Abstract
Parasite Ag-specific T cell unresponsiveness and diminished IFN-γ production are immunologic hallmarks of patent infection with lymph-dwelling filarial nematodes. Although this diminished responsiveness is directed primarily against the intravascular microfilarial (MF) parasite stage and mediated in part by reduced APC function, the mechanisms involved are not fully understood. In this report, we demonstrate that human dendritic cells (DC) exposed to live MF up-regulate both the cell surface and gene expression of CD54 (ICAM-1). Moreover, live MF result in a 3-fold increase in DC death compared with MF-unexposed DC, primarily due to apoptosis. Notably, microarray and real-time RT-PCR data indicate that live MF concurrently up-regulate mRNA expression of proinflammatory molecules such as IL-8, RANTES, IL-1α, TNF-α, and IL-β in DC, the presence of which is also detected at the protein level, while inhibiting the production of IL-12 (p40 and p70) and IL-10. Soluble excretory-secretory products from live MF diminished IL-12 and IL-10 production and induced DC death, although to a lesser degree. Moreover, exposure of DC to live MF resulted in a decrease in the ability of DC to promote CD4+ T cell production of IFN-γ and IL-5. Our findings clearly suggest that the interaction between live MF and DC is complex but contributes to the hyporesponsiveness and parasite persistence associated with the MF+ state in the infected human. These data further suggest that MF induce an orchestrated response in APC that leads to a diminished capacity to function appropriately, which in turn has significant consequences for CD4+ T cells.
Lymphatic filariasis, a mosquito-borne disease caused by Wuchereria bancrofti, Brugia malayi, or B. timori, is a major cause of morbidity in tropical and subtropical regions. Although there are many clinical outcomes associated with infection with these lymphatic-dwelling filariae, the most intriguing from an immunologic point of view is a subclinical condition associated with the inability of T cells to proliferate or produce IFN-γ in response to parasite Ag (1). This lack of T cell responsiveness has been shown to be primarily directed at Ag produced by microfilariae (MF)2 found in the circulation (2).
Although many helminth infections are associated with responses felt to reflect an augmented or skewed Th2 response (3), filarial infections cannot be so easily explained. For example in MF+ patients, although both Ab responses and the IL-4 response to parasite Ag are usually robust, IL-5 production is variable (reviewed in Ref. 1).
Down-regulation of an Ag-specific proliferative response is a hallmark of several different parasitic infections (4, 5, 6) and may reflect a mechanism by which parasite survival is promoted. The mechanisms underlying T cell hyporesponsiveness vary from organism to organism, but factors such as regulatory cytokines (7), altered function of APC (8, 9, 10), T cell apoptosis (11), inducible NO synthase (5, 12), and pro- and anti-inflammatory cytokines have been implicated to be associated with this defect. We have previously shown that MF Ag impairs the production of IL-12 and IL-10 by dendritic cells (DC) (10). In W. bancrofti infection, it has been shown that PBMC from MF individuals produce large amounts of IL-10 spontaneously and in response to Ag in vitro (13). Of interest, neutralizing Ab to TGF-β or IL-10 (7, 13) enhanced Ag-specific proliferative responses, suggesting a role for these cytokines in modulating the immune response in filarial infection. It also has been shown that, using a mouse model of B. pahangi infection, MF can selectively induce lymphocyte apoptosis (11), a process that also might contribute to proliferative defects and parasite survival.
As DC are the major APC involved in initiating the immune response to parasitic worms and other pathogens, and because immature DC can take up Ag in the periphery, mature, and migrate to secondary lymphoid organs in which they prime Ag-specific T cells (14), we hypothesized that in lymphatic filarial infections, MF must encounter DC at different stages of differentiation as they travel from the afferent lymphatic vessels to the peripheral circulation. Although we and other researchers have investigated the effect of MF Ag on various arms of the immune response (10), in this study we have used physiologically relevant numbers of live parasites to examine their influence on human DC. In so doing, we have shown that exposure of monocyte-derived DC to live MF results in generation of ICAM-1-independent DC-parasite aggregates, a process that induces cell death in DC through apoptosis. These extracellular parasites induce mRNA expression and production of proinflammatory molecules IL-1α, IL-1β, IL-6, TNF-α, IL-8, and RANTES at a time when they also impair the production of IL-12 and IL-10. Most important, this impaired DC function results in a diminished ability of DC to promote CD4+ T cell cytokine production, a finding that reflects what is seen clinically among individuals with patent filarial infections.
Materials and Methods
MF preparation
Live B. malayi MF were provided by Dr. J. McCall (University of Georgia, Athens, GA) as previously described (15). Briefly, live MF were collected by peritoneal lavage of infected jirds and separated from peritoneal cells by Ficoll diatrizoate density centrifugation. The MF were then washed repeatedly in RPMI 1640 with antibiotics and cultured overnight at 37°C in 5% CO2.
In vitro generation of DC
CD14+ peripheral blood-derived monocytes were isolated from leukopacks of healthy donors by counterflow centrifugal elutriation. Monocytes were cryopreserved at 5 × 107/vial and thawed for culture in 6-well tissue culture plates at 2–3 × 106/ml (CoStar, Cambridge, MA) in complete RPMI 1640 (BioWhittaker, Walkersville, MD) supplemented with 20 mM glutamine (BioWhittaker), 2% heat-inactivated human AB serum (Gemini Bio-Products, Woodland, CA), 100 IU/ml penicillin, and 100 g/ml streptomycin (Biofluids, Rockville, MD). Recombinant human IL-4 and recombinant human GM-CSF (PeproTech, Rocky Hill, NJ) were added to the culture at 50 ng/ml on days 1, 4, and 6 of culture. Live MF were added on day 6 at final concentrations of 5,000, 20,000, and 50,000 per well (per 1–2 × 106 DC). For transwell experiments, 3.0-micron pore-size transwells were loaded with media or MF and placed in wells containing DC at day 6 of culture. Cells were harvested at day 8 of culture with Versene/EDTA (Biofluids), washed twice with PBS (without Ca2+/Mg2+), counted by trypan blue exclusion, and used for flow cytometric analysis or other functional studies. DC harvested at day 8 were repeatedly shown to be CD1a+, HLA-DR+, CD86+, CD40+, CD3−, CD14−/low, CD19−, and CD56− by flow cytometry (FACSCalibur; BD Biosciences, Mountain View, CA).
In vitro activation of DC
On day 8 of culture, DC were harvested and cultured at 0.5 × 106/ml in a 48-well tissue culture plate in medium alone or activated with Staphylococcus aureus Cowan I bacteria (SAC) (10 μg/ml) plus IFN-γ (1 ng/ml) (SAC/IFN-γ), or with soluble CD40 ligand (2 μg/ml) plus IFN-γ (CD40L/IFN-γ). Supernatants were collected at 48 h.
CD4+ T cell isolation
Blood was obtained from healthy volunteer blood donors at the National Institutes of Health by apheresis, and lymphocytes were isolated using elutriation. They were washed in PBS and cryopreserved in aliquots. When needed, the cells were thawed and washed. Resting CD4+ T cells were subsequently obtained as described (16) using a mixture of mAb and rigorous immunomagnetic negative selection with BioMag beads (Polysciences, Warrington, PA) that were bound to goat anti-mouse IgG (H plus L chain). The purity of isolated cells was shown by flow cytometry to be >97.0%.
In vitro CD4+ T cell activation
Unexposed or MF-exposed DC were harvested at day 8 and cultured with autologous CD4+ T cells (at a 1:5 DC-T cell ratio) either in medium or with 1 μg/ml of staphylococcal enterotoxin B (SEB; Sigma-Aldrich, St. Louis, MO), 10 μg/ml of anti-CD3, or 10 μg/ml of MF Ag in 48-well tissue culture plates (CoStar). Supernatants were collected at both day 2 and day 5 for cytokine measurement.
Flow cytometry
Staining of cells with Ab was conducted according to standard protocols (17). Propidium iodide (PI) (Sigma-Aldrich) was used to exclude nonviable cells from the analysis. DC (0.2–0.5 × 106) were harvested and washed with FACS medium (HBSS) without phenol red and without Ca2+/Mg2+ (BioWhittaker) containing 0.2% human serum albumin (Sigma-Aldrich) and 0.2% sodium azide (Sigma-Aldrich). Cells were incubated with human gammaglobulin (Sigma-Aldrich) at 10 mg/ml for 10 min at 4°C to inhibit subsequent binding of mAb to Fc receptor. Then cells were incubated with specific mAb conjugated with FITC or PE at saturating concentrations for 30 min at 4°C, washed twice with FACS medium, and analyzed using a FACSCalibur and CellQuest software (BD Biosciences). All Ab used were mouse anti-human mAb and consisted of the following: CD1a-PE (clone VIT6B; Caltag Laboratories, Burlingame, CA); CD11a-FITC (clone MEM25; Caltag Laboratories); CD11b-FITC (clone CR3; Caltag Laboratories); CD11c-PE (clone 3.9; Caltag Laboratories); CD14-FITC (clone Tuk4; Caltag Laboratories); CD40-FITC (clone 14G7; Caltag Laboratories); CD54-FITC (ICAM-1) (clone MEM111; Caltag Laboratories); CD58-FITC (clone IC3; BD PharMingen, San Diego, CA); CD80 (B7-1)-FITC (clone L307.4; BD PharMingen); CD86 (B7-2)-FITC (clone 2331; BD PharMingen); CD83-PE (clone HB15e, BD PharMingen); CD95 (Fas)-PE (clone DX2; Caltag Laboratories); CD95 ligand (FasL)-PE (clone Alf2.1; Caltag Laboratories); HLA-A, -B, -C-FITC (clone G46-2.6; BD PharMingen); and HLA-DR-FITC (clone L243; BD PharMingen).
Cytokine assays
TUNEL
Apoptotic cells were detected by FITC-dUTP end-labeling of fragmented DNA and catalyzed by TdT followed by flow cytometry, according to manufacturer’s instructions for the APO-Direct kit (BD PharMingen).
RNA preparation
DC were cultured in media alone or were exposed at day 6 to different numbers of live MF for 6, 24, or 48 h, after which the cells were harvested and total RNA prepared using the RNAeasy mini kit (Qiagen, Valencia, CA) from four independent donors and pooled together to make cDNA for real-time RT-PCR and continued to prepare cRNA for microarray. This RNA was used for both microarray hybridization and real-time RT-PCR (ABI 7700; Applied Biosystems, Fullerton, CA).
Microarray analysis
Total RNA was used to generate cRNA probes. Preparation of cRNA, hybridization, and scanning of the HU95 arrays were performed according to the manufacturer’s protocol (Affymetrix, Santa Clara, CA). Briefly, 12–15 μg RNA was converted into double-stranded cDNA by reverse transcription using a cDNA synthesis kit (SuperScript Choice; Life Technologies, Gaithersburg, MD) with an oligo(dT)24 primer containing a T7 RNA polymerase promoter site added 3′ of poly(T) (Genset Oligos, La Jolla, CA). After second-strand synthesis, labeled cRNA was generated from the cDNA sample by an in vitro transcription reaction supplemented with the Bioarray HighYield RNA transcription labeling kit (Enzo Diagnostics, Farmingdale, NY). The labeled cRNA was purified using RNAeasy spin columns (Qiagen). The labeled cRNA samples were fragmented at 94°C before hybridization. Labeled cRNA was hybridized to the HU95A microarray while rotating at 60 rpm for ∼16 h at 45°C. After hybridization, the microarray was washed using the Affymetrix Fluidics Station in buffer containing biotinylated anti-streptavidin Ab (10 min, 25°C; Vector Laboratories, Burlingame, CA) and stained with streptavidin-PE (final concentration 10 μg/ml; Molecular Probes, Eugene, OR) for 10 min at 25°C. Subsequently, the microarray was washed, restained with streptavidin-PE (10 min, 25°C), and washed again before measuring the fluorescence bound to the microarray at 570 nm in an Affymetrix scanner.
Microarray data processing
Images of scanned Affymetrix GeneChips were processed using the software and parameter settings suggested by the manufacturer (Affymetrix). The processed fluorescence values were placed into an Excel table (Microsoft Excel Analysis Tools, Seattle, WA) in which rows represented genes and columns represented experimental conditions. We performed all data filtering on a Microsoft Excel spreadsheet. We chose a filtering strategy based on present call and fold change of ≥3 for up-regulation and ≤3 for down-regulation of the gene. Under all conditions, MF-exposed values were compared with those of unexposed DC at the given time point. When filtering was complete, the results were clustered using a web-based hierarchic clustering program (18). For categorizing genes, a web-based program (http://www.libgenechip.niaid.nih.gov; Laboratory of Immunopathogenesis and Bioinformatics, National Institute of Allergy and Infectious Diseases, National Institutes of Health) was used.
Real-time RT-PCR (TaqMan)
TaqMan cytokine gene expression plates (Applied Biosystems) were used to detect TNF-α, IL-1α, and IL-1β. For IL-8 and RANTES, separate probes and primers were used. Probes, reagents, and equipment were used as recommended by the manufacturer (Applied Biosystems).
Microscopy
The images of DC and live MF were obtained using Olympus DP10, version 3.0, at ×20 magnification.
Statistical analysis
The nonparametric Wilcoxon signed rank test was used throughout. All statistical analyses were performed with StatView 5 (SAS Institute, Cary, NC).
Results
Live MF up-regulate expression of CD54 (ICAM-1) on DC
To assess the effect of live MF on the expression of cell surface molecules that define DC or their function, DC generated in vitro from elutriated monocytes (in the presence of human GM-CSF and human IL-4 for 8 days) were exposed to live MF at either 50,000 (high), 20,000, or 5,000 (low) per 1–2 × 106 DC during the last 48 h of culture (day 6). The number of live MF was chosen based on the range of MF in clinical settings. After MF exposure, DC were harvested and flow cytometry analysis performed on the viable cells for expression of cell surface molecules.
As expected, monocyte-derived DC were CD14low and had up-regulated CD1a expression, a phenotype typical of immature DC; this phenotype was unaffected by the presence of live MF. Expression of MHC class I and class II molecules was not altered by the presence of parasites (data not shown). In addition, exposure to live MF did not alter the expression of costimulatory molecules CD80, CD86, CD11a, CD11b, CD11c, CD58, or CD40 (data not shown).
However, our data demonstrate that among all the cell surface markers tested, DC exposed to high numbers (50,000) of live MF showed an increase in surface expression of only one adhesion/costimulatory molecule: ICAM-1 (Fig. 1⇓). As shown in Fig. 1⇓b, there was a 1.5- to 2.8-fold increase in the mean fluorescent intensity of ICAM-1 in live MF-exposed DC as compared with unexposed cells (Fig. 1⇓b).
CD54 (ICAM-1) expression is up-regulated by live MF. Flow cytometric analysis of unexposed DC or DC exposed to different numbers of live MF for 48 h. a, Dotted lines, negative control; dashed lines, DC alone; solid lines, DC with 5,000 MF; dark solid line, DC with 50,000 MF. Results shown are from one experiment representative of four independent experiments. b, Mean fluorescent intensity of CD54 expression of four independent donors as a function of MF numbers.
DC exposure to live MF result in aggregate formation
When DC were exposed to live MF for 48 h, large aggregates between the parasite and cells were observed compared with unexposed DC (Fig. 2⇓, a and b). The number of aggregates increased as the number of MF increased. This aggregate formation was not seen when live MF were physically separated from DC using 3.0-micron transwells (Fig. 2⇓c), suggesting involvement of a direct cell-parasite contact in formation of these aggregates. Because ICAM-1 was the only surface molecule up-regulated by live MF in DC and because ICAM-1 plays a key role in inflammation by initiating leukocyte trafficking, its effect on DC-MF aggregation was assessed. ICAM-1 did not appear to play a role in MF-DC parasite aggregates, as neither neutralizing anti-ICAM-1 Ab (Fig. 2⇓d) nor anti-CD18 (data not shown) interfered with aggregate formation.
DC aggregate formation by live MF. Aggregate formation shown 48 h following exposure of DC to 50,000 live MF. a, Unexposed DC. b, DC exposed to 50,000 live MF/well. c, DC exposed to 50,000 MF across 3.0-micron transwells. d, DC exposed to 50,000 live MF in the presence of neutralizing anti-ICAM-1 Ab.
Live MF induce dose-dependent DC death
To assess the effect of live MF on DC survival, DC were harvested and counted by trypan blue exclusion or stained with PI to measure cell viability following 48 h of exposure to different numbers of live MF. MF-exposed DC showed a significant reduction in cell survival as detected by trypan blue exclusion (Fig. 3⇓a). This induction of cell death was shown to be dose dependent, with 50–60% loss of cell viability with high doses of MF (20,000 and 50,000) and 25% loss with the lowest dose (5,000). Furthermore, exposure of DC to high numbers (50,000) of live MF resulted in a 3-fold increase in PI+ cells compared with MF-unexposed DC (p = 0.01; Fig. 3⇓b). To further investigate whether the induced cell death in DC required contact between the parasite and DC, DC were separated from MF using transwells such that DC were cultured in the wells and live MF were placed in the 3.0-micron transwells. Our data indicate that in cases of physical separation between MF and DC, high numbers (50,000) of MF induced DC death (1.4-fold; Fig. 3⇓c), an affect that was less profound than that seen when DC were in contact with MF (2.0-fold; Fig. 3⇓c). These data collectively suggest that cell contact, while contributory, is not required for MF impairment of DC viability and that excretory/secretory products from live MF can mediate cell death in DC.
Live MF induce cell death in DC. DC exposed to different doses of live MF for 48 h were assessed by trypan blue exclusion (a) and by PI staining (b) as a function of MF numbers. c, Cell viability of DC exposed to 50,000 MF across 3.0-micron transwells (DC/MF transwell) compared with unexposed DC (DC) and DC in direct contact with high dose of live MF (DC/MF). Results shown are means + SE of nine independent experiments for data (a) and (b) and of four to six independent experiments (c).
Live MF induce DC apoptosis
To examine the process by which MF induced DC death, the TUNEL method was used to detect apoptosis. DC were exposed to high (50,000) and low (5,000) numbers of live MF for 24 h, then harvested and processed for TUNEL (Fig. 4⇓). As shown, following exposure of DC to low numbers (5,000) of live MF, 9.0% of the cells underwent apoptosis as compared with 2.5% apoptotic cells in unexposed DC. Of interest, with high numbers (50,000) of live MF, 37% of the cells underwent apoptosis, a 15-fold increase in apoptotic cells compared with the MF-unexposed cells (2.5% apoptosis). Several mechanisms may be involved in induction of apoptosis in DC by live MF. Exposure of DC to live MF did not result in up-regulation of Fas or Fas ligand (data not shown). It is also less likely that apoptosis is due to production of NO by DC, as NO was not detected in the supernatant of MF-exposed cells (data not shown).
Live MF induce DC apoptosis. Apoptosis detected by TUNEL-FITC labeling of DC alone and DC exposed to 5,000 or 50,000 live MF for 24 h. Negative control is DC labeled with TUNEL-FITC without TdT. Data shown for one of three representative donors.
Live MF induce expression of both proinflammatory and proapoptotic genes in DC
Having shown that live MF up-regulate ICAM-1 expression in DC and induce DC death, we assessed more globally the effect of live MF on DC using microarray analysis. DC were exposed to live MF at either 50,000 (Fig. 5⇓) or 5,000 (data not shown) for 6, 24, or 48 h. Hierarchic clustering algorithms were performed based on our filtering criteria (see Materials and Methods). Of 12,000 genes expressed on the U95A chip, MF-unexposed DC showed an almost consistent basal expression of ∼5100 genes and lack of expression of 5080 genes across 6, 24, and 48 h (data not shown and Fig. 5⇓). Of these 12,000 genes, 191 genes were altered after 6 h of DC exposure to high doses of MF (107 induced, 84 repressed). After 24 h of exposure to live MF, 54 genes were induced and 27 repressed. Finally, of the 82 genes regulated after 48 h of exposure to high MF, 49 were induced and 33 were repressed in DC (data not shown).
Hierarchic clustering of induced (red) and repressed (green) genes in DC following 6, 24, or 48 h exposure to 50,000 live MF. Basal expression of the clustered genes is shown in blue. Numbers represent fold up-regulation of genes in MF-exposed DC as compared with unexposed DC.
We utilized a website tool based on gene ontologies to categorize the altered genes according to their biologic function (see Materials and Methods). Of interest, the cell surface up-regulation of ICAM-1 already demonstrated by flow cytometry in Fig. 1⇑ was further confirmed at the RNA level by microarray. The mRNA expression of ICAM-1 was up-regulated by 3-fold after 6 h of exposure to high doses of live MF and declined to basal levels by 48 h of exposure (Fig. 5⇑). We further analyzed the genes involved in apoptosis. Among the genes up-regulated, IL-1β, TRAIL, CXCR4, and TNF ligand 14 have been associated with programmed cell death. IL-1β was induced 15-fold at 6 h, an induction that was maintained at 24 h but had declined to basal expression by 48 h (Fig. 5⇑). In addition, both TRAIL and CXCR4 were up-regulated at 6 h, declining to their steady-state level at 24 h.
Other genes up-regulated by high numbers of live MF were involved in inflammatory processes. Proinflammatory genes involved in chemotaxis, such as IL-8, IL-6, MIP-3α, MIP-1β, MCP1, and GRO-β, were up-regulated after 6 h of MF exposure, while RANTES was induced at 24 h.
Live MF up-regulate expression of IL-8, RANTES, TNF-α, IL-1α, and IL-1β
Having shown that live MF up-regulate the gene expression profile of proinflammatory genes in DC, we used real-time quantitative RT-PCR to directly confirm these findings. DC were first exposed to live MF at high or low numbers for 6, 24, and 48 h; at each time point, supernatants were collected and RNA prepared. As shown in Fig. 6⇓a, live MF up-regulated the expression of RANTES, TNF-α, IL-8, IL-1α, and IL-1β in a dose-dependent manner, with high numbers of MF having the most profound effect. Notably, gene expression of TNF-α, IL-8, IL-1α, and IL-1β was detectable as early as 6 h, peaked at 24 h (TNF-α, IL-8, and IL-1β), and declined to almost basal levels by 48 h (TNF-α). Up-regulation of the products of these proinflammatory genes could also be seen, with high numbers of live MF shown to be capable of inducing IL-8 protein by 10- to 18-fold at both 24 and 48 h (Fig. 6⇓b) and production of RANTES at 48 h (4-fold induction).
Live MF up-regulate mRNA expression and production of IL-8, RANTES, TNF-α, IL-1α, and IL-1β. a, mRNA expression in DC exposed to 5,000 (▦) and 50,000 (▪) live MF shown as fold change from unexposed DC using real-time RT-PCR relative to ribosomal RNA (∗ > 100-fold). b, Cytokine production from unexposed DC (□) and those exposed to 5,000 (▦) or 50,000 (▪) live MF for 6, 24, or 48 h. Results shown are the mean + SE of four independent experiments. ND, not done.
Because inflammatory responses seen in animal models of Brugia infection have been associated with a release of LPS-like molecules from the endosymbiont Wolbachia following the death of B. malayi (19), we verified that the proinflammatory response induced by live MF was not due to released LPS. When LPS was blocked using bacterial/permeability increasing protein (BPi) in cultures of live MF-exposed DC, production of IL-8 and TNF-α was not significantly inhibited compared with cultures of live MF-exposed DC without BPi (data not shown).
Live MF suppress the ability of DC to produce IL-12 and IL-10
To evaluate the effect of live MF on DC function, production of IL-12 and IL-10 was measured following activation. Unexposed DC or DC exposed to different doses of live MF were harvested, normalized for the number of viable cells, and cultured in medium alone or with SAC/IFN-γ or CD40L/IFN-γ for an additional 48 h (Fig. 7⇓). As seen, when DC were exposed to high doses (20,000 or 50,000) of live MF, production of IL-12 (p40 and p70) and IL-10 was significantly inhibited (80–90% inhibition; p < 0.05) in response to stimulation by SAC/IFN-γ. In response to CD40L/IFN-γ, DC production of IL-12 p40 (65% inhibition) and IL-10 (80–90% inhibition; p < 0.05) was also significantly reduced. Although we were unable to detect significant production of IL-12 p70 in response to CD40L/IFN-γ in MF-exposed DC, the trend was similar, with higher doses of MF resulting in greater down-regulation. This cytokine inhibition was independent of cell death, as equal numbers of viable DC were used for activation in both MF-exposed and unexposed DC. There were no changes in the number of viable cells between exposed and unexposed DC following activation with SAC/IFN-γ or CD40L/IFN-γ (data not shown). Furthermore, this cytokine inhibition was not due to LPS activity of the live worms, as an LPS blocker such as BPi did not reverse inhibition of IL-12 or IL-10 in MF-exposed cultures (data not shown). Of note, BPi inhibited the production of IL-8, TNF-α, and IL-1β by LPS-exposed DC, confirming that BPi can block LPS activity in our system. In addition, this cytokine inhibition may have been specific for IL-12 and IL-10, because production of other cytokines such as IL-8 was not significantly inhibited by MF following exposure of DC to SAC/IFN-γ or CD40L/IFN-γ (data not shown).
DC exposed to live MF produce significantly less IL-12 p40, IL-12 p70, and IL-10. IL-12 p40 (a), IL-12 p70 (b), and IL-10 (c) production from unexposed DC (□) and those exposed to 5,000 (▦), 20,000 (dark gray bars), or 50,000 (▪) live MF for 48 h following activation with SAC and IFN-γ or with CD40 ligand and IFN-γ. Results shown are the mean and error bars represent + SE of eight independent experiments. ∗, p = 0.02; ∗∗, p = 0.03.
Live MF diminish the ability of DC to induce CD4+ T cell production of IFN-γ and IL-5
Having demonstrated that live MF impair the function of DC to produce IL-12 and IL-10 in response to several different stimuli, we next assessed whether the altered function of DC translates into diminished capacity to induce CD4+ T cell activation and cytokine production. To this end, autologous CD4+ T cells were cultured with MF-exposed or unexposed DC. In the presence of either anti-CD3 or SEB, CD4+ T cells activated with MF-exposed DC had the capacity to expand in culture to the same level as those activated with unexposed DC (data not shown); however, CD4+ T cells cultured with MF-exposed DC had a significantly diminished capacity to produce IFN-γ in response either to SEB (geometric mean, 23% inhibition; p = 0.043) or to anti-CD3 (geometric mean, 40% inhibition; p = 0.043). Of interest, IL-5 production was also diminished but to a lesser degree than that of IFN-γ in MF-exposed DC compared with unexposed DC, with GM inhibitions of 25% in response to anti-CD3 (Fig. 8⇓, a and b). This down-regulation was shown to be specific for IFN-γ under all activation conditions and in response to anti-CD3 and Mf Ag to IL-5, as the production of other cytokines such as IL-10 or IL-3 was not altered in response to any of the activation conditions. Although CD4+ T cells produced small amounts of IFN-γ and IL-5 in response to MF Ag (likely due to the extremely low MF-specific T cell precursor frequency), the production of both of these cytokines was significantly inhibited (geometric mean, 53% inhibition for IFN-γ, p = 0.043; geometric mean, 63% inhibition for IL-5) in the presence of MF-exposed DC as compared with unexposed DC (Fig. 8⇓, a and b). No IL-10 or IL-13 was detected by CD4+ T cells in response to MF Ag (data not shown).
DC exposed to live MF have a diminished capacity to induce CD4+ T cell production of IFN-γ and IL-5. a, Production of IFN-γ and IL-5 from CD4+ T cells cultured with autologous, unexposed DC and those exposed to 50,000 live MF 48 h following activation with anti-CD3, SEB, or MF Ag in five independent donors. b, Geometric mean percentage inhibition of IFN-γ, IL-5, IL-10, and IL-13 production. Results shown are the geometric mean of five to six independent donors.
Discussion
Among the many explanations for the down-regulated parasite-specific T cell response seen in humans with filarial parasite infections, the least well investigated has been the role of the APC. Although attention has turned to the APC in several animal models of filarial infection (8, 9), the function of human DC in response to the filarial parasite has not been fully investigated. We have previously used crude MF Ag to begin to study APC-parasite interaction (10); however, to simulate physiologic conditions more precisely, we have used live parasites to assess the effect of the blood-borne stage (MF) of B. malayi on human monocyte-derived DC.
In the present study, we have thus established that live MF interact with human monocyte-derived DC to form cell-parasite aggregates, induce DC apoptosis, and impair their ability to produce IL-12 and, in turn, limit T cell activation and proliferation.
Granuloma formation is one of the primary pathologic host immune responses seen in many infectious diseases including Brugian filariasis (20). Trafficking of cells from the bloodstream and their firm adhesion to each other is necessary for the organ-like growth of a granuloma designed to destroy an infectious agent. Furthermore, to recruit immune cells to the site of inflammation there must be orchestrated responses that utilize adhesion molecules, cytokines, and chemokines (21). In Schistosoma mansoni, granuloma formation requires the expression of adhesion molecules such as ICAM-1 (22). Moreover, in a model of eye disease in onchocerciasis (“river blindness”), it has been demonstrated that recruitment of eosinophils to the cornea is mediated by ICAM-1 expression on limbal vessels (23). In the present study, we have shown that live MF from lymphatic-dwelling filarial parasites results in at least 2-fold induction of DC surface ICAM-1 expression (Fig. 1⇑). In addition, our microarray data indicate induction of ICAM-1 mRNA as early as 6 h following exposure of DC to live MF (Fig. 5⇑). Although live MF induce cell surface and gene expression of ICAM-1, this molecule did not appear to play a direct role in this DC-MF aggregate formation, nor can it be implicated in granuloma formation in vivo.
Aggregate formation is followed closely by an induction of DC death caused by MF. As measured by trypan blue exclusion and/or PI staining, the number of dead DC increased as the number of live MF in culture increased (Fig. 3⇑). When MF were physically separated from the DC, there was a diminished induction of cell death, suggesting that soluble products released by the parasite do induce cell death but to a lesser degree than when contact is allowed. This induced cell death is, in part, due to apoptosis, as 40% of the cells underwent apoptosis after 24-h exposure of DC to live MF. Although it has been demonstrated that NO promotes apoptotic pathways in numerous cell types through indirect activation of caspases (24), we were unable to detect NO in culture supernatant of MF-exposed DC (data not shown). Although MF did not alter Fas or FasL expression on DC cell surface, Fas-induced apoptosis in MF-exposed DC has not been definitely excluded, even though Fas mediation is not a major feature of DC apoptosis (25). Microarray and real-time RT-PCR data revealed the induction of several genes known to be involved in apoptosis. Indeed, TRAIL, IL-1β, and TNF-α with IL-1β were induced as early as 6 h after exposure of DC to the parasite (Figs. 5⇑ and 6⇑); however, a causal relationship between these genes and the apoptosis observed in the present study has not been shown. Apoptosis induced in filarial infection has found precedent in both murine and human models. In a murine model of Brugia infection, it has been shown that, in response to in vitro restimulation with filarial Ag, CD4+ T cells from MF-infected mice undergo apoptosis (11). Furthermore, it has been shown that filarial sheath proteins induce apoptosis in the human epithelial cell line Hep2 through an apoptotic pathway that can be inhibited through the overexpression of bcl2 (26).
Aggregate formation and ICAM-1 up-regulation may be considered direct signs of inflammation induced by live MF. To investigate whether these changes in DC reflected in up-regulation of proinflammatory genes, we performed both microarray and real-time RT-PCR studies. We found that live MF induced gene expression of chemokines and cytokines (IL-8, IL-6, IL-1α, IL-1β, MIP-1β, MIP-3α, and GRO-β) involved in inflammatory pathways (Fig. 5⇑). Up-regulation of these proinflammatory mediators was further confirmed by ELISA for protein expression and was shown to be dependent on the numbers of parasites, with higher numbers having the most profound effect (Fig. 6⇑b). Because it has been shown that intracellular bacteria (Wolbachiae) present in B. malayi released from worm debris can stimulate macrophages to produce proinflammatory cytokines (27), the potential contribution of this intracellular, rickettsial-like endosymbiont was assessed. The response induced by MF was not caused by LPS or LPS-like molecules, because a strong endotoxin inhibitor, rBPi, neither blocked production of IL-8 and TNF-α in the culture supernatant of MF-exposed DC (data not shown) nor reversed DC cell death (as measured by PI staining and trypan blue exclusion; data not shown). Furthermore, LPS has been shown to induce DC secretion of IL-6 and IL-12 (28), an effect distinct from MF-exposed DC, which were unable to produce IL-12 when left unstimulated (Fig. 6⇑, a and b).
Pathology in lymphatic filariasis can be associated with acute inflammation (29). Infection of immunodeficient mice with Brugia species results in development of lymphedema in the absence of T cells (30) related to production of proinflammatory cytokines (e.g., IL-1α, IL-6, TNF-α, and GM-CSF) in the lymphatic vessels (31). B. malayi contains a homologue of human macrophage inhibitory factor (32) that has also been shown to have chemotactic activity for human monocyte/macrophages and can activate them to produce IL-8 and TNF-α (33). In addition, filarial parasitic sheath proteins have been shown to induce apoptosis in Hep2 cells, a process that results in secretion of IL-6 and IL-8 by this cell line (26).
The importance of IL-8 and RANTES produced by DC after exposure to live MF in attracting neutrophils and T cells, respectively, remains to be investigated. The IL-8 produced by MF-exposed DC was not responsible for DC cell death, as neutralizing anti-IL-8 Ab did not reverse DC cell death (data not shown). Moreover, induction of IL-8 in DC may be unique to the MF stage of the parasite, because the infective stage (L3) of B. malayi diminished the expression of this chemokine in human Langerhans’ cells (R. T. Semnani, manuscript in preparation).
DC are particularly important in priming naive CD4+ T cells and in the development of Th1 response through their production of IL-12. IL-12 can be induced either by bacterial stimuli (34, 35, 36, 37) or by CD40 ligation (34, 35, 38). Although CD40 triggering alone is sufficient to induce production of the p40 subunit of IL-12, induction of biologically active IL-12 p70 requires an additional signal that may be provided by IFN-γ (39). In the present study, we demonstrated that the interaction between DC and live MF 5 days following DC differentiation results in a significant suppression in the production of IL-12 and IL-10 in response to either SAC/IFN-γ or CD40L/IFN-γ. This cytokine inhibition was independent of the increased DC death by MF, as only viable cells were activated to measure of cytokine production. Moreover, this lack of IL-12/IL-10 production was not due to induction of further cell death in MF-exposed DC with either SAC/IFN-γ or CD40L/IFN-γ, as there were equivalent numbers of viable cells under both MF-exposed and unexposed conditions.
It has been reported that a phosphorylcholine-containing glycoprotein, ES-62, secreted by a filarial nematode, has also been shown to induce maturation of the so-called DC2 (9). This ES-62-exposed DC produced significantly less IL-12 but not less IL-10 than did LPS-generated DC following CD4+ T cell interaction. This differs from the results shown in the present study with live MF (and the previous study using MF Ag (10)), which showed a general inhibition in cytokine production associated with a reduced capacity to induce an allogeneic MLR (10). These differences may be attributed to both the source of Ag and the nature of the host cells.
Obviously, the consequence(s) of this MF-induced impairment of DC function on T cell activation and T cell cytokine production becomes the next step for investigation. MF-exposed DC could undergo further cell death after interaction with T cells. Because both MF-exposed and unexposed DC were capable of activating and expanding autologous CD4+ T cells in the presence of anti-CD3, SEB, or MF Ag (data not shown), it is unlikely that there is an additional increase in MF-exposed DC death. Our data indicated, however, that there was a marked reduction in the T cell production of IFN-γ in MF-exposed DC compared with unexposed DC following stimulation of the T cells with either anti-CD3 or SEB (Fig. 8⇑). Regardless of the fact that both anti-CD3 and SEB are strong initial signals, there was a significant difference between MF-exposed and unexposed DC in prompting the production of IFN-γ and, to a smaller extent, IL-5 (Th1 and Th2 surrogates). MF Ag, however, only induced the production of these cytokines by CD4+ T cells to a small degree, possibly because of the known extremely low MF-specific naive T cell precursor frequency (range 1/87,000 to 1/200,000) (40). Nevertheless, in response to MF Ag, there was an enhanced down-regulation of IL-5 and a significant inhibition of IFN-γ (GM 53% inhibition, p = 0.043) by CD4+ T cells when MF-exposed DC were used as APC compared with unexposed DC. Of interest, in response to SEB or anti-CD3, CD4+ T cell production of IL-10 and IL-13 did not significantly change when MF-exposed DC were used as APC, suggesting that this was not a general suppression in the function of CD4+ T cells but rather a specific down-regulation of IFN-γ and IL-5 by these cells. This in vitro down-regulation of IFN-γ and IL-5 by CD4+ T cells that we show in this report is of particular importance, as it closely follows what has been found ex vivo in MF+ patients (41). Our data find support in patient studies where it has been shown that in areas in which Brugian filariasis is endemic, both IFN-γ and IL-5 were suppressed in microfilaremic carriers but IL-4 was not changed (41). It has also been reported that MF carriers have poor Th1-type responses to filarial Ag, with lowered IFN-γ secretion and proliferation, while some (IgG4, IL-4) but not all (IL-5, IgE) Th2-type surrogates remain intact (42). Hence, the MF state cannot easily be explained by a simple Th1/Th2 dichotomy.
Our data indicate that although MF-exposed DC are capable of activating CD4+ T cells to proliferate and expand to the same level as unexposed DC, the mechanism underlying their ability to induce effector function may be impaired. This finding seen using both polyclonal activators (e.g., anti-CD3) and MF Ag suggests that some signaling events may be different in T cells after interacting with MF-exposed DC. Preliminary data using another Ag that requires Ag processing (tetanus toxoid) parallels that seen with anti-CD3 and MF Ag. Indeed, we are in the process of investigating the differences in signaling events in DC after exposure to MF as well as investigating the processing and presentation of multiple soluble Ag to T cells. In addition, we are examining the differences in T cell function after interaction with MF-exposed DC and how exogenous cytokines (e.g., IL-12 or IL-10) might overcome the T cell dysfunction.
Our data suggest, therefore, that MF of B. malayi encountering monocyte-derived DC generally impair viability and function of these cells. Concurrently, MF are capable of inducing production of proinflammatory mediators that may attract other cell types to the site of infection. It is the APC defect, however, that may provide a strong inhibitory cue that leads to the down-regulated Ag-specific T cell response that is the hallmark of chronic infection with lymphatic-dwelling filarial parasites.
Acknowledgments
We thank Richard Lempicki and Jun Young for microarray hybridization, Damien Chaussabel for help in microarray analysis, Shaden Kamhawi for help in microscopy, and Brenda Rae Marshall for editorial assistance.
Footnotes
-
↵1 Address correspondence and reprint requests to Dr. Roshanak Tolouei Semnani, Laboratory of Parasitic Diseases, National Institute of Allergy and Infectious Diseases, National Institutes of Health, 4 Center Drive, Building 4-Room 126, Bethesda, MD 20892. E-mail address: rsemnani{at}niaid.nih.gov
-
↵2 Abbreviations used in this paper: MF, microfilariae; BPi, bacterial/permeability increasing protein; DC, dendritic cell; PI, propidium iodide; SAC, Staphylococcus aureus Cowan I bacteria; SEB, staphylococcal enterotoxin B.
- Received February 14, 2003.
- Accepted May 21, 2003.
- Copyright © 2003 by The American Association of Immunologists
References
In this issue
