|Home | About | Journals | Submit | Contact Us | Français|
Saliva of the blood feeding sand fly Lutzomyia longipalpis was previously shown to inhibit the alternative pathway (AP) of the complement system. Here, we have identified Lufaxin, a protein component in saliva, as the inhibitor of the AP. Lufaxin inhibited the deposition of C3b, Bb, Properdin, C5b, and C9b on agarose-coated plates in a dose-dependent manner. It also inhibited the activation of factor B in normal serum, but had no effect on the components of the membrane attack complex. Surface plasmon resonance (SPR) experiments demonstrated that Lufaxin stabilizes the C3b-B proconvertase complex when passed over a C3b surface in combination with factor B. Lufaxin was also shown to inhibit the activation of factor B by factor D in a reconstituted C3b-B, but did not inhibit the activation of C3 by reconstituted C3b-Bb. Proconvertase stabilization does not require the presence of divalent cations, but addition of Ni2+ increases the stability of complexes formed on SPR surfaces. Stabilization of the C3b-B complex to prevent C3 convertase formation (C3b-Bb formation) is a novel mechanism that differs from previously described strategies used by other organisms to inhibit the AP of the host complement system.
Sand flies are dipteran plant feeders which females must feed on vertebrate blood in order to maturate their eggs. To be able to blood feed on a vertebrate host, the sand fly salivary secretion contains proteins, peptides, and small molecules aimed at modulating hemostatic responses of the host, including coagulation, vasoconstriction, and platelet aggregation, as well as inflammation and immune responses (1). In saliva of Lutzomyia longipalpis, the main vector of visceral leishmaniasis in the Americas, different molecules with antihemostatic properties have been characterized, such as the vasodilator Maxadilan (2, 3), the anticoagulant Lufaxin (4), the endonuclease Lundep (5), and, recently, the inhibitor of the classical pathway of the complement system SALO (6).
The complement system is part of the innate immune response that is responsible for the opsonization of cell surfaces, generation of potent anaphylatoxins, and direct killing of invasive pathogens and altered self-cells (7). Moreover, it is involved in antigen presentation and lymphocyte activation, through facilitation of the adaptive immune response (8, 9). Composed of more than 30 plasma and membrane proteins, the complement system can be triggered through three pathways, the classical (CP), lectin (LP), and alternative (AP). Once activated, these pathways converge at the C3 convertase and continue in a common pathway leading to formation of the membrane attack complex (MAC) that causes cell death (10). The complement components are also responsible for non-lethal effects such as cell activation and proliferation, resistance to subsequent complement lysis and either resistance to or induction of apoptosis (11).
The AP is primarily activated by spontaneous hydrolysis of the C3 component, forming soluble C3-H2O. This form binds factor B, that is then cleaved by factor D, generating C3-H2O-Bb. This soluble convertase cleaves more C3 molecules to C3b that go on to recognize surface hydroxyl or amino groups and covalently bind to them (12, 13). Factor B binds to deposited C3b forming the AP C3 proconvertase (C3b-B) that is then cleaved by factor D, forming the active C3 convertase of the AP (C3b-Bb). This complex activates more C3 molecules but has a short half-life of about 90s, requiring properdin for stabilization (14). Activation of C3 leads to self-amplification and generation of the C5-convertase C3b-Bb-C3b, directing the formation of the MAC (7). As it is being constantly activated in the blood, the AP is a crucial mechanism of immune surveillance. Moreover, it has been shown to play a critical role in amplification of both the CP and LP and is responsible for 80% of the C5a and MAC generated by complement pathways (15).
Lutzomyia longipalpis saliva inhibits both the CP and the AP (16, 17). This inhibition is important for successful blood feeding in that it can diminish the inflammatory response at the bite site and protect the insect’s midgut from deleterious effect of the MAC (18–20). SALO, the L. longipalpis salivary inhibitor of the CP, was recently described as being an 11kDa protein that acts on the first steps of the cascade (6). In this paper, we show that Lufaxin, a known L. longipalpis salivary anticoagulant (4) and a candidate vaccine for leishmaniasis (21), is the inhibitor of the AP in unfractionated saliva. We also demonstrate that Lufaxin binds to the C3b-B complex and inhibits activation of factor B and consequently the formation of the C3 convertase, a unique mode of action not seen in other organisms.
All animal procedures were reviewed and approved by the National Institute of Allergy and Infectious Diseases (NIAID) Animal Care and Use Committee under protocol LMVR4E and handled in accordance to the Guide for the Care and Use of Laboratory Animals and with the NIH OACU ARAC guidelines and also approved by Ethics Committee in Animal Experimentation (CETEA) of Universidade Federal de Minas Gerais (UFMG) under Protocol no. 87/2011.
Sand fly transcripts coding for Lufaxin and 11 other salivary proteins were cloned into the VR2001-TOPO vector as described before (21, 22). Recombinant proteins were produced in Leidos Biomedical Research PEL facility by transfecting HEK 293F cells with the VR2001-TOPO DNA plasmids coding for the different sand fly salivary proteins and incubated for 72h. The supernatant was concentrated and further purified by HPLC (NGC Chromatography system, Bio-Rad Laboratories, Inc.) using a HiTrap chelating HP column (GE Healthcare) charged with Ni2SO4. Imidazole was removed from fractions containing Lufaxin by washing with PBS using a 5,000, 10,000, or 30,000 MWCO Amicon filter (Millipore). Purified Lufaxin was analyzed by NuPage 4–12% gels (4) and stored at −80°C until use. Aliquots did not undergo more than three freeze-thaw cycles.
Twelve recombinant salivary proteins from L. longipalpis were tested on standardized AP-mediated hemolysis assays in order to detect anti-complement activity as previously described (23). The rabbit erythrocytes were acquired from CompTech or collected by venous puncture from a rabbit kept on the animal facility of UFMG. Before the experiments, 500µl of rabbit blood were washed three times in 5ml of Mg-EGTA solution (1mM HEPES, 30mM NaCl, 10mM EGTA, 7mM MgCl2, 3% glucose, and 0.02% gelatin, pH 7.4) as described in Ref. (16). The erythrocyte concentration was adjusted to 1×108 cells/ml. All the experiments were performed at pH 7.4, unless specified. Briefly, in 1.5ml microcentrifuge tubes, 25µl of normal human serum (NHS, CompTech) diluted 1:20 in Mg-EGTA buffer (1mM HEPES, 30mM NaCl, 10mM EGTA, 7mM MgCl2, 3% glucose, and 0.02% gelatin, pH 7.4) were mixed with 12.5µl of PBS containing 1µg of each recombinant protein. Then, 25µl of Mg-EGTA containing 2.5×106 rabbit red blood cells were mixed and the tubes incubated for 30min at 37°C for complement activation. Final concentration of NHS in 62.5µl of buffer was 2%. After incubation, 250µl of cold PBS were added and the tubes rapidly centrifuged. Two hundred microliters of the supernatants were transferred to microplates and read at 415nm. Tubes incubated without any recombinant protein were used as positive controls and tubes without NHS were used as negative controls. Tubes containing red blood cells but without NHS and recombinant proteins were combined with 250µl of distilled water and were used to obtain total hemolysis. The assays were performed in duplicate, and in every test the mean of negative control was subtracted from the mean of the other results. The results were then transformed in percentage of lysis, considering tubes treated with distilled water as 100% of hemolysis.
In some experiments, AP-mediated hemolytic assays were performed with different concentrations of recombinant Lufaxin or LJS169 using Mg-EGTA buffer at pH 7.4 and pH 8.15 [pH in the intestinal lumen of L. longipalpis right after blood meal (24)]. Similar assays were also carried out using properdin-depleted serum (CompTech) diluted 1:2 in Mg-EGTA buffer (pH 7.4) and final concentrations of Lufaxin ranging from 0 to 200nM.
In order to investigate if Lufaxin inhibits classical pathway, CP-mediated hemolytic assays were performed using IgG-sensitized sheep red blood cells (CompTech) as described by Ferreira et al. (6). Possible inhibition of LP by Lufaxin and LJS169 was investigated as described in the study of Mendes-Sousa et al. (23) through detection of the deposition of C4b component on Saccharomyces cerevisiae’s mannan-coated microplates using anti-C4 polyclonal antibodies (CompTech).
Balb/c mice were injected three times every 2weeks, intradermally in the ear with 5µg of recombinant Lufaxin mixed (1:1 volume) with Magic™ Mouse Adjuvant (Creative Diagnostics) as recommended by the manufacturer. Fifteen days after the last inoculation, blood was collected to obtain the Lufaxin antiserum. The ability of Lufaxin antiserum to recognize the native and recombinant Lufaxin was checked by Western blot as previously described (4). Lufaxin (500ng) and SGH (five pairs of L. longipalpis glands) were run in a NuPAGE 4–12% gel (ThermoFisher Scientific) and transferred to nitrocellulose membranes (ThermoFisher Scientific). Membranes were blocked with 5% non-fat milk solution overnight, incubated with Lufaxin antiserum (1:1,000 dilution) or mice pre-immune serum (control; 1:1,000 dilution), followed by anti-mouse secondary antibodies (Promega), and developed by addition of Western Blue® stabilized substrate for alkaline phosphatase (Promega).
Rabbit erythrocytes (5×106) with NHS (1:20) and samples were incubated at 37°C for 30min, and the percentage of hemolysis was quantified. Erythrocytes incubated with distilled water were considered as 100% of hemolysis. Results were expressed as mean±SEM of at least three independent experiments. Lufaxin (100nM) or L. longipalpis salivary gland homogenate (SGH) was incubated at 37°C for 10min with varying dilutions of anti-Lufaxin serum prior to the AP-hemolytic assays. The Lufaxin-serum or SGH/serum mixture was then used in hemolytic assay as described above.
ELISA plate (Corning Inc.) wells were filled with 100µl of 0.1% agarose solution and dried completely by overnight incubation at 37°C. Then, 100µl of HMEBN buffer (5mM HEPES, 7mM MgCl2, 10mM EGTA, 5mg/ml BSA, and 140mM NaCl, pH 7.4) containing 20% NHS and different concentrations of Lufaxin or LJS169 were added to the wells and the plate incubated at 37°C for 30min under agitation. After two washes, the wells were incubated for 30min with antibodies against each specific complement components produced in goat [C3, factor B, properdin, C5, or C9 (CompTech)] diluted in 10mM HEPES and 140mM NaCl solution. After two more washes, the wells were treated with anti-goat peroxidase-conjugated antibodies (Sigma-Aldrich) and then developed with 200µl of developing buffer (50mM Na3C6H5O7, 50mM Na2HPO4, 1mg/ml o-phenylenediamine (Sigma-Aldrich), and 0.075% H2O2, pH 5.0). The plate was read at 450nm at kinetic mode (one read every 30s) at 37°C for 10min. Wells without serum and without recombinant protein were used as negative and positive control, respectively. The assays were performed in triplicate. The mean of negative controls was subtracted from the mean of other results and then transformed in percentage of deposition considering the positive control as 100% deposition.
To detect activation of complement components in the presence of Lufaxin, Western blotting assays were performed as described in the study of Mendes-Sousa et al. (23). First, AP-mediated hemolytic assays were done as described above in the presence of Lufaxin, LJS169, or PBS. After different times of incubation (0, 30, and 60min), an aliquot of the supernatant of each tube was collected, mixed with denaturing sample buffer, and loaded onto SDS gels. Purified factor B (10ng), Bb (10ng), and C3a (5ng) (CompTech) were loaded on the SDS gels as controls.
In the assays for detection of C3 activation, an aliquot with diluted NHS without inhibitor and red blood cells was also loaded onto the gel. After transfer to nitrocelullose membranes (Bio-Rad Laboratories, Inc.), the membranes were blocked for 2h with 0.05% Tween 20 and 10% dried non-fat milk in PBS under agitation and then incubated with goat anti-factor B or rabbit anti-C3a antibodies (CompTech) diluted 1:1,000 in 0.05% Tween 20 and 1% BSA in PBS for 1h. After washing three times with 0.05% Tween 20 in PBS, the membranes were treated with anti-goat or anti-rabbit peroxidase-conjugated antibodies (Sigma-Aldrich) diluted 1:3,000 in the same buffer of primary antibodies for 1h. Visualization of the proteins was possible after treating the membranes with the peroxidase substrate diaminobenzidine kit (Vector Laboratories).
To determine if Lufaxin acts at MAC formation, AP-mediated hemolytic assays were performed with some modifications. Primarily, in 1.5ml microcentrifuge tubes, 25µl of C6-depleted serum (CompTech) diluted 1:20 in Mg-EGTA buffer were mixed with 12.5µl of PBS containing recombinant Lufaxin or LJS169 and 25µl of rabbit red blood cells and incubated at 37°C for 15min. The final concentration of the recombinant proteins was 75nM. Tubes without recombinant proteins (only PBS) were used as negative controls. After incubation, the tubes were centrifuged (1min, 1,700g) and the supernatant discarded. Erythrocytes were resuspended in a mixture of 12.5µl PBS plus 50µl of NHS diluted 1:40 in GHB-EDTA buffer (5mM HEPES, 145mM NaCl, 10mM EDTA, and 0.1% gelatin, pH 7.4) and incubated at 37°C for 30min. After this step, the assay followed as described previously.
For detection of inhibition after the C6 component (MAC assembly), the assay was performed without recombinant protein during the first incubation, and the erythrocytes were resuspended in a mixture of 12.5µl of PBS containing Lufaxin or LJS169 and 50µl of NHS diluted 1:40 in GHB-EDTA buffer. Detection of hemolysis was evaluated as described above.
In order to check if Lufaxin affects factor D activity, enzymatic assays based on Kam et al. (25) were performed. Purified factor D (6µg/ml) (CompTech), Lufaxin (0.1µg), and 0.1mM aldrithiol (Sigma-Aldrich) were combined in 0.1M HEPES, 0.05M NaCl, and 10% DMSO (pH 7.4) buffer (100µl final volume) in 1.5µl microcentrifuge tubes. After homogenization, the content of the tubes was transferred to ELISA microplates and 100µl of 2mM Z-1-Lys-SBzL substrate (Sigma-Aldrich) diluted in the same buffer was added to the wells. The plate was immediately read at the kinetic mode at 37°C for 30min at 324nm. Tubes without Lufaxin and factor D were used as positive and negative controls, respectively. Tubes with only Lufaxin, aldrithiol, and the substrate were also used to check if Lufaxin have any effect over the substrate.
ELISA plates (Corning Inc.) were incubated overnight at 4°C with 50µl of coating buffer (35mM Na2CO3, 15mM NaHCO3, pH 9.6) containing 0.5µM of purified C3b, properdin, factor B, Bb, or factor D (CompTech). Wells incubated with 1% BSA in coating buffer were used as negative controls. Following a 1h blocking with 1% BSA in PBS, the wells were incubated with 120nM of Lufaxin in 50µl of PBS for 30min under agitation. The plate was washed twice with 0.05% Tween 20 in PBS and then treated with mouse anti-Lufaxin serum diluted 1:500 in blocking buffer (1% BSA in PBS). After two more washes, the wells received 50µl of blocking buffer containing alkaline phosphatase-conjugated anti-mouse antibody (Sigma-Aldrich) diluted 1:10,000 for 30min. The plate was washed and the reaction was completed by adding 100µl of p-nitrophenyl phosphate substrate for alkaline phosphatase (Sigma-Aldrich). After 20min of incubation at 37°C, the plate was read at 450nm.
C3b was diluted in 10mM sodium acetate buffer (pH 5.0) and immobilized on CM5 sensor chips (GE Healthcare) by the amine-coupling method. Assays were conducted at 25°C on a Biacore T100 instrument (GE Healthcare) by passing different combinations of complement factors and Lufaxin over the C3b immobilized surface. The buffer for these experiments (HBS-N) contained 10mM HEPES, pH 7.4, 150mM NaCl. In most cases, either MgCl2 or NiCl2 was added to the buffer to a concentration of 2mM.
To assay the ability of Lufaxin to inhibit the activation of C3, the C3b-Bb complex was formed in solution by incubating 200nM C3b, 100nM factor B, and 50nM factor D in HBS-N buffer containing 2mM MgCl2 for 2min at room temperature. After stopping the formation of activated complex by the addition of EDTA (5mM), purified C3 was added at various concentrations in the presence or absence of Lufaxin (1µM), and the mixtures were incubated for 20min at room temperature before the addition of SDS-PAGE sample preparation buffer. Inhibition of factor B activation was evaluated by incubating 200nM C3b, 100nM factor B, and 50nM factor D in the presence or absence of 1µM Lufaxin in HBS-N containing 2mM MgCl2 for 30min. The reactions were stopped by adding EDTA to a concentration of 5mM. In both types of experiments, the reaction products were separated by SDS-PAGE and blotted to nitrocellulose. To detect C3 cleavage, blots were incubated with rabbit anti-C3a, and to detect factor B cleavage they were incubated with goat anti-factor B, followed by incubation with the appropriate alkaline phosphatase conjugate antibody. The blots were then developed using Western Blue alkaline phosphatase substrate reagent (Promega).
Inhibition of samples was analyzed using GraphPad Prism 5.0 software. All assays were carried out in duplicate (hemolytic assays) or triplicate (deposition assays) with at least three biological repetitions. Mean of groups was statistically compared using ANOVA followed by Tukey’s test using 95% as significant level (p<0.05).
To identify the AP-complement inhibitor in L. longipalpis saliva, 12 recombinant salivary proteins, chosen from among the most abundant molecules (26), were tested in AP hemolytic assays. Only one sand fly salivary recombinant protein of 32.5kDa, Lufaxin, produced a complete reduction in AP erythrocyte lysis (Figure (Figure1A).1A). Lufaxin inhibited AP erythrocyte lysis in a dose-dependent manner at host physiological pH (7.4) and sand fly midgut pH (8.15) (24) in comparison to the control protein LJS169, another sand fly recombinant salivary protein (Figure (Figure1B).1B). Statistical significance was observed at Lufaxin concentrations of 40nM or above (p<0.0001), confirming its inhibitory effect on the AP.
The possible activity of Lufaxin against other complement pathways was also examined. The CP-mediated hemolytic assay (Figure (Figure1C)1C) and LP-mediated activation (Figure (Figure1D)1D) were not inhibited by Lufaxin or the control protein LJS169.
We tested whether neutralization of Lufaxin by antibodies would prevent blockage of AP-mediated rabbit erythrocyte hemolysis by Lufaxin or sand fly saliva. Anti-Lufaxin serum produced in mice was able to recognize both the recombinant and native Lufaxin (Figure (Figure2A).2A). Recombinant Lufaxin inhibited AP activation in a dose-dependent manner (Figure (Figure2B),2B), and addition of anti-Lufaxin serum to recombinant Lufaxin partially rescued complement activation at a dilution of 1:25, and completely rescued activation at a 1:10 dilution (Figure (Figure2C).2C). In order to investigate if Lufaxin is the only AP inhibitor in L. longipalpis saliva, we tested the effect of anti-Lufaxin serum on L. longipalpis SGH in the AP hemolysis assay. First, we confirmed that SGH also inhibited AP activation in a dose-dependent manner (Figure (Figure2D)2D) and next that addition of anti-Lufaxin serum at a dilution of 1:10 completely rescued AP complement activation (Figure (Figure2E).2E). These results indicate that Lufaxin accounts for most or all AP-inhibitory activity in L. longipalpis saliva.
Agarose-coated microplates were used to measure the deposition of complement components after activation of the AP. In the presence of Lufaxin, deposition of C3b, the first activated component of the AP, was strongly inhibited in a dose-dependent manner (Figure (Figure3A).3A). This correlates with the inhibition of subsequent components of the complement cascade (Bb, properdin, C5b, and C9) by Lufaxin (Figures (Figures3B–E).3B–E). The negative control (LJS169) showed no effect on the deposition of the tested components (Figures (Figures33A–E).
Inhibition of complement activation by Lufaxin was further confirmed by Western blotting of supernatants from the AP-hemolysis assay using anti-factor B and anti-C3a polyclonal antibodies in the presence of Lufaxin or LJS169. Lufaxin blocked factor B cleavage, since no bands (or very weak bands) corresponding to factor Bb or Ba were detected on immunoblots at 30 or 60min after AP complement initiation as compared to treatment with LJS169 (Figure (Figure3F).3F). Similarly, when Lufaxin was added to tubes with serum and red blood cells, a band corresponding to C3a was not observed, indicating that it inhibited the activation of C3, even after 60min at 37°C (Figure (Figure3G).3G). These results strongly suggest that Lufaxin prevents the formation of C3b and Bb, the components of the AP C3-convertase complex.
Hemolysis assays were then performed to determine if Lufaxin acts exclusively at the early steps of the activation cascade or may act on any MAC component. Hemolytic assays were carried out using C6-depleted serum and red blood cells, and, after removal of the supernatant, MAC formation was elicited by addition of NHS diluted in buffer containing EDTA to prevent the formation of additional C3 convertase complexes. When Lufaxin was added with the C6-depleted serum, hemolysis was significantly reduced (p<0.0001), suggesting that Lufaxin impairs C3 convertase formation (Figure (Figure3H).3H). The strong hemolysis seen in tubes without inhibitor, or with the control recombinant LJS169, indicates that the C6-depleted serum was active (Figure (Figure3H).3H). When Lufaxin was added together with the NHS, after C3 convertase assembly, no effect on MAC formation was observed (p=0.7657) (Figure (Figure3I),3I), providing further evidence that Lufaxin acts exclusively on the early steps of the AP, where C3b and Bb are produced.
To determine how Lufaxin impairs C3-convertase formation, we first tested whether it is a direct inhibitor of factor D, the serine protease that cleaves factor B in the AP pathway. Lufaxin, purified factor D, and aldrithiol were mixed and added to factor D substrate Z-1-Lys-SBzL. Similar levels of substrate cleavage were seen in tubes containing factor D with or without Lufaxin, indicating that Lufaxin does not inhibit the enzymatic activity of factor D (Figure (Figure44A).
As Lufaxin does not act on factor D, we then checked which AP complement components may be affected by Lufaxin. ELISA plates were coated with different AP complement components followed by addition of Lufaxin which was able to bind significantly to properdin (p<0.001). Lufaxin was not able to bind significantly (p>0.05) to C3b, factor B, Bb, and factor D when they were alone on the plate surface (Figure (Figure44B).
The role of Lufaxin binding to properdin was assessed by hemolytic assays using properdin-depleted sera. Lufaxin maintained its AP inhibitory activity even in the absence of properdin (Figure (Figure4C),4C), suggesting properdin binding is not essential for AP inhibition.
The ability of Lufaxin to bind the active C3 convertase C3b-Bb, or the proconvertase C3b-B, was evaluated using SPR. In these experiments, factor D, factor B, and Lufaxin were run in different combinations over surfaces of immobilized C3b in the presence of Mg2+. Factor B bound detectably to immobilized C3b, and stability of the complex was significantly increased by adding factor D to the mobile phase (Figures (Figures5A,B),5A,B), indicating the formation of C3b-Bb. Lufaxin alone did not bind C3b (Figures (Figures5A,B),5A,B), but when it was added to factor B, or to the factor B-factor D mixture prior to injection, a large increase in the stability of the complex formed on immobilized C3b was observed (Figure (Figure5A).5A). Under conditions of constant factor B concentration, complex formation was found to depend on the concentration of Lufaxin (Figure (Figure5C).5C). These results indicate that Lufaxin binds to the C3b-B complex and possibly to the C3b-Bb complex. To distinguish these possibilities, the C3b-Bb complex was formed by injection of a mixture of factor B and factor D over a C3b surface (Figure (Figure5B)5B) and, after formation of the complex, Lufaxin was injected alone while the complex slowly dissociated. No significant Lufaxin binding was observed in this case, suggesting that the inhibitor binds only to the C3b-B complex and not to C3b-Bb (Figure (Figure5B).5B). The fact that complex stabilization by Lufaxin was essentially equivalent in the presence and absence of factor D suggests that the inhibitor stabilizes the C3b-B complex, while preventing activation of factor B by factor D (Figure (Figure55A).
Magnesium ion binds at the metal ion-dependent adhesion site (MIDAS) of the VWA domain of factor B and is coordinated by the C-terminus of C3b (27, 28). Replacement of Mg2+ with Ni2+ stabilizes an open form of C3b-bound factor B that is susceptible to cleavage by factor D (29, 30). To gain insight into the mechanism of complex formation we examined the ion specificity of C3b-B-Lufaxin formation using SPR. Binding of factor B to a C3b surface in the presence of Ni2+ was enhanced, and the addition of Lufaxin led to a further increase in the amplitude of the sensorgram (Figure (Figure5D).5D). However, fitting the dissociation phase data to a single exponential decay function indicated that the half-time for C3b-B-Lufaxin dissociation (~400s) was of similar magnitude to that for C3b-B (~600s) suggesting that Lufaxin binds to C3b-B, but does not provide the degree of stabilization in the presence of Ni2+ seen in the presence of Mg2+ (Figures (Figures5A,F,5A,F, t1/2 ~ 135s). Injection of Lufaxin on the same surface after formation of the C3b-B complex (during the dissociation phase) showed an increase in binding to produce a total sensorgram amplitude that was similar to that seen with coincident injection of factor B and Lufaxin (Figure (Figure5E,5E, B+Lufaxin), again suggesting that Lufaxin binds to the complex but does not further stabilize the interaction of factor B with C3b in the presence of Ni2+. Comparison of assembly on the same C3b surface in Ni2+ and Mg2+ buffers revealed a higher level of C3b-B-Lufaxin complex formation in the presence of Ni2+ than Mg2+ (Figure (Figure5F),5F), but comparison of Figures Figures5A,D,E5A,D,E suggests that Lufaxin enhances the binding of factor B to the C3b surface in the presence of Mg2+, but not Ni2+. Perhaps, Lufaxin binds preferentially to the open form of C3b-bound factor B, thereby increasing its abundance in the presence of Mg2+ but not in the presence of Ni2+ where the open form is already favored. When divalent cations were removed completely (HBS-N buffer alone) neither factor B nor Lufaxin bound detectably to the C3b surface, while binding of the factor B-Lufaxin mixture was similar to that seen in the presence of Mg2+ (Figure (Figure5G).5G). This indicates that the presence of Mg2+ or Ni2+ in the MIDAS of the VDW domain of factor B is not essential for formation of a C3b-B-Lufaxin complex and suggests that Lufaxin may stabilize the open form of factor B independently of metal ions.
The effect of Lufaxin on the enzymatic activity of the proconvertase and convertase complexes was tested using reconstituted systems. The C3b-Bb complex was assembled from soluble purified components by activating C3b-bound factor B with factor D in the absence of Lufaxin. The cleavage of C3 by the complex was then evaluated in the presence and absence of Lufaxin using Western blots to detect the release of C3a (Figure (Figure6A).6A). Under these conditions, Lufaxin did not inhibit C3 cleavage, which is consistent with SPR results showing a lack of binding to the C3b-Bb complex (Figure (Figure6A).6A). However, when C3b was mixed with factor B in the presence and absence of Lufaxin prior to the addition of factor D, cleavage of factor B, as indicated by the appearance of Bb on Western blots, was inhibited in the presence of Lufaxin, but not in its absence (Figure (Figure6B).6B). This result is also consistent with SPR data and supports the idea that Lufaxin binds to the C3b-B complex and inhibits the cleavage of factor B by factor D.
Saliva of L. longipalpis has previously been shown to inhibit both the classical and APs of human complement system, suggesting the presence of a single inhibitor of the common pathway, or of multiple anti-complement factors (16). An 11kDa salivary protein, given the name SALO, was recently found to inhibit the classical pathway but not the AP, indicating that a second component must be present (6). In this study, we show that Lufaxin, a 32kDa protein of 278 amino acids, first described as an inhibitor of coagulation factor Xa (4), is also an inhibitor of the AP. The protein shows no sequence similarity to any other known proteins with the exception of homologs found in the saliva of other sand fly species (31). Along with the previously described classical pathway inhibitor SALO, Lufaxin could potentially protect the feeding sand fly from the effects of anaphylatoxins and the tissue damaging effects of complement deposition and MAC assembly in the gut.
Among the arthropod AP-complement inhibitors, the inhibitory mechanism of Lufaxin is novel. Albicin, an AP inhibitor from saliva of the mosquito Anopheles albimanus, also targets the C3 convertase, but binds and inhibits C3b-Bb rather than C3b-B (23). The mechanism is also different from the AP inhibition elicited by other organisms such as ticks (32–34), mites (35), and bacteria (28).
Although the ligation of Lufaxin to properdin may play a role in the AP inhibition, the ligation is not the central mechanism of the Lufaxin AP inhibitory activity since it remained active in properdin depleted sera. Such fact distinguishes its mode of action from well-known AP-inhibition mechanisms involving properdin in saliva of ticks, which bind and remove properdin from the active AP-C3 convertase, reducing its half-life and accelerating its decay (32–34, 36). Inhibition of the AP C5-convertase by Lufaxin was not directly investigated. If present, this effect could increase the AP inhibition by Lufaxin.
Inhibition of the complement cascade at an early point, such as C3 convertase activation, would be beneficial to a blood feeder having a limited quantity of salivary secretion since the concentrations of these complexes are initially low. Additionally, blockage of the cascade before the production of anaphylatoxins (C3a and C5a) would limit any negative effect of these pro-inflammatory molecules while feeding. In the pool-feeding sand flies, this is apparently an adaptive mechanism, since they possess a limited amount of salivary proteins (less than 1μg/gland) (17) and feed in a complex environment where local inflammatory responses are triggered by tissue damage caused by feeding (37–39).
The inhibition of the host complement system has also been shown to prevent complement-mediated damage to the gut of triatomine bugs (18) and mosquitoes (19), thereby improving the utilization of the blood meal. Considering that the pH in the midgut lumen of L. longipalpis becomes alkaline immediately after a blood meal [rising to 8.15 (24)] and that the complement system remains active in the midgut of hematophagous insects (18, 19), specially the AP (40), Lufaxin would need to be active in the physiological conditions where complement meets the midgut cells in order to have a protective effect. It is known that L. longipalpis ingests saliva with the contents of the feeding pool (41). Thus, Lufaxin starts to inhibit the AP at the bite site (on pH 7.4) and inhibition has to continue at the midgut pH (8.15). Such activity corroborates the strong and similar activity detected for Lufaxin in both pH 7.4 and 8.15.
In addition to blood feeding and midgut protection, Lufaxin may be also important for Leishmania transmission since it can protect recently injected promastigotes in host tissues. Both the alternative and classical pathways of the complement system are activated during Leishmania infection (42, 43). Although the parasite possesses its own mechanisms of evasion of the complement system, such as lipophosphoglycans and the metalloprotease gp63 (44), it is known that promastigotes are significantly sensitive to the complement system and must invade host cells rapidly in order to escape it.
Over the last two decades, several studies with rodent models have shown that protection against cutaneous or visceral leishmaniasis can be achieved by previous immunization with sand fly saliva or salivary proteins (45–49), acting as a transmission-blocking vaccine (TBV). Vaccination of dogs and mice with DNA plasmids encoding Lufaxin induced a delayed-type hypersensitivity (21, 50), that is the response mechanism underlying the adverse effects on Leishmania promastigotes (37). The induction of neutralizing antibodies able to reverse the anticoagulant and/or AP-inhibitory activity, at least in part, could increase the efficiency of Lufaxin as a TBV.
In addition to Lufaxin, SALO (the L. longipalpis CP inhibitor) also induced protective anti-Leishmania immunity in hamsters (47, 51). It is conceivable that a vaccine containing both proteins (Lufaxin and SALO) would potentially potentiate the effectiveness of a TBV.
All animal procedures were reviewed and approved by the National Institute of Allergy and Infectious Diseases (NIAID) Animal Care and Use Committee under protocol LMVR4E and handled in accordance to the Guide for the Care and Use of Laboratory Animals and with the NIH OACU ARAC guidelines and also approved by Ethics Committee in Animal Experimentation (CETEA) of Universidade Federal de Minas Gerais under Protocol no. 87/2011.
MP, MS, FO, SK, JR, JA, JV, and RA conceived the study. AM-S, VV, NS, AG-C, JA, JV, and RA designed experiments. AM-S, VV, NS, AG-C, and JA performed experiments. All the authors contributed to analysis and interpretation of data. AM-S, VV, NS, AG-C, JA, JV, and RA wrote the paper. MP, MS, FO, SK, JR, JA, JV, and RA critically revised the paper. All the authors approved the manuscript.
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
The authors thank Van Pham for technical assistance.
Funding. Support for this work was provided in part by the Intramural Research Program at the National Institute of Allergy and Infectious Diseases, NIH (AG-C, FO, SK, JR, JA, and JV) and the Brazilian agencies Fundação de Amparo à Pesquisa do Estado de Minas Gerais (FAPEMIG), Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) and INCT-Entomologia Molecular (AM-S, VV, NS, MP, MS, and RA).