Published online before print January 13, 2006
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Laboratory for Immunogenetics, Central Institute for Tuberculosis, Moscow, Russia
1Correspondence: Laboratory for Immunogenetics, Central Institute for Tuberculosis, Yauza alley 2, Moscow 107564, Russia. E-mail: a.pichugin{at}mail.ru
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: tuberculosis • susceptibility • T lymphocytes
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Earlier, we reported that there is a profound difference between H2-congenic, IE-negative mouse strains on a B10 background in terms of the severity of M. tuberculosis-induced disease and the ability to maintain a recurrent T cell response to stimulations with mycobacterial sonicate. Whereas T cells from mice of the resistant strain 4R (H2h4) responded readily to repeated stimulations in vitro and expanded to form a stable CD4+ line, T cells from their B10 (H2b) counterparts died rapidly following the second stimulation cycle [6 ]. Another noteworthy difference between cultured T cells from TB-susceptible and -resistant mice was a clear deficiency of CD8+ cells in the former cultures. Neither the specificity of these features for mycobacterial antigens nor the correlation of the persistent T cell response with resistance to infection among a panel of distinct MHC haplotypes has been studied.
Here, we extend our earlier observations by demonstrating that T cells from mice of three different H2-congenic strains resistant to TB infection readily respond by proliferation following repeated stimulations with mycobacterial sonicate. By contrast, T cells from mice of three susceptible strains died after the second antigen stimulation. This difference was antigen-specific, in that T cells from TB-susceptible and -resistant mouse strains did not differ in their capacity to respond against bacterial (staphylococcal) or protein (conalbumin) antigens. The CD4/CD8 ratio in immune lymph nodes strongly correlated with TB susceptibility and was significantly lower in resistant mice as a result of a higher proportion of CD8+ cells. The difference in capacity to maintain a persistent, antigen-specific T cell response between two sets of strains was a result of a significantly elevated level of activation-induced apoptosis in T cells from TB-susceptible mouse strains. Taken together, these results contribute to a better understanding of the mechanisms of MHC involvement in the control of TB susceptibility and of the nature of antigen-induced T cell loss during infection.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Infection of mice
Mice were infected intravenously (i.v.) with 5 x 105 or 2 x 107 colony-forming units (CFUs) of mid-log-phase M. tuberculosis strain H37Rv (original stock was a kind gift of Dr. Gilles Marchal, Institute Pasteur, Paris, France) in 0.5 ml saline, as described earlier in detail [15
]. Mortality of mice was monitored daily, starting at day 14 following infection, and is expressed as mean survival time (MST).
Immunization
To induce T cell responses, mice were immunized in the footpads with 50 µg H37Rv sonicate [1
], 50 µg cytoplasmic staphylococcal antigen [16
], or 250 µg conalbumin (Sigma Chemical Co., St. Louis, MO) per mouse. Antigens were solubilized in saline and mixed 1:1 with incomplete Freund’s adjuvant (Sigma Chemical Co.). Cells from popliteal lymph nodes were collected at days 10–14 following immunization.
Proliferation assays
Cells from popliteal lymph nodes were prepared and cultured exactly as described earlier [6
]. Antigen-specific stimulation was achieved by adding 10 µg/ml H37Rv sonicate, cytoplasmic staphylococcal antigen, or conalbumin; nonstimulated wells served as controls. Cells were cultured for a total of 65 h at 37°C in an atmosphere of 5% CO2 in air. All cultures were performed in triplicates. For the last 18 h, cultures were pulsed with 0.5 µCi/well methyl-[3H]-thymidine. Cultures were harvested onto glass microfiber filters using a semiautomatic cell harvester (Scatron, Oslo, Norway) for liquid scintillation counting. Results are expressed as 
counts per minute (cpm), i.e., cpmAntigen (Ag) – cpmControl.
Stimulation/rest protocol for prolonged lymphocyte cultures
Cells from popliteal lymph nodes were cultured as described previously [6
]. Briefly, 2 x 106 immune cells isolated from popliteal lymph nodes were cultured per well in 1 ml RPMI 1640 containing 10% fetal calf serum, 10 mM HEPES, 4 mM L-glutamine, 5 x 10–5 M 2-mercaptoethanol, vitamins, piruvate, nonessential amino acids, and antibiotics (all components from HiClone, Logan, UT) in 24-well plates (Costar, Badhoevedorp, The Netherlands) for 9–11 days in the presence of antigens (see above). Following a period of culture, live immune cells (>93% viability by trypan blue exclusion) were isolated by centrifugation at 2500 g for 20 min at 23°C on a 1.088-g/ml Lympholyte M gradient (Cedarlane Labs, Ontario, Canada), washed twice, and counted. The next stimulation cycle was accomplished by coculturing 2 x 105 isolated cells with irradiated (12 Gr) 1.5 x 106 splenic antigen-presenting cells (APC) in the presence of antigens for another 9–11 days. This 9- to 11-day cycle was repeated until the cells were lost or started to grow as a stable cell line.
Cytokine assays
Enzyme-linked immunosorbent assay was used to detect interlerukin (IL)-4, IL-10, IL-12, and interferon-
(IFN-) in 48-h culture supernatants as described [16
, 18
]. Capture and detecting (biotinylated) monoclonal antibodies (mAb) specific for mouse cytokines were purchased from PharMingen (San Diego, CA): for IFN-, clones R4-6A2 and XMG1.2 (sensitivity, 312 pg/ml); for IL-4, clones 11B11 and BVD6-24G2 (sensitivity, 62 pg/ml); for IL-10, clones JES5-2A5 and JES5-16E3 (sensitivity, 312 pg/ml); and for IL-12, clones C 17.8 and C 15.6 (sensitivity, 250 pg/ml).
Immunofluorescent staining
Lymph node cells (3–5x105) were washed twice in phosphate-buffered saline (PBS) containing 0.01% NaN3 and 0.5 % bovine serum albumin and incubated for 15 min at 4°C in the presence of CD16/CD32 mAb (clone 2.4G2, PharMingen) to block Fc receptors. Cells were then double-, triple-, or quadruple-stained with antibodies conjugated directly, according to the manufacturer’s instructions. All antibodies were purchased from PharMingen: fluorescein isothiocyanate (FITC)-anti-CD4 (clone H129.19), peridinin chlorophyll protein (PerCP)-anti-CD4 (clone RM4-5), phycoerythrin (PE)-anti-CD8a (clone 53-6.7), PerCP-anti-CD8a (clone 53-6.7), PE-anti-CD3 (clone 17A2), FITC-anti-CD44 (clone IM7), allophycocyanin-anti-CD62 ligand (CD62L; clone MEL-14), and FITC-anti-CD95 (clone Jo2). Stained cells (104 cells per sample) were washed twice, fixed with 1% paraformaldehyde, and analyzed by flow cytometry using FACSCalibur cytometer (Becton Dickinson, San Diego, CA), CellQuestPro (Becton Dickinson), and FlowJo 4.5.9 (Tree Star, Inc., San Carlos, CA) software.
Apoptosis evaluation
To evaluate the apoptotic pathway of cell death, the following techniques were used: (i) assessment of DNA fragmentation exactly as described by Duke and Cohen [18
]; (ii) enumeration of hypoploid cells, as described by Nicoletti et al. [19
]. Briefly, cells were removed from the wells, washed, and incubated in 70% ethanol for 30 min at 4°C. Pellets were washed twice with PBS and incubated for 30 min in 0.5 ml 50 µg/ml propidium iodide (PI) and 100 µg/ml RNAase A at room temperature in the dark. Fluorescence of PI was determined using flow cytometry to assess the percentage of nuclei with DNA content lower than in normal diploid cells. (iii) Annexin V assay for T cells, which were cultured in the presence or absence of 10 µg/ml H37Rv sonicate for different time-points, harvested, stained with FITC-Annexin V (PharMingen) and PI or 7-amino-actinomycin (7-AAD; PharMingen), according to the manufacturer’s instructions, and assayed by flow cytometry.
Statistical analysis
The differences between experimental groups (Student’s t-test and Mann-Whitney U-test) and the values of correlation coefficient (
) were considered statistically significant at P < 0.05. The data were processed by means of BIOSTAT software (Practika, Moscow, Russia).
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
View this table: [in a new window] |
Table 1. Influence of H2 Haplotypes on the Survival Time following TB Challengea
|
As shown in Table 2A , unusually low CD4/CD8 ratios (<0.9) were observed in lymph nodes of immunized mice expressing R phenotype, whereas in TB-susceptible H2 congenics the CD4/CD8 ratios were between 1.5 and 2.0. The difference between the two groups was largely a result of a significantly (P<0.05) higher CD8+ T cell content in lymph nodes of mice of all resistant strains, whereas the numbers of CD4+ T cells were paractically identical in the two susceptible and three resistant mouse strains, and B10.SM mice were an exception (Table 2A) . Furthermore, the difference between TB-susceptible and -resistant H2 congenics was secondary to immunization: in nonimmunized mice of all strains (again, except B10.SM), the CD4/CD8 ratio in lymph nodes was similar (range, 0.95–1.2). B10.SM showed a clear, intrinsic prevalence of CD4+ cells (CD4/CD8 ratio=2:1). To evaluate whether similar differences in CD4/CD8 ratios between TB-susceptible and -resistant mice accompany the development of TB disease, we measured CD4/CD8 ratios in the auxillary lymph nodes of susceptible B10 and resistant 4R mice at Week 3 following i.v. infection with 5 x 105 M. tuberculosis CFU. As shown in Table 2B , the CD4/CD8 ratio again was significantly (P<0.05) higher in susceptible mice, and this was a result of a decreased number of CD8 cells in their lymph nodes. It is interesting that the total number of cells per lymph node did not differ between susceptible and resistant mouse strains following immunization or infection (data not shown), indicating that administration of mycobacterial antigens leads to shifts in the CD4:CD8 ratio per se rather than affects mechanisms of cell homeostasis.
|
View this table: [in a new window] |
Table 2. Influence of H2 Haplotypes on the CD4+/CD8+ Ratio in Lymph Nodes
|
As shown in Figure 1
, neither the cell yield nor the level of antigen-specific proliferation differed between mice of the two strains after immunization and the first stimulation in vitro with the two nonmycobacterial antigens. Moreover, contrary to what has been observed during mycobacteria-specific response, following the second stimulation with conalbumin, significantly (P<0.01) fewer live cells were recovered from cultures of 4R T cells (Fig. 1A)
, and these cells demonstrated lower (P<0.05) proliferative response to conalbumin (Fig. 1B)
. T cells from neither mouse strain responded to the second in vitro stimulation with CSA (Fig. 1C
and 1D)
. As during mycobacteria-specific response, conalbumin- and CSA-specific T cells were biased toward the type 1 cytokine profile (IFN--positive; IL-4- and IL-10-negative, not shown). Overall, these results suggest that the H2-controlled difference in the capacity of T cells to respond to repeated stimulations with mycobacterial antigens, which correlate with the difference in susceptibility to infection, is mycobacteria-specific.
 View larger version (20K): [in a new window] |
Figure 1. Survival (A and C) and proliferative response (B and D) of lymph node cells following repeated stimulations in vitro with conalbumin (Con; A and B) and CSA (C, D). Cells from popliteal lymph nodes were isolated 2 weeks following immunization with 250 µg/mouse conalbumin or 50 µg/mouse cytoplasmatic staphylococcal antigen in Freund’s incomplete adjuvant. cpm = cpmAg – cpmControl. Two independent experiments with mixtures of cells from three mice of each strain in each experiment provided similar results.
|
 View larger version (44K): [in a new window] |
Figure 4. Percentage of CD4+-activated cell (CD4+CD44highCD62Llow) in lymph nodes of TB-susceptible (B10.SM, B10) and TB-resistant (4R, B10.MBR) mice 2 weeks following immunization with H37Rv sonicate. Three independent experiments (Ntotal=9 mice of each strain) gave similar results with <10% variation (see text).
|
50%, quadrants A3) did not depend on the presence of mycobacterial sonicate. By contrast, antigen stimulation promoted apoptosis in B10 T cell cultures (13% vs. 33% of early apoptotic T cells in nonstimulated and antigen-stimulated cultures, respectively; P<0.001, upper panels, quadrants A4) and resulted in an overall, more rapid cell death (33% and 15% of live cells in antigen-stimulated and nonstimulated, 36-h, B10 cultures, respectively; P<0.01, upper panels, quadrants A3). Analogous differences (data not shown) characterized two other strains, B10.SM and B10.MBR, belonging, respectively, to TB-susceptible and TB-resistant groups.
 View larger version (36K): [in a new window] |
Figure 2. Antigen-induced and spontaneous apoptosis in B10 and 4R 36-h T cell cultures. (A) Staining with FITC-labeled Annexin V and PI. (B) Expression of the CD95 marker following second stimulation in vitro with H37Rv sonicate and APC. Numbers represent the percentage of cells in each quadrant. Three independent experiments with mixtures of cells from three mice of each strain in each experiment gave similar results with <10% variation (see text for the statistical analysis, Mann-Whitney U-test). Annexin V– PI–, Live cells; Annexin V+ PI–, early apoptotic cells; Annexin V+ PI+, late apoptotic and necrotic cells.
|
It was important to find out whether similar differences in T cell apoptosis induction between susceptible and resistant mice occur following real infection with virulent mycobacteria, i.e., that TB infection leads to a higher degree of T cell apoptosis in susceptible mice. To this end, mice of TB-susceptible B10 and -resistant 4R strains were infected i.v. with 5 x 105 M. tuberculosis CFU, and after a 2-week interval, T cells from their inguinal lymph nodes were analyzed with respect to the apoptosis induction. As expected, a freshly isolated, ex vivo T cell population (Day 0) was basically free of dead cells, most likely a result of the normal in vivo elimination of the latter, approximately with the speed of appearance for preventing necrotic processes. However, shortly after stimulation of lymph node cells in vitro with mycobacterial sonicate, a significantly (P=0.047) higher proportion of apoptotic cells expressing the CD8 phenotype accumulated in B10 compared with 4R cultures, and this difference increased along the incubation period (Fig. 3B ). Conversely, the proportion of apoptotic CD4+ cells in B10 and 4R cultures remained identical throughout the incubation period (Fig. 3A) . These results suggest that an apoptotic response of CD8+ T cells to prolonged stimulation with mycobacterial antigens during infection really depends on genetic variation in the H2 complex. Moreover, a selective expression of this trait in the CD8 T cell subset explains the difference in the CD4:CD8 ratio between susceptible and resistant H2-congenic mouse strains described above (Table 2) .
 View larger version (15K): [in a new window] |
Figure 3. Progression of apoptotic response in CD4+ (A) and CD8+ (B) lymph node T cells from TB-infected B10 and 4R mice following stimulation with mycobacterial sonicate in vitro. Cells from inguinal lymph nodes of four mice of each strain were cultured individually in the presence of 10 µg/ml antigen, and at indicated time-points, the proportion of apoptotic cells was estimated by flow cytometry after quadruple staining with anti-CD4, anti-CD8, Annexin V, and 7-AAD antibodies. Results of one out of two similar experiments are displayed as mean ± SD. *, P < 0.05.
|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The question of genetic control of the CD4/CD8 ratio has been addressed in a few previous studies. An obvious assumption was that variants of T cell receptor and MHC class I and II genes, i.e., of genetic elements determining T cell selection in the thymus, would affect this ratio, and in certain mouse strains, this was confirmed experimentally [22 , 23 ]. A genome-wide linkage scan for loci determining the CD4/CD8 ratio in the B6-DBA/2 strain combination also suggestively mapped one of the involved QTL within the H2 complex [24 ]. In these studies, H2-dependent variations in the CD4/CD8 ratio were detected in the thymus but not in lymph nodes [24 ] and in the absence of antigenic stimulation [23 , 24 ]. Here, we show that the H2 complex controls a shift in the CD4/CD8 ratio in immune lymph nodes as well. It is important that in our system, this occurs after T cell activation with mycobacterial antigens, probably reflecting secondary, selective events in the periphery.
As the CD4/CD8 ratio could be considered as a potential immunological correlate of anti-TB defense, it should be emphasized that the inverse correlation between that ratio and TB resistance demonstrated in our experiments is by no means universal and may pertain only when genetic variation is limited to the H2 complex. Genetic loci outside the MHC regulate immune reactions to mycobacteria, which differ from those controlled by the MHC. It has been demonstrated convincingly that TB susceptibility [13 ] and CD4/CD8 ratio [23 ] are under control of many epistatically interacting genetic loci located within and outside the H2 complex, so the correlation between the two traits should be accepted with a great deal of caution. Consequently, in genetically diverse human populations, the prognostic potency of the CD4/CD8 ratio estimation in groups of infected individuals will likely be low. Nevertheless, it may be of interest to study possible links among haplotypes of the human leukocyte anitgen system, TB infection, and CD4/CD8 ratio in isolated human populations with a considerable level of natural inbreeding and/or in families with marriages among relatives.
The paradox of simultaneous development of T cell reactivity and immune anergy during the development of TB is generally acknowledged [25 26 27 ]. Almost immediately after the discovery of apoptosis, apoptotic cell death was detected by light and electronic microscopy within specific tuberculous and leprosy granulomas [28 ]. Since then, the possible role of apoptosis in the suppression of immune responses during TB has been discussed widely [29 , 30 ]. It should be emphasized, however, that the apoptotic death of mycobacteria-infected macrophages and immune T cells may have different consequences for protection of the host.
Apoptosis of M. tuberculosis-infected macrophages has a profound influence on survival of the parasite [31 ]. Thus, CD95-CD95L-mediated apoptosis of infected macrophages substantially decreased the viability of virulent H37Rv and attenuated H37Ra bacilli, whereas necrotic death did not [32 ]. Moreover, infected macrophages appeared to be more resistant to CD95L binding-induced apoptosis, probably as a result of active down-regulation of CD95 expression by mycobacteria [32 , 33 ]. It is well known that M. tuberculosis inhibits phagosome-lysosome fusion [34 , 35 ], thus blocking acidation of its niche [36 ]. As the early stages of apoptosis are acccompanied by the acidation of the content of cytoplasmic organelles [37 ], it may serve as a compensatory antimycobacterial effector mechanism. Thus, apoptotic death of infected macrophages is generally considered as an element of host defense [38 39 40 ].
The opposite may be true for the apoptosis of T cells. It has been shown that spontaneous and mycobacterial antigen-induced apoptosis of CD4-positive T cells is increased among TB patients compared with tuberculin-positive, healthy donors [41 ]. In agreement with this finding, infection with M. tuberculosis is accompanied with elevated expression of the CD95 receptor on CD4-positive T cells [42 ], i.e., activation of the pathway involved in CD95-CD95L-mediated AICD of CD4+ cells upon stimulation with high doses of antigens [43 , 44 ]. Thus, elimination of T cells by the AICD mechanism, apparently developed as a mechanism for maintenance of T cell homeostasis, may prevent specific, long-term responsiveness to antigens and exert a deleterious effect on the course of chronic infections including TB [27 , 45 , 46 ]. In the present study, we demonstrate that the capacity to maintain a steady-state, proliferative response to mycobacterial antigens upon repeated stimulation correlates with an increased resistance to infection and is under MHC genetic control.
Received July 18, 2005; revised October 17, 2005; accepted November 21, 2005.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
(-238 and -308) and ß gene polymorphisms in pulmonary tuberculosis: haplotype analysis with HLA-A, B and DR genes Tuberculosis (Edinb.) 81,335-341-irradiation Int. J. Radiat. Biol. 73,179-185[CrossRef][Medline]-chain locus Curr. Biol. 8,701-704[CrossRef][Medline] production at sites of active Mycobacterium tuberculosis infection in human tuberculosis J. Infect. Dis. 183,779-788[CrossRef][Medline]/ß T lymphocytes in a concentration-dependent manner Immunology 105,222-230[CrossRef][Medline]/
T lymphocytes from patients with advanced clinical forms of active tuberculosis Clin. Diagn. Lab. Immunol. 4,14-18
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
