Portal de Periódicos da CAPES

In vitro T‐cell activation of monocyte‐derived macrophages by soluble messengers or cell‐to‐cell contact in bovine tuberculosis

2000; Wiley; Volume: 100; Issue: 2 Linguagem: Inglês

10.1046/j.1365-2567.2000.00027.x

1365-2567

E. Liébana, Alicia Aranaz, Michael Welsh, S.D. Neill, J.M. Pollock,

Immune Cell Function and Interaction

Animal tuberculosis caused by Mycobacterium bovis is a disease with significant economic and public health implications. Organisms in the M. tuberculosis complex, such as M. bovis, are facultative intracellular pathogens which survive and multiply within the macrophages of the host. 1 It is believed that the interactions of these infected cells with T lymphocytes result either in the modulation of the different functions normally assigned to macrophages (mycobactericidal activity, antigen presentation capabilities), or in the destruction of the chronically infected macrophages by cytotoxic T lymphocytes (CTL). Furthermore, it is considered that interferon-γ (IFN-γ) plays an essential role in stimulation of macrophages, and is capable of up-regulating the pathways for several macrophage effector functions, although other factors may also be needed in combination.2,3 The macrophage plays a dual role in tuberculosis; promoting not only protection against the disease (effector role) but also survival of the pathogen. Mycobacterium tuberculosis complex organisms are efficient in entering the macrophage by specific receptor–ligand interactions, 4 and in surviving the hostile environment of phagolysosomes. However, there are some macrophage defences [such as release of tumour necrosis factor-α (TNF-α) and production of reactive nitrogen intermediates] which may restrict proliferation of these pathogens. The balance between mycobacterial virulence and these macrophage defences is fundamental to an understanding of the pathogenesis of tuberculosis. Correlations of in vitro growth rates of mycobacteria with in vivo virulence have been made previously for mice, 5 rabbits, 6 guinea-pigs, 7 humans 8–10 and cattle. 11 However, observations in one system may not necessarily translate to other species. 8 There is significant controversy on the feasibility of in vitro activation of mycobactericidal activity of macrophages by treatment with different cytokines. The ability of murine macrophages to inhibit the growth of mycobacteria following in vitro activation by soluble messengers has been demostrated. 12–16 In contrast, other studies have failed to show an increased ability to restrict mycobacterial growth by treatment with soluble factors in other species including humans. 12,17–19 The situation for cattle is still unclear and contradictory results have been reached.20,21 Some authors have suggested that the primary role of IFN-γ and other cytokines may be to regulate antigen processing and presentation of mycobacterial antigens to T cells rather than to stimulate macrophage activation for mycobacterial killing. 22 In the present study we have investigated the in vitro growth of M. bovis bacillus Calmette–Guérin (BCG) and virulent M. bovis in non-activated bovine monocyte-derived macrophages; also we assayed the use of T-cell supernatants (TCS) for in vitro activation of macrophages leading to changes in morphology, to modification of the antimycobacterial capabilities and to modulation of their potential as antigen-presenting cells (APC). We also evaluated the interaction of different T-cell populations with M. bovis-infected macrophages in terms of capacity to restrict intracellular mycobacterial growth. Field isolates of M. bovis (T/91/1378, Veterinary Sciences Division, Belfast, Northern Ireland) and M. bovis BCG (kindly provided by Dr P. Andersen, Statens Seruminstitut, Copenhagen, Denmark) were prepared for the experiments as previously described. 23 For production of M. bovis sonic extract (MBSE), mid-log phase cultures were treated as previously described. 23 Protein concentration was estimated by bicinchonicic acid (BCA) protein assay (Pierce, Rockford, IL). Friesian-cross, male calves, approximately 6 months old, were obtained from herds with no history of tuberculosis for at least 5 years. During the study all the animals were fed normal diets. The experiments were done in two different sets of animals, consisting of animals 01, 02 and 03, and animals 04 and 05, respectively. Animals 01–03 were experimentally infected as follows: the animals were housed in a high-security isolation house under negative pressure with expelled air filtered through absolute filters and were infected by intranasal instillation of 106 colony-forming units (CFU) of M. bovis. Non-infected animals 04 and 05 were housed in normal farm boxes and were used as controls for the experiments. Blood samples utilized for the experiments were obtained at times varying between 18 and 36 weeks postinfection (p.i.). Monocytes were prepared as CD14+ cells positively selected from freshly isolated peripheral blood mononuclear cells (PBMC) using the magnetic-activated cell sorting system (MACS, Miltenyi Biotec, Bergisch Gladbach, Germany). Pelleted aliquots of 2 × 107 PBMC which had been separated from heparinized venous blood over Ficoll–Histopaque gradients (Amersham Pharmacia Biotech AB, Little Chalfont, UK) were resuspended in 100 µl of MACS flow [2 parts phosphate-buffered saline (PBS) : 1 part fluorescence-activated cell sorter flow (FACS flow; Becton Dickinson, Oxford, UK) : 1% bovine serum albumin (Sigma Chemical Co., Poole, UK)]. These were then incubated with anti-human CD14 MicroBeads (Miltenyi Biotec) for 30 min at 4°. Thereafter, cells bound by MACS microbeads were separated using LS+ Separation Columns attached to Midi Magnets according to the manufacturer's instructions (Miltenyi Biotec). The recovered cell population was resuspended in macrophage medium [RPMI-1640 (BioWhittaker UK Ltd, Wokingham, UK) containing 10 m m HEPES buffer (Gibco, Paisley, UK), 4 m m l-glutamine (Gibco), 1 m m sodium pyruvate (Gibco), 1 m m non-essential amino acids (Sigma), 50 U/ml penicillin G (Sigma) and 10% (v/v) fetal calf serum (Difco, East Molesley, Surrey, UK)] and cell viability was checked by trypan blue exclusion. Assessment of cell purity was carried out by flow cytometry and was found to be greater than 98%. Purified monocytes were plated at 105 cells/well in flat-bottom, 96-well polystyrene microtitre plates (Nunclon, Nunc, Denmark) and incubated at 37° in a 6% CO2 atmosphere for a further 5–7 days with medium changes every 3 days before they were used for the experiments. CD4+ and CD8+ T cells were positively selected by MACS from freshly isolated PBMC following previously described conditions. 23 The final cell population was resuspended in complete macrophage medium and cell viability was checked by trypan blue exclusion. Assessment of cell purity was carried out by flow cytometry after staining with the appropriate monoclonal antibodies and was always found to be greater than 98%. PBMC from animal 03 were resuspended in LTA culture medium [RPMI-1640 containing 10 m m HEPES buffer, 2 m m l-glutamine, 50 U/ml penicillin G, and 10% (v/v) fetal calf serum]. Then, 20-ml aliquots of PBMC at a concentration of 106 cells/ml were incubated in flasks for 2 days with concanavalin A (Con A; 4 µg/ml) or with live M. bovis (106 CFU/ml) at 37° in the presence of 6% CO2. Cultures were then pelleted by centrifugation, and supernatants were filtered through 0·22-µm pores and subsequently frozen at −20° in 0·5 ml aliquots. As a control of the cytokine production by the PBMC, the presence of IFN-γ was determined in the filtered supernatants by using a commercial enzyme-linked immunosorbent assay (ELISA; CSL, Parkville, Australia) and the optical density at 450 nm (OD450) values were always higher than 2·0. At least quadruplicate wells containing approximately 105 macrophages per well were infected with 105 CFU of M. bovis or M. bovis BCG/well to give a multiplicity of infection 1 : 1, and incubated for 4 hr at 37° in 5% CO2. Macrophages were then rinsed with warm PBS to remove possible extracellular bacteria (control wells were stained by Auramine-O acid-fast technique 24 to check that all bacilli were intracellular) and reincubated with fresh media for 24, 48, or 96 hr. After the incubation time, macrophages were lysed by addition of 0·1% saponin (Sigma). Thereafter, samples were pulsed with 2 µCi/well of 3H-labelled uracil (Amersham International, UK) and incubated for 40 hr. Radioactivity incorporated into mycobacterial RNA was determined as counts per minute (c.p.m.) by liquid scintillation counting using a Wallac 1205 Betaplate counter (Wallac, Milton Keynes, UK). Metabolic labelling with 3H-labelled uracil makes it possible to measure intracellular mycobacterial growth, but it should not be considered as equivalent to CFU. The technique does not allow the differentiation of death and metabolic stasis of micro-organisms. At least quadruplicate wells, containing approximately 105 macrophages per well, were treated with TCS (25% v/v final concentration) produced with Con A or live M. bovis for 24 hr prior to the experiments. Thereafter, cells were infected with 3 × 105 CFU of M. bovis per well to give a multiplicity of infection 3 : 1, and incubated for 4 hr at 37° in 5% CO2. Macrophages were then rinsed with warm PBS to remove any extracellular bacteria and reincubated with fresh medium containing 25% TCS for 48 hr at 37° in 5% CO2. Simultaneously, quadruplicate wells of control macrophages were left untreated, and infected and incubated in fresh medium in the same conditions. Also, quadruplicate wells of untreated macrophages were infected, and incubated under the same conditions with autologous T cells (PBMC or sorted cells) in macrophage medium at a T-cell : macrophage ratio of 20 : 1, 10 : 1, or 5 : 1 for PBMC, or 5 : 1 for sorted CD4+ or CD8+ T cells. After the incubation time, macrophages were lysed and viable mycobacteria were pulsed with uracil as above. Microtitre macrophage cultures were treated for 24 hr with Con-A-generated TCS (25%). Subsequently, medium was removed and adherent cells were methanol-fixed to the bottom of the wells, stained with methylene blue and eosine (Speedy-Diff kit, Clin-Tech, Clacton-on-Sea, Essex, UK) and examined microscopically (Nikon TMS, Tokyo, Japan). Untreated control wells, and Con-A-treated (1 µg/ml) wells were also stained for comparison. To determine the effects of TCS on the capacity of macrophages to present antigen, proliferation assays were set up for CD4+ sorted T cells. Microtitre cultures were established with 2·5 × 105 CD4+ T cells, and 105 autologous macrophages in 200 µl of medium. For the comparative experiments, macrophages were pretreated for 24 hr with 25% Con A–TCS, and control macrophages were left untreated. The macrophage cultures were gently washed before the addition of the lymphocytes to prevent any direct effect of the TCS on the proliferation results. For both types of APC (controls and TCS-treated cells) triplicate wells were set up for the antigen (MBSE) as well as negative controls (PBS). Cultures were incubated for 5 days in the same conditions described previously. 23 Aliquots of 5 × 106 CD14+ sorted cells in 10 ml of medium were incubated in 50 ml Teflon Erlenmeyer flasks (Nalge Company, Rochester, NY) at 37° in the presence of 5% CO2 for 6 days; the flasks were then incubated for a further 24 hr with or without Con A–TCS (25%). Control cultures were also treated with Con A (1 µg/ml) to rule out any direct effect of residual mitogen present in the TCS. After the period of culture, the flasks were incubated on ice for 15 min and cells were harvested and incubated with optimal concentrations (in RPMI : 10% normal rabit serum) of anti-ovine CD14+ monoclonal antibody [mouse immunoglobulin G1 (IgG1); Serotec, Oxford, UK] and anti-bovine MHC class II monoclonal antibody (ILA-21, mouse IgG2a) (ECACC, Salisbury, UK) for 30 min at 4°. After washing twice with PBS : 0·1% NaN3, positive cells were identified using a mix of fluorescein isothiocyanate (FITC) -conjugated goat anti-mouse IgG2a, and phycoerythrin (PE) -conjugated goat anti-mouse IgG1 (Southern Biotechnologies Associates Inc., Birmingham, AL). After a further 30 min incubation at 4°, the cells were washed and fixed in 1% paraformaldehyde in PBS before analysis. Flow cytometry analysis was performed using a FACS Vantage (Becton Dickinson) equipped with an Innova Enterprise ion Laser (Coherent Laser Group, San Jose, CA). Using the cell quest programme, the monocyte population was identified on the basis of forward and orthogonal light scatter, and gated appropriately. Labelling of the cells with anti-CD14+ antibody allowed confirmation of a pure macrophage population after the in vitro culture. Log integral fluorescence was measured for the gated population and 5000 cells were counted for each sample. Data were analysed using Kolmogorov–Smirnov statistics (D-value) to determine differences in expression of MHC class II molecules between control and TCS-treated samples. 25 Figure 1 shows the results of experiments with cells from animals 02 and 03. Both strains of mycobacteria survived and grew well in cell-free macrophage medium: the mean c.p.m. in the uracil assay for the initial (24 hr) time-point were 44 334 ± 7291 and 38 759 ± 9033 for M. bovis and M. bovis BCG, respectively; the mean c.p.m. in the uracil assay for the final (96 hr) time-point were 210 656 ± 17 052 and 232 785 ± 13 400 for M. bovis and M. bovis BCG, respectively. When cultured in the presence of bovine monocyte-derived macrophages both strains showed low levels of uracil uptake compared to the cultures in medium alone. The uptake of uracil by bacteria in the presence of macrophages was, on average, for both strains in both animals 20% of the values for 24 hr bacterial growth in medium alone, suggesting that macrophages were partially successful in limiting the survival and early growth of these organisms. In both animals, by 24 hr p.i. uptake of uracil by both strains of M. bovis was very similar. However, at time-point 72 hr, uptake by virulent M. bovis was distinctly greater than by M. bovis BCG, indicating that macrophages were only able to limit the initial survival and growth of virulent M bovis. Over an extended period of 96 hr, M. bovis (but not M. bovis BCG) growth was markedly increased within the macrophage cultures. Therefore, it was apparent that bovine monocytes matured in vitro were able to cause metabolic stasis of M. bovis BCG for up to 96 hr p.i., whereas they could not control metabolism by virulent M. bovis. Microscopic examination of M. bovis-infected cultures at 72 hr p.i. showed that cell membranes were intact with no evidence of extracellular bacteria; however, disruption and lysis of some macrophages was microscopically evident in cultures at 96 hr p.i., resulting in the presence of extracellular bacteria able to grow and metabolize free from the restraining intracellular environment. Therefore the dramatic increase in M. bovis uracil uptake at time 96 hr is likely to reflect both intra- and extracellular growth. Intracellular growth of M. bovis and M. bovis BCG in bovine monocyte-derived macrophages. Cells (105) were infected with mycobacteria at a ratio of 1 : 1 and were cultured for 24, 72 and 96 hr. At the end of each time-point, macrophages were saponin-lysed and mycobacterial growth was determined by incorporation of 3H-labelled uracil. Data are mean values ± SE of experiments with cells from two infected animals. For comparison, the mean c.p.m. values for bacteria in medium alone were: for the 24-hr time-point 44 334 ± 7291 and 38 759 ± 9033 for M. bovis and M. bovis BCG, respectively; for the 96-hr time-point 210 656 ± 17 052 and 232 785 ± 13 400 for M. bovis and M. bovis BCG, respectively. Figure 2 shows a typical example of change in cellular morphology of monocyte cultures after a 24-hr incubation period in the presence of 25% Con-A-derived TCS. Unlike the controls and Con-A-treated cells, TCS-treated cells were of increased cellular size and exhibited increased adherence to plastic surfaces. These monocytes displayed pseudopodic membranes and were found in tight groups of several cells, sometimes with membrane fusion forming multinucleated giant cells. These changes in appearance were suggestive of in vitro activation of the monocytes. Typical examples of control (a) and Con A–TCS-treated (b) bovine moncytes; 105 cells per well were cultured for 6 days in 96-well polystyrene plates before stimulation with TCS (25%) for a further 24 hr. Treated cells showed marked attachment with spreading of cytoplasmic margins and formation of multinucleated cells; untreated cells kept the usual rounded appearance. Speedy–Diff staining; magnification × 200. We wanted to determine if treatment with immunomodulating factors would affect the growth pattern of M. bovis within bovine monocytes. Figure 3 shows the results of a series of experiments with cells from three infected and two non-infected animals. The amounts of 3H-labelled uracil incorporated by bacteria growing within the treated (TCS or T cell) cells were calculated as a percentage of c.p.m. incorporated by bacilli growing in untreated monocytes. In general terms, the use of soluble TCS for the in vitro activation did not enhance the antimycobacterial activity of bovine macrophages compared to the control untreated cells (uracil uptake = 100%). Furthermore, on several occasions the treated cells were more permissive to M. bovis growth, as indicated by a uracil uptake in the cultures greater than 100%. No differences were found between TCS generated with Con A or with live M. bovis infection of the PBMC. Modifying a number of parameters of TCS production (increasing cell number, incubation time prior to harvesting) or per cent v/v (50, 75%) of supernatants was ineffective in terms of enhancing the antimycobacterial effect within the macrophage cultures (data not shown). Intracellular growth of M. bovis in bovine monocyte-derived macrophages from three infected (01,02 and 03) and two non-infected (04 and 05) animals: comparison of the effects of TCS and various T-cell populations on bacterial survival. Macrophages were pretreated with M. bovis–TCS or Con A–TCS for 24 hr and then infected at a multiplicity of infection of 3 : 1. Alternatively, macrophages were infected and subsequently autologous T cells (PBMC, CD4+, CD8+) were added to the wells at the concentrations indicated in the figure. Cultures were incubated for 2 days, then saponin-lysed and incubated for 40 hr in the presence of 3H-labelled uracil. Data are expressed as percentages of incorporated c.p.m. by bacteria growing within the treated (TCS or T-cell) macrophages, calculated in relation to the c.p.m. incorporated by bacilli growing in untreated moncytes (100%). The mean 3H-labelled uracil control values (c.p.m.) were as follows: 50 968 (animal 01), 46 956 (02), 45 550 (03), 88 459 (04) and 777 212 (05). The result are presented as means for repeated experiments ± SE. In contrast to these results, in the present in vitro model, incubation of infected monocytes with T-cell populations resulted in increased inhibition of M. bovis uracil uptake, suggesting the need for cellular contact for activation of antimycobacterial capabilities in bovine monocytes. Several T-cell populations were compared in their ability to promote M. bovis stasis. When a mixed PBMC population was used, a noticeable M. bovis stasis was observed, with values of uracil uptake between 28 and 73% of the values for the control infected cells. In most cases a relation between cellular ratio/inhibition was found: an increase of the number of PBMC per macrophage resulted in a lower uracil uptake. No differences were found between the groups of infected and non-infected animals, indicating that much of this response could be due to innate resistance to the disease. In an attempt to compare the potential of CD4+ and CD8+ cells to promote such activity, MACS-sorted populations were used in the experiments. The M. bovis growth inhibition achieved was considerably lower than with the mixed population, with percentages of 68–82% for the CD4+ cells and 56–86% for the CD8+ cells. While TCS did not affect the ability of macrophages to restrict M. bovis growth, we decided to test whether TCS treatment could enhance antigen presentation of these cells to CD4+ T lymphocytes. Figure 4 illustrates the difference in antigen (MBSE) responsiveness of sorted CD4+ cells in the presence of TCS treated/untreated macrophages. The responses were measured as proliferation (methyl-[3H]thymidine incorporation) of CD4+ cells. A dramatic increase in proliferation was found after treatment of the APC with Con A–TCS; the effect was only found in the presence of antigen and any effect of soluble factors or previous Con A treatment was ruled out by the inclusion of the PBS proliferation controls. Proliferative responses of CD4+ sorted cells to soluble mycobacterial antigen (MBSE). For the comparative experiments, macrophages were pretreated for 24 hr with 25% Con A–TCS, and control macrophages were left untreated. The macrophage cultures were gently washed before the addition of the lymphocytes to prevent any direct effect of the TCS on the proliferation results. Results are the mean ± SE representative of one experiment for animal 02 and four experiments for animal 03. To determine whether an increase in the MHC class II expression was a possible cause of a more proficient antigen-presenting activity, cultures of monocytes from animals 03 and 04 were established in suspension (Teflon-coated flasks) and treated with TCS. Subsequently, MHC class II expression was analysed by FACS. The experiments were repeated twice in both animals with similar results. Data were analysed by Kolmogorov–Smirnov statistics (D-values > 30%). Figure 5 illustrates a representative experiment showing that there is an up-regulation of class II molecules in the treated cultures compared to untreated cells. Flow cytometric analysis of MHC class II expression. Six-day-old CD14+ monocytes from animal 03 cultured in suspension (Teflon-coated flasks), were treated for 24 hr with Con A–TCS or left untreated as controls. Subsequently, MHC class II expression was analysed by FACS. Viable cells were selected on forward scatter (FCS) versus side scatter (SCS) dot plots (a). Log integral fluorescence was measured for the gated position and represented in histograms for control and TCS-treated cultures (b). Results were analysed using cell quest and Kolmogorov–Smirnov statistics (D = 60%) (c). The uracil uptake technique has proved useful previously for assessment of intracellular mycobacterial growth. 26–28 The measurement of uracil incorporation provides a reliable indication of the metabolic status of growing organisms.12,29 The present study is the first to compare the growth of a virulent M. bovis strain and M. bovis BCG in bovine monocyte-derived macrophages. Acid-fast staining of cultures showed that the bacilli were intracellular following infection of monolayers and therefore the uracil counts reflected intracellular growth of bacteria. The dramatic increase in M. bovis uracil uptake at 96 hr is likely to reflect both intra- and extracellular growth due to destruction of macrophages. Our study shows that blood-derived bovine macrophages are capable of controlling M. bovis BCG growth but they allow intracellular growth of virulent M. bovis. We only detected an ability to limit the initial growth of M bovis (first 24 hr) with little effect on surviving intracellular bacilli which were able to replicate over an extended period of study (96 hr). A similar comparison in monocyte-derived ferret macrophages reached the same conclusion. 19 Other studies in cattle have also found that bovine alveolar macrophages are capable of causing metabolic stasis of BCG for up to 96 hr p.i., whereas M. bovis is able to metabolize actively in the cultures.11,26 In contrast with these results, others found that M. bovis BCG was able to grow inside monocyte-derived macrophages. 20 The discrepancy between studies could be due to the higher multiplicity of infection (10 : 1 as opposed to 1 : 1 in the present study) used for those experiments. Several theories have been proposed to explain why virulent M. bovis is able to overwhelm the bacteriostatic activity of macrophages. 26 Recent evidence has strongly suggested that combinations of lymphokines may synergize and that certain soluble factors (such as interleukin-4, TNF-α, IFN-γ) may influence monocyte/macrophage differentiation and fusion of these cells into giant multinucleated cells. 30–32 Our study shows that treatment of cultured monocytes with mitogen-produced TCS promoted morphological changes and giant cell formation which probably reflect an activation status of the cells. A recent study with human monocytes treated with TCS showed fusion rates between 10 and 20%. 33 The role of giant cell formation in an in vivo granuloma might be to retain mycobacteria inside cells which are able to kill the pathogens and to prevent dissemination from the local site of infection. 33 In order to find out the functional significance of the morphological changes originated by TCS treatment we performed experiments to determine antimycobacterial activity, and capacity to present antigen to CD4+ T cells. We found that the use of soluble TCS for the in vitro activation did not enhance the antimycobacterial activity of bovine macrophages compared to the control untreated cells. Furthermore, on several occasions, the treated cells were more permissive to intracellular mycobacterial growth. Several studies have suggested the need for several cytokines acting together in the macrophage activation. In order to simulate a combination of cytokines and other lymphocyte-derived factors mimicking an in vivo situation we decided to test the use of a complex TCS generated with Con A or by infection of PBMC with live M. bovis. It could be argued that stimulation with mitogen did not induce maximal secretion of the putative activating soluble factor, that is why this was also tested by generating supernatants with live M. bovis from cells derived from infected cattle. Neither of these reagents were able to activate antimycobacterial activity. Previous experiments in the human system also revealed that pretreatment or concomitant treatment of M. tuberculosis-infected macrophages with different cytokines did not have an effect on the growth rate of that mycobacterium.12,18,34 A similar situation was encountered for M. bovis in ferret macrophages, where TCS-activated ferret macrophages had heightened oxidative burst performance but failed to control the intracellular growth; 19 and for M. bovis BCG in cattle monocytes. 20 However, other studies with bovine alveolar macrophages found that pretreatment with IFN-γ or sequentially with IFN-γ and lipopolysaccharide resulted in a significant reduction of M. bovis and M. bovis BCG growth.11,21 A similar picture seems to be accepted for the murine model.16,27 These contrasting results may reflect differences in the host species, state of macrophage differentiation, site of origin, strain of mycobacteria and conditions of culture. Macrophages derived from different sources may also be subject to different regulatory controls. 21 The failure in macrophage activation found in our study was not due to inactivity of the TCS because experiments run with similar reagents showed morphological changes as well as increase in the antigen-presenting capacity of the cells. To rule out the possibility that the active lymphokines were at suboptimal concentration, we performed experiments with increasing percentages of TCS but did not get an improvement on the expression of antibacterial activity; in fact a more permissive behaviour of the macrophage to mycobacterial growth was often observed. Lymphokines rendering macrophages more permissive for growth of mycobacteria have been described before.18,19,35 Recent evidence supports the concept that cell contact may be as important for activation of macrophages as the presence of soluble cytokines.22,36 In the present experiments, cell contact of macrophages with autologous T cells (PBMC or sorted cells), probably in combination with the release of factors such as IFN-γ, enabled macrophages to restrict replication of intracellular bacilli more efficiently. In the case of the PBMC, the degree of M. bovis growth inhibition related to lymphocyte concentration; a similar observation has been reported for M. bovis BCG. 20 In the present study and for the same T-cell : macrophage ratio (5 : 1) CD4+ or CD8+ sorted cells were less able to inhibit M. bovis growth compared to PBMC. This could indicate that both subpopulations are needed for effective antimycobacterial activity. Our data from both infected and non-infected animals indicated that PBMC interact in vitro in some way with the M. bovis-infected macrophages, inducing significant stasis of the mycobacterial load in comparison with the control infected macrophages. At present we do not have an explanation for this observation, a possible theory being that there is a degree of innate resistance to the mycobacterial infection in these cellular interactions. In accordance with our results, autologous PBMC from both immunized and non-immunized donors have been found to restrict the growth of BCG inside monocyte-derived macrophages. 20 Macrophages are primary effector cells but also are essential for processing and presentation of antigens to T cells. We have shown that in vitro treatment of blood-derived macrophages with Con A–TCS enhances the antigen presentation capabilities of these cells. In our experiments, presentation of a soluble mycobacterial antigen (MBSE) to CD4+ T cells was substantially improved as indicated by an increased proliferation of the sorted cells. Furthermore, we demonstrated the ability of TCS to up-regulate MHC class II expression on the bovine macrophages. Other authors in the past have found similar results in IFN-γ-treated human monocytes. 37 In summary, although treatment of bovine monocytes with TCS caused morphological changes, increased MHC class II expression, and increased proliferation of CD4+ sorted cells against mycobacterial antigens, we were unable to detect increased in vitro antimycobacterial activity. The observations described may be of future help in the study of protective immunity and in the evaluation of candidate vaccines for bovine tuberculosis. This work was completely performed in the Department of Veterinary Science (Queen's University of Belfast) and was supported by the European Contracts ERBFMBICT972853, and FAIRBM971136; and funds from the Department of Agriculture for Northern Ireland.