Published online before print August 3, 2007
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
* Department of Pathology and Laboratory Medicine, University of Texas Health Sciences Center, and
Institute of Molecular Medicine, University of Texas, Houston, Texas, USA
1 Correspondence: Department of Pathology and Laboratory Medicine, UTHSC, 6431 Fannin, Houston, TX 77030, USA. E-mail: chinnaswamy.jagannath{at}uth.tmc.edu
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
-secreting Th1 T cells. Coincident with increased IL-12p70 levels, BCG-primed C5+/+ DCs cocultured with naive or immune C5+/+ T cells showed a larger increase in CD4+ IFN-/CD8+ IFN-+ T cells compared with cocultured DCs and T cells from C5–/– mice. Thus, BCG-primed C5+/+ DCs were better able to drive a Th1 response. Furthermore, BCG aerosol-infected C5–/– mice showed reduced CD4 and CD8 IFN--secreting T cells in the lungs, concurrent with an increased growth of BCG. Thus, C5a, an innate peptide, appears to play an important role in the generation of acquired immune responses in mice by regulating the Th1 response through modulation of IL-12p70 secretion from DCs.
Key Words: congenic C5 mice • IL-12 • IL-10 • IFN- • tuberculosis
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
-secreting CD4+ or CD8+ T cells, which activate macrophages to kill multiple infectious organisms. Indeed, it is well known that Th1 immunity is protective against many intracellular infections including tuberculosis [3
]. It has been proposed that a Th1-to-Th2 shift can increase susceptibility of humans to tuberculosis [4
]. Recent evidence suggests that such phenomenon may also play a role in determining the vaccine efficacy against tuberculosis [5
]. IL-10-secreting T cells representing Th3 immunity are of relatively recent interest and have been proposed to play a regulatory role [6
]. DCs therefore play a pivotal role in the generation of immune responses, which decide the ultimate fate of infections.
A key feature of the DC-induced Th1 immunity is the production of IL-12p70 from DCs upon phagocytosis of pathogens. Usually, induced by internalization of pathogens through TLRs, secreted IL-12 induces the naive T cells to differentiate into IFN--secreting Th1 T cells [7
]. Furthermore, IL-12 can feedback-activate DCs to produce more IL-12 [8
]. It is well established that decreased IL-12 secretion can lead to a decreased Th1 response and a dominance of a Th2 or Th3 T cell response [9
]. Curiously, IL-12 as well as IL-23 can play a regulatory role on murine DCs themselves and modulate their ability to prime and sensitize T cells [10
]. Mechanisms which regulate IL-12 production in DCs are therefore important during the development of Th1 immunity through T cell- and DC-dependent pathways [11
].
We described previously that complement C5-derived C5a anaphylatoxin is a regulator of IL-12 response in murine macrophages [12 ], and macrophages of mice deficient in C5 (C5–/–), were found to have reduced levels of IL-12 by two groups almost simultaneously, although different models were used to suggest the link [12 , 13 ]. It is interesting that C5–/– mice, such as B.10-congenic mice, A/J, DBA/2, and the SWR strains, have been found to be more susceptible to tuberculosis [14 15 16 ]. This observation suggested a link between C5 deficiency, decreased IL-12 in some mouse strains (A/J, congenic C5), and enhanced susceptibility to tuberculosis. As Th1 immunity can control murine tuberculosis, we sought to investigate the mechanisms of induction of Th1 immunity. Unlike macrophages, DCs can determine the initial immune response in mice after they engulf mycobacteria [17 ]. Recent studies have implicated C5a as a major modulator of DC–T cell interactions in mouse models during allergy [18 , 19 ]. However, relatively little is known about the role of C5a in regulation of murine DC function during mycobacterial infection, where a shift between Th1 and Th2 immunity seems to be of emerging importance. As IL-12 plays a key role in defense against mycobacterial infections, in this study, we have used the C5-congenic mice to understand the regulatory relationships among C5a, IL-12, and development of Th1 immunity. We report here a novel observation that decreased IL-12 secretion from DCs in C5–/– mice affects the ability of DCs to drive the Th1 response in mice.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Mice
Complement C5–/– (B10.D2-Hc°H2dH2-T18c/oSnJ) and complement C5-sufficient, C5+/+ (B10.D2-Hc1H2dH2-T18c/nSnJ), mice were purchased from The Jackson Laboratory (Bar Harbor, ME, USA). They were maintained under BSL-3 conditions with food and water according to approved Institutional Animal Care and Use Committee procedures.
BCG-immunized mice
C5–/– and C5+/+ mice were immunized i.p. with two doses of BCG, given at 106 CFU per mouse, 2 weeks apart. Mice were used 2 weeks later for spleen-derived T cells. Uninfected mice were used as controls.
Antibody to C5a peptide
A mono-specific antibody was produced in rabbits to the C-terminal peptide (CTIANKIRKESPHKPVQLGR) of mouse C5a linked to keyhole limpet hemocyanin (titer 1/240,000 vs. C5a peptide in ELISA, Bethyl Labs, Inc., Montgomery, TX, USA). Antibody was affinity-purified and stored in aliquots at –70°C until use. To detect C5a peptide from DCs, DCs purified as below were infected with BCG and incubated in McCoy's medium, which supports viable cells without serum for 24 h along with uninfected controls. The culture supernatants were concentrated using an amicon ultraconcentrator (m.w. cut-off, 5000 daltons) and tested in Western blot against antibody to the C5a peptide at a dilution of 1/5000. A 14-kDa C5a peptide was recognized in the supernatants of BCG-infected but not naive C5+/+ or C5–/– DCs (data not shown). C5+/+ DCs thus produced an extracellular C5a peptide similar to macrophages [12
].
DCs
Mouse bone marrow (BM) cells from naive mice were collected and treated with acetate kinase (ACK) lysing buffer followed by washing with PBS. The cells were then cultured in IMDM containing 10% heat-inactivated FBS, recombinant mouse (rm)IL-4, and rmGM-CSF (10 ng/mL each) for 5–7 days. Culture-grown DCs were then purified to 95–98% using CD11c (N418) microbeads (Miltenyi Biotec, Auburn, CA, USA). Purified DCs were then maintained in IMDM with 10% zymosan-treated, heat-inactivated pooled mouse serum for all experiments. This deprived an exogenous source of C5a unless secreted by DCs. CD11c+ DCs were infected with BCG, activated with LPS, or treated with C5a peptide as indicated. Mouse lung-derived DCs (LDCs) were purified using CD11c beads after total lung cells were obtained as described below from alveolar-lavaged and PBS-perfused lungs as described below. To confirm the phenotype of DCs, BM-derived DCs (BMDCs) and LDCs were stained for CD11c-PE, CD4-FITC, CD8-FITC, and the macrophage-specific markers F4/80 and MOMA-1 (AbD Serotec, Raleigh, NC, USA) conjugated to FITC. BMDCs and LDCs contained approximately comparable proportions of CD4+ CD11c+ (
30%), CD8+ CD11c+ (40%), and CD11c+ CD4–, CD8– (30%) cells. All were CD11c+ but F4/80-negative, thus excluding alveolar macrophages (data not shown).
Sandwich ELISA
ELISA was performed as recommended by the manufacturer using the duo set kits for mouse IL-12 p70 and IFN- (R&D Systems, Minneapolis, MN, USA).
IL-12 inhibition assay
DCs were purified by using the Miltenyi Biotec CD11c (N-18) kit from C5–/– and C5+/+ mice and stained for the surface expression of the C5a receptor (C5aR) using a mAb described by Soruri et al. [20
] and plated at 1 x 106 cells/well. Fifteen minutes prior to infection, pyrrolidine dithiocarbamate (PDTC) (NF-
B inhibitor, Calbiochem Inc., San Diego, CA, USA; 10 µM) or an anti-C5aR mAb (kind gift of Dr. Jorg Zwirner, University of Gottingen, Germany) or their combination was added to the DC cultures. DCs were then infected with BCG at a multiplicity of infection (MOI) of 1:1 for 4 h, washed, replated, and incubated with reagents as above. Controls consisted of naive DCs treated with the same reagents, vehicle, or mAb alone. Supernatants were collected after 8 h, and IL-12p70 levels were determined using ELISA. DCs were >90% viable at the end of 8 h, as evaluated by the conversion of alamar blue vital dye.
Spleen-derived T cells
Mouse spleens were collected from immune or naive mice, as per standard procedures, and pan T cell kit (Miltenyi Biotec)-purified T cells were suspended in DMEM with pooled mouse serum (as above) with penicillin-streptomycin and 10 mM 2-ME. The T cells were plated onto a rat anti-mouse monoclonal CD3 (1 µg/mL)-coated plate to activate the cells. After 24 h, the T cells were purified using Ficoll-Hypaque columns to remove dead cells, rested for 24 h, and used for coculture experiments as indicated.
DC-T cell coculture
The method described for DC-T cell coculture, which results in Th1 expansion of naive T cells, was followed [21
]. DCs were infected with BCG (1:1), washed after 4 h, and then plated in replicates at 2 x 106 cells/well using six-well plates. Purified, activated T cells from naive C5+/+ and C5–/– mice were mixed with DCs at a DC:T cell ratio of 1:20 and incubated in the presence of 10 ng/mL rmIL-2 per mL for 5 consecutive days. Thus, 40 x 106 T cells were added per well of DCs in replicate cultures. T cells were harvested from cocultures, purified on Ficoll-Hypaque columns to remove dead cells at the end of 5 days, and tested for intracellular IFN- or further tested for cell proliferation as below.
Specificity evaluation of T cells
DCs were cocultured with naive or immune T cells as above. At 5 days coculture after priming, T cells were purified using Ficoll-Hypaque columns to remove dead cells and incubated at 1 x 105 cells per 300 µL in the presence of PPD antigen (10 µg/mL) for 72 h and mitomycin (1 µg/mL)-treated DCs. C5a peptide was added at the same dose as an irrelevant peptide. On Days 1 and 3 after culture, the proliferative response was measured by addition of 3H-thymidine (0.5 µCi/well) to replicate cultures, and cells were harvested and measured for isotope uptake using a liquid scintillation counter [22
].
Flow cytometry
DCs were typed for CD11c and surface markers, and T cells were typed for surface markers (CD4, CD8) as well as intracellular cytokines as follows. DCs were gated based on forward-scatter and CD11c (integrin 
-chain) stain. An isotype control of rat IgG2a (Caltag Laboratories, South San Francisco, CA, USA, R2a04) was used. A panel of various adhesion: CD44 (Pgp-1, Ly-24, BD PharMingen, San Diego, CA, USA, 553133), CD54 (ICAM-1, BD PharMingen, 553252), CD11a (integrin L-chain, LFA-1 -chain, BD PharMingen, 553340), and CD11b (integrin M-chain, membrane-activated complex 1 -chain, BD PharMingen, 557396 or 553310); costimulatory: CD40 (BD PharMingen, 553791), CD80, B7-1 (BD PharMingen, 553769), and CD86, B7-2 (BD PharMingen, 553692); and antigen-presenting: MHC-II mouse I-Ab (BD PharMingen, 553605) and CD1d (CD1.1, Ly-38, BD PharMingen, 553846) molecules was analyzed. DCs were dispensed at 105 cells per pellet, added with FcR blocker (CD16/32, Caltag Laboratories, MFCR00), and incubated for 15 min on ice. The appropriate antibodies were added to each sample and incubated for 30 min in the dark. The cells were washed and then fixed in 2% paraformaldehyde for 30 min before acquisition using a Becton Dickinson FACScan. The intracellular staining was performed as follows: T cells harvested from cocultures were treated with 1 µg/mL brefeldin A or 5 µg/mL monensin (BD Biosciences, GolgiStop 554715). Mouse lung- or spleen-derived T cells were incubated at 37°C and 5% CO2 for 90 min in the presence of PMA (250 ng/mL) and ionomycin (100 ng/mL) and 1 µg/mL brefeldin A or 5 µg/mL monensin before intracellular staining. T cells were added with FcR blocker (CD16/32, Caltag Laboratories, MFCR00) and incubated on ice for 15 min. The appropriate antibodies were added to each sample (CD4, L3T4, Caltag Laboratories, MCD0401 or MCD0404; CD8, Ly-2, Caltag Laboratories, RM2201-3 or MCD0804-3) and incubated for 30 min in the dark. The cells were then washed and fixed with 2% paraformaldehyde for 30 min, followed by staining with anti-IFN- (Caltag Laboratories, RM9001-3) diluted in a perm buffer containing 0.1% Triton-X 100, 0.1% saponin, and PBS for 2 h at room temperature or overnight at 4°C. The samples were washed thrice in PBS with 0.01% Tween-80 and acquired using a Becton Dickinson FACScan and Cellquest software.
Mouse infections with BCG
Previously reported methods were used to enumerate and characterize T cells in BCG-infected mice [23
, 24
].
Growth curves of BCG
BCG was sonicated as a viable suspension and aerosolized to C5+/+ and C5–/– mice using the Middlebrook aerosol apparatus. Mice were housed in cages and sacrificed weekly intervals for 4 weeks. At each time-point, lungs were homogenized in saline and plated for CFUs on 7H11 agar. CFUs were determined after 4 weeks of incubation and plotted against time.
Flow cytometry of lung T cells
Eight mice were sacrificed each week and alveolar-lavaged, and PBS-perfused lungs were removed and teased in saline. Lungs were minced into pieces with scissors, digested in PBS buffer with EDTA (10 mM), EGTA (5 mM), 100 U per mL collagense (Type VII), and 50 µg/mL elastase (both from Sigma Chemical Co.) for 90 min at 37°C and 5% CO2. Cells were strained through a nylon mesh, and red cells were lysed in ACK buffer. Single cell suspensions were prepared by passing through 18- followed by 21-gauge needles, and cells were counted in a hemocytometer. The usual yield of cells from tuberculosis-infected lungs at week 2 postinfection was 50–100 x 106 cells, which increased up to 4 weeks. Viability was tested by trypan blue exclusion, and usually, the viability was >95%. Flow tubes were then dispensed with 106 cells in 100 µL vol and stained for CD4 and CD8 followed by intracellular stains. It should be noted that this cell number represented the total lung cells of which T cells were a proportion.
Absolute numbers of T cells per lungs or spleens
These were done only for 2–4 weeks of infection, as T cells were then relatively abundant. A known number of total lung cells were dispensed to the staining tubes (106/pellet; >95% viable), and cells were stained for CD4 and CD8 phenotype and for intracellular IFN-. At least 100,000 events were acquired for each of the two pools consisting of four mouse lungs each. Absolute numbers of IFN-+ T cells were then calculated from the percent-positive T cells observed per total number of lung cells.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, IL-4, and IL-10, which can skew their ability to prime T cells [25
26
27
]. C5–/– DCs produced less TNF- following infection but more IL-10 in response to BCG infection (Fig. 1D
and 1E)
. Together, these observations suggest that C5–/– DCs have reduced IL-12p70 and TNF-. At the same time, they have enhanced IL-10, which can possibly skew DC-induced priming responses. The levels of IL-4 from BCG-primed DCs were low in cultures and are not shown. BMDCs as well as LDCs produced enormous amounts of IL-6 (>5000 pg/mL) after BCG infection, and no significant differences could be found for this cytokine between C5–/– and C5+/+ DCs (not shown).
 View larger version (25K): [in a new window] |
Figure 1. Complement anaphylatoxin (C5a)-deficient DCs show reduced IL-12p70 and altered pro (TNF-)- and anti (IL-10)-inflammatory cytokine levels after mycobacterial infection. BMDCs from C5–/– and C5+/+ mice were infected with M. bovis BCG, and cytokines were measured using sandwich ELISA at indicated times. (A) C5–/– DCs show reduced IL-12p70 (*, P<0.009; **, P<0.007). (B) Addition of C5a to BCG-infected C5–/– DCs increases IL-12 secretion (***, P<0.01). (C) LDCs show a similar defect in producing IL-12, which is restored after C5a supplementation ($, P<0.01). TNF- and IL-10 levels were also measured in BMDCs infected with BCG. (D) TNF- levels are reduced in C5–/– DCs (*, P<008; **, P<0.009). (E) IL-10 is increased in C5–/– DCs after BCG infection (*, P<0.009; **, P<0.01). Data represent three independent experiments. Significance evaluated using the t-test.
|
B activationB pathway. mAb to C5aR and PDTC were used to block the C5aR and NF-B pathway, respectively. Figure 2B
shows that the blockade of C5aR led to reduction of BCG-induced IL-12p70. As C5a-R mAb or naive DCs alone did not produce IL-12p70, it appears that BCG-induced IL-12p70 occurs through activation of C5aR via the C5a peptide. PDTC likewise blocked BCG-induced IL-12p70. This suggests a link between BCG-induced, C5aR-dependent IL-12 secretion and NF-B activation. These observations appear consistent with previous observations that bacterial products, C5a and chemokines such as IL-8, induce IL-12 through NF-B-dependent mechanisms [28
, 29
].
 View larger version (16K): [in a new window] |
Figure 2. IL-12 secretion in DCs is dependent on signaling via C5aR as well as NF-B activation. (A) DCs purified from C5–/– (left) and C5+/+ (right) mice were stained for C5aR using a mouse mAb (C5aR FITC mAb) after BCG infection (blue/purple; two separate experiments merged). Naive cells (green/black; two separate experiments merged) show a modest C5aR expression, which is up-regulated after BCG infection (unstained or isotype cells are in red). (B) C5+/+ DCs were treated with anti C5aR mAb (100 ng/mL) or pyrollidone dithiocarbamate (PDTC; 10 µM), an NF-B inhibitor, infected with BCG and IL-12, measured 8 h later in three independent experiments. BCG induces copious IL-12p70, which is inhibited by C5aR blockade as well as PDTC treatment. Their combination is not synergistic. A combination of BCG with an isotype IgG1 antibody or vehicle control for PDTC has no effect. PDTC alone or C5aR mAb fail to induce IL-12 (*, P<0.0083; **, P<0.0097; ***, P<0.00.91, vs. BCG alone group, t-test).
|
 View larger version (29K): [in a new window] |
Figure 3. Expression of CD1d and CD40 in DCs after BCG infection and correlation with IL-12p70 levels after DC and T cell coculture. BMDCs were infected with BCG in the presence or absence of C5a peptide (100 ng/mL) and stained for surface expression of CD1d and CD40 before analysis by flow cytometry. (A) Histograms illustrate differential expression of CD1d and CD40 in DCs and represent data from three experiments merged into one profile. They show receptors before (left column) and after addition of C5a peptide (right column). C5–/– DCs show reduced CD40, which is enhanced after addition of the C5a peptide (*, significance, see Table 1
). CD1d expression between C5–/– and C5+/+ DCs is similar. C5+/+ DCs show a comparable expression of CD40 and CD1d before and after addition of C5a peptide. Mean fluorescence intensity (MFI) calculated from three independent experiments is shown in Table 1
. (B) C5+/+ and C5–/– DCs were cocultured with BCG-immune spleen T cells from C5+/+ and C5–/– mice to determine if ligation of CD40 on DC with CD40L on T cells makes a difference in IL-12p70 levels. C5–/– DCs continue to show reduced IL-12 p70 (*, P<0.01, t-test).
|
|
View this table: [in a new window] |
Table 1. C5 Deficiency Affects CD40 Expression in DCs
|
|
View this table: [in a new window] |
Table 2A. Surface Expression of Other Adhesion Molecules in C5–/– and C5+/+ DCs after BCG Infection
|
|
View this table: [in a new window] |
Table 2B. Surface Expression of Costimulatory and MHC II Molecules in C5–/– and C5+/+ DCs after BCG Infection
|
and expand Th1 T cells-secreting T cells in vitro [32
]. Accordingly, DCs were infected with BCG and cocultured with splenic T cells from naive mice. To ensure that priming of naive cells was enhanced, T cells were preactivated with anti-CD3 antibody for 24 h and rested for 24 h before coculture with infected or naive DCs. T cells were purified on Ficoll columns on Day 5 to remove dead cells, and live cells were stained for CD4 and CD8 phenotype and intracellular IFN-. A typical pattern of the expansion of IFN--producing T cells is illustrated in Figure 4A
. The histograms show the limited expansion of IFN-+ CD4 and CD8 T cells over 7 days when cocultured with BCG-infected C5+/+ DCs (left and middle columns). Naive C5+/+ T cells cocultured with naive C5+/+ DCs do not expand over 5 days (right column). Results from three independent experiments summarized in Fig. 4B
and 4C
, confirm that C5+/+ DCs are more efficient than C5–/– DCs to induce the expansion of naive T cells. Furthermore, Figure 4D
shows that C5+/+ cocutures also contain higher levels of IFN-. Finally, T cells harvested on Day 5 were further cultured in the presence of PPD antigen or an irrelevant peptide. PPD antigen induces a strong T cell reaction to antigens 85A and 85B shared within the MTB complex (including BCG) and is thus representative of the immunodominant antigens of the MTB complex [33
]. Naive T cells sensitized by BCG-infected DCs replicated over 3 days and incorporated 3H-thymidine, indicating the presence of antigen-specific T cells in the cocultures (Fig. 4E)
. Such cells did not proliferate in the presence of control peptide. Finally, to determine whether defects in T cell priming could be corrected, cross-priming experiments were performed. C5–/– T cells were cocultured with C5+/+ DCs and vice versa. Figure 4F
and 4G
, shows that cross-priming was effective in expansion of C5–/– CD4 T cells via C5+/+ DCs but not for CD8 T cells. Paradoxically, C5–/– CD4 T cells cocultured with C5+/+ DCs showed reduced expansion into IFN--secreting T cells. Figure 4H
shows that IFN- levels reflected enhanced priming of C5–/– T cells by C5+/+ DCs. The latter studies suggested that a defect in C5–/– DCs is responsible for reduced expansion of T cells.
 View larger version (40K): [in a new window] |
Figure 4. C5+/+ and C5–/– DCs prime and induce IFN--secreting Th1 T cells in vitro. BCG-infected or naive C5+/+ and C5–/– DCs were cocultured, respectively, with C5+/+ and C5–/– naive, splenic T cells in the presence of IL-2, and on Day 5, Ficoll-purified cells were tested for intracellular IFN-, followed by proliferation in the presence of antigen. (A) Dot-plots illustrate the expansion of IFN-+ CD4 and CD8 T cells from naive C5+/+ mice when cocultured with BCG-infected C5+/+ DCs (middle and bottom rows; left and middle columns). Isotype controls are shown in top row. Numbers in upper-right quadrants show percent of T cells positive for intracellular IFN-. CD4+ T cells expand better than CD8 T cells over 5 days. Expansion is specific, as T cells cocultured with naive DCs (right column) do not secrete IFN-. (B and C) Percent of IFN-+ CD4 and CD8 T cells is tabulated from three independent experiments, which show that C5+/+ DC-T cell cocultures contain more IFN-+ CD4 and CD8 cells (*, P<0.01; **, P<0.01, t-test). (D) Supernatants of cocultures shown in Figure 4B
and 4C
, were assayed for IFN- using sandwich ELISA. Cultures of C5+/+ DC-T cell cocultures contain more IFN- (***, P<0.009, t-test). (E) T cells harvested on Day 5 from cocultures were cultured again in the presence of 10 µg/mL PPD or C5a peptide as irrelevant antigen and on days shown, tested for 3H-thymdine uptake. T cells replicate and incorporate isotope only in the presence of PPD antigen (*, P<0.0087; **, P<0.0091, vs. control peptide, t-test). (F and G) Cross-priming studies were done to determine if C5+/+ DCs can prime C5–/– T cells and vice versa. C5+/+ DCs induce a better expression of IFN- in CD4 but not CD8 T cells, and C5–/– DCs are unable to induce IFN- synthesis in C5+/+ T cells (*, P<0.0081). ns, Not significant. (G) Increased IFN- levels occur in cocultures of C5+/+ and C5–/– T cells (**, P<0.0043).
|
+-immune T cells as above. Figure 5A
(left and middle columns) illustrates that C5+/+ DCs were more efficient in inducing the expansion of IFN-+ T cells than C5–/– DCs. Immune T cells cocultured with uninfected DCs did not expand over 5 days of culture, suggesting that expansion was antigen-driven (Fig. 5A
, right column). Results from three independent experiments are shown in Figure 5B
and 5C
, and IFN- levels are shown in Figure 5D
. Figure 5E
demonstrates that T cells primed to expand with C5+/+ DCs strongly incorporate 3H-thymidine in the presence of specific PPD antigen but not an irrelevant control peptide.
 View larger version (47K): [in a new window] |
Figure 5. C5+/+ DCs induce a stronger recall expansion of Th1 response in immune T cells than C5–/– DCs during in vitro coculture. DCs from C5+/+ and C5–/– mice were cocultured as in Figure 4
with the difference that immune T cells from mice immunized with BCG were used. (A) Dot-plots illustrate that DCs induce an expansion of IFN-+ CD4 and CD8 T cells. (Top row) Isotype controls. (Second and third from top rows) C5+/+ DCs induce a better expansion of IFN-+ CD4 and CD8 T cells. Numbers in upper-right quadrants show percent of T cells positive for intracellular IFN-. (Bottom two rows) C5–/– DCs induce a less-significant expansion of IFN-+ CD4 and CD8 T cells. For each row, immune T cells were cultured with naive DCs, which do not induce IFN-+ T cells. (B and C) Data from three experiments confirm better induction of a Th1 response by C5+/+ DCs (*, P<0.0092; **, P<0.01, t-test). (D) C5+/+ DC-T cocultures contain higher levels of IFN- (***, P<0.0056, t-test). (E) T cells harvested on Day 5 were cultured in the presence of 10 µg/mL PPD or C5a peptide and on days shown, tested for 3H-thymdine uptake. T cells replicate and incorporate isotope only in the presence of PPD antigen, indicating antigen-specific expansion (*, P<0.007; **, P<0.0081, vs. control peptide, t-test).
|
 View larger version (23K): [in a new window] |
Figure 6. DC-T cocultures contain different levels of IL-12p70. Culture supernatants of naive and immune T cells cocultured as in Figures 4
and 5
were tested for IL-12p70 using sandwich ELISA. C5+/+ DC-T cocultures uniformly contain higher levels of IL-12p70 (*, P<0.01; **, P<0.0087, t-test)
|
-secreting T cells during the course of infection. Figure 7A
illustrates the growth curve of BCG, and it is apparent that the bacteria grow more in the lungs of C5–/– mice. On Day 28, the peak of infection, there was approximately a log10 increase in the growth of BCG in C5–/– mice over that of C5+/+ mice.
 View larger version (46K): [in a new window] |
Figure 7. C5–/– mice show increased growth of BCG in the lungs after aerosol infection, coincident with the decreased accumulation of IFN--secreting T cells in the lungs. C5+/+ and C5–/– mice were aerosol-infected with BCG with 100 CFU per lungs and killed at intervals for organ counts of bacteria, and the lungs and spleens were analyzed for T cell content. (A) BCG grows more in the lungs of C5–/– mice between Days 21 and 28 (*, P<0.0072, Mann Whitney U test). Four mice killed per time-point. (B) Lung-derived T cells were stained for intracellular IFN-+ T cells. Eight mice killed per time-point for T cell stains and counts. C5+/+ mice show a stronger increase in the percentage of IFN-+ T cells in the lungs compared with C5–/– lungs (*, P<0.005, vs. CD4 T cells of C5–/– mice; **, P<0.01, vs. CD8 T cells of C5–/– mice, t-test). (C) Spleens show minimal changes in the proportion of IFN-+ T cells. (D) Dot-plots illustrate the expansion of CD4+ IFN-+ T cells in C5–/– (bottom row) and C5+/+ mice (middle row) over the time course of BCG infection. IFN-+ T cells increase in frequency in the lungs of C5+/+ mice compared with C5–/– mice. Isotype control for the IFN- stain and the distribution of T cells on week 3 are shown in the top row.
|
+ T cells after BCG aerosol infection+ T cells using flow cytometry. Figure 7B
illustrates the enumeration of T cells in the lungs over the 4 weeks of infection. One week postinfection, there were few T cells in the lungs (not shown). By the 2nd week, the lungs of C5+/+ mice showed significant numbers of IFN-+ CD4+ T cells, which increased progressively over 4 weeks. In the same mice, IFN-+ CD8+ T cells expanded more gradually. It was striking to note that CD4 and CD8 T cells, which were positive for IFN-, were lower in number in the lungs of C5–/– mice up to 4 weeks of infection. The spleens of these mice did not show appreciable differences in the numbers of T cells, indicating that the change in T cell numbers was unique to lungs (Fig. 7C)
. Figure 7D
illustrates representative histograms of IFN-+ CD4+ T cells in the lungs of C5+/+ versus C5–/– mice during the course of infection between the 2nd and 4th weeks.
As C5a is a chemotactic agent and has been reported to chemoattract T cells in mice, reduced numbers of T cells in C5–/– mice could have been a result of decreased recruitment into the lungs from the spleen or hilar lymph nodes [34
]. To ascertain this, the T cell numbers of lungs and spleens were enumerated and tabulated. Table 3A and 3B
, indeed shows that week 2 postinfection, the lungs of C5–/– mice contained reduced numbers of CD4 and CD8 T cells when compared with the lungs of C5+/+ mice (Table 3A)
. The numbers of IFN-+ CD4 and CD8 T cells were also less in C5–/– mice (Table 3B)
. At the same time, however, spleens of both types of mice contained comparable numbers of T cells (Table 4
). The numbers of T cells in hilar lymph nodes were too few to be enumerated reliably. Thus, it appears that between weeks 1 and 2 postinfection, there is reduced influx of T cells into the lungs of C5–/– mice, possibly as a result of the lack of C5a secreted from lung macrophages. It is interesting that between weeks 3 and 4 postinfection, the total lung numbers of CD4 and CD8 T cells were comparable between C5+/+ mice and C5–/– mice, although C5–/– mice continued to have markedly reduced numbers of IFN-+ CD4 and CD8 T cells (Table 3A and 3B)
. Thus, we suggest that there is decreased, early infiltration of T cells into the lungs of C5–/– mice, perhaps as a result of defects in chemotaxis. This defect is not apparent after week 2 and despite comparable T cell influx, C5–/– mice continue to show reduced accumulation of IFN-+ T cells by weeks 3 and 4. We suggest that this is a result of the defect in priming and expansion of Th1 T cells by DCs.
|
View this table: [in a new window] |
Table 3A. Total Numbers of CD4 and CD8 T Cells and IFN-+ CD4 and CD8 T Cell Numbers in the Lungs of C5–/– and C5+/+ Mice during Aerosol BCG Infection
|
|
View this table: [in a new window] |
Table 4. Total Numbers of CD4 and CD8 T Cells in the Spleens of C5–/– and C5+/+ Mice during Aerosol BCG Infection
|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
+ T cells is protective against many intracellular infections including those caused by mycobacteria [36
]. IL-12 is a key determinant of Th1 immunity in human as well as murine systems. Thus, IL-12R deficiency is associated with an increased risk of tuberculosis, and IL-12p40-deficient mice are more susceptible to tuberculosis [37
, 38
]. We described for the first time [12
] along with Karp et al. [13
] that murine C5–/– macrophages produced reduced IL-12 [12
, 13
]. Furthermore, B10-derived, C5–/– mice also showed reduced secretion of IL-12p70 from MTB-infected macrophages. Furthermore, they expressed reduced IL-12p40 mRNA messages in the lungs during MTB infection [15
]. As C5–/– mice are more susceptible to tuberculosis, a link between C5a deficiency, IL-12, and development of Th1 immunity was obvious. In our earlier studies, we investigated the role of C5 in macrophage-mediated immunity. Macrophages secrete C5, which is cleaved by extracellular proteases to result in the C5a peptide, which in turn, binds to the C5aR. Binding of C5a to C5aR triggers multiple events within neutrophils and macrophages, which culminate in proinflammatory responses. C5a induces a respiratory burst in neutrophils as well as macrophages and induces chemotaxis [39
]. It has been found to regulate macrophage production of TNF-, IL-6, and IL-1ß [40
, 41
]. We showed that C5 deficiency affects even chemokine secretion, leading to defective granuloma formation in mice [15
]. Thus, it appears reasonable to propose that the physiologic effects of C5a anaphylatoxin on macrophages and neutrophils have been studied extensively [34
, 42
]. In contrast, the effects of C5a on DCs are relatively uncharacterized. Soruri et al. [43
] reported that rmC5a recruited and enhanced the maturation of human monocytes and immature DCs into mature DCs in the peritoneal cavity of SCID mice. The effect of C5a was indirect and was dependent on TNF- and PGE2. The mature DCs released IL-12 when stimulated with CD40L. Recent studies show that DCs can modulate the Th2 immunity in mice through C5aR interactions [18
, 19
]. We therefore speculated that C5a could also play a regulatory role in the generation of Th1 immunity. Research into the regulatory effects of C5a on DCs is important, as DCs alone seem capable of priming naive T cells to differentiate into effectors [32
], the C5a peptide was reported to be one of the earliest molecules released during cutaneous, hypersensitivity reactions to chemoattract T cells [44
], and abundant receptors for the C5a peptide are expressed on DCs [43
, 45
, 46
].
As IL-12 drives the Th1 response, we initially sought to characterize the cytokine response of murine DCs. C5–/– DCs produced significantly less IL-12 than C5+/+ DCs, confirming that reduced IL-12 is a common feature of C5–/– macrophages as well as DCs. Reduced IL-12 was linked to signaling through C5aR as well as NF-B. Supplementation of DCs with the C5a peptide during BCG infection restored their ability to secrete copious IL-12. Thus, C5a appears to be an innate peptide, which is released by DCs during infection and regulates production of IL-12 through the C5aR in a loop mechanism similar to macrophages. Our findings that C5a affects IL-12 levels in BMDCs are consistent with our own previous observation on macrophages [12
] and that of Karp et al. [13
]. They are in contrast with two recent reports about macrophages. la Sala et al. [47
] used human monocytes to show that C5a negatively affects IL-12 production. Hawlisch et al. [48
] used C5aR-deficient mice to show that there was an increase in Th1 response in the absence of C5a-C5aR interactions and that C5a negatively regulated IL-12 at the level of macrophages. Although we are not certain why murine DCs are positively regulated by C5a in our studies, it is possible that C5a affects human or mouse macrophages and DCs differently. One possibility is that C5a may affect IL-12 levels in a dose-dependent manner, and moderate-to-slight excess of C5a may skew IL-12 production adversely. Indeed, Braun et al. [49
] described a similar, self-regulatory mechanism in murine macrophages. A second possibility is that IL-12p70 production is dependent on the ligation of appropriate TLRs in murine DCs [7
]. Use of ligands specific to TLRs may cause a differential release of IL-12p70 as well as other cytokines, which in turn, may have different effects on the ability of DCs to prime and expand T cells. It should be noted here that besides the regulatory effects of C5a, BCG infection itself can modulate levels of IL-12. For example, BCG reduced IL-12 secretion but increased IL-10. This is consistent with the previous report that BCG inhibits IL-12 production but stimulates IL-10 in hDCs [50
]. It is known that IL-10 down-regulates IL-12 production in macrophages as well as DCs [51
, 52
]. Thus, it is possible that DCs may also have an IL-10-dependent, cross-regulatory mechanism of IL-12 synthesis.
The reduced IL-12p70 response of DCs led to us to examine the role of CD40. It is interesting that C5–/– DCs showed reduced expression of CD40, which could be restored with the addition of a C5a peptide. The molecular basis of reduced CD40 in C5–/– DCs remains unclear. However, TNF- is known to induce maturation of DCs, including up-regulation of CD40 [53
]. We suggest that reduced TNF- levels in C5–/– DCs may lead to reduced CD40 expression. In any case, reduced IL-12p70 and CD40 expression suggested that the C5–/– DCs could be defective in priming T cells. In vitro DC and T cell cocultures were used to examine the expansion of immune as well as naive T cells by BCG-infected DCs. In these experiments, C5–/– DCs were less able to induce an expansion of IFN--secreting T cells than were C5+/+ DCs. As naive DCs failed uniformly to expand immune or naive T cells or induce IFN--secreting T cells, we suggest that reduced IL-12 and reduced CD40 were directly responsible for the decreased amplification of IFN--secreting T cells under in vitro conditions.
Ex vivo DC and T coculture experiments are elegant in that they can identify distinct populations of T cells expanding into cytokine-secreting T cells. However, they reflect only a part of the priming process, which occurs in vivo. For example, DC and T cells may differentiate better under multiple types of cells and cofactors, available only in the lymphoid tissue. Thus, ex vivo experiments were extended to analyze in vivo expansion in T cells. Mice were aerosol-infected, and lung T cells were analyzed for cytokine secretion. These studies confirmed that C5–/– mice showed a poor expansion of IFN--secreting T cells in the lungs compared with C5+/+ T cells. It is important to note that C5–/– mice allowed BCG to grow more in lungs than C5+/+ mice. This suggests that such mice have a decreased ability to kill BCG. As the ability of IFN- to activate and kill mycobacteria is well established in murine models, increased BCG growth appears to be a result of the decreased Th1 response in the lungs of C5–/– mice. It should be noted that early after aerosol infection, C5–/– mice showed a reduced influx of T cells into the lungs, suggesting a C5a-related, chemotatic defect. Thus, decreased Th1 immunity in the lungs of C5–/– mice could be a result of a decreased homing defect as well as a decreased IL-12 response of C5–/– DCs and decreased priming of T cells.
In conclusion, our study suggests a regulatory role for the C5a peptide in the development of murine DC-mediated Th1 immunity. However, some deficiencies remain to be investigated. In general, a suppression of Th1 leads to dominance of the Th2 or Th3 pathway in mice. It remains unclear if C5–/– mice have a change in these pathways. The levels of IL-4-secreting T cells were too low for detection by Day 28 postinfection in the lungs of mice. As chronic infection with BCG has been shown to result in a shift in Th1-to-Th2 immunity in mice, it is possible that a Th1/Th2 shift occurs in C5–/– mice during chronic infection beyond 4 weeks [54 ]. Finally, macrophages and DCs may produce a C5a peptide in distinct, local environments to skew T cell development independent of each other. Thus, the regulatory role of IL-12 on these physiologically and functionally different populations needs further research.
Received February 27, 2006; revised March 5, 2007; accepted May 14, 2007.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
2 J. Immunol. 172,6938-6943B and primes DC for IL-12 production Immunity 9,315-323[CrossRef][Medline] responses Clin. Exp. Immunol. 128,405-410[CrossRef][Medline] receptor control IL-12 p40 synthesis and NF- B activation J. Immunol. 172,2559-2568+ and CD8 – murine dendritic cell subsets J. Immunol. 172,4826-4833}B activation and production of bioactive IL-12 Proc. Natl. Acad. Sci. USA 102,11811-11816 responses if IL-12p70 is available J. Immunol. 175,788-795 and prostaglandin E2-dependent mechanisms J. Immunol. 171,2631-2636 enhances phenotypic and functional maturation of human epidermal Langerhans cells and induces IL-12 p40 and IP-10/CXCL-10 production FEBS Lett. 579,3660-3668[CrossRef][Medline]
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
