated by the strongly inhibited proliferation of LN cells from SialoL-treated mice upon MOG restimulation in vitro. Additionally, cytokines involved in development of EAE, namely IFN-γ and IL-17 (39-42), were suppressed, indicating that a defect in priming of CD4+ T cells was the mechanism by which SialoL treatment suppressed EAE development. To our knowledge, this is the first work showing the preventive role of an ectoparasite protein on an autoimmune disease. This is in line with previous studies relating cathepsin S to autoimmune pathologies (43-45), a novel field that some pharmaceuticals companies are beginning to explore with cathepsin S inhibitors in clinical trials (46). Thus, we propose that the SialoL/cathepsin S axis could be used in designing novel treatments for autoimmune diseases.
The authors thank Drs. Thomas E. Wellems, Robert W. Gwadz, and Kathryn C. Zoon (NIAID, NIH) for support. We also thank Van M. Pham and Roseanne Hearn for technical assistance, NIAID intramural editor Brenda Rae Marshall for assistance and Drs Alan Sher and Dragana Jankovic from LPD-NIAID/NIH for help and critical reading of the manuscript.
The authors have no conflicting financial interests.
1This work was supported by funding from the Intramural Research Program of the Division of Intramural Research, National Institute of Allergy and Infectious Diseases, National Institutes of Health. A.B. received funding from CNPq (472477/2007-2 and 565496/2008-5), FAPESC (04524/2008-1), and WHO/TDR (2008-8734-0).